1.. SPDX-License-Identifier: GPL-2.0 2 3==================== 4The /proc Filesystem 5==================== 6 7===================== ======================================= ================ 8/proc/sys Terrehon Bowden <terrehon@pacbell.net>, October 7 1999 9 Bodo Bauer <bb@ricochet.net> 102.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000 11move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009 12fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009 13===================== ======================================= ================ 14 15 16 17.. Table of Contents 18 19 0 Preface 20 0.1 Introduction/Credits 21 0.2 Legal Stuff 22 23 1 Collecting System Information 24 1.1 Process-Specific Subdirectories 25 1.2 Kernel data 26 1.3 IDE devices in /proc/ide 27 1.4 Networking info in /proc/net 28 1.5 SCSI info 29 1.6 Parallel port info in /proc/parport 30 1.7 TTY info in /proc/tty 31 1.8 Miscellaneous kernel statistics in /proc/stat 32 1.9 Ext4 file system parameters 33 34 2 Modifying System Parameters 35 36 3 Per-Process Parameters 37 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer 38 score 39 3.2 /proc/<pid>/oom_score - Display current oom-killer score 40 3.3 /proc/<pid>/io - Display the IO accounting fields 41 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings 42 3.5 /proc/<pid>/mountinfo - Information about mounts 43 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 44 3.7 /proc/<pid>/task/<tid>/children - Information about task children 45 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 46 3.9 /proc/<pid>/map_files - Information about memory mapped files 47 3.10 /proc/<pid>/timerslack_ns - Task timerslack value 48 3.11 /proc/<pid>/patch_state - Livepatch patch operation state 49 3.12 /proc/<pid>/arch_status - Task architecture specific information 50 51 4 Configuring procfs 52 4.1 Mount options 53 54 5 Filesystem behavior 55 56Preface 57======= 58 590.1 Introduction/Credits 60------------------------ 61 62This documentation is part of a soon (or so we hope) to be released book on 63the SuSE Linux distribution. As there is no complete documentation for the 64/proc file system and we've used many freely available sources to write these 65chapters, it seems only fair to give the work back to the Linux community. 66This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm 67afraid it's still far from complete, but we hope it will be useful. As far as 68we know, it is the first 'all-in-one' document about the /proc file system. It 69is focused on the Intel x86 hardware, so if you are looking for PPC, ARM, 70SPARC, AXP, etc., features, you probably won't find what you are looking for. 71It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But 72additions and patches are welcome and will be added to this document if you 73mail them to Bodo. 74 75We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of 76other people for help compiling this documentation. We'd also like to extend a 77special thank you to Andi Kleen for documentation, which we relied on heavily 78to create this document, as well as the additional information he provided. 79Thanks to everybody else who contributed source or docs to the Linux kernel 80and helped create a great piece of software... :) 81 82If you have any comments, corrections or additions, please don't hesitate to 83contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this 84document. 85 86The latest version of this document is available online at 87http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html 88 89If the above direction does not works for you, you could try the kernel 90mailing list at linux-kernel@vger.kernel.org and/or try to reach me at 91comandante@zaralinux.com. 92 930.2 Legal Stuff 94--------------- 95 96We don't guarantee the correctness of this document, and if you come to us 97complaining about how you screwed up your system because of incorrect 98documentation, we won't feel responsible... 99 100Chapter 1: Collecting System Information 101======================================== 102 103In This Chapter 104--------------- 105* Investigating the properties of the pseudo file system /proc and its 106 ability to provide information on the running Linux system 107* Examining /proc's structure 108* Uncovering various information about the kernel and the processes running 109 on the system 110 111------------------------------------------------------------------------------ 112 113The proc file system acts as an interface to internal data structures in the 114kernel. It can be used to obtain information about the system and to change 115certain kernel parameters at runtime (sysctl). 116 117First, we'll take a look at the read-only parts of /proc. In Chapter 2, we 118show you how you can use /proc/sys to change settings. 119 1201.1 Process-Specific Subdirectories 121----------------------------------- 122 123The directory /proc contains (among other things) one subdirectory for each 124process running on the system, which is named after the process ID (PID). 125 126The link 'self' points to the process reading the file system. Each process 127subdirectory has the entries listed in Table 1-1. 128 129Note that an open file descriptor to /proc/<pid> or to any of its 130contained files or subdirectories does not prevent <pid> being reused 131for some other process in the event that <pid> exits. Operations on 132open /proc/<pid> file descriptors corresponding to dead processes 133never act on any new process that the kernel may, through chance, have 134also assigned the process ID <pid>. Instead, operations on these FDs 135usually fail with ESRCH. 136 137.. table:: Table 1-1: Process specific entries in /proc 138 139 ============= =============================================================== 140 File Content 141 ============= =============================================================== 142 clear_refs Clears page referenced bits shown in smaps output 143 cmdline Command line arguments 144 cpu Current and last cpu in which it was executed (2.4)(smp) 145 cwd Link to the current working directory 146 environ Values of environment variables 147 exe Link to the executable of this process 148 fd Directory, which contains all file descriptors 149 maps Memory maps to executables and library files (2.4) 150 mem Memory held by this process 151 root Link to the root directory of this process 152 stat Process status 153 statm Process memory status information 154 status Process status in human readable form 155 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function 156 symbol the task is blocked in - or "0" if not blocked. 157 pagemap Page table 158 stack Report full stack trace, enable via CONFIG_STACKTRACE 159 smaps An extension based on maps, showing the memory consumption of 160 each mapping and flags associated with it 161 smaps_rollup Accumulated smaps stats for all mappings of the process. This 162 can be derived from smaps, but is faster and more convenient 163 numa_maps An extension based on maps, showing the memory locality and 164 binding policy as well as mem usage (in pages) of each mapping. 165 ============= =============================================================== 166 167For example, to get the status information of a process, all you have to do is 168read the file /proc/PID/status:: 169 170 >cat /proc/self/status 171 Name: cat 172 State: R (running) 173 Tgid: 5452 174 Pid: 5452 175 PPid: 743 176 TracerPid: 0 (2.4) 177 Uid: 501 501 501 501 178 Gid: 100 100 100 100 179 FDSize: 256 180 Groups: 100 14 16 181 VmPeak: 5004 kB 182 VmSize: 5004 kB 183 VmLck: 0 kB 184 VmHWM: 476 kB 185 VmRSS: 476 kB 186 RssAnon: 352 kB 187 RssFile: 120 kB 188 RssShmem: 4 kB 189 VmData: 156 kB 190 VmStk: 88 kB 191 VmExe: 68 kB 192 VmLib: 1412 kB 193 VmPTE: 20 kb 194 VmSwap: 0 kB 195 HugetlbPages: 0 kB 196 CoreDumping: 0 197 THP_enabled: 1 198 Threads: 1 199 SigQ: 0/28578 200 SigPnd: 0000000000000000 201 ShdPnd: 0000000000000000 202 SigBlk: 0000000000000000 203 SigIgn: 0000000000000000 204 SigCgt: 0000000000000000 205 CapInh: 00000000fffffeff 206 CapPrm: 0000000000000000 207 CapEff: 0000000000000000 208 CapBnd: ffffffffffffffff 209 CapAmb: 0000000000000000 210 NoNewPrivs: 0 211 Seccomp: 0 212 Speculation_Store_Bypass: thread vulnerable 213 SpeculationIndirectBranch: conditional enabled 214 voluntary_ctxt_switches: 0 215 nonvoluntary_ctxt_switches: 1 216 217This shows you nearly the same information you would get if you viewed it with 218the ps command. In fact, ps uses the proc file system to obtain its 219information. But you get a more detailed view of the process by reading the 220file /proc/PID/status. It fields are described in table 1-2. 221 222The statm file contains more detailed information about the process 223memory usage. Its seven fields are explained in Table 1-3. The stat file 224contains detailed information about the process itself. Its fields are 225explained in Table 1-4. 226 227(for SMP CONFIG users) 228 229For making accounting scalable, RSS related information are handled in an 230asynchronous manner and the value may not be very precise. To see a precise 231snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table. 232It's slow but very precise. 233 234.. table:: Table 1-2: Contents of the status files (as of 4.19) 235 236 ========================== =================================================== 237 Field Content 238 ========================== =================================================== 239 Name filename of the executable 240 Umask file mode creation mask 241 State state (R is running, S is sleeping, D is sleeping 242 in an uninterruptible wait, Z is zombie, 243 T is traced or stopped) 244 Tgid thread group ID 245 Ngid NUMA group ID (0 if none) 246 Pid process id 247 PPid process id of the parent process 248 TracerPid PID of process tracing this process (0 if not) 249 Uid Real, effective, saved set, and file system UIDs 250 Gid Real, effective, saved set, and file system GIDs 251 FDSize number of file descriptor slots currently allocated 252 Groups supplementary group list 253 NStgid descendant namespace thread group ID hierarchy 254 NSpid descendant namespace process ID hierarchy 255 NSpgid descendant namespace process group ID hierarchy 256 NSsid descendant namespace session ID hierarchy 257 VmPeak peak virtual memory size 258 VmSize total program size 259 VmLck locked memory size 260 VmPin pinned memory size 261 VmHWM peak resident set size ("high water mark") 262 VmRSS size of memory portions. It contains the three 263 following parts 264 (VmRSS = RssAnon + RssFile + RssShmem) 265 RssAnon size of resident anonymous memory 266 RssFile size of resident file mappings 267 RssShmem size of resident shmem memory (includes SysV shm, 268 mapping of tmpfs and shared anonymous mappings) 269 VmData size of private data segments 270 VmStk size of stack segments 271 VmExe size of text segment 272 VmLib size of shared library code 273 VmPTE size of page table entries 274 VmSwap amount of swap used by anonymous private data 275 (shmem swap usage is not included) 276 HugetlbPages size of hugetlb memory portions 277 CoreDumping process's memory is currently being dumped 278 (killing the process may lead to a corrupted core) 279 THP_enabled process is allowed to use THP (returns 0 when 280 PR_SET_THP_DISABLE is set on the process 281 Threads number of threads 282 SigQ number of signals queued/max. number for queue 283 SigPnd bitmap of pending signals for the thread 284 ShdPnd bitmap of shared pending signals for the process 285 SigBlk bitmap of blocked signals 286 SigIgn bitmap of ignored signals 287 SigCgt bitmap of caught signals 288 CapInh bitmap of inheritable capabilities 289 CapPrm bitmap of permitted capabilities 290 CapEff bitmap of effective capabilities 291 CapBnd bitmap of capabilities bounding set 292 CapAmb bitmap of ambient capabilities 293 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...) 294 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...) 295 Speculation_Store_Bypass speculative store bypass mitigation status 296 SpeculationIndirectBranch indirect branch speculation mode 297 Cpus_allowed mask of CPUs on which this process may run 298 Cpus_allowed_list Same as previous, but in "list format" 299 Mems_allowed mask of memory nodes allowed to this process 300 Mems_allowed_list Same as previous, but in "list format" 301 voluntary_ctxt_switches number of voluntary context switches 302 nonvoluntary_ctxt_switches number of non voluntary context switches 303 ========================== =================================================== 304 305 306.. table:: Table 1-3: Contents of the statm files (as of 2.6.8-rc3) 307 308 ======== =============================== ============================== 309 Field Content 310 ======== =============================== ============================== 311 size total program size (pages) (same as VmSize in status) 312 resident size of memory portions (pages) (same as VmRSS in status) 313 shared number of pages that are shared (i.e. backed by a file, same 314 as RssFile+RssShmem in status) 315 trs number of pages that are 'code' (not including libs; broken, 316 includes data segment) 317 lrs number of pages of library (always 0 on 2.6) 318 drs number of pages of data/stack (including libs; broken, 319 includes library text) 320 dt number of dirty pages (always 0 on 2.6) 321 ======== =============================== ============================== 322 323 324.. table:: Table 1-4: Contents of the stat files (as of 2.6.30-rc7) 325 326 ============= =============================================================== 327 Field Content 328 ============= =============================================================== 329 pid process id 330 tcomm filename of the executable 331 state state (R is running, S is sleeping, D is sleeping in an 332 uninterruptible wait, Z is zombie, T is traced or stopped) 333 ppid process id of the parent process 334 pgrp pgrp of the process 335 sid session id 336 tty_nr tty the process uses 337 tty_pgrp pgrp of the tty 338 flags task flags 339 min_flt number of minor faults 340 cmin_flt number of minor faults with child's 341 maj_flt number of major faults 342 cmaj_flt number of major faults with child's 343 utime user mode jiffies 344 stime kernel mode jiffies 345 cutime user mode jiffies with child's 346 cstime kernel mode jiffies with child's 347 priority priority level 348 nice nice level 349 num_threads number of threads 350 it_real_value (obsolete, always 0) 351 start_time time the process started after system boot 352 vsize virtual memory size 353 rss resident set memory size 354 rsslim current limit in bytes on the rss 355 start_code address above which program text can run 356 end_code address below which program text can run 357 start_stack address of the start of the main process stack 358 esp current value of ESP 359 eip current value of EIP 360 pending bitmap of pending signals 361 blocked bitmap of blocked signals 362 sigign bitmap of ignored signals 363 sigcatch bitmap of caught signals 364 0 (place holder, used to be the wchan address, 365 use /proc/PID/wchan instead) 366 0 (place holder) 367 0 (place holder) 368 exit_signal signal to send to parent thread on exit 369 task_cpu which CPU the task is scheduled on 370 rt_priority realtime priority 371 policy scheduling policy (man sched_setscheduler) 372 blkio_ticks time spent waiting for block IO 373 gtime guest time of the task in jiffies 374 cgtime guest time of the task children in jiffies 375 start_data address above which program data+bss is placed 376 end_data address below which program data+bss is placed 377 start_brk address above which program heap can be expanded with brk() 378 arg_start address above which program command line is placed 379 arg_end address below which program command line is placed 380 env_start address above which program environment is placed 381 env_end address below which program environment is placed 382 exit_code the thread's exit_code in the form reported by the waitpid 383 system call 384 ============= =============================================================== 385 386The /proc/PID/maps file contains the currently mapped memory regions and 387their access permissions. 388 389The format is:: 390 391 address perms offset dev inode pathname 392 393 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test 394 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test 395 0804a000-0806b000 rw-p 00000000 00:00 0 [heap] 396 a7cb1000-a7cb2000 ---p 00000000 00:00 0 397 a7cb2000-a7eb2000 rw-p 00000000 00:00 0 398 a7eb2000-a7eb3000 ---p 00000000 00:00 0 399 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 400 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6 401 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6 402 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6 403 a800b000-a800e000 rw-p 00000000 00:00 0 404 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0 405 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0 406 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0 407 a8024000-a8027000 rw-p 00000000 00:00 0 408 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2 409 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2 410 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2 411 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack] 412 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso] 413 414where "address" is the address space in the process that it occupies, "perms" 415is a set of permissions:: 416 417 r = read 418 w = write 419 x = execute 420 s = shared 421 p = private (copy on write) 422 423"offset" is the offset into the mapping, "dev" is the device (major:minor), and 424"inode" is the inode on that device. 0 indicates that no inode is associated 425with the memory region, as the case would be with BSS (uninitialized data). 426The "pathname" shows the name associated file for this mapping. If the mapping 427is not associated with a file: 428 429 ============= ==================================== 430 [heap] the heap of the program 431 [stack] the stack of the main process 432 [vdso] the "virtual dynamic shared object", 433 the kernel system call handler 434 [anon:<name>] an anonymous mapping that has been 435 named by userspace 436 ============= ==================================== 437 438 or if empty, the mapping is anonymous. 439 440The /proc/PID/smaps is an extension based on maps, showing the memory 441consumption for each of the process's mappings. For each mapping (aka Virtual 442Memory Area, or VMA) there is a series of lines such as the following:: 443 444 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash 445 446 Size: 1084 kB 447 KernelPageSize: 4 kB 448 MMUPageSize: 4 kB 449 Rss: 892 kB 450 Pss: 374 kB 451 Shared_Clean: 892 kB 452 Shared_Dirty: 0 kB 453 Private_Clean: 0 kB 454 Private_Dirty: 0 kB 455 Referenced: 892 kB 456 Anonymous: 0 kB 457 LazyFree: 0 kB 458 AnonHugePages: 0 kB 459 ShmemPmdMapped: 0 kB 460 Shared_Hugetlb: 0 kB 461 Private_Hugetlb: 0 kB 462 Swap: 0 kB 463 SwapPss: 0 kB 464 KernelPageSize: 4 kB 465 MMUPageSize: 4 kB 466 Locked: 0 kB 467 THPeligible: 0 468 VmFlags: rd ex mr mw me dw 469 470The first of these lines shows the same information as is displayed for the 471mapping in /proc/PID/maps. Following lines show the size of the mapping 472(size); the size of each page allocated when backing a VMA (KernelPageSize), 473which is usually the same as the size in the page table entries; the page size 474used by the MMU when backing a VMA (in most cases, the same as KernelPageSize); 475the amount of the mapping that is currently resident in RAM (RSS); the 476process' proportional share of this mapping (PSS); and the number of clean and 477dirty shared and private pages in the mapping. 478 479The "proportional set size" (PSS) of a process is the count of pages it has 480in memory, where each page is divided by the number of processes sharing it. 481So if a process has 1000 pages all to itself, and 1000 shared with one other 482process, its PSS will be 1500. 483 484Note that even a page which is part of a MAP_SHARED mapping, but has only 485a single pte mapped, i.e. is currently used by only one process, is accounted 486as private and not as shared. 487 488"Referenced" indicates the amount of memory currently marked as referenced or 489accessed. 490 491"Anonymous" shows the amount of memory that does not belong to any file. Even 492a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE 493and a page is modified, the file page is replaced by a private anonymous copy. 494 495"LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE). 496The memory isn't freed immediately with madvise(). It's freed in memory 497pressure if the memory is clean. Please note that the printed value might 498be lower than the real value due to optimizations used in the current 499implementation. If this is not desirable please file a bug report. 500 501"AnonHugePages" shows the ammount of memory backed by transparent hugepage. 502 503"ShmemPmdMapped" shows the ammount of shared (shmem/tmpfs) memory backed by 504huge pages. 505 506"Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by 507hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical 508reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field. 509 510"Swap" shows how much would-be-anonymous memory is also used, but out on swap. 511 512For shmem mappings, "Swap" includes also the size of the mapped (and not 513replaced by copy-on-write) part of the underlying shmem object out on swap. 514"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this 515does not take into account swapped out page of underlying shmem objects. 516"Locked" indicates whether the mapping is locked in memory or not. 517"THPeligible" indicates whether the mapping is eligible for allocating THP 518pages - 1 if true, 0 otherwise. It just shows the current status. 519 520"VmFlags" field deserves a separate description. This member represents the 521kernel flags associated with the particular virtual memory area in two letter 522encoded manner. The codes are the following: 523 524 == ======================================= 525 rd readable 526 wr writeable 527 ex executable 528 sh shared 529 mr may read 530 mw may write 531 me may execute 532 ms may share 533 gd stack segment growns down 534 pf pure PFN range 535 dw disabled write to the mapped file 536 lo pages are locked in memory 537 io memory mapped I/O area 538 sr sequential read advise provided 539 rr random read advise provided 540 dc do not copy area on fork 541 de do not expand area on remapping 542 ac area is accountable 543 nr swap space is not reserved for the area 544 ht area uses huge tlb pages 545 sf synchronous page fault 546 ar architecture specific flag 547 wf wipe on fork 548 dd do not include area into core dump 549 sd soft dirty flag 550 mm mixed map area 551 hg huge page advise flag 552 nh no huge page advise flag 553 mg mergable advise flag 554 bt arm64 BTI guarded page 555 mt arm64 MTE allocation tags are enabled 556 um userfaultfd missing tracking 557 uw userfaultfd wr-protect tracking 558 == ======================================= 559 560Note that there is no guarantee that every flag and associated mnemonic will 561be present in all further kernel releases. Things get changed, the flags may 562be vanished or the reverse -- new added. Interpretation of their meaning 563might change in future as well. So each consumer of these flags has to 564follow each specific kernel version for the exact semantic. 565 566This file is only present if the CONFIG_MMU kernel configuration option is 567enabled. 568 569Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent 570output can be achieved only in the single read call). 571 572This typically manifests when doing partial reads of these files while the 573memory map is being modified. Despite the races, we do provide the following 574guarantees: 575 5761) The mapped addresses never go backwards, which implies no two 577 regions will ever overlap. 5782) If there is something at a given vaddr during the entirety of the 579 life of the smaps/maps walk, there will be some output for it. 580 581The /proc/PID/smaps_rollup file includes the same fields as /proc/PID/smaps, 582but their values are the sums of the corresponding values for all mappings of 583the process. Additionally, it contains these fields: 584 585- Pss_Anon 586- Pss_File 587- Pss_Shmem 588 589They represent the proportional shares of anonymous, file, and shmem pages, as 590described for smaps above. These fields are omitted in smaps since each 591mapping identifies the type (anon, file, or shmem) of all pages it contains. 592Thus all information in smaps_rollup can be derived from smaps, but at a 593significantly higher cost. 594 595The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG 596bits on both physical and virtual pages associated with a process, and the 597soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst 598for details). 599To clear the bits for all the pages associated with the process:: 600 601 > echo 1 > /proc/PID/clear_refs 602 603To clear the bits for the anonymous pages associated with the process:: 604 605 > echo 2 > /proc/PID/clear_refs 606 607To clear the bits for the file mapped pages associated with the process:: 608 609 > echo 3 > /proc/PID/clear_refs 610 611To clear the soft-dirty bit:: 612 613 > echo 4 > /proc/PID/clear_refs 614 615To reset the peak resident set size ("high water mark") to the process's 616current value:: 617 618 > echo 5 > /proc/PID/clear_refs 619 620Any other value written to /proc/PID/clear_refs will have no effect. 621 622The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags 623using /proc/kpageflags and number of times a page is mapped using 624/proc/kpagecount. For detailed explanation, see 625Documentation/admin-guide/mm/pagemap.rst. 626 627The /proc/pid/numa_maps is an extension based on maps, showing the memory 628locality and binding policy, as well as the memory usage (in pages) of 629each mapping. The output follows a general format where mapping details get 630summarized separated by blank spaces, one mapping per each file line:: 631 632 address policy mapping details 633 634 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4 635 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4 636 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4 637 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 638 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 639 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4 640 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4 641 320698b000 default file=/lib64/libc-2.12.so 642 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4 643 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 644 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4 645 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4 646 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4 647 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048 648 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4 649 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4 650 651Where: 652 653"address" is the starting address for the mapping; 654 655"policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst); 656 657"mapping details" summarizes mapping data such as mapping type, page usage counters, 658node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page 659size, in KB, that is backing the mapping up. 660 6611.2 Kernel data 662--------------- 663 664Similar to the process entries, the kernel data files give information about 665the running kernel. The files used to obtain this information are contained in 666/proc and are listed in Table 1-5. Not all of these will be present in your 667system. It depends on the kernel configuration and the loaded modules, which 668files are there, and which are missing. 669 670.. table:: Table 1-5: Kernel info in /proc 671 672 ============ =============================================================== 673 File Content 674 ============ =============================================================== 675 apm Advanced power management info 676 buddyinfo Kernel memory allocator information (see text) (2.5) 677 bus Directory containing bus specific information 678 cmdline Kernel command line 679 cpuinfo Info about the CPU 680 devices Available devices (block and character) 681 dma Used DMS channels 682 filesystems Supported filesystems 683 driver Various drivers grouped here, currently rtc (2.4) 684 execdomains Execdomains, related to security (2.4) 685 fb Frame Buffer devices (2.4) 686 fs File system parameters, currently nfs/exports (2.4) 687 ide Directory containing info about the IDE subsystem 688 interrupts Interrupt usage 689 iomem Memory map (2.4) 690 ioports I/O port usage 691 irq Masks for irq to cpu affinity (2.4)(smp?) 692 isapnp ISA PnP (Plug&Play) Info (2.4) 693 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4)) 694 kmsg Kernel messages 695 ksyms Kernel symbol table 696 loadavg Load average of last 1, 5 & 15 minutes; 697 number of processes currently runnable (running or on ready queue); 698 total number of processes in system; 699 last pid created. 700 All fields are separated by one space except "number of 701 processes currently runnable" and "total number of processes 702 in system", which are separated by a slash ('/'). Example: 703 0.61 0.61 0.55 3/828 22084 704 locks Kernel locks 705 meminfo Memory info 706 misc Miscellaneous 707 modules List of loaded modules 708 mounts Mounted filesystems 709 net Networking info (see text) 710 pagetypeinfo Additional page allocator information (see text) (2.5) 711 partitions Table of partitions known to the system 712 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/, 713 decoupled by lspci (2.4) 714 rtc Real time clock 715 scsi SCSI info (see text) 716 slabinfo Slab pool info 717 softirqs softirq usage 718 stat Overall statistics 719 swaps Swap space utilization 720 sys See chapter 2 721 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4) 722 tty Info of tty drivers 723 uptime Wall clock since boot, combined idle time of all cpus 724 version Kernel version 725 video bttv info of video resources (2.4) 726 vmallocinfo Show vmalloced areas 727 ============ =============================================================== 728 729You can, for example, check which interrupts are currently in use and what 730they are used for by looking in the file /proc/interrupts:: 731 732 > cat /proc/interrupts 733 CPU0 734 0: 8728810 XT-PIC timer 735 1: 895 XT-PIC keyboard 736 2: 0 XT-PIC cascade 737 3: 531695 XT-PIC aha152x 738 4: 2014133 XT-PIC serial 739 5: 44401 XT-PIC pcnet_cs 740 8: 2 XT-PIC rtc 741 11: 8 XT-PIC i82365 742 12: 182918 XT-PIC PS/2 Mouse 743 13: 1 XT-PIC fpu 744 14: 1232265 XT-PIC ide0 745 15: 7 XT-PIC ide1 746 NMI: 0 747 748In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the 749output of a SMP machine):: 750 751 > cat /proc/interrupts 752 753 CPU0 CPU1 754 0: 1243498 1214548 IO-APIC-edge timer 755 1: 8949 8958 IO-APIC-edge keyboard 756 2: 0 0 XT-PIC cascade 757 5: 11286 10161 IO-APIC-edge soundblaster 758 8: 1 0 IO-APIC-edge rtc 759 9: 27422 27407 IO-APIC-edge 3c503 760 12: 113645 113873 IO-APIC-edge PS/2 Mouse 761 13: 0 0 XT-PIC fpu 762 14: 22491 24012 IO-APIC-edge ide0 763 15: 2183 2415 IO-APIC-edge ide1 764 17: 30564 30414 IO-APIC-level eth0 765 18: 177 164 IO-APIC-level bttv 766 NMI: 2457961 2457959 767 LOC: 2457882 2457881 768 ERR: 2155 769 770NMI is incremented in this case because every timer interrupt generates a NMI 771(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups. 772 773LOC is the local interrupt counter of the internal APIC of every CPU. 774 775ERR is incremented in the case of errors in the IO-APIC bus (the bus that 776connects the CPUs in a SMP system. This means that an error has been detected, 777the IO-APIC automatically retry the transmission, so it should not be a big 778problem, but you should read the SMP-FAQ. 779 780In 2.6.2* /proc/interrupts was expanded again. This time the goal was for 781/proc/interrupts to display every IRQ vector in use by the system, not 782just those considered 'most important'. The new vectors are: 783 784THR 785 interrupt raised when a machine check threshold counter 786 (typically counting ECC corrected errors of memory or cache) exceeds 787 a configurable threshold. Only available on some systems. 788 789TRM 790 a thermal event interrupt occurs when a temperature threshold 791 has been exceeded for the CPU. This interrupt may also be generated 792 when the temperature drops back to normal. 793 794SPU 795 a spurious interrupt is some interrupt that was raised then lowered 796 by some IO device before it could be fully processed by the APIC. Hence 797 the APIC sees the interrupt but does not know what device it came from. 798 For this case the APIC will generate the interrupt with a IRQ vector 799 of 0xff. This might also be generated by chipset bugs. 800 801RES, CAL, TLB 802 rescheduling, call and TLB flush interrupts are 803 sent from one CPU to another per the needs of the OS. Typically, 804 their statistics are used by kernel developers and interested users to 805 determine the occurrence of interrupts of the given type. 806 807The above IRQ vectors are displayed only when relevant. For example, 808the threshold vector does not exist on x86_64 platforms. Others are 809suppressed when the system is a uniprocessor. As of this writing, only 810i386 and x86_64 platforms support the new IRQ vector displays. 811 812Of some interest is the introduction of the /proc/irq directory to 2.4. 813It could be used to set IRQ to CPU affinity. This means that you can "hook" an 814IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the 815irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and 816prof_cpu_mask. 817 818For example:: 819 820 > ls /proc/irq/ 821 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask 822 1 11 13 15 17 19 3 5 7 9 default_smp_affinity 823 > ls /proc/irq/0/ 824 smp_affinity 825 826smp_affinity is a bitmask, in which you can specify which CPUs can handle the 827IRQ. You can set it by doing:: 828 829 > echo 1 > /proc/irq/10/smp_affinity 830 831This means that only the first CPU will handle the IRQ, but you can also echo 8325 which means that only the first and third CPU can handle the IRQ. 833 834The contents of each smp_affinity file is the same by default:: 835 836 > cat /proc/irq/0/smp_affinity 837 ffffffff 838 839There is an alternate interface, smp_affinity_list which allows specifying 840a CPU range instead of a bitmask:: 841 842 > cat /proc/irq/0/smp_affinity_list 843 1024-1031 844 845The default_smp_affinity mask applies to all non-active IRQs, which are the 846IRQs which have not yet been allocated/activated, and hence which lack a 847/proc/irq/[0-9]* directory. 848 849The node file on an SMP system shows the node to which the device using the IRQ 850reports itself as being attached. This hardware locality information does not 851include information about any possible driver locality preference. 852 853prof_cpu_mask specifies which CPUs are to be profiled by the system wide 854profiler. Default value is ffffffff (all CPUs if there are only 32 of them). 855 856The way IRQs are routed is handled by the IO-APIC, and it's Round Robin 857between all the CPUs which are allowed to handle it. As usual the kernel has 858more info than you and does a better job than you, so the defaults are the 859best choice for almost everyone. [Note this applies only to those IO-APIC's 860that support "Round Robin" interrupt distribution.] 861 862There are three more important subdirectories in /proc: net, scsi, and sys. 863The general rule is that the contents, or even the existence of these 864directories, depend on your kernel configuration. If SCSI is not enabled, the 865directory scsi may not exist. The same is true with the net, which is there 866only when networking support is present in the running kernel. 867 868The slabinfo file gives information about memory usage at the slab level. 869Linux uses slab pools for memory management above page level in version 2.2. 870Commonly used objects have their own slab pool (such as network buffers, 871directory cache, and so on). 872 873:: 874 875 > cat /proc/buddyinfo 876 877 Node 0, zone DMA 0 4 5 4 4 3 ... 878 Node 0, zone Normal 1 0 0 1 101 8 ... 879 Node 0, zone HighMem 2 0 0 1 1 0 ... 880 881External fragmentation is a problem under some workloads, and buddyinfo is a 882useful tool for helping diagnose these problems. Buddyinfo will give you a 883clue as to how big an area you can safely allocate, or why a previous 884allocation failed. 885 886Each column represents the number of pages of a certain order which are 887available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in 888ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE 889available in ZONE_NORMAL, etc... 890 891More information relevant to external fragmentation can be found in 892pagetypeinfo:: 893 894 > cat /proc/pagetypeinfo 895 Page block order: 9 896 Pages per block: 512 897 898 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10 899 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0 900 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0 901 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2 902 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0 903 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0 904 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9 905 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0 906 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452 907 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0 908 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0 909 910 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate 911 Node 0, zone DMA 2 0 5 1 0 912 Node 0, zone DMA32 41 6 967 2 0 913 914Fragmentation avoidance in the kernel works by grouping pages of different 915migrate types into the same contiguous regions of memory called page blocks. 916A page block is typically the size of the default hugepage size, e.g. 2MB on 917X86-64. By keeping pages grouped based on their ability to move, the kernel 918can reclaim pages within a page block to satisfy a high-order allocation. 919 920The pagetypinfo begins with information on the size of a page block. It 921then gives the same type of information as buddyinfo except broken down 922by migrate-type and finishes with details on how many page blocks of each 923type exist. 924 925If min_free_kbytes has been tuned correctly (recommendations made by hugeadm 926from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can 927make an estimate of the likely number of huge pages that can be allocated 928at a given point in time. All the "Movable" blocks should be allocatable 929unless memory has been mlock()'d. Some of the Reclaimable blocks should 930also be allocatable although a lot of filesystem metadata may have to be 931reclaimed to achieve this. 932 933 934meminfo 935~~~~~~~ 936 937Provides information about distribution and utilization of memory. This 938varies by architecture and compile options. Some of the counters reported 939here overlap. The memory reported by the non overlapping counters may not 940add up to the overall memory usage and the difference for some workloads 941can be substantial. In many cases there are other means to find out 942additional memory using subsystem specific interfaces, for instance 943/proc/net/sockstat for TCP memory allocations. 944 945The following is from a 16GB PIII, which has highmem enabled. 946You may not have all of these fields. 947 948:: 949 950 > cat /proc/meminfo 951 952 MemTotal: 16344972 kB 953 MemFree: 13634064 kB 954 MemAvailable: 14836172 kB 955 Buffers: 3656 kB 956 Cached: 1195708 kB 957 SwapCached: 0 kB 958 Active: 891636 kB 959 Inactive: 1077224 kB 960 HighTotal: 15597528 kB 961 HighFree: 13629632 kB 962 LowTotal: 747444 kB 963 LowFree: 4432 kB 964 SwapTotal: 0 kB 965 SwapFree: 0 kB 966 Dirty: 968 kB 967 Writeback: 0 kB 968 AnonPages: 861800 kB 969 Mapped: 280372 kB 970 Shmem: 644 kB 971 KReclaimable: 168048 kB 972 Slab: 284364 kB 973 SReclaimable: 159856 kB 974 SUnreclaim: 124508 kB 975 PageTables: 24448 kB 976 SecPageTables: 0 kB 977 NFS_Unstable: 0 kB 978 Bounce: 0 kB 979 WritebackTmp: 0 kB 980 CommitLimit: 7669796 kB 981 Committed_AS: 100056 kB 982 VmallocTotal: 112216 kB 983 VmallocUsed: 428 kB 984 VmallocChunk: 111088 kB 985 Percpu: 62080 kB 986 HardwareCorrupted: 0 kB 987 AnonHugePages: 49152 kB 988 ShmemHugePages: 0 kB 989 ShmemPmdMapped: 0 kB 990 991MemTotal 992 Total usable RAM (i.e. physical RAM minus a few reserved 993 bits and the kernel binary code) 994MemFree 995 The sum of LowFree+HighFree 996MemAvailable 997 An estimate of how much memory is available for starting new 998 applications, without swapping. Calculated from MemFree, 999 SReclaimable, the size of the file LRU lists, and the low 1000 watermarks in each zone. 1001 The estimate takes into account that the system needs some 1002 page cache to function well, and that not all reclaimable 1003 slab will be reclaimable, due to items being in use. The 1004 impact of those factors will vary from system to system. 1005Buffers 1006 Relatively temporary storage for raw disk blocks 1007 shouldn't get tremendously large (20MB or so) 1008Cached 1009 in-memory cache for files read from the disk (the 1010 pagecache). Doesn't include SwapCached 1011SwapCached 1012 Memory that once was swapped out, is swapped back in but 1013 still also is in the swapfile (if memory is needed it 1014 doesn't need to be swapped out AGAIN because it is already 1015 in the swapfile. This saves I/O) 1016Active 1017 Memory that has been used more recently and usually not 1018 reclaimed unless absolutely necessary. 1019Inactive 1020 Memory which has been less recently used. It is more 1021 eligible to be reclaimed for other purposes 1022HighTotal, HighFree 1023 Highmem is all memory above ~860MB of physical memory. 1024 Highmem areas are for use by userspace programs, or 1025 for the pagecache. The kernel must use tricks to access 1026 this memory, making it slower to access than lowmem. 1027LowTotal, LowFree 1028 Lowmem is memory which can be used for everything that 1029 highmem can be used for, but it is also available for the 1030 kernel's use for its own data structures. Among many 1031 other things, it is where everything from the Slab is 1032 allocated. Bad things happen when you're out of lowmem. 1033SwapTotal 1034 total amount of swap space available 1035SwapFree 1036 Memory which has been evicted from RAM, and is temporarily 1037 on the disk 1038Dirty 1039 Memory which is waiting to get written back to the disk 1040Writeback 1041 Memory which is actively being written back to the disk 1042AnonPages 1043 Non-file backed pages mapped into userspace page tables 1044HardwareCorrupted 1045 The amount of RAM/memory in KB, the kernel identifies as 1046 corrupted. 1047AnonHugePages 1048 Non-file backed huge pages mapped into userspace page tables 1049Mapped 1050 files which have been mmaped, such as libraries 1051Shmem 1052 Total memory used by shared memory (shmem) and tmpfs 1053ShmemHugePages 1054 Memory used by shared memory (shmem) and tmpfs allocated 1055 with huge pages 1056ShmemPmdMapped 1057 Shared memory mapped into userspace with huge pages 1058KReclaimable 1059 Kernel allocations that the kernel will attempt to reclaim 1060 under memory pressure. Includes SReclaimable (below), and other 1061 direct allocations with a shrinker. 1062Slab 1063 in-kernel data structures cache 1064SReclaimable 1065 Part of Slab, that might be reclaimed, such as caches 1066SUnreclaim 1067 Part of Slab, that cannot be reclaimed on memory pressure 1068PageTables 1069 amount of memory dedicated to the lowest level of page 1070 tables. 1071SecPageTables 1072 Memory consumed by secondary page tables, this currently 1073 currently includes KVM mmu allocations on x86 and arm64. 1074NFS_Unstable 1075 Always zero. Previous counted pages which had been written to 1076 the server, but has not been committed to stable storage. 1077Bounce 1078 Memory used for block device "bounce buffers" 1079WritebackTmp 1080 Memory used by FUSE for temporary writeback buffers 1081CommitLimit 1082 Based on the overcommit ratio ('vm.overcommit_ratio'), 1083 this is the total amount of memory currently available to 1084 be allocated on the system. This limit is only adhered to 1085 if strict overcommit accounting is enabled (mode 2 in 1086 'vm.overcommit_memory'). 1087 1088 The CommitLimit is calculated with the following formula:: 1089 1090 CommitLimit = ([total RAM pages] - [total huge TLB pages]) * 1091 overcommit_ratio / 100 + [total swap pages] 1092 1093 For example, on a system with 1G of physical RAM and 7G 1094 of swap with a `vm.overcommit_ratio` of 30 it would 1095 yield a CommitLimit of 7.3G. 1096 1097 For more details, see the memory overcommit documentation 1098 in vm/overcommit-accounting. 1099Committed_AS 1100 The amount of memory presently allocated on the system. 1101 The committed memory is a sum of all of the memory which 1102 has been allocated by processes, even if it has not been 1103 "used" by them as of yet. A process which malloc()'s 1G 1104 of memory, but only touches 300M of it will show up as 1105 using 1G. This 1G is memory which has been "committed" to 1106 by the VM and can be used at any time by the allocating 1107 application. With strict overcommit enabled on the system 1108 (mode 2 in 'vm.overcommit_memory'), allocations which would 1109 exceed the CommitLimit (detailed above) will not be permitted. 1110 This is useful if one needs to guarantee that processes will 1111 not fail due to lack of memory once that memory has been 1112 successfully allocated. 1113VmallocTotal 1114 total size of vmalloc memory area 1115VmallocUsed 1116 amount of vmalloc area which is used 1117VmallocChunk 1118 largest contiguous block of vmalloc area which is free 1119Percpu 1120 Memory allocated to the percpu allocator used to back percpu 1121 allocations. This stat excludes the cost of metadata. 1122 1123vmallocinfo 1124~~~~~~~~~~~ 1125 1126Provides information about vmalloced/vmaped areas. One line per area, 1127containing the virtual address range of the area, size in bytes, 1128caller information of the creator, and optional information depending 1129on the kind of area: 1130 1131 ========== =================================================== 1132 pages=nr number of pages 1133 phys=addr if a physical address was specified 1134 ioremap I/O mapping (ioremap() and friends) 1135 vmalloc vmalloc() area 1136 vmap vmap()ed pages 1137 user VM_USERMAP area 1138 vpages buffer for pages pointers was vmalloced (huge area) 1139 N<node>=nr (Only on NUMA kernels) 1140 Number of pages allocated on memory node <node> 1141 ========== =================================================== 1142 1143:: 1144 1145 > cat /proc/vmallocinfo 1146 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ... 1147 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128 1148 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ... 1149 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64 1150 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f... 1151 phys=7fee8000 ioremap 1152 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f... 1153 phys=7fee7000 ioremap 1154 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210 1155 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ... 1156 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3 1157 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ... 1158 pages=2 vmalloc N1=2 1159 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ... 1160 /0x130 [x_tables] pages=4 vmalloc N0=4 1161 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ... 1162 pages=14 vmalloc N2=14 1163 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ... 1164 pages=4 vmalloc N1=4 1165 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ... 1166 pages=2 vmalloc N1=2 1167 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ... 1168 pages=10 vmalloc N0=10 1169 1170 1171softirqs 1172~~~~~~~~ 1173 1174Provides counts of softirq handlers serviced since boot time, for each CPU. 1175 1176:: 1177 1178 > cat /proc/softirqs 1179 CPU0 CPU1 CPU2 CPU3 1180 HI: 0 0 0 0 1181 TIMER: 27166 27120 27097 27034 1182 NET_TX: 0 0 0 17 1183 NET_RX: 42 0 0 39 1184 BLOCK: 0 0 107 1121 1185 TASKLET: 0 0 0 290 1186 SCHED: 27035 26983 26971 26746 1187 HRTIMER: 0 0 0 0 1188 RCU: 1678 1769 2178 2250 1189 1190 11911.3 IDE devices in /proc/ide 1192---------------------------- 1193 1194The subdirectory /proc/ide contains information about all IDE devices of which 1195the kernel is aware. There is one subdirectory for each IDE controller, the 1196file drivers and a link for each IDE device, pointing to the device directory 1197in the controller specific subtree. 1198 1199The file 'drivers' contains general information about the drivers used for the 1200IDE devices:: 1201 1202 > cat /proc/ide/drivers 1203 ide-cdrom version 4.53 1204 ide-disk version 1.08 1205 1206More detailed information can be found in the controller specific 1207subdirectories. These are named ide0, ide1 and so on. Each of these 1208directories contains the files shown in table 1-6. 1209 1210 1211.. table:: Table 1-6: IDE controller info in /proc/ide/ide? 1212 1213 ======= ======================================= 1214 File Content 1215 ======= ======================================= 1216 channel IDE channel (0 or 1) 1217 config Configuration (only for PCI/IDE bridge) 1218 mate Mate name 1219 model Type/Chipset of IDE controller 1220 ======= ======================================= 1221 1222Each device connected to a controller has a separate subdirectory in the 1223controllers directory. The files listed in table 1-7 are contained in these 1224directories. 1225 1226 1227.. table:: Table 1-7: IDE device information 1228 1229 ================ ========================================== 1230 File Content 1231 ================ ========================================== 1232 cache The cache 1233 capacity Capacity of the medium (in 512Byte blocks) 1234 driver driver and version 1235 geometry physical and logical geometry 1236 identify device identify block 1237 media media type 1238 model device identifier 1239 settings device setup 1240 smart_thresholds IDE disk management thresholds 1241 smart_values IDE disk management values 1242 ================ ========================================== 1243 1244The most interesting file is ``settings``. This file contains a nice 1245overview of the drive parameters:: 1246 1247 # cat /proc/ide/ide0/hda/settings 1248 name value min max mode 1249 ---- ----- --- --- ---- 1250 bios_cyl 526 0 65535 rw 1251 bios_head 255 0 255 rw 1252 bios_sect 63 0 63 rw 1253 breada_readahead 4 0 127 rw 1254 bswap 0 0 1 r 1255 file_readahead 72 0 2097151 rw 1256 io_32bit 0 0 3 rw 1257 keepsettings 0 0 1 rw 1258 max_kb_per_request 122 1 127 rw 1259 multcount 0 0 8 rw 1260 nice1 1 0 1 rw 1261 nowerr 0 0 1 rw 1262 pio_mode write-only 0 255 w 1263 slow 0 0 1 rw 1264 unmaskirq 0 0 1 rw 1265 using_dma 0 0 1 rw 1266 1267 12681.4 Networking info in /proc/net 1269-------------------------------- 1270 1271The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the 1272additional values you get for IP version 6 if you configure the kernel to 1273support this. Table 1-9 lists the files and their meaning. 1274 1275 1276.. table:: Table 1-8: IPv6 info in /proc/net 1277 1278 ========== ===================================================== 1279 File Content 1280 ========== ===================================================== 1281 udp6 UDP sockets (IPv6) 1282 tcp6 TCP sockets (IPv6) 1283 raw6 Raw device statistics (IPv6) 1284 igmp6 IP multicast addresses, which this host joined (IPv6) 1285 if_inet6 List of IPv6 interface addresses 1286 ipv6_route Kernel routing table for IPv6 1287 rt6_stats Global IPv6 routing tables statistics 1288 sockstat6 Socket statistics (IPv6) 1289 snmp6 Snmp data (IPv6) 1290 ========== ===================================================== 1291 1292.. table:: Table 1-9: Network info in /proc/net 1293 1294 ============= ================================================================ 1295 File Content 1296 ============= ================================================================ 1297 arp Kernel ARP table 1298 dev network devices with statistics 1299 dev_mcast the Layer2 multicast groups a device is listening too 1300 (interface index, label, number of references, number of bound 1301 addresses). 1302 dev_stat network device status 1303 ip_fwchains Firewall chain linkage 1304 ip_fwnames Firewall chain names 1305 ip_masq Directory containing the masquerading tables 1306 ip_masquerade Major masquerading table 1307 netstat Network statistics 1308 raw raw device statistics 1309 route Kernel routing table 1310 rpc Directory containing rpc info 1311 rt_cache Routing cache 1312 snmp SNMP data 1313 sockstat Socket statistics 1314 tcp TCP sockets 1315 udp UDP sockets 1316 unix UNIX domain sockets 1317 wireless Wireless interface data (Wavelan etc) 1318 igmp IP multicast addresses, which this host joined 1319 psched Global packet scheduler parameters. 1320 netlink List of PF_NETLINK sockets 1321 ip_mr_vifs List of multicast virtual interfaces 1322 ip_mr_cache List of multicast routing cache 1323 ============= ================================================================ 1324 1325You can use this information to see which network devices are available in 1326your system and how much traffic was routed over those devices:: 1327 1328 > cat /proc/net/dev 1329 Inter-|Receive |[... 1330 face |bytes packets errs drop fifo frame compressed multicast|[... 1331 lo: 908188 5596 0 0 0 0 0 0 [... 1332 ppp0:15475140 20721 410 0 0 410 0 0 [... 1333 eth0: 614530 7085 0 0 0 0 0 1 [... 1334 1335 ...] Transmit 1336 ...] bytes packets errs drop fifo colls carrier compressed 1337 ...] 908188 5596 0 0 0 0 0 0 1338 ...] 1375103 17405 0 0 0 0 0 0 1339 ...] 1703981 5535 0 0 0 3 0 0 1340 1341In addition, each Channel Bond interface has its own directory. For 1342example, the bond0 device will have a directory called /proc/net/bond0/. 1343It will contain information that is specific to that bond, such as the 1344current slaves of the bond, the link status of the slaves, and how 1345many times the slaves link has failed. 1346 13471.5 SCSI info 1348------------- 1349 1350If you have a SCSI host adapter in your system, you'll find a subdirectory 1351named after the driver for this adapter in /proc/scsi. You'll also see a list 1352of all recognized SCSI devices in /proc/scsi:: 1353 1354 >cat /proc/scsi/scsi 1355 Attached devices: 1356 Host: scsi0 Channel: 00 Id: 00 Lun: 00 1357 Vendor: IBM Model: DGHS09U Rev: 03E0 1358 Type: Direct-Access ANSI SCSI revision: 03 1359 Host: scsi0 Channel: 00 Id: 06 Lun: 00 1360 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04 1361 Type: CD-ROM ANSI SCSI revision: 02 1362 1363 1364The directory named after the driver has one file for each adapter found in 1365the system. These files contain information about the controller, including 1366the used IRQ and the IO address range. The amount of information shown is 1367dependent on the adapter you use. The example shows the output for an Adaptec 1368AHA-2940 SCSI adapter:: 1369 1370 > cat /proc/scsi/aic7xxx/0 1371 1372 Adaptec AIC7xxx driver version: 5.1.19/3.2.4 1373 Compile Options: 1374 TCQ Enabled By Default : Disabled 1375 AIC7XXX_PROC_STATS : Disabled 1376 AIC7XXX_RESET_DELAY : 5 1377 Adapter Configuration: 1378 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter 1379 Ultra Wide Controller 1380 PCI MMAPed I/O Base: 0xeb001000 1381 Adapter SEEPROM Config: SEEPROM found and used. 1382 Adaptec SCSI BIOS: Enabled 1383 IRQ: 10 1384 SCBs: Active 0, Max Active 2, 1385 Allocated 15, HW 16, Page 255 1386 Interrupts: 160328 1387 BIOS Control Word: 0x18b6 1388 Adapter Control Word: 0x005b 1389 Extended Translation: Enabled 1390 Disconnect Enable Flags: 0xffff 1391 Ultra Enable Flags: 0x0001 1392 Tag Queue Enable Flags: 0x0000 1393 Ordered Queue Tag Flags: 0x0000 1394 Default Tag Queue Depth: 8 1395 Tagged Queue By Device array for aic7xxx host instance 0: 1396 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255} 1397 Actual queue depth per device for aic7xxx host instance 0: 1398 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1} 1399 Statistics: 1400 (scsi0:0:0:0) 1401 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8 1402 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0) 1403 Total transfers 160151 (74577 reads and 85574 writes) 1404 (scsi0:0:6:0) 1405 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15 1406 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0) 1407 Total transfers 0 (0 reads and 0 writes) 1408 1409 14101.6 Parallel port info in /proc/parport 1411--------------------------------------- 1412 1413The directory /proc/parport contains information about the parallel ports of 1414your system. It has one subdirectory for each port, named after the port 1415number (0,1,2,...). 1416 1417These directories contain the four files shown in Table 1-10. 1418 1419 1420.. table:: Table 1-10: Files in /proc/parport 1421 1422 ========= ==================================================================== 1423 File Content 1424 ========= ==================================================================== 1425 autoprobe Any IEEE-1284 device ID information that has been acquired. 1426 devices list of the device drivers using that port. A + will appear by the 1427 name of the device currently using the port (it might not appear 1428 against any). 1429 hardware Parallel port's base address, IRQ line and DMA channel. 1430 irq IRQ that parport is using for that port. This is in a separate 1431 file to allow you to alter it by writing a new value in (IRQ 1432 number or none). 1433 ========= ==================================================================== 1434 14351.7 TTY info in /proc/tty 1436------------------------- 1437 1438Information about the available and actually used tty's can be found in the 1439directory /proc/tty. You'll find entries for drivers and line disciplines in 1440this directory, as shown in Table 1-11. 1441 1442 1443.. table:: Table 1-11: Files in /proc/tty 1444 1445 ============= ============================================== 1446 File Content 1447 ============= ============================================== 1448 drivers list of drivers and their usage 1449 ldiscs registered line disciplines 1450 driver/serial usage statistic and status of single tty lines 1451 ============= ============================================== 1452 1453To see which tty's are currently in use, you can simply look into the file 1454/proc/tty/drivers:: 1455 1456 > cat /proc/tty/drivers 1457 pty_slave /dev/pts 136 0-255 pty:slave 1458 pty_master /dev/ptm 128 0-255 pty:master 1459 pty_slave /dev/ttyp 3 0-255 pty:slave 1460 pty_master /dev/pty 2 0-255 pty:master 1461 serial /dev/cua 5 64-67 serial:callout 1462 serial /dev/ttyS 4 64-67 serial 1463 /dev/tty0 /dev/tty0 4 0 system:vtmaster 1464 /dev/ptmx /dev/ptmx 5 2 system 1465 /dev/console /dev/console 5 1 system:console 1466 /dev/tty /dev/tty 5 0 system:/dev/tty 1467 unknown /dev/tty 4 1-63 console 1468 1469 14701.8 Miscellaneous kernel statistics in /proc/stat 1471------------------------------------------------- 1472 1473Various pieces of information about kernel activity are available in the 1474/proc/stat file. All of the numbers reported in this file are aggregates 1475since the system first booted. For a quick look, simply cat the file:: 1476 1477 > cat /proc/stat 1478 cpu 2255 34 2290 22625563 6290 127 456 0 0 0 1479 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0 1480 cpu1 1123 0 849 11313845 2614 0 18 0 0 0 1481 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...] 1482 ctxt 1990473 1483 btime 1062191376 1484 processes 2915 1485 procs_running 1 1486 procs_blocked 0 1487 softirq 183433 0 21755 12 39 1137 231 21459 2263 1488 1489The very first "cpu" line aggregates the numbers in all of the other "cpuN" 1490lines. These numbers identify the amount of time the CPU has spent performing 1491different kinds of work. Time units are in USER_HZ (typically hundredths of a 1492second). The meanings of the columns are as follows, from left to right: 1493 1494- user: normal processes executing in user mode 1495- nice: niced processes executing in user mode 1496- system: processes executing in kernel mode 1497- idle: twiddling thumbs 1498- iowait: In a word, iowait stands for waiting for I/O to complete. But there 1499 are several problems: 1500 1501 1. CPU will not wait for I/O to complete, iowait is the time that a task is 1502 waiting for I/O to complete. When CPU goes into idle state for 1503 outstanding task I/O, another task will be scheduled on this CPU. 1504 2. In a multi-core CPU, the task waiting for I/O to complete is not running 1505 on any CPU, so the iowait of each CPU is difficult to calculate. 1506 3. The value of iowait field in /proc/stat will decrease in certain 1507 conditions. 1508 1509 So, the iowait is not reliable by reading from /proc/stat. 1510- irq: servicing interrupts 1511- softirq: servicing softirqs 1512- steal: involuntary wait 1513- guest: running a normal guest 1514- guest_nice: running a niced guest 1515 1516The "intr" line gives counts of interrupts serviced since boot time, for each 1517of the possible system interrupts. The first column is the total of all 1518interrupts serviced including unnumbered architecture specific interrupts; 1519each subsequent column is the total for that particular numbered interrupt. 1520Unnumbered interrupts are not shown, only summed into the total. 1521 1522The "ctxt" line gives the total number of context switches across all CPUs. 1523 1524The "btime" line gives the time at which the system booted, in seconds since 1525the Unix epoch. 1526 1527The "processes" line gives the number of processes and threads created, which 1528includes (but is not limited to) those created by calls to the fork() and 1529clone() system calls. 1530 1531The "procs_running" line gives the total number of threads that are 1532running or ready to run (i.e., the total number of runnable threads). 1533 1534The "procs_blocked" line gives the number of processes currently blocked, 1535waiting for I/O to complete. 1536 1537The "softirq" line gives counts of softirqs serviced since boot time, for each 1538of the possible system softirqs. The first column is the total of all 1539softirqs serviced; each subsequent column is the total for that particular 1540softirq. 1541 1542 15431.9 Ext4 file system parameters 1544------------------------------- 1545 1546Information about mounted ext4 file systems can be found in 1547/proc/fs/ext4. Each mounted filesystem will have a directory in 1548/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or 1549/proc/fs/ext4/dm-0). The files in each per-device directory are shown 1550in Table 1-12, below. 1551 1552.. table:: Table 1-12: Files in /proc/fs/ext4/<devname> 1553 1554 ============== ========================================================== 1555 File Content 1556 mb_groups details of multiblock allocator buddy cache of free blocks 1557 ============== ========================================================== 1558 15591.10 /proc/consoles 1560------------------- 1561Shows registered system console lines. 1562 1563To see which character device lines are currently used for the system console 1564/dev/console, you may simply look into the file /proc/consoles:: 1565 1566 > cat /proc/consoles 1567 tty0 -WU (ECp) 4:7 1568 ttyS0 -W- (Ep) 4:64 1569 1570The columns are: 1571 1572+--------------------+-------------------------------------------------------+ 1573| device | name of the device | 1574+====================+=======================================================+ 1575| operations | * R = can do read operations | 1576| | * W = can do write operations | 1577| | * U = can do unblank | 1578+--------------------+-------------------------------------------------------+ 1579| flags | * E = it is enabled | 1580| | * C = it is preferred console | 1581| | * B = it is primary boot console | 1582| | * p = it is used for printk buffer | 1583| | * b = it is not a TTY but a Braille device | 1584| | * a = it is safe to use when cpu is offline | 1585+--------------------+-------------------------------------------------------+ 1586| major:minor | major and minor number of the device separated by a | 1587| | colon | 1588+--------------------+-------------------------------------------------------+ 1589 1590Summary 1591------- 1592 1593The /proc file system serves information about the running system. It not only 1594allows access to process data but also allows you to request the kernel status 1595by reading files in the hierarchy. 1596 1597The directory structure of /proc reflects the types of information and makes 1598it easy, if not obvious, where to look for specific data. 1599 1600Chapter 2: Modifying System Parameters 1601====================================== 1602 1603In This Chapter 1604--------------- 1605 1606* Modifying kernel parameters by writing into files found in /proc/sys 1607* Exploring the files which modify certain parameters 1608* Review of the /proc/sys file tree 1609 1610------------------------------------------------------------------------------ 1611 1612A very interesting part of /proc is the directory /proc/sys. This is not only 1613a source of information, it also allows you to change parameters within the 1614kernel. Be very careful when attempting this. You can optimize your system, 1615but you can also cause it to crash. Never alter kernel parameters on a 1616production system. Set up a development machine and test to make sure that 1617everything works the way you want it to. You may have no alternative but to 1618reboot the machine once an error has been made. 1619 1620To change a value, simply echo the new value into the file. 1621You need to be root to do this. You can create your own boot script 1622to perform this every time your system boots. 1623 1624The files in /proc/sys can be used to fine tune and monitor miscellaneous and 1625general things in the operation of the Linux kernel. Since some of the files 1626can inadvertently disrupt your system, it is advisable to read both 1627documentation and source before actually making adjustments. In any case, be 1628very careful when writing to any of these files. The entries in /proc may 1629change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt 1630review the kernel documentation in the directory /usr/src/linux/Documentation. 1631This chapter is heavily based on the documentation included in the pre 2.2 1632kernels, and became part of it in version 2.2.1 of the Linux kernel. 1633 1634Please see: Documentation/admin-guide/sysctl/ directory for descriptions of these 1635entries. 1636 1637Summary 1638------- 1639 1640Certain aspects of kernel behavior can be modified at runtime, without the 1641need to recompile the kernel, or even to reboot the system. The files in the 1642/proc/sys tree can not only be read, but also modified. You can use the echo 1643command to write value into these files, thereby changing the default settings 1644of the kernel. 1645 1646 1647Chapter 3: Per-process Parameters 1648================================= 1649 16503.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score 1651-------------------------------------------------------------------------------- 1652 1653These files can be used to adjust the badness heuristic used to select which 1654process gets killed in out of memory (oom) conditions. 1655 1656The badness heuristic assigns a value to each candidate task ranging from 0 1657(never kill) to 1000 (always kill) to determine which process is targeted. The 1658units are roughly a proportion along that range of allowed memory the process 1659may allocate from based on an estimation of its current memory and swap use. 1660For example, if a task is using all allowed memory, its badness score will be 16611000. If it is using half of its allowed memory, its score will be 500. 1662 1663The amount of "allowed" memory depends on the context in which the oom killer 1664was called. If it is due to the memory assigned to the allocating task's cpuset 1665being exhausted, the allowed memory represents the set of mems assigned to that 1666cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed 1667memory represents the set of mempolicy nodes. If it is due to a memory 1668limit (or swap limit) being reached, the allowed memory is that configured 1669limit. Finally, if it is due to the entire system being out of memory, the 1670allowed memory represents all allocatable resources. 1671 1672The value of /proc/<pid>/oom_score_adj is added to the badness score before it 1673is used to determine which task to kill. Acceptable values range from -1000 1674(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to 1675polarize the preference for oom killing either by always preferring a certain 1676task or completely disabling it. The lowest possible value, -1000, is 1677equivalent to disabling oom killing entirely for that task since it will always 1678report a badness score of 0. 1679 1680Consequently, it is very simple for userspace to define the amount of memory to 1681consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for 1682example, is roughly equivalent to allowing the remainder of tasks sharing the 1683same system, cpuset, mempolicy, or memory controller resources to use at least 168450% more memory. A value of -500, on the other hand, would be roughly 1685equivalent to discounting 50% of the task's allowed memory from being considered 1686as scoring against the task. 1687 1688For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also 1689be used to tune the badness score. Its acceptable values range from -16 1690(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17 1691(OOM_DISABLE) to disable oom killing entirely for that task. Its value is 1692scaled linearly with /proc/<pid>/oom_score_adj. 1693 1694The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last 1695value set by a CAP_SYS_RESOURCE process. To reduce the value any lower 1696requires CAP_SYS_RESOURCE. 1697 1698 16993.2 /proc/<pid>/oom_score - Display current oom-killer score 1700------------------------------------------------------------- 1701 1702This file can be used to check the current score used by the oom-killer for 1703any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which 1704process should be killed in an out-of-memory situation. 1705 1706Please note that the exported value includes oom_score_adj so it is 1707effectively in range [0,2000]. 1708 1709 17103.3 /proc/<pid>/io - Display the IO accounting fields 1711------------------------------------------------------- 1712 1713This file contains IO statistics for each running process. 1714 1715Example 1716~~~~~~~ 1717 1718:: 1719 1720 test:/tmp # dd if=/dev/zero of=/tmp/test.dat & 1721 [1] 3828 1722 1723 test:/tmp # cat /proc/3828/io 1724 rchar: 323934931 1725 wchar: 323929600 1726 syscr: 632687 1727 syscw: 632675 1728 read_bytes: 0 1729 write_bytes: 323932160 1730 cancelled_write_bytes: 0 1731 1732 1733Description 1734~~~~~~~~~~~ 1735 1736rchar 1737^^^^^ 1738 1739I/O counter: chars read 1740The number of bytes which this task has caused to be read from storage. This 1741is simply the sum of bytes which this process passed to read() and pread(). 1742It includes things like tty IO and it is unaffected by whether or not actual 1743physical disk IO was required (the read might have been satisfied from 1744pagecache). 1745 1746 1747wchar 1748^^^^^ 1749 1750I/O counter: chars written 1751The number of bytes which this task has caused, or shall cause to be written 1752to disk. Similar caveats apply here as with rchar. 1753 1754 1755syscr 1756^^^^^ 1757 1758I/O counter: read syscalls 1759Attempt to count the number of read I/O operations, i.e. syscalls like read() 1760and pread(). 1761 1762 1763syscw 1764^^^^^ 1765 1766I/O counter: write syscalls 1767Attempt to count the number of write I/O operations, i.e. syscalls like 1768write() and pwrite(). 1769 1770 1771read_bytes 1772^^^^^^^^^^ 1773 1774I/O counter: bytes read 1775Attempt to count the number of bytes which this process really did cause to 1776be fetched from the storage layer. Done at the submit_bio() level, so it is 1777accurate for block-backed filesystems. <please add status regarding NFS and 1778CIFS at a later time> 1779 1780 1781write_bytes 1782^^^^^^^^^^^ 1783 1784I/O counter: bytes written 1785Attempt to count the number of bytes which this process caused to be sent to 1786the storage layer. This is done at page-dirtying time. 1787 1788 1789cancelled_write_bytes 1790^^^^^^^^^^^^^^^^^^^^^ 1791 1792The big inaccuracy here is truncate. If a process writes 1MB to a file and 1793then deletes the file, it will in fact perform no writeout. But it will have 1794been accounted as having caused 1MB of write. 1795In other words: The number of bytes which this process caused to not happen, 1796by truncating pagecache. A task can cause "negative" IO too. If this task 1797truncates some dirty pagecache, some IO which another task has been accounted 1798for (in its write_bytes) will not be happening. We _could_ just subtract that 1799from the truncating task's write_bytes, but there is information loss in doing 1800that. 1801 1802 1803.. Note:: 1804 1805 At its current implementation state, this is a bit racy on 32-bit machines: 1806 if process A reads process B's /proc/pid/io while process B is updating one 1807 of those 64-bit counters, process A could see an intermediate result. 1808 1809 1810More information about this can be found within the taskstats documentation in 1811Documentation/accounting. 1812 18133.4 /proc/<pid>/coredump_filter - Core dump filtering settings 1814--------------------------------------------------------------- 1815When a process is dumped, all anonymous memory is written to a core file as 1816long as the size of the core file isn't limited. But sometimes we don't want 1817to dump some memory segments, for example, huge shared memory or DAX. 1818Conversely, sometimes we want to save file-backed memory segments into a core 1819file, not only the individual files. 1820 1821/proc/<pid>/coredump_filter allows you to customize which memory segments 1822will be dumped when the <pid> process is dumped. coredump_filter is a bitmask 1823of memory types. If a bit of the bitmask is set, memory segments of the 1824corresponding memory type are dumped, otherwise they are not dumped. 1825 1826The following 9 memory types are supported: 1827 1828 - (bit 0) anonymous private memory 1829 - (bit 1) anonymous shared memory 1830 - (bit 2) file-backed private memory 1831 - (bit 3) file-backed shared memory 1832 - (bit 4) ELF header pages in file-backed private memory areas (it is 1833 effective only if the bit 2 is cleared) 1834 - (bit 5) hugetlb private memory 1835 - (bit 6) hugetlb shared memory 1836 - (bit 7) DAX private memory 1837 - (bit 8) DAX shared memory 1838 1839 Note that MMIO pages such as frame buffer are never dumped and vDSO pages 1840 are always dumped regardless of the bitmask status. 1841 1842 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is 1843 only affected by bit 5-6, and DAX is only affected by bits 7-8. 1844 1845The default value of coredump_filter is 0x33; this means all anonymous memory 1846segments, ELF header pages and hugetlb private memory are dumped. 1847 1848If you don't want to dump all shared memory segments attached to pid 1234, 1849write 0x31 to the process's proc file:: 1850 1851 $ echo 0x31 > /proc/1234/coredump_filter 1852 1853When a new process is created, the process inherits the bitmask status from its 1854parent. It is useful to set up coredump_filter before the program runs. 1855For example:: 1856 1857 $ echo 0x7 > /proc/self/coredump_filter 1858 $ ./some_program 1859 18603.5 /proc/<pid>/mountinfo - Information about mounts 1861-------------------------------------------------------- 1862 1863This file contains lines of the form:: 1864 1865 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue 1866 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11) 1867 1868 (1) mount ID: unique identifier of the mount (may be reused after umount) 1869 (2) parent ID: ID of parent (or of self for the top of the mount tree) 1870 (3) major:minor: value of st_dev for files on filesystem 1871 (4) root: root of the mount within the filesystem 1872 (5) mount point: mount point relative to the process's root 1873 (6) mount options: per mount options 1874 (7) optional fields: zero or more fields of the form "tag[:value]" 1875 (8) separator: marks the end of the optional fields 1876 (9) filesystem type: name of filesystem of the form "type[.subtype]" 1877 (10) mount source: filesystem specific information or "none" 1878 (11) super options: per super block options 1879 1880Parsers should ignore all unrecognised optional fields. Currently the 1881possible optional fields are: 1882 1883================ ============================================================== 1884shared:X mount is shared in peer group X 1885master:X mount is slave to peer group X 1886propagate_from:X mount is slave and receives propagation from peer group X [#]_ 1887unbindable mount is unbindable 1888================ ============================================================== 1889 1890.. [#] X is the closest dominant peer group under the process's root. If 1891 X is the immediate master of the mount, or if there's no dominant peer 1892 group under the same root, then only the "master:X" field is present 1893 and not the "propagate_from:X" field. 1894 1895For more information on mount propagation see: 1896 1897 Documentation/filesystems/sharedsubtree.rst 1898 1899 19003.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 1901-------------------------------------------------------- 1902These files provide a method to access a task's comm value. It also allows for 1903a task to set its own or one of its thread siblings comm value. The comm value 1904is limited in size compared to the cmdline value, so writing anything longer 1905then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated 1906comm value. 1907 1908 19093.7 /proc/<pid>/task/<tid>/children - Information about task children 1910------------------------------------------------------------------------- 1911This file provides a fast way to retrieve first level children pids 1912of a task pointed by <pid>/<tid> pair. The format is a space separated 1913stream of pids. 1914 1915Note the "first level" here -- if a child has its own children they will 1916not be listed here; one needs to read /proc/<children-pid>/task/<tid>/children 1917to obtain the descendants. 1918 1919Since this interface is intended to be fast and cheap it doesn't 1920guarantee to provide precise results and some children might be 1921skipped, especially if they've exited right after we printed their 1922pids, so one needs to either stop or freeze processes being inspected 1923if precise results are needed. 1924 1925 19263.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 1927--------------------------------------------------------------- 1928This file provides information associated with an opened file. The regular 1929files have at least four fields -- 'pos', 'flags', 'mnt_id' and 'ino'. 1930The 'pos' represents the current offset of the opened file in decimal 1931form [see lseek(2) for details], 'flags' denotes the octal O_xxx mask the 1932file has been created with [see open(2) for details] and 'mnt_id' represents 1933mount ID of the file system containing the opened file [see 3.5 1934/proc/<pid>/mountinfo for details]. 'ino' represents the inode number of 1935the file. 1936 1937A typical output is:: 1938 1939 pos: 0 1940 flags: 0100002 1941 mnt_id: 19 1942 ino: 63107 1943 1944All locks associated with a file descriptor are shown in its fdinfo too:: 1945 1946 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF 1947 1948The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags 1949pair provide additional information particular to the objects they represent. 1950 1951Eventfd files 1952~~~~~~~~~~~~~ 1953 1954:: 1955 1956 pos: 0 1957 flags: 04002 1958 mnt_id: 9 1959 ino: 63107 1960 eventfd-count: 5a 1961 1962where 'eventfd-count' is hex value of a counter. 1963 1964Signalfd files 1965~~~~~~~~~~~~~~ 1966 1967:: 1968 1969 pos: 0 1970 flags: 04002 1971 mnt_id: 9 1972 ino: 63107 1973 sigmask: 0000000000000200 1974 1975where 'sigmask' is hex value of the signal mask associated 1976with a file. 1977 1978Epoll files 1979~~~~~~~~~~~ 1980 1981:: 1982 1983 pos: 0 1984 flags: 02 1985 mnt_id: 9 1986 ino: 63107 1987 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7 1988 1989where 'tfd' is a target file descriptor number in decimal form, 1990'events' is events mask being watched and the 'data' is data 1991associated with a target [see epoll(7) for more details]. 1992 1993The 'pos' is current offset of the target file in decimal form 1994[see lseek(2)], 'ino' and 'sdev' are inode and device numbers 1995where target file resides, all in hex format. 1996 1997Fsnotify files 1998~~~~~~~~~~~~~~ 1999For inotify files the format is the following:: 2000 2001 pos: 0 2002 flags: 02000000 2003 mnt_id: 9 2004 ino: 63107 2005 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d 2006 2007where 'wd' is a watch descriptor in decimal form, i.e. a target file 2008descriptor number, 'ino' and 'sdev' are inode and device where the 2009target file resides and the 'mask' is the mask of events, all in hex 2010form [see inotify(7) for more details]. 2011 2012If the kernel was built with exportfs support, the path to the target 2013file is encoded as a file handle. The file handle is provided by three 2014fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex 2015format. 2016 2017If the kernel is built without exportfs support the file handle won't be 2018printed out. 2019 2020If there is no inotify mark attached yet the 'inotify' line will be omitted. 2021 2022For fanotify files the format is:: 2023 2024 pos: 0 2025 flags: 02 2026 mnt_id: 9 2027 ino: 63107 2028 fanotify flags:10 event-flags:0 2029 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003 2030 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4 2031 2032where fanotify 'flags' and 'event-flags' are values used in fanotify_init 2033call, 'mnt_id' is the mount point identifier, 'mflags' is the value of 2034flags associated with mark which are tracked separately from events 2035mask. 'ino' and 'sdev' are target inode and device, 'mask' is the events 2036mask and 'ignored_mask' is the mask of events which are to be ignored. 2037All are in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask' 2038provide information about flags and mask used in fanotify_mark 2039call [see fsnotify manpage for details]. 2040 2041While the first three lines are mandatory and always printed, the rest is 2042optional and may be omitted if no marks created yet. 2043 2044Timerfd files 2045~~~~~~~~~~~~~ 2046 2047:: 2048 2049 pos: 0 2050 flags: 02 2051 mnt_id: 9 2052 ino: 63107 2053 clockid: 0 2054 ticks: 0 2055 settime flags: 01 2056 it_value: (0, 49406829) 2057 it_interval: (1, 0) 2058 2059where 'clockid' is the clock type and 'ticks' is the number of the timer expirations 2060that have occurred [see timerfd_create(2) for details]. 'settime flags' are 2061flags in octal form been used to setup the timer [see timerfd_settime(2) for 2062details]. 'it_value' is remaining time until the timer expiration. 2063'it_interval' is the interval for the timer. Note the timer might be set up 2064with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value' 2065still exhibits timer's remaining time. 2066 2067DMA Buffer files 2068~~~~~~~~~~~~~~~~ 2069 2070:: 2071 2072 pos: 0 2073 flags: 04002 2074 mnt_id: 9 2075 ino: 63107 2076 size: 32768 2077 count: 2 2078 exp_name: system-heap 2079 2080where 'size' is the size of the DMA buffer in bytes. 'count' is the file count of 2081the DMA buffer file. 'exp_name' is the name of the DMA buffer exporter. 2082 20833.9 /proc/<pid>/map_files - Information about memory mapped files 2084--------------------------------------------------------------------- 2085This directory contains symbolic links which represent memory mapped files 2086the process is maintaining. Example output:: 2087 2088 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so 2089 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so 2090 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so 2091 | ... 2092 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1 2093 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls 2094 2095The name of a link represents the virtual memory bounds of a mapping, i.e. 2096vm_area_struct::vm_start-vm_area_struct::vm_end. 2097 2098The main purpose of the map_files is to retrieve a set of memory mapped 2099files in a fast way instead of parsing /proc/<pid>/maps or 2100/proc/<pid>/smaps, both of which contain many more records. At the same 2101time one can open(2) mappings from the listings of two processes and 2102comparing their inode numbers to figure out which anonymous memory areas 2103are actually shared. 2104 21053.10 /proc/<pid>/timerslack_ns - Task timerslack value 2106--------------------------------------------------------- 2107This file provides the value of the task's timerslack value in nanoseconds. 2108This value specifies an amount of time that normal timers may be deferred 2109in order to coalesce timers and avoid unnecessary wakeups. 2110 2111This allows a task's interactivity vs power consumption tradeoff to be 2112adjusted. 2113 2114Writing 0 to the file will set the task's timerslack to the default value. 2115 2116Valid values are from 0 - ULLONG_MAX 2117 2118An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level 2119permissions on the task specified to change its timerslack_ns value. 2120 21213.11 /proc/<pid>/patch_state - Livepatch patch operation state 2122----------------------------------------------------------------- 2123When CONFIG_LIVEPATCH is enabled, this file displays the value of the 2124patch state for the task. 2125 2126A value of '-1' indicates that no patch is in transition. 2127 2128A value of '0' indicates that a patch is in transition and the task is 2129unpatched. If the patch is being enabled, then the task hasn't been 2130patched yet. If the patch is being disabled, then the task has already 2131been unpatched. 2132 2133A value of '1' indicates that a patch is in transition and the task is 2134patched. If the patch is being enabled, then the task has already been 2135patched. If the patch is being disabled, then the task hasn't been 2136unpatched yet. 2137 21383.12 /proc/<pid>/arch_status - task architecture specific status 2139------------------------------------------------------------------- 2140When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the 2141architecture specific status of the task. 2142 2143Example 2144~~~~~~~ 2145 2146:: 2147 2148 $ cat /proc/6753/arch_status 2149 AVX512_elapsed_ms: 8 2150 2151Description 2152~~~~~~~~~~~ 2153 2154x86 specific entries 2155~~~~~~~~~~~~~~~~~~~~~ 2156 2157AVX512_elapsed_ms 2158^^^^^^^^^^^^^^^^^^ 2159 2160 If AVX512 is supported on the machine, this entry shows the milliseconds 2161 elapsed since the last time AVX512 usage was recorded. The recording 2162 happens on a best effort basis when a task is scheduled out. This means 2163 that the value depends on two factors: 2164 2165 1) The time which the task spent on the CPU without being scheduled 2166 out. With CPU isolation and a single runnable task this can take 2167 several seconds. 2168 2169 2) The time since the task was scheduled out last. Depending on the 2170 reason for being scheduled out (time slice exhausted, syscall ...) 2171 this can be arbitrary long time. 2172 2173 As a consequence the value cannot be considered precise and authoritative 2174 information. The application which uses this information has to be aware 2175 of the overall scenario on the system in order to determine whether a 2176 task is a real AVX512 user or not. Precise information can be obtained 2177 with performance counters. 2178 2179 A special value of '-1' indicates that no AVX512 usage was recorded, thus 2180 the task is unlikely an AVX512 user, but depends on the workload and the 2181 scheduling scenario, it also could be a false negative mentioned above. 2182 2183Chapter 4: Configuring procfs 2184============================= 2185 21864.1 Mount options 2187--------------------- 2188 2189The following mount options are supported: 2190 2191 ========= ======================================================== 2192 hidepid= Set /proc/<pid>/ access mode. 2193 gid= Set the group authorized to learn processes information. 2194 subset= Show only the specified subset of procfs. 2195 ========= ======================================================== 2196 2197hidepid=off or hidepid=0 means classic mode - everybody may access all 2198/proc/<pid>/ directories (default). 2199 2200hidepid=noaccess or hidepid=1 means users may not access any /proc/<pid>/ 2201directories but their own. Sensitive files like cmdline, sched*, status are now 2202protected against other users. This makes it impossible to learn whether any 2203user runs specific program (given the program doesn't reveal itself by its 2204behaviour). As an additional bonus, as /proc/<pid>/cmdline is unaccessible for 2205other users, poorly written programs passing sensitive information via program 2206arguments are now protected against local eavesdroppers. 2207 2208hidepid=invisible or hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be 2209fully invisible to other users. It doesn't mean that it hides a fact whether a 2210process with a specific pid value exists (it can be learned by other means, e.g. 2211by "kill -0 $PID"), but it hides process' uid and gid, which may be learned by 2212stat()'ing /proc/<pid>/ otherwise. It greatly complicates an intruder's task of 2213gathering information about running processes, whether some daemon runs with 2214elevated privileges, whether other user runs some sensitive program, whether 2215other users run any program at all, etc. 2216 2217hidepid=ptraceable or hidepid=4 means that procfs should only contain 2218/proc/<pid>/ directories that the caller can ptrace. 2219 2220gid= defines a group authorized to learn processes information otherwise 2221prohibited by hidepid=. If you use some daemon like identd which needs to learn 2222information about processes information, just add identd to this group. 2223 2224subset=pid hides all top level files and directories in the procfs that 2225are not related to tasks. 2226 2227Chapter 5: Filesystem behavior 2228============================== 2229 2230Originally, before the advent of pid namepsace, procfs was a global file 2231system. It means that there was only one procfs instance in the system. 2232 2233When pid namespace was added, a separate procfs instance was mounted in 2234each pid namespace. So, procfs mount options are global among all 2235mountpoints within the same namespace:: 2236 2237 # grep ^proc /proc/mounts 2238 proc /proc proc rw,relatime,hidepid=2 0 0 2239 2240 # strace -e mount mount -o hidepid=1 -t proc proc /tmp/proc 2241 mount("proc", "/tmp/proc", "proc", 0, "hidepid=1") = 0 2242 +++ exited with 0 +++ 2243 2244 # grep ^proc /proc/mounts 2245 proc /proc proc rw,relatime,hidepid=2 0 0 2246 proc /tmp/proc proc rw,relatime,hidepid=2 0 0 2247 2248and only after remounting procfs mount options will change at all 2249mountpoints:: 2250 2251 # mount -o remount,hidepid=1 -t proc proc /tmp/proc 2252 2253 # grep ^proc /proc/mounts 2254 proc /proc proc rw,relatime,hidepid=1 0 0 2255 proc /tmp/proc proc rw,relatime,hidepid=1 0 0 2256 2257This behavior is different from the behavior of other filesystems. 2258 2259The new procfs behavior is more like other filesystems. Each procfs mount 2260creates a new procfs instance. Mount options affect own procfs instance. 2261It means that it became possible to have several procfs instances 2262displaying tasks with different filtering options in one pid namespace:: 2263 2264 # mount -o hidepid=invisible -t proc proc /proc 2265 # mount -o hidepid=noaccess -t proc proc /tmp/proc 2266 # grep ^proc /proc/mounts 2267 proc /proc proc rw,relatime,hidepid=invisible 0 0 2268 proc /tmp/proc proc rw,relatime,hidepid=noaccess 0 0 2269