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 ======= ==================================== 435 436 or if empty, the mapping is anonymous. 437 438The /proc/PID/smaps is an extension based on maps, showing the memory 439consumption for each of the process's mappings. For each mapping (aka Virtual 440Memory Area, or VMA) there is a series of lines such as the following:: 441 442 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash 443 444 Size: 1084 kB 445 KernelPageSize: 4 kB 446 MMUPageSize: 4 kB 447 Rss: 892 kB 448 Pss: 374 kB 449 Shared_Clean: 892 kB 450 Shared_Dirty: 0 kB 451 Private_Clean: 0 kB 452 Private_Dirty: 0 kB 453 Referenced: 892 kB 454 Anonymous: 0 kB 455 LazyFree: 0 kB 456 AnonHugePages: 0 kB 457 ShmemPmdMapped: 0 kB 458 Shared_Hugetlb: 0 kB 459 Private_Hugetlb: 0 kB 460 Swap: 0 kB 461 SwapPss: 0 kB 462 KernelPageSize: 4 kB 463 MMUPageSize: 4 kB 464 Locked: 0 kB 465 THPeligible: 0 466 VmFlags: rd ex mr mw me dw 467 468The first of these lines shows the same information as is displayed for the 469mapping in /proc/PID/maps. Following lines show the size of the mapping 470(size); the size of each page allocated when backing a VMA (KernelPageSize), 471which is usually the same as the size in the page table entries; the page size 472used by the MMU when backing a VMA (in most cases, the same as KernelPageSize); 473the amount of the mapping that is currently resident in RAM (RSS); the 474process' proportional share of this mapping (PSS); and the number of clean and 475dirty shared and private pages in the mapping. 476 477The "proportional set size" (PSS) of a process is the count of pages it has 478in memory, where each page is divided by the number of processes sharing it. 479So if a process has 1000 pages all to itself, and 1000 shared with one other 480process, its PSS will be 1500. 481 482Note that even a page which is part of a MAP_SHARED mapping, but has only 483a single pte mapped, i.e. is currently used by only one process, is accounted 484as private and not as shared. 485 486"Referenced" indicates the amount of memory currently marked as referenced or 487accessed. 488 489"Anonymous" shows the amount of memory that does not belong to any file. Even 490a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE 491and a page is modified, the file page is replaced by a private anonymous copy. 492 493"LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE). 494The memory isn't freed immediately with madvise(). It's freed in memory 495pressure if the memory is clean. Please note that the printed value might 496be lower than the real value due to optimizations used in the current 497implementation. If this is not desirable please file a bug report. 498 499"AnonHugePages" shows the ammount of memory backed by transparent hugepage. 500 501"ShmemPmdMapped" shows the ammount of shared (shmem/tmpfs) memory backed by 502huge pages. 503 504"Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by 505hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical 506reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field. 507 508"Swap" shows how much would-be-anonymous memory is also used, but out on swap. 509 510For shmem mappings, "Swap" includes also the size of the mapped (and not 511replaced by copy-on-write) part of the underlying shmem object out on swap. 512"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this 513does not take into account swapped out page of underlying shmem objects. 514"Locked" indicates whether the mapping is locked in memory or not. 515"THPeligible" indicates whether the mapping is eligible for allocating THP 516pages - 1 if true, 0 otherwise. It just shows the current status. 517 518"VmFlags" field deserves a separate description. This member represents the 519kernel flags associated with the particular virtual memory area in two letter 520encoded manner. The codes are the following: 521 522 == ======================================= 523 rd readable 524 wr writeable 525 ex executable 526 sh shared 527 mr may read 528 mw may write 529 me may execute 530 ms may share 531 gd stack segment growns down 532 pf pure PFN range 533 dw disabled write to the mapped file 534 lo pages are locked in memory 535 io memory mapped I/O area 536 sr sequential read advise provided 537 rr random read advise provided 538 dc do not copy area on fork 539 de do not expand area on remapping 540 ac area is accountable 541 nr swap space is not reserved for the area 542 ht area uses huge tlb pages 543 ar architecture specific flag 544 dd do not include area into core dump 545 sd soft dirty flag 546 mm mixed map area 547 hg huge page advise flag 548 nh no huge page advise flag 549 mg mergable advise flag 550 bt arm64 BTI guarded page 551 mt arm64 MTE allocation tags are enabled 552 == ======================================= 553 554Note that there is no guarantee that every flag and associated mnemonic will 555be present in all further kernel releases. Things get changed, the flags may 556be vanished or the reverse -- new added. Interpretation of their meaning 557might change in future as well. So each consumer of these flags has to 558follow each specific kernel version for the exact semantic. 559 560This file is only present if the CONFIG_MMU kernel configuration option is 561enabled. 562 563Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent 564output can be achieved only in the single read call). 565 566This typically manifests when doing partial reads of these files while the 567memory map is being modified. Despite the races, we do provide the following 568guarantees: 569 5701) The mapped addresses never go backwards, which implies no two 571 regions will ever overlap. 5722) If there is something at a given vaddr during the entirety of the 573 life of the smaps/maps walk, there will be some output for it. 574 575The /proc/PID/smaps_rollup file includes the same fields as /proc/PID/smaps, 576but their values are the sums of the corresponding values for all mappings of 577the process. Additionally, it contains these fields: 578 579- Pss_Anon 580- Pss_File 581- Pss_Shmem 582 583They represent the proportional shares of anonymous, file, and shmem pages, as 584described for smaps above. These fields are omitted in smaps since each 585mapping identifies the type (anon, file, or shmem) of all pages it contains. 586Thus all information in smaps_rollup can be derived from smaps, but at a 587significantly higher cost. 588 589The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG 590bits on both physical and virtual pages associated with a process, and the 591soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst 592for details). 593To clear the bits for all the pages associated with the process:: 594 595 > echo 1 > /proc/PID/clear_refs 596 597To clear the bits for the anonymous pages associated with the process:: 598 599 > echo 2 > /proc/PID/clear_refs 600 601To clear the bits for the file mapped pages associated with the process:: 602 603 > echo 3 > /proc/PID/clear_refs 604 605To clear the soft-dirty bit:: 606 607 > echo 4 > /proc/PID/clear_refs 608 609To reset the peak resident set size ("high water mark") to the process's 610current value:: 611 612 > echo 5 > /proc/PID/clear_refs 613 614Any other value written to /proc/PID/clear_refs will have no effect. 615 616The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags 617using /proc/kpageflags and number of times a page is mapped using 618/proc/kpagecount. For detailed explanation, see 619Documentation/admin-guide/mm/pagemap.rst. 620 621The /proc/pid/numa_maps is an extension based on maps, showing the memory 622locality and binding policy, as well as the memory usage (in pages) of 623each mapping. The output follows a general format where mapping details get 624summarized separated by blank spaces, one mapping per each file line:: 625 626 address policy mapping details 627 628 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4 629 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4 630 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4 631 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 632 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 633 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4 634 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4 635 320698b000 default file=/lib64/libc-2.12.so 636 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4 637 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 638 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4 639 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4 640 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4 641 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048 642 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4 643 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4 644 645Where: 646 647"address" is the starting address for the mapping; 648 649"policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst); 650 651"mapping details" summarizes mapping data such as mapping type, page usage counters, 652node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page 653size, in KB, that is backing the mapping up. 654 6551.2 Kernel data 656--------------- 657 658Similar to the process entries, the kernel data files give information about 659the running kernel. The files used to obtain this information are contained in 660/proc and are listed in Table 1-5. Not all of these will be present in your 661system. It depends on the kernel configuration and the loaded modules, which 662files are there, and which are missing. 663 664.. table:: Table 1-5: Kernel info in /proc 665 666 ============ =============================================================== 667 File Content 668 ============ =============================================================== 669 apm Advanced power management info 670 buddyinfo Kernel memory allocator information (see text) (2.5) 671 bus Directory containing bus specific information 672 cmdline Kernel command line 673 cpuinfo Info about the CPU 674 devices Available devices (block and character) 675 dma Used DMS channels 676 filesystems Supported filesystems 677 driver Various drivers grouped here, currently rtc (2.4) 678 execdomains Execdomains, related to security (2.4) 679 fb Frame Buffer devices (2.4) 680 fs File system parameters, currently nfs/exports (2.4) 681 ide Directory containing info about the IDE subsystem 682 interrupts Interrupt usage 683 iomem Memory map (2.4) 684 ioports I/O port usage 685 irq Masks for irq to cpu affinity (2.4)(smp?) 686 isapnp ISA PnP (Plug&Play) Info (2.4) 687 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4)) 688 kmsg Kernel messages 689 ksyms Kernel symbol table 690 loadavg Load average of last 1, 5 & 15 minutes; 691 number of processes currently runnable (running or on ready queue); 692 total number of processes in system; 693 last pid created. 694 All fields are separated by one space except "number of 695 processes currently runnable" and "total number of processes 696 in system", which are separated by a slash ('/'). Example: 697 0.61 0.61 0.55 3/828 22084 698 locks Kernel locks 699 meminfo Memory info 700 misc Miscellaneous 701 modules List of loaded modules 702 mounts Mounted filesystems 703 net Networking info (see text) 704 pagetypeinfo Additional page allocator information (see text) (2.5) 705 partitions Table of partitions known to the system 706 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/, 707 decoupled by lspci (2.4) 708 rtc Real time clock 709 scsi SCSI info (see text) 710 slabinfo Slab pool info 711 softirqs softirq usage 712 stat Overall statistics 713 swaps Swap space utilization 714 sys See chapter 2 715 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4) 716 tty Info of tty drivers 717 uptime Wall clock since boot, combined idle time of all cpus 718 version Kernel version 719 video bttv info of video resources (2.4) 720 vmallocinfo Show vmalloced areas 721 ============ =============================================================== 722 723You can, for example, check which interrupts are currently in use and what 724they are used for by looking in the file /proc/interrupts:: 725 726 > cat /proc/interrupts 727 CPU0 728 0: 8728810 XT-PIC timer 729 1: 895 XT-PIC keyboard 730 2: 0 XT-PIC cascade 731 3: 531695 XT-PIC aha152x 732 4: 2014133 XT-PIC serial 733 5: 44401 XT-PIC pcnet_cs 734 8: 2 XT-PIC rtc 735 11: 8 XT-PIC i82365 736 12: 182918 XT-PIC PS/2 Mouse 737 13: 1 XT-PIC fpu 738 14: 1232265 XT-PIC ide0 739 15: 7 XT-PIC ide1 740 NMI: 0 741 742In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the 743output of a SMP machine):: 744 745 > cat /proc/interrupts 746 747 CPU0 CPU1 748 0: 1243498 1214548 IO-APIC-edge timer 749 1: 8949 8958 IO-APIC-edge keyboard 750 2: 0 0 XT-PIC cascade 751 5: 11286 10161 IO-APIC-edge soundblaster 752 8: 1 0 IO-APIC-edge rtc 753 9: 27422 27407 IO-APIC-edge 3c503 754 12: 113645 113873 IO-APIC-edge PS/2 Mouse 755 13: 0 0 XT-PIC fpu 756 14: 22491 24012 IO-APIC-edge ide0 757 15: 2183 2415 IO-APIC-edge ide1 758 17: 30564 30414 IO-APIC-level eth0 759 18: 177 164 IO-APIC-level bttv 760 NMI: 2457961 2457959 761 LOC: 2457882 2457881 762 ERR: 2155 763 764NMI is incremented in this case because every timer interrupt generates a NMI 765(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups. 766 767LOC is the local interrupt counter of the internal APIC of every CPU. 768 769ERR is incremented in the case of errors in the IO-APIC bus (the bus that 770connects the CPUs in a SMP system. This means that an error has been detected, 771the IO-APIC automatically retry the transmission, so it should not be a big 772problem, but you should read the SMP-FAQ. 773 774In 2.6.2* /proc/interrupts was expanded again. This time the goal was for 775/proc/interrupts to display every IRQ vector in use by the system, not 776just those considered 'most important'. The new vectors are: 777 778THR 779 interrupt raised when a machine check threshold counter 780 (typically counting ECC corrected errors of memory or cache) exceeds 781 a configurable threshold. Only available on some systems. 782 783TRM 784 a thermal event interrupt occurs when a temperature threshold 785 has been exceeded for the CPU. This interrupt may also be generated 786 when the temperature drops back to normal. 787 788SPU 789 a spurious interrupt is some interrupt that was raised then lowered 790 by some IO device before it could be fully processed by the APIC. Hence 791 the APIC sees the interrupt but does not know what device it came from. 792 For this case the APIC will generate the interrupt with a IRQ vector 793 of 0xff. This might also be generated by chipset bugs. 794 795RES, CAL, TLB 796 rescheduling, call and TLB flush interrupts are 797 sent from one CPU to another per the needs of the OS. Typically, 798 their statistics are used by kernel developers and interested users to 799 determine the occurrence of interrupts of the given type. 800 801The above IRQ vectors are displayed only when relevant. For example, 802the threshold vector does not exist on x86_64 platforms. Others are 803suppressed when the system is a uniprocessor. As of this writing, only 804i386 and x86_64 platforms support the new IRQ vector displays. 805 806Of some interest is the introduction of the /proc/irq directory to 2.4. 807It could be used to set IRQ to CPU affinity. This means that you can "hook" an 808IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the 809irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and 810prof_cpu_mask. 811 812For example:: 813 814 > ls /proc/irq/ 815 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask 816 1 11 13 15 17 19 3 5 7 9 default_smp_affinity 817 > ls /proc/irq/0/ 818 smp_affinity 819 820smp_affinity is a bitmask, in which you can specify which CPUs can handle the 821IRQ. You can set it by doing:: 822 823 > echo 1 > /proc/irq/10/smp_affinity 824 825This means that only the first CPU will handle the IRQ, but you can also echo 8265 which means that only the first and third CPU can handle the IRQ. 827 828The contents of each smp_affinity file is the same by default:: 829 830 > cat /proc/irq/0/smp_affinity 831 ffffffff 832 833There is an alternate interface, smp_affinity_list which allows specifying 834a CPU range instead of a bitmask:: 835 836 > cat /proc/irq/0/smp_affinity_list 837 1024-1031 838 839The default_smp_affinity mask applies to all non-active IRQs, which are the 840IRQs which have not yet been allocated/activated, and hence which lack a 841/proc/irq/[0-9]* directory. 842 843The node file on an SMP system shows the node to which the device using the IRQ 844reports itself as being attached. This hardware locality information does not 845include information about any possible driver locality preference. 846 847prof_cpu_mask specifies which CPUs are to be profiled by the system wide 848profiler. Default value is ffffffff (all CPUs if there are only 32 of them). 849 850The way IRQs are routed is handled by the IO-APIC, and it's Round Robin 851between all the CPUs which are allowed to handle it. As usual the kernel has 852more info than you and does a better job than you, so the defaults are the 853best choice for almost everyone. [Note this applies only to those IO-APIC's 854that support "Round Robin" interrupt distribution.] 855 856There are three more important subdirectories in /proc: net, scsi, and sys. 857The general rule is that the contents, or even the existence of these 858directories, depend on your kernel configuration. If SCSI is not enabled, the 859directory scsi may not exist. The same is true with the net, which is there 860only when networking support is present in the running kernel. 861 862The slabinfo file gives information about memory usage at the slab level. 863Linux uses slab pools for memory management above page level in version 2.2. 864Commonly used objects have their own slab pool (such as network buffers, 865directory cache, and so on). 866 867:: 868 869 > cat /proc/buddyinfo 870 871 Node 0, zone DMA 0 4 5 4 4 3 ... 872 Node 0, zone Normal 1 0 0 1 101 8 ... 873 Node 0, zone HighMem 2 0 0 1 1 0 ... 874 875External fragmentation is a problem under some workloads, and buddyinfo is a 876useful tool for helping diagnose these problems. Buddyinfo will give you a 877clue as to how big an area you can safely allocate, or why a previous 878allocation failed. 879 880Each column represents the number of pages of a certain order which are 881available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in 882ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE 883available in ZONE_NORMAL, etc... 884 885More information relevant to external fragmentation can be found in 886pagetypeinfo:: 887 888 > cat /proc/pagetypeinfo 889 Page block order: 9 890 Pages per block: 512 891 892 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10 893 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0 894 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0 895 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2 896 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0 897 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0 898 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9 899 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0 900 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452 901 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0 902 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0 903 904 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate 905 Node 0, zone DMA 2 0 5 1 0 906 Node 0, zone DMA32 41 6 967 2 0 907 908Fragmentation avoidance in the kernel works by grouping pages of different 909migrate types into the same contiguous regions of memory called page blocks. 910A page block is typically the size of the default hugepage size, e.g. 2MB on 911X86-64. By keeping pages grouped based on their ability to move, the kernel 912can reclaim pages within a page block to satisfy a high-order allocation. 913 914The pagetypinfo begins with information on the size of a page block. It 915then gives the same type of information as buddyinfo except broken down 916by migrate-type and finishes with details on how many page blocks of each 917type exist. 918 919If min_free_kbytes has been tuned correctly (recommendations made by hugeadm 920from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can 921make an estimate of the likely number of huge pages that can be allocated 922at a given point in time. All the "Movable" blocks should be allocatable 923unless memory has been mlock()'d. Some of the Reclaimable blocks should 924also be allocatable although a lot of filesystem metadata may have to be 925reclaimed to achieve this. 926 927 928meminfo 929~~~~~~~ 930 931Provides information about distribution and utilization of memory. This 932varies by architecture and compile options. The following is from a 93316GB PIII, which has highmem enabled. You may not have all of these fields. 934 935:: 936 937 > cat /proc/meminfo 938 939 MemTotal: 16344972 kB 940 MemFree: 13634064 kB 941 MemAvailable: 14836172 kB 942 Buffers: 3656 kB 943 Cached: 1195708 kB 944 SwapCached: 0 kB 945 Active: 891636 kB 946 Inactive: 1077224 kB 947 HighTotal: 15597528 kB 948 HighFree: 13629632 kB 949 LowTotal: 747444 kB 950 LowFree: 4432 kB 951 SwapTotal: 0 kB 952 SwapFree: 0 kB 953 Dirty: 968 kB 954 Writeback: 0 kB 955 AnonPages: 861800 kB 956 Mapped: 280372 kB 957 Shmem: 644 kB 958 KReclaimable: 168048 kB 959 Slab: 284364 kB 960 SReclaimable: 159856 kB 961 SUnreclaim: 124508 kB 962 PageTables: 24448 kB 963 NFS_Unstable: 0 kB 964 Bounce: 0 kB 965 WritebackTmp: 0 kB 966 CommitLimit: 7669796 kB 967 Committed_AS: 100056 kB 968 VmallocTotal: 112216 kB 969 VmallocUsed: 428 kB 970 VmallocChunk: 111088 kB 971 Percpu: 62080 kB 972 HardwareCorrupted: 0 kB 973 AnonHugePages: 49152 kB 974 ShmemHugePages: 0 kB 975 ShmemPmdMapped: 0 kB 976 977MemTotal 978 Total usable RAM (i.e. physical RAM minus a few reserved 979 bits and the kernel binary code) 980MemFree 981 The sum of LowFree+HighFree 982MemAvailable 983 An estimate of how much memory is available for starting new 984 applications, without swapping. Calculated from MemFree, 985 SReclaimable, the size of the file LRU lists, and the low 986 watermarks in each zone. 987 The estimate takes into account that the system needs some 988 page cache to function well, and that not all reclaimable 989 slab will be reclaimable, due to items being in use. The 990 impact of those factors will vary from system to system. 991Buffers 992 Relatively temporary storage for raw disk blocks 993 shouldn't get tremendously large (20MB or so) 994Cached 995 in-memory cache for files read from the disk (the 996 pagecache). Doesn't include SwapCached 997SwapCached 998 Memory that once was swapped out, is swapped back in but 999 still also is in the swapfile (if memory is needed it 1000 doesn't need to be swapped out AGAIN because it is already 1001 in the swapfile. This saves I/O) 1002Active 1003 Memory that has been used more recently and usually not 1004 reclaimed unless absolutely necessary. 1005Inactive 1006 Memory which has been less recently used. It is more 1007 eligible to be reclaimed for other purposes 1008HighTotal, HighFree 1009 Highmem is all memory above ~860MB of physical memory. 1010 Highmem areas are for use by userspace programs, or 1011 for the pagecache. The kernel must use tricks to access 1012 this memory, making it slower to access than lowmem. 1013LowTotal, LowFree 1014 Lowmem is memory which can be used for everything that 1015 highmem can be used for, but it is also available for the 1016 kernel's use for its own data structures. Among many 1017 other things, it is where everything from the Slab is 1018 allocated. Bad things happen when you're out of lowmem. 1019SwapTotal 1020 total amount of swap space available 1021SwapFree 1022 Memory which has been evicted from RAM, and is temporarily 1023 on the disk 1024Dirty 1025 Memory which is waiting to get written back to the disk 1026Writeback 1027 Memory which is actively being written back to the disk 1028AnonPages 1029 Non-file backed pages mapped into userspace page tables 1030HardwareCorrupted 1031 The amount of RAM/memory in KB, the kernel identifies as 1032 corrupted. 1033AnonHugePages 1034 Non-file backed huge pages mapped into userspace page tables 1035Mapped 1036 files which have been mmaped, such as libraries 1037Shmem 1038 Total memory used by shared memory (shmem) and tmpfs 1039ShmemHugePages 1040 Memory used by shared memory (shmem) and tmpfs allocated 1041 with huge pages 1042ShmemPmdMapped 1043 Shared memory mapped into userspace with huge pages 1044KReclaimable 1045 Kernel allocations that the kernel will attempt to reclaim 1046 under memory pressure. Includes SReclaimable (below), and other 1047 direct allocations with a shrinker. 1048Slab 1049 in-kernel data structures cache 1050SReclaimable 1051 Part of Slab, that might be reclaimed, such as caches 1052SUnreclaim 1053 Part of Slab, that cannot be reclaimed on memory pressure 1054PageTables 1055 amount of memory dedicated to the lowest level of page 1056 tables. 1057NFS_Unstable 1058 Always zero. Previous counted pages which had been written to 1059 the server, but has not been committed to stable storage. 1060Bounce 1061 Memory used for block device "bounce buffers" 1062WritebackTmp 1063 Memory used by FUSE for temporary writeback buffers 1064CommitLimit 1065 Based on the overcommit ratio ('vm.overcommit_ratio'), 1066 this is the total amount of memory currently available to 1067 be allocated on the system. This limit is only adhered to 1068 if strict overcommit accounting is enabled (mode 2 in 1069 'vm.overcommit_memory'). 1070 1071 The CommitLimit is calculated with the following formula:: 1072 1073 CommitLimit = ([total RAM pages] - [total huge TLB pages]) * 1074 overcommit_ratio / 100 + [total swap pages] 1075 1076 For example, on a system with 1G of physical RAM and 7G 1077 of swap with a `vm.overcommit_ratio` of 30 it would 1078 yield a CommitLimit of 7.3G. 1079 1080 For more details, see the memory overcommit documentation 1081 in vm/overcommit-accounting. 1082Committed_AS 1083 The amount of memory presently allocated on the system. 1084 The committed memory is a sum of all of the memory which 1085 has been allocated by processes, even if it has not been 1086 "used" by them as of yet. A process which malloc()'s 1G 1087 of memory, but only touches 300M of it will show up as 1088 using 1G. This 1G is memory which has been "committed" to 1089 by the VM and can be used at any time by the allocating 1090 application. With strict overcommit enabled on the system 1091 (mode 2 in 'vm.overcommit_memory'), allocations which would 1092 exceed the CommitLimit (detailed above) will not be permitted. 1093 This is useful if one needs to guarantee that processes will 1094 not fail due to lack of memory once that memory has been 1095 successfully allocated. 1096VmallocTotal 1097 total size of vmalloc memory area 1098VmallocUsed 1099 amount of vmalloc area which is used 1100VmallocChunk 1101 largest contiguous block of vmalloc area which is free 1102Percpu 1103 Memory allocated to the percpu allocator used to back percpu 1104 allocations. This stat excludes the cost of metadata. 1105 1106vmallocinfo 1107~~~~~~~~~~~ 1108 1109Provides information about vmalloced/vmaped areas. One line per area, 1110containing the virtual address range of the area, size in bytes, 1111caller information of the creator, and optional information depending 1112on the kind of area: 1113 1114 ========== =================================================== 1115 pages=nr number of pages 1116 phys=addr if a physical address was specified 1117 ioremap I/O mapping (ioremap() and friends) 1118 vmalloc vmalloc() area 1119 vmap vmap()ed pages 1120 user VM_USERMAP area 1121 vpages buffer for pages pointers was vmalloced (huge area) 1122 N<node>=nr (Only on NUMA kernels) 1123 Number of pages allocated on memory node <node> 1124 ========== =================================================== 1125 1126:: 1127 1128 > cat /proc/vmallocinfo 1129 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ... 1130 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128 1131 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ... 1132 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64 1133 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f... 1134 phys=7fee8000 ioremap 1135 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f... 1136 phys=7fee7000 ioremap 1137 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210 1138 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ... 1139 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3 1140 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ... 1141 pages=2 vmalloc N1=2 1142 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ... 1143 /0x130 [x_tables] pages=4 vmalloc N0=4 1144 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ... 1145 pages=14 vmalloc N2=14 1146 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ... 1147 pages=4 vmalloc N1=4 1148 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ... 1149 pages=2 vmalloc N1=2 1150 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ... 1151 pages=10 vmalloc N0=10 1152 1153 1154softirqs 1155~~~~~~~~ 1156 1157Provides counts of softirq handlers serviced since boot time, for each CPU. 1158 1159:: 1160 1161 > cat /proc/softirqs 1162 CPU0 CPU1 CPU2 CPU3 1163 HI: 0 0 0 0 1164 TIMER: 27166 27120 27097 27034 1165 NET_TX: 0 0 0 17 1166 NET_RX: 42 0 0 39 1167 BLOCK: 0 0 107 1121 1168 TASKLET: 0 0 0 290 1169 SCHED: 27035 26983 26971 26746 1170 HRTIMER: 0 0 0 0 1171 RCU: 1678 1769 2178 2250 1172 1173 11741.3 IDE devices in /proc/ide 1175---------------------------- 1176 1177The subdirectory /proc/ide contains information about all IDE devices of which 1178the kernel is aware. There is one subdirectory for each IDE controller, the 1179file drivers and a link for each IDE device, pointing to the device directory 1180in the controller specific subtree. 1181 1182The file 'drivers' contains general information about the drivers used for the 1183IDE devices:: 1184 1185 > cat /proc/ide/drivers 1186 ide-cdrom version 4.53 1187 ide-disk version 1.08 1188 1189More detailed information can be found in the controller specific 1190subdirectories. These are named ide0, ide1 and so on. Each of these 1191directories contains the files shown in table 1-6. 1192 1193 1194.. table:: Table 1-6: IDE controller info in /proc/ide/ide? 1195 1196 ======= ======================================= 1197 File Content 1198 ======= ======================================= 1199 channel IDE channel (0 or 1) 1200 config Configuration (only for PCI/IDE bridge) 1201 mate Mate name 1202 model Type/Chipset of IDE controller 1203 ======= ======================================= 1204 1205Each device connected to a controller has a separate subdirectory in the 1206controllers directory. The files listed in table 1-7 are contained in these 1207directories. 1208 1209 1210.. table:: Table 1-7: IDE device information 1211 1212 ================ ========================================== 1213 File Content 1214 ================ ========================================== 1215 cache The cache 1216 capacity Capacity of the medium (in 512Byte blocks) 1217 driver driver and version 1218 geometry physical and logical geometry 1219 identify device identify block 1220 media media type 1221 model device identifier 1222 settings device setup 1223 smart_thresholds IDE disk management thresholds 1224 smart_values IDE disk management values 1225 ================ ========================================== 1226 1227The most interesting file is ``settings``. This file contains a nice 1228overview of the drive parameters:: 1229 1230 # cat /proc/ide/ide0/hda/settings 1231 name value min max mode 1232 ---- ----- --- --- ---- 1233 bios_cyl 526 0 65535 rw 1234 bios_head 255 0 255 rw 1235 bios_sect 63 0 63 rw 1236 breada_readahead 4 0 127 rw 1237 bswap 0 0 1 r 1238 file_readahead 72 0 2097151 rw 1239 io_32bit 0 0 3 rw 1240 keepsettings 0 0 1 rw 1241 max_kb_per_request 122 1 127 rw 1242 multcount 0 0 8 rw 1243 nice1 1 0 1 rw 1244 nowerr 0 0 1 rw 1245 pio_mode write-only 0 255 w 1246 slow 0 0 1 rw 1247 unmaskirq 0 0 1 rw 1248 using_dma 0 0 1 rw 1249 1250 12511.4 Networking info in /proc/net 1252-------------------------------- 1253 1254The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the 1255additional values you get for IP version 6 if you configure the kernel to 1256support this. Table 1-9 lists the files and their meaning. 1257 1258 1259.. table:: Table 1-8: IPv6 info in /proc/net 1260 1261 ========== ===================================================== 1262 File Content 1263 ========== ===================================================== 1264 udp6 UDP sockets (IPv6) 1265 tcp6 TCP sockets (IPv6) 1266 raw6 Raw device statistics (IPv6) 1267 igmp6 IP multicast addresses, which this host joined (IPv6) 1268 if_inet6 List of IPv6 interface addresses 1269 ipv6_route Kernel routing table for IPv6 1270 rt6_stats Global IPv6 routing tables statistics 1271 sockstat6 Socket statistics (IPv6) 1272 snmp6 Snmp data (IPv6) 1273 ========== ===================================================== 1274 1275.. table:: Table 1-9: Network info in /proc/net 1276 1277 ============= ================================================================ 1278 File Content 1279 ============= ================================================================ 1280 arp Kernel ARP table 1281 dev network devices with statistics 1282 dev_mcast the Layer2 multicast groups a device is listening too 1283 (interface index, label, number of references, number of bound 1284 addresses). 1285 dev_stat network device status 1286 ip_fwchains Firewall chain linkage 1287 ip_fwnames Firewall chain names 1288 ip_masq Directory containing the masquerading tables 1289 ip_masquerade Major masquerading table 1290 netstat Network statistics 1291 raw raw device statistics 1292 route Kernel routing table 1293 rpc Directory containing rpc info 1294 rt_cache Routing cache 1295 snmp SNMP data 1296 sockstat Socket statistics 1297 tcp TCP sockets 1298 udp UDP sockets 1299 unix UNIX domain sockets 1300 wireless Wireless interface data (Wavelan etc) 1301 igmp IP multicast addresses, which this host joined 1302 psched Global packet scheduler parameters. 1303 netlink List of PF_NETLINK sockets 1304 ip_mr_vifs List of multicast virtual interfaces 1305 ip_mr_cache List of multicast routing cache 1306 ============= ================================================================ 1307 1308You can use this information to see which network devices are available in 1309your system and how much traffic was routed over those devices:: 1310 1311 > cat /proc/net/dev 1312 Inter-|Receive |[... 1313 face |bytes packets errs drop fifo frame compressed multicast|[... 1314 lo: 908188 5596 0 0 0 0 0 0 [... 1315 ppp0:15475140 20721 410 0 0 410 0 0 [... 1316 eth0: 614530 7085 0 0 0 0 0 1 [... 1317 1318 ...] Transmit 1319 ...] bytes packets errs drop fifo colls carrier compressed 1320 ...] 908188 5596 0 0 0 0 0 0 1321 ...] 1375103 17405 0 0 0 0 0 0 1322 ...] 1703981 5535 0 0 0 3 0 0 1323 1324In addition, each Channel Bond interface has its own directory. For 1325example, the bond0 device will have a directory called /proc/net/bond0/. 1326It will contain information that is specific to that bond, such as the 1327current slaves of the bond, the link status of the slaves, and how 1328many times the slaves link has failed. 1329 13301.5 SCSI info 1331------------- 1332 1333If you have a SCSI host adapter in your system, you'll find a subdirectory 1334named after the driver for this adapter in /proc/scsi. You'll also see a list 1335of all recognized SCSI devices in /proc/scsi:: 1336 1337 >cat /proc/scsi/scsi 1338 Attached devices: 1339 Host: scsi0 Channel: 00 Id: 00 Lun: 00 1340 Vendor: IBM Model: DGHS09U Rev: 03E0 1341 Type: Direct-Access ANSI SCSI revision: 03 1342 Host: scsi0 Channel: 00 Id: 06 Lun: 00 1343 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04 1344 Type: CD-ROM ANSI SCSI revision: 02 1345 1346 1347The directory named after the driver has one file for each adapter found in 1348the system. These files contain information about the controller, including 1349the used IRQ and the IO address range. The amount of information shown is 1350dependent on the adapter you use. The example shows the output for an Adaptec 1351AHA-2940 SCSI adapter:: 1352 1353 > cat /proc/scsi/aic7xxx/0 1354 1355 Adaptec AIC7xxx driver version: 5.1.19/3.2.4 1356 Compile Options: 1357 TCQ Enabled By Default : Disabled 1358 AIC7XXX_PROC_STATS : Disabled 1359 AIC7XXX_RESET_DELAY : 5 1360 Adapter Configuration: 1361 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter 1362 Ultra Wide Controller 1363 PCI MMAPed I/O Base: 0xeb001000 1364 Adapter SEEPROM Config: SEEPROM found and used. 1365 Adaptec SCSI BIOS: Enabled 1366 IRQ: 10 1367 SCBs: Active 0, Max Active 2, 1368 Allocated 15, HW 16, Page 255 1369 Interrupts: 160328 1370 BIOS Control Word: 0x18b6 1371 Adapter Control Word: 0x005b 1372 Extended Translation: Enabled 1373 Disconnect Enable Flags: 0xffff 1374 Ultra Enable Flags: 0x0001 1375 Tag Queue Enable Flags: 0x0000 1376 Ordered Queue Tag Flags: 0x0000 1377 Default Tag Queue Depth: 8 1378 Tagged Queue By Device array for aic7xxx host instance 0: 1379 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255} 1380 Actual queue depth per device for aic7xxx host instance 0: 1381 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1} 1382 Statistics: 1383 (scsi0:0:0:0) 1384 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8 1385 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0) 1386 Total transfers 160151 (74577 reads and 85574 writes) 1387 (scsi0:0:6:0) 1388 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15 1389 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0) 1390 Total transfers 0 (0 reads and 0 writes) 1391 1392 13931.6 Parallel port info in /proc/parport 1394--------------------------------------- 1395 1396The directory /proc/parport contains information about the parallel ports of 1397your system. It has one subdirectory for each port, named after the port 1398number (0,1,2,...). 1399 1400These directories contain the four files shown in Table 1-10. 1401 1402 1403.. table:: Table 1-10: Files in /proc/parport 1404 1405 ========= ==================================================================== 1406 File Content 1407 ========= ==================================================================== 1408 autoprobe Any IEEE-1284 device ID information that has been acquired. 1409 devices list of the device drivers using that port. A + will appear by the 1410 name of the device currently using the port (it might not appear 1411 against any). 1412 hardware Parallel port's base address, IRQ line and DMA channel. 1413 irq IRQ that parport is using for that port. This is in a separate 1414 file to allow you to alter it by writing a new value in (IRQ 1415 number or none). 1416 ========= ==================================================================== 1417 14181.7 TTY info in /proc/tty 1419------------------------- 1420 1421Information about the available and actually used tty's can be found in the 1422directory /proc/tty. You'll find entries for drivers and line disciplines in 1423this directory, as shown in Table 1-11. 1424 1425 1426.. table:: Table 1-11: Files in /proc/tty 1427 1428 ============= ============================================== 1429 File Content 1430 ============= ============================================== 1431 drivers list of drivers and their usage 1432 ldiscs registered line disciplines 1433 driver/serial usage statistic and status of single tty lines 1434 ============= ============================================== 1435 1436To see which tty's are currently in use, you can simply look into the file 1437/proc/tty/drivers:: 1438 1439 > cat /proc/tty/drivers 1440 pty_slave /dev/pts 136 0-255 pty:slave 1441 pty_master /dev/ptm 128 0-255 pty:master 1442 pty_slave /dev/ttyp 3 0-255 pty:slave 1443 pty_master /dev/pty 2 0-255 pty:master 1444 serial /dev/cua 5 64-67 serial:callout 1445 serial /dev/ttyS 4 64-67 serial 1446 /dev/tty0 /dev/tty0 4 0 system:vtmaster 1447 /dev/ptmx /dev/ptmx 5 2 system 1448 /dev/console /dev/console 5 1 system:console 1449 /dev/tty /dev/tty 5 0 system:/dev/tty 1450 unknown /dev/tty 4 1-63 console 1451 1452 14531.8 Miscellaneous kernel statistics in /proc/stat 1454------------------------------------------------- 1455 1456Various pieces of information about kernel activity are available in the 1457/proc/stat file. All of the numbers reported in this file are aggregates 1458since the system first booted. For a quick look, simply cat the file:: 1459 1460 > cat /proc/stat 1461 cpu 2255 34 2290 22625563 6290 127 456 0 0 0 1462 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0 1463 cpu1 1123 0 849 11313845 2614 0 18 0 0 0 1464 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...] 1465 ctxt 1990473 1466 btime 1062191376 1467 processes 2915 1468 procs_running 1 1469 procs_blocked 0 1470 softirq 183433 0 21755 12 39 1137 231 21459 2263 1471 1472The very first "cpu" line aggregates the numbers in all of the other "cpuN" 1473lines. These numbers identify the amount of time the CPU has spent performing 1474different kinds of work. Time units are in USER_HZ (typically hundredths of a 1475second). The meanings of the columns are as follows, from left to right: 1476 1477- user: normal processes executing in user mode 1478- nice: niced processes executing in user mode 1479- system: processes executing in kernel mode 1480- idle: twiddling thumbs 1481- iowait: In a word, iowait stands for waiting for I/O to complete. But there 1482 are several problems: 1483 1484 1. CPU will not wait for I/O to complete, iowait is the time that a task is 1485 waiting for I/O to complete. When CPU goes into idle state for 1486 outstanding task I/O, another task will be scheduled on this CPU. 1487 2. In a multi-core CPU, the task waiting for I/O to complete is not running 1488 on any CPU, so the iowait of each CPU is difficult to calculate. 1489 3. The value of iowait field in /proc/stat will decrease in certain 1490 conditions. 1491 1492 So, the iowait is not reliable by reading from /proc/stat. 1493- irq: servicing interrupts 1494- softirq: servicing softirqs 1495- steal: involuntary wait 1496- guest: running a normal guest 1497- guest_nice: running a niced guest 1498 1499The "intr" line gives counts of interrupts serviced since boot time, for each 1500of the possible system interrupts. The first column is the total of all 1501interrupts serviced including unnumbered architecture specific interrupts; 1502each subsequent column is the total for that particular numbered interrupt. 1503Unnumbered interrupts are not shown, only summed into the total. 1504 1505The "ctxt" line gives the total number of context switches across all CPUs. 1506 1507The "btime" line gives the time at which the system booted, in seconds since 1508the Unix epoch. 1509 1510The "processes" line gives the number of processes and threads created, which 1511includes (but is not limited to) those created by calls to the fork() and 1512clone() system calls. 1513 1514The "procs_running" line gives the total number of threads that are 1515running or ready to run (i.e., the total number of runnable threads). 1516 1517The "procs_blocked" line gives the number of processes currently blocked, 1518waiting for I/O to complete. 1519 1520The "softirq" line gives counts of softirqs serviced since boot time, for each 1521of the possible system softirqs. The first column is the total of all 1522softirqs serviced; each subsequent column is the total for that particular 1523softirq. 1524 1525 15261.9 Ext4 file system parameters 1527------------------------------- 1528 1529Information about mounted ext4 file systems can be found in 1530/proc/fs/ext4. Each mounted filesystem will have a directory in 1531/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or 1532/proc/fs/ext4/dm-0). The files in each per-device directory are shown 1533in Table 1-12, below. 1534 1535.. table:: Table 1-12: Files in /proc/fs/ext4/<devname> 1536 1537 ============== ========================================================== 1538 File Content 1539 mb_groups details of multiblock allocator buddy cache of free blocks 1540 ============== ========================================================== 1541 15421.10 /proc/consoles 1543------------------- 1544Shows registered system console lines. 1545 1546To see which character device lines are currently used for the system console 1547/dev/console, you may simply look into the file /proc/consoles:: 1548 1549 > cat /proc/consoles 1550 tty0 -WU (ECp) 4:7 1551 ttyS0 -W- (Ep) 4:64 1552 1553The columns are: 1554 1555+--------------------+-------------------------------------------------------+ 1556| device | name of the device | 1557+====================+=======================================================+ 1558| operations | * R = can do read operations | 1559| | * W = can do write operations | 1560| | * U = can do unblank | 1561+--------------------+-------------------------------------------------------+ 1562| flags | * E = it is enabled | 1563| | * C = it is preferred console | 1564| | * B = it is primary boot console | 1565| | * p = it is used for printk buffer | 1566| | * b = it is not a TTY but a Braille device | 1567| | * a = it is safe to use when cpu is offline | 1568+--------------------+-------------------------------------------------------+ 1569| major:minor | major and minor number of the device separated by a | 1570| | colon | 1571+--------------------+-------------------------------------------------------+ 1572 1573Summary 1574------- 1575 1576The /proc file system serves information about the running system. It not only 1577allows access to process data but also allows you to request the kernel status 1578by reading files in the hierarchy. 1579 1580The directory structure of /proc reflects the types of information and makes 1581it easy, if not obvious, where to look for specific data. 1582 1583Chapter 2: Modifying System Parameters 1584====================================== 1585 1586In This Chapter 1587--------------- 1588 1589* Modifying kernel parameters by writing into files found in /proc/sys 1590* Exploring the files which modify certain parameters 1591* Review of the /proc/sys file tree 1592 1593------------------------------------------------------------------------------ 1594 1595A very interesting part of /proc is the directory /proc/sys. This is not only 1596a source of information, it also allows you to change parameters within the 1597kernel. Be very careful when attempting this. You can optimize your system, 1598but you can also cause it to crash. Never alter kernel parameters on a 1599production system. Set up a development machine and test to make sure that 1600everything works the way you want it to. You may have no alternative but to 1601reboot the machine once an error has been made. 1602 1603To change a value, simply echo the new value into the file. 1604You need to be root to do this. You can create your own boot script 1605to perform this every time your system boots. 1606 1607The files in /proc/sys can be used to fine tune and monitor miscellaneous and 1608general things in the operation of the Linux kernel. Since some of the files 1609can inadvertently disrupt your system, it is advisable to read both 1610documentation and source before actually making adjustments. In any case, be 1611very careful when writing to any of these files. The entries in /proc may 1612change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt 1613review the kernel documentation in the directory /usr/src/linux/Documentation. 1614This chapter is heavily based on the documentation included in the pre 2.2 1615kernels, and became part of it in version 2.2.1 of the Linux kernel. 1616 1617Please see: Documentation/admin-guide/sysctl/ directory for descriptions of these 1618entries. 1619 1620Summary 1621------- 1622 1623Certain aspects of kernel behavior can be modified at runtime, without the 1624need to recompile the kernel, or even to reboot the system. The files in the 1625/proc/sys tree can not only be read, but also modified. You can use the echo 1626command to write value into these files, thereby changing the default settings 1627of the kernel. 1628 1629 1630Chapter 3: Per-process Parameters 1631================================= 1632 16333.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score 1634-------------------------------------------------------------------------------- 1635 1636These files can be used to adjust the badness heuristic used to select which 1637process gets killed in out of memory (oom) conditions. 1638 1639The badness heuristic assigns a value to each candidate task ranging from 0 1640(never kill) to 1000 (always kill) to determine which process is targeted. The 1641units are roughly a proportion along that range of allowed memory the process 1642may allocate from based on an estimation of its current memory and swap use. 1643For example, if a task is using all allowed memory, its badness score will be 16441000. If it is using half of its allowed memory, its score will be 500. 1645 1646The amount of "allowed" memory depends on the context in which the oom killer 1647was called. If it is due to the memory assigned to the allocating task's cpuset 1648being exhausted, the allowed memory represents the set of mems assigned to that 1649cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed 1650memory represents the set of mempolicy nodes. If it is due to a memory 1651limit (or swap limit) being reached, the allowed memory is that configured 1652limit. Finally, if it is due to the entire system being out of memory, the 1653allowed memory represents all allocatable resources. 1654 1655The value of /proc/<pid>/oom_score_adj is added to the badness score before it 1656is used to determine which task to kill. Acceptable values range from -1000 1657(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to 1658polarize the preference for oom killing either by always preferring a certain 1659task or completely disabling it. The lowest possible value, -1000, is 1660equivalent to disabling oom killing entirely for that task since it will always 1661report a badness score of 0. 1662 1663Consequently, it is very simple for userspace to define the amount of memory to 1664consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for 1665example, is roughly equivalent to allowing the remainder of tasks sharing the 1666same system, cpuset, mempolicy, or memory controller resources to use at least 166750% more memory. A value of -500, on the other hand, would be roughly 1668equivalent to discounting 50% of the task's allowed memory from being considered 1669as scoring against the task. 1670 1671For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also 1672be used to tune the badness score. Its acceptable values range from -16 1673(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17 1674(OOM_DISABLE) to disable oom killing entirely for that task. Its value is 1675scaled linearly with /proc/<pid>/oom_score_adj. 1676 1677The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last 1678value set by a CAP_SYS_RESOURCE process. To reduce the value any lower 1679requires CAP_SYS_RESOURCE. 1680 1681 16823.2 /proc/<pid>/oom_score - Display current oom-killer score 1683------------------------------------------------------------- 1684 1685This file can be used to check the current score used by the oom-killer for 1686any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which 1687process should be killed in an out-of-memory situation. 1688 1689Please note that the exported value includes oom_score_adj so it is 1690effectively in range [0,2000]. 1691 1692 16933.3 /proc/<pid>/io - Display the IO accounting fields 1694------------------------------------------------------- 1695 1696This file contains IO statistics for each running process. 1697 1698Example 1699~~~~~~~ 1700 1701:: 1702 1703 test:/tmp # dd if=/dev/zero of=/tmp/test.dat & 1704 [1] 3828 1705 1706 test:/tmp # cat /proc/3828/io 1707 rchar: 323934931 1708 wchar: 323929600 1709 syscr: 632687 1710 syscw: 632675 1711 read_bytes: 0 1712 write_bytes: 323932160 1713 cancelled_write_bytes: 0 1714 1715 1716Description 1717~~~~~~~~~~~ 1718 1719rchar 1720^^^^^ 1721 1722I/O counter: chars read 1723The number of bytes which this task has caused to be read from storage. This 1724is simply the sum of bytes which this process passed to read() and pread(). 1725It includes things like tty IO and it is unaffected by whether or not actual 1726physical disk IO was required (the read might have been satisfied from 1727pagecache). 1728 1729 1730wchar 1731^^^^^ 1732 1733I/O counter: chars written 1734The number of bytes which this task has caused, or shall cause to be written 1735to disk. Similar caveats apply here as with rchar. 1736 1737 1738syscr 1739^^^^^ 1740 1741I/O counter: read syscalls 1742Attempt to count the number of read I/O operations, i.e. syscalls like read() 1743and pread(). 1744 1745 1746syscw 1747^^^^^ 1748 1749I/O counter: write syscalls 1750Attempt to count the number of write I/O operations, i.e. syscalls like 1751write() and pwrite(). 1752 1753 1754read_bytes 1755^^^^^^^^^^ 1756 1757I/O counter: bytes read 1758Attempt to count the number of bytes which this process really did cause to 1759be fetched from the storage layer. Done at the submit_bio() level, so it is 1760accurate for block-backed filesystems. <please add status regarding NFS and 1761CIFS at a later time> 1762 1763 1764write_bytes 1765^^^^^^^^^^^ 1766 1767I/O counter: bytes written 1768Attempt to count the number of bytes which this process caused to be sent to 1769the storage layer. This is done at page-dirtying time. 1770 1771 1772cancelled_write_bytes 1773^^^^^^^^^^^^^^^^^^^^^ 1774 1775The big inaccuracy here is truncate. If a process writes 1MB to a file and 1776then deletes the file, it will in fact perform no writeout. But it will have 1777been accounted as having caused 1MB of write. 1778In other words: The number of bytes which this process caused to not happen, 1779by truncating pagecache. A task can cause "negative" IO too. If this task 1780truncates some dirty pagecache, some IO which another task has been accounted 1781for (in its write_bytes) will not be happening. We _could_ just subtract that 1782from the truncating task's write_bytes, but there is information loss in doing 1783that. 1784 1785 1786.. Note:: 1787 1788 At its current implementation state, this is a bit racy on 32-bit machines: 1789 if process A reads process B's /proc/pid/io while process B is updating one 1790 of those 64-bit counters, process A could see an intermediate result. 1791 1792 1793More information about this can be found within the taskstats documentation in 1794Documentation/accounting. 1795 17963.4 /proc/<pid>/coredump_filter - Core dump filtering settings 1797--------------------------------------------------------------- 1798When a process is dumped, all anonymous memory is written to a core file as 1799long as the size of the core file isn't limited. But sometimes we don't want 1800to dump some memory segments, for example, huge shared memory or DAX. 1801Conversely, sometimes we want to save file-backed memory segments into a core 1802file, not only the individual files. 1803 1804/proc/<pid>/coredump_filter allows you to customize which memory segments 1805will be dumped when the <pid> process is dumped. coredump_filter is a bitmask 1806of memory types. If a bit of the bitmask is set, memory segments of the 1807corresponding memory type are dumped, otherwise they are not dumped. 1808 1809The following 9 memory types are supported: 1810 1811 - (bit 0) anonymous private memory 1812 - (bit 1) anonymous shared memory 1813 - (bit 2) file-backed private memory 1814 - (bit 3) file-backed shared memory 1815 - (bit 4) ELF header pages in file-backed private memory areas (it is 1816 effective only if the bit 2 is cleared) 1817 - (bit 5) hugetlb private memory 1818 - (bit 6) hugetlb shared memory 1819 - (bit 7) DAX private memory 1820 - (bit 8) DAX shared memory 1821 1822 Note that MMIO pages such as frame buffer are never dumped and vDSO pages 1823 are always dumped regardless of the bitmask status. 1824 1825 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is 1826 only affected by bit 5-6, and DAX is only affected by bits 7-8. 1827 1828The default value of coredump_filter is 0x33; this means all anonymous memory 1829segments, ELF header pages and hugetlb private memory are dumped. 1830 1831If you don't want to dump all shared memory segments attached to pid 1234, 1832write 0x31 to the process's proc file:: 1833 1834 $ echo 0x31 > /proc/1234/coredump_filter 1835 1836When a new process is created, the process inherits the bitmask status from its 1837parent. It is useful to set up coredump_filter before the program runs. 1838For example:: 1839 1840 $ echo 0x7 > /proc/self/coredump_filter 1841 $ ./some_program 1842 18433.5 /proc/<pid>/mountinfo - Information about mounts 1844-------------------------------------------------------- 1845 1846This file contains lines of the form:: 1847 1848 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue 1849 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11) 1850 1851 (1) mount ID: unique identifier of the mount (may be reused after umount) 1852 (2) parent ID: ID of parent (or of self for the top of the mount tree) 1853 (3) major:minor: value of st_dev for files on filesystem 1854 (4) root: root of the mount within the filesystem 1855 (5) mount point: mount point relative to the process's root 1856 (6) mount options: per mount options 1857 (7) optional fields: zero or more fields of the form "tag[:value]" 1858 (8) separator: marks the end of the optional fields 1859 (9) filesystem type: name of filesystem of the form "type[.subtype]" 1860 (10) mount source: filesystem specific information or "none" 1861 (11) super options: per super block options 1862 1863Parsers should ignore all unrecognised optional fields. Currently the 1864possible optional fields are: 1865 1866================ ============================================================== 1867shared:X mount is shared in peer group X 1868master:X mount is slave to peer group X 1869propagate_from:X mount is slave and receives propagation from peer group X [#]_ 1870unbindable mount is unbindable 1871================ ============================================================== 1872 1873.. [#] X is the closest dominant peer group under the process's root. If 1874 X is the immediate master of the mount, or if there's no dominant peer 1875 group under the same root, then only the "master:X" field is present 1876 and not the "propagate_from:X" field. 1877 1878For more information on mount propagation see: 1879 1880 Documentation/filesystems/sharedsubtree.rst 1881 1882 18833.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 1884-------------------------------------------------------- 1885These files provide a method to access a task's comm value. It also allows for 1886a task to set its own or one of its thread siblings comm value. The comm value 1887is limited in size compared to the cmdline value, so writing anything longer 1888then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated 1889comm value. 1890 1891 18923.7 /proc/<pid>/task/<tid>/children - Information about task children 1893------------------------------------------------------------------------- 1894This file provides a fast way to retrieve first level children pids 1895of a task pointed by <pid>/<tid> pair. The format is a space separated 1896stream of pids. 1897 1898Note the "first level" here -- if a child has its own children they will 1899not be listed here; one needs to read /proc/<children-pid>/task/<tid>/children 1900to obtain the descendants. 1901 1902Since this interface is intended to be fast and cheap it doesn't 1903guarantee to provide precise results and some children might be 1904skipped, especially if they've exited right after we printed their 1905pids, so one needs to either stop or freeze processes being inspected 1906if precise results are needed. 1907 1908 19093.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 1910--------------------------------------------------------------- 1911This file provides information associated with an opened file. The regular 1912files have at least three fields -- 'pos', 'flags' and 'mnt_id'. The 'pos' 1913represents the current offset of the opened file in decimal form [see lseek(2) 1914for details], 'flags' denotes the octal O_xxx mask the file has been 1915created with [see open(2) for details] and 'mnt_id' represents mount ID of 1916the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo 1917for details]. 1918 1919A typical output is:: 1920 1921 pos: 0 1922 flags: 0100002 1923 mnt_id: 19 1924 1925All locks associated with a file descriptor are shown in its fdinfo too:: 1926 1927 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF 1928 1929The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags 1930pair provide additional information particular to the objects they represent. 1931 1932Eventfd files 1933~~~~~~~~~~~~~ 1934 1935:: 1936 1937 pos: 0 1938 flags: 04002 1939 mnt_id: 9 1940 eventfd-count: 5a 1941 1942where 'eventfd-count' is hex value of a counter. 1943 1944Signalfd files 1945~~~~~~~~~~~~~~ 1946 1947:: 1948 1949 pos: 0 1950 flags: 04002 1951 mnt_id: 9 1952 sigmask: 0000000000000200 1953 1954where 'sigmask' is hex value of the signal mask associated 1955with a file. 1956 1957Epoll files 1958~~~~~~~~~~~ 1959 1960:: 1961 1962 pos: 0 1963 flags: 02 1964 mnt_id: 9 1965 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7 1966 1967where 'tfd' is a target file descriptor number in decimal form, 1968'events' is events mask being watched and the 'data' is data 1969associated with a target [see epoll(7) for more details]. 1970 1971The 'pos' is current offset of the target file in decimal form 1972[see lseek(2)], 'ino' and 'sdev' are inode and device numbers 1973where target file resides, all in hex format. 1974 1975Fsnotify files 1976~~~~~~~~~~~~~~ 1977For inotify files the format is the following:: 1978 1979 pos: 0 1980 flags: 02000000 1981 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d 1982 1983where 'wd' is a watch descriptor in decimal form, i.e. a target file 1984descriptor number, 'ino' and 'sdev' are inode and device where the 1985target file resides and the 'mask' is the mask of events, all in hex 1986form [see inotify(7) for more details]. 1987 1988If the kernel was built with exportfs support, the path to the target 1989file is encoded as a file handle. The file handle is provided by three 1990fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex 1991format. 1992 1993If the kernel is built without exportfs support the file handle won't be 1994printed out. 1995 1996If there is no inotify mark attached yet the 'inotify' line will be omitted. 1997 1998For fanotify files the format is:: 1999 2000 pos: 0 2001 flags: 02 2002 mnt_id: 9 2003 fanotify flags:10 event-flags:0 2004 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003 2005 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4 2006 2007where fanotify 'flags' and 'event-flags' are values used in fanotify_init 2008call, 'mnt_id' is the mount point identifier, 'mflags' is the value of 2009flags associated with mark which are tracked separately from events 2010mask. 'ino' and 'sdev' are target inode and device, 'mask' is the events 2011mask and 'ignored_mask' is the mask of events which are to be ignored. 2012All are in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask' 2013provide information about flags and mask used in fanotify_mark 2014call [see fsnotify manpage for details]. 2015 2016While the first three lines are mandatory and always printed, the rest is 2017optional and may be omitted if no marks created yet. 2018 2019Timerfd files 2020~~~~~~~~~~~~~ 2021 2022:: 2023 2024 pos: 0 2025 flags: 02 2026 mnt_id: 9 2027 clockid: 0 2028 ticks: 0 2029 settime flags: 01 2030 it_value: (0, 49406829) 2031 it_interval: (1, 0) 2032 2033where 'clockid' is the clock type and 'ticks' is the number of the timer expirations 2034that have occurred [see timerfd_create(2) for details]. 'settime flags' are 2035flags in octal form been used to setup the timer [see timerfd_settime(2) for 2036details]. 'it_value' is remaining time until the timer expiration. 2037'it_interval' is the interval for the timer. Note the timer might be set up 2038with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value' 2039still exhibits timer's remaining time. 2040 20413.9 /proc/<pid>/map_files - Information about memory mapped files 2042--------------------------------------------------------------------- 2043This directory contains symbolic links which represent memory mapped files 2044the process is maintaining. Example output:: 2045 2046 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so 2047 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so 2048 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so 2049 | ... 2050 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1 2051 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls 2052 2053The name of a link represents the virtual memory bounds of a mapping, i.e. 2054vm_area_struct::vm_start-vm_area_struct::vm_end. 2055 2056The main purpose of the map_files is to retrieve a set of memory mapped 2057files in a fast way instead of parsing /proc/<pid>/maps or 2058/proc/<pid>/smaps, both of which contain many more records. At the same 2059time one can open(2) mappings from the listings of two processes and 2060comparing their inode numbers to figure out which anonymous memory areas 2061are actually shared. 2062 20633.10 /proc/<pid>/timerslack_ns - Task timerslack value 2064--------------------------------------------------------- 2065This file provides the value of the task's timerslack value in nanoseconds. 2066This value specifies an amount of time that normal timers may be deferred 2067in order to coalesce timers and avoid unnecessary wakeups. 2068 2069This allows a task's interactivity vs power consumption tradeoff to be 2070adjusted. 2071 2072Writing 0 to the file will set the task's timerslack to the default value. 2073 2074Valid values are from 0 - ULLONG_MAX 2075 2076An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level 2077permissions on the task specified to change its timerslack_ns value. 2078 20793.11 /proc/<pid>/patch_state - Livepatch patch operation state 2080----------------------------------------------------------------- 2081When CONFIG_LIVEPATCH is enabled, this file displays the value of the 2082patch state for the task. 2083 2084A value of '-1' indicates that no patch is in transition. 2085 2086A value of '0' indicates that a patch is in transition and the task is 2087unpatched. If the patch is being enabled, then the task hasn't been 2088patched yet. If the patch is being disabled, then the task has already 2089been unpatched. 2090 2091A value of '1' indicates that a patch is in transition and the task is 2092patched. If the patch is being enabled, then the task has already been 2093patched. If the patch is being disabled, then the task hasn't been 2094unpatched yet. 2095 20963.12 /proc/<pid>/arch_status - task architecture specific status 2097------------------------------------------------------------------- 2098When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the 2099architecture specific status of the task. 2100 2101Example 2102~~~~~~~ 2103 2104:: 2105 2106 $ cat /proc/6753/arch_status 2107 AVX512_elapsed_ms: 8 2108 2109Description 2110~~~~~~~~~~~ 2111 2112x86 specific entries 2113~~~~~~~~~~~~~~~~~~~~~ 2114 2115AVX512_elapsed_ms 2116^^^^^^^^^^^^^^^^^^ 2117 2118 If AVX512 is supported on the machine, this entry shows the milliseconds 2119 elapsed since the last time AVX512 usage was recorded. The recording 2120 happens on a best effort basis when a task is scheduled out. This means 2121 that the value depends on two factors: 2122 2123 1) The time which the task spent on the CPU without being scheduled 2124 out. With CPU isolation and a single runnable task this can take 2125 several seconds. 2126 2127 2) The time since the task was scheduled out last. Depending on the 2128 reason for being scheduled out (time slice exhausted, syscall ...) 2129 this can be arbitrary long time. 2130 2131 As a consequence the value cannot be considered precise and authoritative 2132 information. The application which uses this information has to be aware 2133 of the overall scenario on the system in order to determine whether a 2134 task is a real AVX512 user or not. Precise information can be obtained 2135 with performance counters. 2136 2137 A special value of '-1' indicates that no AVX512 usage was recorded, thus 2138 the task is unlikely an AVX512 user, but depends on the workload and the 2139 scheduling scenario, it also could be a false negative mentioned above. 2140 2141Chapter 4: Configuring procfs 2142============================= 2143 21444.1 Mount options 2145--------------------- 2146 2147The following mount options are supported: 2148 2149 ========= ======================================================== 2150 hidepid= Set /proc/<pid>/ access mode. 2151 gid= Set the group authorized to learn processes information. 2152 subset= Show only the specified subset of procfs. 2153 ========= ======================================================== 2154 2155hidepid=off or hidepid=0 means classic mode - everybody may access all 2156/proc/<pid>/ directories (default). 2157 2158hidepid=noaccess or hidepid=1 means users may not access any /proc/<pid>/ 2159directories but their own. Sensitive files like cmdline, sched*, status are now 2160protected against other users. This makes it impossible to learn whether any 2161user runs specific program (given the program doesn't reveal itself by its 2162behaviour). As an additional bonus, as /proc/<pid>/cmdline is unaccessible for 2163other users, poorly written programs passing sensitive information via program 2164arguments are now protected against local eavesdroppers. 2165 2166hidepid=invisible or hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be 2167fully invisible to other users. It doesn't mean that it hides a fact whether a 2168process with a specific pid value exists (it can be learned by other means, e.g. 2169by "kill -0 $PID"), but it hides process' uid and gid, which may be learned by 2170stat()'ing /proc/<pid>/ otherwise. It greatly complicates an intruder's task of 2171gathering information about running processes, whether some daemon runs with 2172elevated privileges, whether other user runs some sensitive program, whether 2173other users run any program at all, etc. 2174 2175hidepid=ptraceable or hidepid=4 means that procfs should only contain 2176/proc/<pid>/ directories that the caller can ptrace. 2177 2178gid= defines a group authorized to learn processes information otherwise 2179prohibited by hidepid=. If you use some daemon like identd which needs to learn 2180information about processes information, just add identd to this group. 2181 2182subset=pid hides all top level files and directories in the procfs that 2183are not related to tasks. 2184 2185Chapter 5: Filesystem behavior 2186============================== 2187 2188Originally, before the advent of pid namepsace, procfs was a global file 2189system. It means that there was only one procfs instance in the system. 2190 2191When pid namespace was added, a separate procfs instance was mounted in 2192each pid namespace. So, procfs mount options are global among all 2193mountpoints within the same namespace:: 2194 2195 # grep ^proc /proc/mounts 2196 proc /proc proc rw,relatime,hidepid=2 0 0 2197 2198 # strace -e mount mount -o hidepid=1 -t proc proc /tmp/proc 2199 mount("proc", "/tmp/proc", "proc", 0, "hidepid=1") = 0 2200 +++ exited with 0 +++ 2201 2202 # grep ^proc /proc/mounts 2203 proc /proc proc rw,relatime,hidepid=2 0 0 2204 proc /tmp/proc proc rw,relatime,hidepid=2 0 0 2205 2206and only after remounting procfs mount options will change at all 2207mountpoints:: 2208 2209 # mount -o remount,hidepid=1 -t proc proc /tmp/proc 2210 2211 # grep ^proc /proc/mounts 2212 proc /proc proc rw,relatime,hidepid=1 0 0 2213 proc /tmp/proc proc rw,relatime,hidepid=1 0 0 2214 2215This behavior is different from the behavior of other filesystems. 2216 2217The new procfs behavior is more like other filesystems. Each procfs mount 2218creates a new procfs instance. Mount options affect own procfs instance. 2219It means that it became possible to have several procfs instances 2220displaying tasks with different filtering options in one pid namespace:: 2221 2222 # mount -o hidepid=invisible -t proc proc /proc 2223 # mount -o hidepid=noaccess -t proc proc /tmp/proc 2224 # grep ^proc /proc/mounts 2225 proc /proc proc rw,relatime,hidepid=invisible 0 0 2226 proc /tmp/proc proc rw,relatime,hidepid=noaccess 0 0 2227