1------------------------------------------------------------------------------ 2 T H E /proc F I L E S Y S T E M 3------------------------------------------------------------------------------ 4/proc/sys Terrehon Bowden <terrehon@pacbell.net> October 7 1999 5 Bodo Bauer <bb@ricochet.net> 6 72.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000 8move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009 9------------------------------------------------------------------------------ 10Version 1.3 Kernel version 2.2.12 11 Kernel version 2.4.0-test11-pre4 12------------------------------------------------------------------------------ 13fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009 14 15Table of Contents 16----------------- 17 18 0 Preface 19 0.1 Introduction/Credits 20 0.2 Legal Stuff 21 22 1 Collecting System Information 23 1.1 Process-Specific Subdirectories 24 1.2 Kernel data 25 1.3 IDE devices in /proc/ide 26 1.4 Networking info in /proc/net 27 1.5 SCSI info 28 1.6 Parallel port info in /proc/parport 29 1.7 TTY info in /proc/tty 30 1.8 Miscellaneous kernel statistics in /proc/stat 31 1.9 Ext4 file system parameters 32 33 2 Modifying System Parameters 34 35 3 Per-Process Parameters 36 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer 37 score 38 3.2 /proc/<pid>/oom_score - Display current oom-killer score 39 3.3 /proc/<pid>/io - Display the IO accounting fields 40 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings 41 3.5 /proc/<pid>/mountinfo - Information about mounts 42 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 43 3.7 /proc/<pid>/task/<tid>/children - Information about task children 44 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 45 3.9 /proc/<pid>/map_files - Information about memory mapped files 46 3.10 /proc/<pid>/timerslack_ns - Task timerslack value 47 3.11 /proc/<pid>/patch_state - Livepatch patch operation state 48 49 4 Configuring procfs 50 4.1 Mount options 51 52------------------------------------------------------------------------------ 53Preface 54------------------------------------------------------------------------------ 55 560.1 Introduction/Credits 57------------------------ 58 59This documentation is part of a soon (or so we hope) to be released book on 60the SuSE Linux distribution. As there is no complete documentation for the 61/proc file system and we've used many freely available sources to write these 62chapters, it seems only fair to give the work back to the Linux community. 63This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm 64afraid it's still far from complete, but we hope it will be useful. As far as 65we know, it is the first 'all-in-one' document about the /proc file system. It 66is focused on the Intel x86 hardware, so if you are looking for PPC, ARM, 67SPARC, AXP, etc., features, you probably won't find what you are looking for. 68It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But 69additions and patches are welcome and will be added to this document if you 70mail them to Bodo. 71 72We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of 73other people for help compiling this documentation. We'd also like to extend a 74special thank you to Andi Kleen for documentation, which we relied on heavily 75to create this document, as well as the additional information he provided. 76Thanks to everybody else who contributed source or docs to the Linux kernel 77and helped create a great piece of software... :) 78 79If you have any comments, corrections or additions, please don't hesitate to 80contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this 81document. 82 83The latest version of this document is available online at 84http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html 85 86If the above direction does not works for you, you could try the kernel 87mailing list at linux-kernel@vger.kernel.org and/or try to reach me at 88comandante@zaralinux.com. 89 900.2 Legal Stuff 91--------------- 92 93We don't guarantee the correctness of this document, and if you come to us 94complaining about how you screwed up your system because of incorrect 95documentation, we won't feel responsible... 96 97------------------------------------------------------------------------------ 98CHAPTER 1: COLLECTING SYSTEM INFORMATION 99------------------------------------------------------------------------------ 100 101------------------------------------------------------------------------------ 102In This Chapter 103------------------------------------------------------------------------------ 104* Investigating the properties of the pseudo file system /proc and its 105 ability to provide information on the running Linux system 106* Examining /proc's structure 107* Uncovering various information about the kernel and the processes running 108 on the system 109------------------------------------------------------------------------------ 110 111 112The proc file system acts as an interface to internal data structures in the 113kernel. It can be used to obtain information about the system and to change 114certain kernel parameters at runtime (sysctl). 115 116First, we'll take a look at the read-only parts of /proc. In Chapter 2, we 117show you how you can use /proc/sys to change settings. 118 1191.1 Process-Specific Subdirectories 120----------------------------------- 121 122The directory /proc contains (among other things) one subdirectory for each 123process running on the system, which is named after the process ID (PID). 124 125The link self points to the process reading the file system. Each process 126subdirectory has the entries listed in Table 1-1. 127 128 129Table 1-1: Process specific entries in /proc 130.............................................................................. 131 File Content 132 clear_refs Clears page referenced bits shown in smaps output 133 cmdline Command line arguments 134 cpu Current and last cpu in which it was executed (2.4)(smp) 135 cwd Link to the current working directory 136 environ Values of environment variables 137 exe Link to the executable of this process 138 fd Directory, which contains all file descriptors 139 maps Memory maps to executables and library files (2.4) 140 mem Memory held by this process 141 root Link to the root directory of this process 142 stat Process status 143 statm Process memory status information 144 status Process status in human readable form 145 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function 146 symbol the task is blocked in - or "0" if not blocked. 147 pagemap Page table 148 stack Report full stack trace, enable via CONFIG_STACKTRACE 149 smaps an extension based on maps, showing the memory consumption of 150 each mapping and flags associated with it 151 numa_maps an extension based on maps, showing the memory locality and 152 binding policy as well as mem usage (in pages) of each mapping. 153.............................................................................. 154 155For example, to get the status information of a process, all you have to do is 156read the file /proc/PID/status: 157 158 >cat /proc/self/status 159 Name: cat 160 State: R (running) 161 Tgid: 5452 162 Pid: 5452 163 PPid: 743 164 TracerPid: 0 (2.4) 165 Uid: 501 501 501 501 166 Gid: 100 100 100 100 167 FDSize: 256 168 Groups: 100 14 16 169 VmPeak: 5004 kB 170 VmSize: 5004 kB 171 VmLck: 0 kB 172 VmHWM: 476 kB 173 VmRSS: 476 kB 174 RssAnon: 352 kB 175 RssFile: 120 kB 176 RssShmem: 4 kB 177 VmData: 156 kB 178 VmStk: 88 kB 179 VmExe: 68 kB 180 VmLib: 1412 kB 181 VmPTE: 20 kb 182 VmSwap: 0 kB 183 HugetlbPages: 0 kB 184 CoreDumping: 0 185 Threads: 1 186 SigQ: 0/28578 187 SigPnd: 0000000000000000 188 ShdPnd: 0000000000000000 189 SigBlk: 0000000000000000 190 SigIgn: 0000000000000000 191 SigCgt: 0000000000000000 192 CapInh: 00000000fffffeff 193 CapPrm: 0000000000000000 194 CapEff: 0000000000000000 195 CapBnd: ffffffffffffffff 196 NoNewPrivs: 0 197 Seccomp: 0 198 voluntary_ctxt_switches: 0 199 nonvoluntary_ctxt_switches: 1 200 201This shows you nearly the same information you would get if you viewed it with 202the ps command. In fact, ps uses the proc file system to obtain its 203information. But you get a more detailed view of the process by reading the 204file /proc/PID/status. It fields are described in table 1-2. 205 206The statm file contains more detailed information about the process 207memory usage. Its seven fields are explained in Table 1-3. The stat file 208contains details information about the process itself. Its fields are 209explained in Table 1-4. 210 211(for SMP CONFIG users) 212For making accounting scalable, RSS related information are handled in an 213asynchronous manner and the value may not be very precise. To see a precise 214snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table. 215It's slow but very precise. 216 217Table 1-2: Contents of the status files (as of 4.8) 218.............................................................................. 219 Field Content 220 Name filename of the executable 221 Umask file mode creation mask 222 State state (R is running, S is sleeping, D is sleeping 223 in an uninterruptible wait, Z is zombie, 224 T is traced or stopped) 225 Tgid thread group ID 226 Ngid NUMA group ID (0 if none) 227 Pid process id 228 PPid process id of the parent process 229 TracerPid PID of process tracing this process (0 if not) 230 Uid Real, effective, saved set, and file system UIDs 231 Gid Real, effective, saved set, and file system GIDs 232 FDSize number of file descriptor slots currently allocated 233 Groups supplementary group list 234 NStgid descendant namespace thread group ID hierarchy 235 NSpid descendant namespace process ID hierarchy 236 NSpgid descendant namespace process group ID hierarchy 237 NSsid descendant namespace session ID hierarchy 238 VmPeak peak virtual memory size 239 VmSize total program size 240 VmLck locked memory size 241 VmPin pinned memory size 242 VmHWM peak resident set size ("high water mark") 243 VmRSS size of memory portions. It contains the three 244 following parts (VmRSS = RssAnon + RssFile + RssShmem) 245 RssAnon size of resident anonymous memory 246 RssFile size of resident file mappings 247 RssShmem size of resident shmem memory (includes SysV shm, 248 mapping of tmpfs and shared anonymous mappings) 249 VmData size of private data segments 250 VmStk size of stack segments 251 VmExe size of text segment 252 VmLib size of shared library code 253 VmPTE size of page table entries 254 VmSwap amount of swap used by anonymous private data 255 (shmem swap usage is not included) 256 HugetlbPages size of hugetlb memory portions 257 CoreDumping process's memory is currently being dumped 258 (killing the process may lead to a corrupted core) 259 Threads number of threads 260 SigQ number of signals queued/max. number for queue 261 SigPnd bitmap of pending signals for the thread 262 ShdPnd bitmap of shared pending signals for the process 263 SigBlk bitmap of blocked signals 264 SigIgn bitmap of ignored signals 265 SigCgt bitmap of caught signals 266 CapInh bitmap of inheritable capabilities 267 CapPrm bitmap of permitted capabilities 268 CapEff bitmap of effective capabilities 269 CapBnd bitmap of capabilities bounding set 270 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...) 271 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...) 272 Cpus_allowed mask of CPUs on which this process may run 273 Cpus_allowed_list Same as previous, but in "list format" 274 Mems_allowed mask of memory nodes allowed to this process 275 Mems_allowed_list Same as previous, but in "list format" 276 voluntary_ctxt_switches number of voluntary context switches 277 nonvoluntary_ctxt_switches number of non voluntary context switches 278.............................................................................. 279 280Table 1-3: Contents of the statm files (as of 2.6.8-rc3) 281.............................................................................. 282 Field Content 283 size total program size (pages) (same as VmSize in status) 284 resident size of memory portions (pages) (same as VmRSS in status) 285 shared number of pages that are shared (i.e. backed by a file, same 286 as RssFile+RssShmem in status) 287 trs number of pages that are 'code' (not including libs; broken, 288 includes data segment) 289 lrs number of pages of library (always 0 on 2.6) 290 drs number of pages of data/stack (including libs; broken, 291 includes library text) 292 dt number of dirty pages (always 0 on 2.6) 293.............................................................................. 294 295 296Table 1-4: Contents of the stat files (as of 2.6.30-rc7) 297.............................................................................. 298 Field Content 299 pid process id 300 tcomm filename of the executable 301 state state (R is running, S is sleeping, D is sleeping in an 302 uninterruptible wait, Z is zombie, T is traced or stopped) 303 ppid process id of the parent process 304 pgrp pgrp of the process 305 sid session id 306 tty_nr tty the process uses 307 tty_pgrp pgrp of the tty 308 flags task flags 309 min_flt number of minor faults 310 cmin_flt number of minor faults with child's 311 maj_flt number of major faults 312 cmaj_flt number of major faults with child's 313 utime user mode jiffies 314 stime kernel mode jiffies 315 cutime user mode jiffies with child's 316 cstime kernel mode jiffies with child's 317 priority priority level 318 nice nice level 319 num_threads number of threads 320 it_real_value (obsolete, always 0) 321 start_time time the process started after system boot 322 vsize virtual memory size 323 rss resident set memory size 324 rsslim current limit in bytes on the rss 325 start_code address above which program text can run 326 end_code address below which program text can run 327 start_stack address of the start of the main process stack 328 esp current value of ESP 329 eip current value of EIP 330 pending bitmap of pending signals 331 blocked bitmap of blocked signals 332 sigign bitmap of ignored signals 333 sigcatch bitmap of caught signals 334 0 (place holder, used to be the wchan address, use /proc/PID/wchan instead) 335 0 (place holder) 336 0 (place holder) 337 exit_signal signal to send to parent thread on exit 338 task_cpu which CPU the task is scheduled on 339 rt_priority realtime priority 340 policy scheduling policy (man sched_setscheduler) 341 blkio_ticks time spent waiting for block IO 342 gtime guest time of the task in jiffies 343 cgtime guest time of the task children in jiffies 344 start_data address above which program data+bss is placed 345 end_data address below which program data+bss is placed 346 start_brk address above which program heap can be expanded with brk() 347 arg_start address above which program command line is placed 348 arg_end address below which program command line is placed 349 env_start address above which program environment is placed 350 env_end address below which program environment is placed 351 exit_code the thread's exit_code in the form reported by the waitpid system call 352.............................................................................. 353 354The /proc/PID/maps file containing the currently mapped memory regions and 355their access permissions. 356 357The format is: 358 359address perms offset dev inode pathname 360 36108048000-08049000 r-xp 00000000 03:00 8312 /opt/test 36208049000-0804a000 rw-p 00001000 03:00 8312 /opt/test 3630804a000-0806b000 rw-p 00000000 00:00 0 [heap] 364a7cb1000-a7cb2000 ---p 00000000 00:00 0 365a7cb2000-a7eb2000 rw-p 00000000 00:00 0 366a7eb2000-a7eb3000 ---p 00000000 00:00 0 367a7eb3000-a7ed5000 rw-p 00000000 00:00 0 368a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6 369a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6 370a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6 371a800b000-a800e000 rw-p 00000000 00:00 0 372a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0 373a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0 374a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0 375a8024000-a8027000 rw-p 00000000 00:00 0 376a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2 377a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2 378a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2 379aff35000-aff4a000 rw-p 00000000 00:00 0 [stack] 380ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso] 381 382where "address" is the address space in the process that it occupies, "perms" 383is a set of permissions: 384 385 r = read 386 w = write 387 x = execute 388 s = shared 389 p = private (copy on write) 390 391"offset" is the offset into the mapping, "dev" is the device (major:minor), and 392"inode" is the inode on that device. 0 indicates that no inode is associated 393with the memory region, as the case would be with BSS (uninitialized data). 394The "pathname" shows the name associated file for this mapping. If the mapping 395is not associated with a file: 396 397 [heap] = the heap of the program 398 [stack] = the stack of the main process 399 [vdso] = the "virtual dynamic shared object", 400 the kernel system call handler 401 402 or if empty, the mapping is anonymous. 403 404The /proc/PID/smaps is an extension based on maps, showing the memory 405consumption for each of the process's mappings. For each of mappings there 406is a series of lines such as the following: 407 40808048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash 409Size: 1084 kB 410Rss: 892 kB 411Pss: 374 kB 412Shared_Clean: 892 kB 413Shared_Dirty: 0 kB 414Private_Clean: 0 kB 415Private_Dirty: 0 kB 416Referenced: 892 kB 417Anonymous: 0 kB 418LazyFree: 0 kB 419AnonHugePages: 0 kB 420ShmemPmdMapped: 0 kB 421Shared_Hugetlb: 0 kB 422Private_Hugetlb: 0 kB 423Swap: 0 kB 424SwapPss: 0 kB 425KernelPageSize: 4 kB 426MMUPageSize: 4 kB 427Locked: 0 kB 428THPeligible: 0 429VmFlags: rd ex mr mw me dw 430 431the first of these lines shows the same information as is displayed for the 432mapping in /proc/PID/maps. The remaining lines show the size of the mapping 433(size), the amount of the mapping that is currently resident in RAM (RSS), the 434process' proportional share of this mapping (PSS), the number of clean and 435dirty private pages in the mapping. 436 437The "proportional set size" (PSS) of a process is the count of pages it has 438in memory, where each page is divided by the number of processes sharing it. 439So if a process has 1000 pages all to itself, and 1000 shared with one other 440process, its PSS will be 1500. 441Note that even a page which is part of a MAP_SHARED mapping, but has only 442a single pte mapped, i.e. is currently used by only one process, is accounted 443as private and not as shared. 444"Referenced" indicates the amount of memory currently marked as referenced or 445accessed. 446"Anonymous" shows the amount of memory that does not belong to any file. Even 447a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE 448and a page is modified, the file page is replaced by a private anonymous copy. 449"LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE). 450The memory isn't freed immediately with madvise(). It's freed in memory 451pressure if the memory is clean. Please note that the printed value might 452be lower than the real value due to optimizations used in the current 453implementation. If this is not desirable please file a bug report. 454"AnonHugePages" shows the ammount of memory backed by transparent hugepage. 455"ShmemPmdMapped" shows the ammount of shared (shmem/tmpfs) memory backed by 456huge pages. 457"Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by 458hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical 459reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field. 460"Swap" shows how much would-be-anonymous memory is also used, but out on swap. 461For shmem mappings, "Swap" includes also the size of the mapped (and not 462replaced by copy-on-write) part of the underlying shmem object out on swap. 463"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this 464does not take into account swapped out page of underlying shmem objects. 465"Locked" indicates whether the mapping is locked in memory or not. 466"THPeligible" indicates whether the mapping is eligible for THP pages - 1 if 467true, 0 otherwise. 468 469"VmFlags" field deserves a separate description. This member represents the kernel 470flags associated with the particular virtual memory area in two letter encoded 471manner. The codes are the following: 472 rd - readable 473 wr - writeable 474 ex - executable 475 sh - shared 476 mr - may read 477 mw - may write 478 me - may execute 479 ms - may share 480 gd - stack segment growns down 481 pf - pure PFN range 482 dw - disabled write to the mapped file 483 lo - pages are locked in memory 484 io - memory mapped I/O area 485 sr - sequential read advise provided 486 rr - random read advise provided 487 dc - do not copy area on fork 488 de - do not expand area on remapping 489 ac - area is accountable 490 nr - swap space is not reserved for the area 491 ht - area uses huge tlb pages 492 ar - architecture specific flag 493 dd - do not include area into core dump 494 sd - soft-dirty flag 495 mm - mixed map area 496 hg - huge page advise flag 497 nh - no-huge page advise flag 498 mg - mergable advise flag 499 500Note that there is no guarantee that every flag and associated mnemonic will 501be present in all further kernel releases. Things get changed, the flags may 502be vanished or the reverse -- new added. Interpretation of their meaning 503might change in future as well. So each consumer of these flags has to 504follow each specific kernel version for the exact semantic. 505 506This file is only present if the CONFIG_MMU kernel configuration option is 507enabled. 508 509Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent 510output can be achieved only in the single read call). 511This typically manifests when doing partial reads of these files while the 512memory map is being modified. Despite the races, we do provide the following 513guarantees: 514 5151) The mapped addresses never go backwards, which implies no two 516 regions will ever overlap. 5172) If there is something at a given vaddr during the entirety of the 518 life of the smaps/maps walk, there will be some output for it. 519 520 521The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG 522bits on both physical and virtual pages associated with a process, and the 523soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst 524for details). 525To clear the bits for all the pages associated with the process 526 > echo 1 > /proc/PID/clear_refs 527 528To clear the bits for the anonymous pages associated with the process 529 > echo 2 > /proc/PID/clear_refs 530 531To clear the bits for the file mapped pages associated with the process 532 > echo 3 > /proc/PID/clear_refs 533 534To clear the soft-dirty bit 535 > echo 4 > /proc/PID/clear_refs 536 537To reset the peak resident set size ("high water mark") to the process's 538current value: 539 > echo 5 > /proc/PID/clear_refs 540 541Any other value written to /proc/PID/clear_refs will have no effect. 542 543The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags 544using /proc/kpageflags and number of times a page is mapped using 545/proc/kpagecount. For detailed explanation, see 546Documentation/admin-guide/mm/pagemap.rst. 547 548The /proc/pid/numa_maps is an extension based on maps, showing the memory 549locality and binding policy, as well as the memory usage (in pages) of 550each mapping. The output follows a general format where mapping details get 551summarized separated by blank spaces, one mapping per each file line: 552 553address policy mapping details 554 55500400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4 55600600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4 5573206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4 558320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 5593206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 5603206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4 5613206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4 562320698b000 default file=/lib64/libc-2.12.so 5633206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4 5643206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 5653206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4 5667f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4 5677f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4 5687f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048 5697fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4 5707fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4 571 572Where: 573"address" is the starting address for the mapping; 574"policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst); 575"mapping details" summarizes mapping data such as mapping type, page usage counters, 576node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page 577size, in KB, that is backing the mapping up. 578 5791.2 Kernel data 580--------------- 581 582Similar to the process entries, the kernel data files give information about 583the running kernel. The files used to obtain this information are contained in 584/proc and are listed in Table 1-5. Not all of these will be present in your 585system. It depends on the kernel configuration and the loaded modules, which 586files are there, and which are missing. 587 588Table 1-5: Kernel info in /proc 589.............................................................................. 590 File Content 591 apm Advanced power management info 592 buddyinfo Kernel memory allocator information (see text) (2.5) 593 bus Directory containing bus specific information 594 cmdline Kernel command line 595 cpuinfo Info about the CPU 596 devices Available devices (block and character) 597 dma Used DMS channels 598 filesystems Supported filesystems 599 driver Various drivers grouped here, currently rtc (2.4) 600 execdomains Execdomains, related to security (2.4) 601 fb Frame Buffer devices (2.4) 602 fs File system parameters, currently nfs/exports (2.4) 603 ide Directory containing info about the IDE subsystem 604 interrupts Interrupt usage 605 iomem Memory map (2.4) 606 ioports I/O port usage 607 irq Masks for irq to cpu affinity (2.4)(smp?) 608 isapnp ISA PnP (Plug&Play) Info (2.4) 609 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4)) 610 kmsg Kernel messages 611 ksyms Kernel symbol table 612 loadavg Load average of last 1, 5 & 15 minutes 613 locks Kernel locks 614 meminfo Memory info 615 misc Miscellaneous 616 modules List of loaded modules 617 mounts Mounted filesystems 618 net Networking info (see text) 619 pagetypeinfo Additional page allocator information (see text) (2.5) 620 partitions Table of partitions known to the system 621 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/, 622 decoupled by lspci (2.4) 623 rtc Real time clock 624 scsi SCSI info (see text) 625 slabinfo Slab pool info 626 softirqs softirq usage 627 stat Overall statistics 628 swaps Swap space utilization 629 sys See chapter 2 630 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4) 631 tty Info of tty drivers 632 uptime Wall clock since boot, combined idle time of all cpus 633 version Kernel version 634 video bttv info of video resources (2.4) 635 vmallocinfo Show vmalloced areas 636.............................................................................. 637 638You can, for example, check which interrupts are currently in use and what 639they are used for by looking in the file /proc/interrupts: 640 641 > cat /proc/interrupts 642 CPU0 643 0: 8728810 XT-PIC timer 644 1: 895 XT-PIC keyboard 645 2: 0 XT-PIC cascade 646 3: 531695 XT-PIC aha152x 647 4: 2014133 XT-PIC serial 648 5: 44401 XT-PIC pcnet_cs 649 8: 2 XT-PIC rtc 650 11: 8 XT-PIC i82365 651 12: 182918 XT-PIC PS/2 Mouse 652 13: 1 XT-PIC fpu 653 14: 1232265 XT-PIC ide0 654 15: 7 XT-PIC ide1 655 NMI: 0 656 657In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the 658output of a SMP machine): 659 660 > cat /proc/interrupts 661 662 CPU0 CPU1 663 0: 1243498 1214548 IO-APIC-edge timer 664 1: 8949 8958 IO-APIC-edge keyboard 665 2: 0 0 XT-PIC cascade 666 5: 11286 10161 IO-APIC-edge soundblaster 667 8: 1 0 IO-APIC-edge rtc 668 9: 27422 27407 IO-APIC-edge 3c503 669 12: 113645 113873 IO-APIC-edge PS/2 Mouse 670 13: 0 0 XT-PIC fpu 671 14: 22491 24012 IO-APIC-edge ide0 672 15: 2183 2415 IO-APIC-edge ide1 673 17: 30564 30414 IO-APIC-level eth0 674 18: 177 164 IO-APIC-level bttv 675 NMI: 2457961 2457959 676 LOC: 2457882 2457881 677 ERR: 2155 678 679NMI is incremented in this case because every timer interrupt generates a NMI 680(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups. 681 682LOC is the local interrupt counter of the internal APIC of every CPU. 683 684ERR is incremented in the case of errors in the IO-APIC bus (the bus that 685connects the CPUs in a SMP system. This means that an error has been detected, 686the IO-APIC automatically retry the transmission, so it should not be a big 687problem, but you should read the SMP-FAQ. 688 689In 2.6.2* /proc/interrupts was expanded again. This time the goal was for 690/proc/interrupts to display every IRQ vector in use by the system, not 691just those considered 'most important'. The new vectors are: 692 693 THR -- interrupt raised when a machine check threshold counter 694 (typically counting ECC corrected errors of memory or cache) exceeds 695 a configurable threshold. Only available on some systems. 696 697 TRM -- a thermal event interrupt occurs when a temperature threshold 698 has been exceeded for the CPU. This interrupt may also be generated 699 when the temperature drops back to normal. 700 701 SPU -- a spurious interrupt is some interrupt that was raised then lowered 702 by some IO device before it could be fully processed by the APIC. Hence 703 the APIC sees the interrupt but does not know what device it came from. 704 For this case the APIC will generate the interrupt with a IRQ vector 705 of 0xff. This might also be generated by chipset bugs. 706 707 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are 708 sent from one CPU to another per the needs of the OS. Typically, 709 their statistics are used by kernel developers and interested users to 710 determine the occurrence of interrupts of the given type. 711 712The above IRQ vectors are displayed only when relevant. For example, 713the threshold vector does not exist on x86_64 platforms. Others are 714suppressed when the system is a uniprocessor. As of this writing, only 715i386 and x86_64 platforms support the new IRQ vector displays. 716 717Of some interest is the introduction of the /proc/irq directory to 2.4. 718It could be used to set IRQ to CPU affinity, this means that you can "hook" an 719IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the 720irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and 721prof_cpu_mask. 722 723For example 724 > ls /proc/irq/ 725 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask 726 1 11 13 15 17 19 3 5 7 9 default_smp_affinity 727 > ls /proc/irq/0/ 728 smp_affinity 729 730smp_affinity is a bitmask, in which you can specify which CPUs can handle the 731IRQ, you can set it by doing: 732 733 > echo 1 > /proc/irq/10/smp_affinity 734 735This means that only the first CPU will handle the IRQ, but you can also echo 7365 which means that only the first and third CPU can handle the IRQ. 737 738The contents of each smp_affinity file is the same by default: 739 740 > cat /proc/irq/0/smp_affinity 741 ffffffff 742 743There is an alternate interface, smp_affinity_list which allows specifying 744a cpu range instead of a bitmask: 745 746 > cat /proc/irq/0/smp_affinity_list 747 1024-1031 748 749The default_smp_affinity mask applies to all non-active IRQs, which are the 750IRQs which have not yet been allocated/activated, and hence which lack a 751/proc/irq/[0-9]* directory. 752 753The node file on an SMP system shows the node to which the device using the IRQ 754reports itself as being attached. This hardware locality information does not 755include information about any possible driver locality preference. 756 757prof_cpu_mask specifies which CPUs are to be profiled by the system wide 758profiler. Default value is ffffffff (all cpus if there are only 32 of them). 759 760The way IRQs are routed is handled by the IO-APIC, and it's Round Robin 761between all the CPUs which are allowed to handle it. As usual the kernel has 762more info than you and does a better job than you, so the defaults are the 763best choice for almost everyone. [Note this applies only to those IO-APIC's 764that support "Round Robin" interrupt distribution.] 765 766There are three more important subdirectories in /proc: net, scsi, and sys. 767The general rule is that the contents, or even the existence of these 768directories, depend on your kernel configuration. If SCSI is not enabled, the 769directory scsi may not exist. The same is true with the net, which is there 770only when networking support is present in the running kernel. 771 772The slabinfo file gives information about memory usage at the slab level. 773Linux uses slab pools for memory management above page level in version 2.2. 774Commonly used objects have their own slab pool (such as network buffers, 775directory cache, and so on). 776 777.............................................................................. 778 779> cat /proc/buddyinfo 780 781Node 0, zone DMA 0 4 5 4 4 3 ... 782Node 0, zone Normal 1 0 0 1 101 8 ... 783Node 0, zone HighMem 2 0 0 1 1 0 ... 784 785External fragmentation is a problem under some workloads, and buddyinfo is a 786useful tool for helping diagnose these problems. Buddyinfo will give you a 787clue as to how big an area you can safely allocate, or why a previous 788allocation failed. 789 790Each column represents the number of pages of a certain order which are 791available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in 792ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE 793available in ZONE_NORMAL, etc... 794 795More information relevant to external fragmentation can be found in 796pagetypeinfo. 797 798> cat /proc/pagetypeinfo 799Page block order: 9 800Pages per block: 512 801 802Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10 803Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0 804Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0 805Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2 806Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0 807Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0 808Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9 809Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0 810Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452 811Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0 812Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0 813 814Number of blocks type Unmovable Reclaimable Movable Reserve Isolate 815Node 0, zone DMA 2 0 5 1 0 816Node 0, zone DMA32 41 6 967 2 0 817 818Fragmentation avoidance in the kernel works by grouping pages of different 819migrate types into the same contiguous regions of memory called page blocks. 820A page block is typically the size of the default hugepage size e.g. 2MB on 821X86-64. By keeping pages grouped based on their ability to move, the kernel 822can reclaim pages within a page block to satisfy a high-order allocation. 823 824The pagetypinfo begins with information on the size of a page block. It 825then gives the same type of information as buddyinfo except broken down 826by migrate-type and finishes with details on how many page blocks of each 827type exist. 828 829If min_free_kbytes has been tuned correctly (recommendations made by hugeadm 830from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can 831make an estimate of the likely number of huge pages that can be allocated 832at a given point in time. All the "Movable" blocks should be allocatable 833unless memory has been mlock()'d. Some of the Reclaimable blocks should 834also be allocatable although a lot of filesystem metadata may have to be 835reclaimed to achieve this. 836 837.............................................................................. 838 839meminfo: 840 841Provides information about distribution and utilization of memory. This 842varies by architecture and compile options. The following is from a 84316GB PIII, which has highmem enabled. You may not have all of these fields. 844 845> cat /proc/meminfo 846 847MemTotal: 16344972 kB 848MemFree: 13634064 kB 849MemAvailable: 14836172 kB 850Buffers: 3656 kB 851Cached: 1195708 kB 852SwapCached: 0 kB 853Active: 891636 kB 854Inactive: 1077224 kB 855HighTotal: 15597528 kB 856HighFree: 13629632 kB 857LowTotal: 747444 kB 858LowFree: 4432 kB 859SwapTotal: 0 kB 860SwapFree: 0 kB 861Dirty: 968 kB 862Writeback: 0 kB 863AnonPages: 861800 kB 864Mapped: 280372 kB 865Shmem: 644 kB 866Slab: 284364 kB 867SReclaimable: 159856 kB 868SUnreclaim: 124508 kB 869PageTables: 24448 kB 870NFS_Unstable: 0 kB 871Bounce: 0 kB 872WritebackTmp: 0 kB 873CommitLimit: 7669796 kB 874Committed_AS: 100056 kB 875VmallocTotal: 112216 kB 876VmallocUsed: 428 kB 877VmallocChunk: 111088 kB 878Percpu: 62080 kB 879HardwareCorrupted: 0 kB 880AnonHugePages: 49152 kB 881ShmemHugePages: 0 kB 882ShmemPmdMapped: 0 kB 883 884 885 MemTotal: Total usable ram (i.e. physical ram minus a few reserved 886 bits and the kernel binary code) 887 MemFree: The sum of LowFree+HighFree 888MemAvailable: An estimate of how much memory is available for starting new 889 applications, without swapping. Calculated from MemFree, 890 SReclaimable, the size of the file LRU lists, and the low 891 watermarks in each zone. 892 The estimate takes into account that the system needs some 893 page cache to function well, and that not all reclaimable 894 slab will be reclaimable, due to items being in use. The 895 impact of those factors will vary from system to system. 896 Buffers: Relatively temporary storage for raw disk blocks 897 shouldn't get tremendously large (20MB or so) 898 Cached: in-memory cache for files read from the disk (the 899 pagecache). Doesn't include SwapCached 900 SwapCached: Memory that once was swapped out, is swapped back in but 901 still also is in the swapfile (if memory is needed it 902 doesn't need to be swapped out AGAIN because it is already 903 in the swapfile. This saves I/O) 904 Active: Memory that has been used more recently and usually not 905 reclaimed unless absolutely necessary. 906 Inactive: Memory which has been less recently used. It is more 907 eligible to be reclaimed for other purposes 908 HighTotal: 909 HighFree: Highmem is all memory above ~860MB of physical memory 910 Highmem areas are for use by userspace programs, or 911 for the pagecache. The kernel must use tricks to access 912 this memory, making it slower to access than lowmem. 913 LowTotal: 914 LowFree: Lowmem is memory which can be used for everything that 915 highmem can be used for, but it is also available for the 916 kernel's use for its own data structures. Among many 917 other things, it is where everything from the Slab is 918 allocated. Bad things happen when you're out of lowmem. 919 SwapTotal: total amount of swap space available 920 SwapFree: Memory which has been evicted from RAM, and is temporarily 921 on the disk 922 Dirty: Memory which is waiting to get written back to the disk 923 Writeback: Memory which is actively being written back to the disk 924 AnonPages: Non-file backed pages mapped into userspace page tables 925HardwareCorrupted: The amount of RAM/memory in KB, the kernel identifies as 926 corrupted. 927AnonHugePages: Non-file backed huge pages mapped into userspace page tables 928 Mapped: files which have been mmaped, such as libraries 929 Shmem: Total memory used by shared memory (shmem) and tmpfs 930ShmemHugePages: Memory used by shared memory (shmem) and tmpfs allocated 931 with huge pages 932ShmemPmdMapped: Shared memory mapped into userspace with huge pages 933 Slab: in-kernel data structures cache 934SReclaimable: Part of Slab, that might be reclaimed, such as caches 935 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure 936 PageTables: amount of memory dedicated to the lowest level of page 937 tables. 938NFS_Unstable: NFS pages sent to the server, but not yet committed to stable 939 storage 940 Bounce: Memory used for block device "bounce buffers" 941WritebackTmp: Memory used by FUSE for temporary writeback buffers 942 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'), 943 this is the total amount of memory currently available to 944 be allocated on the system. This limit is only adhered to 945 if strict overcommit accounting is enabled (mode 2 in 946 'vm.overcommit_memory'). 947 The CommitLimit is calculated with the following formula: 948 CommitLimit = ([total RAM pages] - [total huge TLB pages]) * 949 overcommit_ratio / 100 + [total swap pages] 950 For example, on a system with 1G of physical RAM and 7G 951 of swap with a `vm.overcommit_ratio` of 30 it would 952 yield a CommitLimit of 7.3G. 953 For more details, see the memory overcommit documentation 954 in vm/overcommit-accounting. 955Committed_AS: The amount of memory presently allocated on the system. 956 The committed memory is a sum of all of the memory which 957 has been allocated by processes, even if it has not been 958 "used" by them as of yet. A process which malloc()'s 1G 959 of memory, but only touches 300M of it will show up as 960 using 1G. This 1G is memory which has been "committed" to 961 by the VM and can be used at any time by the allocating 962 application. With strict overcommit enabled on the system 963 (mode 2 in 'vm.overcommit_memory'),allocations which would 964 exceed the CommitLimit (detailed above) will not be permitted. 965 This is useful if one needs to guarantee that processes will 966 not fail due to lack of memory once that memory has been 967 successfully allocated. 968VmallocTotal: total size of vmalloc memory area 969 VmallocUsed: amount of vmalloc area which is used 970VmallocChunk: largest contiguous block of vmalloc area which is free 971 Percpu: Memory allocated to the percpu allocator used to back percpu 972 allocations. This stat excludes the cost of metadata. 973 974.............................................................................. 975 976vmallocinfo: 977 978Provides information about vmalloced/vmaped areas. One line per area, 979containing the virtual address range of the area, size in bytes, 980caller information of the creator, and optional information depending 981on the kind of area : 982 983 pages=nr number of pages 984 phys=addr if a physical address was specified 985 ioremap I/O mapping (ioremap() and friends) 986 vmalloc vmalloc() area 987 vmap vmap()ed pages 988 user VM_USERMAP area 989 vpages buffer for pages pointers was vmalloced (huge area) 990 N<node>=nr (Only on NUMA kernels) 991 Number of pages allocated on memory node <node> 992 993> cat /proc/vmallocinfo 9940xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ... 995 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128 9960xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ... 997 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64 9980xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f... 999 phys=7fee8000 ioremap 10000xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f... 1001 phys=7fee7000 ioremap 10020xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210 10030xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ... 1004 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3 10050xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ... 1006 pages=2 vmalloc N1=2 10070xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ... 1008 /0x130 [x_tables] pages=4 vmalloc N0=4 10090xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ... 1010 pages=14 vmalloc N2=14 10110xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ... 1012 pages=4 vmalloc N1=4 10130xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ... 1014 pages=2 vmalloc N1=2 10150xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ... 1016 pages=10 vmalloc N0=10 1017 1018.............................................................................. 1019 1020softirqs: 1021 1022Provides counts of softirq handlers serviced since boot time, for each cpu. 1023 1024> cat /proc/softirqs 1025 CPU0 CPU1 CPU2 CPU3 1026 HI: 0 0 0 0 1027 TIMER: 27166 27120 27097 27034 1028 NET_TX: 0 0 0 17 1029 NET_RX: 42 0 0 39 1030 BLOCK: 0 0 107 1121 1031 TASKLET: 0 0 0 290 1032 SCHED: 27035 26983 26971 26746 1033 HRTIMER: 0 0 0 0 1034 RCU: 1678 1769 2178 2250 1035 1036 10371.3 IDE devices in /proc/ide 1038---------------------------- 1039 1040The subdirectory /proc/ide contains information about all IDE devices of which 1041the kernel is aware. There is one subdirectory for each IDE controller, the 1042file drivers and a link for each IDE device, pointing to the device directory 1043in the controller specific subtree. 1044 1045The file drivers contains general information about the drivers used for the 1046IDE devices: 1047 1048 > cat /proc/ide/drivers 1049 ide-cdrom version 4.53 1050 ide-disk version 1.08 1051 1052More detailed information can be found in the controller specific 1053subdirectories. These are named ide0, ide1 and so on. Each of these 1054directories contains the files shown in table 1-6. 1055 1056 1057Table 1-6: IDE controller info in /proc/ide/ide? 1058.............................................................................. 1059 File Content 1060 channel IDE channel (0 or 1) 1061 config Configuration (only for PCI/IDE bridge) 1062 mate Mate name 1063 model Type/Chipset of IDE controller 1064.............................................................................. 1065 1066Each device connected to a controller has a separate subdirectory in the 1067controllers directory. The files listed in table 1-7 are contained in these 1068directories. 1069 1070 1071Table 1-7: IDE device information 1072.............................................................................. 1073 File Content 1074 cache The cache 1075 capacity Capacity of the medium (in 512Byte blocks) 1076 driver driver and version 1077 geometry physical and logical geometry 1078 identify device identify block 1079 media media type 1080 model device identifier 1081 settings device setup 1082 smart_thresholds IDE disk management thresholds 1083 smart_values IDE disk management values 1084.............................................................................. 1085 1086The most interesting file is settings. This file contains a nice overview of 1087the drive parameters: 1088 1089 # cat /proc/ide/ide0/hda/settings 1090 name value min max mode 1091 ---- ----- --- --- ---- 1092 bios_cyl 526 0 65535 rw 1093 bios_head 255 0 255 rw 1094 bios_sect 63 0 63 rw 1095 breada_readahead 4 0 127 rw 1096 bswap 0 0 1 r 1097 file_readahead 72 0 2097151 rw 1098 io_32bit 0 0 3 rw 1099 keepsettings 0 0 1 rw 1100 max_kb_per_request 122 1 127 rw 1101 multcount 0 0 8 rw 1102 nice1 1 0 1 rw 1103 nowerr 0 0 1 rw 1104 pio_mode write-only 0 255 w 1105 slow 0 0 1 rw 1106 unmaskirq 0 0 1 rw 1107 using_dma 0 0 1 rw 1108 1109 11101.4 Networking info in /proc/net 1111-------------------------------- 1112 1113The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the 1114additional values you get for IP version 6 if you configure the kernel to 1115support this. Table 1-9 lists the files and their meaning. 1116 1117 1118Table 1-8: IPv6 info in /proc/net 1119.............................................................................. 1120 File Content 1121 udp6 UDP sockets (IPv6) 1122 tcp6 TCP sockets (IPv6) 1123 raw6 Raw device statistics (IPv6) 1124 igmp6 IP multicast addresses, which this host joined (IPv6) 1125 if_inet6 List of IPv6 interface addresses 1126 ipv6_route Kernel routing table for IPv6 1127 rt6_stats Global IPv6 routing tables statistics 1128 sockstat6 Socket statistics (IPv6) 1129 snmp6 Snmp data (IPv6) 1130.............................................................................. 1131 1132 1133Table 1-9: Network info in /proc/net 1134.............................................................................. 1135 File Content 1136 arp Kernel ARP table 1137 dev network devices with statistics 1138 dev_mcast the Layer2 multicast groups a device is listening too 1139 (interface index, label, number of references, number of bound 1140 addresses). 1141 dev_stat network device status 1142 ip_fwchains Firewall chain linkage 1143 ip_fwnames Firewall chain names 1144 ip_masq Directory containing the masquerading tables 1145 ip_masquerade Major masquerading table 1146 netstat Network statistics 1147 raw raw device statistics 1148 route Kernel routing table 1149 rpc Directory containing rpc info 1150 rt_cache Routing cache 1151 snmp SNMP data 1152 sockstat Socket statistics 1153 tcp TCP sockets 1154 udp UDP sockets 1155 unix UNIX domain sockets 1156 wireless Wireless interface data (Wavelan etc) 1157 igmp IP multicast addresses, which this host joined 1158 psched Global packet scheduler parameters. 1159 netlink List of PF_NETLINK sockets 1160 ip_mr_vifs List of multicast virtual interfaces 1161 ip_mr_cache List of multicast routing cache 1162.............................................................................. 1163 1164You can use this information to see which network devices are available in 1165your system and how much traffic was routed over those devices: 1166 1167 > cat /proc/net/dev 1168 Inter-|Receive |[... 1169 face |bytes packets errs drop fifo frame compressed multicast|[... 1170 lo: 908188 5596 0 0 0 0 0 0 [... 1171 ppp0:15475140 20721 410 0 0 410 0 0 [... 1172 eth0: 614530 7085 0 0 0 0 0 1 [... 1173 1174 ...] Transmit 1175 ...] bytes packets errs drop fifo colls carrier compressed 1176 ...] 908188 5596 0 0 0 0 0 0 1177 ...] 1375103 17405 0 0 0 0 0 0 1178 ...] 1703981 5535 0 0 0 3 0 0 1179 1180In addition, each Channel Bond interface has its own directory. For 1181example, the bond0 device will have a directory called /proc/net/bond0/. 1182It will contain information that is specific to that bond, such as the 1183current slaves of the bond, the link status of the slaves, and how 1184many times the slaves link has failed. 1185 11861.5 SCSI info 1187------------- 1188 1189If you have a SCSI host adapter in your system, you'll find a subdirectory 1190named after the driver for this adapter in /proc/scsi. You'll also see a list 1191of all recognized SCSI devices in /proc/scsi: 1192 1193 >cat /proc/scsi/scsi 1194 Attached devices: 1195 Host: scsi0 Channel: 00 Id: 00 Lun: 00 1196 Vendor: IBM Model: DGHS09U Rev: 03E0 1197 Type: Direct-Access ANSI SCSI revision: 03 1198 Host: scsi0 Channel: 00 Id: 06 Lun: 00 1199 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04 1200 Type: CD-ROM ANSI SCSI revision: 02 1201 1202 1203The directory named after the driver has one file for each adapter found in 1204the system. These files contain information about the controller, including 1205the used IRQ and the IO address range. The amount of information shown is 1206dependent on the adapter you use. The example shows the output for an Adaptec 1207AHA-2940 SCSI adapter: 1208 1209 > cat /proc/scsi/aic7xxx/0 1210 1211 Adaptec AIC7xxx driver version: 5.1.19/3.2.4 1212 Compile Options: 1213 TCQ Enabled By Default : Disabled 1214 AIC7XXX_PROC_STATS : Disabled 1215 AIC7XXX_RESET_DELAY : 5 1216 Adapter Configuration: 1217 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter 1218 Ultra Wide Controller 1219 PCI MMAPed I/O Base: 0xeb001000 1220 Adapter SEEPROM Config: SEEPROM found and used. 1221 Adaptec SCSI BIOS: Enabled 1222 IRQ: 10 1223 SCBs: Active 0, Max Active 2, 1224 Allocated 15, HW 16, Page 255 1225 Interrupts: 160328 1226 BIOS Control Word: 0x18b6 1227 Adapter Control Word: 0x005b 1228 Extended Translation: Enabled 1229 Disconnect Enable Flags: 0xffff 1230 Ultra Enable Flags: 0x0001 1231 Tag Queue Enable Flags: 0x0000 1232 Ordered Queue Tag Flags: 0x0000 1233 Default Tag Queue Depth: 8 1234 Tagged Queue By Device array for aic7xxx host instance 0: 1235 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255} 1236 Actual queue depth per device for aic7xxx host instance 0: 1237 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1} 1238 Statistics: 1239 (scsi0:0:0:0) 1240 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8 1241 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0) 1242 Total transfers 160151 (74577 reads and 85574 writes) 1243 (scsi0:0:6:0) 1244 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15 1245 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0) 1246 Total transfers 0 (0 reads and 0 writes) 1247 1248 12491.6 Parallel port info in /proc/parport 1250--------------------------------------- 1251 1252The directory /proc/parport contains information about the parallel ports of 1253your system. It has one subdirectory for each port, named after the port 1254number (0,1,2,...). 1255 1256These directories contain the four files shown in Table 1-10. 1257 1258 1259Table 1-10: Files in /proc/parport 1260.............................................................................. 1261 File Content 1262 autoprobe Any IEEE-1284 device ID information that has been acquired. 1263 devices list of the device drivers using that port. A + will appear by the 1264 name of the device currently using the port (it might not appear 1265 against any). 1266 hardware Parallel port's base address, IRQ line and DMA channel. 1267 irq IRQ that parport is using for that port. This is in a separate 1268 file to allow you to alter it by writing a new value in (IRQ 1269 number or none). 1270.............................................................................. 1271 12721.7 TTY info in /proc/tty 1273------------------------- 1274 1275Information about the available and actually used tty's can be found in the 1276directory /proc/tty.You'll find entries for drivers and line disciplines in 1277this directory, as shown in Table 1-11. 1278 1279 1280Table 1-11: Files in /proc/tty 1281.............................................................................. 1282 File Content 1283 drivers list of drivers and their usage 1284 ldiscs registered line disciplines 1285 driver/serial usage statistic and status of single tty lines 1286.............................................................................. 1287 1288To see which tty's are currently in use, you can simply look into the file 1289/proc/tty/drivers: 1290 1291 > cat /proc/tty/drivers 1292 pty_slave /dev/pts 136 0-255 pty:slave 1293 pty_master /dev/ptm 128 0-255 pty:master 1294 pty_slave /dev/ttyp 3 0-255 pty:slave 1295 pty_master /dev/pty 2 0-255 pty:master 1296 serial /dev/cua 5 64-67 serial:callout 1297 serial /dev/ttyS 4 64-67 serial 1298 /dev/tty0 /dev/tty0 4 0 system:vtmaster 1299 /dev/ptmx /dev/ptmx 5 2 system 1300 /dev/console /dev/console 5 1 system:console 1301 /dev/tty /dev/tty 5 0 system:/dev/tty 1302 unknown /dev/tty 4 1-63 console 1303 1304 13051.8 Miscellaneous kernel statistics in /proc/stat 1306------------------------------------------------- 1307 1308Various pieces of information about kernel activity are available in the 1309/proc/stat file. All of the numbers reported in this file are aggregates 1310since the system first booted. For a quick look, simply cat the file: 1311 1312 > cat /proc/stat 1313 cpu 2255 34 2290 22625563 6290 127 456 0 0 0 1314 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0 1315 cpu1 1123 0 849 11313845 2614 0 18 0 0 0 1316 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...] 1317 ctxt 1990473 1318 btime 1062191376 1319 processes 2915 1320 procs_running 1 1321 procs_blocked 0 1322 softirq 183433 0 21755 12 39 1137 231 21459 2263 1323 1324The very first "cpu" line aggregates the numbers in all of the other "cpuN" 1325lines. These numbers identify the amount of time the CPU has spent performing 1326different kinds of work. Time units are in USER_HZ (typically hundredths of a 1327second). The meanings of the columns are as follows, from left to right: 1328 1329- user: normal processes executing in user mode 1330- nice: niced processes executing in user mode 1331- system: processes executing in kernel mode 1332- idle: twiddling thumbs 1333- iowait: In a word, iowait stands for waiting for I/O to complete. But there 1334 are several problems: 1335 1. Cpu will not wait for I/O to complete, iowait is the time that a task is 1336 waiting for I/O to complete. When cpu goes into idle state for 1337 outstanding task io, another task will be scheduled on this CPU. 1338 2. In a multi-core CPU, the task waiting for I/O to complete is not running 1339 on any CPU, so the iowait of each CPU is difficult to calculate. 1340 3. The value of iowait field in /proc/stat will decrease in certain 1341 conditions. 1342 So, the iowait is not reliable by reading from /proc/stat. 1343- irq: servicing interrupts 1344- softirq: servicing softirqs 1345- steal: involuntary wait 1346- guest: running a normal guest 1347- guest_nice: running a niced guest 1348 1349The "intr" line gives counts of interrupts serviced since boot time, for each 1350of the possible system interrupts. The first column is the total of all 1351interrupts serviced including unnumbered architecture specific interrupts; 1352each subsequent column is the total for that particular numbered interrupt. 1353Unnumbered interrupts are not shown, only summed into the total. 1354 1355The "ctxt" line gives the total number of context switches across all CPUs. 1356 1357The "btime" line gives the time at which the system booted, in seconds since 1358the Unix epoch. 1359 1360The "processes" line gives the number of processes and threads created, which 1361includes (but is not limited to) those created by calls to the fork() and 1362clone() system calls. 1363 1364The "procs_running" line gives the total number of threads that are 1365running or ready to run (i.e., the total number of runnable threads). 1366 1367The "procs_blocked" line gives the number of processes currently blocked, 1368waiting for I/O to complete. 1369 1370The "softirq" line gives counts of softirqs serviced since boot time, for each 1371of the possible system softirqs. The first column is the total of all 1372softirqs serviced; each subsequent column is the total for that particular 1373softirq. 1374 1375 13761.9 Ext4 file system parameters 1377------------------------------- 1378 1379Information about mounted ext4 file systems can be found in 1380/proc/fs/ext4. Each mounted filesystem will have a directory in 1381/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or 1382/proc/fs/ext4/dm-0). The files in each per-device directory are shown 1383in Table 1-12, below. 1384 1385Table 1-12: Files in /proc/fs/ext4/<devname> 1386.............................................................................. 1387 File Content 1388 mb_groups details of multiblock allocator buddy cache of free blocks 1389.............................................................................. 1390 13912.0 /proc/consoles 1392------------------ 1393Shows registered system console lines. 1394 1395To see which character device lines are currently used for the system console 1396/dev/console, you may simply look into the file /proc/consoles: 1397 1398 > cat /proc/consoles 1399 tty0 -WU (ECp) 4:7 1400 ttyS0 -W- (Ep) 4:64 1401 1402The columns are: 1403 1404 device name of the device 1405 operations R = can do read operations 1406 W = can do write operations 1407 U = can do unblank 1408 flags E = it is enabled 1409 C = it is preferred console 1410 B = it is primary boot console 1411 p = it is used for printk buffer 1412 b = it is not a TTY but a Braille device 1413 a = it is safe to use when cpu is offline 1414 major:minor major and minor number of the device separated by a colon 1415 1416------------------------------------------------------------------------------ 1417Summary 1418------------------------------------------------------------------------------ 1419The /proc file system serves information about the running system. It not only 1420allows access to process data but also allows you to request the kernel status 1421by reading files in the hierarchy. 1422 1423The directory structure of /proc reflects the types of information and makes 1424it easy, if not obvious, where to look for specific data. 1425------------------------------------------------------------------------------ 1426 1427------------------------------------------------------------------------------ 1428CHAPTER 2: MODIFYING SYSTEM PARAMETERS 1429------------------------------------------------------------------------------ 1430 1431------------------------------------------------------------------------------ 1432In This Chapter 1433------------------------------------------------------------------------------ 1434* Modifying kernel parameters by writing into files found in /proc/sys 1435* Exploring the files which modify certain parameters 1436* Review of the /proc/sys file tree 1437------------------------------------------------------------------------------ 1438 1439 1440A very interesting part of /proc is the directory /proc/sys. This is not only 1441a source of information, it also allows you to change parameters within the 1442kernel. Be very careful when attempting this. You can optimize your system, 1443but you can also cause it to crash. Never alter kernel parameters on a 1444production system. Set up a development machine and test to make sure that 1445everything works the way you want it to. You may have no alternative but to 1446reboot the machine once an error has been made. 1447 1448To change a value, simply echo the new value into the file. An example is 1449given below in the section on the file system data. You need to be root to do 1450this. You can create your own boot script to perform this every time your 1451system boots. 1452 1453The files in /proc/sys can be used to fine tune and monitor miscellaneous and 1454general things in the operation of the Linux kernel. Since some of the files 1455can inadvertently disrupt your system, it is advisable to read both 1456documentation and source before actually making adjustments. In any case, be 1457very careful when writing to any of these files. The entries in /proc may 1458change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt 1459review the kernel documentation in the directory /usr/src/linux/Documentation. 1460This chapter is heavily based on the documentation included in the pre 2.2 1461kernels, and became part of it in version 2.2.1 of the Linux kernel. 1462 1463Please see: Documentation/sysctl/ directory for descriptions of these 1464entries. 1465 1466------------------------------------------------------------------------------ 1467Summary 1468------------------------------------------------------------------------------ 1469Certain aspects of kernel behavior can be modified at runtime, without the 1470need to recompile the kernel, or even to reboot the system. The files in the 1471/proc/sys tree can not only be read, but also modified. You can use the echo 1472command to write value into these files, thereby changing the default settings 1473of the kernel. 1474------------------------------------------------------------------------------ 1475 1476------------------------------------------------------------------------------ 1477CHAPTER 3: PER-PROCESS PARAMETERS 1478------------------------------------------------------------------------------ 1479 14803.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score 1481-------------------------------------------------------------------------------- 1482 1483These file can be used to adjust the badness heuristic used to select which 1484process gets killed in out of memory conditions. 1485 1486The badness heuristic assigns a value to each candidate task ranging from 0 1487(never kill) to 1000 (always kill) to determine which process is targeted. The 1488units are roughly a proportion along that range of allowed memory the process 1489may allocate from based on an estimation of its current memory and swap use. 1490For example, if a task is using all allowed memory, its badness score will be 14911000. If it is using half of its allowed memory, its score will be 500. 1492 1493There is an additional factor included in the badness score: the current memory 1494and swap usage is discounted by 3% for root processes. 1495 1496The amount of "allowed" memory depends on the context in which the oom killer 1497was called. If it is due to the memory assigned to the allocating task's cpuset 1498being exhausted, the allowed memory represents the set of mems assigned to that 1499cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed 1500memory represents the set of mempolicy nodes. If it is due to a memory 1501limit (or swap limit) being reached, the allowed memory is that configured 1502limit. Finally, if it is due to the entire system being out of memory, the 1503allowed memory represents all allocatable resources. 1504 1505The value of /proc/<pid>/oom_score_adj is added to the badness score before it 1506is used to determine which task to kill. Acceptable values range from -1000 1507(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to 1508polarize the preference for oom killing either by always preferring a certain 1509task or completely disabling it. The lowest possible value, -1000, is 1510equivalent to disabling oom killing entirely for that task since it will always 1511report a badness score of 0. 1512 1513Consequently, it is very simple for userspace to define the amount of memory to 1514consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for 1515example, is roughly equivalent to allowing the remainder of tasks sharing the 1516same system, cpuset, mempolicy, or memory controller resources to use at least 151750% more memory. A value of -500, on the other hand, would be roughly 1518equivalent to discounting 50% of the task's allowed memory from being considered 1519as scoring against the task. 1520 1521For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also 1522be used to tune the badness score. Its acceptable values range from -16 1523(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17 1524(OOM_DISABLE) to disable oom killing entirely for that task. Its value is 1525scaled linearly with /proc/<pid>/oom_score_adj. 1526 1527The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last 1528value set by a CAP_SYS_RESOURCE process. To reduce the value any lower 1529requires CAP_SYS_RESOURCE. 1530 1531Caveat: when a parent task is selected, the oom killer will sacrifice any first 1532generation children with separate address spaces instead, if possible. This 1533avoids servers and important system daemons from being killed and loses the 1534minimal amount of work. 1535 1536 15373.2 /proc/<pid>/oom_score - Display current oom-killer score 1538------------------------------------------------------------- 1539 1540This file can be used to check the current score used by the oom-killer is for 1541any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which 1542process should be killed in an out-of-memory situation. 1543 1544 15453.3 /proc/<pid>/io - Display the IO accounting fields 1546------------------------------------------------------- 1547 1548This file contains IO statistics for each running process 1549 1550Example 1551------- 1552 1553test:/tmp # dd if=/dev/zero of=/tmp/test.dat & 1554[1] 3828 1555 1556test:/tmp # cat /proc/3828/io 1557rchar: 323934931 1558wchar: 323929600 1559syscr: 632687 1560syscw: 632675 1561read_bytes: 0 1562write_bytes: 323932160 1563cancelled_write_bytes: 0 1564 1565 1566Description 1567----------- 1568 1569rchar 1570----- 1571 1572I/O counter: chars read 1573The number of bytes which this task has caused to be read from storage. This 1574is simply the sum of bytes which this process passed to read() and pread(). 1575It includes things like tty IO and it is unaffected by whether or not actual 1576physical disk IO was required (the read might have been satisfied from 1577pagecache) 1578 1579 1580wchar 1581----- 1582 1583I/O counter: chars written 1584The number of bytes which this task has caused, or shall cause to be written 1585to disk. Similar caveats apply here as with rchar. 1586 1587 1588syscr 1589----- 1590 1591I/O counter: read syscalls 1592Attempt to count the number of read I/O operations, i.e. syscalls like read() 1593and pread(). 1594 1595 1596syscw 1597----- 1598 1599I/O counter: write syscalls 1600Attempt to count the number of write I/O operations, i.e. syscalls like 1601write() and pwrite(). 1602 1603 1604read_bytes 1605---------- 1606 1607I/O counter: bytes read 1608Attempt to count the number of bytes which this process really did cause to 1609be fetched from the storage layer. Done at the submit_bio() level, so it is 1610accurate for block-backed filesystems. <please add status regarding NFS and 1611CIFS at a later time> 1612 1613 1614write_bytes 1615----------- 1616 1617I/O counter: bytes written 1618Attempt to count the number of bytes which this process caused to be sent to 1619the storage layer. This is done at page-dirtying time. 1620 1621 1622cancelled_write_bytes 1623--------------------- 1624 1625The big inaccuracy here is truncate. If a process writes 1MB to a file and 1626then deletes the file, it will in fact perform no writeout. But it will have 1627been accounted as having caused 1MB of write. 1628In other words: The number of bytes which this process caused to not happen, 1629by truncating pagecache. A task can cause "negative" IO too. If this task 1630truncates some dirty pagecache, some IO which another task has been accounted 1631for (in its write_bytes) will not be happening. We _could_ just subtract that 1632from the truncating task's write_bytes, but there is information loss in doing 1633that. 1634 1635 1636Note 1637---- 1638 1639At its current implementation state, this is a bit racy on 32-bit machines: if 1640process A reads process B's /proc/pid/io while process B is updating one of 1641those 64-bit counters, process A could see an intermediate result. 1642 1643 1644More information about this can be found within the taskstats documentation in 1645Documentation/accounting. 1646 16473.4 /proc/<pid>/coredump_filter - Core dump filtering settings 1648--------------------------------------------------------------- 1649When a process is dumped, all anonymous memory is written to a core file as 1650long as the size of the core file isn't limited. But sometimes we don't want 1651to dump some memory segments, for example, huge shared memory or DAX. 1652Conversely, sometimes we want to save file-backed memory segments into a core 1653file, not only the individual files. 1654 1655/proc/<pid>/coredump_filter allows you to customize which memory segments 1656will be dumped when the <pid> process is dumped. coredump_filter is a bitmask 1657of memory types. If a bit of the bitmask is set, memory segments of the 1658corresponding memory type are dumped, otherwise they are not dumped. 1659 1660The following 9 memory types are supported: 1661 - (bit 0) anonymous private memory 1662 - (bit 1) anonymous shared memory 1663 - (bit 2) file-backed private memory 1664 - (bit 3) file-backed shared memory 1665 - (bit 4) ELF header pages in file-backed private memory areas (it is 1666 effective only if the bit 2 is cleared) 1667 - (bit 5) hugetlb private memory 1668 - (bit 6) hugetlb shared memory 1669 - (bit 7) DAX private memory 1670 - (bit 8) DAX shared memory 1671 1672 Note that MMIO pages such as frame buffer are never dumped and vDSO pages 1673 are always dumped regardless of the bitmask status. 1674 1675 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is 1676 only affected by bit 5-6, and DAX is only affected by bits 7-8. 1677 1678The default value of coredump_filter is 0x33; this means all anonymous memory 1679segments, ELF header pages and hugetlb private memory are dumped. 1680 1681If you don't want to dump all shared memory segments attached to pid 1234, 1682write 0x31 to the process's proc file. 1683 1684 $ echo 0x31 > /proc/1234/coredump_filter 1685 1686When a new process is created, the process inherits the bitmask status from its 1687parent. It is useful to set up coredump_filter before the program runs. 1688For example: 1689 1690 $ echo 0x7 > /proc/self/coredump_filter 1691 $ ./some_program 1692 16933.5 /proc/<pid>/mountinfo - Information about mounts 1694-------------------------------------------------------- 1695 1696This file contains lines of the form: 1697 169836 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue 1699(1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11) 1700 1701(1) mount ID: unique identifier of the mount (may be reused after umount) 1702(2) parent ID: ID of parent (or of self for the top of the mount tree) 1703(3) major:minor: value of st_dev for files on filesystem 1704(4) root: root of the mount within the filesystem 1705(5) mount point: mount point relative to the process's root 1706(6) mount options: per mount options 1707(7) optional fields: zero or more fields of the form "tag[:value]" 1708(8) separator: marks the end of the optional fields 1709(9) filesystem type: name of filesystem of the form "type[.subtype]" 1710(10) mount source: filesystem specific information or "none" 1711(11) super options: per super block options 1712 1713Parsers should ignore all unrecognised optional fields. Currently the 1714possible optional fields are: 1715 1716shared:X mount is shared in peer group X 1717master:X mount is slave to peer group X 1718propagate_from:X mount is slave and receives propagation from peer group X (*) 1719unbindable mount is unbindable 1720 1721(*) X is the closest dominant peer group under the process's root. If 1722X is the immediate master of the mount, or if there's no dominant peer 1723group under the same root, then only the "master:X" field is present 1724and not the "propagate_from:X" field. 1725 1726For more information on mount propagation see: 1727 1728 Documentation/filesystems/sharedsubtree.txt 1729 1730 17313.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 1732-------------------------------------------------------- 1733These files provide a method to access a tasks comm value. It also allows for 1734a task to set its own or one of its thread siblings comm value. The comm value 1735is limited in size compared to the cmdline value, so writing anything longer 1736then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated 1737comm value. 1738 1739 17403.7 /proc/<pid>/task/<tid>/children - Information about task children 1741------------------------------------------------------------------------- 1742This file provides a fast way to retrieve first level children pids 1743of a task pointed by <pid>/<tid> pair. The format is a space separated 1744stream of pids. 1745 1746Note the "first level" here -- if a child has own children they will 1747not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children 1748to obtain the descendants. 1749 1750Since this interface is intended to be fast and cheap it doesn't 1751guarantee to provide precise results and some children might be 1752skipped, especially if they've exited right after we printed their 1753pids, so one need to either stop or freeze processes being inspected 1754if precise results are needed. 1755 1756 17573.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 1758--------------------------------------------------------------- 1759This file provides information associated with an opened file. The regular 1760files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos' 1761represents the current offset of the opened file in decimal form [see lseek(2) 1762for details], 'flags' denotes the octal O_xxx mask the file has been 1763created with [see open(2) for details] and 'mnt_id' represents mount ID of 1764the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo 1765for details]. 1766 1767A typical output is 1768 1769 pos: 0 1770 flags: 0100002 1771 mnt_id: 19 1772 1773All locks associated with a file descriptor are shown in its fdinfo too. 1774 1775lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF 1776 1777The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags 1778pair provide additional information particular to the objects they represent. 1779 1780 Eventfd files 1781 ~~~~~~~~~~~~~ 1782 pos: 0 1783 flags: 04002 1784 mnt_id: 9 1785 eventfd-count: 5a 1786 1787 where 'eventfd-count' is hex value of a counter. 1788 1789 Signalfd files 1790 ~~~~~~~~~~~~~~ 1791 pos: 0 1792 flags: 04002 1793 mnt_id: 9 1794 sigmask: 0000000000000200 1795 1796 where 'sigmask' is hex value of the signal mask associated 1797 with a file. 1798 1799 Epoll files 1800 ~~~~~~~~~~~ 1801 pos: 0 1802 flags: 02 1803 mnt_id: 9 1804 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7 1805 1806 where 'tfd' is a target file descriptor number in decimal form, 1807 'events' is events mask being watched and the 'data' is data 1808 associated with a target [see epoll(7) for more details]. 1809 1810 The 'pos' is current offset of the target file in decimal form 1811 [see lseek(2)], 'ino' and 'sdev' are inode and device numbers 1812 where target file resides, all in hex format. 1813 1814 Fsnotify files 1815 ~~~~~~~~~~~~~~ 1816 For inotify files the format is the following 1817 1818 pos: 0 1819 flags: 02000000 1820 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d 1821 1822 where 'wd' is a watch descriptor in decimal form, ie a target file 1823 descriptor number, 'ino' and 'sdev' are inode and device where the 1824 target file resides and the 'mask' is the mask of events, all in hex 1825 form [see inotify(7) for more details]. 1826 1827 If the kernel was built with exportfs support, the path to the target 1828 file is encoded as a file handle. The file handle is provided by three 1829 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex 1830 format. 1831 1832 If the kernel is built without exportfs support the file handle won't be 1833 printed out. 1834 1835 If there is no inotify mark attached yet the 'inotify' line will be omitted. 1836 1837 For fanotify files the format is 1838 1839 pos: 0 1840 flags: 02 1841 mnt_id: 9 1842 fanotify flags:10 event-flags:0 1843 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003 1844 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4 1845 1846 where fanotify 'flags' and 'event-flags' are values used in fanotify_init 1847 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of 1848 flags associated with mark which are tracked separately from events 1849 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events 1850 mask and 'ignored_mask' is the mask of events which are to be ignored. 1851 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask' 1852 does provide information about flags and mask used in fanotify_mark 1853 call [see fsnotify manpage for details]. 1854 1855 While the first three lines are mandatory and always printed, the rest is 1856 optional and may be omitted if no marks created yet. 1857 1858 Timerfd files 1859 ~~~~~~~~~~~~~ 1860 1861 pos: 0 1862 flags: 02 1863 mnt_id: 9 1864 clockid: 0 1865 ticks: 0 1866 settime flags: 01 1867 it_value: (0, 49406829) 1868 it_interval: (1, 0) 1869 1870 where 'clockid' is the clock type and 'ticks' is the number of the timer expirations 1871 that have occurred [see timerfd_create(2) for details]. 'settime flags' are 1872 flags in octal form been used to setup the timer [see timerfd_settime(2) for 1873 details]. 'it_value' is remaining time until the timer exiration. 1874 'it_interval' is the interval for the timer. Note the timer might be set up 1875 with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value' 1876 still exhibits timer's remaining time. 1877 18783.9 /proc/<pid>/map_files - Information about memory mapped files 1879--------------------------------------------------------------------- 1880This directory contains symbolic links which represent memory mapped files 1881the process is maintaining. Example output: 1882 1883 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so 1884 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so 1885 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so 1886 | ... 1887 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1 1888 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls 1889 1890The name of a link represents the virtual memory bounds of a mapping, i.e. 1891vm_area_struct::vm_start-vm_area_struct::vm_end. 1892 1893The main purpose of the map_files is to retrieve a set of memory mapped 1894files in a fast way instead of parsing /proc/<pid>/maps or 1895/proc/<pid>/smaps, both of which contain many more records. At the same 1896time one can open(2) mappings from the listings of two processes and 1897comparing their inode numbers to figure out which anonymous memory areas 1898are actually shared. 1899 19003.10 /proc/<pid>/timerslack_ns - Task timerslack value 1901--------------------------------------------------------- 1902This file provides the value of the task's timerslack value in nanoseconds. 1903This value specifies a amount of time that normal timers may be deferred 1904in order to coalesce timers and avoid unnecessary wakeups. 1905 1906This allows a task's interactivity vs power consumption trade off to be 1907adjusted. 1908 1909Writing 0 to the file will set the tasks timerslack to the default value. 1910 1911Valid values are from 0 - ULLONG_MAX 1912 1913An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level 1914permissions on the task specified to change its timerslack_ns value. 1915 19163.11 /proc/<pid>/patch_state - Livepatch patch operation state 1917----------------------------------------------------------------- 1918When CONFIG_LIVEPATCH is enabled, this file displays the value of the 1919patch state for the task. 1920 1921A value of '-1' indicates that no patch is in transition. 1922 1923A value of '0' indicates that a patch is in transition and the task is 1924unpatched. If the patch is being enabled, then the task hasn't been 1925patched yet. If the patch is being disabled, then the task has already 1926been unpatched. 1927 1928A value of '1' indicates that a patch is in transition and the task is 1929patched. If the patch is being enabled, then the task has already been 1930patched. If the patch is being disabled, then the task hasn't been 1931unpatched yet. 1932 1933 1934------------------------------------------------------------------------------ 1935Configuring procfs 1936------------------------------------------------------------------------------ 1937 19384.1 Mount options 1939--------------------- 1940 1941The following mount options are supported: 1942 1943 hidepid= Set /proc/<pid>/ access mode. 1944 gid= Set the group authorized to learn processes information. 1945 1946hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories 1947(default). 1948 1949hidepid=1 means users may not access any /proc/<pid>/ directories but their 1950own. Sensitive files like cmdline, sched*, status are now protected against 1951other users. This makes it impossible to learn whether any user runs 1952specific program (given the program doesn't reveal itself by its behaviour). 1953As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users, 1954poorly written programs passing sensitive information via program arguments are 1955now protected against local eavesdroppers. 1956 1957hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other 1958users. It doesn't mean that it hides a fact whether a process with a specific 1959pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"), 1960but it hides process' uid and gid, which may be learned by stat()'ing 1961/proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering 1962information about running processes, whether some daemon runs with elevated 1963privileges, whether other user runs some sensitive program, whether other users 1964run any program at all, etc. 1965 1966gid= defines a group authorized to learn processes information otherwise 1967prohibited by hidepid=. If you use some daemon like identd which needs to learn 1968information about processes information, just add identd to this group. 1969