1LSOF(1) General Commands Manual LSOF(1)
2
3
4
6 lsof - list open files
7
9 lsof [ -?abChlnNOPRtUvVX ] [ -A A ] [ -c c ] [ +c c ] [ +|-d d ] [ +|-D
10 D ] [ +|-e s ] [ +|-E ] [ +|-f [cfgGn] ] [ -F [f] ] [ -g [s] ] [ -i [i]
11 ] [ -k k ] [ -K k ] [ +|-L [l] ] [ +|-m m ] [ +|-M ] [ -o [o] ] [ -p s
12 ] [ +|-r [t[m<fmt>]] ] [ -s [p:s] ] [ -S [t] ] [ -T [t] ] [ -u s ] [
13 +|-w ] [ -x [fl] ] [ -z [z] ] [ -Z [Z] ] [ -- ] [names]
14
16 Lsof revision 4.91 lists on its standard output file information about
17 files opened by processes for the following UNIX dialects:
18
19 Apple Darwin 9 and Mac OS X 10.[567]
20 FreeBSD 8.[234], 9.0 and 1[012].0 for AMD64-based systems
21 Linux 2.1.72 and above for x86-based systems
22 Solaris 9, 10 and 11
23
24 (See the DISTRIBUTION section of this manual page for information on
25 how to obtain the latest lsof revision.)
26
27 An open file may be a regular file, a directory, a block special file,
28 a character special file, an executing text reference, a library, a
29 stream or a network file (Internet socket, NFS file or UNIX domain
30 socket.) A specific file or all the files in a file system may be
31 selected by path.
32
33 Instead of a formatted display, lsof will produce output that can be
34 parsed by other programs. See the -F, option description, and the OUT‐
35 PUT FOR OTHER PROGRAMS section for more information.
36
37 In addition to producing a single output list, lsof will run in repeat
38 mode. In repeat mode it will produce output, delay, then repeat the
39 output operation until stopped with an interrupt or quit signal. See
40 the +|-r [t[m<fmt>]] option description for more information.
41
43 In the absence of any options, lsof lists all open files belonging to
44 all active processes.
45
46 If any list request option is specified, other list requests must be
47 specifically requested - e.g., if -U is specified for the listing of
48 UNIX socket files, NFS files won't be listed unless -N is also speci‐
49 fied; or if a user list is specified with the -u option, UNIX domain
50 socket files, belonging to users not in the list, won't be listed
51 unless the -U option is also specified.
52
53 Normally list options that are specifically stated are ORed - i.e.,
54 specifying the -i option without an address and the -ufoo option pro‐
55 duces a listing of all network files OR files belonging to processes
56 owned by user ``foo''. The exceptions are:
57
58 1) the `^' (negated) login name or user ID (UID), specified with the -u
59 option;
60
61 2) the `^' (negated) process ID (PID), specified with the -p option;
62
63 3) the `^' (negated) process group ID (PGID), specified with the -g
64 option;
65
66 4) the `^' (negated) command, specified with the -c option;
67
68 5) the (`^') negated TCP or UDP protocol state names, specified with
69 the -s [p:s] option.
70
71 Since they represent exclusions, they are applied without ORing or AND‐
72 ing and take effect before any other selection criteria are applied.
73
74 The -a option may be used to AND the selections. For example, specify‐
75 ing -a, -U, and -ufoo produces a listing of only UNIX socket files that
76 belong to processes owned by user ``foo''.
77
78 Caution: the -a option causes all list selection options to be ANDed;
79 it can't be used to cause ANDing of selected pairs of selection options
80 by placing it between them, even though its placement there is accept‐
81 able. Wherever -a is placed, it causes the ANDing of all selection
82 options.
83
84 Items of the same selection set - command names, file descriptors, net‐
85 work addresses, process identifiers, user identifiers, zone names,
86 security contexts - are joined in a single ORed set and applied before
87 the result participates in ANDing. Thus, for example, specifying
88 -i@aaa.bbb, -i@ccc.ddd, -a, and -ufff,ggg will select the listing of
89 files that belong to either login ``fff'' OR ``ggg'' AND have network
90 connections to either host aaa.bbb OR ccc.ddd.
91
92 Options may be grouped together following a single prefix -- e.g., the
93 option set ``-a -b -C'' may be stated as -abC. However, since values
94 are optional following +|-f, -F, -g, -i, +|-L, -o, +|-r, -s, -S, -T, -x
95 and -z. when you have no values for them be careful that the following
96 character isn't ambiguous. For example, -Fn might represent the -F and
97 -n options, or it might represent the n field identifier character fol‐
98 lowing the -F option. When ambiguity is possible, start a new option
99 with a `-' character - e.g., ``-F -n''. If the next option is a file
100 name, follow the possibly ambiguous option with ``--'' - e.g., ``-F --
101 name''.
102
103 Either the `+' or the `-' prefix may be applied to a group of options.
104 Options that don't take on separate meanings for each prefix - e.g., -i
105 - may be grouped under either prefix. Thus, for example, ``+M -i'' may
106 be stated as ``+Mi'' and the group means the same as the separate
107 options. Be careful of prefix grouping when one or more options in the
108 group does take on separate meanings under different prefixes - e.g.,
109 +|-M; ``-iM'' is not the same request as ``-i +M''. When in doubt, use
110 separate options with appropriate prefixes.
111
112 -? -h These two equivalent options select a usage (help) output
113 list. Lsof displays a shortened form of this output when it
114 detects an error in the options supplied to it, after it has
115 displayed messages explaining each error. (Escape the `?'
116 character as your shell requires.)
117
118 -a causes list selection options to be ANDed, as described above.
119
120 -A A is available on systems configured for AFS whose AFS kernel
121 code is implemented via dynamic modules. It allows the lsof
122 user to specify A as an alternate name list file where the
123 kernel addresses of the dynamic modules might be found. See
124 the lsof FAQ (The FAQ section gives its location.) for more
125 information about dynamic modules, their symbols, and how they
126 affect lsof.
127
128 -b causes lsof to avoid kernel functions that might block -
129 lstat(2), readlink(2), and stat(2).
130
131 See the BLOCKS AND TIMEOUTS and AVOIDING KERNEL BLOCKS sec‐
132 tions for information on using this option.
133
134 -c c selects the listing of files for processes executing the com‐
135 mand that begins with the characters of c. Multiple commands
136 may be specified, using multiple -c options. They are joined
137 in a single ORed set before participating in AND option selec‐
138 tion.
139
140 If c begins with a `^', then the following characters specify
141 a command name whose processes are to be ignored (excluded.)
142
143 If c begins and ends with a slash ('/'), the characters
144 between the slashes are interpreted as a regular expression.
145 Shell meta-characters in the regular expression must be quoted
146 to prevent their interpretation by the shell. The closing
147 slash may be followed by these modifiers:
148
149 b the regular expression is a basic one.
150 i ignore the case of letters.
151 x the regular expression is an extended one
152 (default).
153
154 See the lsof FAQ (The FAQ section gives its location.) for
155 more information on basic and extended regular expressions.
156
157 The simple command specification is tested first. If that
158 test fails, the command regular expression is applied. If the
159 simple command test succeeds, the command regular expression
160 test isn't made. This may result in ``no command found for
161 regex:'' messages when lsof's -V option is specified.
162
163 +c w defines the maximum number of initial characters of the name,
164 supplied by the UNIX dialect, of the UNIX command associated
165 with a process to be printed in the COMMAND column. (The lsof
166 default is nine.)
167
168 Note that many UNIX dialects do not supply all command name
169 characters to lsof in the files and structures from which lsof
170 obtains command name. Often dialects limit the number of
171 characters supplied in those sources. For example, Linux
172 2.4.27 and Solaris 9 both limit command name length to 16
173 characters.
174
175 If w is zero ('0'), all command characters supplied to lsof by
176 the UNIX dialect will be printed.
177
178 If w is less than the length of the column title, ``COMMAND'',
179 it will be raised to that length.
180
181 -C disables the reporting of any path name components from the
182 kernel's name cache. See the KERNEL NAME CACHE section for
183 more information.
184
185 +d s causes lsof to search for all open instances of directory s
186 and the files and directories it contains at its top level.
187 +d does NOT descend the directory tree, rooted at s. The +D D
188 option may be used to request a full-descent directory tree
189 search, rooted at directory D.
190
191 Processing of the +d option does not follow symbolic links
192 within s unless the -x or -x l option is also specified. Nor
193 does it search for open files on file system mount points on
194 subdirectories of s unless the -x or -x f option is also
195 specified.
196
197 Note: the authority of the user of this option limits it to
198 searching for files that the user has permission to examine
199 with the system stat(2) function.
200
201 -d s specifies a list of file descriptors (FDs) to exclude from or
202 include in the output listing. The file descriptors are spec‐
203 ified in the comma-separated set s - e.g., ``cwd,1,3'',
204 ``^6,^2''. (There should be no spaces in the set.)
205
206 The list is an exclusion list if all entries of the set begin
207 with `^'. It is an inclusion list if no entry begins with
208 `^'. Mixed lists are not permitted.
209
210 A file descriptor number range may be in the set as long as
211 neither member is empty, both members are numbers, and the
212 ending member is larger than the starting one - e.g., ``0-7''
213 or ``3-10''. Ranges may be specified for exclusion if they
214 have the `^' prefix - e.g., ``^0-7'' excludes all file
215 descriptors 0 through 7.
216
217 Multiple file descriptor numbers are joined in a single ORed
218 set before participating in AND option selection.
219
220 When there are exclusion and inclusion members in the set,
221 lsof reports them as errors and exits with a non-zero return
222 code.
223
224 See the description of File Descriptor (FD) output values in
225 the OUTPUT section for more information on file descriptor
226 names.
227
228 +D D causes lsof to search for all open instances of directory D
229 and all the files and directories it contains to its complete
230 depth.
231
232 Processing of the +D option does not follow symbolic links
233 within D unless the -x or -x l option is also specified. Nor
234 does it search for open files on file system mount points on
235 subdirectories of D unless the -x or -x f option is also
236 specified.
237
238 Note: the authority of the user of this option limits it to
239 searching for files that the user has permission to examine
240 with the system stat(2) function.
241
242 Further note: lsof may process this option slowly and require
243 a large amount of dynamic memory to do it. This is because it
244 must descend the entire directory tree, rooted at D, calling
245 stat(2) for each file and directory, building a list of all
246 the files it finds, and searching that list for a match with
247 every open file. When directory D is large, these steps can
248 take a long time, so use this option prudently.
249
250 -D D directs lsof's use of the device cache file. The use of this
251 option is sometimes restricted. See the DEVICE CACHE FILE
252 section and the sections that follow it for more information
253 on this option.
254
255 -D must be followed by a function letter; the function letter
256 may optionally be followed by a path name. Lsof recognizes
257 these function letters:
258
259 ? - report device cache file paths
260 b - build the device cache file
261 i - ignore the device cache file
262 r - read the device cache file
263 u - read and update the device cache file
264
265 The b, r, and u functions, accompanied by a path name, are
266 sometimes restricted. When these functions are restricted,
267 they will not appear in the description of the -D option that
268 accompanies -h or -? option output. See the DEVICE CACHE
269 FILE section and the sections that follow it for more informa‐
270 tion on these functions and when they're restricted.
271
272 The ? function reports the read-only and write paths that
273 lsof can use for the device cache file, the names of any envi‐
274 ronment variables whose values lsof will examine when forming
275 the device cache file path, and the format for the personal
276 device cache file path. (Escape the `?' character as your
277 shell requires.)
278
279 When available, the b, r, and u functions may be followed by
280 the device cache file's path. The standard default is
281 .lsof_hostname in the home directory of the real user ID that
282 executes lsof, but this could have been changed when lsof was
283 configured and compiled. (The output of the -h and -?
284 options show the current default prefix - e.g., ``.lsof''.)
285 The suffix, hostname, is the first component of the host's
286 name returned by gethostname(2).
287
288 When available, the b function directs lsof to build a new
289 device cache file at the default or specified path.
290
291 The i function directs lsof to ignore the default device cache
292 file and obtain its information about devices via direct calls
293 to the kernel.
294
295 The r function directs lsof to read the device cache at the
296 default or specified path, but prevents it from creating a new
297 device cache file when none exists or the existing one is
298 improperly structured. The r function, when specified without
299 a path name, prevents lsof from updating an incorrect or out‐
300 dated device cache file, or creating a new one in its place.
301 The r function is always available when it is specified with‐
302 out a path name argument; it may be restricted by the permis‐
303 sions of the lsof process.
304
305 When available, the u function directs lsof to read the device
306 cache file at the default or specified path, if possible, and
307 to rebuild it, if necessary. This is the default device cache
308 file function when no -D option has been specified.
309
310 +|-e s exempts the file system whose path name is s from being sub‐
311 jected to kernel function calls that might block. The +e
312 option exempts stat(2), lstat(2) and most readlink(2) kernel
313 function calls. The -e option exempts only stat(2) and
314 lstat(2) kernel function calls. Multiple file systems may be
315 specified with separate +|-e specifications and each may have
316 readlink(2) calls exempted or not.
317
318 This option is currently implemented only for Linux.
319
320 CAUTION: this option can easily be mis-applied to other than
321 the file system of interest, because it uses path name rather
322 than the more reliable device and inode numbers. (Device and
323 inode numbers are acquired via the potentially blocking
324 stat(2) kernel call and are thus not available, but see the
325 +|-m m option as a possible alternative way to supply device
326 numbers.) Use this option with great care and fully specify
327 the path name of the file system to be exempted.
328
329 When open files on exempted file systems are reported, it may
330 not be possible to obtain all their information. Therefore,
331 some information columns will be blank, the characters
332 ``UNKN'' preface the values in the TYPE column, and the appli‐
333 cable exemption option is added in parentheses to the end of
334 the NAME column. (Some device number information might be
335 made available via the +|-m m option.)
336
337 +|-E +E specifies that Linux pipe, Linux UNIX socket and Linux
338 pseudoterminal files should be displayed with endpoint infor‐
339 mation and the files of the endpoints should also be dis‐
340 played. Note: UNIX socket file endpoint information is only
341 available when the compile flags line of -v output contains
342 HASUXSOCKEPT, and psudoterminal endpoint information is only
343 available when the compile flags line contains HASPTYEPT.
344
345 Pipe endpoint information is displayed in the NAME column in
346 the form ``PID,cmd,FDmode'', where PID is the endpoint process
347 ID; cmd is the endpoint process command; FD is the endpoint
348 file's descriptor; and mode is the endpoint file's access
349 mode.
350
351 Pseudoterminal endpoint information is displayed in the NAME
352 column as ``->/dev/ptsmin PID,cmd,FDmode'' or
353 ``PID,cmd,FDmode''. The first form is for a master device;
354 the second, for a slave device. min is a slave device's minor
355 device number; and PID, cmd, FD and mode are the same as with
356 pipe endpoint information. Note: psudoterminal endpoint
357 information is only available when the compile flags line of
358 -V output contains HASPTYEPT.
359
360 UNIX socket file endpoint information is displayed in the NAME
361 column in the form
362 ``type=TYPE ->INO=INODE PID,cmd,FDmode'', where TYPE is the
363 socket type; INODE is the i-node number of the connected
364 socket; and PID, cmd, FD and mode are the same as with pipe
365 endpoint information. Note: UNIX socket file endpoint infor‐
366 mation is available only when the compile flags line of -v
367 output contains HASUXSOCKEPT.
368
369 Multiple occurrences of this information can appear in a
370 file's NAME column.
371
372 -E specfies that Linux pipe and Linux UNIX socket files should
373 be displayed with endpoint information, but not the files of
374 the endpoints.
375
376 +|-f [cfgGn]
377 f by itself clarifies how path name arguments are to be inter‐
378 preted. When followed by c, f, g, G, or n in any combination
379 it specifies that the listing of kernel file structure infor‐
380 mation is to be enabled (`+') or inhibited (`-').
381
382 Normally a path name argument is taken to be a file system
383 name if it matches a mounted-on directory name reported by
384 mount(8), or if it represents a block device, named in the
385 mount output and associated with a mounted directory name.
386 When +f is specified, all path name arguments will be taken to
387 be file system names, and lsof will complain if any are not.
388 This can be useful, for example, when the file system name
389 (mounted-on device) isn't a block device. This happens for
390 some CD-ROM file systems.
391
392 When -f is specified by itself, all path name arguments will
393 be taken to be simple files. Thus, for example, the ``-f --
394 /'' arguments direct lsof to search for open files with a `/'
395 path name, not all open files in the `/' (root) file system.
396
397 Be careful to make sure +f and -f are properly terminated and
398 aren't followed by a character (e.g., of the file or file sys‐
399 tem name) that might be taken as a parameter. For example,
400 use ``--'' after +f and -f as in these examples.
401
402 $ lsof +f -- /file/system/name
403 $ lsof -f -- /file/name
404
405 The listing of information from kernel file structures,
406 requested with the +f [cfgGn] option form, is normally inhib‐
407 ited, and is not available in whole or part for some dialects
408 - e.g., /proc-based Linux kernels below 2.6.22. When the pre‐
409 fix to f is a plus sign (`+'), these characters request file
410 structure information:
411
412 c file structure use count (not Linux)
413 f file structure address (not Linux)
414 g file flag abbreviations (Linux 2.6.22 and up)
415 G file flags in hexadecimal (Linux 2.6.22 and up)
416 n file structure node address (not Linux)
417
418 When the prefix is minus (`-') the same characters disable the
419 listing of the indicated values.
420
421 File structure addresses, use counts, flags, and node
422 addresses may be used to detect more readily identical files
423 inherited by child processes and identical files in use by
424 different processes. Lsof column output can be sorted by out‐
425 put columns holding the values and listed to identify identi‐
426 cal file use, or lsof field output can be parsed by an AWK or
427 Perl post-filter script, or by a C program.
428
429 -F f specifies a character list, f, that selects the fields to be
430 output for processing by another program, and the character
431 that terminates each output field. Each field to be output is
432 specified with a single character in f. The field terminator
433 defaults to NL, but may be changed to NUL (000). See the OUT‐
434 PUT FOR OTHER PROGRAMS section for a description of the field
435 identification characters and the field output process.
436
437 When the field selection character list is empty, all standard
438 fields are selected (except the raw device field, security
439 context and zone field for compatibility reasons) and the NL
440 field terminator is used.
441
442 When the field selection character list contains only a zero
443 (`0'), all fields are selected (except the raw device field
444 for compatibility reasons) and the NUL terminator character is
445 used.
446
447 Other combinations of fields and their associated field termi‐
448 nator character must be set with explicit entries in f, as
449 described in the OUTPUT FOR OTHER PROGRAMS section.
450
451 When a field selection character identifies an item lsof does
452 not normally list - e.g., PPID, selected with -R - specifica‐
453 tion of the field character - e.g., ``-FR'' - also selects the
454 listing of the item.
455
456 When the field selection character list contains the single
457 character `?', lsof will display a help list of the field
458 identification characters. (Escape the `?' character as your
459 shell requires.)
460
461 -g [s] excludes or selects the listing of files for the processes
462 whose optional process group IDentification (PGID) numbers are
463 in the comma-separated set s - e.g., ``123'' or ``123,^456''.
464 (There should be no spaces in the set.)
465
466 PGID numbers that begin with `^' (negation) represent exclu‐
467 sions.
468
469 Multiple PGID numbers are joined in a single ORed set before
470 participating in AND option selection. However, PGID exclu‐
471 sions are applied without ORing or ANDing and take effect
472 before other selection criteria are applied.
473
474 The -g option also enables the output display of PGID numbers.
475 When specified without a PGID set that's all it does.
476
477 -i [i] selects the listing of files any of whose Internet address
478 matches the address specified in i. If no address is speci‐
479 fied, this option selects the listing of all Internet and x.25
480 (HP-UX) network files.
481
482 If -i4 or -i6 is specified with no following address, only
483 files of the indicated IP version, IPv4 or IPv6, are dis‐
484 played. (An IPv6 specification may be used only if the
485 dialects supports IPv6, as indicated by ``[46]'' and
486 ``IPv[46]'' in lsof's -h or -? output.) Sequentially speci‐
487 fying -i4, followed by -i6 is the same as specifying -i, and
488 vice-versa. Specifying -i4, or -i6 after -i is the same as
489 specifying -i4 or -i6 by itself.
490
491 Multiple addresses (up to a limit of 100) may be specified
492 with multiple -i options. (A port number or service name
493 range is counted as one address.) They are joined in a single
494 ORed set before participating in AND option selection.
495
496 An Internet address is specified in the form (Items in square
497 brackets are optional.):
498
499 [46][protocol][@hostname|hostaddr][:service|port]
500
501 where:
502 46 specifies the IP version, IPv4 or IPv6
503 that applies to the following address.
504 '6' may be be specified only if the UNIX
505 dialect supports IPv6. If neither '4' nor
506 '6' is specified, the following address
507 applies to all IP versions.
508 protocol is a protocol name - TCP, UDP
509 hostname is an Internet host name. Unless a
510 specific IP version is specified, open
511 network files associated with host names
512 of all versions will be selected.
513 hostaddr is a numeric Internet IPv4 address in
514 dot form; or an IPv6 numeric address in
515 colon form, enclosed in brackets, if the
516 UNIX dialect supports IPv6. When an IP
517 version is selected, only its numeric
518 addresses may be specified.
519 service is an /etc/services name - e.g., smtp -
520 or a list of them.
521 port is a port number, or a list of them.
522
523 IPv6 options may be used only if the UNIX dialect supports
524 IPv6. To see if the dialect supports IPv6, run lsof and spec‐
525 ify the -h or -? (help) option. If the displayed description
526 of the -i option contains ``[46]'' and ``IPv[46]'', IPv6 is
527 supported.
528
529 IPv4 host names and addresses may not be specified if network
530 file selection is limited to IPv6 with -i 6. IPv6 host names
531 and addresses may not be specified if network file selection
532 is limited to IPv4 with -i 4. When an open IPv4 network
533 file's address is mapped in an IPv6 address, the open file's
534 type will be IPv6, not IPv4, and its display will be selected
535 by '6', not '4'.
536
537 At least one address component - 4, 6, protocol, hostname,
538 hostaddr, or service - must be supplied. The `@' character,
539 leading the host specification, is always required; as is the
540 `:', leading the port specification. Specify either hostname
541 or hostaddr. Specify either service name list or port number
542 list. If a service name list is specified, the protocol may
543 also need to be specified if the TCP, UDP and UDPLITE port
544 numbers for the service name are different. Use any case -
545 lower or upper - for protocol.
546
547 Service names and port numbers may be combined in a list whose
548 entries are separated by commas and whose numeric range
549 entries are separated by minus signs. There may be no embed‐
550 ded spaces, and all service names must belong to the specified
551 protocol. Since service names may contain embedded minus
552 signs, the starting entry of a range can't be a service name;
553 it can be a port number, however.
554
555 Here are some sample addresses:
556
557 -i6 - IPv6 only
558 TCP:25 - TCP and port 25
559 @1.2.3.4 - Internet IPv4 host address 1.2.3.4
560 @[3ffe:1ebc::1]:1234 - Internet IPv6 host address
561 3ffe:1ebc::1, port 1234
562 UDP:who - UDP who service port
563 TCP@lsof.itap:513 - TCP, port 513 and host name lsof.itap
564 tcp@foo:1-10,smtp,99 - TCP, ports 1 through 10,
565 service name smtp, port 99, host name foo
566 tcp@bar:1-smtp - TCP, ports 1 through smtp, host bar
567 :time - either TCP, UDP or UDPLITE time service port
568
569 -K k selects the listing of tasks (threads) of processes, on
570 dialects where task (thread) reporting is supported. (If help
571 output - i.e., the output of the -h or -? options - shows
572 this option, then task (thread) reporting is supported by the
573 dialect.)
574
575 If -K is followed by a value, k, it must be ``i''. That
576 causes lsof to ignore tasks, particularly in the default,
577 list-everything case when no other options are specified.
578
579 When -K and -a are both specified on Linux, and the tasks of a
580 main process are selected by other options, the main process
581 will also be listed as though it were a task, but without a
582 task ID. (See the description of the TID column in the OUTPUT
583 section.)
584
585 Where the FreeBSD version supports threads, all threads will
586 be listed with their IDs.
587
588 In general threads and tasks inherit the files of the caller,
589 but may close some and open others, so lsof always reports all
590 the open files of threads and tasks.
591
592 -k k specifies a kernel name list file, k, in place of /vmunix,
593 /mach, etc. -k is not available under AIX on the IBM
594 RISC/System 6000.
595
596 -l inhibits the conversion of user ID numbers to login names. It
597 is also useful when login name lookup is working improperly or
598 slowly.
599
600 +|-L [l] enables (`+') or disables (`-') the listing of file link
601 counts, where they are available - e.g., they aren't available
602 for sockets, or most FIFOs and pipes.
603
604 When +L is specified without a following number, all link
605 counts will be listed. When -L is specified (the default), no
606 link counts will be listed.
607
608 When +L is followed by a number, only files having a link
609 count less than that number will be listed. (No number may
610 follow -L.) A specification of the form ``+L1'' will select
611 open files that have been unlinked. A specification of the
612 form ``+aL1 <file_system>'' will select unlinked open files on
613 the specified file system.
614
615 For other link count comparisons, use field output (-F) and a
616 post-processing script or program.
617
618 +|-m m specifies an alternate kernel memory file or activates mount
619 table supplement processing.
620
621 The option form -m m specifies a kernel memory file, m, in
622 place of /dev/kmem or /dev/mem - e.g., a crash dump file.
623
624 The option form +m requests that a mount supplement file be
625 written to the standard output file. All other options are
626 silently ignored.
627
628 There will be a line in the mount supplement file for each
629 mounted file system, containing the mounted file system direc‐
630 tory, followed by a single space, followed by the device num‐
631 ber in hexadecimal "0x" format - e.g.,
632
633 / 0x801
634
635 Lsof can use the mount supplement file to get device numbers
636 for file systems when it can't get them via stat(2) or
637 lstat(2).
638
639 The option form +m m identifies m as a mount supplement file.
640
641 Note: the +m and +m m options are not available for all sup‐
642 ported dialects. Check the output of lsof's -h or -? options
643 to see if the +m and +m m options are available.
644
645 +|-M Enables (+) or disables (-) the reporting of portmapper regis‐
646 trations for local TCP, UDP and UDPLITE ports, where port map‐
647 ping is supported. (See the last paragraph of this option
648 description for information about where portmapper registra‐
649 tion reporting is supported.)
650
651 The default reporting mode is set by the lsof builder with the
652 HASPMAPENABLED #define in the dialect's machine.h header file;
653 lsof is distributed with the HASPMAPENABLED #define deacti‐
654 vated, so portmapper reporting is disabled by default and must
655 be requested with +M. Specifying lsof's -h or -? option will
656 report the default mode. Disabling portmapper registration
657 when it is already disabled or enabling it when already
658 enabled is acceptable. When portmapper registration reporting
659 is enabled, lsof displays the portmapper registration (if any)
660 for local TCP, UDP or UDPLITE ports in square brackets immedi‐
661 ately following the port numbers or service names - e.g.,
662 ``:1234[name]'' or ``:name[100083]''. The registration infor‐
663 mation may be a name or number, depending on what the regis‐
664 tering program supplied to the portmapper when it registered
665 the port.
666
667 When portmapper registration reporting is enabled, lsof may
668 run a little more slowly or even become blocked when access to
669 the portmapper becomes congested or stopped. Reverse the
670 reporting mode to determine if portmapper registration report‐
671 ing is slowing or blocking lsof.
672
673 For purposes of portmapper registration reporting lsof consid‐
674 ers a TCP, UDP or UDPLITE port local if: it is found in the
675 local part of its containing kernel structure; or if it is
676 located in the foreign part of its containing kernel structure
677 and the local and foreign Internet addresses are the same; or
678 if it is located in the foreign part of its containing kernel
679 structure and the foreign Internet address is INADDR_LOOPBACK
680 (127.0.0.1). This rule may make lsof ignore some foreign
681 ports on machines with multiple interfaces when the foreign
682 Internet address is on a different interface from the local
683 one.
684
685 See the lsof FAQ (The FAQ section gives its location.) for
686 further discussion of portmapper registration reporting
687 issues.
688
689 Portmapper registration reporting is supported only on
690 dialects that have RPC header files. (Some Linux distribu‐
691 tions with GlibC 2.14 do not have them.) When portmapper reg‐
692 istration reporting is supported, the -h or -? help output
693 will show the +|-M option.
694
695 -n inhibits the conversion of network numbers to host names for
696 network files. Inhibiting conversion may make lsof run
697 faster. It is also useful when host name lookup is not work‐
698 ing properly.
699
700 -N selects the listing of NFS files.
701
702 -o directs lsof to display file offset at all times. It causes
703 the SIZE/OFF output column title to be changed to OFFSET.
704 Note: on some UNIX dialects lsof can't obtain accurate or con‐
705 sistent file offset information from its kernel data sources,
706 sometimes just for particular kinds of files (e.g., socket
707 files.) Consult the lsof FAQ (The FAQ section gives its loca‐
708 tion.) for more information.
709
710 The -o and -s options are mutually exclusive; they can't both
711 be specified. When neither is specified, lsof displays what‐
712 ever value - size or offset - is appropriate and available for
713 the type of the file.
714
715 -o o defines the number of decimal digits (o) to be printed after
716 the ``0t'' for a file offset before the form is switched to
717 ``0x...''. An o value of zero (unlimited) directs lsof to use
718 the ``0t'' form for all offset output.
719
720 This option does NOT direct lsof to display offset at all
721 times; specify -o (without a trailing number) to do that. -o
722 o only specifies the number of digits after ``0t'' in either
723 mixed size and offset or offset-only output. Thus, for exam‐
724 ple, to direct lsof to display offset at all times with a dec‐
725 imal digit count of 10, use:
726
727 -o -o 10
728 or
729 -oo10
730
731 The default number of digits allowed after ``0t'' is normally
732 8, but may have been changed by the lsof builder. Consult the
733 description of the -o o option in the output of the -h or -?
734 option to determine the default that is in effect.
735
736 -O directs lsof to bypass the strategy it uses to avoid being
737 blocked by some kernel operations - i.e., doing them in forked
738 child processes. See the BLOCKS AND TIMEOUTS and AVOIDING
739 KERNEL BLOCKS sections for more information on kernel opera‐
740 tions that may block lsof.
741
742 While use of this option will reduce lsof startup overhead, it
743 may also cause lsof to hang when the kernel doesn't respond to
744 a function. Use this option cautiously.
745
746 -p s excludes or selects the listing of files for the processes
747 whose optional process IDentification (PID) numbers are in the
748 comma-separated set s - e.g., ``123'' or ``123,^456''. (There
749 should be no spaces in the set.)
750
751 PID numbers that begin with `^' (negation) represent exclu‐
752 sions.
753
754 Multiple process ID numbers are joined in a single ORed set
755 before participating in AND option selection. However, PID
756 exclusions are applied without ORing or ANDing and take effect
757 before other selection criteria are applied.
758
759 -P inhibits the conversion of port numbers to port names for net‐
760 work files. Inhibiting the conversion may make lsof run a
761 little faster. It is also useful when port name lookup is not
762 working properly.
763
764 +|-r [t[m<fmt>]]
765 puts lsof in repeat mode. There lsof lists open files as
766 selected by other options, delays t seconds (default fifteen),
767 then repeats the listing, delaying and listing repetitively
768 until stopped by a condition defined by the prefix to the
769 option.
770
771 If the prefix is a `-', repeat mode is endless. Lsof must be
772 terminated with an interrupt or quit signal.
773
774 If the prefix is `+', repeat mode will end the first cycle no
775 open files are listed - and of course when lsof is stopped
776 with an interrupt or quit signal. When repeat mode ends
777 because no files are listed, the process exit code will be
778 zero if any open files were ever listed; one, if none were
779 ever listed.
780
781 Lsof marks the end of each listing: if field output is in
782 progress (the -F, option has been specified), the default
783 marker is `m'; otherwise the default marker is ``========''.
784 The marker is followed by a NL character.
785
786 The optional "m<fmt>" argument specifies a format for the
787 marker line. The <fmt> characters following `m' are inter‐
788 preted as a format specification to the strftime(3) function,
789 when both it and the localtime(3) function are available in
790 the dialect's C library. Consult the strftime(3) documenta‐
791 tion for what may appear in its format specification. Note
792 that when field output is requested with the -F option, <fmt>
793 cannot contain the NL format, ``%n''. Note also that when
794 <fmt> contains spaces or other characters that affect the
795 shell's interpretation of arguments, <fmt> must be quoted
796 appropriately.
797
798 Repeat mode reduces lsof startup overhead, so it is more effi‐
799 cient to use this mode than to call lsof repetitively from a
800 shell script, for example.
801
802 To use repeat mode most efficiently, accompany +|-r with spec‐
803 ification of other lsof selection options, so the amount of
804 kernel memory access lsof does will be kept to a minimum.
805 Options that filter at the process level - e.g., -c, -g, -p,
806 -u - are the most efficient selectors.
807
808 Repeat mode is useful when coupled with field output (see the
809 -F, option description) and a supervising awk or Perl script,
810 or a C program.
811
812 -R directs lsof to list the Parent Process IDentification number
813 in the PPID column.
814
815 -s [p:s] s alone directs lsof to display file size at all times. It
816 causes the SIZE/OFF output column title to be changed to SIZE.
817 If the file does not have a size, nothing is displayed.
818
819 The optional -s p:s form is available only for selected
820 dialects, and only when the -h or -? help output lists it.
821
822 When the optional form is available, the s may be followed by
823 a protocol name (p), either TCP or UDP, a colon (`:') and a
824 comma-separated protocol state name list, the option causes
825 open TCP and UDP files to be excluded if their state name(s)
826 are in the list (s) preceded by a `^'; or included if their
827 name(s) are not preceded by a `^'.
828
829 Dialects that support this option may support only one proto‐
830 col. When an unsupported protocol is specified, a message
831 will be displayed indicating state names for the protocol are
832 unavailable.
833
834 When an inclusion list is defined, only network files with
835 state names in the list will be present in the lsof output.
836 Thus, specifying one state name means that only network files
837 with that lone state name will be listed.
838
839 Case is unimportant in the protocol or state names, but there
840 may be no spaces and the colon (`:') separating the protocol
841 name (p) and the state name list (s) is required.
842
843 If only TCP and UDP files are to be listed, as controlled by
844 the specified exclusions and inclusions, the -i option must be
845 specified, too. If only a single protocol's files are to be
846 listed, add its name as an argument to the -i option.
847
848 For example, to list only network files with TCP state LISTEN,
849 use:
850
851 -iTCP -sTCP:LISTEN
852
853 Or, for example, to list network files with all UDP states
854 except Idle, use:
855
856 -iUDP -sUDP:Idle
857
858 State names vary with UNIX dialects, so it's not possible to
859 provide a complete list. Some common TCP state names are:
860 CLOSED, IDLE, BOUND, LISTEN, ESTABLISHED, SYN_SENT, SYN_RCDV,
861 ESTABLISHED, CLOSE_WAIT, FIN_WAIT1, CLOSING, LAST_ACK,
862 FIN_WAIT_2, and TIME_WAIT. Two common UDP state names are
863 Unbound and Idle.
864
865 See the lsof FAQ (The FAQ section gives its location.) for
866 more information on how to use protocol state exclusion and
867 inclusion, including examples.
868
869 The -o (without a following decimal digit count) and -s option
870 (without a following protocol and state name list) are mutu‐
871 ally exclusive; they can't both be specified. When neither is
872 specified, lsof displays whatever value - size or offset - is
873 appropriate and available for the type of file.
874
875 Since some types of files don't have true sizes - sockets,
876 FIFOs, pipes, etc. - lsof displays for their sizes the content
877 amounts in their associated kernel buffers, if possible.
878
879 -S [t] specifies an optional time-out seconds value for kernel func‐
880 tions - lstat(2), readlink(2), and stat(2) - that might other‐
881 wise deadlock. The minimum for t is two; the default, fif‐
882 teen; when no value is specified, the default is used.
883
884 See the BLOCKS AND TIMEOUTS section for more information.
885
886 -T [t] controls the reporting of some TCP/TPI information, also
887 reported by netstat(1), following the network addresses. In
888 normal output the information appears in parentheses, each
889 item except TCP or TPI state name identified by a keyword,
890 followed by `=', separated from others by a single space:
891
892 <TCP or TPI state name>
893 QR=<read queue length>
894 QS=<send queue length>
895 SO=<socket options and values>
896 SS=<socket states>
897 TF=<TCP flags and values>
898 WR=<window read length>
899 WW=<window write length>
900
901 Not all values are reported for all UNIX dialects. Items val‐
902 ues (when available) are reported after the item name and '='.
903
904 When the field output mode is in effect (See OUTPUT FOR OTHER
905 PROGRAMS.) each item appears as a field with a `T' leading
906 character.
907
908 -T with no following key characters disables TCP/TPI informa‐
909 tion reporting.
910
911 -T with following characters selects the reporting of specific
912 TCP/TPI information:
913
914 f selects reporting of socket options,
915 states and values, and TCP flags and
916 values.
917 q selects queue length reporting.
918 s selects connection state reporting.
919 w selects window size reporting.
920
921 Not all selections are enabled for some UNIX dialects. State
922 may be selected for all dialects and is reported by default.
923 The -h or -? help output for the -T option will show what
924 selections may be used with the UNIX dialect.
925
926 When -T is used to select information - i.e., it is followed
927 by one or more selection characters - the displaying of state
928 is disabled by default, and it must be explicitly selected
929 again in the characters following -T. (In effect, then, the
930 default is equivalent to -Ts.) For example, if queue lengths
931 and state are desired, use -Tqs.
932
933 Socket options, socket states, some socket values, TCP flags
934 and one TCP value may be reported (when available in the UNIX
935 dialect) in the form of the names that commonly appear after
936 SO_, so_, SS_, TCP_ and TF_ in the dialect's header files -
937 most often <sys/socket.h>, <sys/socketvar.h> and
938 <netinet/tcp_var.h>. Consult those header files for the mean‐
939 ing of the flags, options, states and values.
940
941 ``SO='' precedes socket options and values; ``SS='', socket
942 states; and ``TF='', TCP flags and values.
943
944 If a flag or option has a value, the value will follow an '='
945 and the name -- e.g., ``SO=LINGER=5'', ``SO=QLIM=5'',
946 ``TF=MSS=512''. The following seven values may be reported:
947
948 Name
949 Reported Description (Common Symbol)
950
951 KEEPALIVE keep alive time (SO_KEEPALIVE)
952 LINGER linger time (SO_LINGER)
953 MSS maximum segment size (TCP_MAXSEG)
954 PQLEN partial listen queue connections
955 QLEN established listen queue connections
956 QLIM established listen queue limit
957 RCVBUF receive buffer length (SO_RCVBUF)
958 SNDBUF send buffer length (SO_SNDBUF)
959
960 Details on what socket options and values, socket states, and
961 TCP flags and values may be displayed for particular UNIX
962 dialects may be found in the answer to the ``Why doesn't lsof
963 report socket options, socket states, and TCP flags and values
964 for my dialect?'' and ``Why doesn't lsof report the partial
965 listen queue connection count for my dialect?'' questions in
966 the lsof FAQ (The FAQ section gives its location.)
967
968 -t specifies that lsof should produce terse output with process
969 identifiers only and no header - e.g., so that the output may
970 be piped to kill(1). -t selects the -w option.
971
972 -u s selects the listing of files for the user whose login names or
973 user ID numbers are in the comma-separated set s - e.g.,
974 ``abe'', or ``548,root''. (There should be no spaces in the
975 set.)
976
977 Multiple login names or user ID numbers are joined in a single
978 ORed set before participating in AND option selection.
979
980 If a login name or user ID is preceded by a `^', it becomes a
981 negation - i.e., files of processes owned by the login name or
982 user ID will never be listed. A negated login name or user ID
983 selection is neither ANDed nor ORed with other selections; it
984 is applied before all other selections and absolutely excludes
985 the listing of the files of the process. For example, to
986 direct lsof to exclude the listing of files belonging to root
987 processes, specify ``-u^root'' or ``-u^0''.
988
989 -U selects the listing of UNIX domain socket files.
990
991 -v selects the listing of lsof version information, including:
992 revision number; when the lsof binary was constructed; who
993 constructed the binary and where; the name of the compiler
994 used to construct the lsof binary; the version number of the
995 compiler when readily available; the compiler and loader flags
996 used to construct the lsof binary; and system information,
997 typically the output of uname's -a option.
998
999 -V directs lsof to indicate the items it was asked to list and
1000 failed to find - command names, file names, Internet addresses
1001 or files, login names, NFS files, PIDs, PGIDs, and UIDs.
1002
1003 When other options are ANDed to search options, or com‐
1004 pile-time options restrict the listing of some files, lsof may
1005 not report that it failed to find a search item when an ANDed
1006 option or compile-time option prevents the listing of the open
1007 file containing the located search item.
1008
1009 For example, ``lsof -V -iTCP@foobar -a -d 999'' may not report
1010 a failure to locate open files at ``TCP@foobar'' and may not
1011 list any, if none have a file descriptor number of 999. A
1012 similar situation arises when HASSECURITY and HASNOSOCKSECU‐
1013 RITY are defined at compile time and they prevent the listing
1014 of open files.
1015
1016 +|-w Enables (+) or disables (-) the suppression of warning mes‐
1017 sages.
1018
1019 The lsof builder may choose to have warning messages disabled
1020 or enabled by default. The default warning message state is
1021 indicated in the output of the -h or -? option. Disabling
1022 warning messages when they are already disabled or enabling
1023 them when already enabled is acceptable.
1024
1025 The -t option selects the -w option.
1026
1027 -x [fl] may accompany the +d and +D options to direct their processing
1028 to cross over symbolic links and|or file system mount points
1029 encountered when scanning the directory (+d) or directory tree
1030 (+D).
1031
1032 If -x is specified by itself without a following parameter,
1033 cross-over processing of both symbolic links and file system
1034 mount points is enabled. Note that when -x is specified with‐
1035 out a parameter, the next argument must begin with '-' or '+'.
1036
1037 The optional 'f' parameter enables file system mount point
1038 cross-over processing; 'l', symbolic link cross-over process‐
1039 ing.
1040
1041 The -x option may not be supplied without also supplying a +d
1042 or +D option.
1043
1044 -X This is a dialect-specific option.
1045
1046 AIX:
1047 This IBM AIX RISC/System 6000 option requests the reporting of
1048 executed text file and shared library references.
1049
1050 WARNING: because this option uses the kernel readx() function,
1051 its use on a busy AIX system might cause an application
1052 process to hang so completely that it can neither be killed
1053 nor stopped. I have never seen this happen or had a report of
1054 its happening, but I think there is a remote possibility it
1055 could happen.
1056
1057 By default use of readx() is disabled. On AIX 5L and above
1058 lsof may need setuid-root permission to perform the actions
1059 this option requests.
1060
1061 The lsof builder may specify that the -X option be restricted
1062 to processes whose real UID is root. If that has been done,
1063 the -X option will not appear in the -h or -? help output
1064 unless the real UID of the lsof process is root. The default
1065 lsof distribution allows any UID to specify -X, so by default
1066 it will appear in the help output.
1067
1068 When AIX readx() use is disabled, lsof may not be able to
1069 report information for all text and loader file references,
1070 but it may also avoid exacerbating an AIX kernel directory
1071 search kernel error, known as the Stale Segment ID bug.
1072
1073 The readx() function, used by lsof or any other program to
1074 access some sections of kernel virtual memory, can trigger the
1075 Stale Segment ID bug. It can cause the kernel's dir_search()
1076 function to believe erroneously that part of an in-memory copy
1077 of a file system directory has been zeroed. Another applica‐
1078 tion process, distinct from lsof, asking the kernel to search
1079 the directory - e.g., by using open(2) - can cause
1080 dir_search() to loop forever, thus hanging the application
1081 process.
1082
1083 Consult the lsof FAQ (The FAQ section gives its location.)
1084 and the 00README file of the lsof distribution for a more com‐
1085 plete description of the Stale Segment ID bug, its APAR, and
1086 methods for defining readx() use when compiling lsof.
1087
1088 Linux:
1089 This Linux option requests that lsof skip the reporting of
1090 information on all open TCP, UDP and UDPLITE IPv4 and IPv6
1091 files.
1092
1093 This Linux option is most useful when the system has an
1094 extremely large number of open TCP, UDP and UDPLITE files, the
1095 processing of whose information in the /proc/net/tcp* and
1096 /proc/net/udp* files would take lsof a long time, and whose
1097 reporting is not of interest.
1098
1099 Use this option with care and only when you are sure that the
1100 information you want lsof to display isn't associated with
1101 open TCP, UDP or UDPLITE socket files.
1102
1103 Solaris 10 and above:
1104 This Solaris 10 and above option requests the reporting of
1105 cached paths for files that have been deleted - i.e., removed
1106 with rm(1) or unlink(2).
1107
1108 The cached path is followed by the string `` (deleted)'' to
1109 indicate that the path by which the file was opened has been
1110 deleted.
1111
1112 Because intervening changes made to the path - i.e., renames
1113 with mv(1) or rename(2) - are not recorded in the cached path,
1114 what lsof reports is only the path by which the file was
1115 opened, not its possibly different final path.
1116
1117 -z [z] specifies how Solaris 10 and higher zone information is to be
1118 handled.
1119
1120 Without a following argument - e.g., NO z - the option speci‐
1121 fies that zone names are to be listed in the ZONE output col‐
1122 umn.
1123
1124 The -z option may be followed by a zone name, z. That causes
1125 lsof to list only open files for processes in that zone. Mul‐
1126 tiple -z z option and argument pairs may be specified to form
1127 a list of named zones. Any open file of any process in any of
1128 the zones will be listed, subject to other conditions speci‐
1129 fied by other options and arguments.
1130
1131 -Z [Z] specifies how SELinux security contexts are to be handled. It
1132 and 'Z' field output character support are inhibited when
1133 SELinux is disabled in the running Linux kernel. See OUTPUT
1134 FOR OTHER PROGRAMS for more information on the 'Z' field out‐
1135 put character.
1136
1137 Without a following argument - e.g., NO Z - the option speci‐
1138 fies that security contexts are to be listed in the SECU‐
1139 RITY-CONTEXT output column.
1140
1141 The -Z option may be followed by a wildcard security context
1142 name, Z. That causes lsof to list only open files for pro‐
1143 cesses in that security context. Multiple -Z Z option and
1144 argument pairs may be specified to form a list of security
1145 contexts. Any open file of any process in any of the security
1146 contexts will be listed, subject to other conditions specified
1147 by other options and arguments. Note that Z can be A:B:C or
1148 *:B:C or A:B:* or *:*:C to match against the A:B:C context.
1149
1150 -- The double minus sign option is a marker that signals the end
1151 of the keyed options. It may be used, for example, when the
1152 first file name begins with a minus sign. It may also be used
1153 when the absence of a value for the last keyed option must be
1154 signified by the presence of a minus sign in the following
1155 option and before the start of the file names.
1156
1157 names These are path names of specific files to list. Symbolic
1158 links are resolved before use. The first name may be sepa‐
1159 rated from the preceding options with the ``--'' option.
1160
1161 If a name is the mounted-on directory of a file system or the
1162 device of the file system, lsof will list all the files open
1163 on the file system. To be considered a file system, the name
1164 must match a mounted-on directory name in mount(8) output, or
1165 match the name of a block device associated with a mounted-on
1166 directory name. The +|-f option may be used to force lsof to
1167 consider a name a file system identifier (+f) or a simple file
1168 (-f).
1169
1170 If name is a path to a directory that is not the mounted-on
1171 directory name of a file system, it is treated just as a regu‐
1172 lar file is treated - i.e., its listing is restricted to pro‐
1173 cesses that have it open as a file or as a process-specific
1174 directory, such as the root or current working directory. To
1175 request that lsof look for open files inside a directory name,
1176 use the +d s and +D D options.
1177
1178 If a name is the base name of a family of multiplexed files -
1179 e.g, AIX's /dev/pt[cs] - lsof will list all the associated
1180 multiplexed files on the device that are open - e.g.,
1181 /dev/pt[cs]/1, /dev/pt[cs]/2, etc.
1182
1183 If a name is a UNIX domain socket name, lsof will usually
1184 search for it by the characters of the name alone - exactly as
1185 it is specified and is recorded in the kernel socket struc‐
1186 ture. (See the next paragraph for an exception to that rule
1187 for Linux.) Specifying a relative path - e.g., ./file - in
1188 place of the file's absolute path - e.g., /tmp/file - won't
1189 work because lsof must match the characters you specify with
1190 what it finds in the kernel UNIX domain socket structures.
1191
1192 If a name is a Linux UNIX domain socket name, in one case lsof
1193 is able to search for it by its device and inode number,
1194 allowing name to be a relative path. The case requires that
1195 the absolute path -- i.e., one beginning with a slash ('/') be
1196 used by the process that created the socket, and hence be
1197 stored in the /proc/net/unix file; and it requires that lsof
1198 be able to obtain the device and node numbers of both the
1199 absolute path in /proc/net/unix and name via successful
1200 stat(2) system calls. When those conditions are met, lsof
1201 will be able to search for the UNIX domain socket when some
1202 path to it is is specified in name. Thus, for example, if the
1203 path is /dev/log, and an lsof search is initiated when the
1204 working directory is /dev, then name could be ./log.
1205
1206 If a name is none of the above, lsof will list any open files
1207 whose device and inode match that of the specified path name.
1208
1209 If you have also specified the -b option, the only names you
1210 may safely specify are file systems for which your mount table
1211 supplies alternate device numbers. See the AVOIDING KERNEL
1212 BLOCKS and ALTERNATE DEVICE NUMBERS sections for more informa‐
1213 tion.
1214
1215 Multiple file names are joined in a single ORed set before
1216 participating in AND option selection.
1217
1219 Lsof supports the recognition of AFS files for these dialects (and AFS
1220 versions):
1221
1222 AIX 4.1.4 (AFS 3.4a)
1223 HP-UX 9.0.5 (AFS 3.4a)
1224 Linux 1.2.13 (AFS 3.3)
1225 Solaris 2.[56] (AFS 3.4a)
1226
1227 It may recognize AFS files on other versions of these dialects, but has
1228 not been tested there. Depending on how AFS is implemented, lsof may
1229 recognize AFS files in other dialects, or may have difficulties recog‐
1230 nizing AFS files in the supported dialects.
1231
1232 Lsof may have trouble identifying all aspects of AFS files in supported
1233 dialects when AFS kernel support is implemented via dynamic modules
1234 whose addresses do not appear in the kernel's variable name list. In
1235 that case, lsof may have to guess at the identity of AFS files, and
1236 might not be able to obtain volume information from the kernel that is
1237 needed for calculating AFS volume node numbers. When lsof can't com‐
1238 pute volume node numbers, it reports blank in the NODE column.
1239
1240 The -A A option is available in some dialect implementations of lsof
1241 for specifying the name list file where dynamic module kernel addresses
1242 may be found. When this option is available, it will be listed in the
1243 lsof help output, presented in response to the -h or -?
1244
1245 See the lsof FAQ (The FAQ section gives its location.) for more infor‐
1246 mation about dynamic modules, their symbols, and how they affect lsof
1247 options.
1248
1249 Because AFS path lookups don't seem to participate in the kernel's name
1250 cache operations, lsof can't identify path name components for AFS
1251 files.
1252
1254 Lsof has three features that may cause security concerns. First, its
1255 default compilation mode allows anyone to list all open files with it.
1256 Second, by default it creates a user-readable and user-writable device
1257 cache file in the home directory of the real user ID that executes
1258 lsof. (The list-all-open-files and device cache features may be dis‐
1259 abled when lsof is compiled.) Third, its -k and -m options name alter‐
1260 nate kernel name list or memory files.
1261
1262 Restricting the listing of all open files is controlled by the com‐
1263 pile-time HASSECURITY and HASNOSOCKSECURITY options. When HASSECURITY
1264 is defined, lsof will allow only the root user to list all open files.
1265 The non-root user may list only open files of processes with the same
1266 user IDentification number as the real user ID number of the lsof
1267 process (the one that its user logged on with).
1268
1269 However, if HASSECURITY and HASNOSOCKSECURITY are both defined, anyone
1270 may list open socket files, provided they are selected with the -i
1271 option.
1272
1273 When HASSECURITY is not defined, anyone may list all open files.
1274
1275 Help output, presented in response to the -h or -? option, gives the
1276 status of the HASSECURITY and HASNOSOCKSECURITY definitions.
1277
1278 See the Security section of the 00README file of the lsof distribution
1279 for information on building lsof with the HASSECURITY and HASNOSOCKSE‐
1280 CURITY options enabled.
1281
1282 Creation and use of a user-readable and user-writable device cache file
1283 is controlled by the compile-time HASDCACHE option. See the DEVICE
1284 CACHE FILE section and the sections that follow it for details on how
1285 its path is formed. For security considerations it is important to
1286 note that in the default lsof distribution, if the real user ID under
1287 which lsof is executed is root, the device cache file will be written
1288 in root's home directory - e.g., / or /root. When HASDCACHE is not
1289 defined, lsof does not write or attempt to read a device cache file.
1290
1291 When HASDCACHE is defined, the lsof help output, presented in response
1292 to the -h, -D?, or -? options, will provide device cache file handling
1293 information. When HASDCACHE is not defined, the -h or -? output will
1294 have no -D option description.
1295
1296 Before you decide to disable the device cache file feature - enabling
1297 it improves the performance of lsof by reducing the startup overhead of
1298 examining all the nodes in /dev (or /devices) - read the discussion of
1299 it in the 00DCACHE file of the lsof distribution and the lsof FAQ (The
1300 FAQ section gives its location.)
1301
1302 WHEN IN DOUBT, YOU CAN TEMPORARILY DISABLE THE USE OF THE DEVICE CACHE
1303 FILE WITH THE -Di OPTION.
1304
1305 When lsof user declares alternate kernel name list or memory files with
1306 the -k and -m options, lsof checks the user's authority to read them
1307 with access(2). This is intended to prevent whatever special power
1308 lsof's modes might confer on it from letting it read files not normally
1309 accessible via the authority of the real user ID.
1310
1312 This section describes the information lsof lists for each open file.
1313 See the OUTPUT FOR OTHER PROGRAMS section for additional information on
1314 output that can be processed by another program.
1315
1316 Lsof only outputs printable (declared so by isprint(3)) 8 bit charac‐
1317 ters. Non-printable characters are printed in one of three forms: the
1318 C ``\[bfrnt]'' form; the control character `^' form (e.g., ``^@''); or
1319 hexadecimal leading ``\x'' form (e.g., ``\xab''). Space is non-print‐
1320 able in the COMMAND column (``\x20'') and printable elsewhere.
1321
1322 For some dialects - if HASSETLOCALE is defined in the dialect's
1323 machine.h header file - lsof will print the extended 8 bit characters
1324 of a language locale. The lsof process must be supplied a language
1325 locale environment variable (e.g., LANG) whose value represents a known
1326 language locale in which the extended characters are considered print‐
1327 able by isprint(3). Otherwise lsof considers the extended characters
1328 non-printable and prints them according to its rules for non-printable
1329 characters, stated above. Consult your dialect's setlocale(3) man page
1330 for the names of other environment variables that may be used in place
1331 of LANG - e.g., LC_ALL, LC_CTYPE, etc.
1332
1333 Lsof's language locale support for a dialect also covers wide charac‐
1334 ters - e.g., UTF-8 - when HASSETLOCALE and HASWIDECHAR are defined in
1335 the dialect's machine.h header file, and when a suitable language
1336 locale has been defined in the appropriate environment variable for the
1337 lsof process. Wide characters are printable under those conditions if
1338 iswprint(3) reports them to be. If HASSETLOCALE, HASWIDECHAR and a
1339 suitable language locale aren't defined, or if iswprint(3) reports wide
1340 characters that aren't printable, lsof considers the wide characters
1341 non-printable and prints each of their 8 bits according to its rules
1342 for non-printable characters, stated above.
1343
1344 Consult the answers to the "Language locale support" questions in the
1345 lsof FAQ (The FAQ section gives its location.) for more information.
1346
1347 Lsof dynamically sizes the output columns each time it runs, guarantee‐
1348 ing that each column is a minimum size. It also guarantees that each
1349 column is separated from its predecessor by at least one space.
1350
1351 COMMAND contains the first nine characters of the name of the UNIX
1352 command associated with the process. If a non-zero w value
1353 is specified to the +c w option, the column contains the
1354 first w characters of the name of the UNIX command associ‐
1355 ated with the process up to the limit of characters supplied
1356 to lsof by the UNIX dialect. (See the description of the +c
1357 w command or the lsof FAQ for more information. The FAQ
1358 section gives its location.)
1359
1360 If w is less than the length of the column title, ``COM‐
1361 MAND'', it will be raised to that length.
1362
1363 If a zero w value is specified to the +c w option, the col‐
1364 umn contains all the characters of the name of the UNIX com‐
1365 mand associated with the process.
1366
1367 All command name characters maintained by the kernel in its
1368 structures are displayed in field output when the command
1369 name descriptor (`c') is specified. See the OUTPUT FOR
1370 OTHER COMMANDS section for information on selecting field
1371 output and the associated command name descriptor.
1372
1373 PID is the Process IDentification number of the process.
1374
1375 TID is the task (thread) IDentification number, if task (thread)
1376 reporting is supported by the dialect and a task (thread) is
1377 being listed. (If help output - i.e., the output of the -h
1378 or -? options - shows this option, then task (thread)
1379 reporting is supported by the dialect.)
1380
1381 A blank TID column in Linux indicates a process - i.e., a
1382 non-task.
1383
1384 TASKCMD is the task command name. Generally this will be the same
1385 as the process named in the COMMAND column, but some task
1386 implementations (e.g., Linux) permit a task to change its
1387 command name.
1388
1389 The TASKCMD column width is subject to the same size limita‐
1390 tion as the COMMAND column.
1391
1392 ZONE is the Solaris 10 and higher zone name. This column must be
1393 selected with the -z option.
1394
1395 SECURITY-CONTEXT
1396 is the SELinux security context. This column must be
1397 selected with the -Z option. Note that the -Z option is
1398 inhibited when SELinux is disabled in the running Linux ker‐
1399 nel.
1400
1401 PPID is the Parent Process IDentification number of the process.
1402 It is only displayed when the -R option has been specified.
1403
1404 PGID is the process group IDentification number associated with
1405 the process. It is only displayed when the -g option has
1406 been specified.
1407
1408 USER is the user ID number or login name of the user to whom the
1409 process belongs, usually the same as reported by ps(1).
1410 However, on Linux USER is the user ID number or login that
1411 owns the directory in /proc where lsof finds information
1412 about the process. Usually that is the same value reported
1413 by ps(1), but may differ when the process has changed its
1414 effective user ID. (See the -l option description for
1415 information on when a user ID number or login name is dis‐
1416 played.)
1417
1418 FD is the File Descriptor number of the file or:
1419
1420 cwd current working directory;
1421 Lnn library references (AIX);
1422 err FD information error (see NAME column);
1423 jld jail directory (FreeBSD);
1424 ltx shared library text (code and data);
1425 Mxx hex memory-mapped type number xx.
1426 m86 DOS Merge mapped file;
1427 mem memory-mapped file;
1428 mmap memory-mapped device;
1429 pd parent directory;
1430 rtd root directory;
1431 tr kernel trace file (OpenBSD);
1432 txt program text (code and data);
1433 v86 VP/ix mapped file;
1434
1435 FD is followed by one of these characters, describing the
1436 mode under which the file is open:
1437
1438 r for read access;
1439 w for write access;
1440 u for read and write access;
1441 space if mode unknown and no lock
1442 character follows;
1443 `-' if mode unknown and lock
1444 character follows.
1445
1446 The mode character is followed by one of these lock charac‐
1447 ters, describing the type of lock applied to the file:
1448
1449 N for a Solaris NFS lock of unknown type;
1450 r for read lock on part of the file;
1451 R for a read lock on the entire file;
1452 w for a write lock on part of the file;
1453 W for a write lock on the entire file;
1454 u for a read and write lock of any length;
1455 U for a lock of unknown type;
1456 x for an SCO OpenServer Xenix lock on part of the
1457 file;
1458 X for an SCO OpenServer Xenix lock on the entire file;
1459 space if there is no lock.
1460
1461 See the LOCKS section for more information on the lock
1462 information character.
1463
1464 The FD column contents constitutes a single field for pars‐
1465 ing in post-processing scripts.
1466
1467 TYPE is the type of the node associated with the file - e.g.,
1468 GDIR, GREG, VDIR, VREG, etc.
1469
1470 or ``IPv4'' for an IPv4 socket;
1471
1472 or ``IPv6'' for an open IPv6 network file - even if its
1473 address is IPv4, mapped in an IPv6 address;
1474
1475 or ``ax25'' for a Linux AX.25 socket;
1476
1477 or ``inet'' for an Internet domain socket;
1478
1479 or ``lla'' for a HP-UX link level access file;
1480
1481 or ``rte'' for an AF_ROUTE socket;
1482
1483 or ``sock'' for a socket of unknown domain;
1484
1485 or ``unix'' for a UNIX domain socket;
1486
1487 or ``x.25'' for an HP-UX x.25 socket;
1488
1489 or ``BLK'' for a block special file;
1490
1491 or ``CHR'' for a character special file;
1492
1493 or ``DEL'' for a Linux map file that has been deleted;
1494
1495 or ``DIR'' for a directory;
1496
1497 or ``DOOR'' for a VDOOR file;
1498
1499 or ``FIFO'' for a FIFO special file;
1500
1501 or ``KQUEUE'' for a BSD style kernel event queue file;
1502
1503 or ``LINK'' for a symbolic link file;
1504
1505 or ``MPB'' for a multiplexed block file;
1506
1507 or ``MPC'' for a multiplexed character file;
1508
1509 or ``NOFD'' for a Linux /proc/<PID>/fd directory that can't
1510 be opened -- the directory path appears in the NAME column,
1511 followed by an error message;
1512
1513 or ``PAS'' for a /proc/as file;
1514
1515 or ``PAXV'' for a /proc/auxv file;
1516
1517 or ``PCRE'' for a /proc/cred file;
1518
1519 or ``PCTL'' for a /proc control file;
1520
1521 or ``PCUR'' for the current /proc process;
1522
1523 or ``PCWD'' for a /proc current working directory;
1524
1525 or ``PDIR'' for a /proc directory;
1526
1527 or ``PETY'' for a /proc executable type (etype);
1528
1529 or ``PFD'' for a /proc file descriptor;
1530
1531 or ``PFDR'' for a /proc file descriptor directory;
1532
1533 or ``PFIL'' for an executable /proc file;
1534
1535 or ``PFPR'' for a /proc FP register set;
1536
1537 or ``PGD'' for a /proc/pagedata file;
1538
1539 or ``PGID'' for a /proc group notifier file;
1540
1541 or ``PIPE'' for pipes;
1542
1543 or ``PLC'' for a /proc/lwpctl file;
1544
1545 or ``PLDR'' for a /proc/lpw directory;
1546
1547 or ``PLDT'' for a /proc/ldt file;
1548
1549 or ``PLPI'' for a /proc/lpsinfo file;
1550
1551 or ``PLST'' for a /proc/lstatus file;
1552
1553 or ``PLU'' for a /proc/lusage file;
1554
1555 or ``PLWG'' for a /proc/gwindows file;
1556
1557 or ``PLWI'' for a /proc/lwpsinfo file;
1558
1559 or ``PLWS'' for a /proc/lwpstatus file;
1560
1561 or ``PLWU'' for a /proc/lwpusage file;
1562
1563 or ``PLWX'' for a /proc/xregs file;
1564
1565 or ``PMAP'' for a /proc map file (map);
1566
1567 or ``PMEM'' for a /proc memory image file;
1568
1569 or ``PNTF'' for a /proc process notifier file;
1570
1571 or ``POBJ'' for a /proc/object file;
1572
1573 or ``PODR'' for a /proc/object directory;
1574
1575 or ``POLP'' for an old format /proc light weight process
1576 file;
1577
1578 or ``POPF'' for an old format /proc PID file;
1579
1580 or ``POPG'' for an old format /proc page data file;
1581
1582 or ``PORT'' for a SYSV named pipe;
1583
1584 or ``PREG'' for a /proc register file;
1585
1586 or ``PRMP'' for a /proc/rmap file;
1587
1588 or ``PRTD'' for a /proc root directory;
1589
1590 or ``PSGA'' for a /proc/sigact file;
1591
1592 or ``PSIN'' for a /proc/psinfo file;
1593
1594 or ``PSTA'' for a /proc status file;
1595
1596 or ``PSXSEM'' for a POSIX semaphore file;
1597
1598 or ``PSXSHM'' for a POSIX shared memory file;
1599
1600 or ``PTS'' for a /dev/pts file;
1601
1602 or ``PUSG'' for a /proc/usage file;
1603
1604 or ``PW'' for a /proc/watch file;
1605
1606 or ``PXMP'' for a /proc/xmap file;
1607
1608 or ``REG'' for a regular file;
1609
1610 or ``SMT'' for a shared memory transport file;
1611
1612 or ``STSO'' for a stream socket;
1613
1614 or ``UNNM'' for an unnamed type file;
1615
1616 or ``XNAM'' for an OpenServer Xenix special file of unknown
1617 type;
1618
1619 or ``XSEM'' for an OpenServer Xenix semaphore file;
1620
1621 or ``XSD'' for an OpenServer Xenix shared data file;
1622
1623 or the four type number octets if the corresponding name
1624 isn't known.
1625
1626 FILE-ADDR contains the kernel file structure address when f has been
1627 specified to +f;
1628
1629 FCT contains the file reference count from the kernel file
1630 structure when c has been specified to +f;
1631
1632 FILE-FLAG when g or G has been specified to +f, this field contains
1633 the contents of the f_flag[s] member of the kernel file
1634 structure and the kernel's per-process open file flags (if
1635 available); `G' causes them to be displayed in hexadecimal;
1636 `g', as short-hand names; two lists may be displayed with
1637 entries separated by commas, the lists separated by a semi‐
1638 colon (`;'); the first list may contain short-hand names for
1639 f_flag[s] values from the following table:
1640
1641 AIO asynchronous I/O (e.g., FAIO)
1642 AP append
1643 ASYN asynchronous I/O (e.g., FASYNC)
1644 BAS block, test, and set in use
1645 BKIU block if in use
1646 BL use block offsets
1647 BSK block seek
1648 CA copy avoid
1649 CIO concurrent I/O
1650 CLON clone
1651 CLRD CL read
1652 CR create
1653 DF defer
1654 DFI defer IND
1655 DFLU data flush
1656 DIR direct
1657 DLY delay
1658 DOCL do clone
1659 DSYN data-only integrity
1660 DTY must be a directory
1661 EVO event only
1662 EX open for exec
1663 EXCL exclusive open
1664 FSYN synchronous writes
1665 GCDF defer during unp_gc() (AIX)
1666 GCMK mark during unp_gc() (AIX)
1667 GTTY accessed via /dev/tty
1668 HUP HUP in progress
1669 KERN kernel
1670 KIOC kernel-issued ioctl
1671 LCK has lock
1672 LG large file
1673 MBLK stream message block
1674 MK mark
1675 MNT mount
1676 MSYN multiplex synchronization
1677 NATM don't update atime
1678 NB non-blocking I/O
1679 NBDR no BDRM check
1680 NBIO SYSV non-blocking I/O
1681 NBF n-buffering in effect
1682 NC no cache
1683 ND no delay
1684 NDSY no data synchronization
1685 NET network
1686 NFLK don't follow links
1687 NMFS NM file system
1688 NOTO disable background stop
1689 NSH no share
1690 NTTY no controlling TTY
1691 OLRM OLR mirror
1692 PAIO POSIX asynchronous I/O
1693 PP POSIX pipe
1694 R read
1695 RC file and record locking cache
1696 REV revoked
1697 RSH shared read
1698 RSYN read synchronization
1699 RW read and write access
1700 SL shared lock
1701 SNAP cooked snapshot
1702 SOCK socket
1703 SQSH Sequent shared set on open
1704 SQSV Sequent SVM set on open
1705 SQR Sequent set repair on open
1706 SQS1 Sequent full shared open
1707 SQS2 Sequent partial shared open
1708 STPI stop I/O
1709 SWR synchronous read
1710 SYN file integrity while writing
1711 TCPM avoid TCP collision
1712 TR truncate
1713 W write
1714 WKUP parallel I/O synchronization
1715 WTG parallel I/O synchronization
1716 VH vhangup pending
1717 VTXT virtual text
1718 XL exclusive lock
1719
1720 this list of names was derived from F* #define's in dialect
1721 header files <fcntl.h>, <linux</fs.h>, <sys/fcntl.c>,
1722 <sys/fcntlcom.h>, and <sys/file.h>; see the lsof.h header
1723 file for a list showing the correspondence between the above
1724 short-hand names and the header file definitions;
1725
1726 the second list (after the semicolon) may contain short-hand
1727 names for kernel per-process open file flags from this ta‐
1728 ble:
1729
1730 ALLC allocated
1731 BR the file has been read
1732 BHUP activity stopped by SIGHUP
1733 BW the file has been written
1734 CLSG closing
1735 CX close-on-exec (see fcntl(F_SETFD))
1736 LCK lock was applied
1737 MP memory-mapped
1738 OPIP open pending - in progress
1739 RSVW reserved wait
1740 SHMT UF_FSHMAT set (AIX)
1741 USE in use (multi-threaded)
1742
1743 NODE-ID (or INODE-ADDR for some dialects) contains a unique identi‐
1744 fier for the file node (usually the kernel vnode or inode
1745 address, but also occasionally a concatenation of device and
1746 node number) when n has been specified to +f;
1747
1748 DEVICE contains the device numbers, separated by commas, for a
1749 character special, block special, regular, directory or NFS
1750 file;
1751
1752 or ``memory'' for a memory file system node under Tru64
1753 UNIX;
1754
1755 or the address of the private data area of a Solaris socket
1756 stream;
1757
1758 or a kernel reference address that identifies the file (The
1759 kernel reference address may be used for FIFO's, for exam‐
1760 ple.);
1761
1762 or the base address or device name of a Linux AX.25 socket
1763 device.
1764
1765 Usually only the lower thirty two bits of Tru64 UNIX kernel
1766 addresses are displayed.
1767
1768 SIZE, SIZE/OFF, or OFFSET
1769 is the size of the file or the file offset in bytes. A
1770 value is displayed in this column only if it is available.
1771 Lsof displays whatever value - size or offset - is appropri‐
1772 ate for the type of the file and the version of lsof.
1773
1774 On some UNIX dialects lsof can't obtain accurate or consis‐
1775 tent file offset information from its kernel data sources,
1776 sometimes just for particular kinds of files (e.g., socket
1777 files.) In other cases, files don't have true sizes - e.g.,
1778 sockets, FIFOs, pipes - so lsof displays for their sizes the
1779 content amounts it finds in their kernel buffer descriptors
1780 (e.g., socket buffer size counts or TCP/IP window sizes.)
1781 Consult the lsof FAQ (The FAQ section gives its location.)
1782 for more information.
1783
1784 The file size is displayed in decimal; the offset is nor‐
1785 mally displayed in decimal with a leading ``0t'' if it con‐
1786 tains 8 digits or less; in hexadecimal with a leading ``0x''
1787 if it is longer than 8 digits. (Consult the -o o option
1788 description for information on when 8 might default to some
1789 other value.)
1790
1791 Thus the leading ``0t'' and ``0x'' identify an offset when
1792 the column may contain both a size and an offset (i.e., its
1793 title is SIZE/OFF).
1794
1795 If the -o option is specified, lsof always displays the file
1796 offset (or nothing if no offset is available) and labels the
1797 column OFFSET. The offset always begins with ``0t'' or
1798 ``0x'' as described above.
1799
1800 The lsof user can control the switch from ``0t'' to ``0x''
1801 with the -o o option. Consult its description for more
1802 information.
1803
1804 If the -s option is specified, lsof always displays the file
1805 size (or nothing if no size is available) and labels the
1806 column SIZE. The -o and -s options are mutually exclusive;
1807 they can't both be specified.
1808
1809 For files that don't have a fixed size - e.g., don't reside
1810 on a disk device - lsof will display appropriate information
1811 about the current size or position of the file if it is
1812 available in the kernel structures that define the file.
1813
1814 NLINK contains the file link count when +L has been specified;
1815
1816 NODE is the node number of a local file;
1817
1818 or the inode number of an NFS file in the server host;
1819
1820 or the Internet protocol type - e.g, ``TCP'';
1821
1822 or ``STR'' for a stream;
1823
1824 or ``CCITT'' for an HP-UX x.25 socket;
1825
1826 or the IRQ or inode number of a Linux AX.25 socket device.
1827
1828 NAME is the name of the mount point and file system on which the
1829 file resides;
1830
1831 or the name of a file specified in the names option (after
1832 any symbolic links have been resolved);
1833
1834 or the name of a character special or block special device;
1835
1836 or the local and remote Internet addresses of a network
1837 file; the local host name or IP number is followed by a
1838 colon (':'), the port, ``->'', and the two-part remote
1839 address; IP addresses may be reported as numbers or names,
1840 depending on the +|-M, -n, and -P options; colon-separated
1841 IPv6 numbers are enclosed in square brackets; IPv4
1842 INADDR_ANY and IPv6 IN6_IS_ADDR_UNSPECIFIED addresses, and
1843 zero port numbers are represented by an asterisk ('*'); a
1844 UDP destination address may be followed by the amount of
1845 time elapsed since the last packet was sent to the destina‐
1846 tion; TCP, UDP and UDPLITE remote addresses may be followed
1847 by TCP/TPI information in parentheses - state (e.g.,
1848 ``(ESTABLISHED)'', ``(Unbound)''), queue sizes, and window
1849 sizes (not all dialects) - in a fashion similar to what net‐
1850 stat(1) reports; see the -T option description or the
1851 description of the TCP/TPI field in OUTPUT FOR OTHER PRO‐
1852 GRAMS for more information on state, queue size, and window
1853 size;
1854
1855 or the address or name of a UNIX domain socket, possibly
1856 including a stream clone device name, a file system object's
1857 path name, local and foreign kernel addresses, socket pair
1858 information, and a bound vnode address;
1859
1860 or the local and remote mount point names of an NFS file;
1861
1862 or ``STR'', followed by the stream name;
1863
1864 or a stream character device name, followed by ``->'' and
1865 the stream name or a list of stream module names, separated
1866 by ``->'';
1867
1868 or ``STR:'' followed by the SCO OpenServer stream device and
1869 module names, separated by ``->'';
1870
1871 or system directory name, `` -- '', and as many components
1872 of the path name as lsof can find in the kernel's name cache
1873 for selected dialects (See the KERNEL NAME CACHE section for
1874 more information.);
1875
1876 or ``PIPE->'', followed by a Solaris kernel pipe destination
1877 address;
1878
1879 or ``COMMON:'', followed by the vnode device information
1880 structure's device name, for a Solaris common vnode;
1881
1882 or the address family, followed by a slash (`/'), followed
1883 by fourteen comma-separated bytes of a non-Internet raw
1884 socket address;
1885
1886 or the HP-UX x.25 local address, followed by the virtual
1887 connection number (if any), followed by the remote address
1888 (if any);
1889
1890 or ``(dead)'' for disassociated Tru64 UNIX files - typically
1891 terminal files that have been flagged with the TIOCNOTTY
1892 ioctl and closed by daemons;
1893
1894 or ``rd=<offset>'' and ``wr=<offset>'' for the values of the
1895 read and write offsets of a FIFO;
1896
1897 or ``clone n:/dev/event'' for SCO OpenServer file clones of
1898 the /dev/event device, where n is the minor device number of
1899 the file;
1900
1901 or ``(socketpair: n)'' for a Solaris 2.6, 8, 9 or 10 UNIX
1902 domain socket, created by the socketpair(3N) network func‐
1903 tion;
1904
1905 or ``no PCB'' for socket files that do not have a protocol
1906 block associated with them, optionally followed by ``,
1907 CANTSENDMORE'' if sending on the socket has been disabled,
1908 or ``, CANTRCVMORE'' if receiving on the socket has been
1909 disabled (e.g., by the shutdown(2) function);
1910
1911 or the local and remote addresses of a Linux IPX socket file
1912 in the form <net>:[<node>:]<port>, followed in parentheses
1913 by the transmit and receive queue sizes, and the connection
1914 state;
1915
1916 or ``dgram'' or ``stream'' for the type UnixWare 7.1.1 and
1917 above in-kernel UNIX domain sockets, followed by a colon
1918 (':') and the local path name when available, followed by
1919 ``->'' and the remote path name or kernel socket address in
1920 hexadecimal when available;
1921
1922 or the association value, association index, endpoint value,
1923 local address, local port, remote address and remote port
1924 for Linux SCTP sockets;
1925
1926 or ``protocol: '' followed by the Linux socket's protocol
1927 attribute.
1928
1929 For dialects that support a ``namefs'' file system, allowing one file
1930 to be attached to another with fattach(3C), lsof will add
1931 ``(FA:<address1><direction><address2>)'' to the NAME column.
1932 <address1> and <address2> are hexadecimal vnode addresses. <direction>
1933 will be ``<-'' if <address2> has been fattach'ed to this vnode whose
1934 address is <address1>; and ``->'' if <address1>, the vnode address of
1935 this vnode, has been fattach'ed to <address2>. <address1> may be omit‐
1936 ted if it already appears in the DEVICE column.
1937
1938 Lsof may add two parenthetical notes to the NAME column for open
1939 Solaris 10 files: ``(?)'' if lsof considers the path name of question‐
1940 able accuracy; and ``(deleted)'' if the -X option has been specified
1941 and lsof detects the open file's path name has been deleted. Consult
1942 the lsof FAQ (The FAQ section gives its location.) for more informa‐
1943 tion on these NAME column additions.
1944
1946 Lsof can't adequately report the wide variety of UNIX dialect file
1947 locks in a single character. What it reports in a single character is
1948 a compromise between the information it finds in the kernel and the
1949 limitations of the reporting format.
1950
1951 Moreover, when a process holds several byte level locks on a file, lsof
1952 only reports the status of the first lock it encounters. If it is a
1953 byte level lock, then the lock character will be reported in lower case
1954 - i.e., `r', `w', or `x' - rather than the upper case equivalent
1955 reported for a full file lock.
1956
1957 Generally lsof can only report on locks held by local processes on
1958 local files. When a local process sets a lock on a remotely mounted
1959 (e.g., NFS) file, the remote server host usually records the lock
1960 state. One exception is Solaris - at some patch levels of 2.3, and in
1961 all versions above 2.4, the Solaris kernel records information on
1962 remote locks in local structures.
1963
1964 Lsof has trouble reporting locks for some UNIX dialects. Consult the
1965 BUGS section of this manual page or the lsof FAQ (The FAQ section gives
1966 its location.) for more information.
1967
1969 When the -F option is specified, lsof produces output that is suitable
1970 for processing by another program - e.g, an awk or Perl script, or a C
1971 program.
1972
1973 Each unit of information is output in a field that is identified with a
1974 leading character and terminated by a NL (012) (or a NUL (000) if the 0
1975 (zero) field identifier character is specified.) The data of the field
1976 follows immediately after the field identification character and
1977 extends to the field terminator.
1978
1979 It is possible to think of field output as process and file sets. A
1980 process set begins with a field whose identifier is `p' (for process
1981 IDentifier (PID)). It extends to the beginning of the next PID field
1982 or the beginning of the first file set of the process, whichever comes
1983 first. Included in the process set are fields that identify the com‐
1984 mand, the process group IDentification (PGID) number, the task (thread)
1985 ID (TID), and the user ID (UID) number or login name.
1986
1987 A file set begins with a field whose identifier is `f' (for file
1988 descriptor). It is followed by lines that describe the file's access
1989 mode, lock state, type, device, size, offset, inode, protocol, name and
1990 stream module names. It extends to the beginning of the next file or
1991 process set, whichever comes first.
1992
1993 When the NUL (000) field terminator has been selected with the 0 (zero)
1994 field identifier character, lsof ends each process and file set with a
1995 NL (012) character.
1996
1997 Lsof always produces one field, the PID (`p') field. All other fields
1998 may be declared optionally in the field identifier character list that
1999 follows the -F option. When a field selection character identifies an
2000 item lsof does not normally list - e.g., PPID, selected with -R - spec‐
2001 ification of the field character - e.g., ``-FR'' - also selects the
2002 listing of the item.
2003
2004 It is entirely possible to select a set of fields that cannot easily be
2005 parsed - e.g., if the field descriptor field is not selected, it may be
2006 difficult to identify file sets. To help you avoid this difficulty,
2007 lsof supports the -F option; it selects the output of all fields with
2008 NL terminators (the -F0 option pair selects the output of all fields
2009 with NUL terminators). For compatibility reasons neither -F nor -F0
2010 select the raw device field.
2011
2012 These are the fields that lsof will produce. The single character
2013 listed first is the field identifier.
2014
2015 a file access mode
2016 c process command name (all characters from proc or
2017 user structure)
2018 C file structure share count
2019 d file's device character code
2020 D file's major/minor device number (0x<hexadecimal>)
2021 f file descriptor (always selected)
2022 F file structure address (0x<hexadecimal>)
2023 G file flaGs (0x<hexadecimal>; names if +fg follows)
2024 g process group ID
2025 i file's inode number
2026 K tasK ID
2027 k link count
2028 l file's lock status
2029 L process login name
2030 m marker between repeated output
2031 M the task comMand name
2032 n file name, comment, Internet address
2033 N node identifier (ox<hexadecimal>
2034 o file's offset (decimal)
2035 p process ID (always selected)
2036 P protocol name
2037 r raw device number (0x<hexadecimal>)
2038 R parent process ID
2039 s file's size (decimal)
2040 S file's stream identification
2041 t file's type
2042 T TCP/TPI information, identified by prefixes (the
2043 `=' is part of the prefix):
2044 QR=<read queue size>
2045 QS=<send queue size>
2046 SO=<socket options and values> (not all dialects)
2047 SS=<socket states> (not all dialects)
2048 ST=<connection state>
2049 TF=<TCP flags and values> (not all dialects)
2050 WR=<window read size> (not all dialects)
2051 WW=<window write size> (not all dialects)
2052 (TCP/TPI information isn't reported for all supported
2053 UNIX dialects. The -h or -? help output for the
2054 -T option will show what TCP/TPI reporting can be
2055 requested.)
2056 u process user ID
2057 z Solaris 10 and higher zone name
2058 Z SELinux security context (inhibited when SELinux is disabled)
2059 0 use NUL field terminator character in place of NL
2060 1-9 dialect-specific field identifiers (The output
2061 of -F? identifies the information to be found
2062 in dialect-specific fields.)
2063
2064 You can get on-line help information on these characters and their
2065 descriptions by specifying the -F? option pair. (Escape the `?' char‐
2066 acter as your shell requires.) Additional information on field content
2067 can be found in the OUTPUT section.
2068
2069 As an example, ``-F pcfn'' will select the process ID (`p'), command
2070 name (`c'), file descriptor (`f') and file name (`n') fields with an NL
2071 field terminator character; ``-F pcfn0'' selects the same output with a
2072 NUL (000) field terminator character.
2073
2074 Lsof doesn't produce all fields for every process or file set, only
2075 those that are available. Some fields are mutually exclusive: file
2076 device characters and file major/minor device numbers; file inode num‐
2077 ber and protocol name; file name and stream identification; file size
2078 and offset. One or the other member of these mutually exclusive sets
2079 will appear in field output, but not both.
2080
2081 Normally lsof ends each field with a NL (012) character. The 0 (zero)
2082 field identifier character may be specified to change the field termi‐
2083 nator character to a NUL (000). A NUL terminator may be easier to
2084 process with xargs [4m(1), for example, or with programs whose quoting
2085 mechanisms may not easily cope with the range of characters in the
2086 field output. When the NUL field terminator is in use, lsof ends each
2087 process and file set with a NL (012).
2088
2089 Three aids to producing programs that can process lsof field output are
2090 included in the lsof distribution. The first is a C header file,
2091 lsof_fields.h, that contains symbols for the field identification char‐
2092 acters, indexes for storing them in a table, and explanation strings
2093 that may be compiled into programs. Lsof uses this header file.
2094
2095 The second aid is a set of sample scripts that process field output,
2096 written in awk, Perl 4, and Perl 5. They're located in the scripts
2097 subdirectory of the lsof distribution.
2098
2099 The third aid is the C library used for the lsof test suite. The test
2100 suite is written in C and uses field output to validate the correct
2101 operation of lsof. The library can be found in the tests/LTlib.c file
2102 of the lsof distribution. The library uses the first aid, the
2103 lsof_fields.h header file.
2104
2106 Lsof can be blocked by some kernel functions that it uses - lstat(2),
2107 readlink(2), and stat(2). These functions are stalled in the kernel,
2108 for example, when the hosts where mounted NFS file systems reside
2109 become inaccessible.
2110
2111 Lsof attempts to break these blocks with timers and child processes,
2112 but the techniques are not wholly reliable. When lsof does manage to
2113 break a block, it will report the break with an error message. The
2114 messages may be suppressed with the -t and -w options.
2115
2116 The default timeout value may be displayed with the -h or -? option,
2117 and it may be changed with the -S [t] option. The minimum for t is two
2118 seconds, but you should avoid small values, since slow system respon‐
2119 siveness can cause short timeouts to expire unexpectedly and perhaps
2120 stop lsof before it can produce any output.
2121
2122 When lsof has to break a block during its access of mounted file system
2123 information, it normally continues, although with less information
2124 available to display about open files.
2125
2126 Lsof can also be directed to avoid the protection of timers and child
2127 processes when using the kernel functions that might block by specify‐
2128 ing the -O option. While this will allow lsof to start up with less
2129 overhead, it exposes lsof completely to the kernel situations that
2130 might block it. Use this option cautiously.
2131
2133 You can use the -b option to tell lsof to avoid using kernel functions
2134 that would block. Some cautions apply.
2135
2136 First, using this option usually requires that your system supply
2137 alternate device numbers in place of the device numbers that lsof would
2138 normally obtain with the lstat(2) and stat(2) kernel functions. See
2139 the ALTERNATE DEVICE NUMBERS section for more information on alternate
2140 device numbers.
2141
2142 Second, you can't specify names for lsof to locate unless they're file
2143 system names. This is because lsof needs to know the device and inode
2144 numbers of files listed with names in the lsof options, and the -b
2145 option prevents lsof from obtaining them. Moreover, since lsof only
2146 has device numbers for the file systems that have alternates, its abil‐
2147 ity to locate files on file systems depends completely on the avail‐
2148 ability and accuracy of the alternates. If no alternates are avail‐
2149 able, or if they're incorrect, lsof won't be able to locate files on
2150 the named file systems.
2151
2152 Third, if the names of your file system directories that lsof obtains
2153 from your system's mount table are symbolic links, lsof won't be able
2154 to resolve the links. This is because the -b option causes lsof to
2155 avoid the kernel readlink(2) function it uses to resolve symbolic
2156 links.
2157
2158 Finally, using the -b option causes lsof to issue warning messages when
2159 it needs to use the kernel functions that the -b option directs it to
2160 avoid. You can suppress these messages by specifying the -w option,
2161 but if you do, you won't see the alternate device numbers reported in
2162 the warning messages.
2163
2165 On some dialects, when lsof has to break a block because it can't get
2166 information about a mounted file system via the lstat(2) and stat(2)
2167 kernel functions, or because you specified the -b option, lsof can
2168 obtain some of the information it needs - the device number and possi‐
2169 bly the file system type - from the system mount table. When that is
2170 possible, lsof will report the device number it obtained. (You can
2171 suppress the report by specifying the -w option.)
2172
2173 You can assist this process if your mount table is supported with an
2174 /etc/mtab or /etc/mnttab file that contains an options field by adding
2175 a ``dev=xxxx'' field for mount points that do not have one in their
2176 options strings. Note: you must be able to edit the file - i.e., some
2177 mount tables like recent Solaris /etc/mnttab or Linux /proc/mounts are
2178 read-only and can't be modified.
2179
2180 You may also be able to supply device numbers using the +m and +m m
2181 options, provided they are supported by your dialect. Check the output
2182 of lsof's -h or -? options to see if the +m and +m m options are
2183 available.
2184
2185 The ``xxxx'' portion of the field is the hexadecimal value of the file
2186 system's device number. (Consult the st_dev field of the output of the
2187 lstat(2) and stat(2) functions for the appropriate values for your file
2188 systems.) Here's an example from a Sun Solaris 2.6 /etc/mnttab for a
2189 file system remotely mounted via NFS:
2190
2191 nfs ignore,noquota,dev=2a40001
2192
2193 There's an advantage to having ``dev=xxxx'' entries in your mount table
2194 file, especially for file systems that are mounted from remote NFS
2195 servers. When a remote server crashes and you want to identify its
2196 users by running lsof on one of its clients, lsof probably won't be
2197 able to get output from the lstat(2) and stat(2) functions for the file
2198 system. If it can obtain the file system's device number from the
2199 mount table, it will be able to display the files open on the crashed
2200 NFS server.
2201
2202 Some dialects that do not use an ASCII /etc/mtab or /etc/mnttab file
2203 for the mount table may still provide an alternative device number in
2204 their internal mount tables. This includes AIX, Apple Darwin, FreeBSD,
2205 NetBSD, OpenBSD, and Tru64 UNIX. Lsof knows how to obtain the alterna‐
2206 tive device number for these dialects and uses it when its attempt to
2207 lstat(2) or stat(2) the file system is blocked.
2208
2209 If you're not sure your dialect supplies alternate device numbers for
2210 file systems from its mount table, use this lsof incantation to see if
2211 it reports any alternate device numbers:
2212
2213 lsof -b
2214
2215 Look for standard error file warning messages that begin ``assuming
2216 "dev=xxxx" from ...''.
2217
2219 Lsof is able to examine the kernel's name cache or use other kernel
2220 facilities (e.g., the ADVFS 4.x tag_to_path() function under Tru64
2221 UNIX) on some dialects for most file system types, excluding AFS, and
2222 extract recently used path name components from it. (AFS file system
2223 path lookups don't use the kernel's name cache; some Solaris VxFS file
2224 system operations apparently don't use it, either.)
2225
2226 Lsof reports the complete paths it finds in the NAME column. If lsof
2227 can't report all components in a path, it reports in the NAME column
2228 the file system name, followed by a space, two `-' characters, another
2229 space, and the name components it has located, separated by the `/'
2230 character.
2231
2232 When lsof is run in repeat mode - i.e., with the -r option specified -
2233 the extent to which it can report path name components for the same
2234 file may vary from cycle to cycle. That's because other running pro‐
2235 cesses can cause the kernel to remove entries from its name cache and
2236 replace them with others.
2237
2238 Lsof's use of the kernel name cache to identify the paths of files can
2239 lead it to report incorrect components under some circumstances. This
2240 can happen when the kernel name cache uses device and node number as a
2241 key (e.g., SCO OpenServer) and a key on a rapidly changing file system
2242 is reused. If the UNIX dialect's kernel doesn't purge the name cache
2243 entry for a file when it is unlinked, lsof may find a reference to the
2244 wrong entry in the cache. The lsof FAQ (The FAQ section gives its
2245 location.) has more information on this situation.
2246
2247 Lsof can report path name components for these dialects:
2248
2249 FreeBSD
2250 HP-UX
2251 Linux
2252 NetBSD
2253 NEXTSTEP
2254 OpenBSD
2255 OPENSTEP
2256 SCO OpenServer
2257 SCO|Caldera UnixWare
2258 Solaris
2259 Tru64 UNIX
2260
2261 Lsof can't report path name components for these dialects:
2262
2263 AIX
2264
2265 If you want to know why lsof can't report path name components for some
2266 dialects, see the lsof FAQ (The FAQ section gives its location.)
2267
2269 Examining all members of the /dev (or /devices) node tree with stat(2)
2270 functions can be time consuming. What's more, the information that
2271 lsof needs - device number, inode number, and path - rarely changes.
2272
2273 Consequently, lsof normally maintains an ASCII text file of cached /dev
2274 (or /devices) information (exception: the /proc-based Linux lsof where
2275 it's not needed.) The local system administrator who builds lsof can
2276 control the way the device cache file path is formed, selecting from
2277 these options:
2278
2279 Path from the -D option;
2280 Path from an environment variable;
2281 System-wide path;
2282 Personal path (the default);
2283 Personal path, modified by an environment variable.
2284
2285 Consult the output of the -h, -D? , or -? help options for the current
2286 state of device cache support. The help output lists the default
2287 read-mode device cache file path that is in effect for the current
2288 invocation of lsof. The -D? option output lists the read-only and
2289 write device cache file paths, the names of any applicable environment
2290 variables, and the personal device cache path format.
2291
2292 Lsof can detect that the current device cache file has been acciden‐
2293 tally or maliciously modified by integrity checks, including the compu‐
2294 tation and verification of a sixteen bit Cyclic Redundancy Check (CRC)
2295 sum on the file's contents. When lsof senses something wrong with the
2296 file, it issues a warning and attempts to remove the current cache file
2297 and create a new copy, but only to a path that the process can legiti‐
2298 mately write.
2299
2300 The path from which a lsof process may attempt to read a device cache
2301 file may not be the same as the path to which it can legitimately
2302 write. Thus when lsof senses that it needs to update the device cache
2303 file, it may choose a different path for writing it from the path from
2304 which it read an incorrect or outdated version.
2305
2306 If available, the -Dr option will inhibit the writing of a new device
2307 cache file. (It's always available when specified without a path name
2308 argument.)
2309
2310 When a new device is added to the system, the device cache file may
2311 need to be recreated. Since lsof compares the mtime of the device
2312 cache file with the mtime and ctime of the /dev (or /devices) direc‐
2313 tory, it usually detects that a new device has been added; in that case
2314 lsof issues a warning message and attempts to rebuild the device cache
2315 file.
2316
2317 Whenever lsof writes a device cache file, it sets its ownership to the
2318 real UID of the executing process, and its permission modes to 0600,
2319 this restricting its reading and writing to the file's owner.
2320
2322 Two permissions of the lsof executable affect its ability to access
2323 device cache files. The permissions are set by the local system admin‐
2324 istrator when lsof is installed.
2325
2326 The first and rarer permission is setuid-root. It comes into effect
2327 when lsof is executed; its effective UID is then root, while its real
2328 (i.e., that of the logged-on user) UID is not. The lsof distribution
2329 recommends that versions for these dialects run setuid-root.
2330
2331 HP-UX 11.11 and 11.23
2332 Linux
2333
2334 The second and more common permission is setgid. It comes into effect
2335 when the effective group IDentification number (GID) of the lsof
2336 process is set to one that can access kernel memory devices - e.g.,
2337 ``kmem'', ``sys'', or ``system''.
2338
2339 An lsof process that has setgid permission usually surrenders the per‐
2340 mission after it has accessed the kernel memory devices. When it does
2341 that, lsof can allow more liberal device cache path formations. The
2342 lsof distribution recommends that versions for these dialects run set‐
2343 gid and be allowed to surrender setgid permission.
2344
2345 AIX 5.[12] and 5.3-ML1
2346 Apple Darwin 7.x Power Macintosh systems
2347 FreeBSD 4.x, 4.1x, 5.x and [6789].x for x86-based systems
2348 FreeBSD 5.x, [6789].x and 1[012].8for Alpha, AMD64 and Sparc64
2349 based systems
2350 HP-UX 11.00
2351 NetBSD 1.[456], 2.x and 3.x for Alpha, x86, and SPARC-based
2352 systems
2353 NEXTSTEP 3.[13] for NEXTSTEP architectures
2354 OpenBSD 2.[89] and 3.[0-9] for x86-based systems
2355 OPENSTEP 4.x
2356 SCO OpenServer Release 5.0.6 for x86-based systems
2357 SCO|Caldera UnixWare 7.1.4 for x86-based systems
2358 Solaris 2.6, 8, 9 and 10
2359 Tru64 UNIX 5.1
2360
2361 (Note: lsof for AIX 5L and above needs setuid-root permission if its -X
2362 option is used.)
2363
2364 Lsof for these dialects does not support a device cache, so the permis‐
2365 sions given to the executable don't apply to the device cache file.
2366
2367 Linux
2368
2370 The -D option provides limited means for specifying the device cache
2371 file path. Its ? function will report the read-only and write device
2372 cache file paths that lsof will use.
2373
2374 When the -D b, r, and u functions are available, you can use them to
2375 request that the cache file be built in a specific location (b[path]);
2376 read but not rebuilt (r[path]); or read and rebuilt (u[path]). The b,
2377 r, and u functions are restricted under some conditions. They are
2378 restricted when the lsof process is setuid-root. The path specified
2379 with the r function is always read-only, even when it is available.
2380
2381 The b, r, and u functions are also restricted when the lsof process
2382 runs setgid and lsof doesn't surrender the setgid permission. (See the
2383 LSOF PERMISSIONS THAT AFFECT DEVICE CACHE FILE ACCESS section for a
2384 list of implementations that normally don't surrender their setgid per‐
2385 mission.)
2386
2387 A further -D function, i (for ignore), is always available.
2388
2389 When available, the b function tells lsof to read device information
2390 from the kernel with the stat(2) function and build a device cache file
2391 at the indicated path.
2392
2393 When available, the r function tells lsof to read the device cache
2394 file, but not update it. When a path argument accompanies -Dr, it
2395 names the device cache file path. The r function is always available
2396 when it is specified without a path name argument. If lsof is not run‐
2397 ning setuid-root and surrenders its setgid permission, a path name
2398 argument may accompany the r function.
2399
2400 When available, the u function tells lsof to attempt to read and use
2401 the device cache file. If it can't read the file, or if it finds the
2402 contents of the file incorrect or outdated, it will read information
2403 from the kernel, and attempt to write an updated version of the device
2404 cache file, but only to a path it considers legitimate for the lsof
2405 process effective and real UIDs.
2406
2408 Lsof's second choice for the device cache file is the contents of the
2409 LSOFDEVCACHE environment variable. It avoids this choice if the lsof
2410 process is setuid-root, or the real UID of the process is root.
2411
2412 A further restriction applies to a device cache file path taken from
2413 the LSOFDEVCACHE environment variable: lsof will not write a device
2414 cache file to the path if the lsof process doesn't surrender its setgid
2415 permission. (See the LSOF PERMISSIONS THAT AFFECT DEVICE CACHE FILE
2416 ACCESS section for information on implementations that don't surrender
2417 their setgid permission.)
2418
2419 The local system administrator can disable the use of the LSOFDEVCACHE
2420 environment variable or change its name when building lsof. Consult
2421 the output of -D? for the environment variable's name.
2422
2424 The local system administrator may choose to have a system-wide device
2425 cache file when building lsof. That file will generally be constructed
2426 by a special system administration procedure when the system is booted
2427 or when the contents of /dev or /devices) changes. If defined, it is
2428 lsof's third device cache file path choice.
2429
2430 You can tell that a system-wide device cache file is in effect for your
2431 local installation by examining the lsof help option output - i.e., the
2432 output from the -h or -? option.
2433
2434 Lsof will never write to the system-wide device cache file path by
2435 default. It must be explicitly named with a -D function in a
2436 root-owned procedure. Once the file has been written, the procedure
2437 must change its permission modes to 0644 (owner-read and owner-write,
2438 group-read, and other-read).
2439
2441 The default device cache file path of the lsof distribution is one
2442 recorded in the home directory of the real UID that executes lsof.
2443 Added to the home directory is a second path component of the form
2444 .lsof_hostname.
2445
2446 This is lsof's fourth device cache file path choice, and is usually the
2447 default. If a system-wide device cache file path was defined when lsof
2448 was built, this fourth choice will be applied when lsof can't find the
2449 system-wide device cache file. This is the only time lsof uses two
2450 paths when reading the device cache file.
2451
2452 The hostname part of the second component is the base name of the exe‐
2453 cuting host, as returned by gethostname(2). The base name is defined
2454 to be the characters preceding the first `.' in the gethostname(2)
2455 output, or all the gethostname(2) output if it contains no `.'.
2456
2457 The device cache file belongs to the user ID and is readable and
2458 writable by the user ID alone - i.e., its modes are 0600. Each dis‐
2459 tinct real user ID on a given host that executes lsof has a distinct
2460 device cache file. The hostname part of the path distinguishes device
2461 cache files in an NFS-mounted home directory into which device cache
2462 files are written from several different hosts.
2463
2464 The personal device cache file path formed by this method represents a
2465 device cache file that lsof will attempt to read, and will attempt to
2466 write should it not exist or should its contents be incorrect or out‐
2467 dated.
2468
2469 The -Dr option without a path name argument will inhibit the writing of
2470 a new device cache file.
2471
2472 The -D? option will list the format specification for constructing the
2473 personal device cache file. The conversions used in the format speci‐
2474 fication are described in the 00DCACHE file of the lsof distribution.
2475
2477 If this option is defined by the local system administrator when lsof
2478 is built, the LSOFPERSDCPATH environment variable contents may be used
2479 to add a component of the personal device cache file path.
2480
2481 The LSOFPERSDCPATH variable contents are inserted in the path at the
2482 place marked by the local system administrator with the ``%p'' conver‐
2483 sion in the HASPERSDC format specification of the dialect's machine.h
2484 header file. (It's placed right after the home directory in the
2485 default lsof distribution.)
2486
2487 Thus, for example, if LSOFPERSDCPATH contains ``LSOF'', the home direc‐
2488 tory is ``/Homes/abe'', the host name is ``lsof.itap.purdue.edu'', and
2489 the HASPERSDC format is the default (``%h/%p.lsof_%L''), the modified
2490 personal device cache file path is:
2491
2492 /Homes/abe/LSOF/.lsof_vic
2493
2494 The LSOFPERSDCPATH environment variable is ignored when the lsof
2495 process is setuid-root or when the real UID of the process is root.
2496
2497 Lsof will not write to a modified personal device cache file path if
2498 the lsof process doesn't surrender setgid permission. (See the LSOF
2499 PERMISSIONS THAT AFFECT DEVICE CACHE FILE ACCESS section for a list of
2500 implementations that normally don't surrender their setgid permission.)
2501
2502 If, for example, you want to create a sub-directory of personal device
2503 cache file paths by using the LSOFPERSDCPATH environment variable to
2504 name it, and lsof doesn't surrender its setgid permission, you will
2505 have to allow lsof to create device cache files at the standard per‐
2506 sonal path and move them to your subdirectory with shell commands.
2507
2508 The local system administrator may: disable this option when lsof is
2509 built; change the name of the environment variable from LSOFPERSDCPATH
2510 to something else; change the HASPERSDC format to include the personal
2511 path component in another place; or exclude the personal path component
2512 entirely. Consult the output of the -D? option for the environment
2513 variable's name and the HASPERSDC format specification.
2514
2516 Errors are identified with messages on the standard error file.
2517
2518 Lsof returns a one (1) if any error was detected, including the failure
2519 to locate command names, file names, Internet addresses or files, login
2520 names, NFS files, PIDs, PGIDs, or UIDs it was asked to list. If the -V
2521 option is specified, lsof will indicate the search items it failed to
2522 list.
2523
2524 It returns a zero (0) if no errors were detected and if it was able to
2525 list some information about all the specified search arguments.
2526
2527 When lsof cannot open access to /dev (or /devices) or one of its subdi‐
2528 rectories, or get information on a file in them with stat(2), it issues
2529 a warning message and continues. That lsof will issue warning messages
2530 about inaccessible files in /dev (or /devices) is indicated in its help
2531 output - requested with the -h or >B -? options - with the message:
2532
2533 Inaccessible /dev warnings are enabled.
2534
2535 The warning message may be suppressed with the -w option. It may also
2536 have been suppressed by the system administrator when lsof was compiled
2537 by the setting of the WARNDEVACCESS definition. In this case, the out‐
2538 put from the help options will include the message:
2539
2540 Inaccessible /dev warnings are disabled.
2541
2542 Inaccessible device warning messages usually disappear after lsof has
2543 created a working device cache file.
2544
2546 For a more extensive set of examples, documented more fully, see the
2547 00QUICKSTART file of the lsof distribution.
2548
2549 To list all open files, use:
2550
2551 lsof
2552
2553 To list all open Internet, x.25 (HP-UX), and UNIX domain files, use:
2554
2555 lsof -i -U
2556
2557 To list all open IPv4 network files in use by the process whose PID is
2558 1234, use:
2559
2560 lsof -i 4 -a -p 1234
2561
2562 Presuming the UNIX dialect supports IPv6, to list only open IPv6 net‐
2563 work files, use:
2564
2565 lsof -i 6
2566
2567 To list all files using any protocol on ports 513, 514, or 515 of host
2568 wonderland.cc.purdue.edu, use:
2569
2570 lsof -i @wonderland.cc.purdue.edu:513-515
2571
2572 To list all files using any protocol on any port of mace.cc.purdue.edu
2573 (cc.purdue.edu is the default domain), use:
2574
2575 lsof -i @mace
2576
2577 To list all open files for login name ``abe'', or user ID 1234, or
2578 process 456, or process 123, or process 789, use:
2579
2580 lsof -p 456,123,789 -u 1234,abe
2581
2582 To list all open files on device /dev/hd4, use:
2583
2584 lsof /dev/hd4
2585
2586 To find the process that has /u/abe/foo open, use:
2587
2588 lsof /u/abe/foo
2589
2590 To send a SIGHUP to the processes that have /u/abe/bar open, use:
2591
2592 kill -HUP `lsof -t /u/abe/bar`
2593
2594 To find any open file, including an open UNIX domain socket file, with
2595 the name /dev/log, use:
2596
2597 lsof /dev/log
2598
2599 To find processes with open files on the NFS file system named
2600 /nfs/mount/point whose server is inaccessible, and presuming your mount
2601 table supplies the device number for /nfs/mount/point, use:
2602
2603 lsof -b /nfs/mount/point
2604
2605 To do the preceding search with warning messages suppressed, use:
2606
2607 lsof -bw /nfs/mount/point
2608
2609 To ignore the device cache file, use:
2610
2611 lsof -Di
2612
2613 To obtain PID and command name field output for each process, file
2614 descriptor, file device number, and file inode number for each file of
2615 each process, use:
2616
2617 lsof -FpcfDi
2618
2619 To list the files at descriptors 1 and 3 of every process running the
2620 lsof command for login ID ``abe'' every 10 seconds, use:
2621
2622 lsof -c lsof -a -d 1 -d 3 -u abe -r10
2623
2624 To list the current working directory of processes running a command
2625 that is exactly four characters long and has an 'o' or 'O' in character
2626 three, use this regular expression form of the -c c option:
2627
2628 lsof -c /^..o.$/i -a -d cwd
2629
2630 To find an IP version 4 socket file by its associated numeric dot-form
2631 address, use:
2632
2633 lsof -i@128.210.15.17
2634
2635 To find an IP version 6 socket file (when the UNIX dialect supports
2636 IPv6) by its associated numeric colon-form address, use:
2637
2638 lsof -i@[0:1:2:3:4:5:6:7]
2639
2640 To find an IP version 6 socket file (when the UNIX dialect supports
2641 IPv6) by an associated numeric colon-form address that has a run of
2642 zeroes in it - e.g., the loop-back address - use:
2643
2644 lsof -i@[::1]
2645
2646 To obtain a repeat mode marker line that contains the current time,
2647 use:
2648
2649 lsof -rm====%T====
2650
2651 To add spaces to the previous marker line, use:
2652
2653 lsof -r "m==== %T ===="
2654
2656 Since lsof reads kernel memory in its search for open files, rapid
2657 changes in kernel memory may produce unpredictable results.
2658
2659 When a file has multiple record locks, the lock status character (fol‐
2660 lowing the file descriptor) is derived from a test of the first lock
2661 structure, not from any combination of the individual record locks that
2662 might be described by multiple lock structures.
2663
2664 Lsof can't search for files with restrictive access permissions by name
2665 unless it is installed with root set-UID permission. Otherwise it is
2666 limited to searching for files to which its user or its set-GID group
2667 (if any) has access permission.
2668
2669 The display of the destination address of a raw socket (e.g., for ping)
2670 depends on the UNIX operating system. Some dialects store the destina‐
2671 tion address in the raw socket's protocol control block, some do not.
2672
2673 Lsof can't always represent Solaris device numbers in the same way that
2674 ls(1) does. For example, the major and minor device numbers that the
2675 lstat(2) and stat(2) functions report for the directory on which CD-ROM
2676 files are mounted (typically /cdrom) are not the same as the ones that
2677 it reports for the device on which CD-ROM files are mounted (typically
2678 /dev/sr0). (Lsof reports the directory numbers.)
2679
2680 The support for /proc file systems is available only for BSD and Tru64
2681 UNIX dialects, Linux, and dialects derived from SYSV R4 - e.g., Free‐
2682 BSD, NetBSD, OpenBSD, Solaris, UnixWare.
2683
2684 Some /proc file items - device number, inode number, and file size -
2685 are unavailable in some dialects. Searching for files in a /proc file
2686 system may require that the full path name be specified.
2687
2688 No text (txt) file descriptors are displayed for Linux processes. All
2689 entries for files other than the current working directory, the root
2690 directory, and numerical file descriptors are labeled mem descriptors.
2691
2692 Lsof can't search for Tru64 UNIX named pipes by name, because their
2693 kernel implementation of lstat(2) returns an improper device number for
2694 a named pipe.
2695
2696 Lsof can't report fully or correctly on HP-UX 9.01, 10.20, and 11.00
2697 locks because of insufficient access to kernel data or errors in the
2698 kernel data. See the lsof FAQ (The FAQ section gives its location.)
2699 for details.
2700
2701 The AIX SMT file type is a fabrication. It's made up for file struc‐
2702 tures whose type (15) isn't defined in the AIX /usr/include/sys/file.h
2703 header file. One way to create such file structures is to run X
2704 clients with the DISPLAY variable set to ``:0.0''.
2705
2706 The +|-f[cfgGn] option is not supported under /proc-based Linux lsof,
2707 because it doesn't read kernel structures from kernel memory.
2708
2710 Lsof may access these environment variables.
2711
2712 LANG defines a language locale. See setlocale(3) for the
2713 names of other variables that can be used in place of
2714 LANG - e.g., LC_ALL, LC_TYPE, etc.
2715
2716 LSOFDEVCACHE defines the path to a device cache file. See the
2717 DEVICE CACHE PATH FROM AN ENVIRONMENT VARIABLE sec‐
2718 tion for more information.
2719
2720 LSOFPERSDCPATH defines the middle component of a modified personal
2721 device cache file path. See the MODIFIED PERSONAL
2722 DEVICE CACHE PATH section for more information.
2723
2725 Frequently-asked questions and their answers (an FAQ) are available in
2726 the 00FAQ file of the lsof distribution.
2727
2728 That file is also available via anonymous ftp from lsof.itap.purdue.edu
2729 at pub/tools/unix/lsofFAQ. The URL is:
2730
2731 ftp://lsof.itap.purdue.edu/pub/tools/unix/lsof/FAQ
2732
2734 /dev/kmem kernel virtual memory device
2735
2736 /dev/mem physical memory device
2737
2738 /dev/swap system paging device
2739
2740 .lsof_hostname lsof's device cache file (The suffix, hostname, is
2741 the first component of the host's name returned by
2742 gethostname(2).)
2743
2745 Lsof was written by Victor A.Abell <abe@purdue.edu> of Purdue Univer‐
2746 sity. Many others have contributed to lsof. They're listed in the
2747 00CREDITS file of the lsof distribution.
2748
2750 The latest distribution of lsof is available via anonymous ftp from the
2751 host lsof.itap.purdue.edu. You'll find the lsof distribution in the
2752 pub/tools/unix/lsof directory.
2753
2754 You can also use this URL:
2755
2756 ftp://lsof.itap.purdue.edu/pub/tools/unix/lsof
2757
2758 Lsof is also mirrored elsewhere. When you access lsof.itap.purdue.edu
2759 and change to its pub/tools/unix/lsof directory, you'll be given a list
2760 of some mirror sites. The pub/tools/unix/lsof directory also contains
2761 a more complete list in its mirrors file. Use mirrors with caution -
2762 not all mirrors always have the latest lsof revision.
2763
2764 Some pre-compiled Lsof executables are available on lsof.itap.pur‐
2765 due.edu, but their use is discouraged - it's better that you build your
2766 own from the sources. If you feel you must use a pre-compiled exe‐
2767 cutable, please read the cautions that appear in the README files of
2768 the pub/tools/unix/lsof/binaries subdirectories and in the 00* files of
2769 the distribution.
2770
2771 More information on the lsof distribution can be found in its
2772 README.lsof_<version> file. If you intend to get the lsof distribution
2773 and build it, please read README.lsof_<version> and the other 00* files
2774 of the distribution before sending questions to the author.
2775
2777 Not all the following manual pages may exist in every UNIX dialect to
2778 which lsof has been ported.
2779
2780 access(2), awk(1), crash(1), fattach(3C), ff(1), fstat(8), fuser(1),
2781 gethostname(2), isprint(3), kill(1), localtime(3), lstat(2), mod‐
2782 load(8), mount(8), netstat(1), ofiles(8L), perl(1), ps(1), readlink(2),
2783 setlocale(3), stat(2), strftime(3), time(2), uname(1).
2784
2785
2786
2787 Revision-4.91 LSOF(1)