utmp(5)
๐ utmp, wtmp โ login records ๐
okay bestie, sit down, hydrate, because this is the lore behind who's on your system right now. This is utmp(5), a File Formats Manual page, and yes it's giving "ancient Unix scroll" but it's also iconic. ๐ซถ
โจ SYNOPSIS โจ
#include <utmp.h>
that's it. that's the whole vibe. you just include the header and the structures below come with it. ๐
๐ธ DESCRIPTION ๐ธ
The utmp file is literally how you find out who is currently using the system. Just know... there may be more users on the system than the ones showing up here, because not every program bothers to do utmp logging. so it's a sample, not the full census, bestie. ๐ญ
โ ๏ธ Warning, and I mean it:utmpmust not be writable by the user class "other", because sooo many system programs (foolishly!! we said it) depend on its integrity. If you leaveutmpwritable to anybody other than the owner and the group owner of the file, you're risking faked system logfiles and modifications of system files. not a cute look. do not. ๐
The file is just a sequence of utmp structures, declared like this in <utmp.h> โ and heads up, this is only one of several definitions floating around; the details depend on the version of libc:
/* Values for ut_type field, below */
#define EMPTY 0 /* Record does not contain valid info
(formerly known as UT_UNKNOWN on Linux) */
#define RUN_LVL 1 /* Change in system run-level (see
init(1)) */
#define BOOT_TIME 2 /* Time of system boot (in ut_tv) */
#define NEW_TIME 3 /* Time after system clock change
(in ut_tv) */
#define OLD_TIME 4 /* Time before system clock change
(in ut_tv) */
#define INIT_PROCESS 5 /* Process spawned by init(1) */
#define LOGIN_PROCESS 6 /* Session leader process for user login */
#define USER_PROCESS 7 /* Normal process */
#define DEAD_PROCESS 8 /* Terminated process */
#define ACCOUNTING 9 /* Not implemented */
#define UT_LINESIZE 32
#define UT_NAMESIZE 32
#define UT_HOSTSIZE 256
struct exit_status { /* Type for ut_exit, below */
short e_termination; /* Process termination status */
short e_exit; /* Process exit status */
};
struct utmp {
short ut_type; /* Type of record */
pid_t ut_pid; /* PID of login process */
char ut_line[UT_LINESIZE]; /* Device name of tty - "/dev/" */
char ut_id[4]; /* Terminal name suffix,
or inittab(5) ID */
char ut_user[UT_NAMESIZE]; /* Username */
char ut_host[UT_HOSTSIZE]; /* Hostname for remote login, or
kernel version for run-level
messages */
struct exit_status ut_exit; /* Exit status of a process
marked as DEAD_PROCESS; not
used by Linux init(1) */
/* The ut_session and ut_tv fields must be the same size when
compiled 32- and 64-bit. This allows data files and shared
memory to be shared between 32- and 64-bit applications. */
#if __WORDSIZE == 64 && defined __WORDSIZE_COMPAT32
int32_t ut_session; /* Session ID (getsid(2)),
used for windowing */
struct {
int32_t tv_sec; /* Seconds */
int32_t tv_usec; /* Microseconds */
} ut_tv; /* Time entry was made */
#else
long ut_session; /* Session ID */
struct timeval ut_tv; /* Time entry was made */
#endif
int32_t ut_addr_v6[4]; /* Internet address of remote
host; IPv4 address uses
just ut_addr_v6[0] */
char __unused[20]; /* Reserved for future use */
};
/* Backward compatibility hacks */
#define ut_name ut_user
#ifndef _NO_UT_TIME
#define ut_time ut_tv.tv_sec
#endif
#define ut_xtime ut_tv.tv_sec
#define ut_addr ut_addr_v6[0]
so what is this thing actually telling you? it gives you the name of the special file tied to the user's terminal, the user's login name, and the time of login in the form of time(2). oh and โ string fields are terminated by a null byte ('\0') if they're shorter than the size of the field. important detail, remember it. โญ
๐ง how the entries actually get made
The first entries ever created come from init(1) processing inittab(5). But before an entry gets processed, init(1) does a little cleanup on utmp: it sets ut_type to DEAD_PROCESS, and clears ut_user, ut_host, and ut_time with null bytes โ for every record whose ut_type is not DEAD_PROCESS or RUN_LVL and where no process with PID ut_pid exists. If no empty record with the needed ut_id can be found, init(1) creates a new one. It sets ut_id from the inittab, ut_pid and ut_time to the current values, and ut_type to INIT_PROCESS. periodt. ๐
Then mingetty(8) (or agetty(8)) shows up and locates the entry by the PID, changes ut_type to LOGIN_PROCESS, changes ut_time, sets ut_line, and waits for the connection to be established. Then login(1), once a user has been authenticated, changes ut_type to USER_PROCESS, changes ut_time, and sets ut_host and ut_addr. Depending on mingetty(8) (or agetty(8)) and login(1), records may be located by ut_line instead of the preferable ut_pid. ๐
When init(1) finds that a process has exited, it locates its utmp entry by ut_pid, sets ut_type to DEAD_PROCESS, and clears ut_user, ut_host, and ut_time with null bytes. bye bestie ๐
xterm(1) and other terminal emulators do it different โ they directly create a USER_PROCESS record and generate the ut_id by using the string that suffixes the terminal name (the characters following /dev/[pt]ty). If they find a DEAD_PROCESS for this ID, they recycle it โ otherwise they create a new entry. If they can, they'll mark it as DEAD_PROCESS on exiting, and it is advised that they null ut_line, ut_time, ut_user, and ut_host as well. โป๏ธโจ
telnetd(8) sets up a LOGIN_PROCESS entry and leaves the rest to login(1) as usual. After the telnet session ends, telnetd(8) cleans up utmp in the described way. teamwork, honestly. ๐ซถ
๐ฆ and now... wtmp
The wtmp file records all logins and logouts. Its format is exactly like utmp except that:
- a null username indicates a logout on the associated terminal.
- the terminal name
~with usernameshutdownorrebootindicates a system shutdown or reboot. - the pair of terminal names
|/}logs the old/new system time whendate(1)changes it.
wtmp is maintained by login(1), init(1), and some versions of getty(8) (e.g. mingetty(8) or agetty(8)). And listen โ none of these programs creates the file, so if it gets removed, record-keeping is just... turned off. gone. no notes. ๐
๐ FILES ๐
/var/run/utmp/var/log/wtmp
๐ VERSIONS ๐
POSIX.1 does not specify a utmp structure โ it specifies one named utmpx instead (as part of the XSI extension), with specs for the fields ut_type, ut_pid, ut_line, ut_id, ut_user, and ut_tv. POSIX.1 also does not specify the lengths of the ut_line and ut_user fields. so yeah, it's a lil lawless in here. ๐
And Linux? Linux defines the utmpx structure to be the same as the utmp structure. one structure, two names, main character energy fr fr. ๐
โญ STANDARDS โญ
Linux.
(yes, that's the whole section. iconic brevity. ๐ )
๐ HISTORY ๐
Linux utmp entries conform neither to v7/BSD nor to System V โ they're a mix of the two. a little chaos never hurt anybody (it did, actually, keep reading).
- v7/BSD has fewer fields; most importantly it lacks
ut_type, which causes native v7/BSD-like programs to display (for example) dead or login entries. Also, there is no configuration file which allocates slots to sessions. BSD does so because it lacksut_idfields. - In Linux (as in System V), the
ut_idfield of a record will never change once it has been set, which reserves that slot without needing a configuration file. > โ ๏ธ Clearingut_idmay result in race conditions leading to corrupted utmp entries and potential security holes. Clearing the abovementioned fields by filling them with null bytes is not required by System V semantics, but makes it possible to run many programs which assume BSD semantics and which do not modify utmp. Linux uses the BSD conventions for line contents, as documented above. - System V has no
ut_hostorut_addr_v6fields. nope. didn't have them. ๐
๐ NOTES ๐
- Unlike various other systems, where utmp logging can be disabled by removing the file,
utmpmust always exist on Linux. If you want to disablewho(1), then do not makeutmpworld readable. that's the move. ๐ - The file format is machine-dependent, so it's recommended that it be processed only on the machine architecture where it was created. don't carry it around like a friendship bracelet, it won't fit. ๐ง
- Note that on biarch platforms โ that is, systems which can run both 32-bit and 64-bit applications (
x86-64,ppc64,s390x, etc.) โut_tvis the same size in 32-bit mode as in 64-bit mode. The same goes forut_sessionandut_timeif they are present. This allows data files and shared memory to be shared between 32-bit and 64-bit applications. This is achieved by changing the type ofut_sessiontoint32_t, and that ofut_tvto a struct with twoint32_tfieldstv_secandtv_usec. Sinceut_tvmay not be the same asstruct timeval, then instead of the call:
gettimeofday((struct timeval *) &ut.ut_tv, NULL);
the following method of setting this field is recommended:
struct utmp ut;
struct timeval tv;
gettimeofday(&tv, NULL);
ut.ut_tv.tv_sec = tv.tv_sec;
ut.ut_tv.tv_usec = tv.tv_usec;
๐ bestie this one is a real gotcha โ the cast shortcut is RIGHT THERE tempting you and the manpage says do it the long way instead. trust.
๐ SEE ALSO ๐
ac(1), date(1), init(1), last(1), login(1), logname(1), lslogins(1), users(1), utmpdump(1), who(1), getutent(3), getutmp(3), login(3), logout(3), logwtmp(3), updwtmp(3)
and that's the whole page, bestie!! ๐ this one came from Linux man-pages 6.20, dated 2025-09-21 โ so go forth, respect the ut_id, never chmod utmp to other, and keep your null bytes where they belong. mwah. ๐โจ
๐ฆ where this page comes from
- arch rolling: man-pages 6.20-1 ยท BSD-2-Clause AND BSD-3-Clause AND BSD-4.3TAHOE AND BSD-4-Clause-UC AND GPL-1.0-or-later AND GPL-2.0-only AND GPL-2.0-or-later AND GPL-3.0-or-later AND LGPL-3.0-linking-exception AND LGPL-3.0-or-later AND LicenseRef-Public-Domain AND LicenseRef-UltraPermissive AND Linux-man-pages-1-para AND Linux-man-pages-copyleft AND Linux-man-pages-copyleft-2-para AND Linux-man-pages-copyleft-var AND MIT AND Spencer-94
- nixpkgs unstable: man-pages 6.20
- ๐ the page's license: GPL-2.0-or-later
- ๐ girlified by deepseek-ai/DeepSeek-V4.1-Flash on : a language model's rewrite of the original, under its terms