The event API provides a mechanism to execute a function when a spe‐
cific event on a file descriptor occurs or after a given time has
The event API needs to be initialized with event_init() before it can
In order to process events, an application needs to call
event_dispatch(). This function only returns on error, and should
replace the event core of the application program.
The function event_set() prepares the event structure ev to be used in
future calls to event_add() and event_del(). The event will be pre‐
pared to call the function specified by the fn argument with an int
argument indicating the file descriptor, a short argument indicating
the type of event, and a void * argument given in the arg argument.
The fd indicates the file descriptor that should be monitored for
events. The events can be either EV_READ, EV_WRITE, or both, indicat‐
ing that an application can read or write from the file descriptor
respectively without blocking.
The function fn will be called with the file descriptor that triggered
the event and the type of event which will be either EV_TIMEOUT,
EV_SIGNAL, EV_READ, or EV_WRITE. Additionally, an event which has reg‐
istered interest in more than one of the preceeding events, via bit‐
wise-OR to event_set(), can provide its callback function with a bit‐
wise-OR of more than one triggered event. The additional flag
EV_PERSIST makes an event_add() persistent until event_del() has been
Once initialized, the ev structure can be used repeatedly with
event_add() and event_del() and does not need to be reinitialized
unless the function called and/or the argument to it are to be changed.
However, when an ev structure has been added to libevent using
event_add() the structure must persist until the event occurs (assuming
EV_PERSIST is not set) or is removed using event_del(). You may not
reuse the same ev structure for multiple monitored descriptors; each
descriptor needs its own ev.
The function event_add() schedules the execution of the ev event when
the event specified in event_set() occurs or in at least the time spec‐
ified in the tv. If tv is NULL, no timeout occurs and the function
will only be called if a matching event occurs on the file descriptor.
The event in the ev argument must be already initialized by event_set()
and may not be used in calls to event_set() until it has timed out or
been removed with event_del(). If the event in the ev argument already
has a scheduled timeout, the old timeout will be replaced by the new
The function event_del() will cancel the event in the argument ev. If
the event has already executed or has never been added the call will
have no effect.
The functions evtimer_set(), evtimer_add(), evtimer_del(),
evtimer_initialized(), and evtimer_pending() are abbreviations for com‐
mon situations where only a timeout is required. The file descriptor
passed will be -1, and the event type will be EV_TIMEOUT.
The functions signal_set(), signal_add(), signal_del(),
signal_initialized(), and signal_pending() are abbreviations. The
event type will be a persistent EV_SIGNAL. That means signal_set()
In order to avoid races in signal handlers, the event API provides two
variables: event_sigcb and event_gotsig. A signal handler sets
event_gotsig to indicate that a signal has been received. The applica‐
tion sets event_sigcb to a callback function. After the signal handler
sets event_gotsig, event_dispatch will execute the callback function to
process received signals. The callback returns 1 when no events are
registered any more. It can return -1 to indicate an error to the
event library, causing event_dispatch() to terminate with errno set to
The function event_once() is similar to event_set(). However, it
schedules a callback to be called exactly once and does not require the
caller to prepare an event structure. This function supports
EV_TIMEOUT, EV_READ, and EV_WRITE.
The event_pending() function can be used to check if the event speci‐
fied by event is pending to run. If EV_TIMEOUT was specified and tv is
not NULL, the expiration time of the event will be returned in tv.
The event_initialized() macro can be used to check if an event has been
The event_loop function provides an interface for single pass execution
of pending events. The flags EVLOOP_ONCE and EVLOOP_NONBLOCK are rec‐
ognized. The event_loopexit function exits from the event loop. The
next event_loop() iteration after the given timer expires will complete
normally (handling all queued events) then exit without blocking for
events again. Subsequent invocations of event_loop() will proceed nor‐
mally. The event_loopbreak function exits from the event loop immedi‐
ately. event_loop() will abort after the next event is completed;
event_loopbreak() is typically invoked from this event's callback. This
behavior is analogous to the "break;" statement. Subsequent invocations
of event_loop() will proceed normally.
It is the responsibility of the caller to provide these functions with
pre-allocated event structures.
By default libevent schedules all active events with the same priority.
However, sometimes it is desirable to process some events with a higher
priority than others. For that reason, libevent supports strict prior‐
ity queues. Active events with a lower priority are always processed
before events with a higher priority.
The number of different priorities can be set initially with the
event_priority_init() function. This function should be called before
the first call to event_dispatch(). The event_priority_set() function
can be used to assign a priority to an event. By default, libevent
assigns the middle priority to all events unless their priority is
Libevent has experimental support for thread-safe events. When ini‐
tializing the library via event_init(), an event base is returned.
This event base can be used in conjunction with calls to
event_base_set(), event_base_dispatch(), event_base_loop(),
event_base_loopexit(), bufferevent_base_set() and event_base_free().
event_base_set() should be called after preparing an event with
event_set(), as event_set() assigns the provided event to the most
recently created event base. bufferevent_base_set() should be called
after preparing a bufferevent with bufferevent_new().
event_base_free() should be used to free memory associated with the
event base when it is no longer needed.
libevent provides an abstraction on top of the regular event callbacks.
This abstraction is called a buffered event. A buffered event provides
input and output buffers that get filled and drained automatically.
The user of a buffered event no longer deals directly with the IO, but
instead is reading from input and writing to output buffers.
A new bufferevent is created by bufferevent_new(). The parameter fd
specifies the file descriptor from which data is read and written to.
This file descriptor is not allowed to be a pipe(2). The next three
parameters are callbacks. The read and write callback have the follow‐
ing form: void(*cb)(struct bufferevent *bufev, void *arg). The error
callback has the following form: void(*cb)(struct bufferevent *bufev,
short what, void *arg). The argument is specified by the fourth param‐
eter cbarg. A bufferevent struct pointer is returned on success, NULL
on error. Both the read and the write callback may be NULL. The error
callback has to be always provided.
Once initialized, the bufferevent structure can be used repeatedly with
bufferevent_enable() and bufferevent_disable(). The flags parameter
can be a combination of EV_READ and EV_WRITE. When read enabled the
bufferevent will try to read from the file descriptor and call the read
callback. The write callback is executed whenever the output buffer is
drained below the write low watermark, which is 0 by default.
The bufferevent_write() function can be used to write data to the file
descriptor. The data is appended to the output buffer and written to
the descriptor automatically as it becomes available for writing.
bufferevent_write() returns 0 on success or -1 on failure. The
bufferevent_read() function is used to read data from the input buffer,
returning the amount of data read.
If multiple bases are in use, bufferevent_base_set() must be called
before enabling the bufferevent for the first time.
libevent provides a very thin HTTP layer that can be used both to host
an HTTP server and also to make HTTP requests. An HTTP server can be
created by calling evhttp_new(). It can be bound to any port and
address with the evhttp_bind_socket() function. When the HTTP server
is no longer used, it can be freed via evhttp_free().
To be notified of HTTP requests, a user needs to register callbacks
with the HTTP server. This can be done by calling evhttp_set_cb().
The second argument is the URI for which a callback is being regis‐
tered. The corresponding callback will receive an structevhttp_request object that contains all information about the request.
This section does not document all the possible function calls; please
check event.h for the public interfaces.
It is possible to disable support for epoll, kqueue, devpoll, poll or
select by setting the environment variable EVENT_NOEPOLL,
EVENT_NOKQUEUE, EVENT_NODEVPOLL, EVENT_NOPOLL or EVENT_NOSELECT,
respectively. By setting the environment variable EVENT_SHOW_METHOD,
libevent displays the kernel notification method that it uses.
This documentation is neither complete nor authoritative. If you are
in doubt about the usage of this API then check the source code to find
out how it works, write up the missing piece of documentation and send
it to me for inclusion in this man page.
This page is part of the libevent (an event notification library)
project. Information about the project can be found at
http://libevent.org/. If you have a bug report for this manual page,
see ⟨http://sourceforge.net/p/levent/bugs/⟩. This page was obtained
from the project's upstream Git repository
⟨https://github.com/libevent/libevent.git⟩ on 2020-08-13. (At that
time, the date of the most recent commit that was found in the reposi‐
tory was 2020-08-11.) If you discover any rendering problems in this
HTML version of the page, or you believe there is a better or more up-
to-date source for the page, or you have corrections or improvements to
the information in this COLOPHON (which is not part of the original
manual page), send a mail to email@example.com
BSD August 8, 2000 BSD