freebsd-skq/lib/libc/sys/kqueue.2

.\" Copyright (c) 2000 Jonathan Lemon
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\"    notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\"    notice, this list of conditions and the following disclaimer in the
.\"    documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
.Dd October 8, 2020
.Dt KQUEUE 2
.Os
.Sh NAME
.Nm kqueue ,
.Nm kevent
.Nd kernel event notification mechanism
.Sh LIBRARY
.Lb libc
.Sh SYNOPSIS
.In sys/event.h
.Ft int
.Fn kqueue "void"
.Ft int
.Fo kevent
.Fa "int kq"
.Fa "const struct kevent *changelist"
.Fa "int nchanges"
.Fa "struct kevent *eventlist"
.Fa "int nevents"
.Fa "const struct timespec *timeout"
.Fc
.Fn EV_SET "kev" ident filter flags fflags data udata
.Sh DESCRIPTION
The
.Fn kqueue
system call
provides a generic method of notifying the user when an event
happens or a condition holds, based on the results of small
pieces of kernel code termed filters.
A kevent is identified by the (ident, filter) pair; there may only
be one unique kevent per kqueue.
.Pp
The filter is executed upon the initial registration of a kevent
in order to detect whether a preexisting condition is present, and is also
executed whenever an event is passed to the filter for evaluation.
If the filter determines that the condition should be reported,
then the kevent is placed on the kqueue for the user to retrieve.
.Pp
The filter is also run when the user attempts to retrieve the kevent
from the kqueue.
If the filter indicates that the condition that triggered
the event no longer holds, the kevent is removed from the kqueue and
is not returned.
.Pp
Multiple events which trigger the filter do not result in multiple
kevents being placed on the kqueue; instead, the filter will aggregate
the events into a single struct kevent.
Calling
.Fn close
on a file descriptor will remove any kevents that reference the descriptor.
.Pp
The
.Fn kqueue
system call
creates a new kernel event queue and returns a descriptor.
The queue is not inherited by a child created with
.Xr fork 2 .
However, if
.Xr rfork 2
is called without the
.Dv RFFDG
flag, then the descriptor table is shared,
which will allow sharing of the kqueue between two processes.
.Pp
The
.Fn kevent
system call
is used to register events with the queue, and return any pending
events to the user.
The
.Fa changelist
argument
is a pointer to an array of
.Va kevent
structures, as defined in
.In sys/event.h .
All changes contained in the
.Fa changelist
are applied before any pending events are read from the queue.
The
.Fa nchanges
argument
gives the size of
.Fa changelist .
The
.Fa eventlist
argument
is a pointer to an array of kevent structures.
The
.Fa nevents
argument
determines the size of
.Fa eventlist .
When
.Fa nevents
is zero,
.Fn kevent
will return immediately even if there is a
.Fa timeout
specified unlike
.Xr select 2 .
If
.Fa timeout
is a non-NULL pointer, it specifies a maximum interval to wait
for an event, which will be interpreted as a struct timespec.
If
.Fa timeout
is a NULL pointer,
.Fn kevent
waits indefinitely.
To effect a poll, the
.Fa timeout
argument should be non-NULL, pointing to a zero-valued
.Va timespec
structure.
The same array may be used for the
.Fa changelist
and
.Fa eventlist .
.Pp
The
.Fn EV_SET
macro is provided for ease of initializing a
kevent structure.
.Pp
The
.Va kevent
structure is defined as:
.Bd -literal
struct kevent {
	uintptr_t  ident;	/* identifier for this event */
	short	  filter;	/* filter for event */
	u_short	  flags;	/* action flags for kqueue */
	u_int	  fflags;	/* filter flag value */
	int64_t   data;		/* filter data value */
	void	  *udata;	/* opaque user data identifier */
	uint64_t  ext[4];	/* extensions */
};
.Ed
.Pp
The fields of
.Fa struct kevent
are:
.Bl -tag -width "Fa filter"
.It Fa ident
Value used to identify this event.
The exact interpretation is determined by the attached filter,
but often is a file descriptor.
.It Fa filter
Identifies the kernel filter used to process this event.
The pre-defined
system filters are described below.
.It Fa flags
Actions to perform on the event.
.It Fa fflags
Filter-specific flags.
.It Fa data
Filter-specific data value.
.It Fa udata
Opaque user-defined value passed through the kernel unchanged.
.It Fa ext
Extended data passed to and from kernel.
The
.Fa ext[0]
and
.Fa ext[1]
members use is defined by the filter.
If the filter does not use them, the members are copied unchanged.
The
.Fa ext[2]
and
.Fa ext[3]
members are always passed through the kernel as-is,
making additional context available to application.
.El
.Pp
The
.Va flags
field can contain the following values:
.Bl -tag -width EV_DISPATCH
.It Dv EV_ADD
Adds the event to the kqueue.
Re-adding an existing event
will modify the parameters of the original event, and not result
in a duplicate entry.
Adding an event automatically enables it,
unless overridden by the EV_DISABLE flag.
.It Dv EV_ENABLE
Permit
.Fn kevent
to return the event if it is triggered.
.It Dv EV_DISABLE
Disable the event so
.Fn kevent
will not return it.
The filter itself is not disabled.
.It Dv EV_DISPATCH
Disable the event source immediately after delivery of an event.
See
.Dv EV_DISABLE
above.
.It Dv EV_DELETE
Removes the event from the kqueue.
Events which are attached to
file descriptors are automatically deleted on the last close of
the descriptor.
.It Dv EV_RECEIPT
This flag is useful for making bulk changes to a kqueue without draining
any pending events.
When passed as input, it forces
.Dv EV_ERROR
to always be returned.
When a filter is successfully added the
.Va data
field will be zero.
Note that if this flag is encountered and there is no remaining space in
.Fa eventlist
to hold the
.Dv EV_ERROR
event, then subsequent changes will not get processed.
.It Dv EV_ONESHOT
Causes the event to return only the first occurrence of the filter
being triggered.
After the user retrieves the event from the kqueue,
it is deleted.
.It Dv EV_CLEAR
After the event is retrieved by the user, its state is reset.
This is useful for filters which report state transitions
instead of the current state.
Note that some filters may automatically
set this flag internally.
.It Dv EV_EOF
Filters may set this flag to indicate filter-specific EOF condition.
.It Dv EV_ERROR
See
.Sx RETURN VALUES
below.
.El
.Pp
The predefined system filters are listed below.
Arguments may be passed to and from the filter via the
.Va fflags
and
.Va data
fields in the kevent structure.
.Bl -tag -width "Dv EVFILT_PROCDESC"
.It Dv EVFILT_READ
Takes a descriptor as the identifier, and returns whenever
there is data available to read.
The behavior of the filter is slightly different depending
on the descriptor type.
.Bl -tag -width 2n
.It Sockets
Sockets which have previously been passed to
.Fn listen
return when there is an incoming connection pending.
.Va data
contains the size of the listen backlog.
.Pp
Other socket descriptors return when there is data to be read,
subject to the
.Dv SO_RCVLOWAT
value of the socket buffer.
This may be overridden with a per-filter low water mark at the
time the filter is added by setting the
.Dv NOTE_LOWAT
flag in
.Va fflags ,
and specifying the new low water mark in
.Va data .
On return,
.Va data
contains the number of bytes of protocol data available to read.
.Pp
If the read direction of the socket has shutdown, then the filter
also sets
.Dv EV_EOF
in
.Va flags ,
and returns the socket error (if any) in
.Va fflags .
It is possible for EOF to be returned (indicating the connection is gone)
while there is still data pending in the socket buffer.
.It Vnodes
Returns when the file pointer is not at the end of file.
.Va data
contains the offset from current position to end of file,
and may be negative.
.Pp
This behavior is different from
.Xr poll 2 ,
where read events are triggered for regular files unconditionally.
This event can be triggered unconditionally by setting the
.Dv NOTE_FILE_POLL
flag in
.Va fflags .
.It "Fifos, Pipes"
Returns when the there is data to read;
.Va data
contains the number of bytes available.
.Pp
When the last writer disconnects, the filter will set
.Dv EV_EOF
in
.Va flags .
This will be cleared by the filter when a new writer connects,
at which point the
filter will resume waiting for data to become available before
returning.
.It "BPF devices"
Returns when the BPF buffer is full, the BPF timeout has expired, or
when the BPF has
.Dq immediate mode
enabled and there is any data to read;
.Va data
contains the number of bytes available.
.It Eventfds
Returns when the counter is greater than 0;
.Va data
contains the counter value, which must be cast to
.Vt uint64_t .
.El
.It Dv EVFILT_WRITE
Takes a descriptor as the identifier, and returns whenever
it is possible to write to the descriptor.
For sockets, pipes
and fifos,
.Va data
will contain the amount of space remaining in the write buffer.
The filter will set
.Dv EV_EOF
when the reader disconnects, and for the fifo case, this will be cleared
when a new reader connects.
Note that this filter is not supported for vnodes or BPF devices.
.Pp
For sockets, the low water mark and socket error handling is
identical to the
.Dv EVFILT_READ
case.
.Pp
For eventfds,
.Va data
will contain the maximum value that can be added to the counter
without blocking.
.It Dv EVFILT_EMPTY
Takes a descriptor as the identifier, and returns whenever
there is no remaining data in the write buffer.
.It Dv EVFILT_AIO
Events for this filter are not registered with
.Fn kevent
directly but are registered via the
.Va aio_sigevent
member of an asynchronous I/O request when it is scheduled via an
asynchronous I/O system call such as
.Fn aio_read .
The filter returns under the same conditions as
.Fn aio_error .
For more details on this filter see
.Xr sigevent 3 and
.Xr aio 4 .
.It Dv EVFILT_VNODE
Takes a file descriptor as the identifier and the events to watch for in
.Va fflags ,
and returns when one or more of the requested events occurs on the descriptor.
The events to monitor are:
.Bl -tag -width "Dv NOTE_CLOSE_WRITE"
.It Dv NOTE_ATTRIB
The file referenced by the descriptor had its attributes changed.
.It Dv NOTE_CLOSE
A file descriptor referencing the monitored file, was closed.
The closed file descriptor did not have write access.
.It Dv NOTE_CLOSE_WRITE
A file descriptor referencing the monitored file, was closed.
The closed file descriptor had write access.
.Pp
This note, as well as
.Dv NOTE_CLOSE ,
are not activated when files are closed forcibly by
.Xr unmount 2 or
.Xr revoke 2 .
Instead,
.Dv NOTE_REVOKE
is sent for such events.
.It Dv NOTE_DELETE
The
.Fn unlink
system call was called on the file referenced by the descriptor.
.It Dv NOTE_EXTEND
For regular file, the file referenced by the descriptor was extended.
.Pp
For directory, reports that a directory entry was added or removed,
as the result of rename operation.
The
.Dv NOTE_EXTEND
event is not reported when a name is changed inside the directory.
.It Dv NOTE_LINK
The link count on the file changed.
In particular, the
.Dv NOTE_LINK
event is reported if a subdirectory was created or deleted inside
the directory referenced by the descriptor.
.It Dv NOTE_OPEN
The file referenced by the descriptor was opened.
.It Dv NOTE_READ
A read occurred on the file referenced by the descriptor.
.It Dv NOTE_RENAME
The file referenced by the descriptor was renamed.
.It Dv NOTE_REVOKE
Access to the file was revoked via
.Xr revoke 2
or the underlying file system was unmounted.
.It Dv NOTE_WRITE
A write occurred on the file referenced by the descriptor.
.El
.Pp
On return,
.Va fflags
contains the events which triggered the filter.
.It Dv EVFILT_PROC
Takes the process ID to monitor as the identifier and the events to watch for
in
.Va fflags ,
and returns when the process performs one or more of the requested events.
If a process can normally see another process, it can attach an event to it.
The events to monitor are:
.Bl -tag -width "Dv NOTE_TRACKERR"
.It Dv NOTE_EXIT
The process has exited.
The exit status will be stored in
.Va data .
.It Dv NOTE_FORK
The process has called
.Fn fork .
.It Dv NOTE_EXEC
The process has executed a new process via
.Xr execve 2
or a similar call.
.It Dv NOTE_TRACK
Follow a process across
.Fn fork
calls.
The parent process registers a new kevent to monitor the child process
using the same
.Va fflags
as the original event.
The child process will signal an event with
.Dv NOTE_CHILD
set in
.Va fflags
and the parent PID in
.Va data .
.Pp
If the parent process fails to register a new kevent
.Pq usually due to resource limitations ,
it will signal an event with
.Dv NOTE_TRACKERR
set in
.Va fflags ,
and the child process will not signal a
.Dv NOTE_CHILD
event.
.El
.Pp
On return,
.Va fflags
contains the events which triggered the filter.
.It Dv EVFILT_PROCDESC
Takes the process descriptor created by
.Xr pdfork 2
to monitor as the identifier and the events to watch for in
.Va fflags ,
and returns when the associated process performs one or more of the
requested events.
The events to monitor are:
.Bl -tag -width "Dv NOTE_EXIT"
.It Dv NOTE_EXIT
The process has exited.
The exit status will be stored in
.Va data .
.El
.Pp
On return,
.Va fflags
contains the events which triggered the filter.
.It Dv EVFILT_SIGNAL
Takes the signal number to monitor as the identifier and returns
when the given signal is delivered to the process.
This coexists with the
.Fn signal
and
.Fn sigaction
facilities, and has a lower precedence.
The filter will record
all attempts to deliver a signal to a process, even if the signal has
been marked as
.Dv SIG_IGN ,
except for the
.Dv SIGCHLD
signal, which, if ignored, will not be recorded by the filter.
Event notification happens after normal
signal delivery processing.
.Va data
returns the number of times the signal has occurred since the last call to
.Fn kevent .
This filter automatically sets the
.Dv EV_CLEAR
flag internally.
.It Dv EVFILT_TIMER
Establishes an arbitrary timer identified by
.Va ident .
When adding a timer,
.Va data
specifies the moment to fire the timer (for
.Dv NOTE_ABSTIME )
or the timeout period.
The timer will be periodic unless
.Dv EV_ONESHOT
or
.Dv NOTE_ABSTIME
is specified.
On return,
.Va data
contains the number of times the timeout has expired since the last call to
.Fn kevent .
For non-monotonic timers, this filter automatically sets the
.Dv EV_CLEAR
flag internally.
.Pp
The filter accepts the following flags in the
.Va fflags
argument:
.Bl -tag -width "Dv NOTE_MSECONDS"
.It Dv NOTE_SECONDS
.Va data
is in seconds.
.It Dv NOTE_MSECONDS
.Va data
is in milliseconds.
.It Dv NOTE_USECONDS
.Va data
is in microseconds.
.It Dv NOTE_NSECONDS
.Va data
is in nanoseconds.
.It Dv NOTE_ABSTIME
The specified expiration time is absolute.
.El
.Pp
If
.Va fflags
is not set, the default is milliseconds.
On return,
.Va fflags
contains the events which triggered the filter.
.Pp
If an existing timer is re-added, the existing timer will be
effectively canceled (throwing away any undelivered record of previous
timer expiration) and re-started using the new parameters contained in
.Va data
and
.Va fflags .
.Pp
There is a system wide limit on the number of timers
which is controlled by the
.Va kern.kq_calloutmax
sysctl.
.It Dv EVFILT_USER
Establishes a user event identified by
.Va ident
which is not associated with any kernel mechanism but is triggered by
user level code.
The lower 24 bits of the
.Va fflags
may be used for user defined flags and manipulated using the following:
.Bl -tag -width "Dv NOTE_FFLAGSMASK"
.It Dv NOTE_FFNOP
Ignore the input
.Va fflags .
.It Dv NOTE_FFAND
Bitwise AND
.Va fflags .
.It Dv NOTE_FFOR
Bitwise OR
.Va fflags .
.It Dv NOTE_FFCOPY
Copy
.Va fflags .
.It Dv NOTE_FFCTRLMASK
Control mask for
.Va fflags .
.It Dv NOTE_FFLAGSMASK
User defined flag mask for
.Va fflags .
.El
.Pp
A user event is triggered for output with the following:
.Bl -tag -width "Dv NOTE_FFLAGSMASK"
.It Dv NOTE_TRIGGER
Cause the event to be triggered.
.El
.Pp
On return,
.Va fflags
contains the users defined flags in the lower 24 bits.
.El
.Sh CANCELLATION BEHAVIOUR
If
.Fa nevents
is non-zero, i.e., the function is potentially blocking, the call
is a cancellation point.
Otherwise, i.e., if
.Fa nevents
is zero, the call is not cancellable.
Cancellation can only occur before any changes are made to the kqueue,
or when the call was blocked and no changes to the queue were requested.
.Sh RETURN VALUES
The
.Fn kqueue
system call
creates a new kernel event queue and returns a file descriptor.
If there was an error creating the kernel event queue, a value of -1 is
returned and errno set.
.Pp
The
.Fn kevent
system call
returns the number of events placed in the
.Fa eventlist ,
up to the value given by
.Fa nevents .
If an error occurs while processing an element of the
.Fa changelist
and there is enough room in the
.Fa eventlist ,
then the event will be placed in the
.Fa eventlist
with
.Dv EV_ERROR
set in
.Va flags
and the system error in
.Va data .
Otherwise,
.Dv -1
will be returned, and
.Dv errno
will be set to indicate the error condition.
If the time limit expires, then
.Fn kevent
returns 0.
.Sh EXAMPLES
.Bd -literal -compact
#include <sys/event.h>
#include <err.h>
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

int
main(int argc, char **argv)
{
    struct kevent event;    /* Event we want to monitor */
    struct kevent tevent;   /* Event triggered */
    int kq, fd, ret;

    if (argc != 2)
	err(EXIT_FAILURE, "Usage: %s path\en", argv[0]);
    fd = open(argv[1], O_RDONLY);
    if (fd == -1)
	err(EXIT_FAILURE, "Failed to open '%s'", argv[1]);

    /* Create kqueue. */
    kq = kqueue();
    if (kq == -1)
	err(EXIT_FAILURE, "kqueue() failed");

    /* Initialize kevent structure. */
    EV_SET(&event, fd, EVFILT_VNODE, EV_ADD | EV_CLEAR, NOTE_WRITE,
	0, NULL);
    /* Attach event to the kqueue. */
    ret = kevent(kq, &event, 1, NULL, 0, NULL);
    if (ret == -1)
	err(EXIT_FAILURE, "kevent register");
    if (event.flags & EV_ERROR)
	errx(EXIT_FAILURE, "Event error: %s", strerror(event.data));

    for (;;) {
	/* Sleep until something happens. */
	ret = kevent(kq, NULL, 0, &tevent, 1, NULL);
	if (ret == -1) {
	    err(EXIT_FAILURE, "kevent wait");
	} else if (ret > 0) {
	    printf("Something was written in '%s'\en", argv[1]);
	}
    }
}
.Ed
.Sh ERRORS
The
.Fn kqueue
system call fails if:
.Bl -tag -width Er
.It Bq Er ENOMEM
The kernel failed to allocate enough memory for the kernel queue.
.It Bq Er ENOMEM
The
.Dv RLIMIT_KQUEUES
rlimit
(see
.Xr getrlimit 2 )
for the current user would be exceeded.
.It Bq Er EMFILE
The per-process descriptor table is full.
.It Bq Er ENFILE
The system file table is full.
.El
.Pp
The
.Fn kevent
system call fails if:
.Bl -tag -width Er
.It Bq Er EACCES
The process does not have permission to register a filter.
.It Bq Er EFAULT
There was an error reading or writing the
.Va kevent
structure.
.It Bq Er EBADF
The specified descriptor is invalid.
.It Bq Er EINTR
A signal was delivered before the timeout expired and before any
events were placed on the kqueue for return.
.It Bq Er EINTR
A cancellation request was delivered to the thread, but not yet handled.
.It Bq Er EINVAL
The specified time limit or filter is invalid.
.It Bq Er ENOENT
The event could not be found to be modified or deleted.
.It Bq Er ENOMEM
No memory was available to register the event
or, in the special case of a timer, the maximum number of
timers has been exceeded.
This maximum is configurable via the
.Va kern.kq_calloutmax
sysctl.
.It Bq Er ESRCH
The specified process to attach to does not exist.
.El
.Pp
When
.Fn kevent
call fails with
.Er EINTR
error, all changes in the
.Fa changelist
have been applied.
.Sh SEE ALSO
.Xr aio_error 2 ,
.Xr aio_read 2 ,
.Xr aio_return 2 ,
.Xr poll 2 ,
.Xr read 2 ,
.Xr select 2 ,
.Xr sigaction 2 ,
.Xr write 2 ,
.Xr pthread_setcancelstate 3 ,
.Xr signal 3
.Rs
.%A Jonathan Lemon
.%T "Kqueue: A Generic and Scalable Event Notification Facility"
.%I USENIX Association
.%B Proceedings of the FREENIX Track: 2001 USENIX Annual Technical Conference
.%D June 25-30, 2001
.\".http://www.usenix.org/event/usenix01/freenix01/full_papers/lemon/lemon.pdf
.Re
.Sh HISTORY
The
.Fn kqueue
and
.Fn kevent
system calls first appeared in
.Fx 4.1 .
.Sh AUTHORS
The
.Fn kqueue
system and this manual page were written by
.An Jonathan Lemon Aq Mt jlemon@FreeBSD.org .
.Sh BUGS
The
.Fa timeout
value is limited to 24 hours; longer timeouts will be silently
reinterpreted as 24 hours.
.Pp
In versions older than
.Fx 12.0 ,
.In sys/event.h
failed to parse without including
.In sys/types.h
manually.