.\" .\" Copyright (c) 2008-2015 Apple Inc. All rights reserved. .\" .\" @APPLE_LICENSE_HEADER_START@ .\" .\" This file contains Original Code and/or Modifications of Original Code .\" as defined in and that are subject to the Apple Public Source License .\" Version 2.0 (the 'License'). You may not use this file except in .\" compliance with the License. Please obtain a copy of the License at .\" http://www.opensource.apple.com/apsl/ and read it before using this .\" file. .\" .\" The Original Code and all software distributed under the License are .\" distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER .\" EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES, .\" INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY, .\" FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT. .\" Please see the License for the specific language governing rights and .\" limitations under the License. .\" .\" @APPLE_LICENSE_HEADER_END@ .\" .\" .\" 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: src/lib/libc/sys/kqueue.2,v 1.32 2002/12/19 09:40:25 ru Exp $ .\" .Dd October 21, 2008 .Dt KQUEUE 2 .Os .Sh NAME .Nm kqueue , .Nm kevent , .Nm kevent64 and .Nm kevent_qos .Nd kernel event notification mechanism .Sh LIBRARY .Lb libc .Sh SYNOPSIS .In sys/types.h .In sys/event.h .In sys/time.h .Ft int .Fn kqueue "void" .Ft int .Fn kevent "int kq" "const struct kevent *changelist" "int nchanges" "struct kevent *eventlist" "int nevents" "const struct timespec *timeout" .Ft int .Fn kevent64 "int kq" "const struct kevent64_s *changelist" "int nchanges" "struct kevent64_s *eventlist" "int nevents" "unsigned int flags" "const struct timespec *timeout" .Ft int .Fn kevent_qos "int kq" "const struct kevent_qos_s *changelist" "int nchanges" "struct kevent_qos_s *eventlist" "int nevents" "void *data_out" "size_t *data_available" "unsigned int flags" .Fn EV_SET "&kev" ident filter flags fflags data udata .Fn EV_SET64 "&kev" ident filter flags fflags data udata "ext[0]" "ext[1]" .Fn EV_SET_QOS "&kev" ident filter flags qos udata fflags xflags data "ext[0]" "ext[1]" "ext[2]" "ext[3]" .Sh DESCRIPTION The .Fn kqueue system call allocates a kqueue file descriptor. This file descriptor provides a generic method of notifying the user when a kernel event (kevent) happens or a condition holds, based on the results of small pieces of kernel code termed filters. .Pp A kevent is identified by an (ident, filter, and optional udata value) tuple. It specifies the interesting conditions to be notified about for that tuple. An (ident, filter, and optional udata value) tuple can only appear once in a given kqueue. Subsequent attempts to register the same tuple for a given kqueue will result in the replacement of the conditions being watched, not an addition. Whether the udata value is considered as part of the tuple is controlled by the EV_UDATA_SPECIFIC flag on the kevent. .Pp The filter identified in a kevent is executed upon the initial registration of that event 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 . .Pp The .Fn kevent, .Fn kevent64 and .Fn kevent_qos system calls are 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, .Va kevent64_s or .Va kevent_qos_s structures, as defined in .Aq Pa 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 out .Va kevent, .Va kevent64_s or .Va kevent_qos_s structures. The .Fa nevents argument determines the size of .Fa eventlist . If the KEVENT_FLAG_STACK_EVENTS flag is provided on the system call, the eventlist array is filled in in stack order (starting in the highest available index) instead of typical array order. The .Fa out_data argument provides space for extra out data provided by specific filters. The .Fa data_available argument's contents specified the space available in the data pool on input, and contains the amount still remaining on output. If the KEVENT_FLAG_STACK_DATA flag is specified on the system call, the data is allocated from the pool in stack order instead of typical heap order. 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, both .Fn kevent and .Fn kevent64 wait indefinitely. To effect a poll, the .Fa flags argument to .Fn kevent64 or .Fn kevent_qos can include the KEVENT_FLAG_IMMEDIATE value to indicate an immediate timeout. Alternatively, 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 .Va kevent structure. Similarly, .Fn EV_SET64 initializes a .Va kevent64_s structure and .Fn EV_SET_QOS initializes a .Va kevent_qos_s structure. .Pp The .Va kevent, .Va kevent64_s and .Va kevent_qos_s structures are defined as: .Bd -literal struct kevent { uintptr_t ident; /* identifier for this event */ int16_t filter; /* filter for event */ uint16_t flags; /* general flags */ uint32_t fflags; /* filter-specific flags */ intptr_t data; /* filter-specific data */ void *udata; /* opaque user data identifier */ }; struct kevent64_s { uint64_t ident; /* identifier for this event */ int16_t filter; /* filter for event */ uint16_t flags; /* general flags */ uint32_t fflags; /* filter-specific flags */ int64_t data; /* filter-specific data */ uint64_t udata; /* opaque user data identifier */ uint64_t ext[2]; /* filter-specific extensions */ }; struct kevent_qos_s { uint64_t ident; /* identifier for this event */ int16_t filter; /* filter for event */ uint16_t flags; /* general flags */ uint32_t qos; /* quality of service when servicing event */ uint64_t udata; /* opaque user data identifier */ uint32_t fflags; /* filter-specific flags */ uint32_t xflags; /* extra filter-specific flags */ int64_t data; /* filter-specific data */ uint64_t ext[4]; /* filter-specific extensions */ }; .Ed .Pp ---- .Pp The fields of .Fa struct kevent, .Fa struct kevent64_s and .Fa struct kevent_qos_s are: .Bl -tag -width XXXfilter .It ident Value used to identify the source of the event. The exact interpretation is determined by the attached filter, but often is a file descriptor. .It filter Identifies the kernel filter used to process this event. The pre-defined system filters are described below. .It flags Actions to perform on the event. .It fflags Filter-specific flags. .It data Filter-specific data value. .It udata Opaque user-defined value passed through the kernel unchanged. It can optionally be part of the uniquing decision of the kevent system .El .Pp In addition, .Fa struct kevent64_s contains: .Bl -tag -width XXXfilter .It ext[2] This field stores extensions for the event's filter. What type of extension depends on what type of filter is being used. .El .Pp In addition, .Fa struct kevent_qos_s contains: .Bl -tag -width XXXfilter .It xflags Extra filter-specific flags. .It ext[4] The QoS variant provides twice as many extension values for filter-specific uses. .El .Pp ---- .Pp The .Va flags field can contain the following values: .Bl -tag -width XXXEV_ONESHOT .It 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 EV_ENABLE Permit .Fn kevent, .Fn kevent64 and .Fn kevent_qos to return the event if it is triggered. .It EV_DISABLE Disable the event so .Fn kevent, .Fn kevent64 and .Fn kevent_qos will not return it. The filter itself is not disabled. .It 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 EV_RECEIPT This flag is useful for making bulk changes to a kqueue without draining any pending events. When passed as input, it forces EV_ERROR to always be returned. When a filter is successfully added, the .Va data field will be zero. .It 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 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 EV_EOF Filters may set this flag to indicate filter-specific EOF condition. .It EV_OOBAND Read filter on socket may set this flag to indicate the presence of out of band data on the descriptor. .It EV_ERROR See .Sx RETURN VALUES below. .El .Pp ---- .Pp The predefined system filters are listed below. Arguments may be passed to and from the filter via the .Va data, .Va fflags and optionally .Va xflags fields in the .Va kevent, .Va kevent64_s or .Va kevent_qos_s structure. .Bl -tag -width EVFILT_MACHPORT .It EVFILT_READ Takes a file 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. .Pp .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 NOTE_LOWAT flag in .Va fflags , and specifying the new low water mark in .Va data . The derived per filter low water mark value is, however, bounded by socket receive buffer's high and low water mark values. On return, .Va data contains the number of bytes of protocol data available to read. .Pp The presence of EV_OOBAND in .Va flags , indicates the presence of out of band data on the socket .Va data equal to the potential number of OOB bytes availble to read. .Pp If the read direction of the socket has shutdown, then the filter also sets 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. .It "Fifos, Pipes" Returns when there is data to read; .Va data contains the number of bytes available. .Pp When the last writer disconnects, the filter will set EV_EOF in .Va flags . This may be cleared by passing in EV_CLEAR, at which point the filter will resume waiting for data to become available before returning. .It "Device nodes" Returns when there is data to read from the device; .Va data contains the number of bytes available. If the device does not support returning number of bytes, it will not allow the filter to be attached. However, if the NOTE_LOWAT flag is specified and the .Va data field contains 1 on input, those devices will attach - but cannot be relied upon to provide an accurate count of bytes to be read on output. .El .It EVFILT_WRITE Takes a file 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 EV_EOF when the reader disconnects, and for the fifo case, this may be cleared by use of EV_CLEAR. Note that this filter is not supported for vnodes. .Pp For sockets, the low water mark and socket error handling is identical to the EVFILT_READ case. .It EVFILT_AIO This filter is currently unsupported. .\"The sigevent portion of the AIO request is filled in, with .\".Va sigev_notify_kqueue .\"containing the descriptor of the kqueue that the event should .\"be attached to, .\".Va sigev_value .\"containing the udata value, and .\".Va sigev_notify .\"set to SIGEV_KEVENT. .\"When the .\".Fn aio_* .\"system call is made, the event will be registered .\"with the specified kqueue, and the .\".Va ident .\"argument set to the .\".Fa struct aiocb .\"returned by the .\".Fn aio_* .\"system call. .\"The filter returns under the same conditions as aio_error. .\".Pp .\"Alternatively, a kevent structure may be initialized, with .\".Va ident .\"containing the descriptor of the kqueue, and the .\"address of the kevent structure placed in the .\".Va aio_lio_opcode .\"field of the AIO request. However, this approach will not work on .\"architectures with 64-bit pointers, and should be considered deprecated. .It 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 XXNOTE_RENAME .It NOTE_DELETE The .Fn unlink system call was called on the file referenced by the descriptor. .It NOTE_WRITE A write occurred on the file referenced by the descriptor. .It NOTE_EXTEND The file referenced by the descriptor was extended. .It NOTE_ATTRIB The file referenced by the descriptor had its attributes changed. .It NOTE_LINK The link count on the file changed. .It NOTE_RENAME The file referenced by the descriptor was renamed. .It NOTE_REVOKE Access to the file was revoked via .Xr revoke 2 or the underlying fileystem was unmounted. .El .Pp On return, .Va fflags contains the filter-specific flags which are associated with the triggered events seen by this filter. .It 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 NOTE_SIGNAL .It NOTE_EXIT The process has exited. .It NOTE_EXITSTATUS The process has exited and its exit status is in filter specific data. Valid only on child processes and to be used along with NOTE_EXIT. .It NOTE_FORK The process created a child process via .Xr fork 2 or similar call. .It NOTE_EXEC The process executed a new process via .Xr execve 2 or similar call. .It NOTE_SIGNAL The process was sent a signal. Status can be checked via .Xr waitpid 2 or similar call. .It NOTE_REAP The process was reaped by the parent via .Xr wait 2 or similar call. Deprecated, use NOTE_EXIT. .El .Pp On return, .Va fflags contains the events which triggered the filter. .It EVFILT_SIGNAL Takes the signal number to monitor as the identifier and returns when the given signal is generated for the process. This coexists with the .Fn signal and .Fn sigaction facilities, and has a lower precedence. Only signals sent to the process, not to a particular thread, will trigger the filter. The filter will record all attempts to deliver a signal to a process, even if the signal has been marked as SIG_IGN. Event notification happens before normal signal delivery processing. .Va data returns the number of times the signal has been generated since the last call to .Fn kevent . This filter automatically sets the EV_CLEAR flag internally. .It EVFILT_MACHPORT Takes the name of a mach port, or port set, in .Va ident and waits until a message is enqueued on the port or port set. When a message is detected, but not directly received by the kevent call, the name of the specific port where the message is enqueued is returned in .Va data . If .Va fflags contains MACH_RCV_MSG, the ext[0] and ext[1] flags are assumed to contain a pointer to the buffer where the message is to be received and the size of the receive buffer, respectively. If MACH_RCV_MSG is specifed, yet the buffer size in ext[1] is zero, The space for the buffer may be carved out of the .Va data_out area provided to .Fn kevent_qos if there is enough space remaining there. .It EVFILT_TIMER Establishes an interval timer with the data timer identified by .Va ident . When adding a timer, .Va data specifies the timeout period and .Va fflags can be set to one of the following: .Bl -tag -width NOTE_ABSOLUTE .It NOTE_SECONDS data is in seconds .It NOTE_USECONDS data is in microseconds .It NOTE_NSECONDS data is in nanoseconds .It NOTE_ABSOLUTE data is an absolute timeout .It NOTE_CRITICAL system makes a best effort to fire this timer as scheduled. .It NOTE_BACKGROUND system has extra leeway to coalesce this timer. .It NOTE_LEEWAY ext[1] holds user-supplied slop in deadline for timer coalescing. .El .Pp If fflags is not set, the default is milliseconds. The timer will be periodic unless EV_ONESHOT is specified. On return, .Va data contains the number of times the timeout has expired since the last call to .Fn kevent , .Fn kevent64 or .Fn kevent_qos . This filter automatically sets the EV_CLEAR flag internally. .El .Pp ---- .Pp In the .Va ext[2] field of the .Va kevent64_s struture, .Va ext[0] is only used with the EVFILT_MACHPORT filter. With other filters, .Va ext[0] is passed through .Fn kevent64 much like .Va udata . .Va ext[1] can always be used like .Va udata . For the use of ext[0], see the EVFILT_MACHPORT filter above. .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 , .Fn kevent64 and .Fn kevent_qos system calls return 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 , .Fn kevent64 and .Fn kevent_qos return 0. .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 EMFILE The per-process descriptor table is full. .It Bq Er ENFILE The system file table is full. .El .Pp The .Fn kevent and .Fn kevent64 system calls fail 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 or .Va kevent64_s 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 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. .It Bq Er ESRCH The specified process to attach to does not exist. .El .Sh SEE ALSO .Xr aio_error 2 , .Xr aio_read 2 , .Xr aio_return 2 , .Xr read 2 , .Xr select 2 , .Xr sigaction 2 , .Xr write 2 , .Xr signal 3 .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 jlemon@FreeBSD.org . .Sh BUGS Not all filesystem types support kqueue-style notifications. And even some that do, like some remote filesystems, may only support a subset of the notification semantics described here.