1.\" Copyright (c) 2000 Jonathan Lemon 2.\" All rights reserved. 3.\" 4.\" Redistribution and use in source and binary forms, with or without 5.\" modification, are permitted provided that the following conditions 6.\" are met: 7.\" 1. Redistributions of source code must retain the above copyright 8.\" notice, this list of conditions and the following disclaimer. 9.\" 2. Redistributions in binary form must reproduce the above copyright 10.\" notice, this list of conditions and the following disclaimer in the 11.\" documentation and/or other materials provided with the distribution. 12.\" 13.\" THIS SOFTWARE IS PROVIDED ``AS IS'' AND 14.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 15.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 16.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 17.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 18.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 19.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 20.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 21.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 22.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 23.\" SUCH DAMAGE. 24.\" 25.\" $FreeBSD$ 26.\" 27.Dd June 4, 2005 28.Dt KQUEUE 2 29.Os 30.Sh NAME 31.Nm kqueue , 32.Nm kevent 33.Nd kernel event notification mechanism 34.Sh LIBRARY 35.Lb libc 36.Sh SYNOPSIS 37.In sys/types.h 38.In sys/event.h 39.In sys/time.h 40.Ft int 41.Fn kqueue "void" 42.Ft int 43.Fn kevent "int kq" "const struct kevent *changelist" "int nchanges" "struct kevent *eventlist" "int nevents" "const struct timespec *timeout" 44.Fn EV_SET "&kev" ident filter flags fflags data udata 45.Sh DESCRIPTION 46The 47.Fn kqueue 48system call 49provides a generic method of notifying the user when an event 50happens or a condition holds, based on the results of small 51pieces of kernel code termed filters. 52A kevent is identified by the (ident, filter) pair; there may only 53be one unique kevent per kqueue. 54.Pp 55The filter is executed upon the initial registration of a kevent 56in order to detect whether a preexisting condition is present, and is also 57executed whenever an event is passed to the filter for evaluation. 58If the filter determines that the condition should be reported, 59then the kevent is placed on the kqueue for the user to retrieve. 60.Pp 61The filter is also run when the user attempts to retrieve the kevent 62from the kqueue. 63If the filter indicates that the condition that triggered 64the event no longer holds, the kevent is removed from the kqueue and 65is not returned. 66.Pp 67Multiple events which trigger the filter do not result in multiple 68kevents being placed on the kqueue; instead, the filter will aggregate 69the events into a single struct kevent. 70Calling 71.Fn close 72on a file descriptor will remove any kevents that reference the descriptor. 73.Pp 74The 75.Fn kqueue 76system call 77creates a new kernel event queue and returns a descriptor. 78The queue is not inherited by a child created with 79.Xr fork 2 . 80However, if 81.Xr rfork 2 82is called without the 83.Dv RFFDG 84flag, then the descriptor table is shared, 85which will allow sharing of the kqueue between two processes. 86.Pp 87The 88.Fn kevent 89system call 90is used to register events with the queue, and return any pending 91events to the user. 92The 93.Fa changelist 94argument 95is a pointer to an array of 96.Va kevent 97structures, as defined in 98.In sys/event.h . 99All changes contained in the 100.Fa changelist 101are applied before any pending events are read from the queue. 102The 103.Fa nchanges 104argument 105gives the size of 106.Fa changelist . 107The 108.Fa eventlist 109argument 110is a pointer to an array of kevent structures. 111The 112.Fa nevents 113argument 114determines the size of 115.Fa eventlist . 116When 117.Fa nevents 118is zero, 119.Fn kevent 120will return immediately even if there is a 121.Fa timeout 122specified unlike 123.Xr select 2 . 124If 125.Fa timeout 126is a non-NULL pointer, it specifies a maximum interval to wait 127for an event, which will be interpreted as a struct timespec. 128If 129.Fa timeout 130is a NULL pointer, 131.Fn kevent 132waits indefinitely. 133To effect a poll, the 134.Fa timeout 135argument should be non-NULL, pointing to a zero-valued 136.Va timespec 137structure. 138The same array may be used for the 139.Fa changelist 140and 141.Fa eventlist . 142.Pp 143The 144.Fn EV_SET 145macro is provided for ease of initializing a 146kevent structure. 147.Pp 148The 149.Va kevent 150structure is defined as: 151.Bd -literal 152struct kevent { 153 uintptr_t ident; /* identifier for this event */ 154 short filter; /* filter for event */ 155 u_short flags; /* action flags for kqueue */ 156 u_int fflags; /* filter flag value */ 157 intptr_t data; /* filter data value */ 158 void *udata; /* opaque user data identifier */ 159}; 160.Ed 161.Pp 162The fields of 163.Fa struct kevent 164are: 165.Bl -tag -width XXXfilter 166.It ident 167Value used to identify this event. 168The exact interpretation is determined by the attached filter, 169but often is a file descriptor. 170.It filter 171Identifies the kernel filter used to process this event. 172The pre-defined 173system filters are described below. 174.It flags 175Actions to perform on the event. 176.It fflags 177Filter-specific flags. 178.It data 179Filter-specific data value. 180.It udata 181Opaque user-defined value passed through the kernel unchanged. 182.El 183.Pp 184The 185.Va flags 186field can contain the following values: 187.Bl -tag -width XXXEV_ONESHOT 188.It EV_ADD 189Adds the event to the kqueue. 190Re-adding an existing event 191will modify the parameters of the original event, and not result 192in a duplicate entry. 193Adding an event automatically enables it, 194unless overridden by the EV_DISABLE flag. 195.It EV_ENABLE 196Permit 197.Fn kevent 198to return the event if it is triggered. 199.It EV_DISABLE 200Disable the event so 201.Fn kevent 202will not return it. 203The filter itself is not disabled. 204.It EV_DELETE 205Removes the event from the kqueue. 206Events which are attached to 207file descriptors are automatically deleted on the last close of 208the descriptor. 209.It EV_ONESHOT 210Causes the event to return only the first occurrence of the filter 211being triggered. 212After the user retrieves the event from the kqueue, 213it is deleted. 214.It EV_CLEAR 215After the event is retrieved by the user, its state is reset. 216This is useful for filters which report state transitions 217instead of the current state. 218Note that some filters may automatically 219set this flag internally. 220.It EV_EOF 221Filters may set this flag to indicate filter-specific EOF condition. 222.It EV_ERROR 223See 224.Sx RETURN VALUES 225below. 226.El 227.Pp 228The predefined system filters are listed below. 229Arguments may be passed to and from the filter via the 230.Va fflags 231and 232.Va data 233fields in the kevent structure. 234.Bl -tag -width EVFILT_SIGNAL 235.It EVFILT_READ 236Takes a descriptor as the identifier, and returns whenever 237there is data available to read. 238The behavior of the filter is slightly different depending 239on the descriptor type. 240.Pp 241.Bl -tag -width 2n 242.It Sockets 243Sockets which have previously been passed to 244.Fn listen 245return when there is an incoming connection pending. 246.Va data 247contains the size of the listen backlog. 248.Pp 249Other socket descriptors return when there is data to be read, 250subject to the 251.Dv SO_RCVLOWAT 252value of the socket buffer. 253This may be overridden with a per-filter low water mark at the 254time the filter is added by setting the 255NOTE_LOWAT 256flag in 257.Va fflags , 258and specifying the new low water mark in 259.Va data . 260On return, 261.Va data 262contains the number of bytes of protocol data available to read. 263.Pp 264If the read direction of the socket has shutdown, then the filter 265also sets EV_EOF in 266.Va flags , 267and returns the socket error (if any) in 268.Va fflags . 269It is possible for EOF to be returned (indicating the connection is gone) 270while there is still data pending in the socket buffer. 271.It Vnodes 272Returns when the file pointer is not at the end of file. 273.Va data 274contains the offset from current position to end of file, 275and may be negative. 276.It "Fifos, Pipes" 277Returns when the there is data to read; 278.Va data 279contains the number of bytes available. 280.Pp 281When the last writer disconnects, the filter will set EV_EOF in 282.Va flags . 283This may be cleared by passing in EV_CLEAR, at which point the 284filter will resume waiting for data to become available before 285returning. 286.It "BPF devices" 287Returns when the BPF buffer is full, the BPF timeout has expired, or 288when the BPF has 289.Dq immediate mode 290enabled and there is any data to read; 291.Va data 292contains the number of bytes available. 293.El 294.It EVFILT_WRITE 295Takes a descriptor as the identifier, and returns whenever 296it is possible to write to the descriptor. 297For sockets, pipes 298and fifos, 299.Va data 300will contain the amount of space remaining in the write buffer. 301The filter will set EV_EOF when the reader disconnects, and for 302the fifo case, this may be cleared by use of EV_CLEAR. 303Note that this filter is not supported for vnodes or BPF devices. 304.Pp 305For sockets, the low water mark and socket error handling is 306identical to the EVFILT_READ case. 307.It EVFILT_AIO 308The sigevent portion of the AIO request is filled in, with 309.Va sigev_notify_kqueue 310containing the descriptor of the kqueue that the event should 311be attached to, 312.Va sigev_value 313containing the udata value, and 314.Va sigev_notify 315set to SIGEV_KEVENT. 316When the 317.Fn aio_* 318system call is made, the event will be registered 319with the specified kqueue, and the 320.Va ident 321argument set to the 322.Fa struct aiocb 323returned by the 324.Fn aio_* 325system call. 326The filter returns under the same conditions as aio_error. 327.It EVFILT_VNODE 328Takes a file descriptor as the identifier and the events to watch for in 329.Va fflags , 330and returns when one or more of the requested events occurs on the descriptor. 331The events to monitor are: 332.Bl -tag -width XXNOTE_RENAME 333.It NOTE_DELETE 334The 335.Fn unlink 336system call 337was called on the file referenced by the descriptor. 338.It NOTE_WRITE 339A write occurred on the file referenced by the descriptor. 340.It NOTE_EXTEND 341The file referenced by the descriptor was extended. 342.It NOTE_ATTRIB 343The file referenced by the descriptor had its attributes changed. 344.It NOTE_LINK 345The link count on the file changed. 346.It NOTE_RENAME 347The file referenced by the descriptor was renamed. 348.It NOTE_REVOKE 349Access to the file was revoked via 350.Xr revoke 2 351or the underlying file system was unmounted. 352.El 353.Pp 354On return, 355.Va fflags 356contains the events which triggered the filter. 357.It EVFILT_PROC 358Takes the process ID to monitor as the identifier and the events to watch for 359in 360.Va fflags , 361and returns when the process performs one or more of the requested events. 362If a process can normally see another process, it can attach an event to it. 363The events to monitor are: 364.Bl -tag -width XXNOTE_TRACKERR 365.It NOTE_EXIT 366The process has exited. 367.It NOTE_FORK 368The process has called 369.Fn fork . 370.It NOTE_EXEC 371The process has executed a new process via 372.Xr execve 2 373or similar call. 374.It NOTE_TRACK 375Follow a process across 376.Fn fork 377calls. 378The parent process will return with NOTE_TRACK set in the 379.Va fflags 380field, while the child process will return with NOTE_CHILD set in 381.Va fflags 382and the parent PID in 383.Va data . 384.It NOTE_TRACKERR 385This flag is returned if the system was unable to attach an event to 386the child process, usually due to resource limitations. 387.El 388.Pp 389On return, 390.Va fflags 391contains the events which triggered the filter. 392.It EVFILT_SIGNAL 393Takes the signal number to monitor as the identifier and returns 394when the given signal is delivered to the process. 395This coexists with the 396.Fn signal 397and 398.Fn sigaction 399facilities, and has a lower precedence. 400The filter will record 401all attempts to deliver a signal to a process, even if the signal has 402been marked as SIG_IGN. 403Event notification happens after normal 404signal delivery processing. 405.Va data 406returns the number of times the signal has occurred since the last call to 407.Fn kevent . 408This filter automatically sets the EV_CLEAR flag internally. 409.It EVFILT_TIMER 410Establishes an arbitrary timer identified by 411.Va ident . 412When adding a timer, 413.Va data 414specifies the timeout period in milliseconds. 415The timer will be periodic unless EV_ONESHOT is specified. 416On return, 417.Va data 418contains the number of times the timeout has expired since the last call to 419.Fn kevent . 420This filter automatically sets the EV_CLEAR flag internally. 421There is a system wide limit on the number of timers 422which is controlled by the 423.Va kern.kq_calloutmax 424sysctl. 425.It Dv EVFILT_NETDEV 426Takes a descriptor to a network interface as the identifier, and the events to watch for in 427.Va fflags . 428It returns, when one or more of the requested events occur on the descriptor. 429The events to monitor are: 430.Bl -tag -width XXNOTE_LINKDOWN 431.It Dv NOTE_LINKUP 432The link is up. 433.It Dv NOTE_LINKDOWN 434The link is down. 435.It Dv NOTE_LINKINV 436The link state is invalid. 437.El 438.Pp 439On return, 440.Va fflags 441contains the events which triggered the filter. 442.El 443.Sh RETURN VALUES 444The 445.Fn kqueue 446system call 447creates a new kernel event queue and returns a file descriptor. 448If there was an error creating the kernel event queue, a value of -1 is 449returned and errno set. 450.Pp 451The 452.Fn kevent 453system call 454returns the number of events placed in the 455.Fa eventlist , 456up to the value given by 457.Fa nevents . 458If an error occurs while processing an element of the 459.Fa changelist 460and there is enough room in the 461.Fa eventlist , 462then the event will be placed in the 463.Fa eventlist 464with 465.Dv EV_ERROR 466set in 467.Va flags 468and the system error in 469.Va data . 470Otherwise, 471.Dv -1 472will be returned, and 473.Dv errno 474will be set to indicate the error condition. 475If the time limit expires, then 476.Fn kevent 477returns 0. 478.Sh ERRORS 479The 480.Fn kqueue 481system call fails if: 482.Bl -tag -width Er 483.It Bq Er ENOMEM 484The kernel failed to allocate enough memory for the kernel queue. 485.It Bq Er EMFILE 486The per-process descriptor table is full. 487.It Bq Er ENFILE 488The system file table is full. 489.El 490.Pp 491The 492.Fn kevent 493system call fails if: 494.Bl -tag -width Er 495.It Bq Er EACCES 496The process does not have permission to register a filter. 497.It Bq Er EFAULT 498There was an error reading or writing the 499.Va kevent 500structure. 501.It Bq Er EBADF 502The specified descriptor is invalid. 503.It Bq Er EINTR 504A signal was delivered before the timeout expired and before any 505events were placed on the kqueue for return. 506.It Bq Er EINVAL 507The specified time limit or filter is invalid. 508.It Bq Er ENOENT 509The event could not be found to be modified or deleted. 510.It Bq Er ENOMEM 511No memory was available to register the event 512or, in the special case of a timer, the maximum number of 513timers has been exceeded. 514This maximum is configurable via the 515.Va kern.kq_calloutmax 516sysctl. 517.It Bq Er ESRCH 518The specified process to attach to does not exist. 519.El 520.Sh SEE ALSO 521.Xr aio_error 2 , 522.Xr aio_read 2 , 523.Xr aio_return 2 , 524.Xr poll 2 , 525.Xr read 2 , 526.Xr select 2 , 527.Xr sigaction 2 , 528.Xr write 2 , 529.Xr signal 3 530.Sh HISTORY 531The 532.Fn kqueue 533and 534.Fn kevent 535system calls first appeared in 536.Fx 4.1 . 537.Sh AUTHORS 538The 539.Fn kqueue 540system and this manual page were written by 541.An Jonathan Lemon Aq [email protected] . 542.Sh BUGS 543It is currently not possible to watch a 544.Xr vnode 9 545that resides on anything but 546a UFS file system. 547.Pp 548The 549.Dv EVFILT_NETDEV 550filter is currently only implemented for devices that use the 551.Xr miibus 4 552driver for LINKUP and LINKDOWN operations. 553Therefore, it will not work with many non-ethernet devices. 554.Pp 555The 556.Fa timeout 557value is limited to 24 hours; longer timeouts will be silently 558reinterpreted as 24 hours. 559