ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/libev/ev.3
(Generate patch)

Comparing libev/ev.3 (file contents):
Revision 1.8 by root, Fri Nov 23 15:26:08 2007 UTC vs.
Revision 1.72 by root, Tue Oct 21 20:06:52 2008 UTC

1.\" Automatically generated by Pod::Man v1.37, Pod::Parser v1.35 1.\" Automatically generated by Pod::Man 2.16 (Pod::Simple 3.05)
2.\" 2.\"
3.\" Standard preamble: 3.\" Standard preamble:
4.\" ======================================================================== 4.\" ========================================================================
5.de Sh \" Subsection heading 5.de Sh \" Subsection heading
6.br 6.br
… …
23.ft R 23.ft R
24.fi 24.fi
25.. 25..
26.\" Set up some character translations and predefined strings. \*(-- will 26.\" Set up some character translations and predefined strings. \*(-- will
27.\" give an unbreakable dash, \*(PI will give pi, \*(L" will give a left 27.\" give an unbreakable dash, \*(PI will give pi, \*(L" will give a left
28.\" double quote, and \*(R" will give a right double quote. | will give a 28.\" double quote, and \*(R" will give a right double quote. \*(C+ will
29.\" real vertical bar. \*(C+ will give a nicer C++. Capital omega is used to 29.\" give a nicer C++. Capital omega is used to do unbreakable dashes and
30.\" do unbreakable dashes and therefore won't be available. \*(C` and \*(C' 30.\" therefore won't be available. \*(C` and \*(C' expand to `' in nroff,
31.\" expand to `' in nroff, nothing in troff, for use with C<>. 31.\" nothing in troff, for use with C<>.
32.tr \(*W-|\(bv\*(Tr 32.tr \(*W-
33.ds C+ C\v'-.1v'\h'-1p'\s-2+\h'-1p'+\s0\v'.1v'\h'-1p' 33.ds C+ C\v'-.1v'\h'-1p'\s-2+\h'-1p'+\s0\v'.1v'\h'-1p'
34.ie n \{\ 34.ie n \{\
35. ds -- \(*W- 35. ds -- \(*W-
36. ds PI pi 36. ds PI pi
37. if (\n(.H=4u)&(1m=24u) .ds -- \(*W\h'-12u'\(*W\h'-12u'-\" diablo 10 pitch 37. if (\n(.H=4u)&(1m=24u) .ds -- \(*W\h'-12u'\(*W\h'-12u'-\" diablo 10 pitch
… …
46. ds PI \(*p 46. ds PI \(*p
47. ds L" `` 47. ds L" ``
48. ds R" '' 48. ds R" ''
49'br\} 49'br\}
50.\" 50.\"
51.\" Escape single quotes in literal strings from groff's Unicode transform.
52.ie \n(.g .ds Aq \(aq
53.el .ds Aq '
54.\"
51.\" If the F register is turned on, we'll generate index entries on stderr for 55.\" If the F register is turned on, we'll generate index entries on stderr for
52.\" titles (.TH), headers (.SH), subsections (.Sh), items (.Ip), and index 56.\" titles (.TH), headers (.SH), subsections (.Sh), items (.Ip), and index
53.\" entries marked with X<> in POD. Of course, you'll have to process the 57.\" entries marked with X<> in POD. Of course, you'll have to process the
54.\" output yourself in some meaningful fashion. 58.\" output yourself in some meaningful fashion.
55.if \nF \{\ 59.ie \nF \{\
56. de IX 60. de IX
57. tm Index:\\$1\t\\n%\t"\\$2" 61. tm Index:\\$1\t\\n%\t"\\$2"
58.. 62..
59. nr % 0 63. nr % 0
60. rr F 64. rr F
61.\} 65.\}
62.\" 66.el \{\
63.\" For nroff, turn off justification. Always turn off hyphenation; it makes 67. de IX
64.\" way too many mistakes in technical documents. 68..
65.hy 0 69.\}
66.if n .na
67.\" 70.\"
68.\" Accent mark definitions (@(#)ms.acc 1.5 88/02/08 SMI; from UCB 4.2). 71.\" Accent mark definitions (@(#)ms.acc 1.5 88/02/08 SMI; from UCB 4.2).
69.\" Fear. Run. Save yourself. No user-serviceable parts. 72.\" Fear. Run. Save yourself. No user-serviceable parts.
70. \" fudge factors for nroff and troff 73. \" fudge factors for nroff and troff
71.if n \{\ 74.if n \{\
… …
126. ds Ae AE 129. ds Ae AE
127.\} 130.\}
128.rm #[ #] #H #V #F C 131.rm #[ #] #H #V #F C
129.\" ======================================================================== 132.\" ========================================================================
130.\" 133.\"
131.IX Title ""<STANDARD INPUT>" 1" 134.IX Title "LIBEV 3"
132.TH "<STANDARD INPUT>" 1 "2007-11-23" "perl v5.8.8" "User Contributed Perl Documentation" 135.TH LIBEV 3 "2008-10-21" "libev-3.45" "libev - high performance full featured event loop"
136.\" For nroff, turn off justification. Always turn off hyphenation; it makes
137.\" way too many mistakes in technical documents.
138.if n .ad l
139.nh
133.SH "NAME" 140.SH "NAME"
134libev \- a high performance full\-featured event loop written in C 141libev \- a high performance full\-featured event loop written in C
135.SH "SYNOPSIS" 142.SH "SYNOPSIS"
136.IX Header "SYNOPSIS" 143.IX Header "SYNOPSIS"
137.Vb 1 144.Vb 1
138\& #include <ev.h> 145\& #include <ev.h>
146.Ve
147.Sh "\s-1EXAMPLE\s0 \s-1PROGRAM\s0"
148.IX Subsection "EXAMPLE PROGRAM"
149.Vb 2
150\& // a single header file is required
151\& #include <ev.h>
152\&
153\& // every watcher type has its own typedef\*(Aqd struct
154\& // with the name ev_<type>
155\& ev_io stdin_watcher;
156\& ev_timer timeout_watcher;
157\&
158\& // all watcher callbacks have a similar signature
159\& // this callback is called when data is readable on stdin
160\& static void
161\& stdin_cb (EV_P_ struct ev_io *w, int revents)
162\& {
163\& puts ("stdin ready");
164\& // for one\-shot events, one must manually stop the watcher
165\& // with its corresponding stop function.
166\& ev_io_stop (EV_A_ w);
167\&
168\& // this causes all nested ev_loop\*(Aqs to stop iterating
169\& ev_unloop (EV_A_ EVUNLOOP_ALL);
170\& }
171\&
172\& // another callback, this time for a time\-out
173\& static void
174\& timeout_cb (EV_P_ struct ev_timer *w, int revents)
175\& {
176\& puts ("timeout");
177\& // this causes the innermost ev_loop to stop iterating
178\& ev_unloop (EV_A_ EVUNLOOP_ONE);
179\& }
180\&
181\& int
182\& main (void)
183\& {
184\& // use the default event loop unless you have special needs
185\& struct ev_loop *loop = ev_default_loop (0);
186\&
187\& // initialise an io watcher, then start it
188\& // this one will watch for stdin to become readable
189\& ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
190\& ev_io_start (loop, &stdin_watcher);
191\&
192\& // initialise a timer watcher, then start it
193\& // simple non\-repeating 5.5 second timeout
194\& ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
195\& ev_timer_start (loop, &timeout_watcher);
196\&
197\& // now wait for events to arrive
198\& ev_loop (loop, 0);
199\&
200\& // unloop was called, so exit
201\& return 0;
202\& }
139.Ve 203.Ve
140.SH "DESCRIPTION" 204.SH "DESCRIPTION"
141.IX Header "DESCRIPTION" 205.IX Header "DESCRIPTION"
206The newest version of this document is also available as an html-formatted
207web page you might find easier to navigate when reading it for the first
208time: <http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
209.PP
142Libev is an event loop: you register interest in certain events (such as a 210Libev is an event loop: you register interest in certain events (such as a
143file descriptor being readable or a timeout occuring), and it will manage 211file descriptor being readable or a timeout occurring), and it will manage
144these event sources and provide your program with events. 212these event sources and provide your program with events.
145.PP 213.PP
146To do this, it must take more or less complete control over your process 214To do this, it must take more or less complete control over your process
147(or thread) by executing the \fIevent loop\fR handler, and will then 215(or thread) by executing the \fIevent loop\fR handler, and will then
148communicate events via a callback mechanism. 216communicate events via a callback mechanism.
149.PP 217.PP
150You register interest in certain events by registering so-called \fIevent 218You register interest in certain events by registering so-called \fIevent
151watchers\fR, which are relatively small C structures you initialise with the 219watchers\fR, which are relatively small C structures you initialise with the
152details of the event, and then hand it over to libev by \fIstarting\fR the 220details of the event, and then hand it over to libev by \fIstarting\fR the
153watcher. 221watcher.
154.SH "FEATURES" 222.Sh "\s-1FEATURES\s0"
155.IX Header "FEATURES" 223.IX Subsection "FEATURES"
156Libev supports select, poll, the linux-specific epoll and the bsd-specific 224Libev supports \f(CW\*(C`select\*(C'\fR, \f(CW\*(C`poll\*(C'\fR, the Linux-specific \f(CW\*(C`epoll\*(C'\fR, the
157kqueue mechanisms for file descriptor events, relative timers, absolute 225BSD-specific \f(CW\*(C`kqueue\*(C'\fR and the Solaris-specific event port mechanisms
158timers with customised rescheduling, signal events, process status change 226for file descriptor events (\f(CW\*(C`ev_io\*(C'\fR), the Linux \f(CW\*(C`inotify\*(C'\fR interface
159events (related to \s-1SIGCHLD\s0), and event watchers dealing with the event 227(for \f(CW\*(C`ev_stat\*(C'\fR), relative timers (\f(CW\*(C`ev_timer\*(C'\fR), absolute timers
160loop mechanism itself (idle, prepare and check watchers). It also is quite 228with customised rescheduling (\f(CW\*(C`ev_periodic\*(C'\fR), synchronous signals
161fast (see this benchmark comparing 229(\f(CW\*(C`ev_signal\*(C'\fR), process status change events (\f(CW\*(C`ev_child\*(C'\fR), and event
162it to libevent for example). 230watchers dealing with the event loop mechanism itself (\f(CW\*(C`ev_idle\*(C'\fR,
231\&\f(CW\*(C`ev_embed\*(C'\fR, \f(CW\*(C`ev_prepare\*(C'\fR and \f(CW\*(C`ev_check\*(C'\fR watchers) as well as
232file watchers (\f(CW\*(C`ev_stat\*(C'\fR) and even limited support for fork events
233(\f(CW\*(C`ev_fork\*(C'\fR).
234.PP
235It also is quite fast (see this
236benchmark comparing it to libevent
237for example).
163.SH "CONVENTIONS" 238.Sh "\s-1CONVENTIONS\s0"
164.IX Header "CONVENTIONS" 239.IX Subsection "CONVENTIONS"
165Libev is very configurable. In this manual the default configuration 240Libev is very configurable. In this manual the default (and most common)
166will be described, which supports multiple event loops. For more info 241configuration will be described, which supports multiple event loops. For
167about various configuration options please have a look at the file 242more info about various configuration options please have a look at
168\&\fI\s-1README\s0.embed\fR in the libev distribution. If libev was configured without 243\&\fB\s-1EMBED\s0\fR section in this manual. If libev was configured without support
169support for multiple event loops, then all functions taking an initial 244for multiple event loops, then all functions taking an initial argument of
170argument of name \f(CW\*(C`loop\*(C'\fR (which is always of type \f(CW\*(C`struct ev_loop *\*(C'\fR) 245name \f(CW\*(C`loop\*(C'\fR (which is always of type \f(CW\*(C`struct ev_loop *\*(C'\fR) will not have
171will not have this argument. 246this argument.
172.SH "TIME REPRESENTATION" 247.Sh "\s-1TIME\s0 \s-1REPRESENTATION\s0"
173.IX Header "TIME REPRESENTATION" 248.IX Subsection "TIME REPRESENTATION"
174Libev represents time as a single floating point number, representing the 249Libev represents time as a single floating point number, representing the
175(fractional) number of seconds since the (\s-1POSIX\s0) epoch (somewhere near 250(fractional) number of seconds since the (\s-1POSIX\s0) epoch (somewhere near
176the beginning of 1970, details are complicated, don't ask). This type is 251the beginning of 1970, details are complicated, don't ask). This type is
177called \f(CW\*(C`ev_tstamp\*(C'\fR, which is what you should use too. It usually aliases 252called \f(CW\*(C`ev_tstamp\*(C'\fR, which is what you should use too. It usually aliases
178to the double type in C. 253to the \f(CW\*(C`double\*(C'\fR type in C, and when you need to do any calculations on
254it, you should treat it as some floating point value. Unlike the name
255component \f(CW\*(C`stamp\*(C'\fR might indicate, it is also used for time differences
256throughout libev.
257.SH "ERROR HANDLING"
258.IX Header "ERROR HANDLING"
259Libev knows three classes of errors: operating system errors, usage errors
260and internal errors (bugs).
261.PP
262When libev catches an operating system error it cannot handle (for example
263a system call indicating a condition libev cannot fix), it calls the callback
264set via \f(CW\*(C`ev_set_syserr_cb\*(C'\fR, which is supposed to fix the problem or
265abort. The default is to print a diagnostic message and to call \f(CW\*(C`abort
266()\*(C'\fR.
267.PP
268When libev detects a usage error such as a negative timer interval, then
269it will print a diagnostic message and abort (via the \f(CW\*(C`assert\*(C'\fR mechanism,
270so \f(CW\*(C`NDEBUG\*(C'\fR will disable this checking): these are programming errors in
271the libev caller and need to be fixed there.
272.PP
273Libev also has a few internal error-checking \f(CW\*(C`assert\*(C'\fRions, and also has
274extensive consistency checking code. These do not trigger under normal
275circumstances, as they indicate either a bug in libev or worse.
179.SH "GLOBAL FUNCTIONS" 276.SH "GLOBAL FUNCTIONS"
180.IX Header "GLOBAL FUNCTIONS" 277.IX Header "GLOBAL FUNCTIONS"
181These functions can be called anytime, even before initialising the 278These functions can be called anytime, even before initialising the
182library in any way. 279library in any way.
183.IP "ev_tstamp ev_time ()" 4 280.IP "ev_tstamp ev_time ()" 4
184.IX Item "ev_tstamp ev_time ()" 281.IX Item "ev_tstamp ev_time ()"
185Returns the current time as libev would use it. Please note that the 282Returns the current time as libev would use it. Please note that the
186\&\f(CW\*(C`ev_now\*(C'\fR function is usually faster and also often returns the timestamp 283\&\f(CW\*(C`ev_now\*(C'\fR function is usually faster and also often returns the timestamp
187you actually want to know. 284you actually want to know.
285.IP "ev_sleep (ev_tstamp interval)" 4
286.IX Item "ev_sleep (ev_tstamp interval)"
287Sleep for the given interval: The current thread will be blocked until
288either it is interrupted or the given time interval has passed. Basically
289this is a sub-second-resolution \f(CW\*(C`sleep ()\*(C'\fR.
188.IP "int ev_version_major ()" 4 290.IP "int ev_version_major ()" 4
189.IX Item "int ev_version_major ()" 291.IX Item "int ev_version_major ()"
190.PD 0 292.PD 0
191.IP "int ev_version_minor ()" 4 293.IP "int ev_version_minor ()" 4
192.IX Item "int ev_version_minor ()" 294.IX Item "int ev_version_minor ()"
193.PD 295.PD
194You can find out the major and minor version numbers of the library 296You can find out the major and minor \s-1ABI\s0 version numbers of the library
195you linked against by calling the functions \f(CW\*(C`ev_version_major\*(C'\fR and 297you linked against by calling the functions \f(CW\*(C`ev_version_major\*(C'\fR and
196\&\f(CW\*(C`ev_version_minor\*(C'\fR. If you want, you can compare against the global 298\&\f(CW\*(C`ev_version_minor\*(C'\fR. If you want, you can compare against the global
197symbols \f(CW\*(C`EV_VERSION_MAJOR\*(C'\fR and \f(CW\*(C`EV_VERSION_MINOR\*(C'\fR, which specify the 299symbols \f(CW\*(C`EV_VERSION_MAJOR\*(C'\fR and \f(CW\*(C`EV_VERSION_MINOR\*(C'\fR, which specify the
198version of the library your program was compiled against. 300version of the library your program was compiled against.
199.Sp 301.Sp
302These version numbers refer to the \s-1ABI\s0 version of the library, not the
303release version.
304.Sp
200Usually, it's a good idea to terminate if the major versions mismatch, 305Usually, it's a good idea to terminate if the major versions mismatch,
201as this indicates an incompatible change. Minor versions are usually 306as this indicates an incompatible change. Minor versions are usually
202compatible to older versions, so a larger minor version alone is usually 307compatible to older versions, so a larger minor version alone is usually
203not a problem. 308not a problem.
309.Sp
310Example: Make sure we haven't accidentally been linked against the wrong
311version.
312.Sp
313.Vb 3
314\& assert (("libev version mismatch",
315\& ev_version_major () == EV_VERSION_MAJOR
316\& && ev_version_minor () >= EV_VERSION_MINOR));
317.Ve
204.IP "unsigned int ev_supported_backends ()" 4 318.IP "unsigned int ev_supported_backends ()" 4
205.IX Item "unsigned int ev_supported_backends ()" 319.IX Item "unsigned int ev_supported_backends ()"
206Return the set of all backends (i.e. their corresponding \f(CW\*(C`EV_BACKEND_*\*(C'\fR 320Return the set of all backends (i.e. their corresponding \f(CW\*(C`EV_BACKEND_*\*(C'\fR
207value) compiled into this binary of libev (independent of their 321value) compiled into this binary of libev (independent of their
208availability on the system you are running on). See \f(CW\*(C`ev_default_loop\*(C'\fR for 322availability on the system you are running on). See \f(CW\*(C`ev_default_loop\*(C'\fR for
209a description of the set values. 323a description of the set values.
324.Sp
325Example: make sure we have the epoll method, because yeah this is cool and
326a must have and can we have a torrent of it please!!!11
327.Sp
328.Vb 2
329\& assert (("sorry, no epoll, no sex",
330\& ev_supported_backends () & EVBACKEND_EPOLL));
331.Ve
210.IP "unsigned int ev_recommended_backends ()" 4 332.IP "unsigned int ev_recommended_backends ()" 4
211.IX Item "unsigned int ev_recommended_backends ()" 333.IX Item "unsigned int ev_recommended_backends ()"
212Return the set of all backends compiled into this binary of libev and also 334Return the set of all backends compiled into this binary of libev and also
213recommended for this platform. This set is often smaller than the one 335recommended for this platform. This set is often smaller than the one
214returned by \f(CW\*(C`ev_supported_backends\*(C'\fR, as for example kqueue is broken on 336returned by \f(CW\*(C`ev_supported_backends\*(C'\fR, as for example kqueue is broken on
215most BSDs and will not be autodetected unless you explicitly request it 337most BSDs and will not be auto-detected unless you explicitly request it
216(assuming you know what you are doing). This is the set of backends that 338(assuming you know what you are doing). This is the set of backends that
217libev will probe for if you specify no backends explicitly. 339libev will probe for if you specify no backends explicitly.
340.IP "unsigned int ev_embeddable_backends ()" 4
341.IX Item "unsigned int ev_embeddable_backends ()"
342Returns the set of backends that are embeddable in other event loops. This
343is the theoretical, all-platform, value. To find which backends
344might be supported on the current system, you would need to look at
345\&\f(CW\*(C`ev_embeddable_backends () & ev_supported_backends ()\*(C'\fR, likewise for
346recommended ones.
347.Sp
348See the description of \f(CW\*(C`ev_embed\*(C'\fR watchers for more info.
218.IP "ev_set_allocator (void *(*cb)(void *ptr, long size))" 4 349.IP "ev_set_allocator (void *(*cb)(void *ptr, long size)) [\s-1NOT\s0 \s-1REENTRANT\s0]" 4
219.IX Item "ev_set_allocator (void *(*cb)(void *ptr, long size))" 350.IX Item "ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]"
220Sets the allocation function to use (the prototype is similar to the 351Sets the allocation function to use (the prototype is similar \- the
221realloc C function, the semantics are identical). It is used to allocate 352semantics are identical to the \f(CW\*(C`realloc\*(C'\fR C89/SuS/POSIX function). It is
222and free memory (no surprises here). If it returns zero when memory 353used to allocate and free memory (no surprises here). If it returns zero
223needs to be allocated, the library might abort or take some potentially 354when memory needs to be allocated (\f(CW\*(C`size != 0\*(C'\fR), the library might abort
224destructive action. The default is your system realloc function. 355or take some potentially destructive action.
356.Sp
357Since some systems (at least OpenBSD and Darwin) fail to implement
358correct \f(CW\*(C`realloc\*(C'\fR semantics, libev will use a wrapper around the system
359\&\f(CW\*(C`realloc\*(C'\fR and \f(CW\*(C`free\*(C'\fR functions by default.
225.Sp 360.Sp
226You could override this function in high-availability programs to, say, 361You could override this function in high-availability programs to, say,
227free some memory if it cannot allocate memory, to use a special allocator, 362free some memory if it cannot allocate memory, to use a special allocator,
228or even to sleep a while and retry until some memory is available. 363or even to sleep a while and retry until some memory is available.
364.Sp
365Example: Replace the libev allocator with one that waits a bit and then
366retries (example requires a standards-compliant \f(CW\*(C`realloc\*(C'\fR).
367.Sp
368.Vb 6
369\& static void *
370\& persistent_realloc (void *ptr, size_t size)
371\& {
372\& for (;;)
373\& {
374\& void *newptr = realloc (ptr, size);
375\&
376\& if (newptr)
377\& return newptr;
378\&
379\& sleep (60);
380\& }
381\& }
382\&
383\& ...
384\& ev_set_allocator (persistent_realloc);
385.Ve
229.IP "ev_set_syserr_cb (void (*cb)(const char *msg));" 4 386.IP "ev_set_syserr_cb (void (*cb)(const char *msg)); [\s-1NOT\s0 \s-1REENTRANT\s0]" 4
230.IX Item "ev_set_syserr_cb (void (*cb)(const char *msg));" 387.IX Item "ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]"
231Set the callback function to call on a retryable syscall error (such 388Set the callback function to call on a retryable system call error (such
232as failed select, poll, epoll_wait). The message is a printable string 389as failed select, poll, epoll_wait). The message is a printable string
233indicating the system call or subsystem causing the problem. If this 390indicating the system call or subsystem causing the problem. If this
234callback is set, then libev will expect it to remedy the sitution, no 391callback is set, then libev will expect it to remedy the situation, no
235matter what, when it returns. That is, libev will generally retry the 392matter what, when it returns. That is, libev will generally retry the
236requested operation, or, if the condition doesn't go away, do bad stuff 393requested operation, or, if the condition doesn't go away, do bad stuff
237(such as abort). 394(such as abort).
395.Sp
396Example: This is basically the same thing that libev does internally, too.
397.Sp
398.Vb 6
399\& static void
400\& fatal_error (const char *msg)
401\& {
402\& perror (msg);
403\& abort ();
404\& }
405\&
406\& ...
407\& ev_set_syserr_cb (fatal_error);
408.Ve
238.SH "FUNCTIONS CONTROLLING THE EVENT LOOP" 409.SH "FUNCTIONS CONTROLLING THE EVENT LOOP"
239.IX Header "FUNCTIONS CONTROLLING THE EVENT LOOP" 410.IX Header "FUNCTIONS CONTROLLING THE EVENT LOOP"
240An event loop is described by a \f(CW\*(C`struct ev_loop *\*(C'\fR. The library knows two 411An event loop is described by a \f(CW\*(C`struct ev_loop *\*(C'\fR. The library knows two
241types of such loops, the \fIdefault\fR loop, which supports signals and child 412types of such loops, the \fIdefault\fR loop, which supports signals and child
242events, and dynamically created loops which do not. 413events, and dynamically created loops which do not.
243.PP
244If you use threads, a common model is to run the default event loop
245in your main thread (or in a separate thread) and for each thread you
246create, you also create another event loop. Libev itself does no locking
247whatsoever, so if you mix calls to the same event loop in different
248threads, make sure you lock (this is usually a bad idea, though, even if
249done correctly, because it's hideous and inefficient).
250.IP "struct ev_loop *ev_default_loop (unsigned int flags)" 4 414.IP "struct ev_loop *ev_default_loop (unsigned int flags)" 4
251.IX Item "struct ev_loop *ev_default_loop (unsigned int flags)" 415.IX Item "struct ev_loop *ev_default_loop (unsigned int flags)"
252This will initialise the default event loop if it hasn't been initialised 416This will initialise the default event loop if it hasn't been initialised
253yet and return it. If the default loop could not be initialised, returns 417yet and return it. If the default loop could not be initialised, returns
254false. If it already was initialised it simply returns it (and ignores the 418false. If it already was initialised it simply returns it (and ignores the
255flags. If that is troubling you, check \f(CW\*(C`ev_backend ()\*(C'\fR afterwards). 419flags. If that is troubling you, check \f(CW\*(C`ev_backend ()\*(C'\fR afterwards).
256.Sp 420.Sp
257If you don't know what event loop to use, use the one returned from this 421If you don't know what event loop to use, use the one returned from this
258function. 422function.
423.Sp
424Note that this function is \fInot\fR thread-safe, so if you want to use it
425from multiple threads, you have to lock (note also that this is unlikely,
426as loops cannot bes hared easily between threads anyway).
427.Sp
428The default loop is the only loop that can handle \f(CW\*(C`ev_signal\*(C'\fR and
429\&\f(CW\*(C`ev_child\*(C'\fR watchers, and to do this, it always registers a handler
430for \f(CW\*(C`SIGCHLD\*(C'\fR. If this is a problem for your application you can either
431create a dynamic loop with \f(CW\*(C`ev_loop_new\*(C'\fR that doesn't do that, or you
432can simply overwrite the \f(CW\*(C`SIGCHLD\*(C'\fR signal handler \fIafter\fR calling
433\&\f(CW\*(C`ev_default_init\*(C'\fR.
259.Sp 434.Sp
260The flags argument can be used to specify special behaviour or specific 435The flags argument can be used to specify special behaviour or specific
261backends to use, and is usually specified as \f(CW0\fR (or \f(CW\*(C`EVFLAG_AUTO\*(C'\fR). 436backends to use, and is usually specified as \f(CW0\fR (or \f(CW\*(C`EVFLAG_AUTO\*(C'\fR).
262.Sp 437.Sp
263The following flags are supported: 438The following flags are supported:
… …
268The default flags value. Use this if you have no clue (it's the right 443The default flags value. Use this if you have no clue (it's the right
269thing, believe me). 444thing, believe me).
270.ie n .IP """EVFLAG_NOENV""" 4 445.ie n .IP """EVFLAG_NOENV""" 4
271.el .IP "\f(CWEVFLAG_NOENV\fR" 4 446.el .IP "\f(CWEVFLAG_NOENV\fR" 4
272.IX Item "EVFLAG_NOENV" 447.IX Item "EVFLAG_NOENV"
273If this flag bit is ored into the flag value (or the program runs setuid 448If this flag bit is or'ed into the flag value (or the program runs setuid
274or setgid) then libev will \fInot\fR look at the environment variable 449or setgid) then libev will \fInot\fR look at the environment variable
275\&\f(CW\*(C`LIBEV_FLAGS\*(C'\fR. Otherwise (the default), this environment variable will 450\&\f(CW\*(C`LIBEV_FLAGS\*(C'\fR. Otherwise (the default), this environment variable will
276override the flags completely if it is found in the environment. This is 451override the flags completely if it is found in the environment. This is
277useful to try out specific backends to test their performance, or to work 452useful to try out specific backends to test their performance, or to work
278around bugs. 453around bugs.
454.ie n .IP """EVFLAG_FORKCHECK""" 4
455.el .IP "\f(CWEVFLAG_FORKCHECK\fR" 4
456.IX Item "EVFLAG_FORKCHECK"
457Instead of calling \f(CW\*(C`ev_default_fork\*(C'\fR or \f(CW\*(C`ev_loop_fork\*(C'\fR manually after
458a fork, you can also make libev check for a fork in each iteration by
459enabling this flag.
460.Sp
461This works by calling \f(CW\*(C`getpid ()\*(C'\fR on every iteration of the loop,
462and thus this might slow down your event loop if you do a lot of loop
463iterations and little real work, but is usually not noticeable (on my
464GNU/Linux system for example, \f(CW\*(C`getpid\*(C'\fR is actually a simple 5\-insn sequence
465without a system call and thus \fIvery\fR fast, but my GNU/Linux system also has
466\&\f(CW\*(C`pthread_atfork\*(C'\fR which is even faster).
467.Sp
468The big advantage of this flag is that you can forget about fork (and
469forget about forgetting to tell libev about forking) when you use this
470flag.
471.Sp
472This flag setting cannot be overridden or specified in the \f(CW\*(C`LIBEV_FLAGS\*(C'\fR
473environment variable.
279.ie n .IP """EVBACKEND_SELECT"" (value 1, portable select backend)" 4 474.ie n .IP """EVBACKEND_SELECT"" (value 1, portable select backend)" 4
280.el .IP "\f(CWEVBACKEND_SELECT\fR (value 1, portable select backend)" 4 475.el .IP "\f(CWEVBACKEND_SELECT\fR (value 1, portable select backend)" 4
281.IX Item "EVBACKEND_SELECT (value 1, portable select backend)" 476.IX Item "EVBACKEND_SELECT (value 1, portable select backend)"
282This is your standard \fIselect\fR\|(2) backend. Not \fIcompletely\fR standard, as 477This is your standard \fIselect\fR\|(2) backend. Not \fIcompletely\fR standard, as
283libev tries to roll its own fd_set with no limits on the number of fds, 478libev tries to roll its own fd_set with no limits on the number of fds,
284but if that fails, expect a fairly low limit on the number of fds when 479but if that fails, expect a fairly low limit on the number of fds when
285using this backend. It doesn't scale too well (O(highest_fd)), but its usually 480using this backend. It doesn't scale too well (O(highest_fd)), but its
286the fastest backend for a low number of fds. 481usually the fastest backend for a low number of (low-numbered :) fds.
482.Sp
483To get good performance out of this backend you need a high amount of
484parallelism (most of the file descriptors should be busy). If you are
485writing a server, you should \f(CW\*(C`accept ()\*(C'\fR in a loop to accept as many
486connections as possible during one iteration. You might also want to have
487a look at \f(CW\*(C`ev_set_io_collect_interval ()\*(C'\fR to increase the amount of
488readiness notifications you get per iteration.
489.Sp
490This backend maps \f(CW\*(C`EV_READ\*(C'\fR to the \f(CW\*(C`readfds\*(C'\fR set and \f(CW\*(C`EV_WRITE\*(C'\fR to the
491\&\f(CW\*(C`writefds\*(C'\fR set (and to work around Microsoft Windows bugs, also onto the
492\&\f(CW\*(C`exceptfds\*(C'\fR set on that platform).
287.ie n .IP """EVBACKEND_POLL"" (value 2, poll backend, available everywhere except on windows)" 4 493.ie n .IP """EVBACKEND_POLL"" (value 2, poll backend, available everywhere except on windows)" 4
288.el .IP "\f(CWEVBACKEND_POLL\fR (value 2, poll backend, available everywhere except on windows)" 4 494.el .IP "\f(CWEVBACKEND_POLL\fR (value 2, poll backend, available everywhere except on windows)" 4
289.IX Item "EVBACKEND_POLL (value 2, poll backend, available everywhere except on windows)" 495.IX Item "EVBACKEND_POLL (value 2, poll backend, available everywhere except on windows)"
290And this is your standard \fIpoll\fR\|(2) backend. It's more complicated than 496And this is your standard \fIpoll\fR\|(2) backend. It's more complicated
291select, but handles sparse fds better and has no artificial limit on the 497than select, but handles sparse fds better and has no artificial
292number of fds you can use (except it will slow down considerably with a 498limit on the number of fds you can use (except it will slow down
293lot of inactive fds). It scales similarly to select, i.e. O(total_fds). 499considerably with a lot of inactive fds). It scales similarly to select,
500i.e. O(total_fds). See the entry for \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR, above, for
501performance tips.
502.Sp
503This backend maps \f(CW\*(C`EV_READ\*(C'\fR to \f(CW\*(C`POLLIN | POLLERR | POLLHUP\*(C'\fR, and
504\&\f(CW\*(C`EV_WRITE\*(C'\fR to \f(CW\*(C`POLLOUT | POLLERR | POLLHUP\*(C'\fR.
294.ie n .IP """EVBACKEND_EPOLL"" (value 4, Linux)" 4 505.ie n .IP """EVBACKEND_EPOLL"" (value 4, Linux)" 4
295.el .IP "\f(CWEVBACKEND_EPOLL\fR (value 4, Linux)" 4 506.el .IP "\f(CWEVBACKEND_EPOLL\fR (value 4, Linux)" 4
296.IX Item "EVBACKEND_EPOLL (value 4, Linux)" 507.IX Item "EVBACKEND_EPOLL (value 4, Linux)"
297For few fds, this backend is a bit little slower than poll and select, 508For few fds, this backend is a bit little slower than poll and select,
298but it scales phenomenally better. While poll and select usually scale like 509but it scales phenomenally better. While poll and select usually scale
299O(total_fds) where n is the total number of fds (or the highest fd), epoll scales 510like O(total_fds) where n is the total number of fds (or the highest fd),
300either O(1) or O(active_fds). 511epoll scales either O(1) or O(active_fds). The epoll design has a number
512of shortcomings, such as silently dropping events in some hard-to-detect
513cases and requiring a system call per fd change, no fork support and bad
514support for dup.
301.Sp 515.Sp
302While stopping and starting an I/O watcher in the same iteration will 516While stopping, setting and starting an I/O watcher in the same iteration
303result in some caching, there is still a syscall per such incident 517will result in some caching, there is still a system call per such incident
304(because the fd could point to a different file description now), so its 518(because the fd could point to a different file description now), so its
305best to avoid that. Also, \fIdup()\fRed file descriptors might not work very 519best to avoid that. Also, \f(CW\*(C`dup ()\*(C'\fR'ed file descriptors might not work
306well if you register events for both fds. 520very well if you register events for both fds.
307.Sp 521.Sp
308Please note that epoll sometimes generates spurious notifications, so you 522Please note that epoll sometimes generates spurious notifications, so you
309need to use non-blocking I/O or other means to avoid blocking when no data 523need to use non-blocking I/O or other means to avoid blocking when no data
310(or space) is available. 524(or space) is available.
525.Sp
526Best performance from this backend is achieved by not unregistering all
527watchers for a file descriptor until it has been closed, if possible,
528i.e. keep at least one watcher active per fd at all times. Stopping and
529starting a watcher (without re-setting it) also usually doesn't cause
530extra overhead.
531.Sp
532While nominally embeddable in other event loops, this feature is broken in
533all kernel versions tested so far.
534.Sp
535This backend maps \f(CW\*(C`EV_READ\*(C'\fR and \f(CW\*(C`EV_WRITE\*(C'\fR in the same way as
536\&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR.
311.ie n .IP """EVBACKEND_KQUEUE"" (value 8, most \s-1BSD\s0 clones)" 4 537.ie n .IP """EVBACKEND_KQUEUE"" (value 8, most \s-1BSD\s0 clones)" 4
312.el .IP "\f(CWEVBACKEND_KQUEUE\fR (value 8, most \s-1BSD\s0 clones)" 4 538.el .IP "\f(CWEVBACKEND_KQUEUE\fR (value 8, most \s-1BSD\s0 clones)" 4
313.IX Item "EVBACKEND_KQUEUE (value 8, most BSD clones)" 539.IX Item "EVBACKEND_KQUEUE (value 8, most BSD clones)"
314Kqueue deserves special mention, as at the time of this writing, it 540Kqueue deserves special mention, as at the time of this writing, it was
315was broken on all BSDs except NetBSD (usually it doesn't work with 541broken on all BSDs except NetBSD (usually it doesn't work reliably with
316anything but sockets and pipes, except on Darwin, where of course its 542anything but sockets and pipes, except on Darwin, where of course it's
317completely useless). For this reason its not being \*(L"autodetected\*(R" 543completely useless). For this reason it's not being \*(L"auto-detected\*(R" unless
318unless you explicitly specify it explicitly in the flags (i.e. using 544you explicitly specify it in the flags (i.e. using \f(CW\*(C`EVBACKEND_KQUEUE\*(C'\fR) or
319\&\f(CW\*(C`EVBACKEND_KQUEUE\*(C'\fR). 545libev was compiled on a known-to-be-good (\-enough) system like NetBSD.
546.Sp
547You still can embed kqueue into a normal poll or select backend and use it
548only for sockets (after having made sure that sockets work with kqueue on
549the target platform). See \f(CW\*(C`ev_embed\*(C'\fR watchers for more info.
320.Sp 550.Sp
321It scales in the same way as the epoll backend, but the interface to the 551It scales in the same way as the epoll backend, but the interface to the
322kernel is more efficient (which says nothing about its actual speed, of 552kernel is more efficient (which says nothing about its actual speed, of
323course). While starting and stopping an I/O watcher does not cause an 553course). While stopping, setting and starting an I/O watcher does never
324extra syscall as with epoll, it still adds up to four event changes per 554cause an extra system call as with \f(CW\*(C`EVBACKEND_EPOLL\*(C'\fR, it still adds up to
325incident, so its best to avoid that. 555two event changes per incident. Support for \f(CW\*(C`fork ()\*(C'\fR is very bad and it
556drops fds silently in similarly hard-to-detect cases.
557.Sp
558This backend usually performs well under most conditions.
559.Sp
560While nominally embeddable in other event loops, this doesn't work
561everywhere, so you might need to test for this. And since it is broken
562almost everywhere, you should only use it when you have a lot of sockets
563(for which it usually works), by embedding it into another event loop
564(e.g. \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR or \f(CW\*(C`EVBACKEND_POLL\*(C'\fR) and, did I mention it,
565using it only for sockets.
566.Sp
567This backend maps \f(CW\*(C`EV_READ\*(C'\fR into an \f(CW\*(C`EVFILT_READ\*(C'\fR kevent with
568\&\f(CW\*(C`NOTE_EOF\*(C'\fR, and \f(CW\*(C`EV_WRITE\*(C'\fR into an \f(CW\*(C`EVFILT_WRITE\*(C'\fR kevent with
569\&\f(CW\*(C`NOTE_EOF\*(C'\fR.
326.ie n .IP """EVBACKEND_DEVPOLL"" (value 16, Solaris 8)" 4 570.ie n .IP """EVBACKEND_DEVPOLL"" (value 16, Solaris 8)" 4
327.el .IP "\f(CWEVBACKEND_DEVPOLL\fR (value 16, Solaris 8)" 4 571.el .IP "\f(CWEVBACKEND_DEVPOLL\fR (value 16, Solaris 8)" 4
328.IX Item "EVBACKEND_DEVPOLL (value 16, Solaris 8)" 572.IX Item "EVBACKEND_DEVPOLL (value 16, Solaris 8)"
329This is not implemented yet (and might never be). 573This is not implemented yet (and might never be, unless you send me an
574implementation). According to reports, \f(CW\*(C`/dev/poll\*(C'\fR only supports sockets
575and is not embeddable, which would limit the usefulness of this backend
576immensely.
330.ie n .IP """EVBACKEND_PORT"" (value 32, Solaris 10)" 4 577.ie n .IP """EVBACKEND_PORT"" (value 32, Solaris 10)" 4
331.el .IP "\f(CWEVBACKEND_PORT\fR (value 32, Solaris 10)" 4 578.el .IP "\f(CWEVBACKEND_PORT\fR (value 32, Solaris 10)" 4
332.IX Item "EVBACKEND_PORT (value 32, Solaris 10)" 579.IX Item "EVBACKEND_PORT (value 32, Solaris 10)"
333This uses the Solaris 10 port mechanism. As with everything on Solaris, 580This uses the Solaris 10 event port mechanism. As with everything on Solaris,
334it's really slow, but it still scales very well (O(active_fds)). 581it's really slow, but it still scales very well (O(active_fds)).
335.Sp 582.Sp
336Please note that solaris ports can result in a lot of spurious 583Please note that Solaris event ports can deliver a lot of spurious
337notifications, so you need to use non-blocking I/O or other means to avoid 584notifications, so you need to use non-blocking I/O or other means to avoid
338blocking when no data (or space) is available. 585blocking when no data (or space) is available.
586.Sp
587While this backend scales well, it requires one system call per active
588file descriptor per loop iteration. For small and medium numbers of file
589descriptors a \*(L"slow\*(R" \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR or \f(CW\*(C`EVBACKEND_POLL\*(C'\fR backend
590might perform better.
591.Sp
592On the positive side, with the exception of the spurious readiness
593notifications, this backend actually performed fully to specification
594in all tests and is fully embeddable, which is a rare feat among the
595OS-specific backends.
596.Sp
597This backend maps \f(CW\*(C`EV_READ\*(C'\fR and \f(CW\*(C`EV_WRITE\*(C'\fR in the same way as
598\&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR.
339.ie n .IP """EVBACKEND_ALL""" 4 599.ie n .IP """EVBACKEND_ALL""" 4
340.el .IP "\f(CWEVBACKEND_ALL\fR" 4 600.el .IP "\f(CWEVBACKEND_ALL\fR" 4
341.IX Item "EVBACKEND_ALL" 601.IX Item "EVBACKEND_ALL"
342Try all backends (even potentially broken ones that wouldn't be tried 602Try all backends (even potentially broken ones that wouldn't be tried
343with \f(CW\*(C`EVFLAG_AUTO\*(C'\fR). Since this is a mask, you can do stuff such as 603with \f(CW\*(C`EVFLAG_AUTO\*(C'\fR). Since this is a mask, you can do stuff such as
344\&\f(CW\*(C`EVBACKEND_ALL & ~EVBACKEND_KQUEUE\*(C'\fR. 604\&\f(CW\*(C`EVBACKEND_ALL & ~EVBACKEND_KQUEUE\*(C'\fR.
605.Sp
606It is definitely not recommended to use this flag.
345.RE 607.RE
346.RS 4 608.RS 4
347.Sp 609.Sp
348If one or more of these are ored into the flags value, then only these 610If one or more of these are or'ed into the flags value, then only these
349backends will be tried (in the reverse order as given here). If none are 611backends will be tried (in the reverse order as listed here). If none are
350specified, most compiled-in backend will be tried, usually in reverse 612specified, all backends in \f(CW\*(C`ev_recommended_backends ()\*(C'\fR will be tried.
351order of their flag values :)
352.Sp 613.Sp
353The most typical usage is like this: 614Example: This is the most typical usage.
354.Sp 615.Sp
355.Vb 2 616.Vb 2
356\& if (!ev_default_loop (0)) 617\& if (!ev_default_loop (0))
357\& fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?"); 618\& fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
358.Ve 619.Ve
359.Sp 620.Sp
360Restrict libev to the select and poll backends, and do not allow 621Example: Restrict libev to the select and poll backends, and do not allow
361environment settings to be taken into account: 622environment settings to be taken into account:
362.Sp 623.Sp
363.Vb 1 624.Vb 1
364\& ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV); 625\& ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
365.Ve 626.Ve
366.Sp 627.Sp
367Use whatever libev has to offer, but make sure that kqueue is used if 628Example: Use whatever libev has to offer, but make sure that kqueue is
368available (warning, breaks stuff, best use only with your own private 629used if available (warning, breaks stuff, best use only with your own
369event loop and only if you know the \s-1OS\s0 supports your types of fds): 630private event loop and only if you know the \s-1OS\s0 supports your types of
631fds):
370.Sp 632.Sp
371.Vb 1 633.Vb 1
372\& ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE); 634\& ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
373.Ve 635.Ve
374.RE 636.RE
375.IP "struct ev_loop *ev_loop_new (unsigned int flags)" 4 637.IP "struct ev_loop *ev_loop_new (unsigned int flags)" 4
376.IX Item "struct ev_loop *ev_loop_new (unsigned int flags)" 638.IX Item "struct ev_loop *ev_loop_new (unsigned int flags)"
377Similar to \f(CW\*(C`ev_default_loop\*(C'\fR, but always creates a new event loop that is 639Similar to \f(CW\*(C`ev_default_loop\*(C'\fR, but always creates a new event loop that is
378always distinct from the default loop. Unlike the default loop, it cannot 640always distinct from the default loop. Unlike the default loop, it cannot
379handle signal and child watchers, and attempts to do so will be greeted by 641handle signal and child watchers, and attempts to do so will be greeted by
380undefined behaviour (or a failed assertion if assertions are enabled). 642undefined behaviour (or a failed assertion if assertions are enabled).
643.Sp
644Note that this function \fIis\fR thread-safe, and the recommended way to use
645libev with threads is indeed to create one loop per thread, and using the
646default loop in the \*(L"main\*(R" or \*(L"initial\*(R" thread.
647.Sp
648Example: Try to create a event loop that uses epoll and nothing else.
649.Sp
650.Vb 3
651\& struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
652\& if (!epoller)
653\& fatal ("no epoll found here, maybe it hides under your chair");
654.Ve
381.IP "ev_default_destroy ()" 4 655.IP "ev_default_destroy ()" 4
382.IX Item "ev_default_destroy ()" 656.IX Item "ev_default_destroy ()"
383Destroys the default loop again (frees all memory and kernel state 657Destroys the default loop again (frees all memory and kernel state
384etc.). This stops all registered event watchers (by not touching them in 658etc.). None of the active event watchers will be stopped in the normal
385any way whatsoever, although you cannot rely on this :). 659sense, so e.g. \f(CW\*(C`ev_is_active\*(C'\fR might still return true. It is your
660responsibility to either stop all watchers cleanly yourself \fIbefore\fR
661calling this function, or cope with the fact afterwards (which is usually
662the easiest thing, you can just ignore the watchers and/or \f(CW\*(C`free ()\*(C'\fR them
663for example).
664.Sp
665Note that certain global state, such as signal state, will not be freed by
666this function, and related watchers (such as signal and child watchers)
667would need to be stopped manually.
668.Sp
669In general it is not advisable to call this function except in the
670rare occasion where you really need to free e.g. the signal handling
671pipe fds. If you need dynamically allocated loops it is better to use
672\&\f(CW\*(C`ev_loop_new\*(C'\fR and \f(CW\*(C`ev_loop_destroy\*(C'\fR).
386.IP "ev_loop_destroy (loop)" 4 673.IP "ev_loop_destroy (loop)" 4
387.IX Item "ev_loop_destroy (loop)" 674.IX Item "ev_loop_destroy (loop)"
388Like \f(CW\*(C`ev_default_destroy\*(C'\fR, but destroys an event loop created by an 675Like \f(CW\*(C`ev_default_destroy\*(C'\fR, but destroys an event loop created by an
389earlier call to \f(CW\*(C`ev_loop_new\*(C'\fR. 676earlier call to \f(CW\*(C`ev_loop_new\*(C'\fR.
390.IP "ev_default_fork ()" 4 677.IP "ev_default_fork ()" 4
391.IX Item "ev_default_fork ()" 678.IX Item "ev_default_fork ()"
679This function sets a flag that causes subsequent \f(CW\*(C`ev_loop\*(C'\fR iterations
392This function reinitialises the kernel state for backends that have 680to reinitialise the kernel state for backends that have one. Despite the
393one. Despite the name, you can call it anytime, but it makes most sense 681name, you can call it anytime, but it makes most sense after forking, in
394after forking, in either the parent or child process (or both, but that 682the child process (or both child and parent, but that again makes little
395again makes little sense). 683sense). You \fImust\fR call it in the child before using any of the libev
684functions, and it will only take effect at the next \f(CW\*(C`ev_loop\*(C'\fR iteration.
396.Sp 685.Sp
397You \fImust\fR call this function in the child process after forking if and 686On the other hand, you only need to call this function in the child
398only if you want to use the event library in both processes. If you just 687process if and only if you want to use the event library in the child. If
399fork+exec, you don't have to call it. 688you just fork+exec, you don't have to call it at all.
400.Sp 689.Sp
401The function itself is quite fast and it's usually not a problem to call 690The function itself is quite fast and it's usually not a problem to call
402it just in case after a fork. To make this easy, the function will fit in 691it just in case after a fork. To make this easy, the function will fit in
403quite nicely into a call to \f(CW\*(C`pthread_atfork\*(C'\fR: 692quite nicely into a call to \f(CW\*(C`pthread_atfork\*(C'\fR:
404.Sp 693.Sp
405.Vb 1 694.Vb 1
406\& pthread_atfork (0, 0, ev_default_fork); 695\& pthread_atfork (0, 0, ev_default_fork);
407.Ve 696.Ve
408.Sp
409At the moment, \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR and \f(CW\*(C`EVBACKEND_POLL\*(C'\fR are safe to use
410without calling this function, so if you force one of those backends you
411do not need to care.
412.IP "ev_loop_fork (loop)" 4 697.IP "ev_loop_fork (loop)" 4
413.IX Item "ev_loop_fork (loop)" 698.IX Item "ev_loop_fork (loop)"
414Like \f(CW\*(C`ev_default_fork\*(C'\fR, but acts on an event loop created by 699Like \f(CW\*(C`ev_default_fork\*(C'\fR, but acts on an event loop created by
415\&\f(CW\*(C`ev_loop_new\*(C'\fR. Yes, you have to call this on every allocated event loop 700\&\f(CW\*(C`ev_loop_new\*(C'\fR. Yes, you have to call this on every allocated event loop
416after fork, and how you do this is entirely your own problem. 701after fork that you want to re-use in the child, and how you do this is
702entirely your own problem.
703.IP "int ev_is_default_loop (loop)" 4
704.IX Item "int ev_is_default_loop (loop)"
705Returns true when the given loop is, in fact, the default loop, and false
706otherwise.
707.IP "unsigned int ev_loop_count (loop)" 4
708.IX Item "unsigned int ev_loop_count (loop)"
709Returns the count of loop iterations for the loop, which is identical to
710the number of times libev did poll for new events. It starts at \f(CW0\fR and
711happily wraps around with enough iterations.
712.Sp
713This value can sometimes be useful as a generation counter of sorts (it
714\&\*(L"ticks\*(R" the number of loop iterations), as it roughly corresponds with
715\&\f(CW\*(C`ev_prepare\*(C'\fR and \f(CW\*(C`ev_check\*(C'\fR calls.
417.IP "unsigned int ev_backend (loop)" 4 716.IP "unsigned int ev_backend (loop)" 4
418.IX Item "unsigned int ev_backend (loop)" 717.IX Item "unsigned int ev_backend (loop)"
419Returns one of the \f(CW\*(C`EVBACKEND_*\*(C'\fR flags indicating the event backend in 718Returns one of the \f(CW\*(C`EVBACKEND_*\*(C'\fR flags indicating the event backend in
420use. 719use.
421.IP "ev_tstamp ev_now (loop)" 4 720.IP "ev_tstamp ev_now (loop)" 4
422.IX Item "ev_tstamp ev_now (loop)" 721.IX Item "ev_tstamp ev_now (loop)"
423Returns the current \*(L"event loop time\*(R", which is the time the event loop 722Returns the current \*(L"event loop time\*(R", which is the time the event loop
424got events and started processing them. This timestamp does not change 723received events and started processing them. This timestamp does not
425as long as callbacks are being processed, and this is also the base time 724change as long as callbacks are being processed, and this is also the base
426used for relative timers. You can treat it as the timestamp of the event 725time used for relative timers. You can treat it as the timestamp of the
427occuring (or more correctly, the mainloop finding out about it). 726event occurring (or more correctly, libev finding out about it).
727.IP "ev_now_update (loop)" 4
728.IX Item "ev_now_update (loop)"
729Establishes the current time by querying the kernel, updating the time
730returned by \f(CW\*(C`ev_now ()\*(C'\fR in the progress. This is a costly operation and
731is usually done automatically within \f(CW\*(C`ev_loop ()\*(C'\fR.
732.Sp
733This function is rarely useful, but when some event callback runs for a
734very long time without entering the event loop, updating libev's idea of
735the current time is a good idea.
736.Sp
737See also \*(L"The special problem of time updates\*(R" in the \f(CW\*(C`ev_timer\*(C'\fR section.
428.IP "ev_loop (loop, int flags)" 4 738.IP "ev_loop (loop, int flags)" 4
429.IX Item "ev_loop (loop, int flags)" 739.IX Item "ev_loop (loop, int flags)"
430Finally, this is it, the event handler. This function usually is called 740Finally, this is it, the event handler. This function usually is called
431after you initialised all your watchers and you want to start handling 741after you initialised all your watchers and you want to start handling
432events. 742events.
433.Sp 743.Sp
434If the flags argument is specified as \f(CW0\fR, it will not return until 744If the flags argument is specified as \f(CW0\fR, it will not return until
435either no event watchers are active anymore or \f(CW\*(C`ev_unloop\*(C'\fR was called. 745either no event watchers are active anymore or \f(CW\*(C`ev_unloop\*(C'\fR was called.
436.Sp 746.Sp
747Please note that an explicit \f(CW\*(C`ev_unloop\*(C'\fR is usually better than
748relying on all watchers to be stopped when deciding when a program has
749finished (especially in interactive programs), but having a program
750that automatically loops as long as it has to and no longer by virtue
751of relying on its watchers stopping correctly, that is truly a thing of
752beauty.
753.Sp
437A flags value of \f(CW\*(C`EVLOOP_NONBLOCK\*(C'\fR will look for new events, will handle 754A flags value of \f(CW\*(C`EVLOOP_NONBLOCK\*(C'\fR will look for new events, will handle
438those events and any outstanding ones, but will not block your process in 755those events and any already outstanding ones, but will not block your
439case there are no events and will return after one iteration of the loop. 756process in case there are no events and will return after one iteration of
757the loop.
440.Sp 758.Sp
441A flags value of \f(CW\*(C`EVLOOP_ONESHOT\*(C'\fR will look for new events (waiting if 759A flags value of \f(CW\*(C`EVLOOP_ONESHOT\*(C'\fR will look for new events (waiting if
442neccessary) and will handle those and any outstanding ones. It will block 760necessary) and will handle those and any already outstanding ones. It
443your process until at least one new event arrives, and will return after 761will block your process until at least one new event arrives (which could
444one iteration of the loop. This is useful if you are waiting for some 762be an event internal to libev itself, so there is no guarentee that a
445external event in conjunction with something not expressible using other 763user-registered callback will be called), and will return after one
764iteration of the loop.
765.Sp
766This is useful if you are waiting for some external event in conjunction
767with something not expressible using other libev watchers (i.e. "roll your
446libev watchers. However, a pair of \f(CW\*(C`ev_prepare\*(C'\fR/\f(CW\*(C`ev_check\*(C'\fR watchers is 768own \f(CW\*(C`ev_loop\*(C'\fR"). However, a pair of \f(CW\*(C`ev_prepare\*(C'\fR/\f(CW\*(C`ev_check\*(C'\fR watchers is
447usually a better approach for this kind of thing. 769usually a better approach for this kind of thing.
448.Sp 770.Sp
449Here are the gory details of what \f(CW\*(C`ev_loop\*(C'\fR does: 771Here are the gory details of what \f(CW\*(C`ev_loop\*(C'\fR does:
450.Sp 772.Sp
451.Vb 18 773.Vb 10
452\& * If there are no active watchers (reference count is zero), return. 774\& \- Before the first iteration, call any pending watchers.
453\& - Queue prepare watchers and then call all outstanding watchers. 775\& * If EVFLAG_FORKCHECK was used, check for a fork.
776\& \- If a fork was detected (by any means), queue and call all fork watchers.
777\& \- Queue and call all prepare watchers.
454\& - If we have been forked, recreate the kernel state. 778\& \- If we have been forked, detach and recreate the kernel state
779\& as to not disturb the other process.
455\& - Update the kernel state with all outstanding changes. 780\& \- Update the kernel state with all outstanding changes.
456\& - Update the "event loop time". 781\& \- Update the "event loop time" (ev_now ()).
457\& - Calculate for how long to block. 782\& \- Calculate for how long to sleep or block, if at all
783\& (active idle watchers, EVLOOP_NONBLOCK or not having
784\& any active watchers at all will result in not sleeping).
785\& \- Sleep if the I/O and timer collect interval say so.
458\& - Block the process, waiting for any events. 786\& \- Block the process, waiting for any events.
459\& - Queue all outstanding I/O (fd) events. 787\& \- Queue all outstanding I/O (fd) events.
460\& - Update the "event loop time" and do time jump handling. 788\& \- Update the "event loop time" (ev_now ()), and do time jump adjustments.
461\& - Queue all outstanding timers. 789\& \- Queue all expired timers.
462\& - Queue all outstanding periodics. 790\& \- Queue all expired periodics.
463\& - If no events are pending now, queue all idle watchers. 791\& \- Unless any events are pending now, queue all idle watchers.
464\& - Queue all check watchers. 792\& \- Queue all check watchers.
465\& - Call all queued watchers in reverse order (i.e. check watchers first). 793\& \- Call all queued watchers in reverse order (i.e. check watchers first).
466\& Signals and child watchers are implemented as I/O watchers, and will 794\& Signals and child watchers are implemented as I/O watchers, and will
467\& be handled here by queueing them when their watcher gets executed. 795\& be handled here by queueing them when their watcher gets executed.
468\& - If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK 796\& \- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
469\& were used, return, otherwise continue with step *. 797\& were used, or there are no active watchers, return, otherwise
798\& continue with step *.
799.Ve
800.Sp
801Example: Queue some jobs and then loop until no events are outstanding
802anymore.
803.Sp
804.Vb 4
805\& ... queue jobs here, make sure they register event watchers as long
806\& ... as they still have work to do (even an idle watcher will do..)
807\& ev_loop (my_loop, 0);
808\& ... jobs done or somebody called unloop. yeah!
470.Ve 809.Ve
471.IP "ev_unloop (loop, how)" 4 810.IP "ev_unloop (loop, how)" 4
472.IX Item "ev_unloop (loop, how)" 811.IX Item "ev_unloop (loop, how)"
473Can be used to make a call to \f(CW\*(C`ev_loop\*(C'\fR return early (but only after it 812Can be used to make a call to \f(CW\*(C`ev_loop\*(C'\fR return early (but only after it
474has processed all outstanding events). The \f(CW\*(C`how\*(C'\fR argument must be either 813has processed all outstanding events). The \f(CW\*(C`how\*(C'\fR argument must be either
475\&\f(CW\*(C`EVUNLOOP_ONE\*(C'\fR, which will make the innermost \f(CW\*(C`ev_loop\*(C'\fR call return, or 814\&\f(CW\*(C`EVUNLOOP_ONE\*(C'\fR, which will make the innermost \f(CW\*(C`ev_loop\*(C'\fR call return, or
476\&\f(CW\*(C`EVUNLOOP_ALL\*(C'\fR, which will make all nested \f(CW\*(C`ev_loop\*(C'\fR calls return. 815\&\f(CW\*(C`EVUNLOOP_ALL\*(C'\fR, which will make all nested \f(CW\*(C`ev_loop\*(C'\fR calls return.
816.Sp
817This \*(L"unloop state\*(R" will be cleared when entering \f(CW\*(C`ev_loop\*(C'\fR again.
818.Sp
819It is safe to call \f(CW\*(C`ev_unloop\*(C'\fR from otuside any \f(CW\*(C`ev_loop\*(C'\fR calls.
477.IP "ev_ref (loop)" 4 820.IP "ev_ref (loop)" 4
478.IX Item "ev_ref (loop)" 821.IX Item "ev_ref (loop)"
479.PD 0 822.PD 0
480.IP "ev_unref (loop)" 4 823.IP "ev_unref (loop)" 4
481.IX Item "ev_unref (loop)" 824.IX Item "ev_unref (loop)"
482.PD 825.PD
483Ref/unref can be used to add or remove a reference count on the event 826Ref/unref can be used to add or remove a reference count on the event
484loop: Every watcher keeps one reference, and as long as the reference 827loop: Every watcher keeps one reference, and as long as the reference
485count is nonzero, \f(CW\*(C`ev_loop\*(C'\fR will not return on its own. If you have 828count is nonzero, \f(CW\*(C`ev_loop\*(C'\fR will not return on its own.
829.Sp
486a watcher you never unregister that should not keep \f(CW\*(C`ev_loop\*(C'\fR from 830If you have a watcher you never unregister that should not keep \f(CW\*(C`ev_loop\*(C'\fR
487returning, \fIev_unref()\fR after starting, and \fIev_ref()\fR before stopping it. For 831from returning, call \fIev_unref()\fR after starting, and \fIev_ref()\fR before
832stopping it.
833.Sp
488example, libev itself uses this for its internal signal pipe: It is not 834As an example, libev itself uses this for its internal signal pipe: It is
489visible to the libev user and should not keep \f(CW\*(C`ev_loop\*(C'\fR from exiting if 835not visible to the libev user and should not keep \f(CW\*(C`ev_loop\*(C'\fR from exiting
490no event watchers registered by it are active. It is also an excellent 836if no event watchers registered by it are active. It is also an excellent
491way to do this for generic recurring timers or from within third-party 837way to do this for generic recurring timers or from within third-party
492libraries. Just remember to \fIunref after start\fR and \fIref before stop\fR. 838libraries. Just remember to \fIunref after start\fR and \fIref before stop\fR
839(but only if the watcher wasn't active before, or was active before,
840respectively).
841.Sp
842Example: Create a signal watcher, but keep it from keeping \f(CW\*(C`ev_loop\*(C'\fR
843running when nothing else is active.
844.Sp
845.Vb 4
846\& struct ev_signal exitsig;
847\& ev_signal_init (&exitsig, sig_cb, SIGINT);
848\& ev_signal_start (loop, &exitsig);
849\& evf_unref (loop);
850.Ve
851.Sp
852Example: For some weird reason, unregister the above signal handler again.
853.Sp
854.Vb 2
855\& ev_ref (loop);
856\& ev_signal_stop (loop, &exitsig);
857.Ve
858.IP "ev_set_io_collect_interval (loop, ev_tstamp interval)" 4
859.IX Item "ev_set_io_collect_interval (loop, ev_tstamp interval)"
860.PD 0
861.IP "ev_set_timeout_collect_interval (loop, ev_tstamp interval)" 4
862.IX Item "ev_set_timeout_collect_interval (loop, ev_tstamp interval)"
863.PD
864These advanced functions influence the time that libev will spend waiting
865for events. Both time intervals are by default \f(CW0\fR, meaning that libev
866will try to invoke timer/periodic callbacks and I/O callbacks with minimum
867latency.
868.Sp
869Setting these to a higher value (the \f(CW\*(C`interval\*(C'\fR \fImust\fR be >= \f(CW0\fR)
870allows libev to delay invocation of I/O and timer/periodic callbacks
871to increase efficiency of loop iterations (or to increase power-saving
872opportunities).
873.Sp
874The idea is that sometimes your program runs just fast enough to handle
875one (or very few) event(s) per loop iteration. While this makes the
876program responsive, it also wastes a lot of \s-1CPU\s0 time to poll for new
877events, especially with backends like \f(CW\*(C`select ()\*(C'\fR which have a high
878overhead for the actual polling but can deliver many events at once.
879.Sp
880By setting a higher \fIio collect interval\fR you allow libev to spend more
881time collecting I/O events, so you can handle more events per iteration,
882at the cost of increasing latency. Timeouts (both \f(CW\*(C`ev_periodic\*(C'\fR and
883\&\f(CW\*(C`ev_timer\*(C'\fR) will be not affected. Setting this to a non-null value will
884introduce an additional \f(CW\*(C`ev_sleep ()\*(C'\fR call into most loop iterations.
885.Sp
886Likewise, by setting a higher \fItimeout collect interval\fR you allow libev
887to spend more time collecting timeouts, at the expense of increased
888latency/jitter/inexactness (the watcher callback will be called
889later). \f(CW\*(C`ev_io\*(C'\fR watchers will not be affected. Setting this to a non-null
890value will not introduce any overhead in libev.
891.Sp
892Many (busy) programs can usually benefit by setting the I/O collect
893interval to a value near \f(CW0.1\fR or so, which is often enough for
894interactive servers (of course not for games), likewise for timeouts. It
895usually doesn't make much sense to set it to a lower value than \f(CW0.01\fR,
896as this approaches the timing granularity of most systems.
897.Sp
898Setting the \fItimeout collect interval\fR can improve the opportunity for
899saving power, as the program will \*(L"bundle\*(R" timer callback invocations that
900are \*(L"near\*(R" in time together, by delaying some, thus reducing the number of
901times the process sleeps and wakes up again. Another useful technique to
902reduce iterations/wake\-ups is to use \f(CW\*(C`ev_periodic\*(C'\fR watchers and make sure
903they fire on, say, one-second boundaries only.
904.IP "ev_loop_verify (loop)" 4
905.IX Item "ev_loop_verify (loop)"
906This function only does something when \f(CW\*(C`EV_VERIFY\*(C'\fR support has been
907compiled in. which is the default for non-minimal builds. It tries to go
908through all internal structures and checks them for validity. If anything
909is found to be inconsistent, it will print an error message to standard
910error and call \f(CW\*(C`abort ()\*(C'\fR.
911.Sp
912This can be used to catch bugs inside libev itself: under normal
913circumstances, this function will never abort as of course libev keeps its
914data structures consistent.
493.SH "ANATOMY OF A WATCHER" 915.SH "ANATOMY OF A WATCHER"
494.IX Header "ANATOMY OF A WATCHER" 916.IX Header "ANATOMY OF A WATCHER"
495A watcher is a structure that you create and register to record your 917A watcher is a structure that you create and register to record your
496interest in some event. For instance, if you want to wait for \s-1STDIN\s0 to 918interest in some event. For instance, if you want to wait for \s-1STDIN\s0 to
497become readable, you would create an \f(CW\*(C`ev_io\*(C'\fR watcher for that: 919become readable, you would create an \f(CW\*(C`ev_io\*(C'\fR watcher for that:
498.PP 920.PP
499.Vb 5 921.Vb 5
500\& static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents) 922\& static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)
501\& { 923\& {
502\& ev_io_stop (w); 924\& ev_io_stop (w);
503\& ev_unloop (loop, EVUNLOOP_ALL); 925\& ev_unloop (loop, EVUNLOOP_ALL);
504\& } 926\& }
505.Ve 927\&
506.PP
507.Vb 6
508\& struct ev_loop *loop = ev_default_loop (0); 928\& struct ev_loop *loop = ev_default_loop (0);
509\& struct ev_io stdin_watcher; 929\& struct ev_io stdin_watcher;
510\& ev_init (&stdin_watcher, my_cb); 930\& ev_init (&stdin_watcher, my_cb);
511\& ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ); 931\& ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
512\& ev_io_start (loop, &stdin_watcher); 932\& ev_io_start (loop, &stdin_watcher);
513\& ev_loop (loop, 0); 933\& ev_loop (loop, 0);
514.Ve 934.Ve
515.PP 935.PP
516As you can see, you are responsible for allocating the memory for your 936As you can see, you are responsible for allocating the memory for your
517watcher structures (and it is usually a bad idea to do this on the stack, 937watcher structures (and it is usually a bad idea to do this on the stack,
518although this can sometimes be quite valid). 938although this can sometimes be quite valid).
519.PP 939.PP
520Each watcher structure must be initialised by a call to \f(CW\*(C`ev_init 940Each watcher structure must be initialised by a call to \f(CW\*(C`ev_init
521(watcher *, callback)\*(C'\fR, which expects a callback to be provided. This 941(watcher *, callback)\*(C'\fR, which expects a callback to be provided. This
522callback gets invoked each time the event occurs (or, in the case of io 942callback gets invoked each time the event occurs (or, in the case of I/O
523watchers, each time the event loop detects that the file descriptor given 943watchers, each time the event loop detects that the file descriptor given
524is readable and/or writable). 944is readable and/or writable).
525.PP 945.PP
526Each watcher type has its own \f(CW\*(C`ev_<type>_set (watcher *, ...)\*(C'\fR macro 946Each watcher type has its own \f(CW\*(C`ev_<type>_set (watcher *, ...)\*(C'\fR macro
527with arguments specific to this watcher type. There is also a macro 947with arguments specific to this watcher type. There is also a macro
… …
533*)\*(C'\fR), and you can stop watching for events at any time by calling the 953*)\*(C'\fR), and you can stop watching for events at any time by calling the
534corresponding stop function (\f(CW\*(C`ev_<type>_stop (loop, watcher *)\*(C'\fR. 954corresponding stop function (\f(CW\*(C`ev_<type>_stop (loop, watcher *)\*(C'\fR.
535.PP 955.PP
536As long as your watcher is active (has been started but not stopped) you 956As long as your watcher is active (has been started but not stopped) you
537must not touch the values stored in it. Most specifically you must never 957must not touch the values stored in it. Most specifically you must never
538reinitialise it or call its set macro. 958reinitialise it or call its \f(CW\*(C`set\*(C'\fR macro.
539.PP
540You can check whether an event is active by calling the \f(CW\*(C`ev_is_active
541(watcher *)\*(C'\fR macro. To see whether an event is outstanding (but the
542callback for it has not been called yet) you can use the \f(CW\*(C`ev_is_pending
543(watcher *)\*(C'\fR macro.
544.PP 959.PP
545Each and every callback receives the event loop pointer as first, the 960Each and every callback receives the event loop pointer as first, the
546registered watcher structure as second, and a bitset of received events as 961registered watcher structure as second, and a bitset of received events as
547third argument. 962third argument.
548.PP 963.PP
… …
573The signal specified in the \f(CW\*(C`ev_signal\*(C'\fR watcher has been received by a thread. 988The signal specified in the \f(CW\*(C`ev_signal\*(C'\fR watcher has been received by a thread.
574.ie n .IP """EV_CHILD""" 4 989.ie n .IP """EV_CHILD""" 4
575.el .IP "\f(CWEV_CHILD\fR" 4 990.el .IP "\f(CWEV_CHILD\fR" 4
576.IX Item "EV_CHILD" 991.IX Item "EV_CHILD"
577The pid specified in the \f(CW\*(C`ev_child\*(C'\fR watcher has received a status change. 992The pid specified in the \f(CW\*(C`ev_child\*(C'\fR watcher has received a status change.
993.ie n .IP """EV_STAT""" 4
994.el .IP "\f(CWEV_STAT\fR" 4
995.IX Item "EV_STAT"
996The path specified in the \f(CW\*(C`ev_stat\*(C'\fR watcher changed its attributes somehow.
578.ie n .IP """EV_IDLE""" 4 997.ie n .IP """EV_IDLE""" 4
579.el .IP "\f(CWEV_IDLE\fR" 4 998.el .IP "\f(CWEV_IDLE\fR" 4
580.IX Item "EV_IDLE" 999.IX Item "EV_IDLE"
581The \f(CW\*(C`ev_idle\*(C'\fR watcher has determined that you have nothing better to do. 1000The \f(CW\*(C`ev_idle\*(C'\fR watcher has determined that you have nothing better to do.
582.ie n .IP """EV_PREPARE""" 4 1001.ie n .IP """EV_PREPARE""" 4
… …
592\&\f(CW\*(C`ev_loop\*(C'\fR has gathered them, but before it invokes any callbacks for any 1011\&\f(CW\*(C`ev_loop\*(C'\fR has gathered them, but before it invokes any callbacks for any
593received events. Callbacks of both watcher types can start and stop as 1012received events. Callbacks of both watcher types can start and stop as
594many watchers as they want, and all of them will be taken into account 1013many watchers as they want, and all of them will be taken into account
595(for example, a \f(CW\*(C`ev_prepare\*(C'\fR watcher might start an idle watcher to keep 1014(for example, a \f(CW\*(C`ev_prepare\*(C'\fR watcher might start an idle watcher to keep
596\&\f(CW\*(C`ev_loop\*(C'\fR from blocking). 1015\&\f(CW\*(C`ev_loop\*(C'\fR from blocking).
1016.ie n .IP """EV_EMBED""" 4
1017.el .IP "\f(CWEV_EMBED\fR" 4
1018.IX Item "EV_EMBED"
1019The embedded event loop specified in the \f(CW\*(C`ev_embed\*(C'\fR watcher needs attention.
1020.ie n .IP """EV_FORK""" 4
1021.el .IP "\f(CWEV_FORK\fR" 4
1022.IX Item "EV_FORK"
1023The event loop has been resumed in the child process after fork (see
1024\&\f(CW\*(C`ev_fork\*(C'\fR).
1025.ie n .IP """EV_ASYNC""" 4
1026.el .IP "\f(CWEV_ASYNC\fR" 4
1027.IX Item "EV_ASYNC"
1028The given async watcher has been asynchronously notified (see \f(CW\*(C`ev_async\*(C'\fR).
597.ie n .IP """EV_ERROR""" 4 1029.ie n .IP """EV_ERROR""" 4
598.el .IP "\f(CWEV_ERROR\fR" 4 1030.el .IP "\f(CWEV_ERROR\fR" 4
599.IX Item "EV_ERROR" 1031.IX Item "EV_ERROR"
600An unspecified error has occured, the watcher has been stopped. This might 1032An unspecified error has occurred, the watcher has been stopped. This might
601happen because the watcher could not be properly started because libev 1033happen because the watcher could not be properly started because libev
602ran out of memory, a file descriptor was found to be closed or any other 1034ran out of memory, a file descriptor was found to be closed or any other
603problem. You best act on it by reporting the problem and somehow coping 1035problem. You best act on it by reporting the problem and somehow coping
604with the watcher being stopped. 1036with the watcher being stopped.
605.Sp 1037.Sp
606Libev will usually signal a few \*(L"dummy\*(R" events together with an error, 1038Libev will usually signal a few \*(L"dummy\*(R" events together with an error, for
607for example it might indicate that a fd is readable or writable, and if 1039example it might indicate that a fd is readable or writable, and if your
608your callbacks is well-written it can just attempt the operation and cope 1040callbacks is well-written it can just attempt the operation and cope with
609with the error from \fIread()\fR or \fIwrite()\fR. This will not work in multithreaded 1041the error from \fIread()\fR or \fIwrite()\fR. This will not work in multi-threaded
610programs, though, so beware. 1042programs, though, as the fd could already be closed and reused for another
1043thing, so beware.
1044.Sh "\s-1GENERIC\s0 \s-1WATCHER\s0 \s-1FUNCTIONS\s0"
1045.IX Subsection "GENERIC WATCHER FUNCTIONS"
1046In the following description, \f(CW\*(C`TYPE\*(C'\fR stands for the watcher type,
1047e.g. \f(CW\*(C`timer\*(C'\fR for \f(CW\*(C`ev_timer\*(C'\fR watchers and \f(CW\*(C`io\*(C'\fR for \f(CW\*(C`ev_io\*(C'\fR watchers.
1048.ie n .IP """ev_init"" (ev_TYPE *watcher, callback)" 4
1049.el .IP "\f(CWev_init\fR (ev_TYPE *watcher, callback)" 4
1050.IX Item "ev_init (ev_TYPE *watcher, callback)"
1051This macro initialises the generic portion of a watcher. The contents
1052of the watcher object can be arbitrary (so \f(CW\*(C`malloc\*(C'\fR will do). Only
1053the generic parts of the watcher are initialised, you \fIneed\fR to call
1054the type-specific \f(CW\*(C`ev_TYPE_set\*(C'\fR macro afterwards to initialise the
1055type-specific parts. For each type there is also a \f(CW\*(C`ev_TYPE_init\*(C'\fR macro
1056which rolls both calls into one.
1057.Sp
1058You can reinitialise a watcher at any time as long as it has been stopped
1059(or never started) and there are no pending events outstanding.
1060.Sp
1061The callback is always of type \f(CW\*(C`void (*)(ev_loop *loop, ev_TYPE *watcher,
1062int revents)\*(C'\fR.
1063.Sp
1064Example: Initialise an \f(CW\*(C`ev_io\*(C'\fR watcher in two steps.
1065.Sp
1066.Vb 3
1067\& ev_io w;
1068\& ev_init (&w, my_cb);
1069\& ev_io_set (&w, STDIN_FILENO, EV_READ);
1070.Ve
1071.ie n .IP """ev_TYPE_set"" (ev_TYPE *, [args])" 4
1072.el .IP "\f(CWev_TYPE_set\fR (ev_TYPE *, [args])" 4
1073.IX Item "ev_TYPE_set (ev_TYPE *, [args])"
1074This macro initialises the type-specific parts of a watcher. You need to
1075call \f(CW\*(C`ev_init\*(C'\fR at least once before you call this macro, but you can
1076call \f(CW\*(C`ev_TYPE_set\*(C'\fR any number of times. You must not, however, call this
1077macro on a watcher that is active (it can be pending, however, which is a
1078difference to the \f(CW\*(C`ev_init\*(C'\fR macro).
1079.Sp
1080Although some watcher types do not have type-specific arguments
1081(e.g. \f(CW\*(C`ev_prepare\*(C'\fR) you still need to call its \f(CW\*(C`set\*(C'\fR macro.
1082.Sp
1083See \f(CW\*(C`ev_init\*(C'\fR, above, for an example.
1084.ie n .IP """ev_TYPE_init"" (ev_TYPE *watcher, callback, [args])" 4
1085.el .IP "\f(CWev_TYPE_init\fR (ev_TYPE *watcher, callback, [args])" 4
1086.IX Item "ev_TYPE_init (ev_TYPE *watcher, callback, [args])"
1087This convenience macro rolls both \f(CW\*(C`ev_init\*(C'\fR and \f(CW\*(C`ev_TYPE_set\*(C'\fR macro
1088calls into a single call. This is the most convenient method to initialise
1089a watcher. The same limitations apply, of course.
1090.Sp
1091Example: Initialise and set an \f(CW\*(C`ev_io\*(C'\fR watcher in one step.
1092.Sp
1093.Vb 1
1094\& ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
1095.Ve
1096.ie n .IP """ev_TYPE_start"" (loop *, ev_TYPE *watcher)" 4
1097.el .IP "\f(CWev_TYPE_start\fR (loop *, ev_TYPE *watcher)" 4
1098.IX Item "ev_TYPE_start (loop *, ev_TYPE *watcher)"
1099Starts (activates) the given watcher. Only active watchers will receive
1100events. If the watcher is already active nothing will happen.
1101.Sp
1102Example: Start the \f(CW\*(C`ev_io\*(C'\fR watcher that is being abused as example in this
1103whole section.
1104.Sp
1105.Vb 1
1106\& ev_io_start (EV_DEFAULT_UC, &w);
1107.Ve
1108.ie n .IP """ev_TYPE_stop"" (loop *, ev_TYPE *watcher)" 4
1109.el .IP "\f(CWev_TYPE_stop\fR (loop *, ev_TYPE *watcher)" 4
1110.IX Item "ev_TYPE_stop (loop *, ev_TYPE *watcher)"
1111Stops the given watcher if active, and clears the pending status (whether
1112the watcher was active or not).
1113.Sp
1114It is possible that stopped watchers are pending \- for example,
1115non-repeating timers are being stopped when they become pending \- but
1116calling \f(CW\*(C`ev_TYPE_stop\*(C'\fR ensures that the watcher is neither active nor
1117pending. If you want to free or reuse the memory used by the watcher it is
1118therefore a good idea to always call its \f(CW\*(C`ev_TYPE_stop\*(C'\fR function.
1119.IP "bool ev_is_active (ev_TYPE *watcher)" 4
1120.IX Item "bool ev_is_active (ev_TYPE *watcher)"
1121Returns a true value iff the watcher is active (i.e. it has been started
1122and not yet been stopped). As long as a watcher is active you must not modify
1123it.
1124.IP "bool ev_is_pending (ev_TYPE *watcher)" 4
1125.IX Item "bool ev_is_pending (ev_TYPE *watcher)"
1126Returns a true value iff the watcher is pending, (i.e. it has outstanding
1127events but its callback has not yet been invoked). As long as a watcher
1128is pending (but not active) you must not call an init function on it (but
1129\&\f(CW\*(C`ev_TYPE_set\*(C'\fR is safe), you must not change its priority, and you must
1130make sure the watcher is available to libev (e.g. you cannot \f(CW\*(C`free ()\*(C'\fR
1131it).
1132.IP "callback ev_cb (ev_TYPE *watcher)" 4
1133.IX Item "callback ev_cb (ev_TYPE *watcher)"
1134Returns the callback currently set on the watcher.
1135.IP "ev_cb_set (ev_TYPE *watcher, callback)" 4
1136.IX Item "ev_cb_set (ev_TYPE *watcher, callback)"
1137Change the callback. You can change the callback at virtually any time
1138(modulo threads).
1139.IP "ev_set_priority (ev_TYPE *watcher, priority)" 4
1140.IX Item "ev_set_priority (ev_TYPE *watcher, priority)"
1141.PD 0
1142.IP "int ev_priority (ev_TYPE *watcher)" 4
1143.IX Item "int ev_priority (ev_TYPE *watcher)"
1144.PD
1145Set and query the priority of the watcher. The priority is a small
1146integer between \f(CW\*(C`EV_MAXPRI\*(C'\fR (default: \f(CW2\fR) and \f(CW\*(C`EV_MINPRI\*(C'\fR
1147(default: \f(CW\*(C`\-2\*(C'\fR). Pending watchers with higher priority will be invoked
1148before watchers with lower priority, but priority will not keep watchers
1149from being executed (except for \f(CW\*(C`ev_idle\*(C'\fR watchers).
1150.Sp
1151This means that priorities are \fIonly\fR used for ordering callback
1152invocation after new events have been received. This is useful, for
1153example, to reduce latency after idling, or more often, to bind two
1154watchers on the same event and make sure one is called first.
1155.Sp
1156If you need to suppress invocation when higher priority events are pending
1157you need to look at \f(CW\*(C`ev_idle\*(C'\fR watchers, which provide this functionality.
1158.Sp
1159You \fImust not\fR change the priority of a watcher as long as it is active or
1160pending.
1161.Sp
1162The default priority used by watchers when no priority has been set is
1163always \f(CW0\fR, which is supposed to not be too high and not be too low :).
1164.Sp
1165Setting a priority outside the range of \f(CW\*(C`EV_MINPRI\*(C'\fR to \f(CW\*(C`EV_MAXPRI\*(C'\fR is
1166fine, as long as you do not mind that the priority value you query might
1167or might not have been adjusted to be within valid range.
1168.IP "ev_invoke (loop, ev_TYPE *watcher, int revents)" 4
1169.IX Item "ev_invoke (loop, ev_TYPE *watcher, int revents)"
1170Invoke the \f(CW\*(C`watcher\*(C'\fR with the given \f(CW\*(C`loop\*(C'\fR and \f(CW\*(C`revents\*(C'\fR. Neither
1171\&\f(CW\*(C`loop\*(C'\fR nor \f(CW\*(C`revents\*(C'\fR need to be valid as long as the watcher callback
1172can deal with that fact, as both are simply passed through to the
1173callback.
1174.IP "int ev_clear_pending (loop, ev_TYPE *watcher)" 4
1175.IX Item "int ev_clear_pending (loop, ev_TYPE *watcher)"
1176If the watcher is pending, this function clears its pending status and
1177returns its \f(CW\*(C`revents\*(C'\fR bitset (as if its callback was invoked). If the
1178watcher isn't pending it does nothing and returns \f(CW0\fR.
1179.Sp
1180Sometimes it can be useful to \*(L"poll\*(R" a watcher instead of waiting for its
1181callback to be invoked, which can be accomplished with this function.
611.Sh "\s-1ASSOCIATING\s0 \s-1CUSTOM\s0 \s-1DATA\s0 \s-1WITH\s0 A \s-1WATCHER\s0" 1182.Sh "\s-1ASSOCIATING\s0 \s-1CUSTOM\s0 \s-1DATA\s0 \s-1WITH\s0 A \s-1WATCHER\s0"
612.IX Subsection "ASSOCIATING CUSTOM DATA WITH A WATCHER" 1183.IX Subsection "ASSOCIATING CUSTOM DATA WITH A WATCHER"
613Each watcher has, by default, a member \f(CW\*(C`void *data\*(C'\fR that you can change 1184Each watcher has, by default, a member \f(CW\*(C`void *data\*(C'\fR that you can change
614and read at any time, libev will completely ignore it. This can be used 1185and read at any time: libev will completely ignore it. This can be used
615to associate arbitrary data with your watcher. If you need more data and 1186to associate arbitrary data with your watcher. If you need more data and
616don't want to allocate memory and store a pointer to it in that data 1187don't want to allocate memory and store a pointer to it in that data
617member, you can also \*(L"subclass\*(R" the watcher type and provide your own 1188member, you can also \*(L"subclass\*(R" the watcher type and provide your own
618data: 1189data:
619.PP 1190.PP
620.Vb 7 1191.Vb 7
621\& struct my_io 1192\& struct my_io
622\& { 1193\& {
623\& struct ev_io io; 1194\& struct ev_io io;
624\& int otherfd; 1195\& int otherfd;
625\& void *somedata; 1196\& void *somedata;
626\& struct whatever *mostinteresting; 1197\& struct whatever *mostinteresting;
627\& } 1198\& };
1199\&
1200\& ...
1201\& struct my_io w;
1202\& ev_io_init (&w.io, my_cb, fd, EV_READ);
628.Ve 1203.Ve
629.PP 1204.PP
630And since your callback will be called with a pointer to the watcher, you 1205And since your callback will be called with a pointer to the watcher, you
631can cast it back to your own type: 1206can cast it back to your own type:
632.PP 1207.PP
633.Vb 5 1208.Vb 5
634\& static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents) 1209\& static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)
635\& { 1210\& {
636\& struct my_io *w = (struct my_io *)w_; 1211\& struct my_io *w = (struct my_io *)w_;
637\& ... 1212\& ...
638\& } 1213\& }
639.Ve 1214.Ve
640.PP 1215.PP
641More interesting and less C\-conformant ways of catsing your callback type 1216More interesting and less C\-conformant ways of casting your callback type
642have been omitted.... 1217instead have been omitted.
1218.PP
1219Another common scenario is to use some data structure with multiple
1220embedded watchers:
1221.PP
1222.Vb 6
1223\& struct my_biggy
1224\& {
1225\& int some_data;
1226\& ev_timer t1;
1227\& ev_timer t2;
1228\& }
1229.Ve
1230.PP
1231In this case getting the pointer to \f(CW\*(C`my_biggy\*(C'\fR is a bit more
1232complicated: Either you store the address of your \f(CW\*(C`my_biggy\*(C'\fR struct
1233in the \f(CW\*(C`data\*(C'\fR member of the watcher (for woozies), or you need to use
1234some pointer arithmetic using \f(CW\*(C`offsetof\*(C'\fR inside your watchers (for real
1235programmers):
1236.PP
1237.Vb 1
1238\& #include <stddef.h>
1239\&
1240\& static void
1241\& t1_cb (EV_P_ struct ev_timer *w, int revents)
1242\& {
1243\& struct my_biggy big = (struct my_biggy *
1244\& (((char *)w) \- offsetof (struct my_biggy, t1));
1245\& }
1246\&
1247\& static void
1248\& t2_cb (EV_P_ struct ev_timer *w, int revents)
1249\& {
1250\& struct my_biggy big = (struct my_biggy *
1251\& (((char *)w) \- offsetof (struct my_biggy, t2));
1252\& }
1253.Ve
643.SH "WATCHER TYPES" 1254.SH "WATCHER TYPES"
644.IX Header "WATCHER TYPES" 1255.IX Header "WATCHER TYPES"
645This section describes each watcher in detail, but will not repeat 1256This section describes each watcher in detail, but will not repeat
646information given in the last section. 1257information given in the last section. Any initialisation/set macros,
1258functions and members specific to the watcher type are explained.
1259.PP
1260Members are additionally marked with either \fI[read\-only]\fR, meaning that,
1261while the watcher is active, you can look at the member and expect some
1262sensible content, but you must not modify it (you can modify it while the
1263watcher is stopped to your hearts content), or \fI[read\-write]\fR, which
1264means you can expect it to have some sensible content while the watcher
1265is active, but you can also modify it. Modifying it may not do something
1266sensible or take immediate effect (or do anything at all), but libev will
1267not crash or malfunction in any way.
647.ie n .Sh """ev_io"" \- is this file descriptor readable or writable" 1268.ie n .Sh """ev_io"" \- is this file descriptor readable or writable?"
648.el .Sh "\f(CWev_io\fP \- is this file descriptor readable or writable" 1269.el .Sh "\f(CWev_io\fP \- is this file descriptor readable or writable?"
649.IX Subsection "ev_io - is this file descriptor readable or writable" 1270.IX Subsection "ev_io - is this file descriptor readable or writable?"
650I/O watchers check whether a file descriptor is readable or writable 1271I/O watchers check whether a file descriptor is readable or writable
651in each iteration of the event loop (This behaviour is called 1272in each iteration of the event loop, or, more precisely, when reading
652level-triggering because you keep receiving events as long as the 1273would not block the process and writing would at least be able to write
653condition persists. Remember you can stop the watcher if you don't want to 1274some data. This behaviour is called level-triggering because you keep
654act on the event and neither want to receive future events). 1275receiving events as long as the condition persists. Remember you can stop
1276the watcher if you don't want to act on the event and neither want to
1277receive future events.
655.PP 1278.PP
656In general you can register as many read and/or write event watchers per 1279In general you can register as many read and/or write event watchers per
657fd as you want (as long as you don't confuse yourself). Setting all file 1280fd as you want (as long as you don't confuse yourself). Setting all file
658descriptors to non-blocking mode is also usually a good idea (but not 1281descriptors to non-blocking mode is also usually a good idea (but not
659required if you know what you are doing). 1282required if you know what you are doing).
660.PP 1283.PP
661You have to be careful with dup'ed file descriptors, though. Some backends 1284If you cannot use non-blocking mode, then force the use of a
662(the linux epoll backend is a notable example) cannot handle dup'ed file 1285known-to-be-good backend (at the time of this writing, this includes only
663descriptors correctly if you register interest in two or more fds pointing 1286\&\f(CW\*(C`EVBACKEND_SELECT\*(C'\fR and \f(CW\*(C`EVBACKEND_POLL\*(C'\fR).
664to the same underlying file/socket etc. description (that is, they share
665the same underlying \*(L"file open\*(R").
666.PP 1287.PP
667If you must do this, then force the use of a known-to-be-good backend 1288Another thing you have to watch out for is that it is quite easy to
668(at the time of this writing, this includes only \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR and 1289receive \*(L"spurious\*(R" readiness notifications, that is your callback might
1290be called with \f(CW\*(C`EV_READ\*(C'\fR but a subsequent \f(CW\*(C`read\*(C'\fR(2) will actually block
1291because there is no data. Not only are some backends known to create a
1292lot of those (for example Solaris ports), it is very easy to get into
1293this situation even with a relatively standard program structure. Thus
1294it is best to always use non-blocking I/O: An extra \f(CW\*(C`read\*(C'\fR(2) returning
1295\&\f(CW\*(C`EAGAIN\*(C'\fR is far preferable to a program hanging until some data arrives.
1296.PP
1297If you cannot run the fd in non-blocking mode (for example you should
1298not play around with an Xlib connection), then you have to separately
1299re-test whether a file descriptor is really ready with a known-to-be good
1300interface such as poll (fortunately in our Xlib example, Xlib already
1301does this on its own, so its quite safe to use). Some people additionally
1302use \f(CW\*(C`SIGALRM\*(C'\fR and an interval timer, just to be sure you won't block
1303indefinitely.
1304.PP
1305But really, best use non-blocking mode.
1306.PP
1307\fIThe special problem of disappearing file descriptors\fR
1308.IX Subsection "The special problem of disappearing file descriptors"
1309.PP
1310Some backends (e.g. kqueue, epoll) need to be told about closing a file
1311descriptor (either due to calling \f(CW\*(C`close\*(C'\fR explicitly or any other means,
1312such as \f(CW\*(C`dup2\*(C'\fR). The reason is that you register interest in some file
1313descriptor, but when it goes away, the operating system will silently drop
1314this interest. If another file descriptor with the same number then is
1315registered with libev, there is no efficient way to see that this is, in
1316fact, a different file descriptor.
1317.PP
1318To avoid having to explicitly tell libev about such cases, libev follows
1319the following policy: Each time \f(CW\*(C`ev_io_set\*(C'\fR is being called, libev
1320will assume that this is potentially a new file descriptor, otherwise
1321it is assumed that the file descriptor stays the same. That means that
1322you \fIhave\fR to call \f(CW\*(C`ev_io_set\*(C'\fR (or \f(CW\*(C`ev_io_init\*(C'\fR) when you change the
1323descriptor even if the file descriptor number itself did not change.
1324.PP
1325This is how one would do it normally anyway, the important point is that
1326the libev application should not optimise around libev but should leave
1327optimisations to libev.
1328.PP
1329\fIThe special problem of dup'ed file descriptors\fR
1330.IX Subsection "The special problem of dup'ed file descriptors"
1331.PP
1332Some backends (e.g. epoll), cannot register events for file descriptors,
1333but only events for the underlying file descriptions. That means when you
1334have \f(CW\*(C`dup ()\*(C'\fR'ed file descriptors or weirder constellations, and register
1335events for them, only one file descriptor might actually receive events.
1336.PP
1337There is no workaround possible except not registering events
1338for potentially \f(CW\*(C`dup ()\*(C'\fR'ed file descriptors, or to resort to
1339\&\f(CW\*(C`EVBACKEND_SELECT\*(C'\fR or \f(CW\*(C`EVBACKEND_POLL\*(C'\fR.
1340.PP
1341\fIThe special problem of fork\fR
1342.IX Subsection "The special problem of fork"
1343.PP
1344Some backends (epoll, kqueue) do not support \f(CW\*(C`fork ()\*(C'\fR at all or exhibit
1345useless behaviour. Libev fully supports fork, but needs to be told about
1346it in the child.
1347.PP
1348To support fork in your programs, you either have to call
1349\&\f(CW\*(C`ev_default_fork ()\*(C'\fR or \f(CW\*(C`ev_loop_fork ()\*(C'\fR after a fork in the child,
1350enable \f(CW\*(C`EVFLAG_FORKCHECK\*(C'\fR, or resort to \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR or
669\&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR). 1351\&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR.
1352.PP
1353\fIThe special problem of \s-1SIGPIPE\s0\fR
1354.IX Subsection "The special problem of SIGPIPE"
1355.PP
1356While not really specific to libev, it is easy to forget about \f(CW\*(C`SIGPIPE\*(C'\fR:
1357when writing to a pipe whose other end has been closed, your program gets
1358sent a \s-1SIGPIPE\s0, which, by default, aborts your program. For most programs
1359this is sensible behaviour, for daemons, this is usually undesirable.
1360.PP
1361So when you encounter spurious, unexplained daemon exits, make sure you
1362ignore \s-1SIGPIPE\s0 (and maybe make sure you log the exit status of your daemon
1363somewhere, as that would have given you a big clue).
1364.PP
1365\fIWatcher-Specific Functions\fR
1366.IX Subsection "Watcher-Specific Functions"
670.IP "ev_io_init (ev_io *, callback, int fd, int events)" 4 1367.IP "ev_io_init (ev_io *, callback, int fd, int events)" 4
671.IX Item "ev_io_init (ev_io *, callback, int fd, int events)" 1368.IX Item "ev_io_init (ev_io *, callback, int fd, int events)"
672.PD 0 1369.PD 0
673.IP "ev_io_set (ev_io *, int fd, int events)" 4 1370.IP "ev_io_set (ev_io *, int fd, int events)" 4
674.IX Item "ev_io_set (ev_io *, int fd, int events)" 1371.IX Item "ev_io_set (ev_io *, int fd, int events)"
675.PD 1372.PD
676Configures an \f(CW\*(C`ev_io\*(C'\fR watcher. The fd is the file descriptor to rceeive 1373Configures an \f(CW\*(C`ev_io\*(C'\fR watcher. The \f(CW\*(C`fd\*(C'\fR is the file descriptor to
677events for and events is either \f(CW\*(C`EV_READ\*(C'\fR, \f(CW\*(C`EV_WRITE\*(C'\fR or \f(CW\*(C`EV_READ | 1374receive events for and \f(CW\*(C`events\*(C'\fR is either \f(CW\*(C`EV_READ\*(C'\fR, \f(CW\*(C`EV_WRITE\*(C'\fR or
678EV_WRITE\*(C'\fR to receive the given events. 1375\&\f(CW\*(C`EV_READ | EV_WRITE\*(C'\fR, to express the desire to receive the given events.
679.Sp 1376.IP "int fd [read\-only]" 4
680Please note that most of the more scalable backend mechanisms (for example 1377.IX Item "int fd [read-only]"
681epoll and solaris ports) can result in spurious readyness notifications 1378The file descriptor being watched.
682for file descriptors, so you practically need to use non-blocking I/O (and 1379.IP "int events [read\-only]" 4
683treat callback invocation as hint only), or retest separately with a safe 1380.IX Item "int events [read-only]"
684interface before doing I/O (XLib can do this), or force the use of either 1381The events being watched.
685\&\f(CW\*(C`EVBACKEND_SELECT\*(C'\fR or \f(CW\*(C`EVBACKEND_POLL\*(C'\fR, which don't suffer from this 1382.PP
686problem. Also note that it is quite easy to have your callback invoked 1383\fIExamples\fR
687when the readyness condition is no longer valid even when employing 1384.IX Subsection "Examples"
688typical ways of handling events, so its a good idea to use non-blocking 1385.PP
689I/O unconditionally. 1386Example: Call \f(CW\*(C`stdin_readable_cb\*(C'\fR when \s-1STDIN_FILENO\s0 has become, well
1387readable, but only once. Since it is likely line-buffered, you could
1388attempt to read a whole line in the callback.
1389.PP
1390.Vb 6
1391\& static void
1392\& stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
1393\& {
1394\& ev_io_stop (loop, w);
1395\& .. read from stdin here (or from w\->fd) and handle any I/O errors
1396\& }
1397\&
1398\& ...
1399\& struct ev_loop *loop = ev_default_init (0);
1400\& struct ev_io stdin_readable;
1401\& ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1402\& ev_io_start (loop, &stdin_readable);
1403\& ev_loop (loop, 0);
1404.Ve
690.ie n .Sh """ev_timer"" \- relative and optionally recurring timeouts" 1405.ie n .Sh """ev_timer"" \- relative and optionally repeating timeouts"
691.el .Sh "\f(CWev_timer\fP \- relative and optionally recurring timeouts" 1406.el .Sh "\f(CWev_timer\fP \- relative and optionally repeating timeouts"
692.IX Subsection "ev_timer - relative and optionally recurring timeouts" 1407.IX Subsection "ev_timer - relative and optionally repeating timeouts"
693Timer watchers are simple relative timers that generate an event after a 1408Timer watchers are simple relative timers that generate an event after a
694given time, and optionally repeating in regular intervals after that. 1409given time, and optionally repeating in regular intervals after that.
695.PP 1410.PP
696The timers are based on real time, that is, if you register an event that 1411The timers are based on real time, that is, if you register an event that
697times out after an hour and you reset your system clock to last years 1412times out after an hour and you reset your system clock to January last
698time, it will still time out after (roughly) and hour. \*(L"Roughly\*(R" because 1413year, it will still time out after (roughly) one hour. \*(L"Roughly\*(R" because
699detecting time jumps is hard, and some inaccuracies are unavoidable (the 1414detecting time jumps is hard, and some inaccuracies are unavoidable (the
700monotonic clock option helps a lot here). 1415monotonic clock option helps a lot here).
1416.PP
1417The callback is guaranteed to be invoked only \fIafter\fR its timeout has
1418passed, but if multiple timers become ready during the same loop iteration
1419then order of execution is undefined.
1420.PP
1421\fIThe special problem of time updates\fR
1422.IX Subsection "The special problem of time updates"
1423.PP
1424Establishing the current time is a costly operation (it usually takes at
1425least two system calls): \s-1EV\s0 therefore updates its idea of the current
1426time only before and after \f(CW\*(C`ev_loop\*(C'\fR collects new events, which causes a
1427growing difference between \f(CW\*(C`ev_now ()\*(C'\fR and \f(CW\*(C`ev_time ()\*(C'\fR when handling
1428lots of events in one iteration.
701.PP 1429.PP
702The relative timeouts are calculated relative to the \f(CW\*(C`ev_now ()\*(C'\fR 1430The relative timeouts are calculated relative to the \f(CW\*(C`ev_now ()\*(C'\fR
703time. This is usually the right thing as this timestamp refers to the time 1431time. This is usually the right thing as this timestamp refers to the time
704of the event triggering whatever timeout you are modifying/starting. If 1432of the event triggering whatever timeout you are modifying/starting. If
705you suspect event processing to be delayed and you \fIneed\fR to base the timeout 1433you suspect event processing to be delayed and you \fIneed\fR to base the
706on the current time, use something like this to adjust for this: 1434timeout on the current time, use something like this to adjust for this:
707.PP 1435.PP
708.Vb 1 1436.Vb 1
709\& ev_timer_set (&timer, after + ev_now () - ev_time (), 0.); 1437\& ev_timer_set (&timer, after + ev_now () \- ev_time (), 0.);
710.Ve 1438.Ve
711.PP 1439.PP
712The callback is guarenteed to be invoked only when its timeout has passed, 1440If the event loop is suspended for a long time, you can also force an
713but if multiple timers become ready during the same loop iteration then 1441update of the time returned by \f(CW\*(C`ev_now ()\*(C'\fR by calling \f(CW\*(C`ev_now_update
714order of execution is undefined. 1442()\*(C'\fR.
1443.PP
1444\fIWatcher-Specific Functions and Data Members\fR
1445.IX Subsection "Watcher-Specific Functions and Data Members"
715.IP "ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)" 4 1446.IP "ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)" 4
716.IX Item "ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)" 1447.IX Item "ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)"
717.PD 0 1448.PD 0
718.IP "ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)" 4 1449.IP "ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)" 4
719.IX Item "ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)" 1450.IX Item "ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)"
720.PD 1451.PD
721Configure the timer to trigger after \f(CW\*(C`after\*(C'\fR seconds. If \f(CW\*(C`repeat\*(C'\fR is 1452Configure the timer to trigger after \f(CW\*(C`after\*(C'\fR seconds. If \f(CW\*(C`repeat\*(C'\fR
722\&\f(CW0.\fR, then it will automatically be stopped. If it is positive, then the 1453is \f(CW0.\fR, then it will automatically be stopped once the timeout is
723timer will automatically be configured to trigger again \f(CW\*(C`repeat\*(C'\fR seconds 1454reached. If it is positive, then the timer will automatically be
724later, again, and again, until stopped manually. 1455configured to trigger again \f(CW\*(C`repeat\*(C'\fR seconds later, again, and again,
1456until stopped manually.
725.Sp 1457.Sp
726The timer itself will do a best-effort at avoiding drift, that is, if you 1458The timer itself will do a best-effort at avoiding drift, that is, if
727configure a timer to trigger every 10 seconds, then it will trigger at 1459you configure a timer to trigger every 10 seconds, then it will normally
728exactly 10 second intervals. If, however, your program cannot keep up with 1460trigger at exactly 10 second intervals. If, however, your program cannot
729the timer (because it takes longer than those 10 seconds to do stuff) the 1461keep up with the timer (because it takes longer than those 10 seconds to
730timer will not fire more than once per event loop iteration. 1462do stuff) the timer will not fire more than once per event loop iteration.
731.IP "ev_timer_again (loop)" 4 1463.IP "ev_timer_again (loop, ev_timer *)" 4
732.IX Item "ev_timer_again (loop)" 1464.IX Item "ev_timer_again (loop, ev_timer *)"
733This will act as if the timer timed out and restart it again if it is 1465This will act as if the timer timed out and restart it again if it is
734repeating. The exact semantics are: 1466repeating. The exact semantics are:
735.Sp 1467.Sp
1468If the timer is pending, its pending status is cleared.
1469.Sp
736If the timer is started but nonrepeating, stop it. 1470If the timer is started but non-repeating, stop it (as if it timed out).
737.Sp 1471.Sp
738If the timer is repeating, either start it if necessary (with the repeat 1472If the timer is repeating, either start it if necessary (with the
739value), or reset the running timer to the repeat value. 1473\&\f(CW\*(C`repeat\*(C'\fR value), or reset the running timer to the \f(CW\*(C`repeat\*(C'\fR value.
740.Sp 1474.Sp
741This sounds a bit complicated, but here is a useful and typical 1475This sounds a bit complicated, but here is a useful and typical
742example: Imagine you have a tcp connection and you want a so-called idle 1476example: Imagine you have a \s-1TCP\s0 connection and you want a so-called idle
743timeout, that is, you want to be called when there have been, say, 60 1477timeout, that is, you want to be called when there have been, say, 60
744seconds of inactivity on the socket. The easiest way to do this is to 1478seconds of inactivity on the socket. The easiest way to do this is to
745configure an \f(CW\*(C`ev_timer\*(C'\fR with after=repeat=60 and calling ev_timer_again each 1479configure an \f(CW\*(C`ev_timer\*(C'\fR with a \f(CW\*(C`repeat\*(C'\fR value of \f(CW60\fR and then call
746time you successfully read or write some data. If you go into an idle 1480\&\f(CW\*(C`ev_timer_again\*(C'\fR each time you successfully read or write some data. If
747state where you do not expect data to travel on the socket, you can stop 1481you go into an idle state where you do not expect data to travel on the
748the timer, and again will automatically restart it if need be. 1482socket, you can \f(CW\*(C`ev_timer_stop\*(C'\fR the timer, and \f(CW\*(C`ev_timer_again\*(C'\fR will
1483automatically restart it if need be.
1484.Sp
1485That means you can ignore the \f(CW\*(C`after\*(C'\fR value and \f(CW\*(C`ev_timer_start\*(C'\fR
1486altogether and only ever use the \f(CW\*(C`repeat\*(C'\fR value and \f(CW\*(C`ev_timer_again\*(C'\fR:
1487.Sp
1488.Vb 8
1489\& ev_timer_init (timer, callback, 0., 5.);
1490\& ev_timer_again (loop, timer);
1491\& ...
1492\& timer\->again = 17.;
1493\& ev_timer_again (loop, timer);
1494\& ...
1495\& timer\->again = 10.;
1496\& ev_timer_again (loop, timer);
1497.Ve
1498.Sp
1499This is more slightly efficient then stopping/starting the timer each time
1500you want to modify its timeout value.
1501.Sp
1502Note, however, that it is often even more efficient to remember the
1503time of the last activity and let the timer time-out naturally. In the
1504callback, you then check whether the time-out is real, or, if there was
1505some activity, you reschedule the watcher to time-out in \*(L"last_activity +
1506timeout \- ev_now ()\*(R" seconds.
1507.IP "ev_tstamp repeat [read\-write]" 4
1508.IX Item "ev_tstamp repeat [read-write]"
1509The current \f(CW\*(C`repeat\*(C'\fR value. Will be used each time the watcher times out
1510or \f(CW\*(C`ev_timer_again\*(C'\fR is called, and determines the next timeout (if any),
1511which is also when any modifications are taken into account.
1512.PP
1513\fIExamples\fR
1514.IX Subsection "Examples"
1515.PP
1516Example: Create a timer that fires after 60 seconds.
1517.PP
1518.Vb 5
1519\& static void
1520\& one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
1521\& {
1522\& .. one minute over, w is actually stopped right here
1523\& }
1524\&
1525\& struct ev_timer mytimer;
1526\& ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1527\& ev_timer_start (loop, &mytimer);
1528.Ve
1529.PP
1530Example: Create a timeout timer that times out after 10 seconds of
1531inactivity.
1532.PP
1533.Vb 5
1534\& static void
1535\& timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
1536\& {
1537\& .. ten seconds without any activity
1538\& }
1539\&
1540\& struct ev_timer mytimer;
1541\& ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1542\& ev_timer_again (&mytimer); /* start timer */
1543\& ev_loop (loop, 0);
1544\&
1545\& // and in some piece of code that gets executed on any "activity":
1546\& // reset the timeout to start ticking again at 10 seconds
1547\& ev_timer_again (&mytimer);
1548.Ve
749.ie n .Sh """ev_periodic"" \- to cron or not to cron" 1549.ie n .Sh """ev_periodic"" \- to cron or not to cron?"
750.el .Sh "\f(CWev_periodic\fP \- to cron or not to cron" 1550.el .Sh "\f(CWev_periodic\fP \- to cron or not to cron?"
751.IX Subsection "ev_periodic - to cron or not to cron" 1551.IX Subsection "ev_periodic - to cron or not to cron?"
752Periodic watchers are also timers of a kind, but they are very versatile 1552Periodic watchers are also timers of a kind, but they are very versatile
753(and unfortunately a bit complex). 1553(and unfortunately a bit complex).
754.PP 1554.PP
755Unlike \f(CW\*(C`ev_timer\*(C'\fR's, they are not based on real time (or relative time) 1555Unlike \f(CW\*(C`ev_timer\*(C'\fR's, they are not based on real time (or relative time)
756but on wallclock time (absolute time). You can tell a periodic watcher 1556but on wall clock time (absolute time). You can tell a periodic watcher
757to trigger \*(L"at\*(R" some specific point in time. For example, if you tell a 1557to trigger after some specific point in time. For example, if you tell a
758periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now () 1558periodic watcher to trigger in 10 seconds (by specifying e.g. \f(CW\*(C`ev_now ()
759+ 10.>) and then reset your system clock to the last year, then it will 1559+ 10.\*(C'\fR, that is, an absolute time not a delay) and then reset your system
1560clock to January of the previous year, then it will take more than year
760take a year to trigger the event (unlike an \f(CW\*(C`ev_timer\*(C'\fR, which would trigger 1561to trigger the event (unlike an \f(CW\*(C`ev_timer\*(C'\fR, which would still trigger
761roughly 10 seconds later and of course not if you reset your system time 1562roughly 10 seconds later as it uses a relative timeout).
762again).
763.PP 1563.PP
764They can also be used to implement vastly more complex timers, such as 1564\&\f(CW\*(C`ev_periodic\*(C'\fRs can also be used to implement vastly more complex timers,
765triggering an event on eahc midnight, local time. 1565such as triggering an event on each \*(L"midnight, local time\*(R", or other
1566complicated rules.
766.PP 1567.PP
767As with timers, the callback is guarenteed to be invoked only when the 1568As with timers, the callback is guaranteed to be invoked only when the
768time (\f(CW\*(C`at\*(C'\fR) has been passed, but if multiple periodic timers become ready 1569time (\f(CW\*(C`at\*(C'\fR) has passed, but if multiple periodic timers become ready
769during the same loop iteration then order of execution is undefined. 1570during the same loop iteration, then order of execution is undefined.
1571.PP
1572\fIWatcher-Specific Functions and Data Members\fR
1573.IX Subsection "Watcher-Specific Functions and Data Members"
770.IP "ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)" 4 1574.IP "ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)" 4
771.IX Item "ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)" 1575.IX Item "ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)"
772.PD 0 1576.PD 0
773.IP "ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)" 4 1577.IP "ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)" 4
774.IX Item "ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)" 1578.IX Item "ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)"
775.PD 1579.PD
776Lots of arguments, lets sort it out... There are basically three modes of 1580Lots of arguments, lets sort it out... There are basically three modes of
777operation, and we will explain them from simplest to complex: 1581operation, and we will explain them from simplest to most complex:
778.RS 4 1582.RS 4
1583.IP "\(bu" 4
779.IP "* absolute timer (interval = reschedule_cb = 0)" 4 1584absolute timer (at = time, interval = reschedule_cb = 0)
780.IX Item "absolute timer (interval = reschedule_cb = 0)" 1585.Sp
781In this configuration the watcher triggers an event at the wallclock time 1586In this configuration the watcher triggers an event after the wall clock
782\&\f(CW\*(C`at\*(C'\fR and doesn't repeat. It will not adjust when a time jump occurs, 1587time \f(CW\*(C`at\*(C'\fR has passed. It will not repeat and will not adjust when a time
783that is, if it is to be run at January 1st 2011 then it will run when the 1588jump occurs, that is, if it is to be run at January 1st 2011 then it will
784system time reaches or surpasses this time. 1589only run when the system clock reaches or surpasses this time.
1590.IP "\(bu" 4
785.IP "* non-repeating interval timer (interval > 0, reschedule_cb = 0)" 4 1591repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
786.IX Item "non-repeating interval timer (interval > 0, reschedule_cb = 0)" 1592.Sp
787In this mode the watcher will always be scheduled to time out at the next 1593In this mode the watcher will always be scheduled to time out at the next
788\&\f(CW\*(C`at + N * interval\*(C'\fR time (for some integer N) and then repeat, regardless 1594\&\f(CW\*(C`at + N * interval\*(C'\fR time (for some integer N, which can also be negative)
789of any time jumps. 1595and then repeat, regardless of any time jumps.
790.Sp 1596.Sp
791This can be used to create timers that do not drift with respect to system 1597This can be used to create timers that do not drift with respect to the
792time: 1598system clock, for example, here is a \f(CW\*(C`ev_periodic\*(C'\fR that triggers each
1599hour, on the hour:
793.Sp 1600.Sp
794.Vb 1 1601.Vb 1
795\& ev_periodic_set (&periodic, 0., 3600., 0); 1602\& ev_periodic_set (&periodic, 0., 3600., 0);
796.Ve 1603.Ve
797.Sp 1604.Sp
798This doesn't mean there will always be 3600 seconds in between triggers, 1605This doesn't mean there will always be 3600 seconds in between triggers,
799but only that the the callback will be called when the system time shows a 1606but only that the callback will be called when the system time shows a
800full hour (\s-1UTC\s0), or more correctly, when the system time is evenly divisible 1607full hour (\s-1UTC\s0), or more correctly, when the system time is evenly divisible
801by 3600. 1608by 3600.
802.Sp 1609.Sp
803Another way to think about it (for the mathematically inclined) is that 1610Another way to think about it (for the mathematically inclined) is that
804\&\f(CW\*(C`ev_periodic\*(C'\fR will try to run the callback in this mode at the next possible 1611\&\f(CW\*(C`ev_periodic\*(C'\fR will try to run the callback in this mode at the next possible
805time where \f(CW\*(C`time = at (mod interval)\*(C'\fR, regardless of any time jumps. 1612time where \f(CW\*(C`time = at (mod interval)\*(C'\fR, regardless of any time jumps.
806.IP "* manual reschedule mode (reschedule_cb = callback)" 4 1613.Sp
807.IX Item "manual reschedule mode (reschedule_cb = callback)" 1614For numerical stability it is preferable that the \f(CW\*(C`at\*(C'\fR value is near
1615\&\f(CW\*(C`ev_now ()\*(C'\fR (the current time), but there is no range requirement for
1616this value, and in fact is often specified as zero.
1617.Sp
1618Note also that there is an upper limit to how often a timer can fire (\s-1CPU\s0
1619speed for example), so if \f(CW\*(C`interval\*(C'\fR is very small then timing stability
1620will of course deteriorate. Libev itself tries to be exact to be about one
1621millisecond (if the \s-1OS\s0 supports it and the machine is fast enough).
1622.IP "\(bu" 4
1623manual reschedule mode (at and interval ignored, reschedule_cb = callback)
1624.Sp
808In this mode the values for \f(CW\*(C`interval\*(C'\fR and \f(CW\*(C`at\*(C'\fR are both being 1625In this mode the values for \f(CW\*(C`interval\*(C'\fR and \f(CW\*(C`at\*(C'\fR are both being
809ignored. Instead, each time the periodic watcher gets scheduled, the 1626ignored. Instead, each time the periodic watcher gets scheduled, the
810reschedule callback will be called with the watcher as first, and the 1627reschedule callback will be called with the watcher as first, and the
811current time as second argument. 1628current time as second argument.
812.Sp 1629.Sp
813\&\s-1NOTE:\s0 \fIThis callback \s-1MUST\s0 \s-1NOT\s0 stop or destroy any periodic watcher, 1630\&\s-1NOTE:\s0 \fIThis callback \s-1MUST\s0 \s-1NOT\s0 stop or destroy any periodic watcher,
814ever, or make any event loop modifications\fR. If you need to stop it, 1631ever, or make \s-1ANY\s0 event loop modifications whatsoever\fR.
815return \f(CW\*(C`now + 1e30\*(C'\fR (or so, fudge fudge) and stop it afterwards (e.g. by
816starting a prepare watcher).
817.Sp 1632.Sp
1633If you need to stop it, return \f(CW\*(C`now + 1e30\*(C'\fR (or so, fudge fudge) and stop
1634it afterwards (e.g. by starting an \f(CW\*(C`ev_prepare\*(C'\fR watcher, which is the
1635only event loop modification you are allowed to do).
1636.Sp
818Its prototype is \f(CW\*(C`ev_tstamp (*reschedule_cb)(struct ev_periodic *w, 1637The callback prototype is \f(CW\*(C`ev_tstamp (*reschedule_cb)(struct ev_periodic
819ev_tstamp now)\*(C'\fR, e.g.: 1638*w, ev_tstamp now)\*(C'\fR, e.g.:
820.Sp 1639.Sp
821.Vb 4 1640.Vb 4
822\& static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now) 1641\& static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
823\& { 1642\& {
824\& return now + 60.; 1643\& return now + 60.;
… …
828It must return the next time to trigger, based on the passed time value 1647It must return the next time to trigger, based on the passed time value
829(that is, the lowest time value larger than to the second argument). It 1648(that is, the lowest time value larger than to the second argument). It
830will usually be called just before the callback will be triggered, but 1649will usually be called just before the callback will be triggered, but
831might be called at other times, too. 1650might be called at other times, too.
832.Sp 1651.Sp
833\&\s-1NOTE:\s0 \fIThis callback must always return a time that is later than the 1652\&\s-1NOTE:\s0 \fIThis callback must always return a time that is higher than or
834passed \f(CI\*(C`now\*(C'\fI value\fR. Not even \f(CW\*(C`now\*(C'\fR itself will do, it \fImust\fR be larger. 1653equal to the passed \f(CI\*(C`now\*(C'\fI value\fR.
835.Sp 1654.Sp
836This can be used to create very complex timers, such as a timer that 1655This can be used to create very complex timers, such as a timer that
837triggers on each midnight, local time. To do this, you would calculate the 1656triggers on \*(L"next midnight, local time\*(R". To do this, you would calculate the
838next midnight after \f(CW\*(C`now\*(C'\fR and return the timestamp value for this. How 1657next midnight after \f(CW\*(C`now\*(C'\fR and return the timestamp value for this. How
839you do this is, again, up to you (but it is not trivial, which is the main 1658you do this is, again, up to you (but it is not trivial, which is the main
840reason I omitted it as an example). 1659reason I omitted it as an example).
841.RE 1660.RE
842.RS 4 1661.RS 4
… …
845.IX Item "ev_periodic_again (loop, ev_periodic *)" 1664.IX Item "ev_periodic_again (loop, ev_periodic *)"
846Simply stops and restarts the periodic watcher again. This is only useful 1665Simply stops and restarts the periodic watcher again. This is only useful
847when you changed some parameters or the reschedule callback would return 1666when you changed some parameters or the reschedule callback would return
848a different time than the last time it was called (e.g. in a crond like 1667a different time than the last time it was called (e.g. in a crond like
849program when the crontabs have changed). 1668program when the crontabs have changed).
1669.IP "ev_tstamp ev_periodic_at (ev_periodic *)" 4
1670.IX Item "ev_tstamp ev_periodic_at (ev_periodic *)"
1671When active, returns the absolute time that the watcher is supposed to
1672trigger next.
1673.IP "ev_tstamp offset [read\-write]" 4
1674.IX Item "ev_tstamp offset [read-write]"
1675When repeating, this contains the offset value, otherwise this is the
1676absolute point in time (the \f(CW\*(C`at\*(C'\fR value passed to \f(CW\*(C`ev_periodic_set\*(C'\fR).
1677.Sp
1678Can be modified any time, but changes only take effect when the periodic
1679timer fires or \f(CW\*(C`ev_periodic_again\*(C'\fR is being called.
1680.IP "ev_tstamp interval [read\-write]" 4
1681.IX Item "ev_tstamp interval [read-write]"
1682The current interval value. Can be modified any time, but changes only
1683take effect when the periodic timer fires or \f(CW\*(C`ev_periodic_again\*(C'\fR is being
1684called.
1685.IP "ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read\-write]" 4
1686.IX Item "ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]"
1687The current reschedule callback, or \f(CW0\fR, if this functionality is
1688switched off. Can be changed any time, but changes only take effect when
1689the periodic timer fires or \f(CW\*(C`ev_periodic_again\*(C'\fR is being called.
1690.PP
1691\fIExamples\fR
1692.IX Subsection "Examples"
1693.PP
1694Example: Call a callback every hour, or, more precisely, whenever the
1695system time is divisible by 3600. The callback invocation times have
1696potentially a lot of jitter, but good long-term stability.
1697.PP
1698.Vb 5
1699\& static void
1700\& clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
1701\& {
1702\& ... its now a full hour (UTC, or TAI or whatever your clock follows)
1703\& }
1704\&
1705\& struct ev_periodic hourly_tick;
1706\& ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1707\& ev_periodic_start (loop, &hourly_tick);
1708.Ve
1709.PP
1710Example: The same as above, but use a reschedule callback to do it:
1711.PP
1712.Vb 1
1713\& #include <math.h>
1714\&
1715\& static ev_tstamp
1716\& my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
1717\& {
1718\& return now + (3600. \- fmod (now, 3600.));
1719\& }
1720\&
1721\& ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1722.Ve
1723.PP
1724Example: Call a callback every hour, starting now:
1725.PP
1726.Vb 4
1727\& struct ev_periodic hourly_tick;
1728\& ev_periodic_init (&hourly_tick, clock_cb,
1729\& fmod (ev_now (loop), 3600.), 3600., 0);
1730\& ev_periodic_start (loop, &hourly_tick);
1731.Ve
850.ie n .Sh """ev_signal"" \- signal me when a signal gets signalled" 1732.ie n .Sh """ev_signal"" \- signal me when a signal gets signalled!"
851.el .Sh "\f(CWev_signal\fP \- signal me when a signal gets signalled" 1733.el .Sh "\f(CWev_signal\fP \- signal me when a signal gets signalled!"
852.IX Subsection "ev_signal - signal me when a signal gets signalled" 1734.IX Subsection "ev_signal - signal me when a signal gets signalled!"
853Signal watchers will trigger an event when the process receives a specific 1735Signal watchers will trigger an event when the process receives a specific
854signal one or more times. Even though signals are very asynchronous, libev 1736signal one or more times. Even though signals are very asynchronous, libev
855will try it's best to deliver signals synchronously, i.e. as part of the 1737will try it's best to deliver signals synchronously, i.e. as part of the
856normal event processing, like any other event. 1738normal event processing, like any other event.
857.PP 1739.PP
1740If you want signals asynchronously, just use \f(CW\*(C`sigaction\*(C'\fR as you would
1741do without libev and forget about sharing the signal. You can even use
1742\&\f(CW\*(C`ev_async\*(C'\fR from a signal handler to synchronously wake up an event loop.
1743.PP
858You can configure as many watchers as you like per signal. Only when the 1744You can configure as many watchers as you like per signal. Only when the
859first watcher gets started will libev actually register a signal watcher 1745first watcher gets started will libev actually register a signal handler
860with the kernel (thus it coexists with your own signal handlers as long 1746with the kernel (thus it coexists with your own signal handlers as long as
861as you don't register any with libev). Similarly, when the last signal 1747you don't register any with libev for the same signal). Similarly, when
862watcher for a signal is stopped libev will reset the signal handler to 1748the last signal watcher for a signal is stopped, libev will reset the
863\&\s-1SIG_DFL\s0 (regardless of what it was set to before). 1749signal handler to \s-1SIG_DFL\s0 (regardless of what it was set to before).
1750.PP
1751If possible and supported, libev will install its handlers with
1752\&\f(CW\*(C`SA_RESTART\*(C'\fR behaviour enabled, so system calls should not be unduly
1753interrupted. If you have a problem with system calls getting interrupted by
1754signals you can block all signals in an \f(CW\*(C`ev_check\*(C'\fR watcher and unblock
1755them in an \f(CW\*(C`ev_prepare\*(C'\fR watcher.
1756.PP
1757\fIWatcher-Specific Functions and Data Members\fR
1758.IX Subsection "Watcher-Specific Functions and Data Members"
864.IP "ev_signal_init (ev_signal *, callback, int signum)" 4 1759.IP "ev_signal_init (ev_signal *, callback, int signum)" 4
865.IX Item "ev_signal_init (ev_signal *, callback, int signum)" 1760.IX Item "ev_signal_init (ev_signal *, callback, int signum)"
866.PD 0 1761.PD 0
867.IP "ev_signal_set (ev_signal *, int signum)" 4 1762.IP "ev_signal_set (ev_signal *, int signum)" 4
868.IX Item "ev_signal_set (ev_signal *, int signum)" 1763.IX Item "ev_signal_set (ev_signal *, int signum)"
869.PD 1764.PD
870Configures the watcher to trigger on the given signal number (usually one 1765Configures the watcher to trigger on the given signal number (usually one
871of the \f(CW\*(C`SIGxxx\*(C'\fR constants). 1766of the \f(CW\*(C`SIGxxx\*(C'\fR constants).
1767.IP "int signum [read\-only]" 4
1768.IX Item "int signum [read-only]"
1769The signal the watcher watches out for.
1770.PP
1771\fIExamples\fR
1772.IX Subsection "Examples"
1773.PP
1774Example: Try to exit cleanly on \s-1SIGINT\s0.
1775.PP
1776.Vb 5
1777\& static void
1778\& sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
1779\& {
1780\& ev_unloop (loop, EVUNLOOP_ALL);
1781\& }
1782\&
1783\& struct ev_signal signal_watcher;
1784\& ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1785\& ev_signal_start (loop, &signal_watcher);
1786.Ve
872.ie n .Sh """ev_child"" \- wait for pid status changes" 1787.ie n .Sh """ev_child"" \- watch out for process status changes"
873.el .Sh "\f(CWev_child\fP \- wait for pid status changes" 1788.el .Sh "\f(CWev_child\fP \- watch out for process status changes"
874.IX Subsection "ev_child - wait for pid status changes" 1789.IX Subsection "ev_child - watch out for process status changes"
875Child watchers trigger when your process receives a \s-1SIGCHLD\s0 in response to 1790Child watchers trigger when your process receives a \s-1SIGCHLD\s0 in response to
876some child status changes (most typically when a child of yours dies). 1791some child status changes (most typically when a child of yours dies or
1792exits). It is permissible to install a child watcher \fIafter\fR the child
1793has been forked (which implies it might have already exited), as long
1794as the event loop isn't entered (or is continued from a watcher), i.e.,
1795forking and then immediately registering a watcher for the child is fine,
1796but forking and registering a watcher a few event loop iterations later is
1797not.
1798.PP
1799Only the default event loop is capable of handling signals, and therefore
1800you can only register child watchers in the default event loop.
1801.PP
1802\fIProcess Interaction\fR
1803.IX Subsection "Process Interaction"
1804.PP
1805Libev grabs \f(CW\*(C`SIGCHLD\*(C'\fR as soon as the default event loop is
1806initialised. This is necessary to guarantee proper behaviour even if
1807the first child watcher is started after the child exits. The occurrence
1808of \f(CW\*(C`SIGCHLD\*(C'\fR is recorded asynchronously, but child reaping is done
1809synchronously as part of the event loop processing. Libev always reaps all
1810children, even ones not watched.
1811.PP
1812\fIOverriding the Built-In Processing\fR
1813.IX Subsection "Overriding the Built-In Processing"
1814.PP
1815Libev offers no special support for overriding the built-in child
1816processing, but if your application collides with libev's default child
1817handler, you can override it easily by installing your own handler for
1818\&\f(CW\*(C`SIGCHLD\*(C'\fR after initialising the default loop, and making sure the
1819default loop never gets destroyed. You are encouraged, however, to use an
1820event-based approach to child reaping and thus use libev's support for
1821that, so other libev users can use \f(CW\*(C`ev_child\*(C'\fR watchers freely.
1822.PP
1823\fIStopping the Child Watcher\fR
1824.IX Subsection "Stopping the Child Watcher"
1825.PP
1826Currently, the child watcher never gets stopped, even when the
1827child terminates, so normally one needs to stop the watcher in the
1828callback. Future versions of libev might stop the watcher automatically
1829when a child exit is detected.
1830.PP
1831\fIWatcher-Specific Functions and Data Members\fR
1832.IX Subsection "Watcher-Specific Functions and Data Members"
877.IP "ev_child_init (ev_child *, callback, int pid)" 4 1833.IP "ev_child_init (ev_child *, callback, int pid, int trace)" 4
878.IX Item "ev_child_init (ev_child *, callback, int pid)" 1834.IX Item "ev_child_init (ev_child *, callback, int pid, int trace)"
879.PD 0 1835.PD 0
880.IP "ev_child_set (ev_child *, int pid)" 4 1836.IP "ev_child_set (ev_child *, int pid, int trace)" 4
881.IX Item "ev_child_set (ev_child *, int pid)" 1837.IX Item "ev_child_set (ev_child *, int pid, int trace)"
882.PD 1838.PD
883Configures the watcher to wait for status changes of process \f(CW\*(C`pid\*(C'\fR (or 1839Configures the watcher to wait for status changes of process \f(CW\*(C`pid\*(C'\fR (or
884\&\fIany\fR process if \f(CW\*(C`pid\*(C'\fR is specified as \f(CW0\fR). The callback can look 1840\&\fIany\fR process if \f(CW\*(C`pid\*(C'\fR is specified as \f(CW0\fR). The callback can look
885at the \f(CW\*(C`rstatus\*(C'\fR member of the \f(CW\*(C`ev_child\*(C'\fR watcher structure to see 1841at the \f(CW\*(C`rstatus\*(C'\fR member of the \f(CW\*(C`ev_child\*(C'\fR watcher structure to see
886the status word (use the macros from \f(CW\*(C`sys/wait.h\*(C'\fR and see your systems 1842the status word (use the macros from \f(CW\*(C`sys/wait.h\*(C'\fR and see your systems
887\&\f(CW\*(C`waitpid\*(C'\fR documentation). The \f(CW\*(C`rpid\*(C'\fR member contains the pid of the 1843\&\f(CW\*(C`waitpid\*(C'\fR documentation). The \f(CW\*(C`rpid\*(C'\fR member contains the pid of the
888process causing the status change. 1844process causing the status change. \f(CW\*(C`trace\*(C'\fR must be either \f(CW0\fR (only
1845activate the watcher when the process terminates) or \f(CW1\fR (additionally
1846activate the watcher when the process is stopped or continued).
1847.IP "int pid [read\-only]" 4
1848.IX Item "int pid [read-only]"
1849The process id this watcher watches out for, or \f(CW0\fR, meaning any process id.
1850.IP "int rpid [read\-write]" 4
1851.IX Item "int rpid [read-write]"
1852The process id that detected a status change.
1853.IP "int rstatus [read\-write]" 4
1854.IX Item "int rstatus [read-write]"
1855The process exit/trace status caused by \f(CW\*(C`rpid\*(C'\fR (see your systems
1856\&\f(CW\*(C`waitpid\*(C'\fR and \f(CW\*(C`sys/wait.h\*(C'\fR documentation for details).
1857.PP
1858\fIExamples\fR
1859.IX Subsection "Examples"
1860.PP
1861Example: \f(CW\*(C`fork()\*(C'\fR a new process and install a child handler to wait for
1862its completion.
1863.PP
1864.Vb 1
1865\& ev_child cw;
1866\&
1867\& static void
1868\& child_cb (EV_P_ struct ev_child *w, int revents)
1869\& {
1870\& ev_child_stop (EV_A_ w);
1871\& printf ("process %d exited with status %x\en", w\->rpid, w\->rstatus);
1872\& }
1873\&
1874\& pid_t pid = fork ();
1875\&
1876\& if (pid < 0)
1877\& // error
1878\& else if (pid == 0)
1879\& {
1880\& // the forked child executes here
1881\& exit (1);
1882\& }
1883\& else
1884\& {
1885\& ev_child_init (&cw, child_cb, pid, 0);
1886\& ev_child_start (EV_DEFAULT_ &cw);
1887\& }
1888.Ve
1889.ie n .Sh """ev_stat"" \- did the file attributes just change?"
1890.el .Sh "\f(CWev_stat\fP \- did the file attributes just change?"
1891.IX Subsection "ev_stat - did the file attributes just change?"
1892This watches a file system path for attribute changes. That is, it calls
1893\&\f(CW\*(C`stat\*(C'\fR regularly (or when the \s-1OS\s0 says it changed) and sees if it changed
1894compared to the last time, invoking the callback if it did.
1895.PP
1896The path does not need to exist: changing from \*(L"path exists\*(R" to \*(L"path does
1897not exist\*(R" is a status change like any other. The condition \*(L"path does
1898not exist\*(R" is signified by the \f(CW\*(C`st_nlink\*(C'\fR field being zero (which is
1899otherwise always forced to be at least one) and all the other fields of
1900the stat buffer having unspecified contents.
1901.PP
1902The path \fIshould\fR be absolute and \fImust not\fR end in a slash. If it is
1903relative and your working directory changes, the behaviour is undefined.
1904.PP
1905Since there is no standard kernel interface to do this, the portable
1906implementation simply calls \f(CW\*(C`stat (2)\*(C'\fR regularly on the path to see if
1907it changed somehow. You can specify a recommended polling interval for
1908this case. If you specify a polling interval of \f(CW0\fR (highly recommended!)
1909then a \fIsuitable, unspecified default\fR value will be used (which
1910you can expect to be around five seconds, although this might change
1911dynamically). Libev will also impose a minimum interval which is currently
1912around \f(CW0.1\fR, but thats usually overkill.
1913.PP
1914This watcher type is not meant for massive numbers of stat watchers,
1915as even with OS-supported change notifications, this can be
1916resource-intensive.
1917.PP
1918At the time of this writing, the only OS-specific interface implemented
1919is the Linux inotify interface (implementing kqueue support is left as
1920an exercise for the reader. Note, however, that the author sees no way
1921of implementing \f(CW\*(C`ev_stat\*(C'\fR semantics with kqueue).
1922.PP
1923\fI\s-1ABI\s0 Issues (Largefile Support)\fR
1924.IX Subsection "ABI Issues (Largefile Support)"
1925.PP
1926Libev by default (unless the user overrides this) uses the default
1927compilation environment, which means that on systems with large file
1928support disabled by default, you get the 32 bit version of the stat
1929structure. When using the library from programs that change the \s-1ABI\s0 to
1930use 64 bit file offsets the programs will fail. In that case you have to
1931compile libev with the same flags to get binary compatibility. This is
1932obviously the case with any flags that change the \s-1ABI\s0, but the problem is
1933most noticeably disabled with ev_stat and large file support.
1934.PP
1935The solution for this is to lobby your distribution maker to make large
1936file interfaces available by default (as e.g. FreeBSD does) and not
1937optional. Libev cannot simply switch on large file support because it has
1938to exchange stat structures with application programs compiled using the
1939default compilation environment.
1940.PP
1941\fIInotify and Kqueue\fR
1942.IX Subsection "Inotify and Kqueue"
1943.PP
1944When \f(CW\*(C`inotify (7)\*(C'\fR support has been compiled into libev (generally
1945only available with Linux 2.6.25 or above due to bugs in earlier
1946implementations) and present at runtime, it will be used to speed up
1947change detection where possible. The inotify descriptor will be created
1948lazily when the first \f(CW\*(C`ev_stat\*(C'\fR watcher is being started.
1949.PP
1950Inotify presence does not change the semantics of \f(CW\*(C`ev_stat\*(C'\fR watchers
1951except that changes might be detected earlier, and in some cases, to avoid
1952making regular \f(CW\*(C`stat\*(C'\fR calls. Even in the presence of inotify support
1953there are many cases where libev has to resort to regular \f(CW\*(C`stat\*(C'\fR polling,
1954but as long as the path exists, libev usually gets away without polling.
1955.PP
1956There is no support for kqueue, as apparently it cannot be used to
1957implement this functionality, due to the requirement of having a file
1958descriptor open on the object at all times, and detecting renames, unlinks
1959etc. is difficult.
1960.PP
1961\fIThe special problem of stat time resolution\fR
1962.IX Subsection "The special problem of stat time resolution"
1963.PP
1964The \f(CW\*(C`stat ()\*(C'\fR system call only supports full-second resolution portably, and
1965even on systems where the resolution is higher, most file systems still
1966only support whole seconds.
1967.PP
1968That means that, if the time is the only thing that changes, you can
1969easily miss updates: on the first update, \f(CW\*(C`ev_stat\*(C'\fR detects a change and
1970calls your callback, which does something. When there is another update
1971within the same second, \f(CW\*(C`ev_stat\*(C'\fR will be unable to detect unless the
1972stat data does change in other ways (e.g. file size).
1973.PP
1974The solution to this is to delay acting on a change for slightly more
1975than a second (or till slightly after the next full second boundary), using
1976a roughly one-second-delay \f(CW\*(C`ev_timer\*(C'\fR (e.g. \f(CW\*(C`ev_timer_set (w, 0., 1.02);
1977ev_timer_again (loop, w)\*(C'\fR).
1978.PP
1979The \f(CW.02\fR offset is added to work around small timing inconsistencies
1980of some operating systems (where the second counter of the current time
1981might be be delayed. One such system is the Linux kernel, where a call to
1982\&\f(CW\*(C`gettimeofday\*(C'\fR might return a timestamp with a full second later than
1983a subsequent \f(CW\*(C`time\*(C'\fR call \- if the equivalent of \f(CW\*(C`time ()\*(C'\fR is used to
1984update file times then there will be a small window where the kernel uses
1985the previous second to update file times but libev might already execute
1986the timer callback).
1987.PP
1988\fIWatcher-Specific Functions and Data Members\fR
1989.IX Subsection "Watcher-Specific Functions and Data Members"
1990.IP "ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)" 4
1991.IX Item "ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)"
1992.PD 0
1993.IP "ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)" 4
1994.IX Item "ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)"
1995.PD
1996Configures the watcher to wait for status changes of the given
1997\&\f(CW\*(C`path\*(C'\fR. The \f(CW\*(C`interval\*(C'\fR is a hint on how quickly a change is expected to
1998be detected and should normally be specified as \f(CW0\fR to let libev choose
1999a suitable value. The memory pointed to by \f(CW\*(C`path\*(C'\fR must point to the same
2000path for as long as the watcher is active.
2001.Sp
2002The callback will receive an \f(CW\*(C`EV_STAT\*(C'\fR event when a change was detected,
2003relative to the attributes at the time the watcher was started (or the
2004last change was detected).
2005.IP "ev_stat_stat (loop, ev_stat *)" 4
2006.IX Item "ev_stat_stat (loop, ev_stat *)"
2007Updates the stat buffer immediately with new values. If you change the
2008watched path in your callback, you could call this function to avoid
2009detecting this change (while introducing a race condition if you are not
2010the only one changing the path). Can also be useful simply to find out the
2011new values.
2012.IP "ev_statdata attr [read\-only]" 4
2013.IX Item "ev_statdata attr [read-only]"
2014The most-recently detected attributes of the file. Although the type is
2015\&\f(CW\*(C`ev_statdata\*(C'\fR, this is usually the (or one of the) \f(CW\*(C`struct stat\*(C'\fR types
2016suitable for your system, but you can only rely on the POSIX-standardised
2017members to be present. If the \f(CW\*(C`st_nlink\*(C'\fR member is \f(CW0\fR, then there was
2018some error while \f(CW\*(C`stat\*(C'\fRing the file.
2019.IP "ev_statdata prev [read\-only]" 4
2020.IX Item "ev_statdata prev [read-only]"
2021The previous attributes of the file. The callback gets invoked whenever
2022\&\f(CW\*(C`prev\*(C'\fR != \f(CW\*(C`attr\*(C'\fR, or, more precisely, one or more of these members
2023differ: \f(CW\*(C`st_dev\*(C'\fR, \f(CW\*(C`st_ino\*(C'\fR, \f(CW\*(C`st_mode\*(C'\fR, \f(CW\*(C`st_nlink\*(C'\fR, \f(CW\*(C`st_uid\*(C'\fR,
2024\&\f(CW\*(C`st_gid\*(C'\fR, \f(CW\*(C`st_rdev\*(C'\fR, \f(CW\*(C`st_size\*(C'\fR, \f(CW\*(C`st_atime\*(C'\fR, \f(CW\*(C`st_mtime\*(C'\fR, \f(CW\*(C`st_ctime\*(C'\fR.
2025.IP "ev_tstamp interval [read\-only]" 4
2026.IX Item "ev_tstamp interval [read-only]"
2027The specified interval.
2028.IP "const char *path [read\-only]" 4
2029.IX Item "const char *path [read-only]"
2030The file system path that is being watched.
2031.PP
2032\fIExamples\fR
2033.IX Subsection "Examples"
2034.PP
2035Example: Watch \f(CW\*(C`/etc/passwd\*(C'\fR for attribute changes.
2036.PP
2037.Vb 10
2038\& static void
2039\& passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
2040\& {
2041\& /* /etc/passwd changed in some way */
2042\& if (w\->attr.st_nlink)
2043\& {
2044\& printf ("passwd current size %ld\en", (long)w\->attr.st_size);
2045\& printf ("passwd current atime %ld\en", (long)w\->attr.st_mtime);
2046\& printf ("passwd current mtime %ld\en", (long)w\->attr.st_mtime);
2047\& }
2048\& else
2049\& /* you shalt not abuse printf for puts */
2050\& puts ("wow, /etc/passwd is not there, expect problems. "
2051\& "if this is windows, they already arrived\en");
2052\& }
2053\&
2054\& ...
2055\& ev_stat passwd;
2056\&
2057\& ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
2058\& ev_stat_start (loop, &passwd);
2059.Ve
2060.PP
2061Example: Like above, but additionally use a one-second delay so we do not
2062miss updates (however, frequent updates will delay processing, too, so
2063one might do the work both on \f(CW\*(C`ev_stat\*(C'\fR callback invocation \fIand\fR on
2064\&\f(CW\*(C`ev_timer\*(C'\fR callback invocation).
2065.PP
2066.Vb 2
2067\& static ev_stat passwd;
2068\& static ev_timer timer;
2069\&
2070\& static void
2071\& timer_cb (EV_P_ ev_timer *w, int revents)
2072\& {
2073\& ev_timer_stop (EV_A_ w);
2074\&
2075\& /* now it\*(Aqs one second after the most recent passwd change */
2076\& }
2077\&
2078\& static void
2079\& stat_cb (EV_P_ ev_stat *w, int revents)
2080\& {
2081\& /* reset the one\-second timer */
2082\& ev_timer_again (EV_A_ &timer);
2083\& }
2084\&
2085\& ...
2086\& ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
2087\& ev_stat_start (loop, &passwd);
2088\& ev_timer_init (&timer, timer_cb, 0., 1.02);
2089.Ve
889.ie n .Sh """ev_idle"" \- when you've got nothing better to do" 2090.ie n .Sh """ev_idle"" \- when you've got nothing better to do..."
890.el .Sh "\f(CWev_idle\fP \- when you've got nothing better to do" 2091.el .Sh "\f(CWev_idle\fP \- when you've got nothing better to do..."
891.IX Subsection "ev_idle - when you've got nothing better to do" 2092.IX Subsection "ev_idle - when you've got nothing better to do..."
892Idle watchers trigger events when there are no other events are pending 2093Idle watchers trigger events when no other events of the same or higher
893(prepare, check and other idle watchers do not count). That is, as long 2094priority are pending (prepare, check and other idle watchers do not count
894as your process is busy handling sockets or timeouts (or even signals, 2095as receiving \*(L"events\*(R").
895imagine) it will not be triggered. But when your process is idle all idle 2096.PP
896watchers are being called again and again, once per event loop iteration \- 2097That is, as long as your process is busy handling sockets or timeouts
2098(or even signals, imagine) of the same or higher priority it will not be
2099triggered. But when your process is idle (or only lower-priority watchers
2100are pending), the idle watchers are being called once per event loop
897until stopped, that is, or your process receives more events and becomes 2101iteration \- until stopped, that is, or your process receives more events
898busy. 2102and becomes busy again with higher priority stuff.
899.PP 2103.PP
900The most noteworthy effect is that as long as any idle watchers are 2104The most noteworthy effect is that as long as any idle watchers are
901active, the process will not block when waiting for new events. 2105active, the process will not block when waiting for new events.
902.PP 2106.PP
903Apart from keeping your process non-blocking (which is a useful 2107Apart from keeping your process non-blocking (which is a useful
904effect on its own sometimes), idle watchers are a good place to do 2108effect on its own sometimes), idle watchers are a good place to do
905\&\*(L"pseudo\-background processing\*(R", or delay processing stuff to after the 2109\&\*(L"pseudo-background processing\*(R", or delay processing stuff to after the
906event loop has handled all outstanding events. 2110event loop has handled all outstanding events.
2111.PP
2112\fIWatcher-Specific Functions and Data Members\fR
2113.IX Subsection "Watcher-Specific Functions and Data Members"
907.IP "ev_idle_init (ev_signal *, callback)" 4 2114.IP "ev_idle_init (ev_signal *, callback)" 4
908.IX Item "ev_idle_init (ev_signal *, callback)" 2115.IX Item "ev_idle_init (ev_signal *, callback)"
909Initialises and configures the idle watcher \- it has no parameters of any 2116Initialises and configures the idle watcher \- it has no parameters of any
910kind. There is a \f(CW\*(C`ev_idle_set\*(C'\fR macro, but using it is utterly pointless, 2117kind. There is a \f(CW\*(C`ev_idle_set\*(C'\fR macro, but using it is utterly pointless,
911believe me. 2118believe me.
2119.PP
2120\fIExamples\fR
2121.IX Subsection "Examples"
2122.PP
2123Example: Dynamically allocate an \f(CW\*(C`ev_idle\*(C'\fR watcher, start it, and in the
2124callback, free it. Also, use no error checking, as usual.
2125.PP
2126.Vb 7
2127\& static void
2128\& idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
2129\& {
2130\& free (w);
2131\& // now do something you wanted to do when the program has
2132\& // no longer anything immediate to do.
2133\& }
2134\&
2135\& struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
2136\& ev_idle_init (idle_watcher, idle_cb);
2137\& ev_idle_start (loop, idle_cb);
2138.Ve
912.ie n .Sh """ev_prepare""\fP and \f(CW""ev_check"" \- customise your event loop" 2139.ie n .Sh """ev_prepare""\fP and \f(CW""ev_check"" \- customise your event loop!"
913.el .Sh "\f(CWev_prepare\fP and \f(CWev_check\fP \- customise your event loop" 2140.el .Sh "\f(CWev_prepare\fP and \f(CWev_check\fP \- customise your event loop!"
914.IX Subsection "ev_prepare and ev_check - customise your event loop" 2141.IX Subsection "ev_prepare and ev_check - customise your event loop!"
915Prepare and check watchers are usually (but not always) used in tandem: 2142Prepare and check watchers are usually (but not always) used in pairs:
916prepare watchers get invoked before the process blocks and check watchers 2143prepare watchers get invoked before the process blocks and check watchers
917afterwards. 2144afterwards.
918.PP 2145.PP
2146You \fImust not\fR call \f(CW\*(C`ev_loop\*(C'\fR or similar functions that enter
2147the current event loop from either \f(CW\*(C`ev_prepare\*(C'\fR or \f(CW\*(C`ev_check\*(C'\fR
2148watchers. Other loops than the current one are fine, however. The
2149rationale behind this is that you do not need to check for recursion in
2150those watchers, i.e. the sequence will always be \f(CW\*(C`ev_prepare\*(C'\fR, blocking,
2151\&\f(CW\*(C`ev_check\*(C'\fR so if you have one watcher of each kind they will always be
2152called in pairs bracketing the blocking call.
2153.PP
919Their main purpose is to integrate other event mechanisms into libev. This 2154Their main purpose is to integrate other event mechanisms into libev and
920could be used, for example, to track variable changes, implement your own 2155their use is somewhat advanced. They could be used, for example, to track
921watchers, integrate net-snmp or a coroutine library and lots more. 2156variable changes, implement your own watchers, integrate net-snmp or a
2157coroutine library and lots more. They are also occasionally useful if
2158you cache some data and want to flush it before blocking (for example,
2159in X programs you might want to do an \f(CW\*(C`XFlush ()\*(C'\fR in an \f(CW\*(C`ev_prepare\*(C'\fR
2160watcher).
922.PP 2161.PP
923This is done by examining in each prepare call which file descriptors need 2162This is done by examining in each prepare call which file descriptors
924to be watched by the other library, registering \f(CW\*(C`ev_io\*(C'\fR watchers for 2163need to be watched by the other library, registering \f(CW\*(C`ev_io\*(C'\fR watchers
925them and starting an \f(CW\*(C`ev_timer\*(C'\fR watcher for any timeouts (many libraries 2164for them and starting an \f(CW\*(C`ev_timer\*(C'\fR watcher for any timeouts (many
926provide just this functionality). Then, in the check watcher you check for 2165libraries provide exactly this functionality). Then, in the check watcher,
927any events that occured (by checking the pending status of all watchers 2166you check for any events that occurred (by checking the pending status
928and stopping them) and call back into the library. The I/O and timer 2167of all watchers and stopping them) and call back into the library. The
929callbacks will never actually be called (but must be valid nevertheless, 2168I/O and timer callbacks will never actually be called (but must be valid
930because you never know, you know?). 2169nevertheless, because you never know, you know?).
931.PP 2170.PP
932As another example, the Perl Coro module uses these hooks to integrate 2171As another example, the Perl Coro module uses these hooks to integrate
933coroutines into libev programs, by yielding to other active coroutines 2172coroutines into libev programs, by yielding to other active coroutines
934during each prepare and only letting the process block if no coroutines 2173during each prepare and only letting the process block if no coroutines
935are ready to run (it's actually more complicated: it only runs coroutines 2174are ready to run (it's actually more complicated: it only runs coroutines
936with priority higher than or equal to the event loop and one coroutine 2175with priority higher than or equal to the event loop and one coroutine
937of lower priority, but only once, using idle watchers to keep the event 2176of lower priority, but only once, using idle watchers to keep the event
938loop from blocking if lower-priority coroutines are active, thus mapping 2177loop from blocking if lower-priority coroutines are active, thus mapping
939low-priority coroutines to idle/background tasks). 2178low-priority coroutines to idle/background tasks).
2179.PP
2180It is recommended to give \f(CW\*(C`ev_check\*(C'\fR watchers highest (\f(CW\*(C`EV_MAXPRI\*(C'\fR)
2181priority, to ensure that they are being run before any other watchers
2182after the poll (this doesn't matter for \f(CW\*(C`ev_prepare\*(C'\fR watchers).
2183.PP
2184Also, \f(CW\*(C`ev_check\*(C'\fR watchers (and \f(CW\*(C`ev_prepare\*(C'\fR watchers, too) should not
2185activate (\*(L"feed\*(R") events into libev. While libev fully supports this, they
2186might get executed before other \f(CW\*(C`ev_check\*(C'\fR watchers did their job. As
2187\&\f(CW\*(C`ev_check\*(C'\fR watchers are often used to embed other (non-libev) event
2188loops those other event loops might be in an unusable state until their
2189\&\f(CW\*(C`ev_check\*(C'\fR watcher ran (always remind yourself to coexist peacefully with
2190others).
2191.PP
2192\fIWatcher-Specific Functions and Data Members\fR
2193.IX Subsection "Watcher-Specific Functions and Data Members"
940.IP "ev_prepare_init (ev_prepare *, callback)" 4 2194.IP "ev_prepare_init (ev_prepare *, callback)" 4
941.IX Item "ev_prepare_init (ev_prepare *, callback)" 2195.IX Item "ev_prepare_init (ev_prepare *, callback)"
942.PD 0 2196.PD 0
943.IP "ev_check_init (ev_check *, callback)" 4 2197.IP "ev_check_init (ev_check *, callback)" 4
944.IX Item "ev_check_init (ev_check *, callback)" 2198.IX Item "ev_check_init (ev_check *, callback)"
945.PD 2199.PD
946Initialises and configures the prepare or check watcher \- they have no 2200Initialises and configures the prepare or check watcher \- they have no
947parameters of any kind. There are \f(CW\*(C`ev_prepare_set\*(C'\fR and \f(CW\*(C`ev_check_set\*(C'\fR 2201parameters of any kind. There are \f(CW\*(C`ev_prepare_set\*(C'\fR and \f(CW\*(C`ev_check_set\*(C'\fR
948macros, but using them is utterly, utterly and completely pointless. 2202macros, but using them is utterly, utterly, utterly and completely
2203pointless.
2204.PP
2205\fIExamples\fR
2206.IX Subsection "Examples"
2207.PP
2208There are a number of principal ways to embed other event loops or modules
2209into libev. Here are some ideas on how to include libadns into libev
2210(there is a Perl module named \f(CW\*(C`EV::ADNS\*(C'\fR that does this, which you could
2211use as a working example. Another Perl module named \f(CW\*(C`EV::Glib\*(C'\fR embeds a
2212Glib main context into libev, and finally, \f(CW\*(C`Glib::EV\*(C'\fR embeds \s-1EV\s0 into the
2213Glib event loop).
2214.PP
2215Method 1: Add \s-1IO\s0 watchers and a timeout watcher in a prepare handler,
2216and in a check watcher, destroy them and call into libadns. What follows
2217is pseudo-code only of course. This requires you to either use a low
2218priority for the check watcher or use \f(CW\*(C`ev_clear_pending\*(C'\fR explicitly, as
2219the callbacks for the IO/timeout watchers might not have been called yet.
2220.PP
2221.Vb 2
2222\& static ev_io iow [nfd];
2223\& static ev_timer tw;
2224\&
2225\& static void
2226\& io_cb (ev_loop *loop, ev_io *w, int revents)
2227\& {
2228\& }
2229\&
2230\& // create io watchers for each fd and a timer before blocking
2231\& static void
2232\& adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
2233\& {
2234\& int timeout = 3600000;
2235\& struct pollfd fds [nfd];
2236\& // actual code will need to loop here and realloc etc.
2237\& adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
2238\&
2239\& /* the callback is illegal, but won\*(Aqt be called as we stop during check */
2240\& ev_timer_init (&tw, 0, timeout * 1e\-3);
2241\& ev_timer_start (loop, &tw);
2242\&
2243\& // create one ev_io per pollfd
2244\& for (int i = 0; i < nfd; ++i)
2245\& {
2246\& ev_io_init (iow + i, io_cb, fds [i].fd,
2247\& ((fds [i].events & POLLIN ? EV_READ : 0)
2248\& | (fds [i].events & POLLOUT ? EV_WRITE : 0)));
2249\&
2250\& fds [i].revents = 0;
2251\& ev_io_start (loop, iow + i);
2252\& }
2253\& }
2254\&
2255\& // stop all watchers after blocking
2256\& static void
2257\& adns_check_cb (ev_loop *loop, ev_check *w, int revents)
2258\& {
2259\& ev_timer_stop (loop, &tw);
2260\&
2261\& for (int i = 0; i < nfd; ++i)
2262\& {
2263\& // set the relevant poll flags
2264\& // could also call adns_processreadable etc. here
2265\& struct pollfd *fd = fds + i;
2266\& int revents = ev_clear_pending (iow + i);
2267\& if (revents & EV_READ ) fd\->revents |= fd\->events & POLLIN;
2268\& if (revents & EV_WRITE) fd\->revents |= fd\->events & POLLOUT;
2269\&
2270\& // now stop the watcher
2271\& ev_io_stop (loop, iow + i);
2272\& }
2273\&
2274\& adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
2275\& }
2276.Ve
2277.PP
2278Method 2: This would be just like method 1, but you run \f(CW\*(C`adns_afterpoll\*(C'\fR
2279in the prepare watcher and would dispose of the check watcher.
2280.PP
2281Method 3: If the module to be embedded supports explicit event
2282notification (libadns does), you can also make use of the actual watcher
2283callbacks, and only destroy/create the watchers in the prepare watcher.
2284.PP
2285.Vb 5
2286\& static void
2287\& timer_cb (EV_P_ ev_timer *w, int revents)
2288\& {
2289\& adns_state ads = (adns_state)w\->data;
2290\& update_now (EV_A);
2291\&
2292\& adns_processtimeouts (ads, &tv_now);
2293\& }
2294\&
2295\& static void
2296\& io_cb (EV_P_ ev_io *w, int revents)
2297\& {
2298\& adns_state ads = (adns_state)w\->data;
2299\& update_now (EV_A);
2300\&
2301\& if (revents & EV_READ ) adns_processreadable (ads, w\->fd, &tv_now);
2302\& if (revents & EV_WRITE) adns_processwriteable (ads, w\->fd, &tv_now);
2303\& }
2304\&
2305\& // do not ever call adns_afterpoll
2306.Ve
2307.PP
2308Method 4: Do not use a prepare or check watcher because the module you
2309want to embed is not flexible enough to support it. Instead, you can
2310override their poll function. The drawback with this solution is that the
2311main loop is now no longer controllable by \s-1EV\s0. The \f(CW\*(C`Glib::EV\*(C'\fR module uses
2312this approach, effectively embedding \s-1EV\s0 as a client into the horrible
2313libglib event loop.
2314.PP
2315.Vb 4
2316\& static gint
2317\& event_poll_func (GPollFD *fds, guint nfds, gint timeout)
2318\& {
2319\& int got_events = 0;
2320\&
2321\& for (n = 0; n < nfds; ++n)
2322\& // create/start io watcher that sets the relevant bits in fds[n] and increment got_events
2323\&
2324\& if (timeout >= 0)
2325\& // create/start timer
2326\&
2327\& // poll
2328\& ev_loop (EV_A_ 0);
2329\&
2330\& // stop timer again
2331\& if (timeout >= 0)
2332\& ev_timer_stop (EV_A_ &to);
2333\&
2334\& // stop io watchers again \- their callbacks should have set
2335\& for (n = 0; n < nfds; ++n)
2336\& ev_io_stop (EV_A_ iow [n]);
2337\&
2338\& return got_events;
2339\& }
2340.Ve
2341.ie n .Sh """ev_embed"" \- when one backend isn't enough..."
2342.el .Sh "\f(CWev_embed\fP \- when one backend isn't enough..."
2343.IX Subsection "ev_embed - when one backend isn't enough..."
2344This is a rather advanced watcher type that lets you embed one event loop
2345into another (currently only \f(CW\*(C`ev_io\*(C'\fR events are supported in the embedded
2346loop, other types of watchers might be handled in a delayed or incorrect
2347fashion and must not be used).
2348.PP
2349There are primarily two reasons you would want that: work around bugs and
2350prioritise I/O.
2351.PP
2352As an example for a bug workaround, the kqueue backend might only support
2353sockets on some platform, so it is unusable as generic backend, but you
2354still want to make use of it because you have many sockets and it scales
2355so nicely. In this case, you would create a kqueue-based loop and embed
2356it into your default loop (which might use e.g. poll). Overall operation
2357will be a bit slower because first libev has to call \f(CW\*(C`poll\*(C'\fR and then
2358\&\f(CW\*(C`kevent\*(C'\fR, but at least you can use both mechanisms for what they are
2359best: \f(CW\*(C`kqueue\*(C'\fR for scalable sockets and \f(CW\*(C`poll\*(C'\fR if you want it to work :)
2360.PP
2361As for prioritising I/O: under rare circumstances you have the case where
2362some fds have to be watched and handled very quickly (with low latency),
2363and even priorities and idle watchers might have too much overhead. In
2364this case you would put all the high priority stuff in one loop and all
2365the rest in a second one, and embed the second one in the first.
2366.PP
2367As long as the watcher is active, the callback will be invoked every time
2368there might be events pending in the embedded loop. The callback must then
2369call \f(CW\*(C`ev_embed_sweep (mainloop, watcher)\*(C'\fR to make a single sweep and invoke
2370their callbacks (you could also start an idle watcher to give the embedded
2371loop strictly lower priority for example). You can also set the callback
2372to \f(CW0\fR, in which case the embed watcher will automatically execute the
2373embedded loop sweep.
2374.PP
2375As long as the watcher is started it will automatically handle events. The
2376callback will be invoked whenever some events have been handled. You can
2377set the callback to \f(CW0\fR to avoid having to specify one if you are not
2378interested in that.
2379.PP
2380Also, there have not currently been made special provisions for forking:
2381when you fork, you not only have to call \f(CW\*(C`ev_loop_fork\*(C'\fR on both loops,
2382but you will also have to stop and restart any \f(CW\*(C`ev_embed\*(C'\fR watchers
2383yourself \- but you can use a fork watcher to handle this automatically,
2384and future versions of libev might do just that.
2385.PP
2386Unfortunately, not all backends are embeddable: only the ones returned by
2387\&\f(CW\*(C`ev_embeddable_backends\*(C'\fR are, which, unfortunately, does not include any
2388portable one.
2389.PP
2390So when you want to use this feature you will always have to be prepared
2391that you cannot get an embeddable loop. The recommended way to get around
2392this is to have a separate variables for your embeddable loop, try to
2393create it, and if that fails, use the normal loop for everything.
2394.PP
2395\fI\f(CI\*(C`ev_embed\*(C'\fI and fork\fR
2396.IX Subsection "ev_embed and fork"
2397.PP
2398While the \f(CW\*(C`ev_embed\*(C'\fR watcher is running, forks in the embedding loop will
2399automatically be applied to the embedded loop as well, so no special
2400fork handling is required in that case. When the watcher is not running,
2401however, it is still the task of the libev user to call \f(CW\*(C`ev_loop_fork ()\*(C'\fR
2402as applicable.
2403.PP
2404\fIWatcher-Specific Functions and Data Members\fR
2405.IX Subsection "Watcher-Specific Functions and Data Members"
2406.IP "ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)" 4
2407.IX Item "ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)"
2408.PD 0
2409.IP "ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)" 4
2410.IX Item "ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)"
2411.PD
2412Configures the watcher to embed the given loop, which must be
2413embeddable. If the callback is \f(CW0\fR, then \f(CW\*(C`ev_embed_sweep\*(C'\fR will be
2414invoked automatically, otherwise it is the responsibility of the callback
2415to invoke it (it will continue to be called until the sweep has been done,
2416if you do not want that, you need to temporarily stop the embed watcher).
2417.IP "ev_embed_sweep (loop, ev_embed *)" 4
2418.IX Item "ev_embed_sweep (loop, ev_embed *)"
2419Make a single, non-blocking sweep over the embedded loop. This works
2420similarly to \f(CW\*(C`ev_loop (embedded_loop, EVLOOP_NONBLOCK)\*(C'\fR, but in the most
2421appropriate way for embedded loops.
2422.IP "struct ev_loop *other [read\-only]" 4
2423.IX Item "struct ev_loop *other [read-only]"
2424The embedded event loop.
2425.PP
2426\fIExamples\fR
2427.IX Subsection "Examples"
2428.PP
2429Example: Try to get an embeddable event loop and embed it into the default
2430event loop. If that is not possible, use the default loop. The default
2431loop is stored in \f(CW\*(C`loop_hi\*(C'\fR, while the embeddable loop is stored in
2432\&\f(CW\*(C`loop_lo\*(C'\fR (which is \f(CW\*(C`loop_hi\*(C'\fR in the case no embeddable loop can be
2433used).
2434.PP
2435.Vb 3
2436\& struct ev_loop *loop_hi = ev_default_init (0);
2437\& struct ev_loop *loop_lo = 0;
2438\& struct ev_embed embed;
2439\&
2440\& // see if there is a chance of getting one that works
2441\& // (remember that a flags value of 0 means autodetection)
2442\& loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2443\& ? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
2444\& : 0;
2445\&
2446\& // if we got one, then embed it, otherwise default to loop_hi
2447\& if (loop_lo)
2448\& {
2449\& ev_embed_init (&embed, 0, loop_lo);
2450\& ev_embed_start (loop_hi, &embed);
2451\& }
2452\& else
2453\& loop_lo = loop_hi;
2454.Ve
2455.PP
2456Example: Check if kqueue is available but not recommended and create
2457a kqueue backend for use with sockets (which usually work with any
2458kqueue implementation). Store the kqueue/socket\-only event loop in
2459\&\f(CW\*(C`loop_socket\*(C'\fR. (One might optionally use \f(CW\*(C`EVFLAG_NOENV\*(C'\fR, too).
2460.PP
2461.Vb 3
2462\& struct ev_loop *loop = ev_default_init (0);
2463\& struct ev_loop *loop_socket = 0;
2464\& struct ev_embed embed;
2465\&
2466\& if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2467\& if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2468\& {
2469\& ev_embed_init (&embed, 0, loop_socket);
2470\& ev_embed_start (loop, &embed);
2471\& }
2472\&
2473\& if (!loop_socket)
2474\& loop_socket = loop;
2475\&
2476\& // now use loop_socket for all sockets, and loop for everything else
2477.Ve
2478.ie n .Sh """ev_fork"" \- the audacity to resume the event loop after a fork"
2479.el .Sh "\f(CWev_fork\fP \- the audacity to resume the event loop after a fork"
2480.IX Subsection "ev_fork - the audacity to resume the event loop after a fork"
2481Fork watchers are called when a \f(CW\*(C`fork ()\*(C'\fR was detected (usually because
2482whoever is a good citizen cared to tell libev about it by calling
2483\&\f(CW\*(C`ev_default_fork\*(C'\fR or \f(CW\*(C`ev_loop_fork\*(C'\fR). The invocation is done before the
2484event loop blocks next and before \f(CW\*(C`ev_check\*(C'\fR watchers are being called,
2485and only in the child after the fork. If whoever good citizen calling
2486\&\f(CW\*(C`ev_default_fork\*(C'\fR cheats and calls it in the wrong process, the fork
2487handlers will be invoked, too, of course.
2488.PP
2489\fIWatcher-Specific Functions and Data Members\fR
2490.IX Subsection "Watcher-Specific Functions and Data Members"
2491.IP "ev_fork_init (ev_signal *, callback)" 4
2492.IX Item "ev_fork_init (ev_signal *, callback)"
2493Initialises and configures the fork watcher \- it has no parameters of any
2494kind. There is a \f(CW\*(C`ev_fork_set\*(C'\fR macro, but using it is utterly pointless,
2495believe me.
2496.ie n .Sh """ev_async"" \- how to wake up another event loop"
2497.el .Sh "\f(CWev_async\fP \- how to wake up another event loop"
2498.IX Subsection "ev_async - how to wake up another event loop"
2499In general, you cannot use an \f(CW\*(C`ev_loop\*(C'\fR from multiple threads or other
2500asynchronous sources such as signal handlers (as opposed to multiple event
2501loops \- those are of course safe to use in different threads).
2502.PP
2503Sometimes, however, you need to wake up another event loop you do not
2504control, for example because it belongs to another thread. This is what
2505\&\f(CW\*(C`ev_async\*(C'\fR watchers do: as long as the \f(CW\*(C`ev_async\*(C'\fR watcher is active, you
2506can signal it by calling \f(CW\*(C`ev_async_send\*(C'\fR, which is thread\- and signal
2507safe.
2508.PP
2509This functionality is very similar to \f(CW\*(C`ev_signal\*(C'\fR watchers, as signals,
2510too, are asynchronous in nature, and signals, too, will be compressed
2511(i.e. the number of callback invocations may be less than the number of
2512\&\f(CW\*(C`ev_async_sent\*(C'\fR calls).
2513.PP
2514Unlike \f(CW\*(C`ev_signal\*(C'\fR watchers, \f(CW\*(C`ev_async\*(C'\fR works with any event loop, not
2515just the default loop.
2516.PP
2517\fIQueueing\fR
2518.IX Subsection "Queueing"
2519.PP
2520\&\f(CW\*(C`ev_async\*(C'\fR does not support queueing of data in any way. The reason
2521is that the author does not know of a simple (or any) algorithm for a
2522multiple-writer-single-reader queue that works in all cases and doesn't
2523need elaborate support such as pthreads.
2524.PP
2525That means that if you want to queue data, you have to provide your own
2526queue. But at least I can tell you how to implement locking around your
2527queue:
2528.IP "queueing from a signal handler context" 4
2529.IX Item "queueing from a signal handler context"
2530To implement race-free queueing, you simply add to the queue in the signal
2531handler but you block the signal handler in the watcher callback. Here is
2532an example that does that for some fictitious \s-1SIGUSR1\s0 handler:
2533.Sp
2534.Vb 1
2535\& static ev_async mysig;
2536\&
2537\& static void
2538\& sigusr1_handler (void)
2539\& {
2540\& sometype data;
2541\&
2542\& // no locking etc.
2543\& queue_put (data);
2544\& ev_async_send (EV_DEFAULT_ &mysig);
2545\& }
2546\&
2547\& static void
2548\& mysig_cb (EV_P_ ev_async *w, int revents)
2549\& {
2550\& sometype data;
2551\& sigset_t block, prev;
2552\&
2553\& sigemptyset (&block);
2554\& sigaddset (&block, SIGUSR1);
2555\& sigprocmask (SIG_BLOCK, &block, &prev);
2556\&
2557\& while (queue_get (&data))
2558\& process (data);
2559\&
2560\& if (sigismember (&prev, SIGUSR1)
2561\& sigprocmask (SIG_UNBLOCK, &block, 0);
2562\& }
2563.Ve
2564.Sp
2565(Note: pthreads in theory requires you to use \f(CW\*(C`pthread_setmask\*(C'\fR
2566instead of \f(CW\*(C`sigprocmask\*(C'\fR when you use threads, but libev doesn't do it
2567either...).
2568.IP "queueing from a thread context" 4
2569.IX Item "queueing from a thread context"
2570The strategy for threads is different, as you cannot (easily) block
2571threads but you can easily preempt them, so to queue safely you need to
2572employ a traditional mutex lock, such as in this pthread example:
2573.Sp
2574.Vb 2
2575\& static ev_async mysig;
2576\& static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
2577\&
2578\& static void
2579\& otherthread (void)
2580\& {
2581\& // only need to lock the actual queueing operation
2582\& pthread_mutex_lock (&mymutex);
2583\& queue_put (data);
2584\& pthread_mutex_unlock (&mymutex);
2585\&
2586\& ev_async_send (EV_DEFAULT_ &mysig);
2587\& }
2588\&
2589\& static void
2590\& mysig_cb (EV_P_ ev_async *w, int revents)
2591\& {
2592\& pthread_mutex_lock (&mymutex);
2593\&
2594\& while (queue_get (&data))
2595\& process (data);
2596\&
2597\& pthread_mutex_unlock (&mymutex);
2598\& }
2599.Ve
2600.PP
2601\fIWatcher-Specific Functions and Data Members\fR
2602.IX Subsection "Watcher-Specific Functions and Data Members"
2603.IP "ev_async_init (ev_async *, callback)" 4
2604.IX Item "ev_async_init (ev_async *, callback)"
2605Initialises and configures the async watcher \- it has no parameters of any
2606kind. There is a \f(CW\*(C`ev_asynd_set\*(C'\fR macro, but using it is utterly pointless,
2607trust me.
2608.IP "ev_async_send (loop, ev_async *)" 4
2609.IX Item "ev_async_send (loop, ev_async *)"
2610Sends/signals/activates the given \f(CW\*(C`ev_async\*(C'\fR watcher, that is, feeds
2611an \f(CW\*(C`EV_ASYNC\*(C'\fR event on the watcher into the event loop. Unlike
2612\&\f(CW\*(C`ev_feed_event\*(C'\fR, this call is safe to do from other threads, signal or
2613similar contexts (see the discussion of \f(CW\*(C`EV_ATOMIC_T\*(C'\fR in the embedding
2614section below on what exactly this means).
2615.Sp
2616This call incurs the overhead of a system call only once per loop iteration,
2617so while the overhead might be noticeable, it doesn't apply to repeated
2618calls to \f(CW\*(C`ev_async_send\*(C'\fR.
2619.IP "bool = ev_async_pending (ev_async *)" 4
2620.IX Item "bool = ev_async_pending (ev_async *)"
2621Returns a non-zero value when \f(CW\*(C`ev_async_send\*(C'\fR has been called on the
2622watcher but the event has not yet been processed (or even noted) by the
2623event loop.
2624.Sp
2625\&\f(CW\*(C`ev_async_send\*(C'\fR sets a flag in the watcher and wakes up the loop. When
2626the loop iterates next and checks for the watcher to have become active,
2627it will reset the flag again. \f(CW\*(C`ev_async_pending\*(C'\fR can be used to very
2628quickly check whether invoking the loop might be a good idea.
2629.Sp
2630Not that this does \fInot\fR check whether the watcher itself is pending, only
2631whether it has been requested to make this watcher pending.
949.SH "OTHER FUNCTIONS" 2632.SH "OTHER FUNCTIONS"
950.IX Header "OTHER FUNCTIONS" 2633.IX Header "OTHER FUNCTIONS"
951There are some other functions of possible interest. Described. Here. Now. 2634There are some other functions of possible interest. Described. Here. Now.
952.IP "ev_once (loop, int fd, int events, ev_tstamp timeout, callback)" 4 2635.IP "ev_once (loop, int fd, int events, ev_tstamp timeout, callback)" 4
953.IX Item "ev_once (loop, int fd, int events, ev_tstamp timeout, callback)" 2636.IX Item "ev_once (loop, int fd, int events, ev_tstamp timeout, callback)"
954This function combines a simple timer and an I/O watcher, calls your 2637This function combines a simple timer and an I/O watcher, calls your
955callback on whichever event happens first and automatically stop both 2638callback on whichever event happens first and automatically stops both
956watchers. This is useful if you want to wait for a single event on an fd 2639watchers. This is useful if you want to wait for a single event on an fd
957or timeout without having to allocate/configure/start/stop/free one or 2640or timeout without having to allocate/configure/start/stop/free one or
958more watchers yourself. 2641more watchers yourself.
959.Sp 2642.Sp
960If \f(CW\*(C`fd\*(C'\fR is less than 0, then no I/O watcher will be started and events 2643If \f(CW\*(C`fd\*(C'\fR is less than 0, then no I/O watcher will be started and the
961is being ignored. Otherwise, an \f(CW\*(C`ev_io\*(C'\fR watcher for the given \f(CW\*(C`fd\*(C'\fR and 2644\&\f(CW\*(C`events\*(C'\fR argument is being ignored. Otherwise, an \f(CW\*(C`ev_io\*(C'\fR watcher for
962\&\f(CW\*(C`events\*(C'\fR set will be craeted and started. 2645the given \f(CW\*(C`fd\*(C'\fR and \f(CW\*(C`events\*(C'\fR set will be created and started.
963.Sp 2646.Sp
964If \f(CW\*(C`timeout\*(C'\fR is less than 0, then no timeout watcher will be 2647If \f(CW\*(C`timeout\*(C'\fR is less than 0, then no timeout watcher will be
965started. Otherwise an \f(CW\*(C`ev_timer\*(C'\fR watcher with after = \f(CW\*(C`timeout\*(C'\fR (and 2648started. Otherwise an \f(CW\*(C`ev_timer\*(C'\fR watcher with after = \f(CW\*(C`timeout\*(C'\fR (and
966repeat = 0) will be started. While \f(CW0\fR is a valid timeout, it is of 2649repeat = 0) will be started. \f(CW0\fR is a valid timeout.
967dubious value.
968.Sp 2650.Sp
969The callback has the type \f(CW\*(C`void (*cb)(int revents, void *arg)\*(C'\fR and gets 2651The callback has the type \f(CW\*(C`void (*cb)(int revents, void *arg)\*(C'\fR and gets
970passed an \f(CW\*(C`revents\*(C'\fR set like normal event callbacks (a combination of 2652passed an \f(CW\*(C`revents\*(C'\fR set like normal event callbacks (a combination of
971\&\f(CW\*(C`EV_ERROR\*(C'\fR, \f(CW\*(C`EV_READ\*(C'\fR, \f(CW\*(C`EV_WRITE\*(C'\fR or \f(CW\*(C`EV_TIMEOUT\*(C'\fR) and the \f(CW\*(C`arg\*(C'\fR 2653\&\f(CW\*(C`EV_ERROR\*(C'\fR, \f(CW\*(C`EV_READ\*(C'\fR, \f(CW\*(C`EV_WRITE\*(C'\fR or \f(CW\*(C`EV_TIMEOUT\*(C'\fR) and the \f(CW\*(C`arg\*(C'\fR
972value passed to \f(CW\*(C`ev_once\*(C'\fR: 2654value passed to \f(CW\*(C`ev_once\*(C'\fR. Note that it is possible to receive \fIboth\fR
2655a timeout and an io event at the same time \- you probably should give io
2656events precedence.
2657.Sp
2658Example: wait up to ten seconds for data to appear on \s-1STDIN_FILENO\s0.
973.Sp 2659.Sp
974.Vb 7 2660.Vb 7
975\& static void stdin_ready (int revents, void *arg) 2661\& static void stdin_ready (int revents, void *arg)
976\& { 2662\& {
977\& if (revents & EV_TIMEOUT)
978\& /* doh, nothing entered */;
979\& else if (revents & EV_READ) 2663\& if (revents & EV_READ)
980\& /* stdin might have data for us, joy! */; 2664\& /* stdin might have data for us, joy! */;
2665\& else if (revents & EV_TIMEOUT)
2666\& /* doh, nothing entered */;
981\& } 2667\& }
982.Ve 2668\&
983.Sp
984.Vb 1
985\& ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0); 2669\& ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
986.Ve 2670.Ve
987.IP "ev_feed_event (loop, watcher, int events)" 4 2671.IP "ev_feed_event (ev_loop *, watcher *, int revents)" 4
988.IX Item "ev_feed_event (loop, watcher, int events)" 2672.IX Item "ev_feed_event (ev_loop *, watcher *, int revents)"
989Feeds the given event set into the event loop, as if the specified event 2673Feeds the given event set into the event loop, as if the specified event
990had happened for the specified watcher (which must be a pointer to an 2674had happened for the specified watcher (which must be a pointer to an
991initialised but not necessarily started event watcher). 2675initialised but not necessarily started event watcher).
992.IP "ev_feed_fd_event (loop, int fd, int revents)" 4 2676.IP "ev_feed_fd_event (ev_loop *, int fd, int revents)" 4
993.IX Item "ev_feed_fd_event (loop, int fd, int revents)" 2677.IX Item "ev_feed_fd_event (ev_loop *, int fd, int revents)"
994Feed an event on the given fd, as if a file descriptor backend detected 2678Feed an event on the given fd, as if a file descriptor backend detected
995the given events it. 2679the given events it.
996.IP "ev_feed_signal_event (loop, int signum)" 4 2680.IP "ev_feed_signal_event (ev_loop *loop, int signum)" 4
997.IX Item "ev_feed_signal_event (loop, int signum)" 2681.IX Item "ev_feed_signal_event (ev_loop *loop, int signum)"
998Feed an event as if the given signal occured (loop must be the default loop!). 2682Feed an event as if the given signal occurred (\f(CW\*(C`loop\*(C'\fR must be the default
2683loop!).
999.SH "LIBEVENT EMULATION" 2684.SH "LIBEVENT EMULATION"
1000.IX Header "LIBEVENT EMULATION" 2685.IX Header "LIBEVENT EMULATION"
1001Libev offers a compatibility emulation layer for libevent. It cannot 2686Libev offers a compatibility emulation layer for libevent. It cannot
1002emulate the internals of libevent, so here are some usage hints: 2687emulate the internals of libevent, so here are some usage hints:
2688.IP "\(bu" 4
1003.IP "* Use it by including <event.h>, as usual." 4 2689Use it by including <event.h>, as usual.
1004.IX Item "Use it by including <event.h>, as usual." 2690.IP "\(bu" 4
1005.PD 0 2691The following members are fully supported: ev_base, ev_callback,
1006.IP "* The following members are fully supported: ev_base, ev_callback, ev_arg, ev_fd, ev_res, ev_events." 4 2692ev_arg, ev_fd, ev_res, ev_events.
1007.IX Item "The following members are fully supported: ev_base, ev_callback, ev_arg, ev_fd, ev_res, ev_events." 2693.IP "\(bu" 4
1008.IP "* Avoid using ev_flags and the EVLIST_*\-macros, while it is maintained by libev, it does not work exactly the same way as in libevent (consider it a private \s-1API\s0)." 4 2694Avoid using ev_flags and the EVLIST_*\-macros, while it is
1009.IX Item "Avoid using ev_flags and the EVLIST_*-macros, while it is maintained by libev, it does not work exactly the same way as in libevent (consider it a private API)." 2695maintained by libev, it does not work exactly the same way as in libevent (consider
1010.IP "* Priorities are not currently supported. Initialising priorities will fail and all watchers will have the same priority, even though there is an ev_pri field." 4 2696it a private \s-1API\s0).
1011.IX Item "Priorities are not currently supported. Initialising priorities will fail and all watchers will have the same priority, even though there is an ev_pri field." 2697.IP "\(bu" 4
2698Priorities are not currently supported. Initialising priorities
2699will fail and all watchers will have the same priority, even though there
2700is an ev_pri field.
2701.IP "\(bu" 4
2702In libevent, the last base created gets the signals, in libev, the
2703first base created (== the default loop) gets the signals.
2704.IP "\(bu" 4
1012.IP "* Other members are not supported." 4 2705Other members are not supported.
1013.IX Item "Other members are not supported." 2706.IP "\(bu" 4
1014.IP "* The libev emulation is \fInot\fR \s-1ABI\s0 compatible to libevent, you need to use the libev header file and library." 4 2707The libev emulation is \fInot\fR \s-1ABI\s0 compatible to libevent, you need
1015.IX Item "The libev emulation is not ABI compatible to libevent, you need to use the libev header file and library." 2708to use the libev header file and library.
1016.PD
1017.SH "\*(C+ SUPPORT" 2709.SH "\*(C+ SUPPORT"
1018.IX Header " SUPPORT" 2710.IX Header " SUPPORT"
1019\&\s-1TBD\s0. 2711Libev comes with some simplistic wrapper classes for \*(C+ that mainly allow
2712you to use some convenience methods to start/stop watchers and also change
2713the callback model to a model using method callbacks on objects.
2714.PP
2715To use it,
2716.PP
2717.Vb 1
2718\& #include <ev++.h>
2719.Ve
2720.PP
2721This automatically includes \fIev.h\fR and puts all of its definitions (many
2722of them macros) into the global namespace. All \*(C+ specific things are
2723put into the \f(CW\*(C`ev\*(C'\fR namespace. It should support all the same embedding
2724options as \fIev.h\fR, most notably \f(CW\*(C`EV_MULTIPLICITY\*(C'\fR.
2725.PP
2726Care has been taken to keep the overhead low. The only data member the \*(C+
2727classes add (compared to plain C\-style watchers) is the event loop pointer
2728that the watcher is associated with (or no additional members at all if
2729you disable \f(CW\*(C`EV_MULTIPLICITY\*(C'\fR when embedding libev).
2730.PP
2731Currently, functions, and static and non-static member functions can be
2732used as callbacks. Other types should be easy to add as long as they only
2733need one additional pointer for context. If you need support for other
2734types of functors please contact the author (preferably after implementing
2735it).
2736.PP
2737Here is a list of things available in the \f(CW\*(C`ev\*(C'\fR namespace:
2738.ie n .IP """ev::READ""\fR, \f(CW""ev::WRITE"" etc." 4
2739.el .IP "\f(CWev::READ\fR, \f(CWev::WRITE\fR etc." 4
2740.IX Item "ev::READ, ev::WRITE etc."
2741These are just enum values with the same values as the \f(CW\*(C`EV_READ\*(C'\fR etc.
2742macros from \fIev.h\fR.
2743.ie n .IP """ev::tstamp""\fR, \f(CW""ev::now""" 4
2744.el .IP "\f(CWev::tstamp\fR, \f(CWev::now\fR" 4
2745.IX Item "ev::tstamp, ev::now"
2746Aliases to the same types/functions as with the \f(CW\*(C`ev_\*(C'\fR prefix.
2747.ie n .IP """ev::io""\fR, \f(CW""ev::timer""\fR, \f(CW""ev::periodic""\fR, \f(CW""ev::idle""\fR, \f(CW""ev::sig"" etc." 4
2748.el .IP "\f(CWev::io\fR, \f(CWev::timer\fR, \f(CWev::periodic\fR, \f(CWev::idle\fR, \f(CWev::sig\fR etc." 4
2749.IX Item "ev::io, ev::timer, ev::periodic, ev::idle, ev::sig etc."
2750For each \f(CW\*(C`ev_TYPE\*(C'\fR watcher in \fIev.h\fR there is a corresponding class of
2751the same name in the \f(CW\*(C`ev\*(C'\fR namespace, with the exception of \f(CW\*(C`ev_signal\*(C'\fR
2752which is called \f(CW\*(C`ev::sig\*(C'\fR to avoid clashes with the \f(CW\*(C`signal\*(C'\fR macro
2753defines by many implementations.
2754.Sp
2755All of those classes have these methods:
2756.RS 4
2757.IP "ev::TYPE::TYPE ()" 4
2758.IX Item "ev::TYPE::TYPE ()"
2759.PD 0
2760.IP "ev::TYPE::TYPE (struct ev_loop *)" 4
2761.IX Item "ev::TYPE::TYPE (struct ev_loop *)"
2762.IP "ev::TYPE::~TYPE" 4
2763.IX Item "ev::TYPE::~TYPE"
2764.PD
2765The constructor (optionally) takes an event loop to associate the watcher
2766with. If it is omitted, it will use \f(CW\*(C`EV_DEFAULT\*(C'\fR.
2767.Sp
2768The constructor calls \f(CW\*(C`ev_init\*(C'\fR for you, which means you have to call the
2769\&\f(CW\*(C`set\*(C'\fR method before starting it.
2770.Sp
2771It will not set a callback, however: You have to call the templated \f(CW\*(C`set\*(C'\fR
2772method to set a callback before you can start the watcher.
2773.Sp
2774(The reason why you have to use a method is a limitation in \*(C+ which does
2775not allow explicit template arguments for constructors).
2776.Sp
2777The destructor automatically stops the watcher if it is active.
2778.IP "w\->set<class, &class::method> (object *)" 4
2779.IX Item "w->set<class, &class::method> (object *)"
2780This method sets the callback method to call. The method has to have a
2781signature of \f(CW\*(C`void (*)(ev_TYPE &, int)\*(C'\fR, it receives the watcher as
2782first argument and the \f(CW\*(C`revents\*(C'\fR as second. The object must be given as
2783parameter and is stored in the \f(CW\*(C`data\*(C'\fR member of the watcher.
2784.Sp
2785This method synthesizes efficient thunking code to call your method from
2786the C callback that libev requires. If your compiler can inline your
2787callback (i.e. it is visible to it at the place of the \f(CW\*(C`set\*(C'\fR call and
2788your compiler is good :), then the method will be fully inlined into the
2789thunking function, making it as fast as a direct C callback.
2790.Sp
2791Example: simple class declaration and watcher initialisation
2792.Sp
2793.Vb 4
2794\& struct myclass
2795\& {
2796\& void io_cb (ev::io &w, int revents) { }
2797\& }
2798\&
2799\& myclass obj;
2800\& ev::io iow;
2801\& iow.set <myclass, &myclass::io_cb> (&obj);
2802.Ve
2803.IP "w\->set<function> (void *data = 0)" 4
2804.IX Item "w->set<function> (void *data = 0)"
2805Also sets a callback, but uses a static method or plain function as
2806callback. The optional \f(CW\*(C`data\*(C'\fR argument will be stored in the watcher's
2807\&\f(CW\*(C`data\*(C'\fR member and is free for you to use.
2808.Sp
2809The prototype of the \f(CW\*(C`function\*(C'\fR must be \f(CW\*(C`void (*)(ev::TYPE &w, int)\*(C'\fR.
2810.Sp
2811See the method\-\f(CW\*(C`set\*(C'\fR above for more details.
2812.Sp
2813Example: Use a plain function as callback.
2814.Sp
2815.Vb 2
2816\& static void io_cb (ev::io &w, int revents) { }
2817\& iow.set <io_cb> ();
2818.Ve
2819.IP "w\->set (struct ev_loop *)" 4
2820.IX Item "w->set (struct ev_loop *)"
2821Associates a different \f(CW\*(C`struct ev_loop\*(C'\fR with this watcher. You can only
2822do this when the watcher is inactive (and not pending either).
2823.IP "w\->set ([arguments])" 4
2824.IX Item "w->set ([arguments])"
2825Basically the same as \f(CW\*(C`ev_TYPE_set\*(C'\fR, with the same arguments. Must be
2826called at least once. Unlike the C counterpart, an active watcher gets
2827automatically stopped and restarted when reconfiguring it with this
2828method.
2829.IP "w\->start ()" 4
2830.IX Item "w->start ()"
2831Starts the watcher. Note that there is no \f(CW\*(C`loop\*(C'\fR argument, as the
2832constructor already stores the event loop.
2833.IP "w\->stop ()" 4
2834.IX Item "w->stop ()"
2835Stops the watcher if it is active. Again, no \f(CW\*(C`loop\*(C'\fR argument.
2836.ie n .IP "w\->again () (""ev::timer""\fR, \f(CW""ev::periodic"" only)" 4
2837.el .IP "w\->again () (\f(CWev::timer\fR, \f(CWev::periodic\fR only)" 4
2838.IX Item "w->again () (ev::timer, ev::periodic only)"
2839For \f(CW\*(C`ev::timer\*(C'\fR and \f(CW\*(C`ev::periodic\*(C'\fR, this invokes the corresponding
2840\&\f(CW\*(C`ev_TYPE_again\*(C'\fR function.
2841.ie n .IP "w\->sweep () (""ev::embed"" only)" 4
2842.el .IP "w\->sweep () (\f(CWev::embed\fR only)" 4
2843.IX Item "w->sweep () (ev::embed only)"
2844Invokes \f(CW\*(C`ev_embed_sweep\*(C'\fR.
2845.ie n .IP "w\->update () (""ev::stat"" only)" 4
2846.el .IP "w\->update () (\f(CWev::stat\fR only)" 4
2847.IX Item "w->update () (ev::stat only)"
2848Invokes \f(CW\*(C`ev_stat_stat\*(C'\fR.
2849.RE
2850.RS 4
2851.RE
2852.PP
2853Example: Define a class with an \s-1IO\s0 and idle watcher, start one of them in
2854the constructor.
2855.PP
2856.Vb 4
2857\& class myclass
2858\& {
2859\& ev::io io ; void io_cb (ev::io &w, int revents);
2860\& ev::idle idle; void idle_cb (ev::idle &w, int revents);
2861\&
2862\& myclass (int fd)
2863\& {
2864\& io .set <myclass, &myclass::io_cb > (this);
2865\& idle.set <myclass, &myclass::idle_cb> (this);
2866\&
2867\& io.start (fd, ev::READ);
2868\& }
2869\& };
2870.Ve
2871.SH "OTHER LANGUAGE BINDINGS"
2872.IX Header "OTHER LANGUAGE BINDINGS"
2873Libev does not offer other language bindings itself, but bindings for a
2874number of languages exist in the form of third-party packages. If you know
2875any interesting language binding in addition to the ones listed here, drop
2876me a note.
2877.IP "Perl" 4
2878.IX Item "Perl"
2879The \s-1EV\s0 module implements the full libev \s-1API\s0 and is actually used to test
2880libev. \s-1EV\s0 is developed together with libev. Apart from the \s-1EV\s0 core module,
2881there are additional modules that implement libev-compatible interfaces
2882to \f(CW\*(C`libadns\*(C'\fR (\f(CW\*(C`EV::ADNS\*(C'\fR, but \f(CW\*(C`AnyEvent::DNS\*(C'\fR is preferred nowadays),
2883\&\f(CW\*(C`Net::SNMP\*(C'\fR (\f(CW\*(C`Net::SNMP::EV\*(C'\fR) and the \f(CW\*(C`libglib\*(C'\fR event core (\f(CW\*(C`Glib::EV\*(C'\fR
2884and \f(CW\*(C`EV::Glib\*(C'\fR).
2885.Sp
2886It can be found and installed via \s-1CPAN\s0, its homepage is at
2887<http://software.schmorp.de/pkg/EV>.
2888.IP "Python" 4
2889.IX Item "Python"
2890Python bindings can be found at <http://code.google.com/p/pyev/>. It
2891seems to be quite complete and well-documented. Note, however, that the
2892patch they require for libev is outright dangerous as it breaks the \s-1ABI\s0
2893for everybody else, and therefore, should never be applied in an installed
2894libev (if python requires an incompatible \s-1ABI\s0 then it needs to embed
2895libev).
2896.IP "Ruby" 4
2897.IX Item "Ruby"
2898Tony Arcieri has written a ruby extension that offers access to a subset
2899of the libev \s-1API\s0 and adds file handle abstractions, asynchronous \s-1DNS\s0 and
2900more on top of it. It can be found via gem servers. Its homepage is at
2901<http://rev.rubyforge.org/>.
2902.IP "D" 4
2903.IX Item "D"
2904Leandro Lucarella has written a D language binding (\fIev.d\fR) for libev, to
2905be found at <http://proj.llucax.com.ar/wiki/evd>.
2906.SH "MACRO MAGIC"
2907.IX Header "MACRO MAGIC"
2908Libev can be compiled with a variety of options, the most fundamental
2909of which is \f(CW\*(C`EV_MULTIPLICITY\*(C'\fR. This option determines whether (most)
2910functions and callbacks have an initial \f(CW\*(C`struct ev_loop *\*(C'\fR argument.
2911.PP
2912To make it easier to write programs that cope with either variant, the
2913following macros are defined:
2914.ie n .IP """EV_A""\fR, \f(CW""EV_A_""" 4
2915.el .IP "\f(CWEV_A\fR, \f(CWEV_A_\fR" 4
2916.IX Item "EV_A, EV_A_"
2917This provides the loop \fIargument\fR for functions, if one is required (\*(L"ev
2918loop argument\*(R"). The \f(CW\*(C`EV_A\*(C'\fR form is used when this is the sole argument,
2919\&\f(CW\*(C`EV_A_\*(C'\fR is used when other arguments are following. Example:
2920.Sp
2921.Vb 3
2922\& ev_unref (EV_A);
2923\& ev_timer_add (EV_A_ watcher);
2924\& ev_loop (EV_A_ 0);
2925.Ve
2926.Sp
2927It assumes the variable \f(CW\*(C`loop\*(C'\fR of type \f(CW\*(C`struct ev_loop *\*(C'\fR is in scope,
2928which is often provided by the following macro.
2929.ie n .IP """EV_P""\fR, \f(CW""EV_P_""" 4
2930.el .IP "\f(CWEV_P\fR, \f(CWEV_P_\fR" 4
2931.IX Item "EV_P, EV_P_"
2932This provides the loop \fIparameter\fR for functions, if one is required (\*(L"ev
2933loop parameter\*(R"). The \f(CW\*(C`EV_P\*(C'\fR form is used when this is the sole parameter,
2934\&\f(CW\*(C`EV_P_\*(C'\fR is used when other parameters are following. Example:
2935.Sp
2936.Vb 2
2937\& // this is how ev_unref is being declared
2938\& static void ev_unref (EV_P);
2939\&
2940\& // this is how you can declare your typical callback
2941\& static void cb (EV_P_ ev_timer *w, int revents)
2942.Ve
2943.Sp
2944It declares a parameter \f(CW\*(C`loop\*(C'\fR of type \f(CW\*(C`struct ev_loop *\*(C'\fR, quite
2945suitable for use with \f(CW\*(C`EV_A\*(C'\fR.
2946.ie n .IP """EV_DEFAULT""\fR, \f(CW""EV_DEFAULT_""" 4
2947.el .IP "\f(CWEV_DEFAULT\fR, \f(CWEV_DEFAULT_\fR" 4
2948.IX Item "EV_DEFAULT, EV_DEFAULT_"
2949Similar to the other two macros, this gives you the value of the default
2950loop, if multiple loops are supported (\*(L"ev loop default\*(R").
2951.ie n .IP """EV_DEFAULT_UC""\fR, \f(CW""EV_DEFAULT_UC_""" 4
2952.el .IP "\f(CWEV_DEFAULT_UC\fR, \f(CWEV_DEFAULT_UC_\fR" 4
2953.IX Item "EV_DEFAULT_UC, EV_DEFAULT_UC_"
2954Usage identical to \f(CW\*(C`EV_DEFAULT\*(C'\fR and \f(CW\*(C`EV_DEFAULT_\*(C'\fR, but requires that the
2955default loop has been initialised (\f(CW\*(C`UC\*(C'\fR == unchecked). Their behaviour
2956is undefined when the default loop has not been initialised by a previous
2957execution of \f(CW\*(C`EV_DEFAULT\*(C'\fR, \f(CW\*(C`EV_DEFAULT_\*(C'\fR or \f(CW\*(C`ev_default_init (...)\*(C'\fR.
2958.Sp
2959It is often prudent to use \f(CW\*(C`EV_DEFAULT\*(C'\fR when initialising the first
2960watcher in a function but use \f(CW\*(C`EV_DEFAULT_UC\*(C'\fR afterwards.
2961.PP
2962Example: Declare and initialise a check watcher, utilising the above
2963macros so it will work regardless of whether multiple loops are supported
2964or not.
2965.PP
2966.Vb 5
2967\& static void
2968\& check_cb (EV_P_ ev_timer *w, int revents)
2969\& {
2970\& ev_check_stop (EV_A_ w);
2971\& }
2972\&
2973\& ev_check check;
2974\& ev_check_init (&check, check_cb);
2975\& ev_check_start (EV_DEFAULT_ &check);
2976\& ev_loop (EV_DEFAULT_ 0);
2977.Ve
2978.SH "EMBEDDING"
2979.IX Header "EMBEDDING"
2980Libev can (and often is) directly embedded into host
2981applications. Examples of applications that embed it include the Deliantra
2982Game Server, the \s-1EV\s0 perl module, the \s-1GNU\s0 Virtual Private Ethernet (gvpe)
2983and rxvt-unicode.
2984.PP
2985The goal is to enable you to just copy the necessary files into your
2986source directory without having to change even a single line in them, so
2987you can easily upgrade by simply copying (or having a checked-out copy of
2988libev somewhere in your source tree).
2989.Sh "\s-1FILESETS\s0"
2990.IX Subsection "FILESETS"
2991Depending on what features you need you need to include one or more sets of files
2992in your application.
2993.PP
2994\fI\s-1CORE\s0 \s-1EVENT\s0 \s-1LOOP\s0\fR
2995.IX Subsection "CORE EVENT LOOP"
2996.PP
2997To include only the libev core (all the \f(CW\*(C`ev_*\*(C'\fR functions), with manual
2998configuration (no autoconf):
2999.PP
3000.Vb 2
3001\& #define EV_STANDALONE 1
3002\& #include "ev.c"
3003.Ve
3004.PP
3005This will automatically include \fIev.h\fR, too, and should be done in a
3006single C source file only to provide the function implementations. To use
3007it, do the same for \fIev.h\fR in all files wishing to use this \s-1API\s0 (best
3008done by writing a wrapper around \fIev.h\fR that you can include instead and
3009where you can put other configuration options):
3010.PP
3011.Vb 2
3012\& #define EV_STANDALONE 1
3013\& #include "ev.h"
3014.Ve
3015.PP
3016Both header files and implementation files can be compiled with a \*(C+
3017compiler (at least, thats a stated goal, and breakage will be treated
3018as a bug).
3019.PP
3020You need the following files in your source tree, or in a directory
3021in your include path (e.g. in libev/ when using \-Ilibev):
3022.PP
3023.Vb 4
3024\& ev.h
3025\& ev.c
3026\& ev_vars.h
3027\& ev_wrap.h
3028\&
3029\& ev_win32.c required on win32 platforms only
3030\&
3031\& ev_select.c only when select backend is enabled (which is enabled by default)
3032\& ev_poll.c only when poll backend is enabled (disabled by default)
3033\& ev_epoll.c only when the epoll backend is enabled (disabled by default)
3034\& ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
3035\& ev_port.c only when the solaris port backend is enabled (disabled by default)
3036.Ve
3037.PP
3038\&\fIev.c\fR includes the backend files directly when enabled, so you only need
3039to compile this single file.
3040.PP
3041\fI\s-1LIBEVENT\s0 \s-1COMPATIBILITY\s0 \s-1API\s0\fR
3042.IX Subsection "LIBEVENT COMPATIBILITY API"
3043.PP
3044To include the libevent compatibility \s-1API\s0, also include:
3045.PP
3046.Vb 1
3047\& #include "event.c"
3048.Ve
3049.PP
3050in the file including \fIev.c\fR, and:
3051.PP
3052.Vb 1
3053\& #include "event.h"
3054.Ve
3055.PP
3056in the files that want to use the libevent \s-1API\s0. This also includes \fIev.h\fR.
3057.PP
3058You need the following additional files for this:
3059.PP
3060.Vb 2
3061\& event.h
3062\& event.c
3063.Ve
3064.PP
3065\fI\s-1AUTOCONF\s0 \s-1SUPPORT\s0\fR
3066.IX Subsection "AUTOCONF SUPPORT"
3067.PP
3068Instead of using \f(CW\*(C`EV_STANDALONE=1\*(C'\fR and providing your configuration in
3069whatever way you want, you can also \f(CW\*(C`m4_include([libev.m4])\*(C'\fR in your
3070\&\fIconfigure.ac\fR and leave \f(CW\*(C`EV_STANDALONE\*(C'\fR undefined. \fIev.c\fR will then
3071include \fIconfig.h\fR and configure itself accordingly.
3072.PP
3073For this of course you need the m4 file:
3074.PP
3075.Vb 1
3076\& libev.m4
3077.Ve
3078.Sh "\s-1PREPROCESSOR\s0 \s-1SYMBOLS/MACROS\s0"
3079.IX Subsection "PREPROCESSOR SYMBOLS/MACROS"
3080Libev can be configured via a variety of preprocessor symbols you have to
3081define before including any of its files. The default in the absence of
3082autoconf is documented for every option.
3083.IP "\s-1EV_STANDALONE\s0" 4
3084.IX Item "EV_STANDALONE"
3085Must always be \f(CW1\fR if you do not use autoconf configuration, which
3086keeps libev from including \fIconfig.h\fR, and it also defines dummy
3087implementations for some libevent functions (such as logging, which is not
3088supported). It will also not define any of the structs usually found in
3089\&\fIevent.h\fR that are not directly supported by the libev core alone.
3090.IP "\s-1EV_USE_MONOTONIC\s0" 4
3091.IX Item "EV_USE_MONOTONIC"
3092If defined to be \f(CW1\fR, libev will try to detect the availability of the
3093monotonic clock option at both compile time and runtime. Otherwise no use
3094of the monotonic clock option will be attempted. If you enable this, you
3095usually have to link against librt or something similar. Enabling it when
3096the functionality isn't available is safe, though, although you have
3097to make sure you link against any libraries where the \f(CW\*(C`clock_gettime\*(C'\fR
3098function is hiding in (often \fI\-lrt\fR).
3099.IP "\s-1EV_USE_REALTIME\s0" 4
3100.IX Item "EV_USE_REALTIME"
3101If defined to be \f(CW1\fR, libev will try to detect the availability of the
3102real-time clock option at compile time (and assume its availability at
3103runtime if successful). Otherwise no use of the real-time clock option will
3104be attempted. This effectively replaces \f(CW\*(C`gettimeofday\*(C'\fR by \f(CW\*(C`clock_get
3105(CLOCK_REALTIME, ...)\*(C'\fR and will not normally affect correctness. See the
3106note about libraries in the description of \f(CW\*(C`EV_USE_MONOTONIC\*(C'\fR, though.
3107.IP "\s-1EV_USE_NANOSLEEP\s0" 4
3108.IX Item "EV_USE_NANOSLEEP"
3109If defined to be \f(CW1\fR, libev will assume that \f(CW\*(C`nanosleep ()\*(C'\fR is available
3110and will use it for delays. Otherwise it will use \f(CW\*(C`select ()\*(C'\fR.
3111.IP "\s-1EV_USE_EVENTFD\s0" 4
3112.IX Item "EV_USE_EVENTFD"
3113If defined to be \f(CW1\fR, then libev will assume that \f(CW\*(C`eventfd ()\*(C'\fR is
3114available and will probe for kernel support at runtime. This will improve
3115\&\f(CW\*(C`ev_signal\*(C'\fR and \f(CW\*(C`ev_async\*(C'\fR performance and reduce resource consumption.
3116If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
31172.7 or newer, otherwise disabled.
3118.IP "\s-1EV_USE_SELECT\s0" 4
3119.IX Item "EV_USE_SELECT"
3120If undefined or defined to be \f(CW1\fR, libev will compile in support for the
3121\&\f(CW\*(C`select\*(C'\fR(2) backend. No attempt at auto-detection will be done: if no
3122other method takes over, select will be it. Otherwise the select backend
3123will not be compiled in.
3124.IP "\s-1EV_SELECT_USE_FD_SET\s0" 4
3125.IX Item "EV_SELECT_USE_FD_SET"
3126If defined to \f(CW1\fR, then the select backend will use the system \f(CW\*(C`fd_set\*(C'\fR
3127structure. This is useful if libev doesn't compile due to a missing
3128\&\f(CW\*(C`NFDBITS\*(C'\fR or \f(CW\*(C`fd_mask\*(C'\fR definition or it mis-guesses the bitset layout on
3129exotic systems. This usually limits the range of file descriptors to some
3130low limit such as 1024 or might have other limitations (winsocket only
3131allows 64 sockets). The \f(CW\*(C`FD_SETSIZE\*(C'\fR macro, set before compilation, might
3132influence the size of the \f(CW\*(C`fd_set\*(C'\fR used.
3133.IP "\s-1EV_SELECT_IS_WINSOCKET\s0" 4
3134.IX Item "EV_SELECT_IS_WINSOCKET"
3135When defined to \f(CW1\fR, the select backend will assume that
3136select/socket/connect etc. don't understand file descriptors but
3137wants osf handles on win32 (this is the case when the select to
3138be used is the winsock select). This means that it will call
3139\&\f(CW\*(C`_get_osfhandle\*(C'\fR on the fd to convert it to an \s-1OS\s0 handle. Otherwise,
3140it is assumed that all these functions actually work on fds, even
3141on win32. Should not be defined on non\-win32 platforms.
3142.IP "\s-1EV_FD_TO_WIN32_HANDLE\s0" 4
3143.IX Item "EV_FD_TO_WIN32_HANDLE"
3144If \f(CW\*(C`EV_SELECT_IS_WINSOCKET\*(C'\fR is enabled, then libev needs a way to map
3145file descriptors to socket handles. When not defining this symbol (the
3146default), then libev will call \f(CW\*(C`_get_osfhandle\*(C'\fR, which is usually
3147correct. In some cases, programs use their own file descriptor management,
3148in which case they can provide this function to map fds to socket handles.
3149.IP "\s-1EV_USE_POLL\s0" 4
3150.IX Item "EV_USE_POLL"
3151If defined to be \f(CW1\fR, libev will compile in support for the \f(CW\*(C`poll\*(C'\fR(2)
3152backend. Otherwise it will be enabled on non\-win32 platforms. It
3153takes precedence over select.
3154.IP "\s-1EV_USE_EPOLL\s0" 4
3155.IX Item "EV_USE_EPOLL"
3156If defined to be \f(CW1\fR, libev will compile in support for the Linux
3157\&\f(CW\*(C`epoll\*(C'\fR(7) backend. Its availability will be detected at runtime,
3158otherwise another method will be used as fallback. This is the preferred
3159backend for GNU/Linux systems. If undefined, it will be enabled if the
3160headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
3161.IP "\s-1EV_USE_KQUEUE\s0" 4
3162.IX Item "EV_USE_KQUEUE"
3163If defined to be \f(CW1\fR, libev will compile in support for the \s-1BSD\s0 style
3164\&\f(CW\*(C`kqueue\*(C'\fR(2) backend. Its actual availability will be detected at runtime,
3165otherwise another method will be used as fallback. This is the preferred
3166backend for \s-1BSD\s0 and BSD-like systems, although on most BSDs kqueue only
3167supports some types of fds correctly (the only platform we found that
3168supports ptys for example was NetBSD), so kqueue might be compiled in, but
3169not be used unless explicitly requested. The best way to use it is to find
3170out whether kqueue supports your type of fd properly and use an embedded
3171kqueue loop.
3172.IP "\s-1EV_USE_PORT\s0" 4
3173.IX Item "EV_USE_PORT"
3174If defined to be \f(CW1\fR, libev will compile in support for the Solaris
317510 port style backend. Its availability will be detected at runtime,
3176otherwise another method will be used as fallback. This is the preferred
3177backend for Solaris 10 systems.
3178.IP "\s-1EV_USE_DEVPOLL\s0" 4
3179.IX Item "EV_USE_DEVPOLL"
3180Reserved for future expansion, works like the \s-1USE\s0 symbols above.
3181.IP "\s-1EV_USE_INOTIFY\s0" 4
3182.IX Item "EV_USE_INOTIFY"
3183If defined to be \f(CW1\fR, libev will compile in support for the Linux inotify
3184interface to speed up \f(CW\*(C`ev_stat\*(C'\fR watchers. Its actual availability will
3185be detected at runtime. If undefined, it will be enabled if the headers
3186indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
3187.IP "\s-1EV_ATOMIC_T\s0" 4
3188.IX Item "EV_ATOMIC_T"
3189Libev requires an integer type (suitable for storing \f(CW0\fR or \f(CW1\fR) whose
3190access is atomic with respect to other threads or signal contexts. No such
3191type is easily found in the C language, so you can provide your own type
3192that you know is safe for your purposes. It is used both for signal handler \*(L"locking\*(R"
3193as well as for signal and thread safety in \f(CW\*(C`ev_async\*(C'\fR watchers.
3194.Sp
3195In the absence of this define, libev will use \f(CW\*(C`sig_atomic_t volatile\*(C'\fR
3196(from \fIsignal.h\fR), which is usually good enough on most platforms.
3197.IP "\s-1EV_H\s0" 4
3198.IX Item "EV_H"
3199The name of the \fIev.h\fR header file used to include it. The default if
3200undefined is \f(CW"ev.h"\fR in \fIevent.h\fR, \fIev.c\fR and \fIev++.h\fR. This can be
3201used to virtually rename the \fIev.h\fR header file in case of conflicts.
3202.IP "\s-1EV_CONFIG_H\s0" 4
3203.IX Item "EV_CONFIG_H"
3204If \f(CW\*(C`EV_STANDALONE\*(C'\fR isn't \f(CW1\fR, this variable can be used to override
3205\&\fIev.c\fR's idea of where to find the \fIconfig.h\fR file, similarly to
3206\&\f(CW\*(C`EV_H\*(C'\fR, above.
3207.IP "\s-1EV_EVENT_H\s0" 4
3208.IX Item "EV_EVENT_H"
3209Similarly to \f(CW\*(C`EV_H\*(C'\fR, this macro can be used to override \fIevent.c\fR's idea
3210of how the \fIevent.h\fR header can be found, the default is \f(CW"event.h"\fR.
3211.IP "\s-1EV_PROTOTYPES\s0" 4
3212.IX Item "EV_PROTOTYPES"
3213If defined to be \f(CW0\fR, then \fIev.h\fR will not define any function
3214prototypes, but still define all the structs and other symbols. This is
3215occasionally useful if you want to provide your own wrapper functions
3216around libev functions.
3217.IP "\s-1EV_MULTIPLICITY\s0" 4
3218.IX Item "EV_MULTIPLICITY"
3219If undefined or defined to \f(CW1\fR, then all event-loop-specific functions
3220will have the \f(CW\*(C`struct ev_loop *\*(C'\fR as first argument, and you can create
3221additional independent event loops. Otherwise there will be no support
3222for multiple event loops and there is no first event loop pointer
3223argument. Instead, all functions act on the single default loop.
3224.IP "\s-1EV_MINPRI\s0" 4
3225.IX Item "EV_MINPRI"
3226.PD 0
3227.IP "\s-1EV_MAXPRI\s0" 4
3228.IX Item "EV_MAXPRI"
3229.PD
3230The range of allowed priorities. \f(CW\*(C`EV_MINPRI\*(C'\fR must be smaller or equal to
3231\&\f(CW\*(C`EV_MAXPRI\*(C'\fR, but otherwise there are no non-obvious limitations. You can
3232provide for more priorities by overriding those symbols (usually defined
3233to be \f(CW\*(C`\-2\*(C'\fR and \f(CW2\fR, respectively).
3234.Sp
3235When doing priority-based operations, libev usually has to linearly search
3236all the priorities, so having many of them (hundreds) uses a lot of space
3237and time, so using the defaults of five priorities (\-2 .. +2) is usually
3238fine.
3239.Sp
3240If your embedding application does not need any priorities, defining these
3241both to \f(CW0\fR will save some memory and \s-1CPU\s0.
3242.IP "\s-1EV_PERIODIC_ENABLE\s0" 4
3243.IX Item "EV_PERIODIC_ENABLE"
3244If undefined or defined to be \f(CW1\fR, then periodic timers are supported. If
3245defined to be \f(CW0\fR, then they are not. Disabling them saves a few kB of
3246code.
3247.IP "\s-1EV_IDLE_ENABLE\s0" 4
3248.IX Item "EV_IDLE_ENABLE"
3249If undefined or defined to be \f(CW1\fR, then idle watchers are supported. If
3250defined to be \f(CW0\fR, then they are not. Disabling them saves a few kB of
3251code.
3252.IP "\s-1EV_EMBED_ENABLE\s0" 4
3253.IX Item "EV_EMBED_ENABLE"
3254If undefined or defined to be \f(CW1\fR, then embed watchers are supported. If
3255defined to be \f(CW0\fR, then they are not. Embed watchers rely on most other
3256watcher types, which therefore must not be disabled.
3257.IP "\s-1EV_STAT_ENABLE\s0" 4
3258.IX Item "EV_STAT_ENABLE"
3259If undefined or defined to be \f(CW1\fR, then stat watchers are supported. If
3260defined to be \f(CW0\fR, then they are not.
3261.IP "\s-1EV_FORK_ENABLE\s0" 4
3262.IX Item "EV_FORK_ENABLE"
3263If undefined or defined to be \f(CW1\fR, then fork watchers are supported. If
3264defined to be \f(CW0\fR, then they are not.
3265.IP "\s-1EV_ASYNC_ENABLE\s0" 4
3266.IX Item "EV_ASYNC_ENABLE"
3267If undefined or defined to be \f(CW1\fR, then async watchers are supported. If
3268defined to be \f(CW0\fR, then they are not.
3269.IP "\s-1EV_MINIMAL\s0" 4
3270.IX Item "EV_MINIMAL"
3271If you need to shave off some kilobytes of code at the expense of some
3272speed, define this symbol to \f(CW1\fR. Currently this is used to override some
3273inlining decisions, saves roughly 30% code size on amd64. It also selects a
3274much smaller 2\-heap for timer management over the default 4\-heap.
3275.IP "\s-1EV_PID_HASHSIZE\s0" 4
3276.IX Item "EV_PID_HASHSIZE"
3277\&\f(CW\*(C`ev_child\*(C'\fR watchers use a small hash table to distribute workload by
3278pid. The default size is \f(CW16\fR (or \f(CW1\fR with \f(CW\*(C`EV_MINIMAL\*(C'\fR), usually more
3279than enough. If you need to manage thousands of children you might want to
3280increase this value (\fImust\fR be a power of two).
3281.IP "\s-1EV_INOTIFY_HASHSIZE\s0" 4
3282.IX Item "EV_INOTIFY_HASHSIZE"
3283\&\f(CW\*(C`ev_stat\*(C'\fR watchers use a small hash table to distribute workload by
3284inotify watch id. The default size is \f(CW16\fR (or \f(CW1\fR with \f(CW\*(C`EV_MINIMAL\*(C'\fR),
3285usually more than enough. If you need to manage thousands of \f(CW\*(C`ev_stat\*(C'\fR
3286watchers you might want to increase this value (\fImust\fR be a power of
3287two).
3288.IP "\s-1EV_USE_4HEAP\s0" 4
3289.IX Item "EV_USE_4HEAP"
3290Heaps are not very cache-efficient. To improve the cache-efficiency of the
3291timer and periodics heaps, libev uses a 4\-heap when this symbol is defined
3292to \f(CW1\fR. The 4\-heap uses more complicated (longer) code but has noticeably
3293faster performance with many (thousands) of watchers.
3294.Sp
3295The default is \f(CW1\fR unless \f(CW\*(C`EV_MINIMAL\*(C'\fR is set in which case it is \f(CW0\fR
3296(disabled).
3297.IP "\s-1EV_HEAP_CACHE_AT\s0" 4
3298.IX Item "EV_HEAP_CACHE_AT"
3299Heaps are not very cache-efficient. To improve the cache-efficiency of the
3300timer and periodics heaps, libev can cache the timestamp (\fIat\fR) within
3301the heap structure (selected by defining \f(CW\*(C`EV_HEAP_CACHE_AT\*(C'\fR to \f(CW1\fR),
3302which uses 8\-12 bytes more per watcher and a few hundred bytes more code,
3303but avoids random read accesses on heap changes. This improves performance
3304noticeably with many (hundreds) of watchers.
3305.Sp
3306The default is \f(CW1\fR unless \f(CW\*(C`EV_MINIMAL\*(C'\fR is set in which case it is \f(CW0\fR
3307(disabled).
3308.IP "\s-1EV_VERIFY\s0" 4
3309.IX Item "EV_VERIFY"
3310Controls how much internal verification (see \f(CW\*(C`ev_loop_verify ()\*(C'\fR) will
3311be done: If set to \f(CW0\fR, no internal verification code will be compiled
3312in. If set to \f(CW1\fR, then verification code will be compiled in, but not
3313called. If set to \f(CW2\fR, then the internal verification code will be
3314called once per loop, which can slow down libev. If set to \f(CW3\fR, then the
3315verification code will be called very frequently, which will slow down
3316libev considerably.
3317.Sp
3318The default is \f(CW1\fR, unless \f(CW\*(C`EV_MINIMAL\*(C'\fR is set, in which case it will be
3319\&\f(CW0\fR.
3320.IP "\s-1EV_COMMON\s0" 4
3321.IX Item "EV_COMMON"
3322By default, all watchers have a \f(CW\*(C`void *data\*(C'\fR member. By redefining
3323this macro to a something else you can include more and other types of
3324members. You have to define it each time you include one of the files,
3325though, and it must be identical each time.
3326.Sp
3327For example, the perl \s-1EV\s0 module uses something like this:
3328.Sp
3329.Vb 3
3330\& #define EV_COMMON \e
3331\& SV *self; /* contains this struct */ \e
3332\& SV *cb_sv, *fh /* note no trailing ";" */
3333.Ve
3334.IP "\s-1EV_CB_DECLARE\s0 (type)" 4
3335.IX Item "EV_CB_DECLARE (type)"
3336.PD 0
3337.IP "\s-1EV_CB_INVOKE\s0 (watcher, revents)" 4
3338.IX Item "EV_CB_INVOKE (watcher, revents)"
3339.IP "ev_set_cb (ev, cb)" 4
3340.IX Item "ev_set_cb (ev, cb)"
3341.PD
3342Can be used to change the callback member declaration in each watcher,
3343and the way callbacks are invoked and set. Must expand to a struct member
3344definition and a statement, respectively. See the \fIev.h\fR header file for
3345their default definitions. One possible use for overriding these is to
3346avoid the \f(CW\*(C`struct ev_loop *\*(C'\fR as first argument in all cases, or to use
3347method calls instead of plain function calls in \*(C+.
3348.Sh "\s-1EXPORTED\s0 \s-1API\s0 \s-1SYMBOLS\s0"
3349.IX Subsection "EXPORTED API SYMBOLS"
3350If you need to re-export the \s-1API\s0 (e.g. via a \s-1DLL\s0) and you need a list of
3351exported symbols, you can use the provided \fISymbol.*\fR files which list
3352all public symbols, one per line:
3353.PP
3354.Vb 2
3355\& Symbols.ev for libev proper
3356\& Symbols.event for the libevent emulation
3357.Ve
3358.PP
3359This can also be used to rename all public symbols to avoid clashes with
3360multiple versions of libev linked together (which is obviously bad in
3361itself, but sometimes it is inconvenient to avoid this).
3362.PP
3363A sed command like this will create wrapper \f(CW\*(C`#define\*(C'\fR's that you need to
3364include before including \fIev.h\fR:
3365.PP
3366.Vb 1
3367\& <Symbols.ev sed \-e "s/.*/#define & myprefix_&/" >wrap.h
3368.Ve
3369.PP
3370This would create a file \fIwrap.h\fR which essentially looks like this:
3371.PP
3372.Vb 4
3373\& #define ev_backend myprefix_ev_backend
3374\& #define ev_check_start myprefix_ev_check_start
3375\& #define ev_check_stop myprefix_ev_check_stop
3376\& ...
3377.Ve
3378.Sh "\s-1EXAMPLES\s0"
3379.IX Subsection "EXAMPLES"
3380For a real-world example of a program the includes libev
3381verbatim, you can have a look at the \s-1EV\s0 perl module
3382(<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
3383the \fIlibev/\fR subdirectory and includes them in the \fI\s-1EV/EVAPI\s0.h\fR (public
3384interface) and \fI\s-1EV\s0.xs\fR (implementation) files. Only the \fI\s-1EV\s0.xs\fR file
3385will be compiled. It is pretty complex because it provides its own header
3386file.
3387.PP
3388The usage in rxvt-unicode is simpler. It has a \fIev_cpp.h\fR header file
3389that everybody includes and which overrides some configure choices:
3390.PP
3391.Vb 9
3392\& #define EV_MINIMAL 1
3393\& #define EV_USE_POLL 0
3394\& #define EV_MULTIPLICITY 0
3395\& #define EV_PERIODIC_ENABLE 0
3396\& #define EV_STAT_ENABLE 0
3397\& #define EV_FORK_ENABLE 0
3398\& #define EV_CONFIG_H <config.h>
3399\& #define EV_MINPRI 0
3400\& #define EV_MAXPRI 0
3401\&
3402\& #include "ev++.h"
3403.Ve
3404.PP
3405And a \fIev_cpp.C\fR implementation file that contains libev proper and is compiled:
3406.PP
3407.Vb 2
3408\& #include "ev_cpp.h"
3409\& #include "ev.c"
3410.Ve
3411.SH "INTERACTION WITH OTHER PROGRAMS OR LIBRARIES"
3412.IX Header "INTERACTION WITH OTHER PROGRAMS OR LIBRARIES"
3413.Sh "\s-1THREADS\s0 \s-1AND\s0 \s-1COROUTINES\s0"
3414.IX Subsection "THREADS AND COROUTINES"
3415\fI\s-1THREADS\s0\fR
3416.IX Subsection "THREADS"
3417.PP
3418All libev functions are reentrant and thread-safe unless explicitly
3419documented otherwise, but libev implements no locking itself. This means
3420that you can use as many loops as you want in parallel, as long as there
3421are no concurrent calls into any libev function with the same loop
3422parameter (\f(CW\*(C`ev_default_*\*(C'\fR calls have an implicit default loop parameter,
3423of course): libev guarantees that different event loops share no data
3424structures that need any locking.
3425.PP
3426Or to put it differently: calls with different loop parameters can be done
3427concurrently from multiple threads, calls with the same loop parameter
3428must be done serially (but can be done from different threads, as long as
3429only one thread ever is inside a call at any point in time, e.g. by using
3430a mutex per loop).
3431.PP
3432Specifically to support threads (and signal handlers), libev implements
3433so-called \f(CW\*(C`ev_async\*(C'\fR watchers, which allow some limited form of
3434concurrency on the same event loop, namely waking it up \*(L"from the
3435outside\*(R".
3436.PP
3437If you want to know which design (one loop, locking, or multiple loops
3438without or something else still) is best for your problem, then I cannot
3439help you, but here is some generic advice:
3440.IP "\(bu" 4
3441most applications have a main thread: use the default libev loop
3442in that thread, or create a separate thread running only the default loop.
3443.Sp
3444This helps integrating other libraries or software modules that use libev
3445themselves and don't care/know about threading.
3446.IP "\(bu" 4
3447one loop per thread is usually a good model.
3448.Sp
3449Doing this is almost never wrong, sometimes a better-performance model
3450exists, but it is always a good start.
3451.IP "\(bu" 4
3452other models exist, such as the leader/follower pattern, where one
3453loop is handed through multiple threads in a kind of round-robin fashion.
3454.Sp
3455Choosing a model is hard \- look around, learn, know that usually you can do
3456better than you currently do :\-)
3457.IP "\(bu" 4
3458often you need to talk to some other thread which blocks in the
3459event loop.
3460.Sp
3461\&\f(CW\*(C`ev_async\*(C'\fR watchers can be used to wake them up from other threads safely
3462(or from signal contexts...).
3463.Sp
3464An example use would be to communicate signals or other events that only
3465work in the default loop by registering the signal watcher with the
3466default loop and triggering an \f(CW\*(C`ev_async\*(C'\fR watcher from the default loop
3467watcher callback into the event loop interested in the signal.
3468.PP
3469\fI\s-1COROUTINES\s0\fR
3470.IX Subsection "COROUTINES"
3471.PP
3472Libev is very accommodating to coroutines (\*(L"cooperative threads\*(R"):
3473libev fully supports nesting calls to its functions from different
3474coroutines (e.g. you can call \f(CW\*(C`ev_loop\*(C'\fR on the same loop from two
3475different coroutines, and switch freely between both coroutines running the
3476loop, as long as you don't confuse yourself). The only exception is that
3477you must not do this from \f(CW\*(C`ev_periodic\*(C'\fR reschedule callbacks.
3478.PP
3479Care has been taken to ensure that libev does not keep local state inside
3480\&\f(CW\*(C`ev_loop\*(C'\fR, and other calls do not usually allow for coroutine switches as
3481they do not clal any callbacks.
3482.Sh "\s-1COMPILER\s0 \s-1WARNINGS\s0"
3483.IX Subsection "COMPILER WARNINGS"
3484Depending on your compiler and compiler settings, you might get no or a
3485lot of warnings when compiling libev code. Some people are apparently
3486scared by this.
3487.PP
3488However, these are unavoidable for many reasons. For one, each compiler
3489has different warnings, and each user has different tastes regarding
3490warning options. \*(L"Warn-free\*(R" code therefore cannot be a goal except when
3491targeting a specific compiler and compiler-version.
3492.PP
3493Another reason is that some compiler warnings require elaborate
3494workarounds, or other changes to the code that make it less clear and less
3495maintainable.
3496.PP
3497And of course, some compiler warnings are just plain stupid, or simply
3498wrong (because they don't actually warn about the condition their message
3499seems to warn about). For example, certain older gcc versions had some
3500warnings that resulted an extreme number of false positives. These have
3501been fixed, but some people still insist on making code warn-free with
3502such buggy versions.
3503.PP
3504While libev is written to generate as few warnings as possible,
3505\&\*(L"warn-free\*(R" code is not a goal, and it is recommended not to build libev
3506with any compiler warnings enabled unless you are prepared to cope with
3507them (e.g. by ignoring them). Remember that warnings are just that:
3508warnings, not errors, or proof of bugs.
3509.Sh "\s-1VALGRIND\s0"
3510.IX Subsection "VALGRIND"
3511Valgrind has a special section here because it is a popular tool that is
3512highly useful. Unfortunately, valgrind reports are very hard to interpret.
3513.PP
3514If you think you found a bug (memory leak, uninitialised data access etc.)
3515in libev, then check twice: If valgrind reports something like:
3516.PP
3517.Vb 3
3518\& ==2274== definitely lost: 0 bytes in 0 blocks.
3519\& ==2274== possibly lost: 0 bytes in 0 blocks.
3520\& ==2274== still reachable: 256 bytes in 1 blocks.
3521.Ve
3522.PP
3523Then there is no memory leak, just as memory accounted to global variables
3524is not a memleak \- the memory is still being refernced, and didn't leak.
3525.PP
3526Similarly, under some circumstances, valgrind might report kernel bugs
3527as if it were a bug in libev (e.g. in realloc or in the poll backend,
3528although an acceptable workaround has been found here), or it might be
3529confused.
3530.PP
3531Keep in mind that valgrind is a very good tool, but only a tool. Don't
3532make it into some kind of religion.
3533.PP
3534If you are unsure about something, feel free to contact the mailing list
3535with the full valgrind report and an explanation on why you think this
3536is a bug in libev (best check the archives, too :). However, don't be
3537annoyed when you get a brisk \*(L"this is no bug\*(R" answer and take the chance
3538of learning how to interpret valgrind properly.
3539.PP
3540If you need, for some reason, empty reports from valgrind for your project
3541I suggest using suppression lists.
3542.SH "PORTABILITY NOTES"
3543.IX Header "PORTABILITY NOTES"
3544.Sh "\s-1WIN32\s0 \s-1PLATFORM\s0 \s-1LIMITATIONS\s0 \s-1AND\s0 \s-1WORKAROUNDS\s0"
3545.IX Subsection "WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS"
3546Win32 doesn't support any of the standards (e.g. \s-1POSIX\s0) that libev
3547requires, and its I/O model is fundamentally incompatible with the \s-1POSIX\s0
3548model. Libev still offers limited functionality on this platform in
3549the form of the \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR backend, and only supports socket
3550descriptors. This only applies when using Win32 natively, not when using
3551e.g. cygwin.
3552.PP
3553Lifting these limitations would basically require the full
3554re-implementation of the I/O system. If you are into these kinds of
3555things, then note that glib does exactly that for you in a very portable
3556way (note also that glib is the slowest event library known to man).
3557.PP
3558There is no supported compilation method available on windows except
3559embedding it into other applications.
3560.PP
3561Not a libev limitation but worth mentioning: windows apparently doesn't
3562accept large writes: instead of resulting in a partial write, windows will
3563either accept everything or return \f(CW\*(C`ENOBUFS\*(C'\fR if the buffer is too large,
3564so make sure you only write small amounts into your sockets (less than a
3565megabyte seems safe, but this apparently depends on the amount of memory
3566available).
3567.PP
3568Due to the many, low, and arbitrary limits on the win32 platform and
3569the abysmal performance of winsockets, using a large number of sockets
3570is not recommended (and not reasonable). If your program needs to use
3571more than a hundred or so sockets, then likely it needs to use a totally
3572different implementation for windows, as libev offers the \s-1POSIX\s0 readiness
3573notification model, which cannot be implemented efficiently on windows
3574(Microsoft monopoly games).
3575.PP
3576A typical way to use libev under windows is to embed it (see the embedding
3577section for details) and use the following \fIevwrap.h\fR header file instead
3578of \fIev.h\fR:
3579.PP
3580.Vb 2
3581\& #define EV_STANDALONE /* keeps ev from requiring config.h */
3582\& #define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
3583\&
3584\& #include "ev.h"
3585.Ve
3586.PP
3587And compile the following \fIevwrap.c\fR file into your project (make sure
3588you do \fInot\fR compile the \fIev.c\fR or any other embedded source files!):
3589.PP
3590.Vb 2
3591\& #include "evwrap.h"
3592\& #include "ev.c"
3593.Ve
3594.IP "The winsocket select function" 4
3595.IX Item "The winsocket select function"
3596The winsocket \f(CW\*(C`select\*(C'\fR function doesn't follow \s-1POSIX\s0 in that it
3597requires socket \fIhandles\fR and not socket \fIfile descriptors\fR (it is
3598also extremely buggy). This makes select very inefficient, and also
3599requires a mapping from file descriptors to socket handles (the Microsoft
3600C runtime provides the function \f(CW\*(C`_open_osfhandle\*(C'\fR for this). See the
3601discussion of the \f(CW\*(C`EV_SELECT_USE_FD_SET\*(C'\fR, \f(CW\*(C`EV_SELECT_IS_WINSOCKET\*(C'\fR and
3602\&\f(CW\*(C`EV_FD_TO_WIN32_HANDLE\*(C'\fR preprocessor symbols for more info.
3603.Sp
3604The configuration for a \*(L"naked\*(R" win32 using the Microsoft runtime
3605libraries and raw winsocket select is:
3606.Sp
3607.Vb 2
3608\& #define EV_USE_SELECT 1
3609\& #define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
3610.Ve
3611.Sp
3612Note that winsockets handling of fd sets is O(n), so you can easily get a
3613complexity in the O(nA\*^X) range when using win32.
3614.IP "Limited number of file descriptors" 4
3615.IX Item "Limited number of file descriptors"
3616Windows has numerous arbitrary (and low) limits on things.
3617.Sp
3618Early versions of winsocket's select only supported waiting for a maximum
3619of \f(CW64\fR handles (probably owning to the fact that all windows kernels
3620can only wait for \f(CW64\fR things at the same time internally; Microsoft
3621recommends spawning a chain of threads and wait for 63 handles and the
3622previous thread in each. Great).
3623.Sp
3624Newer versions support more handles, but you need to define \f(CW\*(C`FD_SETSIZE\*(C'\fR
3625to some high number (e.g. \f(CW2048\fR) before compiling the winsocket select
3626call (which might be in libev or elsewhere, for example, perl does its own
3627select emulation on windows).
3628.Sp
3629Another limit is the number of file descriptors in the Microsoft runtime
3630libraries, which by default is \f(CW64\fR (there must be a hidden \fI64\fR fetish
3631or something like this inside Microsoft). You can increase this by calling
3632\&\f(CW\*(C`_setmaxstdio\*(C'\fR, which can increase this limit to \f(CW2048\fR (another
3633arbitrary limit), but is broken in many versions of the Microsoft runtime
3634libraries.
3635.Sp
3636This might get you to about \f(CW512\fR or \f(CW2048\fR sockets (depending on
3637windows version and/or the phase of the moon). To get more, you need to
3638wrap all I/O functions and provide your own fd management, but the cost of
3639calling select (O(nA\*^X)) will likely make this unworkable.
3640.Sh "\s-1PORTABILITY\s0 \s-1REQUIREMENTS\s0"
3641.IX Subsection "PORTABILITY REQUIREMENTS"
3642In addition to a working ISO-C implementation and of course the
3643backend-specific APIs, libev relies on a few additional extensions:
3644.ie n .IP """void (*)(ev_watcher_type *, int revents)""\fR must have compatible calling conventions regardless of \f(CW""ev_watcher_type *""." 4
3645.el .IP "\f(CWvoid (*)(ev_watcher_type *, int revents)\fR must have compatible calling conventions regardless of \f(CWev_watcher_type *\fR." 4
3646.IX Item "void (*)(ev_watcher_type *, int revents) must have compatible calling conventions regardless of ev_watcher_type *."
3647Libev assumes not only that all watcher pointers have the same internal
3648structure (guaranteed by \s-1POSIX\s0 but not by \s-1ISO\s0 C for example), but it also
3649assumes that the same (machine) code can be used to call any watcher
3650callback: The watcher callbacks have different type signatures, but libev
3651calls them using an \f(CW\*(C`ev_watcher *\*(C'\fR internally.
3652.ie n .IP """sig_atomic_t volatile"" must be thread-atomic as well" 4
3653.el .IP "\f(CWsig_atomic_t volatile\fR must be thread-atomic as well" 4
3654.IX Item "sig_atomic_t volatile must be thread-atomic as well"
3655The type \f(CW\*(C`sig_atomic_t volatile\*(C'\fR (or whatever is defined as
3656\&\f(CW\*(C`EV_ATOMIC_T\*(C'\fR) must be atomic with respect to accesses from different
3657threads. This is not part of the specification for \f(CW\*(C`sig_atomic_t\*(C'\fR, but is
3658believed to be sufficiently portable.
3659.ie n .IP """sigprocmask"" must work in a threaded environment" 4
3660.el .IP "\f(CWsigprocmask\fR must work in a threaded environment" 4
3661.IX Item "sigprocmask must work in a threaded environment"
3662Libev uses \f(CW\*(C`sigprocmask\*(C'\fR to temporarily block signals. This is not
3663allowed in a threaded program (\f(CW\*(C`pthread_sigmask\*(C'\fR has to be used). Typical
3664pthread implementations will either allow \f(CW\*(C`sigprocmask\*(C'\fR in the \*(L"main
3665thread\*(R" or will block signals process-wide, both behaviours would
3666be compatible with libev. Interaction between \f(CW\*(C`sigprocmask\*(C'\fR and
3667\&\f(CW\*(C`pthread_sigmask\*(C'\fR could complicate things, however.
3668.Sp
3669The most portable way to handle signals is to block signals in all threads
3670except the initial one, and run the default loop in the initial thread as
3671well.
3672.ie n .IP """long"" must be large enough for common memory allocation sizes" 4
3673.el .IP "\f(CWlong\fR must be large enough for common memory allocation sizes" 4
3674.IX Item "long must be large enough for common memory allocation sizes"
3675To improve portability and simplify its \s-1API\s0, libev uses \f(CW\*(C`long\*(C'\fR internally
3676instead of \f(CW\*(C`size_t\*(C'\fR when allocating its data structures. On non-POSIX
3677systems (Microsoft...) this might be unexpectedly low, but is still at
3678least 31 bits everywhere, which is enough for hundreds of millions of
3679watchers.
3680.ie n .IP """double"" must hold a time value in seconds with enough accuracy" 4
3681.el .IP "\f(CWdouble\fR must hold a time value in seconds with enough accuracy" 4
3682.IX Item "double must hold a time value in seconds with enough accuracy"
3683The type \f(CW\*(C`double\*(C'\fR is used to represent timestamps. It is required to
3684have at least 51 bits of mantissa (and 9 bits of exponent), which is good
3685enough for at least into the year 4000. This requirement is fulfilled by
3686implementations implementing \s-1IEEE\s0 754 (basically all existing ones).
3687.PP
3688If you know of other additional requirements drop me a note.
3689.SH "ALGORITHMIC COMPLEXITIES"
3690.IX Header "ALGORITHMIC COMPLEXITIES"
3691In this section the complexities of (many of) the algorithms used inside
3692libev will be documented. For complexity discussions about backends see
3693the documentation for \f(CW\*(C`ev_default_init\*(C'\fR.
3694.PP
3695All of the following are about amortised time: If an array needs to be
3696extended, libev needs to realloc and move the whole array, but this
3697happens asymptotically rarer with higher number of elements, so O(1) might
3698mean that libev does a lengthy realloc operation in rare cases, but on
3699average it is much faster and asymptotically approaches constant time.
3700.IP "Starting and stopping timer/periodic watchers: O(log skipped_other_timers)" 4
3701.IX Item "Starting and stopping timer/periodic watchers: O(log skipped_other_timers)"
3702This means that, when you have a watcher that triggers in one hour and
3703there are 100 watchers that would trigger before that, then inserting will
3704have to skip roughly seven (\f(CW\*(C`ld 100\*(C'\fR) of these watchers.
3705.IP "Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)" 4
3706.IX Item "Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)"
3707That means that changing a timer costs less than removing/adding them,
3708as only the relative motion in the event queue has to be paid for.
3709.IP "Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)" 4
3710.IX Item "Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)"
3711These just add the watcher into an array or at the head of a list.
3712.IP "Stopping check/prepare/idle/fork/async watchers: O(1)" 4
3713.IX Item "Stopping check/prepare/idle/fork/async watchers: O(1)"
3714.PD 0
3715.IP "Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % \s-1EV_PID_HASHSIZE\s0))" 4
3716.IX Item "Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))"
3717.PD
3718These watchers are stored in lists, so they need to be walked to find the
3719correct watcher to remove. The lists are usually short (you don't usually
3720have many watchers waiting for the same fd or signal: one is typical, two
3721is rare).
3722.IP "Finding the next timer in each loop iteration: O(1)" 4
3723.IX Item "Finding the next timer in each loop iteration: O(1)"
3724By virtue of using a binary or 4\-heap, the next timer is always found at a
3725fixed position in the storage array.
3726.IP "Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)" 4
3727.IX Item "Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)"
3728A change means an I/O watcher gets started or stopped, which requires
3729libev to recalculate its status (and possibly tell the kernel, depending
3730on backend and whether \f(CW\*(C`ev_io_set\*(C'\fR was used).
3731.IP "Activating one watcher (putting it into the pending state): O(1)" 4
3732.IX Item "Activating one watcher (putting it into the pending state): O(1)"
3733.PD 0
3734.IP "Priority handling: O(number_of_priorities)" 4
3735.IX Item "Priority handling: O(number_of_priorities)"
3736.PD
3737Priorities are implemented by allocating some space for each
3738priority. When doing priority-based operations, libev usually has to
3739linearly search all the priorities, but starting/stopping and activating
3740watchers becomes O(1) with respect to priority handling.
3741.IP "Sending an ev_async: O(1)" 4
3742.IX Item "Sending an ev_async: O(1)"
3743.PD 0
3744.IP "Processing ev_async_send: O(number_of_async_watchers)" 4
3745.IX Item "Processing ev_async_send: O(number_of_async_watchers)"
3746.IP "Processing signals: O(max_signal_number)" 4
3747.IX Item "Processing signals: O(max_signal_number)"
3748.PD
3749Sending involves a system call \fIiff\fR there were no other \f(CW\*(C`ev_async_send\*(C'\fR
3750calls in the current loop iteration. Checking for async and signal events
3751involves iterating over all running async watchers or all signal numbers.
1020.SH "AUTHOR" 3752.SH "AUTHOR"
1021.IX Header "AUTHOR" 3753.IX Header "AUTHOR"
1022Marc Lehmann <libev@schmorp.de>. 3754Marc Lehmann <libev@schmorp.de>.

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines