ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/AnyEvent/README
(Generate patch)

Comparing AnyEvent/README (file contents):
Revision 1.19 by root, Mon Apr 28 08:02:14 2008 UTC vs.
Revision 1.69 by root, Tue Oct 4 17:45:04 2011 UTC

1NAME 1NAME
2 AnyEvent - provide framework for multiple event loops 2 AnyEvent - the DBI of event loop programming
3 3
4 EV, Event, Coro::EV, Coro::Event, Glib, Tk, Perl, Event::Lib, Qt, POE - 4 EV, Event, Glib, Tk, Perl, Event::Lib, Irssi, rxvt-unicode, IO::Async,
5 various supported event loops 5 Qt, FLTK and POE are various supported event loops/environments.
6 6
7SYNOPSIS 7SYNOPSIS
8 use AnyEvent; 8 use AnyEvent;
9 9
10 # if you prefer function calls, look at the AE manpage for
11 # an alternative API.
12
13 # file handle or descriptor readable
10 my $w = AnyEvent->io (fh => $fh, poll => "r|w", cb => sub { 14 my $w = AnyEvent->io (fh => $fh, poll => "r", cb => sub { ... });
15
16 # one-shot or repeating timers
17 my $w = AnyEvent->timer (after => $seconds, cb => sub { ... });
18 my $w = AnyEvent->timer (after => $seconds, interval => $seconds, cb => ...);
19
20 print AnyEvent->now; # prints current event loop time
21 print AnyEvent->time; # think Time::HiRes::time or simply CORE::time.
22
23 # POSIX signal
24 my $w = AnyEvent->signal (signal => "TERM", cb => sub { ... });
25
26 # child process exit
27 my $w = AnyEvent->child (pid => $pid, cb => sub {
28 my ($pid, $status) = @_;
11 ... 29 ...
12 }); 30 });
13 31
14 my $w = AnyEvent->timer (after => $seconds, cb => sub { 32 # called when event loop idle (if applicable)
15 ... 33 my $w = AnyEvent->idle (cb => sub { ... });
16 });
17 34
18 my $w = AnyEvent->condvar; # stores whether a condition was flagged 35 my $w = AnyEvent->condvar; # stores whether a condition was flagged
36 $w->send; # wake up current and all future recv's
19 $w->wait; # enters "main loop" till $condvar gets ->broadcast 37 $w->recv; # enters "main loop" till $condvar gets ->send
20 $w->broadcast; # wake up current and all future wait's 38 # use a condvar in callback mode:
39 $w->cb (sub { $_[0]->recv });
40
41INTRODUCTION/TUTORIAL
42 This manpage is mainly a reference manual. If you are interested in a
43 tutorial or some gentle introduction, have a look at the AnyEvent::Intro
44 manpage.
45
46SUPPORT
47 An FAQ document is available as AnyEvent::FAQ.
48
49 There also is a mailinglist for discussing all things AnyEvent, and an
50 IRC channel, too.
51
52 See the AnyEvent project page at the Schmorpforge Ta-Sa Software
53 Repository, at <http://anyevent.schmorp.de>, for more info.
21 54
22WHY YOU SHOULD USE THIS MODULE (OR NOT) 55WHY YOU SHOULD USE THIS MODULE (OR NOT)
23 Glib, POE, IO::Async, Event... CPAN offers event models by the dozen 56 Glib, POE, IO::Async, Event... CPAN offers event models by the dozen
24 nowadays. So what is different about AnyEvent? 57 nowadays. So what is different about AnyEvent?
25 58
26 Executive Summary: AnyEvent is *compatible*, AnyEvent is *free of 59 Executive Summary: AnyEvent is *compatible*, AnyEvent is *free of
27 policy* and AnyEvent is *small and efficient*. 60 policy* and AnyEvent is *small and efficient*.
28 61
29 First and foremost, *AnyEvent is not an event model* itself, it only 62 First and foremost, *AnyEvent is not an event model* itself, it only
30 interfaces to whatever event model the main program happens to use in a 63 interfaces to whatever event model the main program happens to use, in a
31 pragmatic way. For event models and certain classes of immortals alike, 64 pragmatic way. For event models and certain classes of immortals alike,
32 the statement "there can only be one" is a bitter reality: In general, 65 the statement "there can only be one" is a bitter reality: In general,
33 only one event loop can be active at the same time in a process. 66 only one event loop can be active at the same time in a process.
34 AnyEvent helps hiding the differences between those event loops. 67 AnyEvent cannot change this, but it can hide the differences between
68 those event loops.
35 69
36 The goal of AnyEvent is to offer module authors the ability to do event 70 The goal of AnyEvent is to offer module authors the ability to do event
37 programming (waiting for I/O or timer events) without subscribing to a 71 programming (waiting for I/O or timer events) without subscribing to a
38 religion, a way of living, and most importantly: without forcing your 72 religion, a way of living, and most importantly: without forcing your
39 module users into the same thing by forcing them to use the same event 73 module users into the same thing by forcing them to use the same event
40 model you use. 74 model you use.
41 75
42 For modules like POE or IO::Async (which is a total misnomer as it is 76 For modules like POE or IO::Async (which is a total misnomer as it is
43 actually doing all I/O *synchronously*...), using them in your module is 77 actually doing all I/O *synchronously*...), using them in your module is
44 like joining a cult: After you joined, you are dependent on them and you 78 like joining a cult: After you join, you are dependent on them and you
45 cannot use anything else, as it is simply incompatible to everything 79 cannot use anything else, as they are simply incompatible to everything
46 that isn't itself. What's worse, all the potential users of your module 80 that isn't them. What's worse, all the potential users of your module
47 are *also* forced to use the same event loop you use. 81 are *also* forced to use the same event loop you use.
48 82
49 AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works 83 AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works
50 fine. AnyEvent + Tk works fine etc. etc. but none of these work together 84 fine. AnyEvent + Tk works fine etc. etc. but none of these work together
51 with the rest: POE + IO::Async? no go. Tk + Event? no go. Again: if your 85 with the rest: POE + EV? No go. Tk + Event? No go. Again: if your module
52 module uses one of those, every user of your module has to use it, too. 86 uses one of those, every user of your module has to use it, too. But if
53 But if your module uses AnyEvent, it works transparently with all event 87 your module uses AnyEvent, it works transparently with all event models
54 models it supports (including stuff like POE and IO::Async, as long as 88 it supports (including stuff like IO::Async, as long as those use one of
55 those use one of the supported event loops. It is trivial to add new 89 the supported event loops. It is easy to add new event loops to
56 event loops to AnyEvent, too, so it is future-proof). 90 AnyEvent, too, so it is future-proof).
57 91
58 In addition to being free of having to use *the one and only true event 92 In addition to being free of having to use *the one and only true event
59 model*, AnyEvent also is free of bloat and policy: with POE or similar 93 model*, AnyEvent also is free of bloat and policy: with POE or similar
60 modules, you get an enourmous amount of code and strict rules you have 94 modules, you get an enormous amount of code and strict rules you have to
61 to follow. AnyEvent, on the other hand, is lean and up to the point, by 95 follow. AnyEvent, on the other hand, is lean and to the point, by only
62 only offering the functionality that is necessary, in as thin as a 96 offering the functionality that is necessary, in as thin as a wrapper as
63 wrapper as technically possible. 97 technically possible.
64 98
99 Of course, AnyEvent comes with a big (and fully optional!) toolbox of
100 useful functionality, such as an asynchronous DNS resolver, 100%
101 non-blocking connects (even with TLS/SSL, IPv6 and on broken platforms
102 such as Windows) and lots of real-world knowledge and workarounds for
103 platform bugs and differences.
104
65 Of course, if you want lots of policy (this can arguably be somewhat 105 Now, if you *do want* lots of policy (this can arguably be somewhat
66 useful) and you want to force your users to use the one and only event 106 useful) and you want to force your users to use the one and only event
67 model, you should *not* use this module. 107 model, you should *not* use this module.
68 108
69DESCRIPTION 109DESCRIPTION
70 AnyEvent provides an identical interface to multiple event loops. This 110 AnyEvent provides a uniform interface to various event loops. This
71 allows module authors to utilise an event loop without forcing module 111 allows module authors to use event loop functionality without forcing
72 users to use the same event loop (as only a single event loop can 112 module users to use a specific event loop implementation (since more
73 coexist peacefully at any one time). 113 than one event loop cannot coexist peacefully).
74 114
75 The interface itself is vaguely similar, but not identical to the Event 115 The interface itself is vaguely similar, but not identical to the Event
76 module. 116 module.
77 117
78 During the first call of any watcher-creation method, the module tries 118 During the first call of any watcher-creation method, the module tries
79 to detect the currently loaded event loop by probing whether one of the 119 to detect the currently loaded event loop by probing whether one of the
80 following modules is already loaded: Coro::EV, Coro::Event, EV, Event, 120 following modules is already loaded: EV, AnyEvent::Loop, Event, Glib,
81 Glib, AnyEvent::Impl::Perl, Tk, Event::Lib, Qt, POE. The first one found 121 Tk, Event::Lib, Qt, POE. The first one found is used. If none are
82 is used. If none are found, the module tries to load these modules 122 detected, the module tries to load the first four modules in the order
83 (excluding Tk, Event::Lib, Qt and POE as the pure perl adaptor should 123 given; but note that if EV is not available, the pure-perl
84 always succeed) in the order given. The first one that can be 124 AnyEvent::Loop should always work, so the other two are not normally
85 successfully loaded will be used. If, after this, still none could be 125 tried.
86 found, AnyEvent will fall back to a pure-perl event loop, which is not
87 very efficient, but should work everywhere.
88 126
89 Because AnyEvent first checks for modules that are already loaded, 127 Because AnyEvent first checks for modules that are already loaded,
90 loading an event model explicitly before first using AnyEvent will 128 loading an event model explicitly before first using AnyEvent will
91 likely make that model the default. For example: 129 likely make that model the default. For example:
92 130
… …
94 use AnyEvent; 132 use AnyEvent;
95 133
96 # .. AnyEvent will likely default to Tk 134 # .. AnyEvent will likely default to Tk
97 135
98 The *likely* means that, if any module loads another event model and 136 The *likely* means that, if any module loads another event model and
99 starts using it, all bets are off. Maybe you should tell their authors 137 starts using it, all bets are off - this case should be very rare
100 to use AnyEvent so their modules work together with others seamlessly... 138 though, as very few modules hardcode event loops without announcing this
139 very loudly.
101 140
102 The pure-perl implementation of AnyEvent is called 141 The pure-perl implementation of AnyEvent is called "AnyEvent::Loop".
103 "AnyEvent::Impl::Perl". Like other event modules you can load it 142 Like other event modules you can load it explicitly and enjoy the high
104 explicitly. 143 availability of that event loop :)
105 144
106WATCHERS 145WATCHERS
107 AnyEvent has the central concept of a *watcher*, which is an object that 146 AnyEvent has the central concept of a *watcher*, which is an object that
108 stores relevant data for each kind of event you are waiting for, such as 147 stores relevant data for each kind of event you are waiting for, such as
109 the callback to call, the filehandle to watch, etc. 148 the callback to call, the file handle to watch, etc.
110 149
111 These watchers are normal Perl objects with normal Perl lifetime. After 150 These watchers are normal Perl objects with normal Perl lifetime. After
112 creating a watcher it will immediately "watch" for events and invoke the 151 creating a watcher it will immediately "watch" for events and invoke the
113 callback when the event occurs (of course, only when the event model is 152 callback when the event occurs (of course, only when the event model is
114 in control). 153 in control).
115 154
155 Note that callbacks must not permanently change global variables
156 potentially in use by the event loop (such as $_ or $[) and that
157 callbacks must not "die". The former is good programming practice in
158 Perl and the latter stems from the fact that exception handling differs
159 widely between event loops.
160
116 To disable the watcher you have to destroy it (e.g. by setting the 161 To disable a watcher you have to destroy it (e.g. by setting the
117 variable you store it in to "undef" or otherwise deleting all references 162 variable you store it in to "undef" or otherwise deleting all references
118 to it). 163 to it).
119 164
120 All watchers are created by calling a method on the "AnyEvent" class. 165 All watchers are created by calling a method on the "AnyEvent" class.
121 166
122 Many watchers either are used with "recursion" (repeating timers for 167 Many watchers either are used with "recursion" (repeating timers for
123 example), or need to refer to their watcher object in other ways. 168 example), or need to refer to their watcher object in other ways.
124 169
125 An any way to achieve that is this pattern: 170 One way to achieve that is this pattern:
126 171
127 my $w; $w = AnyEvent->type (arg => value ..., cb => sub { 172 my $w; $w = AnyEvent->type (arg => value ..., cb => sub {
128 # you can use $w here, for example to undef it 173 # you can use $w here, for example to undef it
129 undef $w; 174 undef $w;
130 }); 175 });
131 176
132 Note that "my $w; $w =" combination. This is necessary because in Perl, 177 Note that "my $w; $w =" combination. This is necessary because in Perl,
133 my variables are only visible after the statement in which they are 178 my variables are only visible after the statement in which they are
134 declared. 179 declared.
135 180
136 I/O WATCHERS 181 I/O WATCHERS
182 $w = AnyEvent->io (
183 fh => <filehandle_or_fileno>,
184 poll => <"r" or "w">,
185 cb => <callback>,
186 );
187
137 You can create an I/O watcher by calling the "AnyEvent->io" method with 188 You can create an I/O watcher by calling the "AnyEvent->io" method with
138 the following mandatory key-value pairs as arguments: 189 the following mandatory key-value pairs as arguments:
139 190
140 "fh" the Perl *file handle* (*not* file descriptor) to watch for events. 191 "fh" is the Perl *file handle* (or a naked file descriptor) to watch for
192 events (AnyEvent might or might not keep a reference to this file
193 handle). Note that only file handles pointing to things for which
194 non-blocking operation makes sense are allowed. This includes sockets,
195 most character devices, pipes, fifos and so on, but not for example
196 files or block devices.
197
141 "poll" must be a string that is either "r" or "w", which creates a 198 "poll" must be a string that is either "r" or "w", which creates a
142 watcher waiting for "r"eadable or "w"ritable events, respectively. "cb" 199 watcher waiting for "r"eadable or "w"ritable events, respectively.
200
143 is the callback to invoke each time the file handle becomes ready. 201 "cb" is the callback to invoke each time the file handle becomes ready.
144 202
145 Although the callback might get passed parameters, their value and 203 Although the callback might get passed parameters, their value and
146 presence is undefined and you cannot rely on them. Portable AnyEvent 204 presence is undefined and you cannot rely on them. Portable AnyEvent
147 callbacks cannot use arguments passed to I/O watcher callbacks. 205 callbacks cannot use arguments passed to I/O watcher callbacks.
148 206
149 The I/O watcher might use the underlying file descriptor or a copy of 207 The I/O watcher might use the underlying file descriptor or a copy of
150 it. You must not close a file handle as long as any watcher is active on 208 it. You must not close a file handle as long as any watcher is active on
151 the underlying file descriptor. 209 the underlying file descriptor.
152 210
153 Some event loops issue spurious readyness notifications, so you should 211 Some event loops issue spurious readiness notifications, so you should
154 always use non-blocking calls when reading/writing from/to your file 212 always use non-blocking calls when reading/writing from/to your file
155 handles. 213 handles.
156 214
157 Example:
158
159 # wait for readability of STDIN, then read a line and disable the watcher 215 Example: wait for readability of STDIN, then read a line and disable the
216 watcher.
217
160 my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub { 218 my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
161 chomp (my $input = <STDIN>); 219 chomp (my $input = <STDIN>);
162 warn "read: $input\n"; 220 warn "read: $input\n";
163 undef $w; 221 undef $w;
164 }); 222 });
165 223
166 TIME WATCHERS 224 TIME WATCHERS
225 $w = AnyEvent->timer (after => <seconds>, cb => <callback>);
226
227 $w = AnyEvent->timer (
228 after => <fractional_seconds>,
229 interval => <fractional_seconds>,
230 cb => <callback>,
231 );
232
167 You can create a time watcher by calling the "AnyEvent->timer" method 233 You can create a time watcher by calling the "AnyEvent->timer" method
168 with the following mandatory arguments: 234 with the following mandatory arguments:
169 235
170 "after" specifies after how many seconds (fractional values are 236 "after" specifies after how many seconds (fractional values are
171 supported) the callback should be invoked. "cb" is the callback to 237 supported) the callback should be invoked. "cb" is the callback to
… …
173 239
174 Although the callback might get passed parameters, their value and 240 Although the callback might get passed parameters, their value and
175 presence is undefined and you cannot rely on them. Portable AnyEvent 241 presence is undefined and you cannot rely on them. Portable AnyEvent
176 callbacks cannot use arguments passed to time watcher callbacks. 242 callbacks cannot use arguments passed to time watcher callbacks.
177 243
178 The timer callback will be invoked at most once: if you want a repeating 244 The callback will normally be invoked only once. If you specify another
179 timer you have to create a new watcher (this is a limitation by both Tk 245 parameter, "interval", as a strictly positive number (> 0), then the
180 and Glib). 246 callback will be invoked regularly at that interval (in fractional
247 seconds) after the first invocation. If "interval" is specified with a
248 false value, then it is treated as if it were not specified at all.
181 249
182 Example: 250 The callback will be rescheduled before invoking the callback, but no
251 attempt is made to avoid timer drift in most backends, so the interval
252 is only approximate.
183 253
184 # fire an event after 7.7 seconds 254 Example: fire an event after 7.7 seconds.
255
185 my $w = AnyEvent->timer (after => 7.7, cb => sub { 256 my $w = AnyEvent->timer (after => 7.7, cb => sub {
186 warn "timeout\n"; 257 warn "timeout\n";
187 }); 258 });
188 259
189 # to cancel the timer: 260 # to cancel the timer:
190 undef $w; 261 undef $w;
191 262
192 Example 2:
193
194 # fire an event after 0.5 seconds, then roughly every second 263 Example 2: fire an event after 0.5 seconds, then roughly every second.
195 my $w;
196 264
197 my $cb = sub {
198 # cancel the old timer while creating a new one
199 $w = AnyEvent->timer (after => 1, cb => $cb); 265 my $w = AnyEvent->timer (after => 0.5, interval => 1, cb => sub {
266 warn "timeout\n";
200 }; 267 };
201
202 # start the "loop" by creating the first watcher
203 $w = AnyEvent->timer (after => 0.5, cb => $cb);
204 268
205 TIMING ISSUES 269 TIMING ISSUES
206 There are two ways to handle timers: based on real time (relative, "fire 270 There are two ways to handle timers: based on real time (relative, "fire
207 in 10 seconds") and based on wallclock time (absolute, "fire at 12 271 in 10 seconds") and based on wallclock time (absolute, "fire at 12
208 o'clock"). 272 o'clock").
209 273
210 While most event loops expect timers to specified in a relative way, 274 While most event loops expect timers to specified in a relative way,
211 they use absolute time internally. This makes a difference when your 275 they use absolute time internally. This makes a difference when your
212 clock "jumps", for example, when ntp decides to set your clock backwards 276 clock "jumps", for example, when ntp decides to set your clock backwards
213 from the wrong date of 2014-01-01 to 2008-01-01, a watcher that is 277 from the wrong date of 2014-01-01 to 2008-01-01, a watcher that is
214 supposed to fire "after" a second might actually take six years to 278 supposed to fire "after a second" might actually take six years to
215 finally fire. 279 finally fire.
216 280
217 AnyEvent cannot compensate for this. The only event loop that is 281 AnyEvent cannot compensate for this. The only event loop that is
218 conscious about these issues is EV, which offers both relative 282 conscious of these issues is EV, which offers both relative (ev_timer,
219 (ev_timer, based on true relative time) and absolute (ev_periodic, based 283 based on true relative time) and absolute (ev_periodic, based on
220 on wallclock time) timers. 284 wallclock time) timers.
221 285
222 AnyEvent always prefers relative timers, if available, matching the 286 AnyEvent always prefers relative timers, if available, matching the
223 AnyEvent API. 287 AnyEvent API.
224 288
289 AnyEvent has two additional methods that return the "current time":
290
291 AnyEvent->time
292 This returns the "current wallclock time" as a fractional number of
293 seconds since the Epoch (the same thing as "time" or
294 "Time::HiRes::time" return, and the result is guaranteed to be
295 compatible with those).
296
297 It progresses independently of any event loop processing, i.e. each
298 call will check the system clock, which usually gets updated
299 frequently.
300
301 AnyEvent->now
302 This also returns the "current wallclock time", but unlike "time",
303 above, this value might change only once per event loop iteration,
304 depending on the event loop (most return the same time as "time",
305 above). This is the time that AnyEvent's timers get scheduled
306 against.
307
308 *In almost all cases (in all cases if you don't care), this is the
309 function to call when you want to know the current time.*
310
311 This function is also often faster then "AnyEvent->time", and thus
312 the preferred method if you want some timestamp (for example,
313 AnyEvent::Handle uses this to update its activity timeouts).
314
315 The rest of this section is only of relevance if you try to be very
316 exact with your timing; you can skip it without a bad conscience.
317
318 For a practical example of when these times differ, consider
319 Event::Lib and EV and the following set-up:
320
321 The event loop is running and has just invoked one of your callbacks
322 at time=500 (assume no other callbacks delay processing). In your
323 callback, you wait a second by executing "sleep 1" (blocking the
324 process for a second) and then (at time=501) you create a relative
325 timer that fires after three seconds.
326
327 With Event::Lib, "AnyEvent->time" and "AnyEvent->now" will both
328 return 501, because that is the current time, and the timer will be
329 scheduled to fire at time=504 (501 + 3).
330
331 With EV, "AnyEvent->time" returns 501 (as that is the current time),
332 but "AnyEvent->now" returns 500, as that is the time the last event
333 processing phase started. With EV, your timer gets scheduled to run
334 at time=503 (500 + 3).
335
336 In one sense, Event::Lib is more exact, as it uses the current time
337 regardless of any delays introduced by event processing. However,
338 most callbacks do not expect large delays in processing, so this
339 causes a higher drift (and a lot more system calls to get the
340 current time).
341
342 In another sense, EV is more exact, as your timer will be scheduled
343 at the same time, regardless of how long event processing actually
344 took.
345
346 In either case, if you care (and in most cases, you don't), then you
347 can get whatever behaviour you want with any event loop, by taking
348 the difference between "AnyEvent->time" and "AnyEvent->now" into
349 account.
350
351 AnyEvent->now_update
352 Some event loops (such as EV or AnyEvent::Loop) cache the current
353 time for each loop iteration (see the discussion of AnyEvent->now,
354 above).
355
356 When a callback runs for a long time (or when the process sleeps),
357 then this "current" time will differ substantially from the real
358 time, which might affect timers and time-outs.
359
360 When this is the case, you can call this method, which will update
361 the event loop's idea of "current time".
362
363 A typical example would be a script in a web server (e.g.
364 "mod_perl") - when mod_perl executes the script, then the event loop
365 will have the wrong idea about the "current time" (being potentially
366 far in the past, when the script ran the last time). In that case
367 you should arrange a call to "AnyEvent->now_update" each time the
368 web server process wakes up again (e.g. at the start of your script,
369 or in a handler).
370
371 Note that updating the time *might* cause some events to be handled.
372
225 SIGNAL WATCHERS 373 SIGNAL WATCHERS
374 $w = AnyEvent->signal (signal => <uppercase_signal_name>, cb => <callback>);
375
226 You can watch for signals using a signal watcher, "signal" is the signal 376 You can watch for signals using a signal watcher, "signal" is the signal
227 *name* without any "SIG" prefix, "cb" is the Perl callback to be invoked 377 *name* in uppercase and without any "SIG" prefix, "cb" is the Perl
228 whenever a signal occurs. 378 callback to be invoked whenever a signal occurs.
229 379
230 Although the callback might get passed parameters, their value and 380 Although the callback might get passed parameters, their value and
231 presence is undefined and you cannot rely on them. Portable AnyEvent 381 presence is undefined and you cannot rely on them. Portable AnyEvent
232 callbacks cannot use arguments passed to signal watcher callbacks. 382 callbacks cannot use arguments passed to signal watcher callbacks.
233 383
234 Multiple signal occurances can be clumped together into one callback 384 Multiple signal occurrences can be clumped together into one callback
235 invocation, and callback invocation will be synchronous. synchronous 385 invocation, and callback invocation will be synchronous. Synchronous
236 means that it might take a while until the signal gets handled by the 386 means that it might take a while until the signal gets handled by the
237 process, but it is guarenteed not to interrupt any other callbacks. 387 process, but it is guaranteed not to interrupt any other callbacks.
238 388
239 The main advantage of using these watchers is that you can share a 389 The main advantage of using these watchers is that you can share a
240 signal between multiple watchers. 390 signal between multiple watchers, and AnyEvent will ensure that signals
391 will not interrupt your program at bad times.
241 392
242 This watcher might use %SIG, so programs overwriting those signals 393 This watcher might use %SIG (depending on the event loop used), so
243 directly will likely not work correctly. 394 programs overwriting those signals directly will likely not work
395 correctly.
244 396
245 Example: exit on SIGINT 397 Example: exit on SIGINT
246 398
247 my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 }); 399 my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 });
248 400
401 Restart Behaviour
402 While restart behaviour is up to the event loop implementation, most
403 will not restart syscalls (that includes Async::Interrupt and AnyEvent's
404 pure perl implementation).
405
406 Safe/Unsafe Signals
407 Perl signals can be either "safe" (synchronous to opcode handling) or
408 "unsafe" (asynchronous) - the former might delay signal delivery
409 indefinitely, the latter might corrupt your memory.
410
411 AnyEvent signal handlers are, in addition, synchronous to the event
412 loop, i.e. they will not interrupt your running perl program but will
413 only be called as part of the normal event handling (just like timer,
414 I/O etc. callbacks, too).
415
416 Signal Races, Delays and Workarounds
417 Many event loops (e.g. Glib, Tk, Qt, IO::Async) do not support attaching
418 callbacks to signals in a generic way, which is a pity, as you cannot do
419 race-free signal handling in perl, requiring C libraries for this.
420 AnyEvent will try to do its best, which means in some cases, signals
421 will be delayed. The maximum time a signal might be delayed is 10
422 seconds by default, but can be overriden via
423 $ENV{PERL_ANYEVENT_MAX_SIGNAL_LATENCY} or $AnyEvent::MAX_SIGNAL_LATENCY
424 - see the Ö<ENVIRONMENT VARIABLES> section for details.
425
426 All these problems can be avoided by installing the optional
427 Async::Interrupt module, which works with most event loops. It will not
428 work with inherently broken event loops such as Event or Event::Lib (and
429 not with POE currently). For those, you just have to suffer the delays.
430
249 CHILD PROCESS WATCHERS 431 CHILD PROCESS WATCHERS
432 $w = AnyEvent->child (pid => <process id>, cb => <callback>);
433
250 You can also watch on a child process exit and catch its exit status. 434 You can also watch for a child process exit and catch its exit status.
251 435
252 The child process is specified by the "pid" argument (if set to 0, it 436 The child process is specified by the "pid" argument (on some backends,
253 watches for any child process exit). The watcher will trigger as often 437 using 0 watches for any child process exit, on others this will croak).
254 as status change for the child are received. This works by installing a 438 The watcher will be triggered only when the child process has finished
255 signal handler for "SIGCHLD". The callback will be called with the pid 439 and an exit status is available, not on any trace events
256 and exit status (as returned by waitpid), so unlike other watcher types, 440 (stopped/continued).
257 you *can* rely on child watcher callback arguments. 441
442 The callback will be called with the pid and exit status (as returned by
443 waitpid), so unlike other watcher types, you *can* rely on child watcher
444 callback arguments.
445
446 This watcher type works by installing a signal handler for "SIGCHLD",
447 and since it cannot be shared, nothing else should use SIGCHLD or reap
448 random child processes (waiting for specific child processes, e.g.
449 inside "system", is just fine).
258 450
259 There is a slight catch to child watchers, however: you usually start 451 There is a slight catch to child watchers, however: you usually start
260 them *after* the child process was created, and this means the process 452 them *after* the child process was created, and this means the process
261 could have exited already (and no SIGCHLD will be sent anymore). 453 could have exited already (and no SIGCHLD will be sent anymore).
262 454
263 Not all event models handle this correctly (POE doesn't), but even for 455 Not all event models handle this correctly (neither POE nor IO::Async
456 do, see their AnyEvent::Impl manpages for details), but even for event
264 event models that *do* handle this correctly, they usually need to be 457 models that *do* handle this correctly, they usually need to be loaded
265 loaded before the process exits (i.e. before you fork in the first 458 before the process exits (i.e. before you fork in the first place).
266 place). 459 AnyEvent's pure perl event loop handles all cases correctly regardless
460 of when you start the watcher.
267 461
268 This means you cannot create a child watcher as the very first thing in 462 This means you cannot create a child watcher as the very first thing in
269 an AnyEvent program, you *have* to create at least one watcher before 463 an AnyEvent program, you *have* to create at least one watcher before
270 you "fork" the child (alternatively, you can call "AnyEvent::detect"). 464 you "fork" the child (alternatively, you can call "AnyEvent::detect").
271 465
466 As most event loops do not support waiting for child events, they will
467 be emulated by AnyEvent in most cases, in which case the latency and
468 race problems mentioned in the description of signal watchers apply.
469
272 Example: fork a process and wait for it 470 Example: fork a process and wait for it
273 471
274 my $done = AnyEvent->condvar; 472 my $done = AnyEvent->condvar;
275 473
276 AnyEvent::detect; # force event module to be initialised
277
278 my $pid = fork or exit 5; 474 my $pid = fork or exit 5;
279 475
280 my $w = AnyEvent->child ( 476 my $w = AnyEvent->child (
281 pid => $pid, 477 pid => $pid,
282 cb => sub { 478 cb => sub {
283 my ($pid, $status) = @_; 479 my ($pid, $status) = @_;
284 warn "pid $pid exited with status $status"; 480 warn "pid $pid exited with status $status";
285 $done->broadcast; 481 $done->send;
286 }, 482 },
287 ); 483 );
288 484
289 # do something else, then wait for process exit 485 # do something else, then wait for process exit
290 $done->wait; 486 $done->recv;
487
488 IDLE WATCHERS
489 $w = AnyEvent->idle (cb => <callback>);
490
491 This will repeatedly invoke the callback after the process becomes idle,
492 until either the watcher is destroyed or new events have been detected.
493
494 Idle watchers are useful when there is a need to do something, but it is
495 not so important (or wise) to do it instantly. The callback will be
496 invoked only when there is "nothing better to do", which is usually
497 defined as "all outstanding events have been handled and no new events
498 have been detected". That means that idle watchers ideally get invoked
499 when the event loop has just polled for new events but none have been
500 detected. Instead of blocking to wait for more events, the idle watchers
501 will be invoked.
502
503 Unfortunately, most event loops do not really support idle watchers
504 (only EV, Event and Glib do it in a usable fashion) - for the rest,
505 AnyEvent will simply call the callback "from time to time".
506
507 Example: read lines from STDIN, but only process them when the program
508 is otherwise idle:
509
510 my @lines; # read data
511 my $idle_w;
512 my $io_w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
513 push @lines, scalar <STDIN>;
514
515 # start an idle watcher, if not already done
516 $idle_w ||= AnyEvent->idle (cb => sub {
517 # handle only one line, when there are lines left
518 if (my $line = shift @lines) {
519 print "handled when idle: $line";
520 } else {
521 # otherwise disable the idle watcher again
522 undef $idle_w;
523 }
524 });
525 });
291 526
292 CONDITION VARIABLES 527 CONDITION VARIABLES
528 $cv = AnyEvent->condvar;
529
530 $cv->send (<list>);
531 my @res = $cv->recv;
532
533 If you are familiar with some event loops you will know that all of them
534 require you to run some blocking "loop", "run" or similar function that
535 will actively watch for new events and call your callbacks.
536
537 AnyEvent is slightly different: it expects somebody else to run the
538 event loop and will only block when necessary (usually when told by the
539 user).
540
541 The tool to do that is called a "condition variable", so called because
542 they represent a condition that must become true.
543
544 Now is probably a good time to look at the examples further below.
545
293 Condition variables can be created by calling the "AnyEvent->condvar" 546 Condition variables can be created by calling the "AnyEvent->condvar"
294 method without any arguments. 547 method, usually without arguments. The only argument pair allowed is
548 "cb", which specifies a callback to be called when the condition
549 variable becomes true, with the condition variable as the first argument
550 (but not the results).
295 551
296 A condition variable waits for a condition - precisely that the 552 After creation, the condition variable is "false" until it becomes
297 "->broadcast" method has been called. 553 "true" by calling the "send" method (or calling the condition variable
554 as if it were a callback, read about the caveats in the description for
555 the "->send" method).
298 556
299 They are very useful to signal that a condition has been fulfilled, for 557 Since condition variables are the most complex part of the AnyEvent API,
558 here are some different mental models of what they are - pick the ones
559 you can connect to:
560
561 * Condition variables are like callbacks - you can call them (and pass
562 them instead of callbacks). Unlike callbacks however, you can also
563 wait for them to be called.
564
565 * Condition variables are signals - one side can emit or send them,
566 the other side can wait for them, or install a handler that is
567 called when the signal fires.
568
569 * Condition variables are like "Merge Points" - points in your program
570 where you merge multiple independent results/control flows into one.
571
572 * Condition variables represent a transaction - functions that start
573 some kind of transaction can return them, leaving the caller the
574 choice between waiting in a blocking fashion, or setting a callback.
575
576 * Condition variables represent future values, or promises to deliver
577 some result, long before the result is available.
578
579 Condition variables are very useful to signal that something has
300 example, if you write a module that does asynchronous http requests, 580 finished, for example, if you write a module that does asynchronous http
301 then a condition variable would be the ideal candidate to signal the 581 requests, then a condition variable would be the ideal candidate to
302 availability of results. 582 signal the availability of results. The user can either act when the
583 callback is called or can synchronously "->recv" for the results.
303 584
304 You can also use condition variables to block your main program until an 585 You can also use them to simulate traditional event loops - for example,
305 event occurs - for example, you could "->wait" in your main program 586 you can block your main program until an event occurs - for example, you
306 until the user clicks the Quit button in your app, which would 587 could "->recv" in your main program until the user clicks the Quit
307 "->broadcast" the "quit" event. 588 button of your app, which would "->send" the "quit" event.
308 589
309 Note that condition variables recurse into the event loop - if you have 590 Note that condition variables recurse into the event loop - if you have
310 two pirces of code that call "->wait" in a round-robbin fashion, you 591 two pieces of code that call "->recv" in a round-robin fashion, you
311 lose. Therefore, condition variables are good to export to your caller, 592 lose. Therefore, condition variables are good to export to your caller,
312 but you should avoid making a blocking wait yourself, at least in 593 but you should avoid making a blocking wait yourself, at least in
313 callbacks, as this asks for trouble. 594 callbacks, as this asks for trouble.
314 595
315 This object has two methods: 596 Condition variables are represented by hash refs in perl, and the keys
597 used by AnyEvent itself are all named "_ae_XXX" to make subclassing easy
598 (it is often useful to build your own transaction class on top of
599 AnyEvent). To subclass, use "AnyEvent::CondVar" as base class and call
600 its "new" method in your own "new" method.
316 601
317 $cv->wait 602 There are two "sides" to a condition variable - the "producer side"
318 Wait (blocking if necessary) until the "->broadcast" method has been 603 which eventually calls "-> send", and the "consumer side", which waits
319 called on c<$cv>, while servicing other watchers normally. 604 for the send to occur.
320 605
321 You can only wait once on a condition - additional calls will return 606 Example: wait for a timer.
322 immediately.
323 607
324 Not all event models support a blocking wait - some die in that case 608 # condition: "wait till the timer is fired"
325 (programs might want to do that to stay interactive), so *if you are
326 using this from a module, never require a blocking wait*, but let
327 the caller decide whether the call will block or not (for example,
328 by coupling condition variables with some kind of request results
329 and supporting callbacks so the caller knows that getting the result
330 will not block, while still suppporting blocking waits if the caller
331 so desires).
332
333 Another reason *never* to "->wait" in a module is that you cannot
334 sensibly have two "->wait"'s in parallel, as that would require
335 multiple interpreters or coroutines/threads, none of which
336 "AnyEvent" can supply (the coroutine-aware backends
337 AnyEvent::Impl::CoroEV and AnyEvent::Impl::CoroEvent explicitly
338 support concurrent "->wait"'s from different coroutines, however).
339
340 $cv->broadcast
341 Flag the condition as ready - a running "->wait" and all further
342 calls to "wait" will (eventually) return after this method has been
343 called. If nobody is waiting the broadcast will be remembered..
344
345 Example:
346
347 # wait till the result is ready
348 my $result_ready = AnyEvent->condvar; 609 my $timer_fired = AnyEvent->condvar;
349 610
350 # do something such as adding a timer 611 # create the timer - we could wait for, say
351 # or socket watcher the calls $result_ready->broadcast 612 # a handle becomign ready, or even an
352 # when the "result" is ready. 613 # AnyEvent::HTTP request to finish, but
353 # in this case, we simply use a timer: 614 # in this case, we simply use a timer:
354 my $w = AnyEvent->timer ( 615 my $w = AnyEvent->timer (
355 after => 1, 616 after => 1,
356 cb => sub { $result_ready->broadcast }, 617 cb => sub { $timer_fired->send },
357 ); 618 );
358 619
359 # this "blocks" (while handling events) till the watcher 620 # this "blocks" (while handling events) till the callback
360 # calls broadcast 621 # calls ->send
361 $result_ready->wait; 622 $timer_fired->recv;
623
624 Example: wait for a timer, but take advantage of the fact that condition
625 variables are also callable directly.
626
627 my $done = AnyEvent->condvar;
628 my $delay = AnyEvent->timer (after => 5, cb => $done);
629 $done->recv;
630
631 Example: Imagine an API that returns a condvar and doesn't support
632 callbacks. This is how you make a synchronous call, for example from the
633 main program:
634
635 use AnyEvent::CouchDB;
636
637 ...
638
639 my @info = $couchdb->info->recv;
640
641 And this is how you would just set a callback to be called whenever the
642 results are available:
643
644 $couchdb->info->cb (sub {
645 my @info = $_[0]->recv;
646 });
647
648 METHODS FOR PRODUCERS
649 These methods should only be used by the producing side, i.e. the
650 code/module that eventually sends the signal. Note that it is also the
651 producer side which creates the condvar in most cases, but it isn't
652 uncommon for the consumer to create it as well.
653
654 $cv->send (...)
655 Flag the condition as ready - a running "->recv" and all further
656 calls to "recv" will (eventually) return after this method has been
657 called. If nobody is waiting the send will be remembered.
658
659 If a callback has been set on the condition variable, it is called
660 immediately from within send.
661
662 Any arguments passed to the "send" call will be returned by all
663 future "->recv" calls.
664
665 Condition variables are overloaded so one can call them directly (as
666 if they were a code reference). Calling them directly is the same as
667 calling "send".
668
669 $cv->croak ($error)
670 Similar to send, but causes all calls to "->recv" to invoke
671 "Carp::croak" with the given error message/object/scalar.
672
673 This can be used to signal any errors to the condition variable
674 user/consumer. Doing it this way instead of calling "croak" directly
675 delays the error detection, but has the overwhelming advantage that
676 it diagnoses the error at the place where the result is expected,
677 and not deep in some event callback with no connection to the actual
678 code causing the problem.
679
680 $cv->begin ([group callback])
681 $cv->end
682 These two methods can be used to combine many transactions/events
683 into one. For example, a function that pings many hosts in parallel
684 might want to use a condition variable for the whole process.
685
686 Every call to "->begin" will increment a counter, and every call to
687 "->end" will decrement it. If the counter reaches 0 in "->end", the
688 (last) callback passed to "begin" will be executed, passing the
689 condvar as first argument. That callback is *supposed* to call
690 "->send", but that is not required. If no group callback was set,
691 "send" will be called without any arguments.
692
693 You can think of "$cv->send" giving you an OR condition (one call
694 sends), while "$cv->begin" and "$cv->end" giving you an AND
695 condition (all "begin" calls must be "end"'ed before the condvar
696 sends).
697
698 Let's start with a simple example: you have two I/O watchers (for
699 example, STDOUT and STDERR for a program), and you want to wait for
700 both streams to close before activating a condvar:
701
702 my $cv = AnyEvent->condvar;
703
704 $cv->begin; # first watcher
705 my $w1 = AnyEvent->io (fh => $fh1, cb => sub {
706 defined sysread $fh1, my $buf, 4096
707 or $cv->end;
708 });
709
710 $cv->begin; # second watcher
711 my $w2 = AnyEvent->io (fh => $fh2, cb => sub {
712 defined sysread $fh2, my $buf, 4096
713 or $cv->end;
714 });
715
716 $cv->recv;
717
718 This works because for every event source (EOF on file handle),
719 there is one call to "begin", so the condvar waits for all calls to
720 "end" before sending.
721
722 The ping example mentioned above is slightly more complicated, as
723 the there are results to be passwd back, and the number of tasks
724 that are begun can potentially be zero:
725
726 my $cv = AnyEvent->condvar;
727
728 my %result;
729 $cv->begin (sub { shift->send (\%result) });
730
731 for my $host (@list_of_hosts) {
732 $cv->begin;
733 ping_host_then_call_callback $host, sub {
734 $result{$host} = ...;
735 $cv->end;
736 };
737 }
738
739 $cv->end;
740
741 This code fragment supposedly pings a number of hosts and calls
742 "send" after results for all then have have been gathered - in any
743 order. To achieve this, the code issues a call to "begin" when it
744 starts each ping request and calls "end" when it has received some
745 result for it. Since "begin" and "end" only maintain a counter, the
746 order in which results arrive is not relevant.
747
748 There is an additional bracketing call to "begin" and "end" outside
749 the loop, which serves two important purposes: first, it sets the
750 callback to be called once the counter reaches 0, and second, it
751 ensures that "send" is called even when "no" hosts are being pinged
752 (the loop doesn't execute once).
753
754 This is the general pattern when you "fan out" into multiple (but
755 potentially zero) subrequests: use an outer "begin"/"end" pair to
756 set the callback and ensure "end" is called at least once, and then,
757 for each subrequest you start, call "begin" and for each subrequest
758 you finish, call "end".
759
760 METHODS FOR CONSUMERS
761 These methods should only be used by the consuming side, i.e. the code
762 awaits the condition.
763
764 $cv->recv
765 Wait (blocking if necessary) until the "->send" or "->croak" methods
766 have been called on $cv, while servicing other watchers normally.
767
768 You can only wait once on a condition - additional calls are valid
769 but will return immediately.
770
771 If an error condition has been set by calling "->croak", then this
772 function will call "croak".
773
774 In list context, all parameters passed to "send" will be returned,
775 in scalar context only the first one will be returned.
776
777 Note that doing a blocking wait in a callback is not supported by
778 any event loop, that is, recursive invocation of a blocking "->recv"
779 is not allowed, and the "recv" call will "croak" if such a condition
780 is detected. This condition can be slightly loosened by using
781 Coro::AnyEvent, which allows you to do a blocking "->recv" from any
782 thread that doesn't run the event loop itself.
783
784 Not all event models support a blocking wait - some die in that case
785 (programs might want to do that to stay interactive), so *if you are
786 using this from a module, never require a blocking wait*. Instead,
787 let the caller decide whether the call will block or not (for
788 example, by coupling condition variables with some kind of request
789 results and supporting callbacks so the caller knows that getting
790 the result will not block, while still supporting blocking waits if
791 the caller so desires).
792
793 You can ensure that "->recv" never blocks by setting a callback and
794 only calling "->recv" from within that callback (or at a later
795 time). This will work even when the event loop does not support
796 blocking waits otherwise.
797
798 $bool = $cv->ready
799 Returns true when the condition is "true", i.e. whether "send" or
800 "croak" have been called.
801
802 $cb = $cv->cb ($cb->($cv))
803 This is a mutator function that returns the callback set and
804 optionally replaces it before doing so.
805
806 The callback will be called when the condition becomes "true", i.e.
807 when "send" or "croak" are called, with the only argument being the
808 condition variable itself. If the condition is already true, the
809 callback is called immediately when it is set. Calling "recv" inside
810 the callback or at any later time is guaranteed not to block.
811
812SUPPORTED EVENT LOOPS/BACKENDS
813 The available backend classes are (every class has its own manpage):
814
815 Backends that are autoprobed when no other event loop can be found.
816 EV is the preferred backend when no other event loop seems to be in
817 use. If EV is not installed, then AnyEvent will fall back to its own
818 pure-perl implementation, which is available everywhere as it comes
819 with AnyEvent itself.
820
821 AnyEvent::Impl::EV based on EV (interface to libev, best choice).
822 AnyEvent::Impl::Perl pure-perl AnyEvent::Loop, fast and portable.
823
824 Backends that are transparently being picked up when they are used.
825 These will be used if they are already loaded when the first watcher
826 is created, in which case it is assumed that the application is
827 using them. This means that AnyEvent will automatically pick the
828 right backend when the main program loads an event module before
829 anything starts to create watchers. Nothing special needs to be done
830 by the main program.
831
832 AnyEvent::Impl::Event based on Event, very stable, few glitches.
833 AnyEvent::Impl::Glib based on Glib, slow but very stable.
834 AnyEvent::Impl::Tk based on Tk, very broken.
835 AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
836 AnyEvent::Impl::POE based on POE, very slow, some limitations.
837 AnyEvent::Impl::Irssi used when running within irssi.
838 AnyEvent::Impl::IOAsync based on IO::Async.
839 AnyEvent::Impl::Cocoa based on Cocoa::EventLoop.
840 AnyEvent::Impl::FLTK based on FLTK (fltk 2 binding).
841
842 Backends with special needs.
843 Qt requires the Qt::Application to be instantiated first, but will
844 otherwise be picked up automatically. As long as the main program
845 instantiates the application before any AnyEvent watchers are
846 created, everything should just work.
847
848 AnyEvent::Impl::Qt based on Qt.
849
850 Event loops that are indirectly supported via other backends.
851 Some event loops can be supported via other modules:
852
853 There is no direct support for WxWidgets (Wx) or Prima.
854
855 WxWidgets has no support for watching file handles. However, you can
856 use WxWidgets through the POE adaptor, as POE has a Wx backend that
857 simply polls 20 times per second, which was considered to be too
858 horrible to even consider for AnyEvent.
859
860 Prima is not supported as nobody seems to be using it, but it has a
861 POE backend, so it can be supported through POE.
862
863 AnyEvent knows about both Prima and Wx, however, and will try to
864 load POE when detecting them, in the hope that POE will pick them
865 up, in which case everything will be automatic.
362 866
363GLOBAL VARIABLES AND FUNCTIONS 867GLOBAL VARIABLES AND FUNCTIONS
868 These are not normally required to use AnyEvent, but can be useful to
869 write AnyEvent extension modules.
870
364 $AnyEvent::MODEL 871 $AnyEvent::MODEL
365 Contains "undef" until the first watcher is being created. Then it 872 Contains "undef" until the first watcher is being created, before
873 the backend has been autodetected.
874
366 contains the event model that is being used, which is the name of 875 Afterwards it contains the event model that is being used, which is
367 the Perl class implementing the model. This class is usually one of 876 the name of the Perl class implementing the model. This class is
368 the "AnyEvent::Impl:xxx" modules, but can be any other class in the 877 usually one of the "AnyEvent::Impl::xxx" modules, but can be any
369 case AnyEvent has been extended at runtime (e.g. in *rxvt-unicode*). 878 other class in the case AnyEvent has been extended at runtime (e.g.
370 879 in *rxvt-unicode* it will be "urxvt::anyevent").
371 The known classes so far are:
372
373 AnyEvent::Impl::CoroEV based on Coro::EV, best choice.
374 AnyEvent::Impl::CoroEvent based on Coro::Event, second best choice.
375 AnyEvent::Impl::EV based on EV (an interface to libev, best choice).
376 AnyEvent::Impl::Event based on Event, second best choice.
377 AnyEvent::Impl::Glib based on Glib, third-best choice.
378 AnyEvent::Impl::Perl pure-perl implementation, inefficient but portable.
379 AnyEvent::Impl::Tk based on Tk, very bad choice.
380 AnyEvent::Impl::Qt based on Qt, cannot be autoprobed (see its docs).
381 AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
382 AnyEvent::Impl::POE based on POE, not generic enough for full support.
383
384 There is no support for WxWidgets, as WxWidgets has no support for
385 watching file handles. However, you can use WxWidgets through the
386 POE Adaptor, as POE has a Wx backend that simply polls 20 times per
387 second, which was considered to be too horrible to even consider for
388 AnyEvent. Likewise, other POE backends can be used by AnyEvent by
389 using it's adaptor.
390
391 AnyEvent knows about Prima and Wx and will try to use POE when
392 autodetecting them.
393 880
394 AnyEvent::detect 881 AnyEvent::detect
395 Returns $AnyEvent::MODEL, forcing autodetection of the event model 882 Returns $AnyEvent::MODEL, forcing autodetection of the event model
396 if necessary. You should only call this function right before you 883 if necessary. You should only call this function right before you
397 would have created an AnyEvent watcher anyway, that is, as late as 884 would have created an AnyEvent watcher anyway, that is, as late as
398 possible at runtime. 885 possible at runtime, and not e.g. during initialisation of your
886 module.
887
888 The effect of calling this function is as if a watcher had been
889 created (specifically, actions that happen "when the first watcher
890 is created" happen when calling detetc as well).
891
892 If you need to do some initialisation before AnyEvent watchers are
893 created, use "post_detect".
894
895 $guard = AnyEvent::post_detect { BLOCK }
896 Arranges for the code block to be executed as soon as the event
897 model is autodetected (or immediately if that has already happened).
898
899 The block will be executed *after* the actual backend has been
900 detected ($AnyEvent::MODEL is set), but *before* any watchers have
901 been created, so it is possible to e.g. patch @AnyEvent::ISA or do
902 other initialisations - see the sources of AnyEvent::Strict or
903 AnyEvent::AIO to see how this is used.
904
905 The most common usage is to create some global watchers, without
906 forcing event module detection too early, for example, AnyEvent::AIO
907 creates and installs the global IO::AIO watcher in a "post_detect"
908 block to avoid autodetecting the event module at load time.
909
910 If called in scalar or list context, then it creates and returns an
911 object that automatically removes the callback again when it is
912 destroyed (or "undef" when the hook was immediately executed). See
913 AnyEvent::AIO for a case where this is useful.
914
915 Example: Create a watcher for the IO::AIO module and store it in
916 $WATCHER, but do so only do so after the event loop is initialised.
917
918 our WATCHER;
919
920 my $guard = AnyEvent::post_detect {
921 $WATCHER = AnyEvent->io (fh => IO::AIO::poll_fileno, poll => 'r', cb => \&IO::AIO::poll_cb);
922 };
923
924 # the ||= is important in case post_detect immediately runs the block,
925 # as to not clobber the newly-created watcher. assigning both watcher and
926 # post_detect guard to the same variable has the advantage of users being
927 # able to just C<undef $WATCHER> if the watcher causes them grief.
928
929 $WATCHER ||= $guard;
930
931 @AnyEvent::post_detect
932 If there are any code references in this array (you can "push" to it
933 before or after loading AnyEvent), then they will be called directly
934 after the event loop has been chosen.
935
936 You should check $AnyEvent::MODEL before adding to this array,
937 though: if it is defined then the event loop has already been
938 detected, and the array will be ignored.
939
940 Best use "AnyEvent::post_detect { BLOCK }" when your application
941 allows it, as it takes care of these details.
942
943 This variable is mainly useful for modules that can do something
944 useful when AnyEvent is used and thus want to know when it is
945 initialised, but do not need to even load it by default. This array
946 provides the means to hook into AnyEvent passively, without loading
947 it.
948
949 Example: To load Coro::AnyEvent whenever Coro and AnyEvent are used
950 together, you could put this into Coro (this is the actual code used
951 by Coro to accomplish this):
952
953 if (defined $AnyEvent::MODEL) {
954 # AnyEvent already initialised, so load Coro::AnyEvent
955 require Coro::AnyEvent;
956 } else {
957 # AnyEvent not yet initialised, so make sure to load Coro::AnyEvent
958 # as soon as it is
959 push @AnyEvent::post_detect, sub { require Coro::AnyEvent };
960 }
961
962 AnyEvent::postpone { BLOCK }
963 Arranges for the block to be executed as soon as possible, but not
964 before the call itself returns. In practise, the block will be
965 executed just before the event loop polls for new events, or shortly
966 afterwards.
967
968 This function never returns anything (to make the "return postpone {
969 ... }" idiom more useful.
970
971 To understand the usefulness of this function, consider a function
972 that asynchronously does something for you and returns some
973 transaction object or guard to let you cancel the operation. For
974 example, "AnyEvent::Socket::tcp_connect":
975
976 # start a conenction attempt unless one is active
977 $self->{connect_guard} ||= AnyEvent::Socket::tcp_connect "www.example.net", 80, sub {
978 delete $self->{connect_guard};
979 ...
980 };
981
982 Imagine that this function could instantly call the callback, for
983 example, because it detects an obvious error such as a negative port
984 number. Invoking the callback before the function returns causes
985 problems however: the callback will be called and will try to delete
986 the guard object. But since the function hasn't returned yet, there
987 is nothing to delete. When the function eventually returns it will
988 assign the guard object to "$self->{connect_guard}", where it will
989 likely never be deleted, so the program thinks it is still trying to
990 connect.
991
992 This is where "AnyEvent::postpone" should be used. Instead of
993 calling the callback directly on error:
994
995 $cb->(undef), return # signal error to callback, BAD!
996 if $some_error_condition;
997
998 It should use "postpone":
999
1000 AnyEvent::postpone { $cb->(undef) }, return # signal error to callback, later
1001 if $some_error_condition;
1002
1003 AnyEvent::log $level, $msg[, @args]
1004 Log the given $msg at the given $level.
1005
1006 If AnyEvent::Log is not loaded then this function makes a simple
1007 test to see whether the message will be logged. If the test succeeds
1008 it will load AnyEvent::Log and call "AnyEvent::Log::log" -
1009 consequently, look at the AnyEvent::Log documentation for details.
1010
1011 If the test fails it will simply return. Right now this happens when
1012 a numerical loglevel is used and it is larger than the level
1013 specified via $ENV{PERL_ANYEVENT_VERBOSE}.
1014
1015 If you want to sprinkle loads of logging calls around your code,
1016 consider creating a logger callback with the "AnyEvent::Log::logger"
1017 function, which can reduce typing, codesize and can reduce the
1018 logging overhead enourmously.
399 1019
400WHAT TO DO IN A MODULE 1020WHAT TO DO IN A MODULE
401 As a module author, you should "use AnyEvent" and call AnyEvent methods 1021 As a module author, you should "use AnyEvent" and call AnyEvent methods
402 freely, but you should not load a specific event module or rely on it. 1022 freely, but you should not load a specific event module or rely on it.
403 1023
404 Be careful when you create watchers in the module body - AnyEvent will 1024 Be careful when you create watchers in the module body - AnyEvent will
405 decide which event module to use as soon as the first method is called, 1025 decide which event module to use as soon as the first method is called,
406 so by calling AnyEvent in your module body you force the user of your 1026 so by calling AnyEvent in your module body you force the user of your
407 module to load the event module first. 1027 module to load the event module first.
408 1028
409 Never call "->wait" on a condition variable unless you *know* that the 1029 Never call "->recv" on a condition variable unless you *know* that the
410 "->broadcast" method has been called on it already. This is because it 1030 "->send" method has been called on it already. This is because it will
411 will stall the whole program, and the whole point of using events is to 1031 stall the whole program, and the whole point of using events is to stay
412 stay interactive. 1032 interactive.
413 1033
414 It is fine, however, to call "->wait" when the user of your module 1034 It is fine, however, to call "->recv" when the user of your module
415 requests it (i.e. if you create a http request object ad have a method 1035 requests it (i.e. if you create a http request object ad have a method
416 called "results" that returns the results, it should call "->wait" 1036 called "results" that returns the results, it may call "->recv" freely,
417 freely, as the user of your module knows what she is doing. always). 1037 as the user of your module knows what she is doing. Always).
418 1038
419WHAT TO DO IN THE MAIN PROGRAM 1039WHAT TO DO IN THE MAIN PROGRAM
420 There will always be a single main program - the only place that should 1040 There will always be a single main program - the only place that should
421 dictate which event model to use. 1041 dictate which event model to use.
422 1042
423 If it doesn't care, it can just "use AnyEvent" and use it itself, or not 1043 If the program is not event-based, it need not do anything special, even
424 do anything special (it does not need to be event-based) and let 1044 when it depends on a module that uses an AnyEvent. If the program itself
425 AnyEvent decide which implementation to chose if some module relies on 1045 uses AnyEvent, but does not care which event loop is used, all it needs
426 it. 1046 to do is "use AnyEvent". In either case, AnyEvent will choose the best
1047 available loop implementation.
427 1048
428 If the main program relies on a specific event model. For example, in 1049 If the main program relies on a specific event model - for example, in
429 Gtk2 programs you have to rely on the Glib module. You should load the 1050 Gtk2 programs you have to rely on the Glib module - you should load the
430 event module before loading AnyEvent or any module that uses it: 1051 event module before loading AnyEvent or any module that uses it:
431 generally speaking, you should load it as early as possible. The reason 1052 generally speaking, you should load it as early as possible. The reason
432 is that modules might create watchers when they are loaded, and AnyEvent 1053 is that modules might create watchers when they are loaded, and AnyEvent
433 will decide on the event model to use as soon as it creates watchers, 1054 will decide on the event model to use as soon as it creates watchers,
434 and it might chose the wrong one unless you load the correct one 1055 and it might choose the wrong one unless you load the correct one
435 yourself. 1056 yourself.
436 1057
437 You can chose to use a rather inefficient pure-perl implementation by 1058 You can chose to use a pure-perl implementation by loading the
438 loading the "AnyEvent::Impl::Perl" module, which gives you similar 1059 "AnyEvent::Loop" module, which gives you similar behaviour everywhere,
439 behaviour everywhere, but letting AnyEvent chose is generally better. 1060 but letting AnyEvent chose the model is generally better.
1061
1062 MAINLOOP EMULATION
1063 Sometimes (often for short test scripts, or even standalone programs who
1064 only want to use AnyEvent), you do not want to run a specific event
1065 loop.
1066
1067 In that case, you can use a condition variable like this:
1068
1069 AnyEvent->condvar->recv;
1070
1071 This has the effect of entering the event loop and looping forever.
1072
1073 Note that usually your program has some exit condition, in which case it
1074 is better to use the "traditional" approach of storing a condition
1075 variable somewhere, waiting for it, and sending it when the program
1076 should exit cleanly.
440 1077
441OTHER MODULES 1078OTHER MODULES
442 The following is a non-exhaustive list of additional modules that use 1079 The following is a non-exhaustive list of additional modules that use
443 AnyEvent and can therefore be mixed easily with other AnyEvent modules 1080 AnyEvent as a client and can therefore be mixed easily with other
444 in the same program. Some of the modules come with AnyEvent, some are 1081 AnyEvent modules and other event loops in the same program. Some of the
445 available via CPAN. 1082 modules come as part of AnyEvent, the others are available via CPAN (see
1083 <http://search.cpan.org/search?m=module&q=anyevent%3A%3A*> for a longer
1084 non-exhaustive list), and the list is heavily biased towards modules of
1085 the AnyEvent author himself :)
446 1086
447 AnyEvent::Util 1087 AnyEvent::Util
448 Contains various utility functions that replace often-used but 1088 Contains various utility functions that replace often-used blocking
449 blocking functions such as "inet_aton" by event-/callback-based 1089 functions such as "inet_aton" with event/callback-based versions.
450 versions. 1090
1091 AnyEvent::Socket
1092 Provides various utility functions for (internet protocol) sockets,
1093 addresses and name resolution. Also functions to create non-blocking
1094 tcp connections or tcp servers, with IPv6 and SRV record support and
1095 more.
451 1096
452 AnyEvent::Handle 1097 AnyEvent::Handle
453 Provide read and write buffers and manages watchers for reads and 1098 Provide read and write buffers, manages watchers for reads and
454 writes. 1099 writes, supports raw and formatted I/O, I/O queued and fully
1100 transparent and non-blocking SSL/TLS (via AnyEvent::TLS).
455 1101
456 AnyEvent::Socket 1102 AnyEvent::DNS
457 Provides a means to do non-blocking connects, accepts etc. 1103 Provides rich asynchronous DNS resolver capabilities.
1104
1105 AnyEvent::HTTP, AnyEvent::IRC, AnyEvent::XMPP, AnyEvent::GPSD,
1106 AnyEvent::IGS, AnyEvent::FCP
1107 Implement event-based interfaces to the protocols of the same name
1108 (for the curious, IGS is the International Go Server and FCP is the
1109 Freenet Client Protocol).
1110
1111 AnyEvent::AIO
1112 Truly asynchronous (as opposed to non-blocking) I/O, should be in
1113 the toolbox of every event programmer. AnyEvent::AIO transparently
1114 fuses IO::AIO and AnyEvent together, giving AnyEvent access to
1115 event-based file I/O, and much more.
1116
1117 AnyEvent::Filesys::Notify
1118 AnyEvent is good for non-blocking stuff, but it can't detect file or
1119 path changes (e.g. "watch this directory for new files", "watch this
1120 file for changes"). The AnyEvent::Filesys::Notify module promises to
1121 do just that in a portbale fashion, supporting inotify on GNU/Linux
1122 and some weird, without doubt broken, stuff on OS X to monitor
1123 files. It can fall back to blocking scans at regular intervals
1124 transparently on other platforms, so it's about as portable as it
1125 gets.
1126
1127 (I haven't used it myself, but I haven't heard anybody complaining
1128 about it yet).
1129
1130 AnyEvent::DBI
1131 Executes DBI requests asynchronously in a proxy process for you,
1132 notifying you in an event-based way when the operation is finished.
458 1133
459 AnyEvent::HTTPD 1134 AnyEvent::HTTPD
460 Provides a simple web application server framework. 1135 A simple embedded webserver.
461
462 AnyEvent::DNS
463 Provides asynchronous DNS resolver capabilities, beyond what
464 AnyEvent::Util offers.
465 1136
466 AnyEvent::FastPing 1137 AnyEvent::FastPing
467 The fastest ping in the west. 1138 The fastest ping in the west.
468 1139
469 Net::IRC3
470 AnyEvent based IRC client module family.
471
472 Net::XMPP2
473 AnyEvent based XMPP (Jabber protocol) module family.
474
475 Net::FCP
476 AnyEvent-based implementation of the Freenet Client Protocol,
477 birthplace of AnyEvent.
478
479 Event::ExecFlow
480 High level API for event-based execution flow control.
481
482 Coro 1140 Coro
483 Has special support for AnyEvent. 1141 Has special support for AnyEvent via Coro::AnyEvent, which allows
1142 you to simply invert the flow control - don't call us, we will call
1143 you:
484 1144
485 IO::Lambda 1145 async {
486 The lambda approach to I/O - don't ask, look there. Can use 1146 Coro::AnyEvent::sleep 5; # creates a 5s timer and waits for it
487 AnyEvent. 1147 print "5 seconds later!\n";
488 1148
489 IO::AIO 1149 Coro::AnyEvent::readable *STDIN; # uses an I/O watcher
490 Truly asynchronous I/O, should be in the toolbox of every event 1150 my $line = <STDIN>; # works for ttys
491 programmer. Can be trivially made to use AnyEvent.
492 1151
493 BDB Truly asynchronous Berkeley DB access. Can be trivially made to use 1152 AnyEvent::HTTP::http_get "url", Coro::rouse_cb;
494 AnyEvent. 1153 my ($body, $hdr) = Coro::rouse_wait;
1154 };
1155
1156SIMPLIFIED AE API
1157 Starting with version 5.0, AnyEvent officially supports a second, much
1158 simpler, API that is designed to reduce the calling, typing and memory
1159 overhead by using function call syntax and a fixed number of parameters.
1160
1161 See the AE manpage for details.
1162
1163ERROR AND EXCEPTION HANDLING
1164 In general, AnyEvent does not do any error handling - it relies on the
1165 caller to do that if required. The AnyEvent::Strict module (see also the
1166 "PERL_ANYEVENT_STRICT" environment variable, below) provides strict
1167 checking of all AnyEvent methods, however, which is highly useful during
1168 development.
1169
1170 As for exception handling (i.e. runtime errors and exceptions thrown
1171 while executing a callback), this is not only highly event-loop
1172 specific, but also not in any way wrapped by this module, as this is the
1173 job of the main program.
1174
1175 The pure perl event loop simply re-throws the exception (usually within
1176 "condvar->recv"), the Event and EV modules call "$Event/EV::DIED->()",
1177 Glib uses "install_exception_handler" and so on.
1178
1179ENVIRONMENT VARIABLES
1180 AnyEvent supports a number of environment variables that tune the
1181 runtime behaviour. They are usually evaluated when AnyEvent is loaded,
1182 initialised, or a submodule that uses them is loaded. Many of them also
1183 cause AnyEvent to load additional modules - for example,
1184 "PERL_ANYEVENT_DEBUG_WRAP" causes the AnyEvent::Debug module to be
1185 loaded.
1186
1187 All the environment variables documented here start with
1188 "PERL_ANYEVENT_", which is what AnyEvent considers its own namespace.
1189 Other modules are encouraged (but by no means required) to use
1190 "PERL_ANYEVENT_SUBMODULE" if they have registered the
1191 AnyEvent::Submodule namespace on CPAN, for any submodule. For example,
1192 AnyEvent::HTTP could be expected to use "PERL_ANYEVENT_HTTP_PROXY" (it
1193 should not access env variables starting with "AE_", see below).
1194
1195 All variables can also be set via the "AE_" prefix, that is, instead of
1196 setting "PERL_ANYEVENT_VERBOSE" you can also set "AE_VERBOSE". In case
1197 there is a clash btween anyevent and another program that uses
1198 "AE_something" you can set the corresponding "PERL_ANYEVENT_something"
1199 variable to the empty string, as those variables take precedence.
1200
1201 When AnyEvent is first loaded, it copies all "AE_xxx" env variables to
1202 their "PERL_ANYEVENT_xxx" counterpart unless that variable already
1203 exists. If taint mode is on, then AnyEvent will remove *all* environment
1204 variables starting with "PERL_ANYEVENT_" from %ENV (or replace them with
1205 "undef" or the empty string, if the corresaponding "AE_" variable is
1206 set).
1207
1208 The exact algorithm is currently:
1209
1210 1. if taint mode enabled, delete all PERL_ANYEVENT_xyz variables from %ENV
1211 2. copy over AE_xyz to PERL_ANYEVENT_xyz unless the latter alraedy exists
1212 3. if taint mode enabled, set all PERL_ANYEVENT_xyz variables to undef.
1213
1214 This ensures that child processes will not see the "AE_" variables.
1215
1216 The following environment variables are currently known to AnyEvent:
1217
1218 "PERL_ANYEVENT_VERBOSE"
1219 By default, AnyEvent will only log messages with loglevel 3
1220 ("critical") or higher (see AnyEvent::Log). You can set this
1221 environment variable to a numerical loglevel to make AnyEvent more
1222 (or less) talkative.
1223
1224 If you want to do more than just set the global logging level you
1225 should have a look at "PERL_ANYEVENT_LOG", which allows much more
1226 complex specifications.
1227
1228 When set to 0 ("off"), then no messages whatsoever will be logged
1229 with the default logging settings.
1230
1231 When set to 5 or higher ("warn"), causes AnyEvent to warn about
1232 unexpected conditions, such as not being able to load the event
1233 model specified by "PERL_ANYEVENT_MODEL", or a guard callback
1234 throwing an exception - this is the minimum recommended level.
1235
1236 When set to 7 or higher (info), cause AnyEvent to report which event
1237 model it chooses.
1238
1239 When set to 8 or higher (debug), then AnyEvent will report extra
1240 information on which optional modules it loads and how it implements
1241 certain features.
1242
1243 "PERL_ANYEVENT_LOG"
1244 Accepts rather complex logging specifications. For example, you
1245 could log all "debug" messages of some module to stderr, warnings
1246 and above to stderr, and errors and above to syslog, with:
1247
1248 PERL_ANYEVENT_LOG=Some::Module=debug,+log:filter=warn,+%syslog:%syslog=error,syslog
1249
1250 For the rather extensive details, see AnyEvent::Log.
1251
1252 This variable is evaluated when AnyEvent (or AnyEvent::Log) is
1253 loaded, so will take effect even before AnyEvent has initialised
1254 itself.
1255
1256 Note that specifying this environment variable causes the
1257 AnyEvent::Log module to be loaded, while "PERL_ANYEVENT_VERBOSE"
1258 does not, so only using the latter saves a few hundred kB of memory
1259 until the first message is being logged.
1260
1261 "PERL_ANYEVENT_STRICT"
1262 AnyEvent does not do much argument checking by default, as thorough
1263 argument checking is very costly. Setting this variable to a true
1264 value will cause AnyEvent to load "AnyEvent::Strict" and then to
1265 thoroughly check the arguments passed to most method calls. If it
1266 finds any problems, it will croak.
1267
1268 In other words, enables "strict" mode.
1269
1270 Unlike "use strict" (or its modern cousin, "use common::sense", it
1271 is definitely recommended to keep it off in production. Keeping
1272 "PERL_ANYEVENT_STRICT=1" in your environment while developing
1273 programs can be very useful, however.
1274
1275 "PERL_ANYEVENT_DEBUG_SHELL"
1276 If this env variable is nonempty, then its contents will be
1277 interpreted by "AnyEvent::Socket::parse_hostport" and
1278 "AnyEvent::Debug::shell" (after replacing every occurance of $$ by
1279 the process pid). The shell object is saved in
1280 $AnyEvent::Debug::SHELL.
1281
1282 This happens when the first watcher is created.
1283
1284 For example, to bind a debug shell on a unix domain socket in
1285 /tmp/debug<pid>.sock, you could use this:
1286
1287 PERL_ANYEVENT_DEBUG_SHELL=/tmp/debug\$\$.sock perlprog
1288 # connect with e.g.: socat readline /tmp/debug123.sock
1289
1290 Or to bind to tcp port 4545 on localhost:
1291
1292 PERL_ANYEVENT_DEBUG_SHELL=127.0.0.1:4545 perlprog
1293 # connect with e.g.: telnet localhost 4545
1294
1295 Note that creating sockets in /tmp or on localhost is very unsafe on
1296 multiuser systems.
1297
1298 "PERL_ANYEVENT_DEBUG_WRAP"
1299 Can be set to 0, 1 or 2 and enables wrapping of all watchers for
1300 debugging purposes. See "AnyEvent::Debug::wrap" for details.
1301
1302 "PERL_ANYEVENT_MODEL"
1303 This can be used to specify the event model to be used by AnyEvent,
1304 before auto detection and -probing kicks in.
1305
1306 It normally is a string consisting entirely of ASCII letters (e.g.
1307 "EV" or "IOAsync"). The string "AnyEvent::Impl::" gets prepended and
1308 the resulting module name is loaded and - if the load was successful
1309 - used as event model backend. If it fails to load then AnyEvent
1310 will proceed with auto detection and -probing.
1311
1312 If the string ends with "::" instead (e.g. "AnyEvent::Impl::EV::")
1313 then nothing gets prepended and the module name is used as-is (hint:
1314 "::" at the end of a string designates a module name and quotes it
1315 appropriately).
1316
1317 For example, to force the pure perl model (AnyEvent::Loop::Perl) you
1318 could start your program like this:
1319
1320 PERL_ANYEVENT_MODEL=Perl perl ...
1321
1322 "PERL_ANYEVENT_PROTOCOLS"
1323 Used by both AnyEvent::DNS and AnyEvent::Socket to determine
1324 preferences for IPv4 or IPv6. The default is unspecified (and might
1325 change, or be the result of auto probing).
1326
1327 Must be set to a comma-separated list of protocols or address
1328 families, current supported: "ipv4" and "ipv6". Only protocols
1329 mentioned will be used, and preference will be given to protocols
1330 mentioned earlier in the list.
1331
1332 This variable can effectively be used for denial-of-service attacks
1333 against local programs (e.g. when setuid), although the impact is
1334 likely small, as the program has to handle conenction and other
1335 failures anyways.
1336
1337 Examples: "PERL_ANYEVENT_PROTOCOLS=ipv4,ipv6" - prefer IPv4 over
1338 IPv6, but support both and try to use both.
1339 "PERL_ANYEVENT_PROTOCOLS=ipv4" - only support IPv4, never try to
1340 resolve or contact IPv6 addresses.
1341 "PERL_ANYEVENT_PROTOCOLS=ipv6,ipv4" support either IPv4 or IPv6, but
1342 prefer IPv6 over IPv4.
1343
1344 "PERL_ANYEVENT_HOSTS"
1345 This variable, if specified, overrides the /etc/hosts file used by
1346 AnyEvent::Socket"::resolve_sockaddr", i.e. hosts aliases will be
1347 read from that file instead.
1348
1349 "PERL_ANYEVENT_EDNS0"
1350 Used by AnyEvent::DNS to decide whether to use the EDNS0 extension
1351 for DNS. This extension is generally useful to reduce DNS traffic,
1352 especially when DNSSEC is involved, but some (broken) firewalls drop
1353 such DNS packets, which is why it is off by default.
1354
1355 Setting this variable to 1 will cause AnyEvent::DNS to announce
1356 EDNS0 in its DNS requests.
1357
1358 "PERL_ANYEVENT_MAX_FORKS"
1359 The maximum number of child processes that
1360 "AnyEvent::Util::fork_call" will create in parallel.
1361
1362 "PERL_ANYEVENT_MAX_OUTSTANDING_DNS"
1363 The default value for the "max_outstanding" parameter for the
1364 default DNS resolver - this is the maximum number of parallel DNS
1365 requests that are sent to the DNS server.
1366
1367 "PERL_ANYEVENT_MAX_SIGNAL_LATENCY"
1368 Perl has inherently racy signal handling (you can basically choose
1369 between losing signals and memory corruption) - pure perl event
1370 loops (including "AnyEvent::Loop", when "Async::Interrupt" isn't
1371 available) therefore have to poll regularly to avoid losing signals.
1372
1373 Some event loops are racy, but don't poll regularly, and some event
1374 loops are written in C but are still racy. For those event loops,
1375 AnyEvent installs a timer that regularly wakes up the event loop.
1376
1377 By default, the interval for this timer is 10 seconds, but you can
1378 override this delay with this environment variable (or by setting
1379 the $AnyEvent::MAX_SIGNAL_LATENCY variable before creating signal
1380 watchers).
1381
1382 Lower values increase CPU (and energy) usage, higher values can
1383 introduce long delays when reaping children or waiting for signals.
1384
1385 The AnyEvent::Async module, if available, will be used to avoid this
1386 polling (with most event loops).
1387
1388 "PERL_ANYEVENT_RESOLV_CONF"
1389 The absolute path to a resolv.conf-style file to use instead of
1390 /etc/resolv.conf (or the OS-specific configuration) in the default
1391 resolver, or the empty string to select the default configuration.
1392
1393 "PERL_ANYEVENT_CA_FILE", "PERL_ANYEVENT_CA_PATH".
1394 When neither "ca_file" nor "ca_path" was specified during
1395 AnyEvent::TLS context creation, and either of these environment
1396 variables are nonempty, they will be used to specify CA certificate
1397 locations instead of a system-dependent default.
1398
1399 "PERL_ANYEVENT_AVOID_GUARD" and "PERL_ANYEVENT_AVOID_ASYNC_INTERRUPT"
1400 When these are set to 1, then the respective modules are not loaded.
1401 Mostly good for testing AnyEvent itself.
495 1402
496SUPPLYING YOUR OWN EVENT MODEL INTERFACE 1403SUPPLYING YOUR OWN EVENT MODEL INTERFACE
497 This is an advanced topic that you do not normally need to use AnyEvent 1404 This is an advanced topic that you do not normally need to use AnyEvent
498 in a module. This section is only of use to event loop authors who want 1405 in a module. This section is only of use to event loop authors who want
499 to provide AnyEvent compatibility. 1406 to provide AnyEvent compatibility.
… …
533 1440
534 *rxvt-unicode* also cheats a bit by not providing blocking access to 1441 *rxvt-unicode* also cheats a bit by not providing blocking access to
535 condition variables: code blocking while waiting for a condition will 1442 condition variables: code blocking while waiting for a condition will
536 "die". This still works with most modules/usages, and blocking calls 1443 "die". This still works with most modules/usages, and blocking calls
537 must not be done in an interactive application, so it makes sense. 1444 must not be done in an interactive application, so it makes sense.
538
539ENVIRONMENT VARIABLES
540 The following environment variables are used by this module:
541
542 "PERL_ANYEVENT_VERBOSE"
543 By default, AnyEvent will be completely silent except in fatal
544 conditions. You can set this environment variable to make AnyEvent
545 more talkative.
546
547 When set to 1 or higher, causes AnyEvent to warn about unexpected
548 conditions, such as not being able to load the event model specified
549 by "PERL_ANYEVENT_MODEL".
550
551 When set to 2 or higher, cause AnyEvent to report to STDERR which
552 event model it chooses.
553
554 "PERL_ANYEVENT_MODEL"
555 This can be used to specify the event model to be used by AnyEvent,
556 before autodetection and -probing kicks in. It must be a string
557 consisting entirely of ASCII letters. The string "AnyEvent::Impl::"
558 gets prepended and the resulting module name is loaded and if the
559 load was successful, used as event model. If it fails to load
560 AnyEvent will proceed with autodetection and -probing.
561
562 This functionality might change in future versions.
563
564 For example, to force the pure perl model (AnyEvent::Impl::Perl) you
565 could start your program like this:
566
567 PERL_ANYEVENT_MODEL=Perl perl ...
568 1445
569EXAMPLE PROGRAM 1446EXAMPLE PROGRAM
570 The following program uses an I/O watcher to read data from STDIN, a 1447 The following program uses an I/O watcher to read data from STDIN, a
571 timer to display a message once per second, and a condition variable to 1448 timer to display a message once per second, and a condition variable to
572 quit the program when the user enters quit: 1449 quit the program when the user enters quit:
… …
580 poll => 'r', 1457 poll => 'r',
581 cb => sub { 1458 cb => sub {
582 warn "io event <$_[0]>\n"; # will always output <r> 1459 warn "io event <$_[0]>\n"; # will always output <r>
583 chomp (my $input = <STDIN>); # read a line 1460 chomp (my $input = <STDIN>); # read a line
584 warn "read: $input\n"; # output what has been read 1461 warn "read: $input\n"; # output what has been read
585 $cv->broadcast if $input =~ /^q/i; # quit program if /^q/i 1462 $cv->send if $input =~ /^q/i; # quit program if /^q/i
586 }, 1463 },
587 ); 1464 );
588 1465
589 my $time_watcher; # can only be used once
590
591 sub new_timer {
592 $timer = AnyEvent->timer (after => 1, cb => sub { 1466 my $time_watcher = AnyEvent->timer (after => 1, interval => 1, cb => sub {
593 warn "timeout\n"; # print 'timeout' about every second 1467 warn "timeout\n"; # print 'timeout' at most every second
594 &new_timer; # and restart the time
595 });
596 } 1468 });
597 1469
598 new_timer; # create first timer
599
600 $cv->wait; # wait until user enters /^q/i 1470 $cv->recv; # wait until user enters /^q/i
601 1471
602REAL-WORLD EXAMPLE 1472REAL-WORLD EXAMPLE
603 Consider the Net::FCP module. It features (among others) the following 1473 Consider the Net::FCP module. It features (among others) the following
604 API calls, which are to freenet what HTTP GET requests are to http: 1474 API calls, which are to freenet what HTTP GET requests are to http:
605 1475
… …
654 syswrite $txn->{fh}, $txn->{request} 1524 syswrite $txn->{fh}, $txn->{request}
655 or die "connection or write error"; 1525 or die "connection or write error";
656 $txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r }); 1526 $txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r });
657 1527
658 Again, "fh_ready_r" waits till all data has arrived, and then stores the 1528 Again, "fh_ready_r" waits till all data has arrived, and then stores the
659 result and signals any possible waiters that the request ahs finished: 1529 result and signals any possible waiters that the request has finished:
660 1530
661 sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf}; 1531 sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf};
662 1532
663 if (end-of-file or data complete) { 1533 if (end-of-file or data complete) {
664 $txn->{result} = $txn->{buf}; 1534 $txn->{result} = $txn->{buf};
665 $txn->{finished}->broadcast; 1535 $txn->{finished}->send;
666 $txb->{cb}->($txn) of $txn->{cb}; # also call callback 1536 $txb->{cb}->($txn) of $txn->{cb}; # also call callback
667 } 1537 }
668 1538
669 The "result" method, finally, just waits for the finished signal (if the 1539 The "result" method, finally, just waits for the finished signal (if the
670 request was already finished, it doesn't wait, of course, and returns 1540 request was already finished, it doesn't wait, of course, and returns
671 the data: 1541 the data:
672 1542
673 $txn->{finished}->wait; 1543 $txn->{finished}->recv;
674 return $txn->{result}; 1544 return $txn->{result};
675 1545
676 The actual code goes further and collects all errors ("die"s, 1546 The actual code goes further and collects all errors ("die"s,
677 exceptions) that occured during request processing. The "result" method 1547 exceptions) that occurred during request processing. The "result" method
678 detects whether an exception as thrown (it is stored inside the $txn 1548 detects whether an exception as thrown (it is stored inside the $txn
679 object) and just throws the exception, which means connection errors and 1549 object) and just throws the exception, which means connection errors and
680 other problems get reported tot he code that tries to use the result, 1550 other problems get reported to the code that tries to use the result,
681 not in a random callback. 1551 not in a random callback.
682 1552
683 All of this enables the following usage styles: 1553 All of this enables the following usage styles:
684 1554
685 1. Blocking: 1555 1. Blocking:
… …
713 1583
714 my $quit = AnyEvent->condvar; 1584 my $quit = AnyEvent->condvar;
715 1585
716 $fcp->txn_client_get ($url)->cb (sub { 1586 $fcp->txn_client_get ($url)->cb (sub {
717 ... 1587 ...
718 $quit->broadcast; 1588 $quit->send;
719 }); 1589 });
720 1590
721 $quit->wait; 1591 $quit->recv;
722 1592
723BENCHMARKS 1593BENCHMARKS
724 To give you an idea of the performance and overheads that AnyEvent adds 1594 To give you an idea of the performance and overheads that AnyEvent adds
725 over the event loops themselves and to give you an impression of the 1595 over the event loops themselves and to give you an impression of the
726 speed of various event loops I prepared some benchmarks. 1596 speed of various event loops I prepared some benchmarks.
727 1597
728 BENCHMARKING ANYEVENT OVERHEAD 1598 BENCHMARKING ANYEVENT OVERHEAD
729 Here is a benchmark of various supported event models used natively and 1599 Here is a benchmark of various supported event models used natively and
730 through anyevent. The benchmark creates a lot of timers (with a zero 1600 through AnyEvent. The benchmark creates a lot of timers (with a zero
731 timeout) and I/O watchers (watching STDOUT, a pty, to become writable, 1601 timeout) and I/O watchers (watching STDOUT, a pty, to become writable,
732 which it is), lets them fire exactly once and destroys them again. 1602 which it is), lets them fire exactly once and destroys them again.
733 1603
734 Source code for this benchmark is found as eg/bench in the AnyEvent 1604 Source code for this benchmark is found as eg/bench in the AnyEvent
735 distribution. 1605 distribution. It uses the AE interface, which makes a real difference
1606 for the EV and Perl backends only.
736 1607
737 Explanation of the columns 1608 Explanation of the columns
738 *watcher* is the number of event watchers created/destroyed. Since 1609 *watcher* is the number of event watchers created/destroyed. Since
739 different event models feature vastly different performances, each event 1610 different event models feature vastly different performances, each event
740 loop was given a number of watchers so that overall runtime is 1611 loop was given a number of watchers so that overall runtime is
… …
751 between all watchers, to avoid adding memory overhead. That means 1622 between all watchers, to avoid adding memory overhead. That means
752 closure creation and memory usage is not included in the figures. 1623 closure creation and memory usage is not included in the figures.
753 1624
754 *invoke* is the time, in microseconds, used to invoke a simple callback. 1625 *invoke* is the time, in microseconds, used to invoke a simple callback.
755 The callback simply counts down a Perl variable and after it was invoked 1626 The callback simply counts down a Perl variable and after it was invoked
756 "watcher" times, it would "->broadcast" a condvar once to signal the end 1627 "watcher" times, it would "->send" a condvar once to signal the end of
757 of this phase. 1628 this phase.
758 1629
759 *destroy* is the time, in microseconds, that it takes to destroy a 1630 *destroy* is the time, in microseconds, that it takes to destroy a
760 single watcher. 1631 single watcher.
761 1632
762 Results 1633 Results
763 name watchers bytes create invoke destroy comment 1634 name watchers bytes create invoke destroy comment
764 EV/EV 400000 244 0.56 0.46 0.31 EV native interface 1635 EV/EV 100000 223 0.47 0.43 0.27 EV native interface
765 EV/Any 100000 244 2.50 0.46 0.29 EV + AnyEvent watchers 1636 EV/Any 100000 223 0.48 0.42 0.26 EV + AnyEvent watchers
766 CoroEV/Any 100000 244 2.49 0.44 0.29 coroutines + Coro::Signal 1637 Coro::EV/Any 100000 223 0.47 0.42 0.26 coroutines + Coro::Signal
767 Perl/Any 100000 513 4.92 0.87 1.12 pure perl implementation 1638 Perl/Any 100000 431 2.70 0.74 0.92 pure perl implementation
768 Event/Event 16000 516 31.88 31.30 0.85 Event native interface 1639 Event/Event 16000 516 31.16 31.84 0.82 Event native interface
769 Event/Any 16000 590 35.75 31.42 1.08 Event + AnyEvent watchers 1640 Event/Any 16000 1203 42.61 34.79 1.80 Event + AnyEvent watchers
1641 IOAsync/Any 16000 1911 41.92 27.45 16.81 via IO::Async::Loop::IO_Poll
1642 IOAsync/Any 16000 1726 40.69 26.37 15.25 via IO::Async::Loop::Epoll
770 Glib/Any 16000 1357 98.22 12.41 54.00 quadratic behaviour 1643 Glib/Any 16000 1118 89.00 12.57 51.17 quadratic behaviour
771 Tk/Any 2000 1860 26.97 67.98 14.00 SEGV with >> 2000 watchers 1644 Tk/Any 2000 1346 20.96 10.75 8.00 SEGV with >> 2000 watchers
772 POE/Event 2000 6644 108.64 736.02 14.73 via POE::Loop::Event 1645 POE/Any 2000 6951 108.97 795.32 14.24 via POE::Loop::Event
773 POE/Select 2000 6343 94.13 809.12 565.96 via POE::Loop::Select 1646 POE/Any 2000 6648 94.79 774.40 575.51 via POE::Loop::Select
774 1647
775 Discussion 1648 Discussion
776 The benchmark does *not* measure scalability of the event loop very 1649 The benchmark does *not* measure scalability of the event loop very
777 well. For example, a select-based event loop (such as the pure perl one) 1650 well. For example, a select-based event loop (such as the pure perl one)
778 can never compete with an event loop that uses epoll when the number of 1651 can never compete with an event loop that uses epoll when the number of
… …
789 benchmark machine, handling an event takes roughly 1600 CPU cycles with 1662 benchmark machine, handling an event takes roughly 1600 CPU cycles with
790 EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 1663 EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000
791 CPU cycles with POE. 1664 CPU cycles with POE.
792 1665
793 "EV" is the sole leader regarding speed and memory use, which are both 1666 "EV" is the sole leader regarding speed and memory use, which are both
794 maximal/minimal, respectively. Even when going through AnyEvent, it uses 1667 maximal/minimal, respectively. When using the AE API there is zero
1668 overhead (when going through the AnyEvent API create is about 5-6 times
1669 slower, with other times being equal, so still uses far less memory than
795 far less memory than any other event loop and is still faster than Event 1670 any other event loop and is still faster than Event natively).
796 natively.
797 1671
798 The pure perl implementation is hit in a few sweet spots (both the 1672 The pure perl implementation is hit in a few sweet spots (both the
799 constant timeout and the use of a single fd hit optimisations in the 1673 constant timeout and the use of a single fd hit optimisations in the
800 perl interpreter and the backend itself). Nevertheless this shows that 1674 perl interpreter and the backend itself). Nevertheless this shows that
801 it adds very little overhead in itself. Like any select-based backend 1675 it adds very little overhead in itself. Like any select-based backend
… …
804 benchmark. 1678 benchmark.
805 1679
806 The "Event" module has a relatively high setup and callback invocation 1680 The "Event" module has a relatively high setup and callback invocation
807 cost, but overall scores in on the third place. 1681 cost, but overall scores in on the third place.
808 1682
1683 "IO::Async" performs admirably well, about on par with "Event", even
1684 when using its pure perl backend.
1685
809 "Glib"'s memory usage is quite a bit higher, but it features a faster 1686 "Glib"'s memory usage is quite a bit higher, but it features a faster
810 callback invocation and overall ends up in the same class as "Event". 1687 callback invocation and overall ends up in the same class as "Event".
811 However, Glib scales extremely badly, doubling the number of watchers 1688 However, Glib scales extremely badly, doubling the number of watchers
812 increases the processing time by more than a factor of four, making it 1689 increases the processing time by more than a factor of four, making it
813 completely unusable when using larger numbers of watchers (note that 1690 completely unusable when using larger numbers of watchers (note that
… …
823 the figures above). 1700 the figures above).
824 1701
825 "POE", regardless of underlying event loop (whether using its pure perl 1702 "POE", regardless of underlying event loop (whether using its pure perl
826 select-based backend or the Event module, the POE-EV backend couldn't be 1703 select-based backend or the Event module, the POE-EV backend couldn't be
827 tested because it wasn't working) shows abysmal performance and memory 1704 tested because it wasn't working) shows abysmal performance and memory
828 usage: Watchers use almost 30 times as much memory as EV watchers, and 1705 usage with AnyEvent: Watchers use almost 30 times as much memory as EV
829 10 times as much memory as Event (the high memory requirements are 1706 watchers, and 10 times as much memory as Event (the high memory
830 caused by requiring a session for each watcher). Watcher invocation 1707 requirements are caused by requiring a session for each watcher).
831 speed is almost 900 times slower than with AnyEvent's pure perl 1708 Watcher invocation speed is almost 900 times slower than with AnyEvent's
1709 pure perl implementation.
1710
832 implementation. The design of the POE adaptor class in AnyEvent can not 1711 The design of the POE adaptor class in AnyEvent can not really account
833 really account for this, as session creation overhead is small compared 1712 for the performance issues, though, as session creation overhead is
834 to execution of the state machine, which is coded pretty optimally 1713 small compared to execution of the state machine, which is coded pretty
835 within AnyEvent::Impl::POE. POE simply seems to be abysmally slow. 1714 optimally within AnyEvent::Impl::POE (and while everybody agrees that
1715 using multiple sessions is not a good approach, especially regarding
1716 memory usage, even the author of POE could not come up with a faster
1717 design).
836 1718
837 Summary 1719 Summary
838 * Using EV through AnyEvent is faster than any other event loop (even 1720 * Using EV through AnyEvent is faster than any other event loop (even
839 when used without AnyEvent), but most event loops have acceptable 1721 when used without AnyEvent), but most event loops have acceptable
840 performance with or without AnyEvent. 1722 performance with or without AnyEvent.
841 1723
842 * The overhead AnyEvent adds is usually much smaller than the overhead 1724 * The overhead AnyEvent adds is usually much smaller than the overhead
843 of the actual event loop, only with extremely fast event loops such 1725 of the actual event loop, only with extremely fast event loops such
844 as EV adds AnyEvent significant overhead. 1726 as EV does AnyEvent add significant overhead.
845 1727
846 * You should avoid POE like the plague if you want performance or 1728 * You should avoid POE like the plague if you want performance or
847 reasonable memory usage. 1729 reasonable memory usage.
848 1730
849 BENCHMARKING THE LARGE SERVER CASE 1731 BENCHMARKING THE LARGE SERVER CASE
850 This benchmark atcually benchmarks the event loop itself. It works by 1732 This benchmark actually benchmarks the event loop itself. It works by
851 creating a number of "servers": each server consists of a socketpair, a 1733 creating a number of "servers": each server consists of a socket pair, a
852 timeout watcher that gets reset on activity (but never fires), and an 1734 timeout watcher that gets reset on activity (but never fires), and an
853 I/O watcher waiting for input on one side of the socket. Each time the 1735 I/O watcher waiting for input on one side of the socket. Each time the
854 socket watcher reads a byte it will write that byte to a random other 1736 socket watcher reads a byte it will write that byte to a random other
855 "server". 1737 "server".
856 1738
857 The effect is that there will be a lot of I/O watchers, only part of 1739 The effect is that there will be a lot of I/O watchers, only part of
858 which are active at any one point (so there is a constant number of 1740 which are active at any one point (so there is a constant number of
859 active fds for each loop iterstaion, but which fds these are is random). 1741 active fds for each loop iteration, but which fds these are is random).
860 The timeout is reset each time something is read because that reflects 1742 The timeout is reset each time something is read because that reflects
861 how most timeouts work (and puts extra pressure on the event loops). 1743 how most timeouts work (and puts extra pressure on the event loops).
862 1744
863 In this benchmark, we use 10000 socketpairs (20000 sockets), of which 1745 In this benchmark, we use 10000 socket pairs (20000 sockets), of which
864 100 (1%) are active. This mirrors the activity of large servers with 1746 100 (1%) are active. This mirrors the activity of large servers with
865 many connections, most of which are idle at any one point in time. 1747 many connections, most of which are idle at any one point in time.
866 1748
867 Source code for this benchmark is found as eg/bench2 in the AnyEvent 1749 Source code for this benchmark is found as eg/bench2 in the AnyEvent
868 distribution. 1750 distribution. It uses the AE interface, which makes a real difference
1751 for the EV and Perl backends only.
869 1752
870 Explanation of the columns 1753 Explanation of the columns
871 *sockets* is the number of sockets, and twice the number of "servers" 1754 *sockets* is the number of sockets, and twice the number of "servers"
872 (as each server has a read and write socket end). 1755 (as each server has a read and write socket end).
873 1756
874 *create* is the time it takes to create a socketpair (which is 1757 *create* is the time it takes to create a socket pair (which is
875 nontrivial) and two watchers: an I/O watcher and a timeout watcher. 1758 nontrivial) and two watchers: an I/O watcher and a timeout watcher.
876 1759
877 *request*, the most important value, is the time it takes to handle a 1760 *request*, the most important value, is the time it takes to handle a
878 single "request", that is, reading the token from the pipe and 1761 single "request", that is, reading the token from the pipe and
879 forwarding it to another server. This includes deleting the old timeout 1762 forwarding it to another server. This includes deleting the old timeout
880 and creating a new one that moves the timeout into the future. 1763 and creating a new one that moves the timeout into the future.
881 1764
882 Results 1765 Results
883 name sockets create request 1766 name sockets create request
884 EV 20000 69.01 11.16 1767 EV 20000 62.66 7.99
885 Perl 20000 73.32 35.87 1768 Perl 20000 68.32 32.64
886 Event 20000 212.62 257.32 1769 IOAsync 20000 174.06 101.15 epoll
887 Glib 20000 651.16 1896.30 1770 IOAsync 20000 174.67 610.84 poll
1771 Event 20000 202.69 242.91
1772 Glib 20000 557.01 1689.52
888 POE 20000 349.67 12317.24 uses POE::Loop::Event 1773 POE 20000 341.54 12086.32 uses POE::Loop::Event
889 1774
890 Discussion 1775 Discussion
891 This benchmark *does* measure scalability and overall performance of the 1776 This benchmark *does* measure scalability and overall performance of the
892 particular event loop. 1777 particular event loop.
893 1778
894 EV is again fastest. Since it is using epoll on my system, the setup 1779 EV is again fastest. Since it is using epoll on my system, the setup
895 time is relatively high, though. 1780 time is relatively high, though.
896 1781
897 Perl surprisingly comes second. It is much faster than the C-based event 1782 Perl surprisingly comes second. It is much faster than the C-based event
898 loops Event and Glib. 1783 loops Event and Glib.
1784
1785 IO::Async performs very well when using its epoll backend, and still
1786 quite good compared to Glib when using its pure perl backend.
899 1787
900 Event suffers from high setup time as well (look at its code and you 1788 Event suffers from high setup time as well (look at its code and you
901 will understand why). Callback invocation also has a high overhead 1789 will understand why). Callback invocation also has a high overhead
902 compared to the "$_->() for .."-style loop that the Perl event loop 1790 compared to the "$_->() for .."-style loop that the Perl event loop
903 uses. Event uses select or poll in basically all documented 1791 uses. Event uses select or poll in basically all documented
… …
909 POE is still completely out of the picture, taking over 1000 times as 1797 POE is still completely out of the picture, taking over 1000 times as
910 long as EV, and over 100 times as long as the Perl implementation, even 1798 long as EV, and over 100 times as long as the Perl implementation, even
911 though it uses a C-based event loop in this case. 1799 though it uses a C-based event loop in this case.
912 1800
913 Summary 1801 Summary
914 * The pure perl implementation performs extremely well, considering 1802 * The pure perl implementation performs extremely well.
915 that it uses select.
916 1803
917 * Avoid Glib or POE in large projects where performance matters. 1804 * Avoid Glib or POE in large projects where performance matters.
918 1805
919 BENCHMARKING SMALL SERVERS 1806 BENCHMARKING SMALL SERVERS
920 While event loops should scale (and select-based ones do not...) even to 1807 While event loops should scale (and select-based ones do not...) even to
… …
944 and speed most when you have lots of watchers, not when you only have a 1831 and speed most when you have lots of watchers, not when you only have a
945 few of them). 1832 few of them).
946 1833
947 EV is again fastest. 1834 EV is again fastest.
948 1835
949 Perl again comes second. It is noticably faster than the C-based event 1836 Perl again comes second. It is noticeably faster than the C-based event
950 loops Event and Glib, although the difference is too small to really 1837 loops Event and Glib, although the difference is too small to really
951 matter. 1838 matter.
952 1839
953 POE also performs much better in this case, but is is still far behind 1840 POE also performs much better in this case, but is is still far behind
954 the others. 1841 the others.
955 1842
956 Summary 1843 Summary
957 * C-based event loops perform very well with small number of watchers, 1844 * C-based event loops perform very well with small number of watchers,
958 as the management overhead dominates. 1845 as the management overhead dominates.
959 1846
1847 THE IO::Lambda BENCHMARK
1848 Recently I was told about the benchmark in the IO::Lambda manpage, which
1849 could be misinterpreted to make AnyEvent look bad. In fact, the
1850 benchmark simply compares IO::Lambda with POE, and IO::Lambda looks
1851 better (which shouldn't come as a surprise to anybody). As such, the
1852 benchmark is fine, and mostly shows that the AnyEvent backend from
1853 IO::Lambda isn't very optimal. But how would AnyEvent compare when used
1854 without the extra baggage? To explore this, I wrote the equivalent
1855 benchmark for AnyEvent.
1856
1857 The benchmark itself creates an echo-server, and then, for 500 times,
1858 connects to the echo server, sends a line, waits for the reply, and then
1859 creates the next connection. This is a rather bad benchmark, as it
1860 doesn't test the efficiency of the framework or much non-blocking I/O,
1861 but it is a benchmark nevertheless.
1862
1863 name runtime
1864 Lambda/select 0.330 sec
1865 + optimized 0.122 sec
1866 Lambda/AnyEvent 0.327 sec
1867 + optimized 0.138 sec
1868 Raw sockets/select 0.077 sec
1869 POE/select, components 0.662 sec
1870 POE/select, raw sockets 0.226 sec
1871 POE/select, optimized 0.404 sec
1872
1873 AnyEvent/select/nb 0.085 sec
1874 AnyEvent/EV/nb 0.068 sec
1875 +state machine 0.134 sec
1876
1877 The benchmark is also a bit unfair (my fault): the IO::Lambda/POE
1878 benchmarks actually make blocking connects and use 100% blocking I/O,
1879 defeating the purpose of an event-based solution. All of the newly
1880 written AnyEvent benchmarks use 100% non-blocking connects (using
1881 AnyEvent::Socket::tcp_connect and the asynchronous pure perl DNS
1882 resolver), so AnyEvent is at a disadvantage here, as non-blocking
1883 connects generally require a lot more bookkeeping and event handling
1884 than blocking connects (which involve a single syscall only).
1885
1886 The last AnyEvent benchmark additionally uses AnyEvent::Handle, which
1887 offers similar expressive power as POE and IO::Lambda, using
1888 conventional Perl syntax. This means that both the echo server and the
1889 client are 100% non-blocking, further placing it at a disadvantage.
1890
1891 As you can see, the AnyEvent + EV combination even beats the
1892 hand-optimised "raw sockets benchmark", while AnyEvent + its pure perl
1893 backend easily beats IO::Lambda and POE.
1894
1895 And even the 100% non-blocking version written using the high-level (and
1896 slow :) AnyEvent::Handle abstraction beats both POE and IO::Lambda
1897 higher level ("unoptimised") abstractions by a large margin, even though
1898 it does all of DNS, tcp-connect and socket I/O in a non-blocking way.
1899
1900 The two AnyEvent benchmarks programs can be found as eg/ae0.pl and
1901 eg/ae2.pl in the AnyEvent distribution, the remaining benchmarks are
1902 part of the IO::Lambda distribution and were used without any changes.
1903
1904SIGNALS
1905 AnyEvent currently installs handlers for these signals:
1906
1907 SIGCHLD
1908 A handler for "SIGCHLD" is installed by AnyEvent's child watcher
1909 emulation for event loops that do not support them natively. Also,
1910 some event loops install a similar handler.
1911
1912 Additionally, when AnyEvent is loaded and SIGCHLD is set to IGNORE,
1913 then AnyEvent will reset it to default, to avoid losing child exit
1914 statuses.
1915
1916 SIGPIPE
1917 A no-op handler is installed for "SIGPIPE" when $SIG{PIPE} is
1918 "undef" when AnyEvent gets loaded.
1919
1920 The rationale for this is that AnyEvent users usually do not really
1921 depend on SIGPIPE delivery (which is purely an optimisation for
1922 shell use, or badly-written programs), but "SIGPIPE" can cause
1923 spurious and rare program exits as a lot of people do not expect
1924 "SIGPIPE" when writing to some random socket.
1925
1926 The rationale for installing a no-op handler as opposed to ignoring
1927 it is that this way, the handler will be restored to defaults on
1928 exec.
1929
1930 Feel free to install your own handler, or reset it to defaults.
1931
1932RECOMMENDED/OPTIONAL MODULES
1933 One of AnyEvent's main goals is to be 100% Pure-Perl(tm): only perl (and
1934 its built-in modules) are required to use it.
1935
1936 That does not mean that AnyEvent won't take advantage of some additional
1937 modules if they are installed.
1938
1939 This section explains which additional modules will be used, and how
1940 they affect AnyEvent's operation.
1941
1942 Async::Interrupt
1943 This slightly arcane module is used to implement fast signal
1944 handling: To my knowledge, there is no way to do completely
1945 race-free and quick signal handling in pure perl. To ensure that
1946 signals still get delivered, AnyEvent will start an interval timer
1947 to wake up perl (and catch the signals) with some delay (default is
1948 10 seconds, look for $AnyEvent::MAX_SIGNAL_LATENCY).
1949
1950 If this module is available, then it will be used to implement
1951 signal catching, which means that signals will not be delayed, and
1952 the event loop will not be interrupted regularly, which is more
1953 efficient (and good for battery life on laptops).
1954
1955 This affects not just the pure-perl event loop, but also other event
1956 loops that have no signal handling on their own (e.g. Glib, Tk, Qt).
1957
1958 Some event loops (POE, Event, Event::Lib) offer signal watchers
1959 natively, and either employ their own workarounds (POE) or use
1960 AnyEvent's workaround (using $AnyEvent::MAX_SIGNAL_LATENCY).
1961 Installing Async::Interrupt does nothing for those backends.
1962
1963 EV This module isn't really "optional", as it is simply one of the
1964 backend event loops that AnyEvent can use. However, it is simply the
1965 best event loop available in terms of features, speed and stability:
1966 It supports the AnyEvent API optimally, implements all the watcher
1967 types in XS, does automatic timer adjustments even when no monotonic
1968 clock is available, can take avdantage of advanced kernel interfaces
1969 such as "epoll" and "kqueue", and is the fastest backend *by far*.
1970 You can even embed Glib/Gtk2 in it (or vice versa, see EV::Glib and
1971 Glib::EV).
1972
1973 If you only use backends that rely on another event loop (e.g.
1974 "Tk"), then this module will do nothing for you.
1975
1976 Guard
1977 The guard module, when used, will be used to implement
1978 "AnyEvent::Util::guard". This speeds up guards considerably (and
1979 uses a lot less memory), but otherwise doesn't affect guard
1980 operation much. It is purely used for performance.
1981
1982 JSON and JSON::XS
1983 One of these modules is required when you want to read or write JSON
1984 data via AnyEvent::Handle. JSON is also written in pure-perl, but
1985 can take advantage of the ultra-high-speed JSON::XS module when it
1986 is installed.
1987
1988 Net::SSLeay
1989 Implementing TLS/SSL in Perl is certainly interesting, but not very
1990 worthwhile: If this module is installed, then AnyEvent::Handle (with
1991 the help of AnyEvent::TLS), gains the ability to do TLS/SSL.
1992
1993 Time::HiRes
1994 This module is part of perl since release 5.008. It will be used
1995 when the chosen event library does not come with a timing source of
1996 its own. The pure-perl event loop (AnyEvent::Loop) will additionally
1997 load it to try to use a monotonic clock for timing stability.
1998
960FORK 1999FORK
961 Most event libraries are not fork-safe. The ones who are usually are 2000 Most event libraries are not fork-safe. The ones who are usually are
962 because they are so inefficient. Only EV is fully fork-aware. 2001 because they rely on inefficient but fork-safe "select" or "poll" calls
2002 - higher performance APIs such as BSD's kqueue or the dreaded Linux
2003 epoll are usually badly thought-out hacks that are incompatible with
2004 fork in one way or another. Only EV is fully fork-aware and ensures that
2005 you continue event-processing in both parent and child (or both, if you
2006 know what you are doing).
2007
2008 This means that, in general, you cannot fork and do event processing in
2009 the child if the event library was initialised before the fork (which
2010 usually happens when the first AnyEvent watcher is created, or the
2011 library is loaded).
963 2012
964 If you have to fork, you must either do so *before* creating your first 2013 If you have to fork, you must either do so *before* creating your first
965 watcher OR you must not use AnyEvent at all in the child. 2014 watcher OR you must not use AnyEvent at all in the child OR you must do
2015 something completely out of the scope of AnyEvent.
2016
2017 The problem of doing event processing in the parent *and* the child is
2018 much more complicated: even for backends that *are* fork-aware or
2019 fork-safe, their behaviour is not usually what you want: fork clones all
2020 watchers, that means all timers, I/O watchers etc. are active in both
2021 parent and child, which is almost never what you want. USing "exec" to
2022 start worker children from some kind of manage rprocess is usually
2023 preferred, because it is much easier and cleaner, at the expense of
2024 having to have another binary.
966 2025
967SECURITY CONSIDERATIONS 2026SECURITY CONSIDERATIONS
968 AnyEvent can be forced to load any event model via 2027 AnyEvent can be forced to load any event model via
969 $ENV{PERL_ANYEVENT_MODEL}. While this cannot (to my knowledge) be used 2028 $ENV{PERL_ANYEVENT_MODEL}. While this cannot (to my knowledge) be used
970 to execute arbitrary code or directly gain access, it can easily be used 2029 to execute arbitrary code or directly gain access, it can easily be used
… …
973 model than specified in the variable. 2032 model than specified in the variable.
974 2033
975 You can make AnyEvent completely ignore this variable by deleting it 2034 You can make AnyEvent completely ignore this variable by deleting it
976 before the first watcher gets created, e.g. with a "BEGIN" block: 2035 before the first watcher gets created, e.g. with a "BEGIN" block:
977 2036
978 BEGIN { delete $ENV{PERL_ANYEVENT_MODEL} } 2037 BEGIN { delete $ENV{PERL_ANYEVENT_MODEL} }
979 2038
980 use AnyEvent; 2039 use AnyEvent;
2040
2041 Similar considerations apply to $ENV{PERL_ANYEVENT_VERBOSE}, as that can
2042 be used to probe what backend is used and gain other information (which
2043 is probably even less useful to an attacker than PERL_ANYEVENT_MODEL),
2044 and $ENV{PERL_ANYEVENT_STRICT}.
2045
2046 Note that AnyEvent will remove *all* environment variables starting with
2047 "PERL_ANYEVENT_" from %ENV when it is loaded while taint mode is
2048 enabled.
2049
2050BUGS
2051 Perl 5.8 has numerous memleaks that sometimes hit this module and are
2052 hard to work around. If you suffer from memleaks, first upgrade to Perl
2053 5.10 and check wether the leaks still show up. (Perl 5.10.0 has other
2054 annoying memleaks, such as leaking on "map" and "grep" but it is usually
2055 not as pronounced).
981 2056
982SEE ALSO 2057SEE ALSO
983 Event modules: Coro::EV, EV, EV::Glib, Glib::EV, Coro::Event, Event, 2058 Tutorial/Introduction: AnyEvent::Intro.
2059
2060 FAQ: AnyEvent::FAQ.
2061
2062 Utility functions: AnyEvent::Util (misc. grab-bag), AnyEvent::Log
2063 (simply logging).
2064
2065 Development/Debugging: AnyEvent::Strict (stricter checking),
2066 AnyEvent::Debug (interactive shell, watcher tracing).
2067
2068 Supported event modules: AnyEvent::Loop, EV, EV::Glib, Glib::EV, Event,
984 Glib::Event, Glib, Coro, Tk, Event::Lib, Qt, POE. 2069 Glib::Event, Glib, Tk, Event::Lib, Qt, POE, FLTK.
985 2070
986 Implementations: AnyEvent::Impl::CoroEV, AnyEvent::Impl::EV, 2071 Implementations: AnyEvent::Impl::EV, AnyEvent::Impl::Event,
987 AnyEvent::Impl::CoroEvent, AnyEvent::Impl::Event, AnyEvent::Impl::Glib, 2072 AnyEvent::Impl::Glib, AnyEvent::Impl::Tk, AnyEvent::Impl::Perl,
988 AnyEvent::Impl::Tk, AnyEvent::Impl::Perl, AnyEvent::Impl::EventLib,
989 AnyEvent::Impl::Qt, AnyEvent::Impl::POE. 2073 AnyEvent::Impl::EventLib, AnyEvent::Impl::Qt, AnyEvent::Impl::POE,
2074 AnyEvent::Impl::IOAsync, Anyevent::Impl::Irssi, AnyEvent::Impl::FLTK.
990 2075
991 Nontrivial usage examples: Net::FCP, Net::XMPP2. 2076 Non-blocking handles, pipes, stream sockets, TCP clients and servers:
2077 AnyEvent::Handle, AnyEvent::Socket, AnyEvent::TLS.
2078
2079 Asynchronous DNS: AnyEvent::DNS.
2080
2081 Thread support: Coro, Coro::AnyEvent, Coro::EV, Coro::Event.
2082
2083 Nontrivial usage examples: AnyEvent::GPSD, AnyEvent::IRC,
2084 AnyEvent::HTTP.
992 2085
993AUTHOR 2086AUTHOR
994 Marc Lehmann <schmorp@schmorp.de> 2087 Marc Lehmann <schmorp@schmorp.de>
995 http://home.schmorp.de/ 2088 http://home.schmorp.de/
996 2089

Diff Legend

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