1 | NAME |
1 | NAME |
2 | EV - perl interface to libevent, monkey.org/~provos/libevent/ |
2 | EV - perl interface to libev, a high performance full-featured event |
|
|
3 | loop |
3 | |
4 | |
4 | SYNOPSIS |
5 | SYNOPSIS |
5 | use EV; |
6 | use EV; |
6 | |
7 | |
7 | # TIMER |
8 | # TIMERS |
8 | |
9 | |
9 | my $w = EV::timer 2, 0, sub { |
10 | my $w = EV::timer 2, 0, sub { |
10 | warn "is called after 2s"; |
11 | warn "is called after 2s"; |
11 | }; |
12 | }; |
12 | |
13 | |
13 | my $w = EV::timer 2, 1, sub { |
14 | my $w = EV::timer 2, 2, sub { |
14 | warn "is called roughly every 2s (repeat = 1)"; |
15 | warn "is called roughly every 2s (repeat = 2)"; |
15 | }; |
16 | }; |
16 | |
17 | |
17 | undef $w; # destroy event watcher again |
18 | undef $w; # destroy event watcher again |
18 | |
19 | |
19 | my $w = EV::timer_abs 0, 60, sub { |
20 | my $w = EV::periodic 0, 60, 0, sub { |
20 | warn "is called every minute, on the minute, exactly"; |
21 | warn "is called every minute, on the minute, exactly"; |
21 | }; |
22 | }; |
22 | |
23 | |
23 | # IO |
24 | # IO |
24 | |
25 | |
25 | my $w = EV::io \*STDIN, EV::READ | EV::PERSIST, sub { |
26 | my $w = EV::io *STDIN, EV::READ, sub { |
26 | my ($w, $revents) = @_; # all callbacks get the watcher object and event mask |
27 | my ($w, $revents) = @_; # all callbacks receive the watcher and event mask |
27 | if ($revents & EV::TIMEOUT) { |
|
|
28 | warn "nothing received on stdin for 10 seconds, retrying"; |
|
|
29 | } else { |
|
|
30 | warn "stdin is readable, you entered: ", <STDIN>; |
28 | warn "stdin is readable, you entered: ", <STDIN>; |
31 | } |
|
|
32 | }; |
29 | }; |
33 | $w->timeout (10); |
|
|
34 | |
|
|
35 | my $w = EV::timed_io \*STDIN, EV::READ, 30, sub { |
|
|
36 | my ($w, $revents) = @_; |
|
|
37 | if ($revents & EV::TIMEOUT) { |
|
|
38 | warn "nothing entered within 30 seconds, bye bye.\n"; |
|
|
39 | $w->stop; |
|
|
40 | } else { |
|
|
41 | my $line = <STDIN>; |
|
|
42 | warn "you entered something, you again have 30 seconds.\n"; |
|
|
43 | } |
30 | |
44 | }; |
|
|
45 | |
|
|
46 | # SIGNALS |
31 | # SIGNALS |
47 | |
32 | |
48 | my $w = EV::signal 'QUIT', sub { |
33 | my $w = EV::signal 'QUIT', sub { |
49 | warn "sigquit received\n"; |
34 | warn "sigquit received\n"; |
50 | }; |
35 | }; |
51 | |
|
|
52 | my $w = EV::signal 3, sub { |
|
|
53 | warn "sigquit received (this is GNU/Linux, right?)\n"; |
|
|
54 | }; |
36 | |
55 | |
|
|
56 | # CHILD/PID STATUS CHANGES |
37 | # CHILD/PID STATUS CHANGES |
57 | |
38 | |
58 | my $w = EV::child 666, sub { |
39 | my $w = EV::child 666, 0, sub { |
59 | my ($w, $revents, $status) = @_; |
40 | my ($w, $revents) = @_; |
|
|
41 | my $status = $w->rstatus; |
60 | }; |
42 | }; |
61 | |
43 | |
|
|
44 | # STAT CHANGES |
|
|
45 | my $w = EV::stat "/etc/passwd", 10, sub { |
|
|
46 | my ($w, $revents) = @_; |
|
|
47 | warn $w->path, " has changed somehow.\n"; |
|
|
48 | }; |
|
|
49 | |
62 | # MAINLOOP |
50 | # MAINLOOP |
63 | EV::dispatch; # loop as long as watchers are active |
51 | EV::loop; # loop until EV::unloop is called or all watchers stop |
64 | EV::loop; # the same thing |
|
|
65 | EV::loop EV::LOOP_ONESHOT; # block until some events could be handles |
52 | EV::loop EV::LOOP_ONESHOT; # block until at least one event could be handled |
66 | EV::loop EV::LOOP_NONBLOCK; # check and handle some events, but do not wait |
53 | EV::loop EV::LOOP_NONBLOCK; # try to handle same events, but do not block |
67 | |
54 | |
68 | DESCRIPTION |
55 | DESCRIPTION |
69 | This module provides an interface to libev |
56 | This module provides an interface to libev |
70 | (<http://software.schmorp.de/pkg/libev.html>). You probably should |
57 | (<http://software.schmorp.de/pkg/libev.html>). While the documentation |
71 | acquaint yourself with its documentation and source code to be able to |
58 | below is comprehensive, one might also consult the documentation of |
72 | use this module fully. |
59 | libev itself (<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>) |
|
|
60 | for more subtle details on watcher semantics or some discussion on the |
|
|
61 | available backends, or how to force a specific backend with |
|
|
62 | "LIBEV_FLAGS", or just about in any case because it has much more |
|
|
63 | detailed information. |
|
|
64 | |
|
|
65 | This module is very fast and scalable. It is actually so fast that you |
|
|
66 | can use it through the AnyEvent module, stay portable to other event |
|
|
67 | loops (if you don't rely on any watcher types not available through it) |
|
|
68 | and still be faster than with any other event loop currently supported |
|
|
69 | in Perl. |
|
|
70 | |
|
|
71 | EVENT LOOPS |
|
|
72 | EV supports multiple event loops: There is a single "default event loop" |
|
|
73 | that can handle everything including signals and child watchers, and any |
|
|
74 | number of "dynamic event loops" that can use different backends (with |
|
|
75 | various limitations), but no child and signal watchers. |
|
|
76 | |
|
|
77 | You do not have to do anything to create the default event loop: When |
|
|
78 | the module is loaded a suitable backend is selected on the premise of |
|
|
79 | selecting a working backend (which for example rules out kqueue on most |
|
|
80 | BSDs). Modules should, unless they have "special needs" always use the |
|
|
81 | default loop as this is fastest (perl-wise), best supported by other |
|
|
82 | modules (e.g. AnyEvent or Coro) and most portable event loop. |
|
|
83 | |
|
|
84 | For specific programs you can create additional event loops dynamically. |
|
|
85 | |
|
|
86 | If you want to take avdantage of kqueue (which often works properly for |
|
|
87 | sockets only) even though the default loop doesn't enable it, you can |
|
|
88 | *embed* a kqueue loop into the default loop: running the default loop |
|
|
89 | will then also service the kqueue loop to some extent. See the example |
|
|
90 | in the section about embed watchers for an example on how to achieve |
|
|
91 | that. |
|
|
92 | |
|
|
93 | $loop = new EV::loop [$flags] |
|
|
94 | Create a new event loop as per the specified flags. Please refer to |
|
|
95 | the "ev_loop_new ()" function description in the libev documentation |
|
|
96 | (<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod#GLOBAL_FUNCTI |
|
|
97 | ONS>) for more info. |
|
|
98 | |
|
|
99 | The loop will automatically be destroyed when it is no longer |
|
|
100 | referenced by any watcher and the loop object goes out of scope. |
|
|
101 | |
|
|
102 | Using "EV::FLAG_FORKCHECK" is recommended, as only the default event |
|
|
103 | loop is protected by this module. |
|
|
104 | |
|
|
105 | $loop->loop_fork |
|
|
106 | Must be called after a fork in the child, before entering or |
|
|
107 | continuing the event loop. An alternative is to use |
|
|
108 | "EV::FLAG_FORKCHECK" which calls this function automatically, at |
|
|
109 | some performance loss (refer to the libev documentation). |
|
|
110 | |
|
|
111 | $loop->loop_verify |
|
|
112 | Calls "ev_verify" to make internal consistency checks (for debugging |
|
|
113 | libev) and abort the program if any data structures were found to be |
|
|
114 | corrupted. |
|
|
115 | |
|
|
116 | $loop = EV::default_loop [$flags] |
|
|
117 | Return the default loop (which is a singleton object). Since this |
|
|
118 | module already creates the default loop with default flags, |
|
|
119 | specifying flags here will not have any effect unless you destroy |
|
|
120 | the default loop first, which isn't supported. So in short: don't do |
|
|
121 | it, and if you break it, you get to keep the pieces. |
73 | |
122 | |
74 | BASIC INTERFACE |
123 | BASIC INTERFACE |
75 | $EV::NPRI |
|
|
76 | How many priority levels are available. |
|
|
77 | |
|
|
78 | $EV::DIED |
124 | $EV::DIED |
79 | Must contain a reference to a function that is called when a |
125 | Must contain a reference to a function that is called when a |
80 | callback throws an exception (with $@ containing thr error). The |
126 | callback throws an exception (with $@ containing the error). The |
81 | default prints an informative message and continues. |
127 | default prints an informative message and continues. |
82 | |
128 | |
83 | If this callback throws an exception it will be silently ignored. |
129 | If this callback throws an exception it will be silently ignored. |
84 | |
130 | |
|
|
131 | $flags = EV::supported_backends |
|
|
132 | $flags = EV::recommended_backends |
|
|
133 | $flags = EV::embeddable_backends |
|
|
134 | Returns the set (see "EV::BACKEND_*" flags) of backends supported by |
|
|
135 | this instance of EV, the set of recommended backends (supposed to be |
|
|
136 | good) for this platform and the set of embeddable backends (see |
|
|
137 | EMBED WATCHERS). |
|
|
138 | |
|
|
139 | EV::sleep $seconds |
|
|
140 | Block the process for the given number of (fractional) seconds. |
|
|
141 | |
|
|
142 | $time = EV::time |
|
|
143 | Returns the current time in (fractional) seconds since the epoch. |
|
|
144 | |
85 | $time = EV::now |
145 | $time = EV::now |
86 | Returns the time in (fractional) seconds since the epoch. |
146 | $time = $loop->now |
|
|
147 | Returns the time the last event loop iteration has been started. |
|
|
148 | This is the time that (relative) timers are based on, and refering |
|
|
149 | to it is usually faster then calling EV::time. |
87 | |
150 | |
88 | $version = EV::version |
151 | $backend = EV::backend |
89 | $method = EV::method |
152 | $backend = $loop->backend |
90 | Return version string and event polling method used. |
153 | Returns an integer describing the backend used by libev |
|
|
154 | (EV::METHOD_SELECT or EV::METHOD_EPOLL). |
91 | |
155 | |
92 | EV::loop $flags # EV::LOOP_ONCE, EV::LOOP_ONESHOT |
156 | EV::loop [$flags] |
93 | EV::loopexit $after |
157 | $loop->loop ([$flags]) |
94 | Exit any active loop or dispatch after $after seconds or immediately |
158 | Begin checking for events and calling callbacks. It returns when a |
95 | if $after is missing or zero. |
159 | callback calls EV::unloop. |
96 | |
160 | |
97 | EV::dispatch |
161 | The $flags argument can be one of the following: |
98 | Same as "EV::loop 0". |
|
|
99 | |
162 | |
100 | EV::event $callback |
163 | 0 as above |
101 | Creates a new event watcher waiting for nothing, calling the given |
164 | EV::LOOP_ONESHOT block at most once (wait, but do not loop) |
102 | callback. |
165 | EV::LOOP_NONBLOCK do not block at all (fetch/handle events but do not wait) |
103 | |
166 | |
|
|
167 | EV::unloop [$how] |
|
|
168 | $loop->unloop ([$how]) |
|
|
169 | When called with no arguments or an argument of EV::UNLOOP_ONE, |
|
|
170 | makes the innermost call to EV::loop return. |
|
|
171 | |
|
|
172 | When called with an argument of EV::UNLOOP_ALL, all calls to |
|
|
173 | EV::loop will return as fast as possible. |
|
|
174 | |
|
|
175 | $count = EV::loop_count |
|
|
176 | $count = $loop->loop_count |
|
|
177 | Return the number of times the event loop has polled for new events. |
|
|
178 | Sometiems useful as a generation counter. |
|
|
179 | |
|
|
180 | EV::once $fh_or_undef, $events, $timeout, $cb->($revents) |
|
|
181 | $loop->once ($fh_or_undef, $events, $timeout, $cb->($revents)) |
|
|
182 | This function rolls together an I/O and a timer watcher for a single |
|
|
183 | one-shot event without the need for managing a watcher object. |
|
|
184 | |
|
|
185 | If $fh_or_undef is a filehandle or file descriptor, then $events |
|
|
186 | must be a bitset containing either "EV::READ", "EV::WRITE" or |
|
|
187 | "EV::READ | EV::WRITE", indicating the type of I/O event you want to |
|
|
188 | wait for. If you do not want to wait for some I/O event, specify |
|
|
189 | "undef" for $fh_or_undef and 0 for $events). |
|
|
190 | |
|
|
191 | If timeout is "undef" or negative, then there will be no timeout. |
|
|
192 | Otherwise a EV::timer with this value will be started. |
|
|
193 | |
|
|
194 | When an error occurs or either the timeout or I/O watcher triggers, |
|
|
195 | then the callback will be called with the received event set (in |
|
|
196 | general you can expect it to be a combination of "EV::ERROR", |
|
|
197 | "EV::READ", "EV::WRITE" and "EV::TIMEOUT"). |
|
|
198 | |
|
|
199 | EV::once doesn't return anything: the watchers stay active till |
|
|
200 | either of them triggers, then they will be stopped and freed, and |
|
|
201 | the callback invoked. |
|
|
202 | |
|
|
203 | EV::feed_fd_event ($fd, $revents) |
|
|
204 | $loop->feed_fd_event ($fd, $revents) |
|
|
205 | Feed an event on a file descriptor into EV. EV will react to this |
|
|
206 | call as if the readyness notifications specified by $revents (a |
|
|
207 | combination of "EV::READ" and "EV::WRITE") happened on the file |
|
|
208 | descriptor $fd. |
|
|
209 | |
|
|
210 | EV::feed_signal_event ($signal) |
|
|
211 | Feed a signal event into EV. EV will react to this call as if the |
|
|
212 | signal specified by $signal had occured. |
|
|
213 | |
|
|
214 | EV::set_io_collect_interval $time |
|
|
215 | $loop->set_io_collect_interval ($time) |
|
|
216 | EV::set_timeout_collect_interval $time |
|
|
217 | $loop->set_timeout_collect_interval ($time) |
|
|
218 | These advanced functions set the minimum block interval when polling |
|
|
219 | for I/O events and the minimum wait interval for timer events. See |
|
|
220 | the libev documentation at |
|
|
221 | <http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod#FUNCTIONS_CONT |
|
|
222 | ROLLING_THE_EVENT_LOOP> for a more detailed discussion. |
|
|
223 | |
|
|
224 | WATCHER OBJECTS |
|
|
225 | A watcher is an object that gets created to record your interest in some |
|
|
226 | event. For instance, if you want to wait for STDIN to become readable, |
|
|
227 | you would create an EV::io watcher for that: |
|
|
228 | |
|
|
229 | my $watcher = EV::io *STDIN, EV::READ, sub { |
|
|
230 | my ($watcher, $revents) = @_; |
|
|
231 | warn "yeah, STDIN should now be readable without blocking!\n" |
|
|
232 | }; |
|
|
233 | |
|
|
234 | All watchers can be active (waiting for events) or inactive (paused). |
|
|
235 | Only active watchers will have their callbacks invoked. All callbacks |
|
|
236 | will be called with at least two arguments: the watcher and a bitmask of |
|
|
237 | received events. |
|
|
238 | |
|
|
239 | Each watcher type has its associated bit in revents, so you can use the |
|
|
240 | same callback for multiple watchers. The event mask is named after the |
|
|
241 | type, i..e. EV::child sets EV::CHILD, EV::prepare sets EV::PREPARE, |
|
|
242 | EV::periodic sets EV::PERIODIC and so on, with the exception of I/O |
|
|
243 | events (which can set both EV::READ and EV::WRITE bits), and EV::timer |
|
|
244 | (which uses EV::TIMEOUT). |
|
|
245 | |
|
|
246 | In the rare case where one wants to create a watcher but not start it at |
|
|
247 | the same time, each constructor has a variant with a trailing "_ns" in |
|
|
248 | its name, e.g. EV::io has a non-starting variant EV::io_ns and so on. |
|
|
249 | |
|
|
250 | Please note that a watcher will automatically be stopped when the |
|
|
251 | watcher object is destroyed, so you *need* to keep the watcher objects |
|
|
252 | returned by the constructors. |
|
|
253 | |
|
|
254 | Also, all methods changing some aspect of a watcher (->set, ->priority, |
|
|
255 | ->fh and so on) automatically stop and start it again if it is active, |
|
|
256 | which means pending events get lost. |
|
|
257 | |
|
|
258 | COMMON WATCHER METHODS |
|
|
259 | This section lists methods common to all watchers. |
|
|
260 | |
|
|
261 | $w->start |
|
|
262 | Starts a watcher if it isn't active already. Does nothing to an |
|
|
263 | already active watcher. By default, all watchers start out in the |
|
|
264 | active state (see the description of the "_ns" variants if you need |
|
|
265 | stopped watchers). |
|
|
266 | |
|
|
267 | $w->stop |
|
|
268 | Stop a watcher if it is active. Also clear any pending events |
|
|
269 | (events that have been received but that didn't yet result in a |
|
|
270 | callback invocation), regardless of whether the watcher was active |
|
|
271 | or not. |
|
|
272 | |
|
|
273 | $bool = $w->is_active |
|
|
274 | Returns true if the watcher is active, false otherwise. |
|
|
275 | |
|
|
276 | $current_data = $w->data |
|
|
277 | $old_data = $w->data ($new_data) |
|
|
278 | Queries a freely usable data scalar on the watcher and optionally |
|
|
279 | changes it. This is a way to associate custom data with a watcher: |
|
|
280 | |
|
|
281 | my $w = EV::timer 60, 0, sub { |
|
|
282 | warn $_[0]->data; |
|
|
283 | }; |
|
|
284 | $w->data ("print me!"); |
|
|
285 | |
|
|
286 | $current_cb = $w->cb |
|
|
287 | $old_cb = $w->cb ($new_cb) |
|
|
288 | Queries the callback on the watcher and optionally changes it. You |
|
|
289 | can do this at any time without the watcher restarting. |
|
|
290 | |
|
|
291 | $current_priority = $w->priority |
|
|
292 | $old_priority = $w->priority ($new_priority) |
|
|
293 | Queries the priority on the watcher and optionally changes it. |
|
|
294 | Pending watchers with higher priority will be invoked first. The |
|
|
295 | valid range of priorities lies between EV::MAXPRI (default 2) and |
|
|
296 | EV::MINPRI (default -2). If the priority is outside this range it |
|
|
297 | will automatically be normalised to the nearest valid priority. |
|
|
298 | |
|
|
299 | The default priority of any newly-created watcher is 0. |
|
|
300 | |
|
|
301 | Note that the priority semantics have not yet been fleshed out and |
|
|
302 | are subject to almost certain change. |
|
|
303 | |
|
|
304 | $w->invoke ($revents) |
|
|
305 | Call the callback *now* with the given event mask. |
|
|
306 | |
|
|
307 | $w->feed_event ($revents) |
|
|
308 | Feed some events on this watcher into EV. EV will react to this call |
|
|
309 | as if the watcher had received the given $revents mask. |
|
|
310 | |
|
|
311 | $revents = $w->clear_pending |
|
|
312 | If the watcher is pending, this function clears its pending status |
|
|
313 | and returns its $revents bitset (as if its callback was invoked). If |
|
|
314 | the watcher isn't pending it does nothing and returns 0. |
|
|
315 | |
|
|
316 | $previous_state = $w->keepalive ($bool) |
|
|
317 | Normally, "EV::loop" will return when there are no active watchers |
|
|
318 | (which is a "deadlock" because no progress can be made anymore). |
|
|
319 | This is convinient because it allows you to start your watchers (and |
|
|
320 | your jobs), call "EV::loop" once and when it returns you know that |
|
|
321 | all your jobs are finished (or they forgot to register some watchers |
|
|
322 | for their task :). |
|
|
323 | |
|
|
324 | Sometimes, however, this gets in your way, for example when the |
|
|
325 | module that calls "EV::loop" (usually the main program) is not the |
|
|
326 | same module as a long-living watcher (for example a DNS client |
|
|
327 | module written by somebody else even). Then you might want any |
|
|
328 | outstanding requests to be handled, but you would not want to keep |
|
|
329 | "EV::loop" from returning just because you happen to have this |
|
|
330 | long-running UDP port watcher. |
|
|
331 | |
|
|
332 | In this case you can clear the keepalive status, which means that |
|
|
333 | even though your watcher is active, it won't keep "EV::loop" from |
|
|
334 | returning. |
|
|
335 | |
|
|
336 | The initial value for keepalive is true (enabled), and you cna |
|
|
337 | change it any time. |
|
|
338 | |
|
|
339 | Example: Register an I/O watcher for some UDP socket but do not keep |
|
|
340 | the event loop from running just because of that watcher. |
|
|
341 | |
|
|
342 | my $udp_socket = ... |
|
|
343 | my $udp_watcher = EV::io $udp_socket, EV::READ, sub { ... }; |
|
|
344 | $udp_watcher->keepalive (0); |
|
|
345 | |
|
|
346 | $loop = $w->loop |
|
|
347 | Return the loop that this watcher is attached to. |
|
|
348 | |
|
|
349 | WATCHER TYPES |
|
|
350 | Each of the following subsections describes a single watcher type. |
|
|
351 | |
|
|
352 | I/O WATCHERS - is this file descriptor readable or writable? |
104 | my $w = EV::io $fileno_or_fh, $eventmask, $callback |
353 | $w = EV::io $fileno_or_fh, $eventmask, $callback |
105 | my $w = EV::io_ns $fileno_or_fh, $eventmask, $callback |
354 | $w = EV::io_ns $fileno_or_fh, $eventmask, $callback |
|
|
355 | $w = $loop->io ($fileno_or_fh, $eventmask, $callback) |
|
|
356 | $w = $loop->io_ns ($fileno_or_fh, $eventmask, $callback) |
106 | As long as the returned watcher object is alive, call the $callback |
357 | As long as the returned watcher object is alive, call the $callback |
107 | when the events specified in $eventmask happen. Initially, the |
358 | when at least one of events specified in $eventmask occurs. |
108 | timeout is disabled. |
|
|
109 | |
359 | |
110 | You can additionall set a timeout to occur on the watcher, but note |
|
|
111 | that this timeout will not be reset when you get an I/O event in the |
|
|
112 | EV::PERSIST case, and reaching a timeout will always stop the |
|
|
113 | watcher even in the EV::PERSIST case. |
|
|
114 | |
|
|
115 | If you want a timeout to occur only after a specific time of |
|
|
116 | inactivity, set a repeating timeout and do NOT use EV::PERSIST. |
|
|
117 | |
|
|
118 | Eventmask can be one or more of these constants ORed together: |
360 | The $eventmask can be one or more of these constants ORed together: |
119 | |
361 | |
120 | EV::READ wait until read() wouldn't block anymore |
362 | EV::READ wait until read() wouldn't block anymore |
121 | EV::WRITE wait until write() wouldn't block anymore |
363 | EV::WRITE wait until write() wouldn't block anymore |
122 | EV::PERSIST stay active after a (non-timeout) event occured |
|
|
123 | |
364 | |
124 | The "io_ns" variant doesn't add/start the newly created watcher. |
365 | The "io_ns" variant doesn't start (activate) the newly created |
125 | |
|
|
126 | my $w = EV::timed_io $fileno_or_fh, $eventmask, $timeout, $callback |
|
|
127 | my $w = EV::timed_io_ns $fileno_or_fh, $eventmask, $timeout, $callback |
|
|
128 | Same as "io" and "io_ns", but also specifies a timeout (as if there |
|
|
129 | was a call to "$w->timeout ($timout, 1)". The persist flag is not |
|
|
130 | allowed and will automatically be cleared. The watcher will be |
|
|
131 | restarted after each event. |
|
|
132 | |
|
|
133 | If the timeout is zero or undef, no timeout will be set, and a |
|
|
134 | normal watcher (with the persist flag set!) will be created. |
|
|
135 | |
|
|
136 | This has the effect of timing out after the specified period of |
|
|
137 | inactivity has happened. |
|
|
138 | |
|
|
139 | Due to the design of libevent, this is also relatively inefficient, |
|
|
140 | having one or two io watchers and a separate timeout watcher that |
|
|
141 | you reset on activity (by calling its "start" method) is usually |
|
|
142 | more efficient. |
|
|
143 | |
|
|
144 | my $w = EV::timer $after, $repeat, $callback |
|
|
145 | my $w = EV::timer_ns $after, $repeat, $callback |
|
|
146 | Calls the callback after $after seconds. If $repeat is true, the |
|
|
147 | timer will be restarted after the callback returns. This means that |
|
|
148 | the callback would be called roughly every $after seconds, prolonged |
|
|
149 | by the time the callback takes. |
|
|
150 | |
|
|
151 | The "timer_ns" variant doesn't add/start the newly created watcher. |
|
|
152 | |
|
|
153 | my $w = EV::timer_abs $at, $interval, $callback |
|
|
154 | my $w = EV::timer_abs_ns $at, $interval, $callback |
|
|
155 | Similar to EV::timer, but the time is given as an absolute point in |
|
|
156 | time ($at), plus an optional $interval. |
|
|
157 | |
|
|
158 | If the $interval is zero, then the callback will be called at the |
|
|
159 | time $at if that is in the future, or as soon as possible if its in |
|
|
160 | the past. It will not automatically repeat. |
|
|
161 | |
|
|
162 | If the $interval is nonzero, then the watcher will always be |
|
|
163 | scheduled to time out at the next "$at + integer * $interval" time. |
|
|
164 | |
|
|
165 | This can be used to schedule a callback to run at very regular |
|
|
166 | intervals, as long as the processing time is less then the interval |
|
|
167 | (otherwise obviously events will be skipped). |
|
|
168 | |
|
|
169 | Another way to think about it (for the mathematically inclined) is |
|
|
170 | that "timer_abs" will try to tun the callback at the next possible |
|
|
171 | time where "$time = $at (mod $interval)", regardless of any time |
|
|
172 | jumps. |
|
|
173 | |
|
|
174 | The "timer_abs_ns" variant doesn't add/start the newly created |
|
|
175 | watcher. |
366 | watcher. |
176 | |
367 | |
177 | my $w = EV::signal $signal, $callback |
368 | $w->set ($fileno_or_fh, $eventmask) |
178 | my $w = EV::signal_ns $signal, $callback |
369 | Reconfigures the watcher, see the constructor above for details. Can |
179 | Call the callback when $signal is received (the signal can be |
370 | be called at any time. |
180 | specified by number or by name, just as with kill or %SIG). Signal |
|
|
181 | watchers are persistent no natter what. |
|
|
182 | |
|
|
183 | EV will grab the signal for the process (the kernel only allows one |
|
|
184 | component to receive signals) when you start a signal watcher, and |
|
|
185 | removes it again when you stop it. Pelr does the same when you |
|
|
186 | add/remove callbacks to %SIG, so watch out. |
|
|
187 | |
|
|
188 | Unfortunately, only one handler can be registered per signal. Screw |
|
|
189 | libevent. |
|
|
190 | |
|
|
191 | The "signal_ns" variant doesn't add/start the newly created watcher. |
|
|
192 | |
|
|
193 | THE EV::Event CLASS |
|
|
194 | All EV functions creating an event watcher (designated by "my $w =" |
|
|
195 | above) support the following methods on the returned watcher object: |
|
|
196 | |
|
|
197 | $w->add ($timeout) |
|
|
198 | Stops and (re-)starts the event watcher, setting the optional |
|
|
199 | timeout to the given value, or clearing the timeout if none is |
|
|
200 | given. |
|
|
201 | |
|
|
202 | $w->start |
|
|
203 | Stops and (re-)starts the event watcher without touching the |
|
|
204 | timeout. |
|
|
205 | |
|
|
206 | $w->del |
|
|
207 | $w->stop |
|
|
208 | Stop the event watcher if it was started. |
|
|
209 | |
|
|
210 | $current_callback = $w->cb |
|
|
211 | $old_callback = $w->cb ($new_callback) |
|
|
212 | Return the previously set callback and optionally set a new one. |
|
|
213 | |
371 | |
214 | $current_fh = $w->fh |
372 | $current_fh = $w->fh |
215 | $old_fh = $w->fh ($new_fh) |
373 | $old_fh = $w->fh ($new_fh) |
216 | Returns the previously set filehandle and optionally set a new one |
374 | Returns the previously set filehandle and optionally set a new one. |
217 | (also clears the EV::SIGNAL flag when setting a filehandle). |
|
|
218 | |
|
|
219 | $current_signal = $w->signal |
|
|
220 | $old_signal = $w->signal ($new_signal) |
|
|
221 | Returns the previously set signal number and optionally set a new |
|
|
222 | one (also sets the EV::SIGNAL flag when setting a signal). |
|
|
223 | |
375 | |
224 | $current_eventmask = $w->events |
376 | $current_eventmask = $w->events |
225 | $old_eventmask = $w->events ($new_eventmask) |
377 | $old_eventmask = $w->events ($new_eventmask) |
226 | Returns the previously set event mask and optionally set a new one. |
378 | Returns the previously set event mask and optionally set a new one. |
227 | |
379 | |
|
|
380 | TIMER WATCHERS - relative and optionally repeating timeouts |
|
|
381 | $w = EV::timer $after, $repeat, $callback |
|
|
382 | $w = EV::timer_ns $after, $repeat, $callback |
|
|
383 | $w = $loop->timer ($after, $repeat, $callback) |
|
|
384 | $w = $loop->timer_ns ($after, $repeat, $callback) |
|
|
385 | Calls the callback after $after seconds (which may be fractional). |
|
|
386 | If $repeat is non-zero, the timer will be restarted (with the |
|
|
387 | $repeat value as $after) after the callback returns. |
|
|
388 | |
|
|
389 | This means that the callback would be called roughly after $after |
|
|
390 | seconds, and then every $repeat seconds. The timer does his best not |
|
|
391 | to drift, but it will not invoke the timer more often then once per |
|
|
392 | event loop iteration, and might drift in other cases. If that isn't |
|
|
393 | acceptable, look at EV::periodic, which can provide long-term stable |
|
|
394 | timers. |
|
|
395 | |
|
|
396 | The timer is based on a monotonic clock, that is, if somebody is |
|
|
397 | sitting in front of the machine while the timer is running and |
|
|
398 | changes the system clock, the timer will nevertheless run (roughly) |
|
|
399 | the same time. |
|
|
400 | |
|
|
401 | The "timer_ns" variant doesn't start (activate) the newly created |
|
|
402 | watcher. |
|
|
403 | |
228 | $w->timeout ($after, $repeat) |
404 | $w->set ($after, $repeat) |
229 | Resets the timeout (see "EV::timer" for details). |
405 | Reconfigures the watcher, see the constructor above for details. Can |
|
|
406 | be called at any time. |
230 | |
407 | |
|
|
408 | $w->again |
|
|
409 | Similar to the "start" method, but has special semantics for |
|
|
410 | repeating timers: |
|
|
411 | |
|
|
412 | If the timer is active and non-repeating, it will be stopped. |
|
|
413 | |
|
|
414 | If the timer is active and repeating, reset the timeout to occur |
|
|
415 | $repeat seconds after now. |
|
|
416 | |
|
|
417 | If the timer is inactive and repeating, start it using the repeat |
|
|
418 | value. |
|
|
419 | |
|
|
420 | Otherwise do nothing. |
|
|
421 | |
|
|
422 | This behaviour is useful when you have a timeout for some IO |
|
|
423 | operation. You create a timer object with the same value for $after |
|
|
424 | and $repeat, and then, in the read/write watcher, run the "again" |
|
|
425 | method on the timeout. |
|
|
426 | |
|
|
427 | PERIODIC WATCHERS - to cron or not to cron? |
|
|
428 | $w = EV::periodic $at, $interval, $reschedule_cb, $callback |
|
|
429 | $w = EV::periodic_ns $at, $interval, $reschedule_cb, $callback |
|
|
430 | $w = $loop->periodic ($at, $interval, $reschedule_cb, $callback) |
|
|
431 | $w = $loop->periodic_ns ($at, $interval, $reschedule_cb, $callback) |
|
|
432 | Similar to EV::timer, but is not based on relative timeouts but on |
|
|
433 | absolute times. Apart from creating "simple" timers that trigger |
|
|
434 | "at" the specified time, it can also be used for non-drifting |
|
|
435 | absolute timers and more complex, cron-like, setups that are not |
|
|
436 | adversely affected by time jumps (i.e. when the system clock is |
|
|
437 | changed by explicit date -s or other means such as ntpd). It is also |
|
|
438 | the most complex watcher type in EV. |
|
|
439 | |
|
|
440 | It has three distinct "modes": |
|
|
441 | |
|
|
442 | * absolute timer ($interval = $reschedule_cb = 0) |
|
|
443 | |
|
|
444 | This time simply fires at the wallclock time $at and doesn't |
|
|
445 | repeat. It will not adjust when a time jump occurs, that is, if |
|
|
446 | it is to be run at January 1st 2011 then it will run when the |
|
|
447 | system time reaches or surpasses this time. |
|
|
448 | |
|
|
449 | * repeating interval timer ($interval > 0, $reschedule_cb = 0) |
|
|
450 | |
|
|
451 | In this mode the watcher will always be scheduled to time out at |
|
|
452 | the next "$at + N * $interval" time (for some integer N) and |
|
|
453 | then repeat, regardless of any time jumps. |
|
|
454 | |
|
|
455 | This can be used to create timers that do not drift with respect |
|
|
456 | to system time: |
|
|
457 | |
|
|
458 | my $hourly = EV::periodic 0, 3600, 0, sub { print "once/hour\n" }; |
|
|
459 | |
|
|
460 | That doesn't mean there will always be 3600 seconds in between |
|
|
461 | triggers, but only that the the clalback will be called when the |
|
|
462 | system time shows a full hour (UTC). |
|
|
463 | |
|
|
464 | Another way to think about it (for the mathematically inclined) |
|
|
465 | is that EV::periodic will try to run the callback in this mode |
|
|
466 | at the next possible time where "$time = $at (mod $interval)", |
|
|
467 | regardless of any time jumps. |
|
|
468 | |
|
|
469 | * manual reschedule mode ($reschedule_cb = coderef) |
|
|
470 | |
|
|
471 | In this mode $interval and $at are both being ignored. Instead, |
|
|
472 | each time the periodic watcher gets scheduled, the reschedule |
|
|
473 | callback ($reschedule_cb) will be called with the watcher as |
|
|
474 | first, and the current time as second argument. |
|
|
475 | |
|
|
476 | *This callback MUST NOT stop or destroy this or any other |
|
|
477 | periodic watcher, ever, and MUST NOT call any event loop |
|
|
478 | functions or methods*. If you need to stop it, return 1e30 and |
|
|
479 | stop it afterwards. You may create and start a "EV::prepare" |
|
|
480 | watcher for this task. |
|
|
481 | |
|
|
482 | It must return the next time to trigger, based on the passed |
|
|
483 | time value (that is, the lowest time value larger than or equal |
|
|
484 | to to the second argument). It will usually be called just |
|
|
485 | before the callback will be triggered, but might be called at |
|
|
486 | other times, too. |
|
|
487 | |
|
|
488 | This can be used to create very complex timers, such as a timer |
|
|
489 | that triggers on each midnight, local time (actually 24 hours |
|
|
490 | after the last midnight, to keep the example simple. If you know |
|
|
491 | a way to do it correctly in about the same space (without |
|
|
492 | requiring elaborate modules), drop me a note :): |
|
|
493 | |
|
|
494 | my $daily = EV::periodic 0, 0, sub { |
|
|
495 | my ($w, $now) = @_; |
|
|
496 | |
|
|
497 | use Time::Local (); |
|
|
498 | my (undef, undef, undef, $d, $m, $y) = localtime $now; |
|
|
499 | 86400 + Time::Local::timelocal 0, 0, 0, $d, $m, $y |
|
|
500 | }, sub { |
|
|
501 | print "it's midnight or likely shortly after, now\n"; |
|
|
502 | }; |
|
|
503 | |
|
|
504 | The "periodic_ns" variant doesn't start (activate) the newly created |
|
|
505 | watcher. |
|
|
506 | |
|
|
507 | $w->set ($at, $interval, $reschedule_cb) |
|
|
508 | Reconfigures the watcher, see the constructor above for details. Can |
|
|
509 | be called at any time. |
|
|
510 | |
|
|
511 | $w->again |
|
|
512 | Simply stops and starts the watcher again. |
|
|
513 | |
|
|
514 | $time = $w->at |
|
|
515 | Return the time that the watcher is expected to trigger next. |
|
|
516 | |
|
|
517 | SIGNAL WATCHERS - signal me when a signal gets signalled! |
|
|
518 | $w = EV::signal $signal, $callback |
|
|
519 | $w = EV::signal_ns $signal, $callback |
|
|
520 | Call the callback when $signal is received (the signal can be |
|
|
521 | specified by number or by name, just as with "kill" or %SIG). |
|
|
522 | |
|
|
523 | EV will grab the signal for the process (the kernel only allows one |
|
|
524 | component to receive a signal at a time) when you start a signal |
|
|
525 | watcher, and removes it again when you stop it. Perl does the same |
|
|
526 | when you add/remove callbacks to %SIG, so watch out. |
|
|
527 | |
|
|
528 | You can have as many signal watchers per signal as you want. |
|
|
529 | |
|
|
530 | The "signal_ns" variant doesn't start (activate) the newly created |
|
|
531 | watcher. |
|
|
532 | |
|
|
533 | $w->set ($signal) |
|
|
534 | Reconfigures the watcher, see the constructor above for details. Can |
|
|
535 | be called at any time. |
|
|
536 | |
|
|
537 | $current_signum = $w->signal |
|
|
538 | $old_signum = $w->signal ($new_signal) |
|
|
539 | Returns the previously set signal (always as a number not name) and |
|
|
540 | optionally set a new one. |
|
|
541 | |
|
|
542 | CHILD WATCHERS - watch out for process status changes |
|
|
543 | $w = EV::child $pid, $trace, $callback |
|
|
544 | $w = EV::child_ns $pid, $trace, $callback |
|
|
545 | $w = $loop->child ($pid, $trace, $callback) |
|
|
546 | $w = $loop->child_ns ($pid, $trace, $callback) |
|
|
547 | Call the callback when a status change for pid $pid (or any pid if |
|
|
548 | $pid is 0) has been received (a status change happens when the |
|
|
549 | process terminates or is killed, or, when trace is true, |
|
|
550 | additionally when it is stopped or continued). More precisely: when |
|
|
551 | the process receives a "SIGCHLD", EV will fetch the outstanding |
|
|
552 | exit/wait status for all changed/zombie children and call the |
|
|
553 | callback. |
|
|
554 | |
|
|
555 | It is valid (and fully supported) to install a child watcher after a |
|
|
556 | child has exited but before the event loop has started its next |
|
|
557 | iteration (for example, first you "fork", then the new child process |
|
|
558 | might exit, and only then do you install a child watcher in the |
|
|
559 | parent for the new pid). |
|
|
560 | |
|
|
561 | You can access both exit (or tracing) status and pid by using the |
|
|
562 | "rstatus" and "rpid" methods on the watcher object. |
|
|
563 | |
|
|
564 | You can have as many pid watchers per pid as you want, they will all |
|
|
565 | be called. |
|
|
566 | |
|
|
567 | The "child_ns" variant doesn't start (activate) the newly created |
|
|
568 | watcher. |
|
|
569 | |
|
|
570 | $w->set ($pid, $trace) |
|
|
571 | Reconfigures the watcher, see the constructor above for details. Can |
|
|
572 | be called at any time. |
|
|
573 | |
|
|
574 | $current_pid = $w->pid |
|
|
575 | Returns the previously set process id and optionally set a new one. |
|
|
576 | |
|
|
577 | $exit_status = $w->rstatus |
|
|
578 | Return the exit/wait status (as returned by waitpid, see the waitpid |
|
|
579 | entry in perlfunc). |
|
|
580 | |
|
|
581 | $pid = $w->rpid |
|
|
582 | Return the pid of the awaited child (useful when you have installed |
|
|
583 | a watcher for all pids). |
|
|
584 | |
|
|
585 | STAT WATCHERS - did the file attributes just change? |
|
|
586 | $w = EV::stat $path, $interval, $callback |
|
|
587 | $w = EV::stat_ns $path, $interval, $callback |
|
|
588 | $w = $loop->stat ($path, $interval, $callback) |
|
|
589 | $w = $loop->stat_ns ($path, $interval, $callback) |
|
|
590 | Call the callback when a file status change has been detected on |
|
|
591 | $path. The $path does not need to exist, changing from "path exists" |
|
|
592 | to "path does not exist" is a status change like any other. |
|
|
593 | |
|
|
594 | The $interval is a recommended polling interval for systems where |
|
|
595 | OS-supported change notifications don't exist or are not supported. |
|
|
596 | If you use 0 then an unspecified default is used (which is highly |
|
|
597 | recommended!), which is to be expected to be around five seconds |
|
|
598 | usually. |
|
|
599 | |
|
|
600 | This watcher type is not meant for massive numbers of stat watchers, |
|
|
601 | as even with OS-supported change notifications, this can be |
|
|
602 | resource-intensive. |
|
|
603 | |
|
|
604 | The "stat_ns" variant doesn't start (activate) the newly created |
|
|
605 | watcher. |
|
|
606 | |
|
|
607 | ... = $w->stat |
|
|
608 | This call is very similar to the perl "stat" built-in: It stats |
|
|
609 | (using "lstat") the path specified in the watcher and sets perls |
|
|
610 | stat cache (as well as EV's idea of the current stat values) to the |
|
|
611 | values found. |
|
|
612 | |
|
|
613 | In scalar context, a boolean is return indicating success or failure |
|
|
614 | of the stat. In list context, the same 13-value list as with stat is |
|
|
615 | returned (except that the blksize and blocks fields are not |
|
|
616 | reliable). |
|
|
617 | |
|
|
618 | In the case of an error, errno is set to "ENOENT" (regardless of the |
|
|
619 | actual error value) and the "nlink" value is forced to zero (if the |
|
|
620 | stat was successful then nlink is guaranteed to be non-zero). |
|
|
621 | |
|
|
622 | See also the next two entries for more info. |
|
|
623 | |
|
|
624 | ... = $w->attr |
|
|
625 | Just like "$w->stat", but without the initial stat'ing: this returns |
|
|
626 | the values most recently detected by EV. See the next entry for more |
|
|
627 | info. |
|
|
628 | |
|
|
629 | ... = $w->prev |
|
|
630 | Just like "$w->stat", but without the initial stat'ing: this returns |
|
|
631 | the previous set of values, before the change. |
|
|
632 | |
|
|
633 | That is, when the watcher callback is invoked, "$w->prev" will be |
|
|
634 | set to the values found *before* a change was detected, while |
|
|
635 | "$w->attr" returns the values found leading to the change detection. |
|
|
636 | The difference (if any) between "prev" and "attr" is what triggered |
|
|
637 | the callback. |
|
|
638 | |
|
|
639 | If you did something to the filesystem object and do not want to |
|
|
640 | trigger yet another change, you can call "stat" to update EV's idea |
|
|
641 | of what the current attributes are. |
|
|
642 | |
231 | $w->timeout_abs ($at, $interval) |
643 | $w->set ($path, $interval) |
232 | Resets the timeout (see "EV::timer_abs" for details). |
644 | Reconfigures the watcher, see the constructor above for details. Can |
|
|
645 | be called at any time. |
233 | |
646 | |
234 | $w->priority_set ($priority) |
647 | $current_path = $w->path |
235 | Set the priority of the watcher to $priority (0 <= $priority < |
648 | $old_path = $w->path ($new_path) |
236 | $EV::NPRI). |
649 | Returns the previously set path and optionally set a new one. |
|
|
650 | |
|
|
651 | $current_interval = $w->interval |
|
|
652 | $old_interval = $w->interval ($new_interval) |
|
|
653 | Returns the previously set interval and optionally set a new one. |
|
|
654 | Can be used to query the actual interval used. |
|
|
655 | |
|
|
656 | IDLE WATCHERS - when you've got nothing better to do... |
|
|
657 | $w = EV::idle $callback |
|
|
658 | $w = EV::idle_ns $callback |
|
|
659 | $w = $loop->idle ($callback) |
|
|
660 | $w = $loop->idle_ns ($callback) |
|
|
661 | Call the callback when there are no other pending watchers of the |
|
|
662 | same or higher priority (excluding check, prepare and other idle |
|
|
663 | watchers of the same or lower priority, of course). They are called |
|
|
664 | idle watchers because when the watcher is the highest priority |
|
|
665 | pending event in the process, the process is considered to be idle |
|
|
666 | at that priority. |
|
|
667 | |
|
|
668 | If you want a watcher that is only ever called when *no* other |
|
|
669 | events are outstanding you have to set the priority to "EV::MINPRI". |
|
|
670 | |
|
|
671 | The process will not block as long as any idle watchers are active, |
|
|
672 | and they will be called repeatedly until stopped. |
|
|
673 | |
|
|
674 | For example, if you have idle watchers at priority 0 and 1, and an |
|
|
675 | I/O watcher at priority 0, then the idle watcher at priority 1 and |
|
|
676 | the I/O watcher will always run when ready. Only when the idle |
|
|
677 | watcher at priority 1 is stopped and the I/O watcher at priority 0 |
|
|
678 | is not pending with the 0-priority idle watcher be invoked. |
|
|
679 | |
|
|
680 | The "idle_ns" variant doesn't start (activate) the newly created |
|
|
681 | watcher. |
|
|
682 | |
|
|
683 | PREPARE WATCHERS - customise your event loop! |
|
|
684 | $w = EV::prepare $callback |
|
|
685 | $w = EV::prepare_ns $callback |
|
|
686 | $w = $loop->prepare ($callback) |
|
|
687 | $w = $loop->prepare_ns ($callback) |
|
|
688 | Call the callback just before the process would block. You can still |
|
|
689 | create/modify any watchers at this point. |
|
|
690 | |
|
|
691 | See the EV::check watcher, below, for explanations and an example. |
|
|
692 | |
|
|
693 | The "prepare_ns" variant doesn't start (activate) the newly created |
|
|
694 | watcher. |
|
|
695 | |
|
|
696 | CHECK WATCHERS - customise your event loop even more! |
|
|
697 | $w = EV::check $callback |
|
|
698 | $w = EV::check_ns $callback |
|
|
699 | $w = $loop->check ($callback) |
|
|
700 | $w = $loop->check_ns ($callback) |
|
|
701 | Call the callback just after the process wakes up again (after it |
|
|
702 | has gathered events), but before any other callbacks have been |
|
|
703 | invoked. |
|
|
704 | |
|
|
705 | This is used to integrate other event-based software into the EV |
|
|
706 | mainloop: You register a prepare callback and in there, you create |
|
|
707 | io and timer watchers as required by the other software. Here is a |
|
|
708 | real-world example of integrating Net::SNMP (with some details left |
|
|
709 | out): |
|
|
710 | |
|
|
711 | our @snmp_watcher; |
|
|
712 | |
|
|
713 | our $snmp_prepare = EV::prepare sub { |
|
|
714 | # do nothing unless active |
|
|
715 | $dispatcher->{_event_queue_h} |
|
|
716 | or return; |
|
|
717 | |
|
|
718 | # make the dispatcher handle any outstanding stuff |
|
|
719 | ... not shown |
|
|
720 | |
|
|
721 | # create an I/O watcher for each and every socket |
|
|
722 | @snmp_watcher = ( |
|
|
723 | (map { EV::io $_, EV::READ, sub { } } |
|
|
724 | keys %{ $dispatcher->{_descriptors} }), |
|
|
725 | |
|
|
726 | EV::timer +($event->[Net::SNMP::Dispatcher::_ACTIVE] |
|
|
727 | ? $event->[Net::SNMP::Dispatcher::_TIME] - EV::now : 0), |
|
|
728 | 0, sub { }, |
|
|
729 | ); |
|
|
730 | }; |
|
|
731 | |
|
|
732 | The callbacks are irrelevant (and are not even being called), the |
|
|
733 | only purpose of those watchers is to wake up the process as soon as |
|
|
734 | one of those events occurs (socket readable, or timer timed out). |
|
|
735 | The corresponding EV::check watcher will then clean up: |
|
|
736 | |
|
|
737 | our $snmp_check = EV::check sub { |
|
|
738 | # destroy all watchers |
|
|
739 | @snmp_watcher = (); |
|
|
740 | |
|
|
741 | # make the dispatcher handle any new stuff |
|
|
742 | ... not shown |
|
|
743 | }; |
|
|
744 | |
|
|
745 | The callbacks of the created watchers will not be called as the |
|
|
746 | watchers are destroyed before this cna happen (remember EV::check |
|
|
747 | gets called first). |
|
|
748 | |
|
|
749 | The "check_ns" variant doesn't start (activate) the newly created |
|
|
750 | watcher. |
|
|
751 | |
|
|
752 | FORK WATCHERS - the audacity to resume the event loop after a fork |
|
|
753 | Fork watchers are called when a "fork ()" was detected. The invocation |
|
|
754 | is done before the event loop blocks next and before "check" watchers |
|
|
755 | are being called, and only in the child after the fork. |
|
|
756 | |
|
|
757 | $w = EV::fork $callback |
|
|
758 | $w = EV::fork_ns $callback |
|
|
759 | $w = $loop->fork ($callback) |
|
|
760 | $w = $loop->fork_ns ($callback) |
|
|
761 | Call the callback before the event loop is resumed in the child |
|
|
762 | process after a fork. |
|
|
763 | |
|
|
764 | The "fork_ns" variant doesn't start (activate) the newly created |
|
|
765 | watcher. |
|
|
766 | |
|
|
767 | EMBED WATCHERS - when one backend isn't enough... |
|
|
768 | This is a rather advanced watcher type that lets you embed one event |
|
|
769 | loop into another (currently only IO events are supported in the |
|
|
770 | embedded loop, other types of watchers might be handled in a delayed or |
|
|
771 | incorrect fashion and must not be used). |
|
|
772 | |
|
|
773 | See the libev documentation at |
|
|
774 | <http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod#code_ev_embed_code |
|
|
775 | _when_one_backend_> for more details. |
|
|
776 | |
|
|
777 | In short, this watcher is most useful on BSD systems without working |
|
|
778 | kqueue to still be able to handle a large number of sockets: |
|
|
779 | |
|
|
780 | my $socket_loop; |
|
|
781 | |
|
|
782 | # check wether we use SELECT or POLL _and_ KQUEUE is supported |
|
|
783 | if ( |
|
|
784 | (EV::backend & (EV::BACKEND_POLL | EV::BACKEND_SELECT)) |
|
|
785 | && (EV::supported_backends & EV::embeddable_backends & EV::BACKEND_KQUEUE) |
|
|
786 | ) { |
|
|
787 | # use kqueue for sockets |
|
|
788 | $socket_loop = new EV::Loop EV::BACKEND_KQUEUE | EV::FLAG_NOENV; |
|
|
789 | } |
|
|
790 | |
|
|
791 | # use the default loop otherwise |
|
|
792 | $socket_loop ||= EV::default_loop; |
|
|
793 | |
|
|
794 | $w = EV::embed $otherloop, $callback |
|
|
795 | $w = EV::embed_ns $otherloop, $callback |
|
|
796 | $w = $loop->embed ($otherloop, $callback) |
|
|
797 | $w = $loop->embed_ns ($otherloop, $callback) |
|
|
798 | Call the callback when the embedded event loop ($otherloop) has any |
|
|
799 | I/O activity. The $callback should alwas be specified as "undef" in |
|
|
800 | this version of EV, which means the embedded event loop will be |
|
|
801 | managed automatically. |
|
|
802 | |
|
|
803 | The "embed_ns" variant doesn't start (activate) the newly created |
|
|
804 | watcher. |
|
|
805 | |
|
|
806 | ASYNC WATCHERS - how to wake up another event loop |
|
|
807 | Async watchers are provided by EV, but have little use in perl directly, |
|
|
808 | as perl neither supports threads nor direct access to signal handlers or |
|
|
809 | other contexts where they could be of value. |
|
|
810 | |
|
|
811 | It is, however, possible to use them from the XS level. |
|
|
812 | |
|
|
813 | Please see the libev documentation for further details. |
|
|
814 | |
|
|
815 | $w = EV::async $callback |
|
|
816 | $w = EV::async_ns $callback |
|
|
817 | $w->send |
|
|
818 | $bool = $w->async_pending |
|
|
819 | |
|
|
820 | PERL SIGNALS |
|
|
821 | While Perl signal handling (%SIG) is not affected by EV, the behaviour |
|
|
822 | with EV is as the same as any other C library: Perl-signals will only be |
|
|
823 | handled when Perl runs, which means your signal handler might be invoked |
|
|
824 | only the next time an event callback is invoked. |
|
|
825 | |
|
|
826 | The solution is to use EV signal watchers (see "EV::signal"), which will |
|
|
827 | ensure proper operations with regards to other event watchers. |
|
|
828 | |
|
|
829 | If you cannot do this for whatever reason, you can also force a watcher |
|
|
830 | to be called on every event loop iteration by installing a "EV::check" |
|
|
831 | watcher: |
|
|
832 | |
|
|
833 | my $async_check = EV::check sub { }; |
|
|
834 | |
|
|
835 | This ensures that perl gets into control for a short time to handle any |
|
|
836 | pending signals, and also ensures (slightly) slower overall operation. |
237 | |
837 | |
238 | THREADS |
838 | THREADS |
239 | Threads are not supported by this in any way. Perl pseudo-threads is |
839 | Threads are not supported by this module in any way. Perl pseudo-threads |
240 | evil and must die. |
840 | is evil stuff and must die. As soon as Perl gains real threads I will |
|
|
841 | work on thread support for it. |
|
|
842 | |
|
|
843 | FORK |
|
|
844 | Most of the "improved" event delivering mechanisms of modern operating |
|
|
845 | systems have quite a few problems with fork(2) (to put it bluntly: it is |
|
|
846 | not supported and usually destructive). Libev makes it possible to work |
|
|
847 | around this by having a function that recreates the kernel state after |
|
|
848 | fork in the child. |
|
|
849 | |
|
|
850 | On non-win32 platforms, this module requires the pthread_atfork |
|
|
851 | functionality to do this automatically for you. This function is quite |
|
|
852 | buggy on most BSDs, though, so YMMV. The overhead for this is quite |
|
|
853 | negligible, because everything the function currently does is set a flag |
|
|
854 | that is checked only when the event loop gets used the next time, so |
|
|
855 | when you do fork but not use EV, the overhead is minimal. |
|
|
856 | |
|
|
857 | On win32, there is no notion of fork so all this doesn't apply, of |
|
|
858 | course. |
241 | |
859 | |
242 | SEE ALSO |
860 | SEE ALSO |
243 | L<EV::DNS>, L<event(3)>, L<event.h>, L<evdns.h>. |
861 | EV::ADNS (asynchronous DNS), Glib::EV (makes Glib/Gtk2 use EV as event |
244 | L<EV::AnyEvent>. |
862 | loop), EV::Glib (embed Glib into EV), Coro::EV (efficient coroutines |
|
|
863 | with EV), Net::SNMP::EV (asynchronous SNMP), AnyEvent for event-loop |
|
|
864 | agnostic and portable event driven programming. |
245 | |
865 | |
246 | AUTHOR |
866 | AUTHOR |
247 | Marc Lehmann <schmorp@schmorp.de> |
867 | Marc Lehmann <schmorp@schmorp.de> |
248 | http://home.schmorp.de/ |
868 | http://home.schmorp.de/ |
249 | |
869 | |