1 | NAME |
1 | NAME |
2 | Coro - coroutine process abstraction |
2 | Coro - the only real threads in perl |
3 | |
3 | |
4 | SYNOPSIS |
4 | SYNOPSIS |
5 | use Coro; |
5 | use Coro; |
6 | |
6 | |
7 | async { |
7 | async { |
8 | # some asynchronous thread of execution |
8 | # some asynchronous thread of execution |
|
|
9 | print "2\n"; |
|
|
10 | cede; # yield back to main |
|
|
11 | print "4\n"; |
9 | }; |
12 | }; |
10 | |
13 | print "1\n"; |
11 | # alternatively create an async coroutine like this: |
14 | cede; # yield to coro |
12 | |
15 | print "3\n"; |
13 | sub some_func : Coro { |
16 | cede; # and again |
14 | # some more async code |
17 | |
15 | } |
18 | # use locking |
16 | |
19 | use Coro::Semaphore; |
17 | cede; |
20 | my $lock = new Coro::Semaphore; |
|
|
21 | my $locked; |
|
|
22 | |
|
|
23 | $lock->down; |
|
|
24 | $locked = 1; |
|
|
25 | $lock->up; |
18 | |
26 | |
19 | DESCRIPTION |
27 | DESCRIPTION |
20 | This module collection manages coroutines. Coroutines are similar to |
28 | For a tutorial-style introduction, please read the Coro::Intro manpage. |
|
|
29 | This manpage mainly contains reference information. |
|
|
30 | |
|
|
31 | This module collection manages continuations in general, most often in |
|
|
32 | the form of cooperative threads (also called coros, or simply "coro" in |
|
|
33 | the documentation). They are similar to kernel threads but don't (in |
21 | threads but don't run in parallel at the same time even on SMP machines. |
34 | general) run in parallel at the same time even on SMP machines. The |
22 | The specific flavor of coroutine used in this module also guarantees you |
35 | specific flavor of thread offered by this module also guarantees you |
23 | that it will not switch between coroutines unless necessary, at |
36 | that it will not switch between threads unless necessary, at |
24 | easily-identified points in your program, so locking and parallel access |
37 | easily-identified points in your program, so locking and parallel access |
25 | are rarely an issue, making coroutine programming much safer than |
38 | are rarely an issue, making thread programming much safer and easier |
26 | threads programming. |
39 | than using other thread models. |
27 | |
40 | |
28 | (Perl, however, does not natively support real threads but instead does |
41 | Unlike the so-called "Perl threads" (which are not actually real threads |
29 | a very slow and memory-intensive emulation of processes using threads. |
42 | but only the windows process emulation ported to unix, and as such act |
30 | This is a performance win on Windows machines, and a loss everywhere |
43 | as processes), Coro provides a full shared address space, which makes |
31 | else). |
44 | communication between threads very easy. And Coro's threads are fast, |
|
|
45 | too: disabling the Windows process emulation code in your perl and using |
|
|
46 | Coro can easily result in a two to four times speed increase for your |
|
|
47 | programs. A parallel matrix multiplication benchmark runs over 300 times |
|
|
48 | faster on a single core than perl's pseudo-threads on a quad core using |
|
|
49 | all four cores. |
32 | |
50 | |
|
|
51 | Coro achieves that by supporting multiple running interpreters that |
|
|
52 | share data, which is especially useful to code pseudo-parallel processes |
|
|
53 | and for event-based programming, such as multiple HTTP-GET requests |
|
|
54 | running concurrently. See Coro::AnyEvent to learn more on how to |
|
|
55 | integrate Coro into an event-based environment. |
|
|
56 | |
33 | In this module, coroutines are defined as "callchain + lexical variables |
57 | In this module, a thread is defined as "callchain + lexical variables + |
34 | + @_ + $_ + $@ + $/ + C stack), that is, a coroutine has its own |
58 | some package variables + C stack), that is, a thread has its own |
35 | callchain, its own set of lexicals and its own set of perls most |
59 | callchain, its own set of lexicals and its own set of perls most |
36 | important global variables. |
60 | important global variables (see Coro::State for more configuration and |
|
|
61 | background info). |
37 | |
62 | |
|
|
63 | See also the "SEE ALSO" section at the end of this document - the Coro |
|
|
64 | module family is quite large. |
|
|
65 | |
|
|
66 | GLOBAL VARIABLES |
38 | $main |
67 | $Coro::main |
39 | This coroutine represents the main program. |
68 | This variable stores the Coro object that represents the main |
|
|
69 | program. While you cna "ready" it and do most other things you can |
|
|
70 | do to coro, it is mainly useful to compare again $Coro::current, to |
|
|
71 | see whether you are running in the main program or not. |
40 | |
72 | |
41 | $current (or as function: current) |
73 | $Coro::current |
42 | The current coroutine (the last coroutine switched to). The initial |
74 | The Coro object representing the current coro (the last coro that |
43 | value is $main (of course). |
75 | the Coro scheduler switched to). The initial value is $Coro::main |
|
|
76 | (of course). |
44 | |
77 | |
45 | This variable is strictly *read-only*. It is provided for |
78 | This variable is strictly *read-only*. You can take copies of the |
46 | performance reasons. If performance is not essential you are |
79 | value stored in it and use it as any other Coro object, but you must |
47 | encouraged to use the "Coro::current" function instead. |
80 | not otherwise modify the variable itself. |
48 | |
81 | |
49 | $idle |
82 | $Coro::idle |
|
|
83 | This variable is mainly useful to integrate Coro into event loops. |
|
|
84 | It is usually better to rely on Coro::AnyEvent or Coro::EV, as this |
|
|
85 | is pretty low-level functionality. |
|
|
86 | |
|
|
87 | This variable stores either a Coro object or a callback. |
|
|
88 | |
50 | A callback that is called whenever the scheduler finds no ready |
89 | If it is a callback, the it is called whenever the scheduler finds |
51 | coroutines to run. The default implementation prints "FATAL: |
90 | no ready coros to run. The default implementation prints "FATAL: |
52 | deadlock detected" and exits, because the program has no other way |
91 | deadlock detected" and exits, because the program has no other way |
53 | to continue. |
92 | to continue. |
54 | |
93 | |
|
|
94 | If it is a coro object, then this object will be readied (without |
|
|
95 | invoking any ready hooks, however) when the scheduler finds no other |
|
|
96 | ready coros to run. |
|
|
97 | |
55 | This hook is overwritten by modules such as "Coro::Timer" and |
98 | This hook is overwritten by modules such as "Coro::EV" and |
56 | "Coro::Event" to wait on an external event that hopefully wake up a |
99 | "Coro::AnyEvent" to wait on an external event that hopefully wake up |
57 | coroutine so the scheduler can run it. |
100 | a coro so the scheduler can run it. |
|
|
101 | |
|
|
102 | Note that the callback *must not*, under any circumstances, block |
|
|
103 | the current coro. Normally, this is achieved by having an "idle |
|
|
104 | coro" that calls the event loop and then blocks again, and then |
|
|
105 | readying that coro in the idle handler, or by simply placing the |
|
|
106 | idle coro in this variable. |
|
|
107 | |
|
|
108 | See Coro::Event or Coro::AnyEvent for examples of using this |
|
|
109 | technique. |
58 | |
110 | |
59 | Please note that if your callback recursively invokes perl (e.g. for |
111 | Please note that if your callback recursively invokes perl (e.g. for |
60 | event handlers), then it must be prepared to be called recursively. |
112 | event handlers), then it must be prepared to be called recursively |
|
|
113 | itself. |
61 | |
114 | |
62 | STATIC METHODS |
115 | SIMPLE CORO CREATION |
63 | Static methods are actually functions that operate on the current |
|
|
64 | coroutine only. |
|
|
65 | |
|
|
66 | async { ... } [@args...] |
116 | async { ... } [@args...] |
67 | Create a new asynchronous coroutine and return it's coroutine object |
117 | Create a new coro and return its Coro object (usually unused). The |
68 | (usually unused). When the sub returns the new coroutine is |
118 | coro will be put into the ready queue, so it will start running |
69 | automatically terminated. |
119 | automatically on the next scheduler run. |
70 | |
120 | |
|
|
121 | The first argument is a codeblock/closure that should be executed in |
|
|
122 | the coro. When it returns argument returns the coro is automatically |
|
|
123 | terminated. |
|
|
124 | |
|
|
125 | The remaining arguments are passed as arguments to the closure. |
|
|
126 | |
71 | See the "Coro::State::new" constructor for info about the coroutine |
127 | See the "Coro::State::new" constructor for info about the coro |
72 | environment. |
128 | environment in which coro are executed. |
73 | |
129 | |
74 | Calling "exit" in a coroutine will do the same as calling exit |
130 | Calling "exit" in a coro will do the same as calling exit outside |
75 | outside the coroutine. Likewise, when the coroutine dies, the |
131 | the coro. Likewise, when the coro dies, the program will exit, just |
76 | program will exit, just as it would in the main program. |
132 | as it would in the main program. |
77 | |
133 | |
|
|
134 | If you do not want that, you can provide a default "die" handler, or |
|
|
135 | simply avoid dieing (by use of "eval"). |
|
|
136 | |
78 | # create a new coroutine that just prints its arguments |
137 | Example: Create a new coro that just prints its arguments. |
|
|
138 | |
79 | async { |
139 | async { |
80 | print "@_\n"; |
140 | print "@_\n"; |
81 | } 1,2,3,4; |
141 | } 1,2,3,4; |
82 | |
142 | |
83 | async_pool { ... } [@args...] |
143 | async_pool { ... } [@args...] |
84 | Similar to "async", but uses a coroutine pool, so you should not |
144 | Similar to "async", but uses a coro pool, so you should not call |
85 | call terminate or join (although you are allowed to), and you get a |
145 | terminate or join on it (although you are allowed to), and you get a |
86 | coroutine that might have executed other code already (which can be |
146 | coro that might have executed other code already (which can be good |
87 | good or bad :). |
147 | or bad :). |
88 | |
148 | |
|
|
149 | On the plus side, this function is about twice as fast as creating |
|
|
150 | (and destroying) a completely new coro, so if you need a lot of |
|
|
151 | generic coros in quick successsion, use "async_pool", not "async". |
|
|
152 | |
89 | Also, the block is executed in an "eval" context and a warning will |
153 | The code block is executed in an "eval" context and a warning will |
90 | be issued in case of an exception instead of terminating the |
154 | be issued in case of an exception instead of terminating the |
91 | program, as "async" does. As the coroutine is being reused, stuff |
155 | program, as "async" does. As the coro is being reused, stuff like |
92 | like "on_destroy" will not work in the expected way, unless you call |
156 | "on_destroy" will not work in the expected way, unless you call |
93 | terminate or cancel, which somehow defeats the purpose of pooling. |
157 | terminate or cancel, which somehow defeats the purpose of pooling |
|
|
158 | (but is fine in the exceptional case). |
94 | |
159 | |
95 | The priority will be reset to 0 after each job, tracing will be |
160 | The priority will be reset to 0 after each run, tracing will be |
96 | disabled, the description will be reset and the default output |
161 | disabled, the description will be reset and the default output |
97 | filehandle gets restored, so you can change alkl these. Otherwise |
162 | filehandle gets restored, so you can change all these. Otherwise the |
98 | the coroutine will be re-used "as-is": most notably if you change |
163 | coro will be re-used "as-is": most notably if you change other |
99 | other per-coroutine global stuff such as $/ you need to revert that |
164 | per-coro global stuff such as $/ you *must needs* revert that |
100 | change, which is most simply done by using local as in " local $/ ". |
165 | change, which is most simply done by using local as in: "local $/". |
101 | |
166 | |
102 | The pool size is limited to 8 idle coroutines (this can be adjusted |
167 | The idle pool size is limited to 8 idle coros (this can be adjusted |
103 | by changing $Coro::POOL_SIZE), and there can be as many non-idle |
168 | by changing $Coro::POOL_SIZE), but there can be as many non-idle |
104 | coros as required. |
169 | coros as required. |
105 | |
170 | |
106 | If you are concerned about pooled coroutines growing a lot because a |
171 | If you are concerned about pooled coros growing a lot because a |
107 | single "async_pool" used a lot of stackspace you can e.g. |
172 | single "async_pool" used a lot of stackspace you can e.g. |
108 | "async_pool { terminate }" once per second or so to slowly replenish |
173 | "async_pool { terminate }" once per second or so to slowly replenish |
109 | the pool. In addition to that, when the stacks used by a handler |
174 | the pool. In addition to that, when the stacks used by a handler |
110 | grows larger than 16kb (adjustable with $Coro::POOL_RSS) it will |
175 | grows larger than 32kb (adjustable via $Coro::POOL_RSS) it will also |
111 | also exit. |
176 | be destroyed. |
|
|
177 | |
|
|
178 | STATIC METHODS |
|
|
179 | Static methods are actually functions that implicitly operate on the |
|
|
180 | current coro. |
112 | |
181 | |
113 | schedule |
182 | schedule |
114 | Calls the scheduler. Please note that the current coroutine will not |
183 | Calls the scheduler. The scheduler will find the next coro that is |
115 | be put into the ready queue, so calling this function usually means |
184 | to be run from the ready queue and switches to it. The next coro to |
|
|
185 | be run is simply the one with the highest priority that is longest |
|
|
186 | in its ready queue. If there is no coro ready, it will clal the |
|
|
187 | $Coro::idle hook. |
|
|
188 | |
|
|
189 | Please note that the current coro will *not* be put into the ready |
|
|
190 | queue, so calling this function usually means you will never be |
116 | you will never be called again unless something else (e.g. an event |
191 | called again unless something else (e.g. an event handler) calls |
117 | handler) calls ready. |
192 | "->ready", thus waking you up. |
118 | |
193 | |
119 | The canonical way to wait on external events is this: |
194 | This makes "schedule" *the* generic method to use to block the |
|
|
195 | current coro and wait for events: first you remember the current |
|
|
196 | coro in a variable, then arrange for some callback of yours to call |
|
|
197 | "->ready" on that once some event happens, and last you call |
|
|
198 | "schedule" to put yourself to sleep. Note that a lot of things can |
|
|
199 | wake your coro up, so you need to check whether the event indeed |
|
|
200 | happened, e.g. by storing the status in a variable. |
120 | |
201 | |
|
|
202 | See HOW TO WAIT FOR A CALLBACK, below, for some ways to wait for |
|
|
203 | callbacks. |
|
|
204 | |
|
|
205 | cede |
|
|
206 | "Cede" to other coros. This function puts the current coro into the |
|
|
207 | ready queue and calls "schedule", which has the effect of giving up |
|
|
208 | the current "timeslice" to other coros of the same or higher |
|
|
209 | priority. Once your coro gets its turn again it will automatically |
|
|
210 | be resumed. |
|
|
211 | |
|
|
212 | This function is often called "yield" in other languages. |
|
|
213 | |
|
|
214 | Coro::cede_notself |
|
|
215 | Works like cede, but is not exported by default and will cede to |
|
|
216 | *any* coro, regardless of priority. This is useful sometimes to |
|
|
217 | ensure progress is made. |
|
|
218 | |
|
|
219 | terminate [arg...] |
|
|
220 | Terminates the current coro with the given status values (see |
|
|
221 | cancel). |
|
|
222 | |
|
|
223 | Coro::on_enter BLOCK, Coro::on_leave BLOCK |
|
|
224 | These function install enter and leave winders in the current scope. |
|
|
225 | The enter block will be executed when on_enter is called and |
|
|
226 | whenever the current coro is re-entered by the scheduler, while the |
|
|
227 | leave block is executed whenever the current coro is blocked by the |
|
|
228 | scheduler, and also when the containing scope is exited (by whatever |
|
|
229 | means, be it exit, die, last etc.). |
|
|
230 | |
|
|
231 | *Neither invoking the scheduler, nor exceptions, are allowed within |
|
|
232 | those BLOCKs*. That means: do not even think about calling "die" |
|
|
233 | without an eval, and do not even think of entering the scheduler in |
|
|
234 | any way. |
|
|
235 | |
|
|
236 | Since both BLOCKs are tied to the current scope, they will |
|
|
237 | automatically be removed when the current scope exits. |
|
|
238 | |
|
|
239 | These functions implement the same concept as "dynamic-wind" in |
|
|
240 | scheme does, and are useful when you want to localise some resource |
|
|
241 | to a specific coro. |
|
|
242 | |
|
|
243 | They slow down thread switching considerably for coros that use them |
|
|
244 | (about 40% for a BLOCK with a single assignment, so thread switching |
|
|
245 | is still reasonably fast if the handlers are fast). |
|
|
246 | |
|
|
247 | These functions are best understood by an example: The following |
|
|
248 | function will change the current timezone to |
|
|
249 | "Antarctica/South_Pole", which requires a call to "tzset", but by |
|
|
250 | using "on_enter" and "on_leave", which remember/change the current |
|
|
251 | timezone and restore the previous value, respectively, the timezone |
|
|
252 | is only changed for the coro that installed those handlers. |
|
|
253 | |
|
|
254 | use POSIX qw(tzset); |
|
|
255 | |
121 | { |
256 | async { |
122 | # remember current coroutine |
257 | my $old_tz; # store outside TZ value here |
123 | my $current = $Coro::current; |
|
|
124 | |
258 | |
125 | # register a hypothetical event handler |
259 | Coro::on_enter { |
126 | on_event_invoke sub { |
260 | $old_tz = $ENV{TZ}; # remember the old value |
127 | # wake up sleeping coroutine |
261 | |
128 | $current->ready; |
262 | $ENV{TZ} = "Antarctica/South_Pole"; |
129 | undef $current; |
263 | tzset; # enable new value |
130 | }; |
264 | }; |
131 | |
265 | |
132 | # call schedule until event occurred. |
266 | Coro::on_leave { |
133 | # in case we are woken up for other reasons |
267 | $ENV{TZ} = $old_tz; |
134 | # (current still defined), loop. |
268 | tzset; # restore old value |
135 | Coro::schedule while $current; |
269 | }; |
|
|
270 | |
|
|
271 | # at this place, the timezone is Antarctica/South_Pole, |
|
|
272 | # without disturbing the TZ of any other coro. |
136 | } |
273 | }; |
137 | |
274 | |
138 | cede |
275 | This can be used to localise about any resource (locale, uid, |
139 | "Cede" to other coroutines. This function puts the current coroutine |
276 | current working directory etc.) to a block, despite the existance of |
140 | into the ready queue and calls "schedule", which has the effect of |
277 | other coros. |
141 | giving up the current "timeslice" to other coroutines of the same or |
|
|
142 | higher priority. |
|
|
143 | |
278 | |
144 | Returns true if at least one coroutine switch has happened. |
279 | Another interesting example implements time-sliced multitasking |
|
|
280 | using interval timers (this could obviously be optimised, but does |
|
|
281 | the job): |
145 | |
282 | |
146 | Coro::cede_notself |
283 | # "timeslice" the given block |
147 | Works like cede, but is not exported by default and will cede to any |
284 | sub timeslice(&) { |
148 | coroutine, regardless of priority, once. |
285 | use Time::HiRes (); |
149 | |
286 | |
150 | Returns true if at least one coroutine switch has happened. |
287 | Coro::on_enter { |
|
|
288 | # on entering the thread, we set an VTALRM handler to cede |
|
|
289 | $SIG{VTALRM} = sub { cede }; |
|
|
290 | # and then start the interval timer |
|
|
291 | Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0.01, 0.01; |
|
|
292 | }; |
|
|
293 | Coro::on_leave { |
|
|
294 | # on leaving the thread, we stop the interval timer again |
|
|
295 | Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0, 0; |
|
|
296 | }; |
151 | |
297 | |
152 | terminate [arg...] |
298 | &{+shift}; |
153 | Terminates the current coroutine with the given status values (see |
299 | } |
154 | cancel). |
300 | |
|
|
301 | # use like this: |
|
|
302 | timeslice { |
|
|
303 | # The following is an endless loop that would normally |
|
|
304 | # monopolise the process. Since it runs in a timesliced |
|
|
305 | # environment, it will regularly cede to other threads. |
|
|
306 | while () { } |
|
|
307 | }; |
155 | |
308 | |
156 | killall |
309 | killall |
157 | Kills/terminates/cancels all coroutines except the currently running |
310 | Kills/terminates/cancels all coros except the currently running one. |
158 | one. This is useful after a fork, either in the child or the parent, |
|
|
159 | as usually only one of them should inherit the running coroutines. |
|
|
160 | |
311 | |
161 | # dynamic methods |
312 | Note that while this will try to free some of the main interpreter |
|
|
313 | resources if the calling coro isn't the main coro, but one cannot |
|
|
314 | free all of them, so if a coro that is not the main coro calls this |
|
|
315 | function, there will be some one-time resource leak. |
162 | |
316 | |
163 | COROUTINE METHODS |
317 | CORO OBJECT METHODS |
164 | These are the methods you can call on coroutine objects. |
318 | These are the methods you can call on coro objects (or to create them). |
165 | |
319 | |
166 | new Coro \&sub [, @args...] |
320 | new Coro \&sub [, @args...] |
167 | Create a new coroutine and return it. When the sub returns the |
321 | Create a new coro and return it. When the sub returns, the coro |
168 | coroutine automatically terminates as if "terminate" with the |
322 | automatically terminates as if "terminate" with the returned values |
169 | returned values were called. To make the coroutine run you must |
323 | were called. To make the coro run you must first put it into the |
170 | first put it into the ready queue by calling the ready method. |
324 | ready queue by calling the ready method. |
171 | |
325 | |
172 | See "async" and "Coro::State::new" for additional info about the |
326 | See "async" and "Coro::State::new" for additional info about the |
173 | coroutine environment. |
327 | coro environment. |
174 | |
328 | |
175 | $success = $coroutine->ready |
329 | $success = $coro->ready |
176 | Put the given coroutine into the ready queue (according to it's |
330 | Put the given coro into the end of its ready queue (there is one |
177 | priority) and return true. If the coroutine is already in the ready |
331 | queue for each priority) and return true. If the coro is already in |
178 | queue, do nothing and return false. |
332 | the ready queue, do nothing and return false. |
179 | |
333 | |
|
|
334 | This ensures that the scheduler will resume this coro automatically |
|
|
335 | once all the coro of higher priority and all coro of the same |
|
|
336 | priority that were put into the ready queue earlier have been |
|
|
337 | resumed. |
|
|
338 | |
|
|
339 | $coro->suspend |
|
|
340 | Suspends the specified coro. A suspended coro works just like any |
|
|
341 | other coro, except that the scheduler will not select a suspended |
|
|
342 | coro for execution. |
|
|
343 | |
|
|
344 | Suspending a coro can be useful when you want to keep the coro from |
|
|
345 | running, but you don't want to destroy it, or when you want to |
|
|
346 | temporarily freeze a coro (e.g. for debugging) to resume it later. |
|
|
347 | |
|
|
348 | A scenario for the former would be to suspend all (other) coros |
|
|
349 | after a fork and keep them alive, so their destructors aren't |
|
|
350 | called, but new coros can be created. |
|
|
351 | |
|
|
352 | $coro->resume |
|
|
353 | If the specified coro was suspended, it will be resumed. Note that |
|
|
354 | when the coro was in the ready queue when it was suspended, it might |
|
|
355 | have been unreadied by the scheduler, so an activation might have |
|
|
356 | been lost. |
|
|
357 | |
|
|
358 | To avoid this, it is best to put a suspended coro into the ready |
|
|
359 | queue unconditionally, as every synchronisation mechanism must |
|
|
360 | protect itself against spurious wakeups, and the one in the Coro |
|
|
361 | family certainly do that. |
|
|
362 | |
180 | $is_ready = $coroutine->is_ready |
363 | $is_ready = $coro->is_ready |
181 | Return wether the coroutine is currently the ready queue or not, |
364 | Returns true iff the Coro object is in the ready queue. Unless the |
|
|
365 | Coro object gets destroyed, it will eventually be scheduled by the |
|
|
366 | scheduler. |
182 | |
367 | |
|
|
368 | $is_running = $coro->is_running |
|
|
369 | Returns true iff the Coro object is currently running. Only one Coro |
|
|
370 | object can ever be in the running state (but it currently is |
|
|
371 | possible to have multiple running Coro::States). |
|
|
372 | |
|
|
373 | $is_suspended = $coro->is_suspended |
|
|
374 | Returns true iff this Coro object has been suspended. Suspended |
|
|
375 | Coros will not ever be scheduled. |
|
|
376 | |
183 | $coroutine->cancel (arg...) |
377 | $coro->cancel (arg...) |
184 | Terminates the given coroutine and makes it return the given |
378 | Terminates the given Coro and makes it return the given arguments as |
185 | arguments as status (default: the empty list). Never returns if the |
379 | status (default: the empty list). Never returns if the Coro is the |
186 | coroutine is the current coroutine. |
380 | current Coro. |
187 | |
381 | |
188 | $coroutine->join |
382 | $coro->schedule_to |
189 | Wait until the coroutine terminates and return any values given to |
383 | Puts the current coro to sleep (like "Coro::schedule"), but instead |
190 | the "terminate" or "cancel" functions. "join" can be called |
384 | of continuing with the next coro from the ready queue, always switch |
191 | concurrently from multiple coroutines. |
385 | to the given coro object (regardless of priority etc.). The |
|
|
386 | readyness state of that coro isn't changed. |
192 | |
387 | |
193 | $coroutine->on_destroy (\&cb) |
388 | This is an advanced method for special cases - I'd love to hear |
194 | Registers a callback that is called when this coroutine gets |
389 | about any uses for this one. |
195 | destroyed, but before it is joined. The callback gets passed the |
|
|
196 | terminate arguments, if any. |
|
|
197 | |
390 | |
198 | $oldprio = $coroutine->prio ($newprio) |
391 | $coro->cede_to |
199 | Sets (or gets, if the argument is missing) the priority of the |
392 | Like "schedule_to", but puts the current coro into the ready queue. |
200 | coroutine. Higher priority coroutines get run before lower priority |
393 | This has the effect of temporarily switching to the given coro, and |
201 | coroutines. Priorities are small signed integers (currently -4 .. |
394 | continuing some time later. |
202 | +3), that you can refer to using PRIO_xxx constants (use the import |
|
|
203 | tag :prio to get then): |
|
|
204 | |
395 | |
205 | PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN |
396 | This is an advanced method for special cases - I'd love to hear |
206 | 3 > 1 > 0 > -1 > -3 > -4 |
397 | about any uses for this one. |
207 | |
398 | |
208 | # set priority to HIGH |
|
|
209 | current->prio(PRIO_HIGH); |
|
|
210 | |
|
|
211 | The idle coroutine ($Coro::idle) always has a lower priority than |
|
|
212 | any existing coroutine. |
|
|
213 | |
|
|
214 | Changing the priority of the current coroutine will take effect |
|
|
215 | immediately, but changing the priority of coroutines in the ready |
|
|
216 | queue (but not running) will only take effect after the next |
|
|
217 | schedule (of that coroutine). This is a bug that will be fixed in |
|
|
218 | some future version. |
|
|
219 | |
|
|
220 | $newprio = $coroutine->nice ($change) |
|
|
221 | Similar to "prio", but subtract the given value from the priority |
|
|
222 | (i.e. higher values mean lower priority, just as in unix). |
|
|
223 | |
|
|
224 | $olddesc = $coroutine->desc ($newdesc) |
|
|
225 | Sets (or gets in case the argument is missing) the description for |
|
|
226 | this coroutine. This is just a free-form string you can associate |
|
|
227 | with a coroutine. |
|
|
228 | |
|
|
229 | This method simply sets the "$coroutine->{desc}" member to the given |
|
|
230 | string. You can modify this member directly if you wish. |
|
|
231 | |
|
|
232 | $coroutine->throw ([$scalar]) |
399 | $coro->throw ([$scalar]) |
233 | If $throw is specified and defined, it will be thrown as an |
400 | If $throw is specified and defined, it will be thrown as an |
234 | exception inside the coroutine at the next convinient point in time |
401 | exception inside the coro at the next convenient point in time. |
235 | (usually after it gains control at the next schedule/transfer/cede). |
|
|
236 | Otherwise clears the exception object. |
402 | Otherwise clears the exception object. |
|
|
403 | |
|
|
404 | Coro will check for the exception each time a schedule-like-function |
|
|
405 | returns, i.e. after each "schedule", "cede", |
|
|
406 | "Coro::Semaphore->down", "Coro::Handle->readable" and so on. Most of |
|
|
407 | these functions detect this case and return early in case an |
|
|
408 | exception is pending. |
237 | |
409 | |
238 | The exception object will be thrown "as is" with the specified |
410 | The exception object will be thrown "as is" with the specified |
239 | scalar in $@, i.e. if it is a string, no line number or newline will |
411 | scalar in $@, i.e. if it is a string, no line number or newline will |
240 | be appended (unlike with "die"). |
412 | be appended (unlike with "die"). |
241 | |
413 | |
242 | This can be used as a softer means than "cancel" to ask a coroutine |
414 | This can be used as a softer means than "cancel" to ask a coro to |
243 | to end itself, although there is no guarentee that the exception |
415 | end itself, although there is no guarantee that the exception will |
244 | will lead to termination, and if the exception isn't caught it might |
416 | lead to termination, and if the exception isn't caught it might well |
245 | well end the whole program. |
417 | end the whole program. |
246 | |
418 | |
|
|
419 | You might also think of "throw" as being the moral equivalent of |
|
|
420 | "kill"ing a coro with a signal (in this case, a scalar). |
|
|
421 | |
|
|
422 | $coro->join |
|
|
423 | Wait until the coro terminates and return any values given to the |
|
|
424 | "terminate" or "cancel" functions. "join" can be called concurrently |
|
|
425 | from multiple coro, and all will be resumed and given the status |
|
|
426 | return once the $coro terminates. |
|
|
427 | |
|
|
428 | $coro->on_destroy (\&cb) |
|
|
429 | Registers a callback that is called when this coro gets destroyed, |
|
|
430 | but before it is joined. The callback gets passed the terminate |
|
|
431 | arguments, if any, and *must not* die, under any circumstances. |
|
|
432 | |
|
|
433 | $oldprio = $coro->prio ($newprio) |
|
|
434 | Sets (or gets, if the argument is missing) the priority of the coro. |
|
|
435 | Higher priority coro get run before lower priority coro. Priorities |
|
|
436 | are small signed integers (currently -4 .. +3), that you can refer |
|
|
437 | to using PRIO_xxx constants (use the import tag :prio to get then): |
|
|
438 | |
|
|
439 | PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN |
|
|
440 | 3 > 1 > 0 > -1 > -3 > -4 |
|
|
441 | |
|
|
442 | # set priority to HIGH |
|
|
443 | current->prio (PRIO_HIGH); |
|
|
444 | |
|
|
445 | The idle coro ($Coro::idle) always has a lower priority than any |
|
|
446 | existing coro. |
|
|
447 | |
|
|
448 | Changing the priority of the current coro will take effect |
|
|
449 | immediately, but changing the priority of coro in the ready queue |
|
|
450 | (but not running) will only take effect after the next schedule (of |
|
|
451 | that coro). This is a bug that will be fixed in some future version. |
|
|
452 | |
|
|
453 | $newprio = $coro->nice ($change) |
|
|
454 | Similar to "prio", but subtract the given value from the priority |
|
|
455 | (i.e. higher values mean lower priority, just as in unix). |
|
|
456 | |
|
|
457 | $olddesc = $coro->desc ($newdesc) |
|
|
458 | Sets (or gets in case the argument is missing) the description for |
|
|
459 | this coro. This is just a free-form string you can associate with a |
|
|
460 | coro. |
|
|
461 | |
|
|
462 | This method simply sets the "$coro->{desc}" member to the given |
|
|
463 | string. You can modify this member directly if you wish. |
|
|
464 | |
247 | GLOBAL FUNCTIONS |
465 | GLOBAL FUNCTIONS |
248 | Coro::nready |
466 | Coro::nready |
249 | Returns the number of coroutines that are currently in the ready |
467 | Returns the number of coro that are currently in the ready state, |
250 | state, i.e. that can be switched to. The value 0 means that the only |
468 | i.e. that can be switched to by calling "schedule" directory or |
251 | runnable coroutine is the currently running one, so "cede" would |
469 | indirectly. The value 0 means that the only runnable coro is the |
252 | have no effect, and "schedule" would cause a deadlock unless there |
470 | currently running one, so "cede" would have no effect, and |
253 | is an idle handler that wakes up some coroutines. |
471 | "schedule" would cause a deadlock unless there is an idle handler |
|
|
472 | that wakes up some coro. |
254 | |
473 | |
255 | my $guard = Coro::guard { ... } |
474 | my $guard = Coro::guard { ... } |
256 | This creates and returns a guard object. Nothing happens until the |
475 | This function still exists, but is deprecated. Please use the |
257 | object gets destroyed, in which case the codeblock given as argument |
476 | "Guard::guard" function instead. |
258 | will be executed. This is useful to free locks or other resources in |
|
|
259 | case of a runtime error or when the coroutine gets canceled, as in |
|
|
260 | both cases the guard block will be executed. The guard object |
|
|
261 | supports only one method, "->cancel", which will keep the codeblock |
|
|
262 | from being executed. |
|
|
263 | |
|
|
264 | Example: set some flag and clear it again when the coroutine gets |
|
|
265 | canceled or the function returns: |
|
|
266 | |
|
|
267 | sub do_something { |
|
|
268 | my $guard = Coro::guard { $busy = 0 }; |
|
|
269 | $busy = 1; |
|
|
270 | |
|
|
271 | # do something that requires $busy to be true |
|
|
272 | } |
|
|
273 | |
477 | |
274 | unblock_sub { ... } |
478 | unblock_sub { ... } |
275 | This utility function takes a BLOCK or code reference and "unblocks" |
479 | This utility function takes a BLOCK or code reference and "unblocks" |
276 | it, returning the new coderef. This means that the new coderef will |
480 | it, returning a new coderef. Unblocking means that calling the new |
277 | return immediately without blocking, returning nothing, while the |
481 | coderef will return immediately without blocking, returning nothing, |
278 | original code ref will be called (with parameters) from within its |
482 | while the original code ref will be called (with parameters) from |
279 | own coroutine. |
483 | within another coro. |
280 | |
484 | |
281 | The reason this function exists is that many event libraries (such |
485 | The reason this function exists is that many event libraries (such |
282 | as the venerable Event module) are not coroutine-safe (a weaker form |
486 | as the venerable Event module) are not thread-safe (a weaker form of |
283 | of thread-safety). This means you must not block within event |
487 | reentrancy). This means you must not block within event callbacks, |
284 | callbacks, otherwise you might suffer from crashes or worse. |
488 | otherwise you might suffer from crashes or worse. The only event |
|
|
489 | library currently known that is safe to use without "unblock_sub" is |
|
|
490 | EV. |
285 | |
491 | |
286 | This function allows your callbacks to block by executing them in |
492 | This function allows your callbacks to block by executing them in |
287 | another coroutine where it is safe to block. One example where |
493 | another coro where it is safe to block. One example where blocking |
288 | blocking is handy is when you use the Coro::AIO functions to save |
494 | is handy is when you use the Coro::AIO functions to save results to |
289 | results to disk. |
495 | disk, for example. |
290 | |
496 | |
291 | In short: simply use "unblock_sub { ... }" instead of "sub { ... }" |
497 | In short: simply use "unblock_sub { ... }" instead of "sub { ... }" |
292 | when creating event callbacks that want to block. |
498 | when creating event callbacks that want to block. |
293 | |
499 | |
|
|
500 | If your handler does not plan to block (e.g. simply sends a message |
|
|
501 | to another coro, or puts some other coro into the ready queue), |
|
|
502 | there is no reason to use "unblock_sub". |
|
|
503 | |
|
|
504 | Note that you also need to use "unblock_sub" for any other callbacks |
|
|
505 | that are indirectly executed by any C-based event loop. For example, |
|
|
506 | when you use a module that uses AnyEvent (and you use |
|
|
507 | Coro::AnyEvent) and it provides callbacks that are the result of |
|
|
508 | some event callback, then you must not block either, or use |
|
|
509 | "unblock_sub". |
|
|
510 | |
|
|
511 | $cb = Coro::rouse_cb |
|
|
512 | Create and return a "rouse callback". That's a code reference that, |
|
|
513 | when called, will remember a copy of its arguments and notify the |
|
|
514 | owner coro of the callback. |
|
|
515 | |
|
|
516 | See the next function. |
|
|
517 | |
|
|
518 | @args = Coro::rouse_wait [$cb] |
|
|
519 | Wait for the specified rouse callback (or the last one that was |
|
|
520 | created in this coro). |
|
|
521 | |
|
|
522 | As soon as the callback is invoked (or when the callback was invoked |
|
|
523 | before "rouse_wait"), it will return the arguments originally passed |
|
|
524 | to the rouse callback. In scalar context, that means you get the |
|
|
525 | *last* argument, just as if "rouse_wait" had a "return ($a1, $a2, |
|
|
526 | $a3...)" statement at the end. |
|
|
527 | |
|
|
528 | See the section HOW TO WAIT FOR A CALLBACK for an actual usage |
|
|
529 | example. |
|
|
530 | |
|
|
531 | HOW TO WAIT FOR A CALLBACK |
|
|
532 | It is very common for a coro to wait for some callback to be called. |
|
|
533 | This occurs naturally when you use coro in an otherwise event-based |
|
|
534 | program, or when you use event-based libraries. |
|
|
535 | |
|
|
536 | These typically register a callback for some event, and call that |
|
|
537 | callback when the event occured. In a coro, however, you typically want |
|
|
538 | to just wait for the event, simplyifying things. |
|
|
539 | |
|
|
540 | For example "AnyEvent->child" registers a callback to be called when a |
|
|
541 | specific child has exited: |
|
|
542 | |
|
|
543 | my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... }); |
|
|
544 | |
|
|
545 | But from within a coro, you often just want to write this: |
|
|
546 | |
|
|
547 | my $status = wait_for_child $pid; |
|
|
548 | |
|
|
549 | Coro offers two functions specifically designed to make this easy, |
|
|
550 | "Coro::rouse_cb" and "Coro::rouse_wait". |
|
|
551 | |
|
|
552 | The first function, "rouse_cb", generates and returns a callback that, |
|
|
553 | when invoked, will save its arguments and notify the coro that created |
|
|
554 | the callback. |
|
|
555 | |
|
|
556 | The second function, "rouse_wait", waits for the callback to be called |
|
|
557 | (by calling "schedule" to go to sleep) and returns the arguments |
|
|
558 | originally passed to the callback. |
|
|
559 | |
|
|
560 | Using these functions, it becomes easy to write the "wait_for_child" |
|
|
561 | function mentioned above: |
|
|
562 | |
|
|
563 | sub wait_for_child($) { |
|
|
564 | my ($pid) = @_; |
|
|
565 | |
|
|
566 | my $watcher = AnyEvent->child (pid => $pid, cb => Coro::rouse_cb); |
|
|
567 | |
|
|
568 | my ($rpid, $rstatus) = Coro::rouse_wait; |
|
|
569 | $rstatus |
|
|
570 | } |
|
|
571 | |
|
|
572 | In the case where "rouse_cb" and "rouse_wait" are not flexible enough, |
|
|
573 | you can roll your own, using "schedule": |
|
|
574 | |
|
|
575 | sub wait_for_child($) { |
|
|
576 | my ($pid) = @_; |
|
|
577 | |
|
|
578 | # store the current coro in $current, |
|
|
579 | # and provide result variables for the closure passed to ->child |
|
|
580 | my $current = $Coro::current; |
|
|
581 | my ($done, $rstatus); |
|
|
582 | |
|
|
583 | # pass a closure to ->child |
|
|
584 | my $watcher = AnyEvent->child (pid => $pid, cb => sub { |
|
|
585 | $rstatus = $_[1]; # remember rstatus |
|
|
586 | $done = 1; # mark $rstatus as valud |
|
|
587 | }); |
|
|
588 | |
|
|
589 | # wait until the closure has been called |
|
|
590 | schedule while !$done; |
|
|
591 | |
|
|
592 | $rstatus |
|
|
593 | } |
|
|
594 | |
294 | BUGS/LIMITATIONS |
595 | BUGS/LIMITATIONS |
295 | - you must make very sure that no coro is still active on global |
596 | fork with pthread backend |
296 | destruction. very bad things might happen otherwise (usually segfaults). |
597 | When Coro is compiled using the pthread backend (which isn't |
|
|
598 | recommended but required on many BSDs as their libcs are completely |
|
|
599 | broken), then coro will not survive a fork. There is no known |
|
|
600 | workaround except to fix your libc and use a saner backend. |
297 | |
601 | |
|
|
602 | perl process emulation ("threads") |
298 | - this module is not thread-safe. You should only ever use this module |
603 | This module is not perl-pseudo-thread-safe. You should only ever use |
299 | from the same thread (this requirement might be loosened in the future |
604 | this module from the first thread (this requirement might be removed |
300 | to allow per-thread schedulers, but Coro::State does not yet allow |
605 | in the future to allow per-thread schedulers, but Coro::State does |
301 | this). |
606 | not yet allow this). I recommend disabling thread support and using |
|
|
607 | processes, as having the windows process emulation enabled under |
|
|
608 | unix roughly halves perl performance, even when not used. |
|
|
609 | |
|
|
610 | coro switching is not signal safe |
|
|
611 | You must not switch to another coro from within a signal handler |
|
|
612 | (only relevant with %SIG - most event libraries provide safe |
|
|
613 | signals). |
|
|
614 | |
|
|
615 | That means you *MUST NOT* call any function that might "block" the |
|
|
616 | current coro - "cede", "schedule" "Coro::Semaphore->down" or |
|
|
617 | anything that calls those. Everything else, including calling |
|
|
618 | "ready", works. |
302 | |
619 | |
303 | SEE ALSO |
620 | SEE ALSO |
|
|
621 | Event-Loop integration: Coro::AnyEvent, Coro::EV, Coro::Event. |
|
|
622 | |
|
|
623 | Debugging: Coro::Debug. |
|
|
624 | |
304 | Support/Utility: Coro::Specific, Coro::State, Coro::Util. |
625 | Support/Utility: Coro::Specific, Coro::Util. |
305 | |
626 | |
306 | Locking/IPC: Coro::Signal, Coro::Channel, Coro::Semaphore, |
627 | Locking and IPC: Coro::Signal, Coro::Channel, Coro::Semaphore, |
307 | Coro::SemaphoreSet, Coro::RWLock. |
628 | Coro::SemaphoreSet, Coro::RWLock. |
308 | |
629 | |
309 | Event/IO: Coro::Timer, Coro::Event, Coro::Handle, Coro::Socket, |
630 | I/O and Timers: Coro::Timer, Coro::Handle, Coro::Socket, Coro::AIO. |
|
|
631 | |
|
|
632 | Compatibility with other modules: Coro::LWP (but see also AnyEvent::HTTP |
|
|
633 | for a better-working alternative), Coro::BDB, Coro::Storable, |
310 | Coro::Select. |
634 | Coro::Select. |
311 | |
635 | |
312 | Embedding: <Coro:MakeMaker> |
636 | XS API: Coro::MakeMaker. |
|
|
637 | |
|
|
638 | Low level Configuration, Thread Environment, Continuations: Coro::State. |
313 | |
639 | |
314 | AUTHOR |
640 | AUTHOR |
315 | Marc Lehmann <schmorp@schmorp.de> |
641 | Marc Lehmann <schmorp@schmorp.de> |
316 | http://home.schmorp.de/ |
642 | http://home.schmorp.de/ |
317 | |
643 | |