1 |
root |
1.1 |
NAME |
2 |
root |
1.19 |
Coro - the only real threads in perl |
3 |
root |
1.1 |
|
4 |
|
|
SYNOPSIS |
5 |
root |
1.14 |
use Coro; |
6 |
|
|
|
7 |
|
|
async { |
8 |
|
|
# some asynchronous thread of execution |
9 |
|
|
print "2\n"; |
10 |
|
|
cede; # yield back to main |
11 |
|
|
print "4\n"; |
12 |
|
|
}; |
13 |
|
|
print "1\n"; |
14 |
root |
1.21 |
cede; # yield to coro |
15 |
root |
1.14 |
print "3\n"; |
16 |
|
|
cede; # and again |
17 |
|
|
|
18 |
|
|
# use locking |
19 |
root |
1.15 |
use Coro::Semaphore; |
20 |
root |
1.14 |
my $lock = new Coro::Semaphore; |
21 |
|
|
my $locked; |
22 |
|
|
|
23 |
|
|
$lock->down; |
24 |
|
|
$locked = 1; |
25 |
|
|
$lock->up; |
26 |
root |
1.1 |
|
27 |
|
|
DESCRIPTION |
28 |
root |
1.19 |
For a tutorial-style introduction, please read the Coro::Intro manpage. |
29 |
|
|
This manpage mainly contains reference information. |
30 |
root |
1.1 |
|
31 |
root |
1.19 |
This module collection manages continuations in general, most often in |
32 |
root |
1.21 |
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 |
34 |
root |
1.19 |
general) run in parallel at the same time even on SMP machines. The |
35 |
|
|
specific flavor of thread offered by this module also guarantees you |
36 |
|
|
that it will not switch between threads unless necessary, at |
37 |
|
|
easily-identified points in your program, so locking and parallel access |
38 |
|
|
are rarely an issue, making thread programming much safer and easier |
39 |
|
|
than using other thread models. |
40 |
|
|
|
41 |
|
|
Unlike the so-called "Perl threads" (which are not actually real threads |
42 |
root |
1.21 |
but only the windows process emulation ported to unix, and as such act |
43 |
|
|
as processes), Coro provides a full shared address space, which makes |
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. |
50 |
root |
1.19 |
|
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 |
|
|
|
57 |
|
|
In this module, a thread is defined as "callchain + lexical variables + |
58 |
root |
1.21 |
some package variables + C stack), that is, a thread has its own |
59 |
|
|
callchain, its own set of lexicals and its own set of perls most |
60 |
|
|
important global variables (see Coro::State for more configuration and |
61 |
|
|
background info). |
62 |
root |
1.19 |
|
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 |
67 |
root |
1.14 |
$Coro::main |
68 |
root |
1.21 |
This variable stores the Coro object that represents the main |
69 |
root |
1.14 |
program. While you cna "ready" it and do most other things you can |
70 |
root |
1.21 |
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. |
72 |
root |
1.14 |
|
73 |
|
|
$Coro::current |
74 |
root |
1.21 |
The Coro object representing the current coro (the last coro that |
75 |
|
|
the Coro scheduler switched to). The initial value is $Coro::main |
76 |
|
|
(of course). |
77 |
root |
1.14 |
|
78 |
|
|
This variable is strictly *read-only*. You can take copies of the |
79 |
root |
1.21 |
value stored in it and use it as any other Coro object, but you must |
80 |
|
|
not otherwise modify the variable itself. |
81 |
root |
1.14 |
|
82 |
|
|
$Coro::idle |
83 |
|
|
This variable is mainly useful to integrate Coro into event loops. |
84 |
root |
1.19 |
It is usually better to rely on Coro::AnyEvent or Coro::EV, as this |
85 |
|
|
is pretty low-level functionality. |
86 |
|
|
|
87 |
root |
1.21 |
This variable stores either a Coro object or a callback. |
88 |
root |
1.19 |
|
89 |
|
|
If it is a callback, the it is called whenever the scheduler finds |
90 |
root |
1.21 |
no ready coros to run. The default implementation prints "FATAL: |
91 |
|
|
deadlock detected" and exits, because the program has no other way |
92 |
|
|
to continue. |
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 |
root |
1.14 |
|
98 |
root |
1.19 |
This hook is overwritten by modules such as "Coro::EV" and |
99 |
root |
1.14 |
"Coro::AnyEvent" to wait on an external event that hopefully wake up |
100 |
root |
1.21 |
a coro so the scheduler can run it. |
101 |
root |
1.1 |
|
102 |
root |
1.14 |
Note that the callback *must not*, under any circumstances, block |
103 |
root |
1.21 |
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 |
root |
1.4 |
|
108 |
root |
1.14 |
See Coro::Event or Coro::AnyEvent for examples of using this |
109 |
|
|
technique. |
110 |
root |
1.4 |
|
111 |
|
|
Please note that if your callback recursively invokes perl (e.g. for |
112 |
root |
1.12 |
event handlers), then it must be prepared to be called recursively |
113 |
|
|
itself. |
114 |
root |
1.1 |
|
115 |
root |
1.21 |
SIMPLE CORO CREATION |
116 |
root |
1.14 |
async { ... } [@args...] |
117 |
root |
1.21 |
Create a new coro and return its Coro object (usually unused). The |
118 |
|
|
coro will be put into the ready queue, so it will start running |
119 |
|
|
automatically on the next scheduler run. |
120 |
root |
1.1 |
|
121 |
root |
1.14 |
The first argument is a codeblock/closure that should be executed in |
122 |
root |
1.21 |
the coro. When it returns argument returns the coro is automatically |
123 |
|
|
terminated. |
124 |
root |
1.1 |
|
125 |
root |
1.14 |
The remaining arguments are passed as arguments to the closure. |
126 |
|
|
|
127 |
root |
1.21 |
See the "Coro::State::new" constructor for info about the coro |
128 |
|
|
environment in which coro are executed. |
129 |
root |
1.10 |
|
130 |
root |
1.21 |
Calling "exit" in a coro will do the same as calling exit outside |
131 |
|
|
the coro. Likewise, when the coro dies, the program will exit, just |
132 |
|
|
as it would in the main program. |
133 |
root |
1.3 |
|
134 |
root |
1.14 |
If you do not want that, you can provide a default "die" handler, or |
135 |
|
|
simply avoid dieing (by use of "eval"). |
136 |
|
|
|
137 |
root |
1.21 |
Example: Create a new coro that just prints its arguments. |
138 |
root |
1.14 |
|
139 |
root |
1.1 |
async { |
140 |
|
|
print "@_\n"; |
141 |
|
|
} 1,2,3,4; |
142 |
|
|
|
143 |
root |
1.6 |
async_pool { ... } [@args...] |
144 |
root |
1.21 |
Similar to "async", but uses a coro pool, so you should not call |
145 |
|
|
terminate or join on it (although you are allowed to), and you get a |
146 |
|
|
coro that might have executed other code already (which can be good |
147 |
|
|
or bad :). |
148 |
root |
1.14 |
|
149 |
root |
1.18 |
On the plus side, this function is about twice as fast as creating |
150 |
root |
1.21 |
(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 |
root |
1.6 |
|
153 |
root |
1.14 |
The code block is executed in an "eval" context and a warning will |
154 |
root |
1.6 |
be issued in case of an exception instead of terminating the |
155 |
root |
1.21 |
program, as "async" does. As the coro is being reused, stuff like |
156 |
|
|
"on_destroy" will not work in the expected way, unless you call |
157 |
root |
1.14 |
terminate or cancel, which somehow defeats the purpose of pooling |
158 |
|
|
(but is fine in the exceptional case). |
159 |
root |
1.6 |
|
160 |
root |
1.14 |
The priority will be reset to 0 after each run, tracing will be |
161 |
root |
1.10 |
disabled, the description will be reset and the default output |
162 |
root |
1.14 |
filehandle gets restored, so you can change all these. Otherwise the |
163 |
root |
1.21 |
coro will be re-used "as-is": most notably if you change other |
164 |
|
|
per-coro global stuff such as $/ you *must needs* revert that |
165 |
root |
1.16 |
change, which is most simply done by using local as in: "local $/". |
166 |
|
|
|
167 |
root |
1.21 |
The idle pool size is limited to 8 idle coros (this can be adjusted |
168 |
|
|
by changing $Coro::POOL_SIZE), but there can be as many non-idle |
169 |
|
|
coros as required. |
170 |
root |
1.6 |
|
171 |
root |
1.21 |
If you are concerned about pooled coros growing a lot because a |
172 |
root |
1.6 |
single "async_pool" used a lot of stackspace you can e.g. |
173 |
|
|
"async_pool { terminate }" once per second or so to slowly replenish |
174 |
root |
1.9 |
the pool. In addition to that, when the stacks used by a handler |
175 |
root |
1.19 |
grows larger than 32kb (adjustable via $Coro::POOL_RSS) it will also |
176 |
root |
1.14 |
be destroyed. |
177 |
|
|
|
178 |
root |
1.19 |
STATIC METHODS |
179 |
|
|
Static methods are actually functions that implicitly operate on the |
180 |
root |
1.21 |
current coro. |
181 |
root |
1.6 |
|
182 |
root |
1.1 |
schedule |
183 |
root |
1.21 |
Calls the scheduler. The scheduler will find the next coro that is |
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 |
191 |
|
|
called again unless something else (e.g. an event handler) calls |
192 |
root |
1.14 |
"->ready", thus waking you up. |
193 |
|
|
|
194 |
|
|
This makes "schedule" *the* generic method to use to block the |
195 |
root |
1.21 |
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. |
201 |
root |
1.4 |
|
202 |
root |
1.18 |
See HOW TO WAIT FOR A CALLBACK, below, for some ways to wait for |
203 |
|
|
callbacks. |
204 |
root |
1.1 |
|
205 |
|
|
cede |
206 |
root |
1.21 |
"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 |
root |
1.14 |
|
212 |
|
|
This function is often called "yield" in other languages. |
213 |
root |
1.1 |
|
214 |
root |
1.6 |
Coro::cede_notself |
215 |
root |
1.14 |
Works like cede, but is not exported by default and will cede to |
216 |
root |
1.21 |
*any* coro, regardless of priority. This is useful sometimes to |
217 |
root |
1.14 |
ensure progress is made. |
218 |
root |
1.6 |
|
219 |
root |
1.1 |
terminate [arg...] |
220 |
root |
1.21 |
Terminates the current coro with the given status values (see |
221 |
root |
1.1 |
cancel). |
222 |
|
|
|
223 |
root |
1.21 |
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 coro switching considerably for coros that use them |
244 |
|
|
(But coro switching is still reasonably fast if the handlers are |
245 |
|
|
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 changes for the coro that installed those handlers. |
253 |
|
|
|
254 |
|
|
use POSIX qw(tzset); |
255 |
|
|
|
256 |
|
|
async { |
257 |
|
|
my $old_tz; # store outside TZ value here |
258 |
|
|
|
259 |
|
|
Coro::on_enter { |
260 |
|
|
$old_tz = $ENV{TZ}; # remember the old value |
261 |
|
|
|
262 |
|
|
$ENV{TZ} = "Antarctica/South_Pole"; |
263 |
|
|
tzset; # enable new value |
264 |
|
|
}; |
265 |
|
|
|
266 |
|
|
Coro::on_leave { |
267 |
|
|
$ENV{TZ} = $old_tz; |
268 |
|
|
tzset; # restore old value |
269 |
|
|
}; |
270 |
|
|
|
271 |
|
|
# at this place, the timezone is Antarctica/South_Pole, |
272 |
|
|
# without disturbing the TZ of any other coro. |
273 |
|
|
}; |
274 |
|
|
|
275 |
|
|
This can be used to localise about any resource (locale, uid, |
276 |
|
|
current working directory etc.) to a block, despite the existance of |
277 |
|
|
other coros. |
278 |
|
|
|
279 |
root |
1.10 |
killall |
280 |
root |
1.21 |
Kills/terminates/cancels all coros except the currently running one. |
281 |
|
|
|
282 |
|
|
Note that while this will try to free some of the main interpreter |
283 |
|
|
resources if the calling coro isn't the main coro, but one cannot |
284 |
|
|
free all of them, so if a coro that is not the main coro calls this |
285 |
|
|
function, there will be some one-time resource leak. |
286 |
|
|
|
287 |
|
|
CORO OBJECT METHODS |
288 |
|
|
These are the methods you can call on coro objects (or to create them). |
289 |
root |
1.1 |
|
290 |
|
|
new Coro \&sub [, @args...] |
291 |
root |
1.21 |
Create a new coro and return it. When the sub returns, the coro |
292 |
|
|
automatically terminates as if "terminate" with the returned values |
293 |
|
|
were called. To make the coro run you must first put it into the |
294 |
|
|
ready queue by calling the ready method. |
295 |
root |
1.4 |
|
296 |
root |
1.10 |
See "async" and "Coro::State::new" for additional info about the |
297 |
root |
1.21 |
coro environment. |
298 |
root |
1.4 |
|
299 |
root |
1.21 |
$success = $coro->ready |
300 |
|
|
Put the given coro into the end of its ready queue (there is one |
301 |
|
|
queue for each priority) and return true. If the coro is already in |
302 |
|
|
the ready queue, do nothing and return false. |
303 |
|
|
|
304 |
|
|
This ensures that the scheduler will resume this coro automatically |
305 |
|
|
once all the coro of higher priority and all coro of the same |
306 |
|
|
priority that were put into the ready queue earlier have been |
307 |
|
|
resumed. |
308 |
|
|
|
309 |
root |
1.22 |
$coro->suspend |
310 |
|
|
Suspends the specified coro. A suspended coro works just like any |
311 |
|
|
other coro, except that the scheduler will not select a suspended |
312 |
|
|
coro for execution. |
313 |
|
|
|
314 |
|
|
Suspending a coro can be useful when you want to keep the coro from |
315 |
|
|
running, but you don't want to destroy it, or when you want to |
316 |
|
|
temporarily freeze a coro (e.g. for debugging) to resume it later. |
317 |
|
|
|
318 |
|
|
A scenario for the former would be to suspend all (other) coros |
319 |
|
|
after a fork and keep them alive, so their destructors aren't |
320 |
|
|
called, but new coros can be created. |
321 |
|
|
|
322 |
|
|
$coro->resume |
323 |
|
|
If the specified coro was suspended, it will be resumed. Note that |
324 |
|
|
when the coro was in the ready queue when it was suspended, it might |
325 |
|
|
have been unreadied by the scheduler, so an activation might have |
326 |
|
|
been lost. |
327 |
|
|
|
328 |
|
|
To avoid this, it is best to put a suspended coro into the ready |
329 |
|
|
queue unconditionally, as every synchronisation mechanism must |
330 |
|
|
protect itself against spurious wakeups, and the one in the Coro |
331 |
|
|
family certainly do that. |
332 |
|
|
|
333 |
root |
1.21 |
$is_ready = $coro->is_ready |
334 |
|
|
Returns true iff the Coro object is in the ready queue. Unless the |
335 |
|
|
Coro object gets destroyed, it will eventually be scheduled by the |
336 |
|
|
scheduler. |
337 |
|
|
|
338 |
|
|
$is_running = $coro->is_running |
339 |
|
|
Returns true iff the Coro object is currently running. Only one Coro |
340 |
|
|
object can ever be in the running state (but it currently is |
341 |
|
|
possible to have multiple running Coro::States). |
342 |
|
|
|
343 |
|
|
$is_suspended = $coro->is_suspended |
344 |
|
|
Returns true iff this Coro object has been suspended. Suspended |
345 |
|
|
Coros will not ever be scheduled. |
346 |
|
|
|
347 |
|
|
$coro->cancel (arg...) |
348 |
|
|
Terminates the given Coro and makes it return the given arguments as |
349 |
|
|
status (default: the empty list). Never returns if the Coro is the |
350 |
|
|
current Coro. |
351 |
|
|
|
352 |
|
|
$coro->schedule_to |
353 |
|
|
Puts the current coro to sleep (like "Coro::schedule"), but instead |
354 |
|
|
of continuing with the next coro from the ready queue, always switch |
355 |
|
|
to the given coro object (regardless of priority etc.). The |
356 |
|
|
readyness state of that coro isn't changed. |
357 |
root |
1.18 |
|
358 |
|
|
This is an advanced method for special cases - I'd love to hear |
359 |
|
|
about any uses for this one. |
360 |
|
|
|
361 |
root |
1.21 |
$coro->cede_to |
362 |
|
|
Like "schedule_to", but puts the current coro into the ready queue. |
363 |
|
|
This has the effect of temporarily switching to the given coro, and |
364 |
|
|
continuing some time later. |
365 |
root |
1.18 |
|
366 |
|
|
This is an advanced method for special cases - I'd love to hear |
367 |
|
|
about any uses for this one. |
368 |
|
|
|
369 |
root |
1.21 |
$coro->throw ([$scalar]) |
370 |
root |
1.17 |
If $throw is specified and defined, it will be thrown as an |
371 |
root |
1.21 |
exception inside the coro at the next convenient point in time. |
372 |
root |
1.17 |
Otherwise clears the exception object. |
373 |
|
|
|
374 |
root |
1.18 |
Coro will check for the exception each time a schedule-like-function |
375 |
|
|
returns, i.e. after each "schedule", "cede", |
376 |
|
|
"Coro::Semaphore->down", "Coro::Handle->readable" and so on. Most of |
377 |
|
|
these functions detect this case and return early in case an |
378 |
|
|
exception is pending. |
379 |
|
|
|
380 |
root |
1.17 |
The exception object will be thrown "as is" with the specified |
381 |
|
|
scalar in $@, i.e. if it is a string, no line number or newline will |
382 |
|
|
be appended (unlike with "die"). |
383 |
|
|
|
384 |
root |
1.21 |
This can be used as a softer means than "cancel" to ask a coro to |
385 |
|
|
end itself, although there is no guarantee that the exception will |
386 |
|
|
lead to termination, and if the exception isn't caught it might well |
387 |
|
|
end the whole program. |
388 |
root |
1.17 |
|
389 |
|
|
You might also think of "throw" as being the moral equivalent of |
390 |
root |
1.21 |
"kill"ing a coro with a signal (in this case, a scalar). |
391 |
root |
1.17 |
|
392 |
root |
1.21 |
$coro->join |
393 |
|
|
Wait until the coro terminates and return any values given to the |
394 |
|
|
"terminate" or "cancel" functions. "join" can be called concurrently |
395 |
|
|
from multiple coro, and all will be resumed and given the status |
396 |
|
|
return once the $coro terminates. |
397 |
|
|
|
398 |
|
|
$coro->on_destroy (\&cb) |
399 |
|
|
Registers a callback that is called when this coro gets destroyed, |
400 |
|
|
but before it is joined. The callback gets passed the terminate |
401 |
|
|
arguments, if any, and *must not* die, under any circumstances. |
402 |
|
|
|
403 |
|
|
$oldprio = $coro->prio ($newprio) |
404 |
|
|
Sets (or gets, if the argument is missing) the priority of the coro. |
405 |
|
|
Higher priority coro get run before lower priority coro. Priorities |
406 |
|
|
are small signed integers (currently -4 .. +3), that you can refer |
407 |
|
|
to using PRIO_xxx constants (use the import tag :prio to get then): |
408 |
root |
1.1 |
|
409 |
|
|
PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN |
410 |
|
|
3 > 1 > 0 > -1 > -3 > -4 |
411 |
|
|
|
412 |
|
|
# set priority to HIGH |
413 |
root |
1.21 |
current->prio (PRIO_HIGH); |
414 |
root |
1.1 |
|
415 |
root |
1.21 |
The idle coro ($Coro::idle) always has a lower priority than any |
416 |
|
|
existing coro. |
417 |
root |
1.1 |
|
418 |
root |
1.21 |
Changing the priority of the current coro will take effect |
419 |
|
|
immediately, but changing the priority of coro in the ready queue |
420 |
|
|
(but not running) will only take effect after the next schedule (of |
421 |
|
|
that coro). This is a bug that will be fixed in some future version. |
422 |
root |
1.1 |
|
423 |
root |
1.21 |
$newprio = $coro->nice ($change) |
424 |
root |
1.1 |
Similar to "prio", but subtract the given value from the priority |
425 |
|
|
(i.e. higher values mean lower priority, just as in unix). |
426 |
|
|
|
427 |
root |
1.21 |
$olddesc = $coro->desc ($newdesc) |
428 |
root |
1.1 |
Sets (or gets in case the argument is missing) the description for |
429 |
root |
1.21 |
this coro. This is just a free-form string you can associate with a |
430 |
|
|
coro. |
431 |
root |
1.4 |
|
432 |
root |
1.21 |
This method simply sets the "$coro->{desc}" member to the given |
433 |
root |
1.10 |
string. You can modify this member directly if you wish. |
434 |
|
|
|
435 |
root |
1.19 |
GLOBAL FUNCTIONS |
436 |
root |
1.5 |
Coro::nready |
437 |
root |
1.21 |
Returns the number of coro that are currently in the ready state, |
438 |
|
|
i.e. that can be switched to by calling "schedule" directory or |
439 |
|
|
indirectly. The value 0 means that the only runnable coro is the |
440 |
|
|
currently running one, so "cede" would have no effect, and |
441 |
root |
1.14 |
"schedule" would cause a deadlock unless there is an idle handler |
442 |
root |
1.21 |
that wakes up some coro. |
443 |
root |
1.5 |
|
444 |
root |
1.6 |
my $guard = Coro::guard { ... } |
445 |
root |
1.21 |
This function still exists, but is deprecated. Please use the |
446 |
|
|
"Guard::guard" function instead. |
447 |
root |
1.6 |
|
448 |
root |
1.4 |
unblock_sub { ... } |
449 |
|
|
This utility function takes a BLOCK or code reference and "unblocks" |
450 |
root |
1.14 |
it, returning a new coderef. Unblocking means that calling the new |
451 |
|
|
coderef will return immediately without blocking, returning nothing, |
452 |
|
|
while the original code ref will be called (with parameters) from |
453 |
root |
1.21 |
within another coro. |
454 |
root |
1.4 |
|
455 |
root |
1.8 |
The reason this function exists is that many event libraries (such |
456 |
root |
1.21 |
as the venerable Event module) are not thread-safe (a weaker form of |
457 |
|
|
reentrancy). This means you must not block within event callbacks, |
458 |
|
|
otherwise you might suffer from crashes or worse. The only event |
459 |
|
|
library currently known that is safe to use without "unblock_sub" is |
460 |
|
|
EV. |
461 |
root |
1.4 |
|
462 |
|
|
This function allows your callbacks to block by executing them in |
463 |
root |
1.21 |
another coro where it is safe to block. One example where blocking |
464 |
|
|
is handy is when you use the Coro::AIO functions to save results to |
465 |
|
|
disk, for example. |
466 |
root |
1.4 |
|
467 |
|
|
In short: simply use "unblock_sub { ... }" instead of "sub { ... }" |
468 |
|
|
when creating event callbacks that want to block. |
469 |
root |
1.1 |
|
470 |
root |
1.14 |
If your handler does not plan to block (e.g. simply sends a message |
471 |
root |
1.21 |
to another coro, or puts some other coro into the ready queue), |
472 |
|
|
there is no reason to use "unblock_sub". |
473 |
root |
1.14 |
|
474 |
|
|
Note that you also need to use "unblock_sub" for any other callbacks |
475 |
|
|
that are indirectly executed by any C-based event loop. For example, |
476 |
|
|
when you use a module that uses AnyEvent (and you use |
477 |
|
|
Coro::AnyEvent) and it provides callbacks that are the result of |
478 |
|
|
some event callback, then you must not block either, or use |
479 |
|
|
"unblock_sub". |
480 |
|
|
|
481 |
root |
1.18 |
$cb = Coro::rouse_cb |
482 |
|
|
Create and return a "rouse callback". That's a code reference that, |
483 |
root |
1.19 |
when called, will remember a copy of its arguments and notify the |
484 |
root |
1.21 |
owner coro of the callback. |
485 |
root |
1.18 |
|
486 |
|
|
See the next function. |
487 |
|
|
|
488 |
|
|
@args = Coro::rouse_wait [$cb] |
489 |
root |
1.19 |
Wait for the specified rouse callback (or the last one that was |
490 |
root |
1.21 |
created in this coro). |
491 |
root |
1.18 |
|
492 |
root |
1.19 |
As soon as the callback is invoked (or when the callback was invoked |
493 |
|
|
before "rouse_wait"), it will return the arguments originally passed |
494 |
|
|
to the rouse callback. |
495 |
root |
1.18 |
|
496 |
|
|
See the section HOW TO WAIT FOR A CALLBACK for an actual usage |
497 |
|
|
example. |
498 |
|
|
|
499 |
|
|
HOW TO WAIT FOR A CALLBACK |
500 |
root |
1.21 |
It is very common for a coro to wait for some callback to be called. |
501 |
|
|
This occurs naturally when you use coro in an otherwise event-based |
502 |
|
|
program, or when you use event-based libraries. |
503 |
root |
1.18 |
|
504 |
|
|
These typically register a callback for some event, and call that |
505 |
root |
1.21 |
callback when the event occured. In a coro, however, you typically want |
506 |
|
|
to just wait for the event, simplyifying things. |
507 |
root |
1.18 |
|
508 |
|
|
For example "AnyEvent->child" registers a callback to be called when a |
509 |
|
|
specific child has exited: |
510 |
|
|
|
511 |
|
|
my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... }); |
512 |
|
|
|
513 |
root |
1.21 |
But from within a coro, you often just want to write this: |
514 |
root |
1.18 |
|
515 |
|
|
my $status = wait_for_child $pid; |
516 |
|
|
|
517 |
|
|
Coro offers two functions specifically designed to make this easy, |
518 |
|
|
"Coro::rouse_cb" and "Coro::rouse_wait". |
519 |
|
|
|
520 |
|
|
The first function, "rouse_cb", generates and returns a callback that, |
521 |
root |
1.21 |
when invoked, will save its arguments and notify the coro that created |
522 |
|
|
the callback. |
523 |
root |
1.18 |
|
524 |
|
|
The second function, "rouse_wait", waits for the callback to be called |
525 |
|
|
(by calling "schedule" to go to sleep) and returns the arguments |
526 |
|
|
originally passed to the callback. |
527 |
|
|
|
528 |
|
|
Using these functions, it becomes easy to write the "wait_for_child" |
529 |
|
|
function mentioned above: |
530 |
|
|
|
531 |
|
|
sub wait_for_child($) { |
532 |
|
|
my ($pid) = @_; |
533 |
|
|
|
534 |
|
|
my $watcher = AnyEvent->child (pid => $pid, cb => Coro::rouse_cb); |
535 |
|
|
|
536 |
|
|
my ($rpid, $rstatus) = Coro::rouse_wait; |
537 |
|
|
$rstatus |
538 |
|
|
} |
539 |
|
|
|
540 |
|
|
In the case where "rouse_cb" and "rouse_wait" are not flexible enough, |
541 |
|
|
you can roll your own, using "schedule": |
542 |
|
|
|
543 |
|
|
sub wait_for_child($) { |
544 |
|
|
my ($pid) = @_; |
545 |
|
|
|
546 |
root |
1.21 |
# store the current coro in $current, |
547 |
root |
1.18 |
# and provide result variables for the closure passed to ->child |
548 |
|
|
my $current = $Coro::current; |
549 |
|
|
my ($done, $rstatus); |
550 |
|
|
|
551 |
|
|
# pass a closure to ->child |
552 |
|
|
my $watcher = AnyEvent->child (pid => $pid, cb => sub { |
553 |
|
|
$rstatus = $_[1]; # remember rstatus |
554 |
|
|
$done = 1; # mark $rstatus as valud |
555 |
|
|
}); |
556 |
|
|
|
557 |
|
|
# wait until the closure has been called |
558 |
|
|
schedule while !$done; |
559 |
|
|
|
560 |
|
|
$rstatus |
561 |
|
|
} |
562 |
|
|
|
563 |
root |
1.1 |
BUGS/LIMITATIONS |
564 |
root |
1.18 |
fork with pthread backend |
565 |
|
|
When Coro is compiled using the pthread backend (which isn't |
566 |
|
|
recommended but required on many BSDs as their libcs are completely |
567 |
root |
1.21 |
broken), then coro will not survive a fork. There is no known |
568 |
root |
1.18 |
workaround except to fix your libc and use a saner backend. |
569 |
|
|
|
570 |
|
|
perl process emulation ("threads") |
571 |
|
|
This module is not perl-pseudo-thread-safe. You should only ever use |
572 |
root |
1.19 |
this module from the first thread (this requirement might be removed |
573 |
root |
1.18 |
in the future to allow per-thread schedulers, but Coro::State does |
574 |
|
|
not yet allow this). I recommend disabling thread support and using |
575 |
|
|
processes, as having the windows process emulation enabled under |
576 |
|
|
unix roughly halves perl performance, even when not used. |
577 |
|
|
|
578 |
root |
1.21 |
coro switching is not signal safe |
579 |
|
|
You must not switch to another coro from within a signal handler |
580 |
|
|
(only relevant with %SIG - most event libraries provide safe |
581 |
root |
1.18 |
signals). |
582 |
|
|
|
583 |
|
|
That means you *MUST NOT* call any function that might "block" the |
584 |
root |
1.21 |
current coro - "cede", "schedule" "Coro::Semaphore->down" or |
585 |
root |
1.18 |
anything that calls those. Everything else, including calling |
586 |
|
|
"ready", works. |
587 |
root |
1.1 |
|
588 |
|
|
SEE ALSO |
589 |
root |
1.14 |
Event-Loop integration: Coro::AnyEvent, Coro::EV, Coro::Event. |
590 |
root |
1.12 |
|
591 |
|
|
Debugging: Coro::Debug. |
592 |
|
|
|
593 |
|
|
Support/Utility: Coro::Specific, Coro::Util. |
594 |
root |
1.2 |
|
595 |
root |
1.19 |
Locking and IPC: Coro::Signal, Coro::Channel, Coro::Semaphore, |
596 |
root |
1.2 |
Coro::SemaphoreSet, Coro::RWLock. |
597 |
|
|
|
598 |
root |
1.19 |
I/O and Timers: Coro::Timer, Coro::Handle, Coro::Socket, Coro::AIO. |
599 |
root |
1.14 |
|
600 |
root |
1.19 |
Compatibility with other modules: Coro::LWP (but see also AnyEvent::HTTP |
601 |
|
|
for a better-working alternative), Coro::BDB, Coro::Storable, |
602 |
|
|
Coro::Select. |
603 |
root |
1.12 |
|
604 |
root |
1.14 |
XS API: Coro::MakeMaker. |
605 |
root |
1.2 |
|
606 |
root |
1.19 |
Low level Configuration, Thread Environment, Continuations: Coro::State. |
607 |
root |
1.1 |
|
608 |
|
|
AUTHOR |
609 |
|
|
Marc Lehmann <schmorp@schmorp.de> |
610 |
|
|
http://home.schmorp.de/ |
611 |
|
|
|