1 |
=head1 NAME |
2 |
|
3 |
Coro - the only real threads in perl |
4 |
|
5 |
=head1 SYNOPSIS |
6 |
|
7 |
use Coro; |
8 |
|
9 |
async { |
10 |
# some asynchronous thread of execution |
11 |
print "2\n"; |
12 |
cede; # yield back to main |
13 |
print "4\n"; |
14 |
}; |
15 |
print "1\n"; |
16 |
cede; # yield to coro |
17 |
print "3\n"; |
18 |
cede; # and again |
19 |
|
20 |
# use locking |
21 |
use Coro::Semaphore; |
22 |
my $lock = new Coro::Semaphore; |
23 |
my $locked; |
24 |
|
25 |
$lock->down; |
26 |
$locked = 1; |
27 |
$lock->up; |
28 |
|
29 |
=head1 DESCRIPTION |
30 |
|
31 |
For a tutorial-style introduction, please read the L<Coro::Intro> |
32 |
manpage. This manpage mainly contains reference information. |
33 |
|
34 |
This module collection manages continuations in general, most often in |
35 |
the form of cooperative threads (also called coros, or simply "coro" |
36 |
in the documentation). They are similar to kernel threads but don't (in |
37 |
general) run in parallel at the same time even on SMP machines. The |
38 |
specific flavor of thread offered by this module also guarantees you that |
39 |
it will not switch between threads unless necessary, at easily-identified |
40 |
points in your program, so locking and parallel access are rarely an |
41 |
issue, making thread programming much safer and easier than using other |
42 |
thread models. |
43 |
|
44 |
Unlike the so-called "Perl threads" (which are not actually real threads |
45 |
but only the windows process emulation ported to unix), Coro provides a |
46 |
full shared address space, which makes communication between threads |
47 |
very easy. And threads are fast, too: disabling the Windows process |
48 |
emulation code in your perl and using Coro can easily result in a two to |
49 |
four times speed increase for your programs. |
50 |
|
51 |
Coro achieves that by supporting multiple running interpreters that share |
52 |
data, which is especially useful to code pseudo-parallel processes and |
53 |
for event-based programming, such as multiple HTTP-GET requests running |
54 |
concurrently. See L<Coro::AnyEvent> to learn more on how to integrate Coro |
55 |
into an event-based environment. |
56 |
|
57 |
In this module, a thread is defined as "callchain + lexical variables + |
58 |
@_ + $_ + $@ + $/ + C stack), that is, a thread has its own callchain, |
59 |
its own set of lexicals and its own set of perls most important global |
60 |
variables (see L<Coro::State> for more configuration and background info). |
61 |
|
62 |
See also the C<SEE ALSO> section at the end of this document - the Coro |
63 |
module family is quite large. |
64 |
|
65 |
=cut |
66 |
|
67 |
package Coro; |
68 |
|
69 |
use strict qw(vars subs); |
70 |
no warnings "uninitialized"; |
71 |
|
72 |
use Guard (); |
73 |
|
74 |
use Coro::State; |
75 |
|
76 |
use base qw(Coro::State Exporter); |
77 |
|
78 |
our $idle; # idle handler |
79 |
our $main; # main coro |
80 |
our $current; # current coro |
81 |
|
82 |
our $VERSION = 5.13; |
83 |
|
84 |
our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub); |
85 |
our %EXPORT_TAGS = ( |
86 |
prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)], |
87 |
); |
88 |
our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready)); |
89 |
|
90 |
=head1 GLOBAL VARIABLES |
91 |
|
92 |
=over 4 |
93 |
|
94 |
=item $Coro::main |
95 |
|
96 |
This variable stores the Coro object that represents the main |
97 |
program. While you cna C<ready> it and do most other things you can do to |
98 |
coro, it is mainly useful to compare again C<$Coro::current>, to see |
99 |
whether you are running in the main program or not. |
100 |
|
101 |
=cut |
102 |
|
103 |
# $main is now being initialised by Coro::State |
104 |
|
105 |
=item $Coro::current |
106 |
|
107 |
The Coro object representing the current coro (the last |
108 |
coro that the Coro scheduler switched to). The initial value is |
109 |
C<$Coro::main> (of course). |
110 |
|
111 |
This variable is B<strictly> I<read-only>. You can take copies of the |
112 |
value stored in it and use it as any other Coro object, but you must |
113 |
not otherwise modify the variable itself. |
114 |
|
115 |
=cut |
116 |
|
117 |
sub current() { $current } # [DEPRECATED] |
118 |
|
119 |
=item $Coro::idle |
120 |
|
121 |
This variable is mainly useful to integrate Coro into event loops. It is |
122 |
usually better to rely on L<Coro::AnyEvent> or L<Coro::EV>, as this is |
123 |
pretty low-level functionality. |
124 |
|
125 |
This variable stores either a Coro object or a callback. |
126 |
|
127 |
If it is a callback, the it is called whenever the scheduler finds no |
128 |
ready coros to run. The default implementation prints "FATAL: |
129 |
deadlock detected" and exits, because the program has no other way to |
130 |
continue. |
131 |
|
132 |
If it is a coro object, then this object will be readied (without |
133 |
invoking any ready hooks, however) when the scheduler finds no other ready |
134 |
coros to run. |
135 |
|
136 |
This hook is overwritten by modules such as C<Coro::EV> and |
137 |
C<Coro::AnyEvent> to wait on an external event that hopefully wake up a |
138 |
coro so the scheduler can run it. |
139 |
|
140 |
Note that the callback I<must not>, under any circumstances, block |
141 |
the current coro. Normally, this is achieved by having an "idle |
142 |
coro" that calls the event loop and then blocks again, and then |
143 |
readying that coro in the idle handler, or by simply placing the idle |
144 |
coro in this variable. |
145 |
|
146 |
See L<Coro::Event> or L<Coro::AnyEvent> for examples of using this |
147 |
technique. |
148 |
|
149 |
Please note that if your callback recursively invokes perl (e.g. for event |
150 |
handlers), then it must be prepared to be called recursively itself. |
151 |
|
152 |
=cut |
153 |
|
154 |
$idle = sub { |
155 |
require Carp; |
156 |
Carp::croak ("FATAL: deadlock detected"); |
157 |
}; |
158 |
|
159 |
# this coro is necessary because a coro |
160 |
# cannot destroy itself. |
161 |
our @destroy; |
162 |
our $manager; |
163 |
|
164 |
$manager = new Coro sub { |
165 |
while () { |
166 |
Coro::State::cancel shift @destroy |
167 |
while @destroy; |
168 |
|
169 |
&schedule; |
170 |
} |
171 |
}; |
172 |
$manager->{desc} = "[coro manager]"; |
173 |
$manager->prio (PRIO_MAX); |
174 |
|
175 |
=back |
176 |
|
177 |
=head1 SIMPLE CORO CREATION |
178 |
|
179 |
=over 4 |
180 |
|
181 |
=item async { ... } [@args...] |
182 |
|
183 |
Create a new coro and return its Coro object (usually |
184 |
unused). The coro will be put into the ready queue, so |
185 |
it will start running automatically on the next scheduler run. |
186 |
|
187 |
The first argument is a codeblock/closure that should be executed in the |
188 |
coro. When it returns argument returns the coro is automatically |
189 |
terminated. |
190 |
|
191 |
The remaining arguments are passed as arguments to the closure. |
192 |
|
193 |
See the C<Coro::State::new> constructor for info about the coro |
194 |
environment in which coro are executed. |
195 |
|
196 |
Calling C<exit> in a coro will do the same as calling exit outside |
197 |
the coro. Likewise, when the coro dies, the program will exit, |
198 |
just as it would in the main program. |
199 |
|
200 |
If you do not want that, you can provide a default C<die> handler, or |
201 |
simply avoid dieing (by use of C<eval>). |
202 |
|
203 |
Example: Create a new coro that just prints its arguments. |
204 |
|
205 |
async { |
206 |
print "@_\n"; |
207 |
} 1,2,3,4; |
208 |
|
209 |
=cut |
210 |
|
211 |
sub async(&@) { |
212 |
my $coro = new Coro @_; |
213 |
$coro->ready; |
214 |
$coro |
215 |
} |
216 |
|
217 |
=item async_pool { ... } [@args...] |
218 |
|
219 |
Similar to C<async>, but uses a coro pool, so you should not call |
220 |
terminate or join on it (although you are allowed to), and you get a |
221 |
coro that might have executed other code already (which can be good |
222 |
or bad :). |
223 |
|
224 |
On the plus side, this function is about twice as fast as creating (and |
225 |
destroying) a completely new coro, so if you need a lot of generic |
226 |
coros in quick successsion, use C<async_pool>, not C<async>. |
227 |
|
228 |
The code block is executed in an C<eval> context and a warning will be |
229 |
issued in case of an exception instead of terminating the program, as |
230 |
C<async> does. As the coro is being reused, stuff like C<on_destroy> |
231 |
will not work in the expected way, unless you call terminate or cancel, |
232 |
which somehow defeats the purpose of pooling (but is fine in the |
233 |
exceptional case). |
234 |
|
235 |
The priority will be reset to C<0> after each run, tracing will be |
236 |
disabled, the description will be reset and the default output filehandle |
237 |
gets restored, so you can change all these. Otherwise the coro will |
238 |
be re-used "as-is": most notably if you change other per-coro global |
239 |
stuff such as C<$/> you I<must needs> revert that change, which is most |
240 |
simply done by using local as in: C<< local $/ >>. |
241 |
|
242 |
The idle pool size is limited to C<8> idle coros (this can be |
243 |
adjusted by changing $Coro::POOL_SIZE), but there can be as many non-idle |
244 |
coros as required. |
245 |
|
246 |
If you are concerned about pooled coros growing a lot because a |
247 |
single C<async_pool> used a lot of stackspace you can e.g. C<async_pool |
248 |
{ terminate }> once per second or so to slowly replenish the pool. In |
249 |
addition to that, when the stacks used by a handler grows larger than 32kb |
250 |
(adjustable via $Coro::POOL_RSS) it will also be destroyed. |
251 |
|
252 |
=cut |
253 |
|
254 |
our $POOL_SIZE = 8; |
255 |
our $POOL_RSS = 32 * 1024; |
256 |
our @async_pool; |
257 |
|
258 |
sub pool_handler { |
259 |
while () { |
260 |
eval { |
261 |
&{&_pool_handler} while 1; |
262 |
}; |
263 |
|
264 |
warn $@ if $@; |
265 |
} |
266 |
} |
267 |
|
268 |
=back |
269 |
|
270 |
=head1 STATIC METHODS |
271 |
|
272 |
Static methods are actually functions that implicitly operate on the |
273 |
current coro. |
274 |
|
275 |
=over 4 |
276 |
|
277 |
=item schedule |
278 |
|
279 |
Calls the scheduler. The scheduler will find the next coro that is |
280 |
to be run from the ready queue and switches to it. The next coro |
281 |
to be run is simply the one with the highest priority that is longest |
282 |
in its ready queue. If there is no coro ready, it will clal the |
283 |
C<$Coro::idle> hook. |
284 |
|
285 |
Please note that the current coro will I<not> be put into the ready |
286 |
queue, so calling this function usually means you will never be called |
287 |
again unless something else (e.g. an event handler) calls C<< ->ready >>, |
288 |
thus waking you up. |
289 |
|
290 |
This makes C<schedule> I<the> generic method to use to block the current |
291 |
coro and wait for events: first you remember the current coro in |
292 |
a variable, then arrange for some callback of yours to call C<< ->ready |
293 |
>> on that once some event happens, and last you call C<schedule> to put |
294 |
yourself to sleep. Note that a lot of things can wake your coro up, |
295 |
so you need to check whether the event indeed happened, e.g. by storing the |
296 |
status in a variable. |
297 |
|
298 |
See B<HOW TO WAIT FOR A CALLBACK>, below, for some ways to wait for callbacks. |
299 |
|
300 |
=item cede |
301 |
|
302 |
"Cede" to other coros. This function puts the current coro into |
303 |
the ready queue and calls C<schedule>, which has the effect of giving |
304 |
up the current "timeslice" to other coros of the same or higher |
305 |
priority. Once your coro gets its turn again it will automatically be |
306 |
resumed. |
307 |
|
308 |
This function is often called C<yield> in other languages. |
309 |
|
310 |
=item Coro::cede_notself |
311 |
|
312 |
Works like cede, but is not exported by default and will cede to I<any> |
313 |
coro, regardless of priority. This is useful sometimes to ensure |
314 |
progress is made. |
315 |
|
316 |
=item terminate [arg...] |
317 |
|
318 |
Terminates the current coro with the given status values (see L<cancel>). |
319 |
|
320 |
=item Coro::on_enter BLOCK, Coro::on_leave BLOCK |
321 |
|
322 |
These function install enter and leave winders in the current scope. The |
323 |
enter block will be executed when on_enter is called and whenever the |
324 |
current coro is re-entered by the scheduler, while the leave block is |
325 |
executed whenever the current coro is blocked by the scheduler, and |
326 |
also when the containing scope is exited (by whatever means, be it exit, |
327 |
die, last etc.). |
328 |
|
329 |
I<Neither invoking the scheduler, nor exceptions, are allowed within those |
330 |
BLOCKs>. That means: do not even think about calling C<die> without an |
331 |
eval, and do not even think of entering the scheduler in any way. |
332 |
|
333 |
Since both BLOCKs are tied to the current scope, they will automatically |
334 |
be removed when the current scope exits. |
335 |
|
336 |
These functions implement the same concept as C<dynamic-wind> in scheme |
337 |
does, and are useful when you want to localise some resource to a specific |
338 |
coro. |
339 |
|
340 |
They slow down coro switching considerably for coros that use |
341 |
them (But coro switching is still reasonably fast if the handlers are |
342 |
fast). |
343 |
|
344 |
These functions are best understood by an example: The following function |
345 |
will change the current timezone to "Antarctica/South_Pole", which |
346 |
requires a call to C<tzset>, but by using C<on_enter> and C<on_leave>, |
347 |
which remember/change the current timezone and restore the previous |
348 |
value, respectively, the timezone is only changes for the coro that |
349 |
installed those handlers. |
350 |
|
351 |
use POSIX qw(tzset); |
352 |
|
353 |
async { |
354 |
my $old_tz; # store outside TZ value here |
355 |
|
356 |
Coro::on_enter { |
357 |
$old_tz = $ENV{TZ}; # remember the old value |
358 |
|
359 |
$ENV{TZ} = "Antarctica/South_Pole"; |
360 |
tzset; # enable new value |
361 |
}; |
362 |
|
363 |
Coro::on_leave { |
364 |
$ENV{TZ} = $old_tz; |
365 |
tzset; # restore old value |
366 |
}; |
367 |
|
368 |
# at this place, the timezone is Antarctica/South_Pole, |
369 |
# without disturbing the TZ of any other coro. |
370 |
}; |
371 |
|
372 |
This can be used to localise about any resource (locale, uid, current |
373 |
working directory etc.) to a block, despite the existance of other |
374 |
coros. |
375 |
|
376 |
=item killall |
377 |
|
378 |
Kills/terminates/cancels all coros except the currently running one. |
379 |
|
380 |
Note that while this will try to free some of the main interpreter |
381 |
resources if the calling coro isn't the main coro, but one |
382 |
cannot free all of them, so if a coro that is not the main coro |
383 |
calls this function, there will be some one-time resource leak. |
384 |
|
385 |
=cut |
386 |
|
387 |
sub killall { |
388 |
for (Coro::State::list) { |
389 |
$_->cancel |
390 |
if $_ != $current && UNIVERSAL::isa $_, "Coro"; |
391 |
} |
392 |
} |
393 |
|
394 |
=back |
395 |
|
396 |
=head1 CORO OBJECT METHODS |
397 |
|
398 |
These are the methods you can call on coro objects (or to create |
399 |
them). |
400 |
|
401 |
=over 4 |
402 |
|
403 |
=item new Coro \&sub [, @args...] |
404 |
|
405 |
Create a new coro and return it. When the sub returns, the coro |
406 |
automatically terminates as if C<terminate> with the returned values were |
407 |
called. To make the coro run you must first put it into the ready |
408 |
queue by calling the ready method. |
409 |
|
410 |
See C<async> and C<Coro::State::new> for additional info about the |
411 |
coro environment. |
412 |
|
413 |
=cut |
414 |
|
415 |
sub _coro_run { |
416 |
terminate &{+shift}; |
417 |
} |
418 |
|
419 |
=item $success = $coro->ready |
420 |
|
421 |
Put the given coro into the end of its ready queue (there is one |
422 |
queue for each priority) and return true. If the coro is already in |
423 |
the ready queue, do nothing and return false. |
424 |
|
425 |
This ensures that the scheduler will resume this coro automatically |
426 |
once all the coro of higher priority and all coro of the same |
427 |
priority that were put into the ready queue earlier have been resumed. |
428 |
|
429 |
=item $is_ready = $coro->is_ready |
430 |
|
431 |
Returns true iff the Coro object is in the ready queue. Unless the Coro |
432 |
object gets destroyed, it will eventually be scheduled by the scheduler. |
433 |
|
434 |
=item $is_running = $coro->is_running |
435 |
|
436 |
Returns true iff the Coro object is currently running. Only one Coro object |
437 |
can ever be in the running state (but it currently is possible to have |
438 |
multiple running Coro::States). |
439 |
|
440 |
=item $is_suspended = $coro->is_suspended |
441 |
|
442 |
Returns true iff this Coro object has been suspended. Suspended Coros will |
443 |
not ever be scheduled. |
444 |
|
445 |
=item $coro->cancel (arg...) |
446 |
|
447 |
Terminates the given Coro and makes it return the given arguments as |
448 |
status (default: the empty list). Never returns if the Coro is the |
449 |
current Coro. |
450 |
|
451 |
=cut |
452 |
|
453 |
sub cancel { |
454 |
my $self = shift; |
455 |
|
456 |
if ($current == $self) { |
457 |
terminate @_; |
458 |
} else { |
459 |
$self->{_status} = [@_]; |
460 |
Coro::State::cancel $self; |
461 |
} |
462 |
} |
463 |
|
464 |
=item $coro->schedule_to |
465 |
|
466 |
Puts the current coro to sleep (like C<Coro::schedule>), but instead |
467 |
of continuing with the next coro from the ready queue, always switch to |
468 |
the given coro object (regardless of priority etc.). The readyness |
469 |
state of that coro isn't changed. |
470 |
|
471 |
This is an advanced method for special cases - I'd love to hear about any |
472 |
uses for this one. |
473 |
|
474 |
=item $coro->cede_to |
475 |
|
476 |
Like C<schedule_to>, but puts the current coro into the ready |
477 |
queue. This has the effect of temporarily switching to the given |
478 |
coro, and continuing some time later. |
479 |
|
480 |
This is an advanced method for special cases - I'd love to hear about any |
481 |
uses for this one. |
482 |
|
483 |
=item $coro->throw ([$scalar]) |
484 |
|
485 |
If C<$throw> is specified and defined, it will be thrown as an exception |
486 |
inside the coro at the next convenient point in time. Otherwise |
487 |
clears the exception object. |
488 |
|
489 |
Coro will check for the exception each time a schedule-like-function |
490 |
returns, i.e. after each C<schedule>, C<cede>, C<< Coro::Semaphore->down |
491 |
>>, C<< Coro::Handle->readable >> and so on. Most of these functions |
492 |
detect this case and return early in case an exception is pending. |
493 |
|
494 |
The exception object will be thrown "as is" with the specified scalar in |
495 |
C<$@>, i.e. if it is a string, no line number or newline will be appended |
496 |
(unlike with C<die>). |
497 |
|
498 |
This can be used as a softer means than C<cancel> to ask a coro to |
499 |
end itself, although there is no guarantee that the exception will lead to |
500 |
termination, and if the exception isn't caught it might well end the whole |
501 |
program. |
502 |
|
503 |
You might also think of C<throw> as being the moral equivalent of |
504 |
C<kill>ing a coro with a signal (in this case, a scalar). |
505 |
|
506 |
=item $coro->join |
507 |
|
508 |
Wait until the coro terminates and return any values given to the |
509 |
C<terminate> or C<cancel> functions. C<join> can be called concurrently |
510 |
from multiple coro, and all will be resumed and given the status |
511 |
return once the C<$coro> terminates. |
512 |
|
513 |
=cut |
514 |
|
515 |
sub join { |
516 |
my $self = shift; |
517 |
|
518 |
unless ($self->{_status}) { |
519 |
my $current = $current; |
520 |
|
521 |
push @{$self->{_on_destroy}}, sub { |
522 |
$current->ready; |
523 |
undef $current; |
524 |
}; |
525 |
|
526 |
&schedule while $current; |
527 |
} |
528 |
|
529 |
wantarray ? @{$self->{_status}} : $self->{_status}[0]; |
530 |
} |
531 |
|
532 |
=item $coro->on_destroy (\&cb) |
533 |
|
534 |
Registers a callback that is called when this coro gets destroyed, |
535 |
but before it is joined. The callback gets passed the terminate arguments, |
536 |
if any, and I<must not> die, under any circumstances. |
537 |
|
538 |
=cut |
539 |
|
540 |
sub on_destroy { |
541 |
my ($self, $cb) = @_; |
542 |
|
543 |
push @{ $self->{_on_destroy} }, $cb; |
544 |
} |
545 |
|
546 |
=item $oldprio = $coro->prio ($newprio) |
547 |
|
548 |
Sets (or gets, if the argument is missing) the priority of the |
549 |
coro. Higher priority coro get run before lower priority |
550 |
coro. Priorities are small signed integers (currently -4 .. +3), |
551 |
that you can refer to using PRIO_xxx constants (use the import tag :prio |
552 |
to get then): |
553 |
|
554 |
PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN |
555 |
3 > 1 > 0 > -1 > -3 > -4 |
556 |
|
557 |
# set priority to HIGH |
558 |
current->prio (PRIO_HIGH); |
559 |
|
560 |
The idle coro ($Coro::idle) always has a lower priority than any |
561 |
existing coro. |
562 |
|
563 |
Changing the priority of the current coro will take effect immediately, |
564 |
but changing the priority of coro in the ready queue (but not |
565 |
running) will only take effect after the next schedule (of that |
566 |
coro). This is a bug that will be fixed in some future version. |
567 |
|
568 |
=item $newprio = $coro->nice ($change) |
569 |
|
570 |
Similar to C<prio>, but subtract the given value from the priority (i.e. |
571 |
higher values mean lower priority, just as in unix). |
572 |
|
573 |
=item $olddesc = $coro->desc ($newdesc) |
574 |
|
575 |
Sets (or gets in case the argument is missing) the description for this |
576 |
coro. This is just a free-form string you can associate with a |
577 |
coro. |
578 |
|
579 |
This method simply sets the C<< $coro->{desc} >> member to the given |
580 |
string. You can modify this member directly if you wish. |
581 |
|
582 |
=cut |
583 |
|
584 |
sub desc { |
585 |
my $old = $_[0]{desc}; |
586 |
$_[0]{desc} = $_[1] if @_ > 1; |
587 |
$old; |
588 |
} |
589 |
|
590 |
sub transfer { |
591 |
require Carp; |
592 |
Carp::croak ("You must not call ->transfer on Coro objects. Use Coro::State objects or the ->schedule_to method. Caught"); |
593 |
} |
594 |
|
595 |
=back |
596 |
|
597 |
=head1 GLOBAL FUNCTIONS |
598 |
|
599 |
=over 4 |
600 |
|
601 |
=item Coro::nready |
602 |
|
603 |
Returns the number of coro that are currently in the ready state, |
604 |
i.e. that can be switched to by calling C<schedule> directory or |
605 |
indirectly. The value C<0> means that the only runnable coro is the |
606 |
currently running one, so C<cede> would have no effect, and C<schedule> |
607 |
would cause a deadlock unless there is an idle handler that wakes up some |
608 |
coro. |
609 |
|
610 |
=item my $guard = Coro::guard { ... } |
611 |
|
612 |
This function still exists, but is deprecated. Please use the |
613 |
C<Guard::guard> function instead. |
614 |
|
615 |
=cut |
616 |
|
617 |
BEGIN { *guard = \&Guard::guard } |
618 |
|
619 |
=item unblock_sub { ... } |
620 |
|
621 |
This utility function takes a BLOCK or code reference and "unblocks" it, |
622 |
returning a new coderef. Unblocking means that calling the new coderef |
623 |
will return immediately without blocking, returning nothing, while the |
624 |
original code ref will be called (with parameters) from within another |
625 |
coro. |
626 |
|
627 |
The reason this function exists is that many event libraries (such as the |
628 |
venerable L<Event|Event> module) are not thread-safe (a weaker form |
629 |
of reentrancy). This means you must not block within event callbacks, |
630 |
otherwise you might suffer from crashes or worse. The only event library |
631 |
currently known that is safe to use without C<unblock_sub> is L<EV>. |
632 |
|
633 |
This function allows your callbacks to block by executing them in another |
634 |
coro where it is safe to block. One example where blocking is handy |
635 |
is when you use the L<Coro::AIO|Coro::AIO> functions to save results to |
636 |
disk, for example. |
637 |
|
638 |
In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when |
639 |
creating event callbacks that want to block. |
640 |
|
641 |
If your handler does not plan to block (e.g. simply sends a message to |
642 |
another coro, or puts some other coro into the ready queue), there is |
643 |
no reason to use C<unblock_sub>. |
644 |
|
645 |
Note that you also need to use C<unblock_sub> for any other callbacks that |
646 |
are indirectly executed by any C-based event loop. For example, when you |
647 |
use a module that uses L<AnyEvent> (and you use L<Coro::AnyEvent>) and it |
648 |
provides callbacks that are the result of some event callback, then you |
649 |
must not block either, or use C<unblock_sub>. |
650 |
|
651 |
=cut |
652 |
|
653 |
our @unblock_queue; |
654 |
|
655 |
# we create a special coro because we want to cede, |
656 |
# to reduce pressure on the coro pool (because most callbacks |
657 |
# return immediately and can be reused) and because we cannot cede |
658 |
# inside an event callback. |
659 |
our $unblock_scheduler = new Coro sub { |
660 |
while () { |
661 |
while (my $cb = pop @unblock_queue) { |
662 |
&async_pool (@$cb); |
663 |
|
664 |
# for short-lived callbacks, this reduces pressure on the coro pool |
665 |
# as the chance is very high that the async_poll coro will be back |
666 |
# in the idle state when cede returns |
667 |
cede; |
668 |
} |
669 |
schedule; # sleep well |
670 |
} |
671 |
}; |
672 |
$unblock_scheduler->{desc} = "[unblock_sub scheduler]"; |
673 |
|
674 |
sub unblock_sub(&) { |
675 |
my $cb = shift; |
676 |
|
677 |
sub { |
678 |
unshift @unblock_queue, [$cb, @_]; |
679 |
$unblock_scheduler->ready; |
680 |
} |
681 |
} |
682 |
|
683 |
=item $cb = Coro::rouse_cb |
684 |
|
685 |
Create and return a "rouse callback". That's a code reference that, |
686 |
when called, will remember a copy of its arguments and notify the owner |
687 |
coro of the callback. |
688 |
|
689 |
See the next function. |
690 |
|
691 |
=item @args = Coro::rouse_wait [$cb] |
692 |
|
693 |
Wait for the specified rouse callback (or the last one that was created in |
694 |
this coro). |
695 |
|
696 |
As soon as the callback is invoked (or when the callback was invoked |
697 |
before C<rouse_wait>), it will return the arguments originally passed to |
698 |
the rouse callback. |
699 |
|
700 |
See the section B<HOW TO WAIT FOR A CALLBACK> for an actual usage example. |
701 |
|
702 |
=back |
703 |
|
704 |
=cut |
705 |
|
706 |
1; |
707 |
|
708 |
=head1 HOW TO WAIT FOR A CALLBACK |
709 |
|
710 |
It is very common for a coro to wait for some callback to be |
711 |
called. This occurs naturally when you use coro in an otherwise |
712 |
event-based program, or when you use event-based libraries. |
713 |
|
714 |
These typically register a callback for some event, and call that callback |
715 |
when the event occured. In a coro, however, you typically want to |
716 |
just wait for the event, simplyifying things. |
717 |
|
718 |
For example C<< AnyEvent->child >> registers a callback to be called when |
719 |
a specific child has exited: |
720 |
|
721 |
my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... }); |
722 |
|
723 |
But from within a coro, you often just want to write this: |
724 |
|
725 |
my $status = wait_for_child $pid; |
726 |
|
727 |
Coro offers two functions specifically designed to make this easy, |
728 |
C<Coro::rouse_cb> and C<Coro::rouse_wait>. |
729 |
|
730 |
The first function, C<rouse_cb>, generates and returns a callback that, |
731 |
when invoked, will save its arguments and notify the coro that |
732 |
created the callback. |
733 |
|
734 |
The second function, C<rouse_wait>, waits for the callback to be called |
735 |
(by calling C<schedule> to go to sleep) and returns the arguments |
736 |
originally passed to the callback. |
737 |
|
738 |
Using these functions, it becomes easy to write the C<wait_for_child> |
739 |
function mentioned above: |
740 |
|
741 |
sub wait_for_child($) { |
742 |
my ($pid) = @_; |
743 |
|
744 |
my $watcher = AnyEvent->child (pid => $pid, cb => Coro::rouse_cb); |
745 |
|
746 |
my ($rpid, $rstatus) = Coro::rouse_wait; |
747 |
$rstatus |
748 |
} |
749 |
|
750 |
In the case where C<rouse_cb> and C<rouse_wait> are not flexible enough, |
751 |
you can roll your own, using C<schedule>: |
752 |
|
753 |
sub wait_for_child($) { |
754 |
my ($pid) = @_; |
755 |
|
756 |
# store the current coro in $current, |
757 |
# and provide result variables for the closure passed to ->child |
758 |
my $current = $Coro::current; |
759 |
my ($done, $rstatus); |
760 |
|
761 |
# pass a closure to ->child |
762 |
my $watcher = AnyEvent->child (pid => $pid, cb => sub { |
763 |
$rstatus = $_[1]; # remember rstatus |
764 |
$done = 1; # mark $rstatus as valud |
765 |
}); |
766 |
|
767 |
# wait until the closure has been called |
768 |
schedule while !$done; |
769 |
|
770 |
$rstatus |
771 |
} |
772 |
|
773 |
|
774 |
=head1 BUGS/LIMITATIONS |
775 |
|
776 |
=over 4 |
777 |
|
778 |
=item fork with pthread backend |
779 |
|
780 |
When Coro is compiled using the pthread backend (which isn't recommended |
781 |
but required on many BSDs as their libcs are completely broken), then |
782 |
coro will not survive a fork. There is no known workaround except to |
783 |
fix your libc and use a saner backend. |
784 |
|
785 |
=item perl process emulation ("threads") |
786 |
|
787 |
This module is not perl-pseudo-thread-safe. You should only ever use this |
788 |
module from the first thread (this requirement might be removed in the |
789 |
future to allow per-thread schedulers, but Coro::State does not yet allow |
790 |
this). I recommend disabling thread support and using processes, as having |
791 |
the windows process emulation enabled under unix roughly halves perl |
792 |
performance, even when not used. |
793 |
|
794 |
=item coro switching is not signal safe |
795 |
|
796 |
You must not switch to another coro from within a signal handler |
797 |
(only relevant with %SIG - most event libraries provide safe signals). |
798 |
|
799 |
That means you I<MUST NOT> call any function that might "block" the |
800 |
current coro - C<cede>, C<schedule> C<< Coro::Semaphore->down >> or |
801 |
anything that calls those. Everything else, including calling C<ready>, |
802 |
works. |
803 |
|
804 |
=back |
805 |
|
806 |
|
807 |
=head1 SEE ALSO |
808 |
|
809 |
Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>. |
810 |
|
811 |
Debugging: L<Coro::Debug>. |
812 |
|
813 |
Support/Utility: L<Coro::Specific>, L<Coro::Util>. |
814 |
|
815 |
Locking and IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>, |
816 |
L<Coro::SemaphoreSet>, L<Coro::RWLock>. |
817 |
|
818 |
I/O and Timers: L<Coro::Timer>, L<Coro::Handle>, L<Coro::Socket>, L<Coro::AIO>. |
819 |
|
820 |
Compatibility with other modules: L<Coro::LWP> (but see also L<AnyEvent::HTTP> for |
821 |
a better-working alternative), L<Coro::BDB>, L<Coro::Storable>, |
822 |
L<Coro::Select>. |
823 |
|
824 |
XS API: L<Coro::MakeMaker>. |
825 |
|
826 |
Low level Configuration, Thread Environment, Continuations: L<Coro::State>. |
827 |
|
828 |
=head1 AUTHOR |
829 |
|
830 |
Marc Lehmann <schmorp@schmorp.de> |
831 |
http://home.schmorp.de/ |
832 |
|
833 |
=cut |
834 |
|