[ViewVC] Diff of: cvs/Coro/Coro.pm

Comparing Coro/Coro.pm (file contents):
Revision 1.246 by root, Mon Dec 15 00:30:40 2008 UTC vs.
Revision 1.292 by root, Sat Apr 30 05:17:42 2011 UTC

…		…
11	print "2\n";	11	print "2\n";
12	cede; # yield back to main	12	cede; # yield back to main
13	print "4\n";	13	print "4\n";
14	};	14	};
15	print "1\n";	15	print "1\n";
16	cede; # yield to coroutine	16	cede; # yield to coro
17	print "3\n";	17	print "3\n";
18	cede; # and again	18	cede; # and again
19		19
20	# use locking	20	# use locking
21	use Coro::Semaphore;	21	use Coro::Semaphore;
…		…
29	=head1 DESCRIPTION	29	=head1 DESCRIPTION
30		30
31	For a tutorial-style introduction, please read the L<Coro::Intro>	31	For a tutorial-style introduction, please read the L<Coro::Intro>
32	manpage. This manpage mainly contains reference information.	32	manpage. This manpage mainly contains reference information.
33		33
34	This module collection manages continuations in general, most often	34	This module collection manages continuations in general, most often in
35	in the form of cooperative threads (also called coroutines in the	35	the form of cooperative threads (also called coros, or simply "coro"
36	documentation). They are similar to kernel threads but don't (in general)	36	in the documentation). They are similar to kernel threads but don't (in
37	run in parallel at the same time even on SMP machines. The specific flavor	37	general) run in parallel at the same time even on SMP machines. The
38	of thread offered by this module also guarantees you that it will not	38	specific flavor of thread offered by this module also guarantees you that
39	switch between threads unless necessary, at easily-identified points in	39	it will not switch between threads unless necessary, at easily-identified
40	your program, so locking and parallel access are rarely an issue, making	40	points in your program, so locking and parallel access are rarely an
41	thread programming much safer and easier than using other thread models.	41	issue, making thread programming much safer and easier than using other
		42	thread models.
42		43
43	Unlike the so-called "Perl threads" (which are not actually real threads	44	Unlike the so-called "Perl threads" (which are not actually real threads
44	but only the windows process emulation ported to unix), Coro provides a	45	but only the windows process emulation (see section of same name for
		46	more details) ported to UNIX, and as such act as processes), Coro
45	full shared address space, which makes communication between threads	47	provides a full shared address space, which makes communication between
46	very easy. And threads are fast, too: disabling the Windows process	48	threads very easy. And coro threads are fast, too: disabling the Windows
47	emulation code in your perl and using Coro can easily result in a two to	49	process emulation code in your perl and using Coro can easily result in
48	four times speed increase for your programs.	50	a two to four times speed increase for your programs. A parallel matrix
		51	multiplication benchmark (very communication-intensive) runs over 300
		52	times faster on a single core than perls pseudo-threads on a quad core
		53	using all four cores.
49		54
50	Coro achieves that by supporting multiple running interpreters that share	55	Coro achieves that by supporting multiple running interpreters that share
51	data, which is especially useful to code pseudo-parallel processes and	56	data, which is especially useful to code pseudo-parallel processes and
52	for event-based programming, such as multiple HTTP-GET requests running	57	for event-based programming, such as multiple HTTP-GET requests running
53	concurrently. See L<Coro::AnyEvent> to learn more on how to integrate Coro	58	concurrently. See L<Coro::AnyEvent> to learn more on how to integrate Coro
54	into an event-based environment.	59	into an event-based environment.
55		60
56	In this module, a thread is defined as "callchain + lexical variables +	61	In this module, a thread is defined as "callchain + lexical variables +
57	@_ + $_ + $@ + $/ + C stack), that is, a thread has its own callchain,	62	some package variables + C stack), that is, a thread has its own callchain,
58	its own set of lexicals and its own set of perls most important global	63	its own set of lexicals and its own set of perls most important global
59	variables (see L<Coro::State> for more configuration and background info).	64	variables (see L<Coro::State> for more configuration and background info).
60		65
61	See also the C<SEE ALSO> section at the end of this document - the Coro	66	See also the C<SEE ALSO> section at the end of this document - the Coro
62	module family is quite large.	67	module family is quite large.
63		68
		69	=head1 CORO THREAD LIFE CYCLE
		70
		71	During the long and exciting (or not) life of a coro thread, it goes
		72	through a number of states:
		73
		74	=over 4
		75
		76	=item 1. Creation
		77
		78	The first thing in the life of a coro thread is it's creation -
		79	obviously. The typical way to create a thread is to call the C<async
		80	BLOCK> function:
		81
		82	async {
		83	# thread code goes here
		84	};
		85
		86	You can also pass arguments, which are put in C<@_>:
		87
		88	async {
		89	print $_[1]; # prints 2
		90	} 1, 2, 3;
		91
		92	This creates a new coro thread and puts it into the ready queue, meaning
		93	it will run as soon as the CPU is free for it.
		94
		95	C<async> will return a coro object - you can store this for future
		96	reference or ignore it, the thread itself will keep a reference to it's
		97	thread object - threads are alive on their own.
		98
		99	Another way to create a thread is to call the C<new> constructor with a
		100	code-reference:
		101
		102	new Coro sub {
		103	# thread code goes here
		104	}, @optional_arguments;
		105
		106	This is quite similar to calling C<async>, but the important difference is
		107	that the new thread is not put into the ready queue, so the thread will
		108	not run until somebody puts it there. C<async> is, therefore, identical to
		109	this sequence:
		110
		111	my $coro = new Coro sub {
		112	# thread code goes here
		113	};
		114	$coro->ready;
		115	return $coro;
		116
		117	=item 2. Startup
		118
		119	When a new coro thread is created, only a copy of the code reference
		120	and the arguments are stored, no extra memory for stacks and so on is
		121	allocated, keeping the coro thread in a low-memory state.
		122
		123	Only when it actually starts executing will all the resources be finally
		124	allocated.
		125
		126	The optional arguments specified at coro creation are available in C<@_>,
		127	similar to function calls.
		128
		129	=item 3. Running / Blocking
		130
		131	A lot can happen after the coro thread has started running. Quite usually,
		132	it will not run to the end in one go (because you could use a function
		133	instead), but it will give up the CPU regularly because it waits for
		134	external events.
		135
		136	As long as a coro thread runs, it's coro object is available in the global
		137	variable C<$Coro::current>.
		138
		139	The low-level way to give up the CPU is to call the scheduler, which
		140	selects a new coro thread to run:
		141
		142	Coro::schedule;
		143
		144	Since running threads are not in the ready queue, calling the scheduler
		145	without doing anything else will block the coro thread forever - you need
		146	to arrange either for the coro to put woken up (readied) by some other
		147	event or some other thread, or you can put it into the ready queue before
		148	scheduling:
		149
		150	# this is exactly what Coro::cede does
		151	$Coro::current->ready;
		152	Coro::schedule;
		153
		154	All the higher-level synchronisation methods (Coro::Semaphore,
		155	Coro::rouse_*...) are actually implemented via C<< ->ready >> and C<<
		156	Coro::schedule >>.
		157
		158	While the coro thread is running it also might get assigned a C-level
		159	thread, or the C-level thread might be unassigned from it, as the Coro
		160	runtime wishes. A C-level thread needs to be assigned when your perl
		161	thread calls into some C-level function and that function in turn calls
		162	perl and perl then wants to switch coroutines. This happens most often
		163	when you run an event loop and block in the callback, or when perl
		164	itself calls some function such as C<AUTOLOAD> or methods via the C<tie>
		165	mechanism.
		166
		167	=item 4. Termination
		168
		169	Many threads actually terminate after some time. There are a number of
		170	ways to terminate a coro thread, the simplest is returning from the
		171	top-level code reference:
		172
		173	async {
		174	# after returning from here, the coro thread is terminated
		175	};
		176
		177	async {
		178	return if 0.5 < rand; # terminate a little earlier, maybe
		179	print "got a chance to print this\n";
		180	# or here
		181	};
		182
		183	Any values returned from the coroutine can be recovered using C<< ->join
		184	>>:
		185
		186	my $coro = async {
		187	"hello, world\n" # return a string
		188	};
		189
		190	my $hello_world = $coro->join;
		191
		192	print $hello_world;
		193
		194	Another way to terminate is to call C<< Coro::terminate >>, which at any
		195	subroutine call nesting level:
		196
		197	async {
		198	Coro::terminate "return value 1", "return value 2";
		199	};
		200
		201	And yet another way is to C<< ->cancel >> the coro thread from another
		202	thread:
		203
		204	my $coro = async {
		205	exit 1;
		206	};
		207
		208	$coro->cancel; # an also accept values for ->join to retrieve
		209
		210	Cancellation I<can> be dangerous - it's a bit like calling C<exit> without
		211	actually exiting, and might leave C libraries and XS modules in a weird
		212	state. Unlike other thread implementations, however, Coro is exceptionally
		213	safe with regards to cancellation, as perl will always be in a consistent
		214	state.
		215
		216	So, cancelling a thread that runs in an XS event loop might not be the
		217	best idea, but any other combination that deals with perl only (cancelling
		218	when a thread is in a C<tie> method or an C<AUTOLOAD> for example) is
		219	safe.
		220
		221	=item 5. Cleanup
		222
		223	Threads will allocate various resources. Most but not all will be returned
		224	when a thread terminates, during clean-up.
		225
		226	Cleanup is quite similar to throwing an uncaught exception: perl will
		227	work it's way up through all subroutine calls and blocks. On it's way, it
		228	will release all C<my> variables, undo all C<local>'s and free any other
		229	resources truly local to the thread.
		230
		231	So, a common way to free resources is to keep them referenced only by my
		232	variables:
		233
		234	async {
		235	my $big_cache = new Cache ...;
		236	};
		237
		238	If there are no other references, then the C<$big_cache> object will be
		239	freed when the thread terminates, regardless of how it does so.
		240
		241	What it does C<NOT> do is unlock any Coro::Semaphores or similar
		242	resources, but that's where the C<guard> methods come in handy:
		243
		244	my $sem = new Coro::Semaphore;
		245
		246	async {
		247	my $lock_guard = $sem->guard;
		248	# if we reutrn, or die or get cancelled, here,
		249	# then the semaphore will be "up"ed.
		250	};
		251
		252	The C<Guard::guard> function comes in handy for any custom cleanup you
		253	might want to do:
		254
		255	async {
		256	my $window = new Gtk2::Window "toplevel";
		257	# The window will not be cleaned up automatically, even when $window
		258	# gets freed, so use a guard to ensure it's destruction
		259	# in case of an error:
		260	my $window_guard = Guard::guard { $window->destroy };
		261
		262	# we are safe here
		263	};
		264
		265	Last not least, C<local> can often be handy, too, e.g. when temporarily
		266	replacing the coro thread description:
		267
		268	sub myfunction {
		269	local $Coro::current->{desc} = "inside myfunction(@_)";
		270
		271	# if we return or die here, the description will be restored
		272	}
		273
		274	=item 6. Viva La Zombie Muerte
		275
		276	Even after a thread has terminated and cleaned up it's resources, the coro
		277	object still is there and stores the return values of the thread. Only in
		278	this state will the coro object be "reference counted" in the normal perl
		279	sense: the thread code keeps a reference to it when it is active, but not
		280	after it has terminated.
		281
		282	The means the coro object gets freed automatically when the thread has
		283	terminated and cleaned up and there arenot other references.
		284
		285	If there are, the coro object will stay around, and you can call C<<
		286	->join >> as many times as you wish to retrieve the result values:
		287
		288	async {
		289	print "hi\n";
		290	1
		291	};
		292
		293	# run the async above, and free everything before returning
		294	# from Coro::cede:
		295	Coro::cede;
		296
		297	{
		298	my $coro = async {
		299	print "hi\n";
		300	1
		301	};
		302
		303	# run the async above, and clean up, but do not free the coro
		304	# object:
		305	Coro::cede;
		306
		307	# optionally retrieve the result values
		308	my @results = $coro->join;
		309
		310	# now $coro goes out of scope, and presumably gets freed
		311	};
		312
		313	=back
		314
64	=cut	315	=cut
65		316
66	package Coro;	317	package Coro;
67		318
68	use strict qw(vars subs);	319	use common::sense;
69	no warnings "uninitialized";	320
		321	use Carp ();
70		322
71	use Guard ();	323	use Guard ();
72		324
73	use Coro::State;	325	use Coro::State;
74		326
75	use base qw(Coro::State Exporter);	327	use base qw(Coro::State Exporter);
76		328
77	our $idle; # idle handler	329	our $idle; # idle handler
78	our $main; # main coroutine	330	our $main; # main coro
79	our $current; # current coroutine	331	our $current; # current coro
80		332
81	our $VERSION = 5.13;	333	our $VERSION = 5.372;
82		334
83	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub);	335	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub rouse_cb rouse_wait);
84	our %EXPORT_TAGS = (	336	our %EXPORT_TAGS = (
85	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],	337	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],
86	);	338	);
87	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));	339	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));
88		340
…		…
90		342
91	=over 4	343	=over 4
92		344
93	=item $Coro::main	345	=item $Coro::main
94		346
95	This variable stores the coroutine object that represents the main	347	This variable stores the Coro object that represents the main
96	program. While you cna C<ready> it and do most other things you can do to	348	program. While you cna C<ready> it and do most other things you can do to
97	coroutines, it is mainly useful to compare again C<$Coro::current>, to see	349	coro, it is mainly useful to compare again C<$Coro::current>, to see
98	whether you are running in the main program or not.	350	whether you are running in the main program or not.
99		351
100	=cut	352	=cut
101		353
102	# $main is now being initialised by Coro::State	354	# $main is now being initialised by Coro::State
103		355
104	=item $Coro::current	356	=item $Coro::current
105		357
106	The coroutine object representing the current coroutine (the last	358	The Coro object representing the current coro (the last
107	coroutine that the Coro scheduler switched to). The initial value is	359	coro that the Coro scheduler switched to). The initial value is
108	C<$Coro::main> (of course).	360	C<$Coro::main> (of course).
109		361
110	This variable is B<strictly> I<read-only>. You can take copies of the	362	This variable is B<strictly> I<read-only>. You can take copies of the
111	value stored in it and use it as any other coroutine object, but you must	363	value stored in it and use it as any other Coro object, but you must
112	not otherwise modify the variable itself.	364	not otherwise modify the variable itself.
113		365
114	=cut	366	=cut
115		367
116	sub current() { $current } # [DEPRECATED]	368	sub current() { $current } # [DEPRECATED]
…		…
119		371
120	This variable is mainly useful to integrate Coro into event loops. It is	372	This variable is mainly useful to integrate Coro into event loops. It is
121	usually better to rely on L<Coro::AnyEvent> or L<Coro::EV>, as this is	373	usually better to rely on L<Coro::AnyEvent> or L<Coro::EV>, as this is
122	pretty low-level functionality.	374	pretty low-level functionality.
123		375
124	This variable stores either a coroutine or a callback.	376	This variable stores a Coro object that is put into the ready queue when
		377	there are no other ready threads (without invoking any ready hooks).
125		378
126	If it is a callback, the it is called whenever the scheduler finds no	379	The default implementation dies with "FATAL: deadlock detected.", followed
127	ready coroutines to run. The default implementation prints "FATAL:	380	by a thread listing, because the program has no other way to continue.
128	deadlock detected" and exits, because the program has no other way to
129	continue.
130
131	If it is a coroutine object, then this object will be readied (without
132	invoking any ready hooks, however) when the scheduler finds no other ready
133	coroutines to run.
134		381
135	This hook is overwritten by modules such as C<Coro::EV> and	382	This hook is overwritten by modules such as C<Coro::EV> and
136	C<Coro::AnyEvent> to wait on an external event that hopefully wake up a	383	C<Coro::AnyEvent> to wait on an external event that hopefully wakes up a
137	coroutine so the scheduler can run it.	384	coro so the scheduler can run it.
138		385
139	Note that the callback I<must not>, under any circumstances, block
140	the current coroutine. Normally, this is achieved by having an "idle
141	coroutine" that calls the event loop and then blocks again, and then
142	readying that coroutine in the idle handler, or by simply placing the idle
143	coroutine in this variable.
144
145	See L<Coro::Event> or L<Coro::AnyEvent> for examples of using this	386	See L<Coro::EV> or L<Coro::AnyEvent> for examples of using this technique.
146	technique.
147		387
148	Please note that if your callback recursively invokes perl (e.g. for event
149	handlers), then it must be prepared to be called recursively itself.
150
151	=cut	388	=cut
152		389
153	$idle = sub {	390	# \|\|= because other modules could have provided their own by now
154	require Carp;	391	$idle \|\|= new Coro sub {
155	Carp::croak ("FATAL: deadlock detected");	392	require Coro::Debug;
		393	die "FATAL: deadlock detected.\n"
		394	. Coro::Debug::ps_listing ();
156	};	395	};
157		396
158	# this coroutine is necessary because a coroutine	397	# this coro is necessary because a coro
159	# cannot destroy itself.	398	# cannot destroy itself.
160	our @destroy;	399	our @destroy;
161	our $manager;	400	our $manager;
162		401
163	$manager = new Coro sub {	402	$manager = new Coro sub {
164	while () {	403	while () {
165	Coro::_cancel shift @destroy	404	_destroy shift @destroy
166	while @destroy;	405	while @destroy;
167		406
168	&schedule;	407	&schedule;
169	}	408	}
170	};	409	};
171	$manager->{desc} = "[coro manager]";	410	$manager->{desc} = "[coro manager]";
172	$manager->prio (PRIO_MAX);	411	$manager->prio (PRIO_MAX);
173		412
174	=back	413	=back
175		414
176	=head1 SIMPLE COROUTINE CREATION	415	=head1 SIMPLE CORO CREATION
177		416
178	=over 4	417	=over 4
179		418
180	=item async { ... } [@args...]	419	=item async { ... } [@args...]
181		420
182	Create a new coroutine and return its coroutine object (usually	421	Create a new coro and return its Coro object (usually
183	unused). The coroutine will be put into the ready queue, so	422	unused). The coro will be put into the ready queue, so
184	it will start running automatically on the next scheduler run.	423	it will start running automatically on the next scheduler run.
185		424
186	The first argument is a codeblock/closure that should be executed in the	425	The first argument is a codeblock/closure that should be executed in the
187	coroutine. When it returns argument returns the coroutine is automatically	426	coro. When it returns argument returns the coro is automatically
188	terminated.	427	terminated.
189		428
190	The remaining arguments are passed as arguments to the closure.	429	The remaining arguments are passed as arguments to the closure.
191		430
192	See the C<Coro::State::new> constructor for info about the coroutine	431	See the C<Coro::State::new> constructor for info about the coro
193	environment in which coroutines are executed.	432	environment in which coro are executed.
194		433
195	Calling C<exit> in a coroutine will do the same as calling exit outside	434	Calling C<exit> in a coro will do the same as calling exit outside
196	the coroutine. Likewise, when the coroutine dies, the program will exit,	435	the coro. Likewise, when the coro dies, the program will exit,
197	just as it would in the main program.	436	just as it would in the main program.
198		437
199	If you do not want that, you can provide a default C<die> handler, or	438	If you do not want that, you can provide a default C<die> handler, or
200	simply avoid dieing (by use of C<eval>).	439	simply avoid dieing (by use of C<eval>).
201		440
202	Example: Create a new coroutine that just prints its arguments.	441	Example: Create a new coro that just prints its arguments.
203		442
204	async {	443	async {
205	print "@_\n";	444	print "@_\n";
206	} 1,2,3,4;	445	} 1,2,3,4;
207		446
208	=cut
209
210	sub async(&@) {
211	my $coro = new Coro @_;
212	$coro->ready;
213	$coro
214	}
215
216	=item async_pool { ... } [@args...]	447	=item async_pool { ... } [@args...]
217		448
218	Similar to C<async>, but uses a coroutine pool, so you should not call	449	Similar to C<async>, but uses a coro pool, so you should not call
219	terminate or join on it (although you are allowed to), and you get a	450	terminate or join on it (although you are allowed to), and you get a
220	coroutine that might have executed other code already (which can be good	451	coro that might have executed other code already (which can be good
221	or bad :).	452	or bad :).
222		453
223	On the plus side, this function is about twice as fast as creating (and	454	On the plus side, this function is about twice as fast as creating (and
224	destroying) a completely new coroutine, so if you need a lot of generic	455	destroying) a completely new coro, so if you need a lot of generic
225	coroutines in quick successsion, use C<async_pool>, not C<async>.	456	coros in quick successsion, use C<async_pool>, not C<async>.
226		457
227	The code block is executed in an C<eval> context and a warning will be	458	The code block is executed in an C<eval> context and a warning will be
228	issued in case of an exception instead of terminating the program, as	459	issued in case of an exception instead of terminating the program, as
229	C<async> does. As the coroutine is being reused, stuff like C<on_destroy>	460	C<async> does. As the coro is being reused, stuff like C<on_destroy>
230	will not work in the expected way, unless you call terminate or cancel,	461	will not work in the expected way, unless you call terminate or cancel,
231	which somehow defeats the purpose of pooling (but is fine in the	462	which somehow defeats the purpose of pooling (but is fine in the
232	exceptional case).	463	exceptional case).
233		464
234	The priority will be reset to C<0> after each run, tracing will be	465	The priority will be reset to C<0> after each run, tracing will be
235	disabled, the description will be reset and the default output filehandle	466	disabled, the description will be reset and the default output filehandle
236	gets restored, so you can change all these. Otherwise the coroutine will	467	gets restored, so you can change all these. Otherwise the coro will
237	be re-used "as-is": most notably if you change other per-coroutine global	468	be re-used "as-is": most notably if you change other per-coro global
238	stuff such as C<$/> you I<must needs> revert that change, which is most	469	stuff such as C<$/> you I<must needs> revert that change, which is most
239	simply done by using local as in: C<< local $/ >>.	470	simply done by using local as in: C<< local $/ >>.
240		471
241	The idle pool size is limited to C<8> idle coroutines (this can be	472	The idle pool size is limited to C<8> idle coros (this can be
242	adjusted by changing $Coro::POOL_SIZE), but there can be as many non-idle	473	adjusted by changing $Coro::POOL_SIZE), but there can be as many non-idle
243	coros as required.	474	coros as required.
244		475
245	If you are concerned about pooled coroutines growing a lot because a	476	If you are concerned about pooled coros growing a lot because a
246	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool	477	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool
247	{ terminate }> once per second or so to slowly replenish the pool. In	478	{ terminate }> once per second or so to slowly replenish the pool. In
248	addition to that, when the stacks used by a handler grows larger than 32kb	479	addition to that, when the stacks used by a handler grows larger than 32kb
249	(adjustable via $Coro::POOL_RSS) it will also be destroyed.	480	(adjustable via $Coro::POOL_RSS) it will also be destroyed.
250		481
…		…
267	=back	498	=back
268		499
269	=head1 STATIC METHODS	500	=head1 STATIC METHODS
270		501
271	Static methods are actually functions that implicitly operate on the	502	Static methods are actually functions that implicitly operate on the
272	current coroutine.	503	current coro.
273		504
274	=over 4	505	=over 4
275		506
276	=item schedule	507	=item schedule
277		508
278	Calls the scheduler. The scheduler will find the next coroutine that is	509	Calls the scheduler. The scheduler will find the next coro that is
279	to be run from the ready queue and switches to it. The next coroutine	510	to be run from the ready queue and switches to it. The next coro
280	to be run is simply the one with the highest priority that is longest	511	to be run is simply the one with the highest priority that is longest
281	in its ready queue. If there is no coroutine ready, it will clal the	512	in its ready queue. If there is no coro ready, it will call the
282	C<$Coro::idle> hook.	513	C<$Coro::idle> hook.
283		514
284	Please note that the current coroutine will I<not> be put into the ready	515	Please note that the current coro will I<not> be put into the ready
285	queue, so calling this function usually means you will never be called	516	queue, so calling this function usually means you will never be called
286	again unless something else (e.g. an event handler) calls C<< ->ready >>,	517	again unless something else (e.g. an event handler) calls C<< ->ready >>,
287	thus waking you up.	518	thus waking you up.
288		519
289	This makes C<schedule> I<the> generic method to use to block the current	520	This makes C<schedule> I<the> generic method to use to block the current
290	coroutine and wait for events: first you remember the current coroutine in	521	coro and wait for events: first you remember the current coro in
291	a variable, then arrange for some callback of yours to call C<< ->ready	522	a variable, then arrange for some callback of yours to call C<< ->ready
292	>> on that once some event happens, and last you call C<schedule> to put	523	>> on that once some event happens, and last you call C<schedule> to put
293	yourself to sleep. Note that a lot of things can wake your coroutine up,	524	yourself to sleep. Note that a lot of things can wake your coro up,
294	so you need to check whether the event indeed happened, e.g. by storing the	525	so you need to check whether the event indeed happened, e.g. by storing the
295	status in a variable.	526	status in a variable.
296		527
297	See B<HOW TO WAIT FOR A CALLBACK>, below, for some ways to wait for callbacks.	528	See B<HOW TO WAIT FOR A CALLBACK>, below, for some ways to wait for callbacks.
298		529
299	=item cede	530	=item cede
300		531
301	"Cede" to other coroutines. This function puts the current coroutine into	532	"Cede" to other coros. This function puts the current coro into
302	the ready queue and calls C<schedule>, which has the effect of giving	533	the ready queue and calls C<schedule>, which has the effect of giving
303	up the current "timeslice" to other coroutines of the same or higher	534	up the current "timeslice" to other coros of the same or higher
304	priority. Once your coroutine gets its turn again it will automatically be	535	priority. Once your coro gets its turn again it will automatically be
305	resumed.	536	resumed.
306		537
307	This function is often called C<yield> in other languages.	538	This function is often called C<yield> in other languages.
308		539
309	=item Coro::cede_notself	540	=item Coro::cede_notself
310		541
311	Works like cede, but is not exported by default and will cede to I<any>	542	Works like cede, but is not exported by default and will cede to I<any>
312	coroutine, regardless of priority. This is useful sometimes to ensure	543	coro, regardless of priority. This is useful sometimes to ensure
313	progress is made.	544	progress is made.
314		545
315	=item terminate [arg...]	546	=item terminate [arg...]
316		547
317	Terminates the current coroutine with the given status values (see L<cancel>).	548	Terminates the current coro with the given status values (see
		549	L<cancel>). The values will not be copied, but referenced directly.
		550
		551	=item Coro::on_enter BLOCK, Coro::on_leave BLOCK
		552
		553	These function install enter and leave winders in the current scope. The
		554	enter block will be executed when on_enter is called and whenever the
		555	current coro is re-entered by the scheduler, while the leave block is
		556	executed whenever the current coro is blocked by the scheduler, and
		557	also when the containing scope is exited (by whatever means, be it exit,
		558	die, last etc.).
		559
		560	I<Neither invoking the scheduler, nor exceptions, are allowed within those
		561	BLOCKs>. That means: do not even think about calling C<die> without an
		562	eval, and do not even think of entering the scheduler in any way.
		563
		564	Since both BLOCKs are tied to the current scope, they will automatically
		565	be removed when the current scope exits.
		566
		567	These functions implement the same concept as C<dynamic-wind> in scheme
		568	does, and are useful when you want to localise some resource to a specific
		569	coro.
		570
		571	They slow down thread switching considerably for coros that use them
		572	(about 40% for a BLOCK with a single assignment, so thread switching is
		573	still reasonably fast if the handlers are fast).
		574
		575	These functions are best understood by an example: The following function
		576	will change the current timezone to "Antarctica/South_Pole", which
		577	requires a call to C<tzset>, but by using C<on_enter> and C<on_leave>,
		578	which remember/change the current timezone and restore the previous
		579	value, respectively, the timezone is only changed for the coro that
		580	installed those handlers.
		581
		582	use POSIX qw(tzset);
		583
		584	async {
		585	my $old_tz; # store outside TZ value here
		586
		587	Coro::on_enter {
		588	$old_tz = $ENV{TZ}; # remember the old value
		589
		590	$ENV{TZ} = "Antarctica/South_Pole";
		591	tzset; # enable new value
		592	};
		593
		594	Coro::on_leave {
		595	$ENV{TZ} = $old_tz;
		596	tzset; # restore old value
		597	};
		598
		599	# at this place, the timezone is Antarctica/South_Pole,
		600	# without disturbing the TZ of any other coro.
		601	};
		602
		603	This can be used to localise about any resource (locale, uid, current
		604	working directory etc.) to a block, despite the existance of other
		605	coros.
		606
		607	Another interesting example implements time-sliced multitasking using
		608	interval timers (this could obviously be optimised, but does the job):
		609
		610	# "timeslice" the given block
		611	sub timeslice(&) {
		612	use Time::HiRes ();
		613
		614	Coro::on_enter {
		615	# on entering the thread, we set an VTALRM handler to cede
		616	$SIG{VTALRM} = sub { cede };
		617	# and then start the interval timer
		618	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0.01, 0.01;
		619	};
		620	Coro::on_leave {
		621	# on leaving the thread, we stop the interval timer again
		622	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0, 0;
		623	};
		624
		625	&{+shift};
		626	}
		627
		628	# use like this:
		629	timeslice {
		630	# The following is an endless loop that would normally
		631	# monopolise the process. Since it runs in a timesliced
		632	# environment, it will regularly cede to other threads.
		633	while () { }
		634	};
		635
318		636
319	=item killall	637	=item killall
320		638
321	Kills/terminates/cancels all coroutines except the currently running	639	Kills/terminates/cancels all coros except the currently running one.
322	one. This can be useful after a fork, either in the child or the parent,
323	as usually only one of them should inherit the running coroutines.
324		640
325	Note that in the implementation, destructors run as normal, making this
326	function not so useful after a fork. Future versions of this function
327	might try to free resources without running any code.
328
329	Note that while this will try to free some of the main programs resources,	641	Note that while this will try to free some of the main interpreter
		642	resources if the calling coro isn't the main coro, but one
330	you cannot free all of them, so if a coroutine that is not the main	643	cannot free all of them, so if a coro that is not the main coro
331	program calls this function, there will be some one-time resource leak.	644	calls this function, there will be some one-time resource leak.
332		645
333	=cut	646	=cut
334		647
335	sub killall {	648	sub killall {
336	for (Coro::State::list) {	649	for (Coro::State::list) {
…		…
339	}	652	}
340	}	653	}
341		654
342	=back	655	=back
343		656
344	=head1 COROUTINE OBJECT METHODS	657	=head1 CORO OBJECT METHODS
345		658
346	These are the methods you can call on coroutine objects (or to create	659	These are the methods you can call on coro objects (or to create
347	them).	660	them).
348		661
349	=over 4	662	=over 4
350		663
351	=item new Coro \&sub [, @args...]	664	=item new Coro \&sub [, @args...]
352		665
353	Create a new coroutine and return it. When the sub returns, the coroutine	666	Create a new coro and return it. When the sub returns, the coro
354	automatically terminates as if C<terminate> with the returned values were	667	automatically terminates as if C<terminate> with the returned values were
355	called. To make the coroutine run you must first put it into the ready	668	called. To make the coro run you must first put it into the ready
356	queue by calling the ready method.	669	queue by calling the ready method.
357		670
358	See C<async> and C<Coro::State::new> for additional info about the	671	See C<async> and C<Coro::State::new> for additional info about the
359	coroutine environment.	672	coro environment.
360		673
361	=cut	674	=cut
362		675
363	sub _coro_run {	676	sub _coro_run {
364	terminate &{+shift};	677	terminate &{+shift};
365	}	678	}
366		679
367	=item $success = $coroutine->ready	680	=item $success = $coro->ready
368		681
369	Put the given coroutine into the end of its ready queue (there is one	682	Put the given coro into the end of its ready queue (there is one
370	queue for each priority) and return true. If the coroutine is already in	683	queue for each priority) and return true. If the coro is already in
371	the ready queue, do nothing and return false.	684	the ready queue, do nothing and return false.
372		685
373	This ensures that the scheduler will resume this coroutine automatically	686	This ensures that the scheduler will resume this coro automatically
374	once all the coroutines of higher priority and all coroutines of the same	687	once all the coro of higher priority and all coro of the same
375	priority that were put into the ready queue earlier have been resumed.	688	priority that were put into the ready queue earlier have been resumed.
376		689
		690	=item $coro->suspend
		691
		692	Suspends the specified coro. A suspended coro works just like any other
		693	coro, except that the scheduler will not select a suspended coro for
		694	execution.
		695
		696	Suspending a coro can be useful when you want to keep the coro from
		697	running, but you don't want to destroy it, or when you want to temporarily
		698	freeze a coro (e.g. for debugging) to resume it later.
		699
		700	A scenario for the former would be to suspend all (other) coros after a
		701	fork and keep them alive, so their destructors aren't called, but new
		702	coros can be created.
		703
		704	=item $coro->resume
		705
		706	If the specified coro was suspended, it will be resumed. Note that when
		707	the coro was in the ready queue when it was suspended, it might have been
		708	unreadied by the scheduler, so an activation might have been lost.
		709
		710	To avoid this, it is best to put a suspended coro into the ready queue
		711	unconditionally, as every synchronisation mechanism must protect itself
		712	against spurious wakeups, and the one in the Coro family certainly do
		713	that.
		714
377	=item $is_ready = $coroutine->is_ready	715	=item $is_ready = $coro->is_ready
378		716
379	Return whether the coroutine is currently the ready queue or not,	717	Returns true iff the Coro object is in the ready queue. Unless the Coro
		718	object gets destroyed, it will eventually be scheduled by the scheduler.
380		719
		720	=item $is_running = $coro->is_running
		721
		722	Returns true iff the Coro object is currently running. Only one Coro object
		723	can ever be in the running state (but it currently is possible to have
		724	multiple running Coro::States).
		725
		726	=item $is_suspended = $coro->is_suspended
		727
		728	Returns true iff this Coro object has been suspended. Suspended Coros will
		729	not ever be scheduled.
		730
381	=item $coroutine->cancel (arg...)	731	=item $coro->cancel (arg...)
382		732
383	Terminates the given coroutine and makes it return the given arguments as	733	Terminates the given Coro object and makes it return the given arguments as
384	status (default: the empty list). Never returns if the coroutine is the	734	status (default: an empty list). Never returns if the Coro is the
385	current coroutine.	735	current Coro.
386		736
387	=cut	737	The arguments are not copied, but instead will be referenced directly
		738	(e.g. if you pass C<$var> and after the call change that variable, then
		739	you might change the return values passed to e.g. C<join>, so don't do
		740	that).
388		741
389	sub cancel {	742	The resources of the Coro are usually freed (or destructed) before this
390	my $self = shift;	743	call returns, but this can be delayed for an indefinite amount of time, as
		744	in some cases the manager thread has to run first to actually destruct the
		745	Coro object.
391		746
392	if ($current == $self) {
393	terminate @_;
394	} else {
395	$self->{_status} = [@_];
396	$self->_cancel;
397	}
398	}
399
400	=item $coroutine->schedule_to	747	=item $coro->schedule_to
401		748
402	Puts the current coroutine to sleep (like C<Coro::schedule>), but instead	749	Puts the current coro to sleep (like C<Coro::schedule>), but instead
403	of continuing with the next coro from the ready queue, always switch to	750	of continuing with the next coro from the ready queue, always switch to
404	the given coroutine object (regardless of priority etc.). The readyness	751	the given coro object (regardless of priority etc.). The readyness
405	state of that coroutine isn't changed.	752	state of that coro isn't changed.
406		753
407	This is an advanced method for special cases - I'd love to hear about any	754	This is an advanced method for special cases - I'd love to hear about any
408	uses for this one.	755	uses for this one.
409		756
410	=item $coroutine->cede_to	757	=item $coro->cede_to
411		758
412	Like C<schedule_to>, but puts the current coroutine into the ready	759	Like C<schedule_to>, but puts the current coro into the ready
413	queue. This has the effect of temporarily switching to the given	760	queue. This has the effect of temporarily switching to the given
414	coroutine, and continuing some time later.	761	coro, and continuing some time later.
415		762
416	This is an advanced method for special cases - I'd love to hear about any	763	This is an advanced method for special cases - I'd love to hear about any
417	uses for this one.	764	uses for this one.
418		765
419	=item $coroutine->throw ([$scalar])	766	=item $coro->throw ([$scalar])
420		767
421	If C<$throw> is specified and defined, it will be thrown as an exception	768	If C<$throw> is specified and defined, it will be thrown as an exception
422	inside the coroutine at the next convenient point in time. Otherwise	769	inside the coro at the next convenient point in time. Otherwise
423	clears the exception object.	770	clears the exception object.
424		771
425	Coro will check for the exception each time a schedule-like-function	772	Coro will check for the exception each time a schedule-like-function
426	returns, i.e. after each C<schedule>, C<cede>, C<< Coro::Semaphore->down	773	returns, i.e. after each C<schedule>, C<cede>, C<< Coro::Semaphore->down
427	>>, C<< Coro::Handle->readable >> and so on. Most of these functions	774	>>, C<< Coro::Handle->readable >> and so on. Most of these functions
…		…
429		776
430	The exception object will be thrown "as is" with the specified scalar in	777	The exception object will be thrown "as is" with the specified scalar in
431	C<$@>, i.e. if it is a string, no line number or newline will be appended	778	C<$@>, i.e. if it is a string, no line number or newline will be appended
432	(unlike with C<die>).	779	(unlike with C<die>).
433		780
434	This can be used as a softer means than C<cancel> to ask a coroutine to	781	This can be used as a softer means than C<cancel> to ask a coro to
435	end itself, although there is no guarantee that the exception will lead to	782	end itself, although there is no guarantee that the exception will lead to
436	termination, and if the exception isn't caught it might well end the whole	783	termination, and if the exception isn't caught it might well end the whole
437	program.	784	program.
438		785
439	You might also think of C<throw> as being the moral equivalent of	786	You might also think of C<throw> as being the moral equivalent of
440	C<kill>ing a coroutine with a signal (in this case, a scalar).	787	C<kill>ing a coro with a signal (in this case, a scalar).
441		788
442	=item $coroutine->join	789	=item $coro->join
443		790
444	Wait until the coroutine terminates and return any values given to the	791	Wait until the coro terminates and return any values given to the
445	C<terminate> or C<cancel> functions. C<join> can be called concurrently	792	C<terminate> or C<cancel> functions. C<join> can be called concurrently
446	from multiple coroutines, and all will be resumed and given the status	793	from multiple threads, and all will be resumed and given the status
447	return once the C<$coroutine> terminates.	794	return once the C<$coro> terminates.
448		795
449	=cut	796	=cut
450		797
451	sub join {	798	sub join {
452	my $self = shift;	799	my $self = shift;
…		…
463	}	810	}
464		811
465	wantarray ? @{$self->{_status}} : $self->{_status}[0];	812	wantarray ? @{$self->{_status}} : $self->{_status}[0];
466	}	813	}
467		814
468	=item $coroutine->on_destroy (\&cb)	815	=item $coro->on_destroy (\&cb)
469		816
470	Registers a callback that is called when this coroutine gets destroyed,	817	Registers a callback that is called when this coro thread gets destroyed,
471	but before it is joined. The callback gets passed the terminate arguments,	818	but before it is joined. The callback gets passed the terminate arguments,
472	if any, and I<must not> die, under any circumstances.	819	if any, and I<must not> die, under any circumstances.
473		820
		821	There can be any number of C<on_destroy> callbacks per coro.
		822
474	=cut	823	=cut
475		824
476	sub on_destroy {	825	sub on_destroy {
477	my ($self, $cb) = @_;	826	my ($self, $cb) = @_;
478		827
479	push @{ $self->{_on_destroy} }, $cb;	828	push @{ $self->{_on_destroy} }, $cb;
480	}	829	}
481		830
482	=item $oldprio = $coroutine->prio ($newprio)	831	=item $oldprio = $coro->prio ($newprio)
483		832
484	Sets (or gets, if the argument is missing) the priority of the	833	Sets (or gets, if the argument is missing) the priority of the
485	coroutine. Higher priority coroutines get run before lower priority	834	coro thread. Higher priority coro get run before lower priority
486	coroutines. Priorities are small signed integers (currently -4 .. +3),	835	coros. Priorities are small signed integers (currently -4 .. +3),
487	that you can refer to using PRIO_xxx constants (use the import tag :prio	836	that you can refer to using PRIO_xxx constants (use the import tag :prio
488	to get then):	837	to get then):
489		838
490	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN	839	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN
491	3 > 1 > 0 > -1 > -3 > -4	840	3 > 1 > 0 > -1 > -3 > -4
492		841
493	# set priority to HIGH	842	# set priority to HIGH
494	current->prio(PRIO_HIGH);	843	current->prio (PRIO_HIGH);
495		844
496	The idle coroutine ($Coro::idle) always has a lower priority than any	845	The idle coro thread ($Coro::idle) always has a lower priority than any
497	existing coroutine.	846	existing coro.
498		847
499	Changing the priority of the current coroutine will take effect immediately,	848	Changing the priority of the current coro will take effect immediately,
500	but changing the priority of coroutines in the ready queue (but not	849	but changing the priority of a coro in the ready queue (but not running)
501	running) will only take effect after the next schedule (of that	850	will only take effect after the next schedule (of that coro). This is a
502	coroutine). This is a bug that will be fixed in some future version.	851	bug that will be fixed in some future version.
503		852
504	=item $newprio = $coroutine->nice ($change)	853	=item $newprio = $coro->nice ($change)
505		854
506	Similar to C<prio>, but subtract the given value from the priority (i.e.	855	Similar to C<prio>, but subtract the given value from the priority (i.e.
507	higher values mean lower priority, just as in unix).	856	higher values mean lower priority, just as in UNIX's nice command).
508		857
509	=item $olddesc = $coroutine->desc ($newdesc)	858	=item $olddesc = $coro->desc ($newdesc)
510		859
511	Sets (or gets in case the argument is missing) the description for this	860	Sets (or gets in case the argument is missing) the description for this
512	coroutine. This is just a free-form string you can associate with a	861	coro thread. This is just a free-form string you can associate with a
513	coroutine.	862	coro.
514		863
515	This method simply sets the C<< $coroutine->{desc} >> member to the given	864	This method simply sets the C<< $coro->{desc} >> member to the given
516	string. You can modify this member directly if you wish.	865	string. You can modify this member directly if you wish, and in fact, this
		866	is often preferred to indicate major processing states that cna then be
		867	seen for example in a L<Coro::Debug> session:
		868
		869	sub my_long_function {
		870	local $Coro::current->{desc} = "now in my_long_function";
		871	...
		872	$Coro::current->{desc} = "my_long_function: phase 1";
		873	...
		874	$Coro::current->{desc} = "my_long_function: phase 2";
		875	...
		876	}
517		877
518	=cut	878	=cut
519		879
520	sub desc {	880	sub desc {
521	my $old = $_[0]{desc};	881	my $old = $_[0]{desc};
…		…
534		894
535	=over 4	895	=over 4
536		896
537	=item Coro::nready	897	=item Coro::nready
538		898
539	Returns the number of coroutines that are currently in the ready state,	899	Returns the number of coro that are currently in the ready state,
540	i.e. that can be switched to by calling C<schedule> directory or	900	i.e. that can be switched to by calling C<schedule> directory or
541	indirectly. The value C<0> means that the only runnable coroutine is the	901	indirectly. The value C<0> means that the only runnable coro is the
542	currently running one, so C<cede> would have no effect, and C<schedule>	902	currently running one, so C<cede> would have no effect, and C<schedule>
543	would cause a deadlock unless there is an idle handler that wakes up some	903	would cause a deadlock unless there is an idle handler that wakes up some
544	coroutines.	904	coro.
545		905
546	=item my $guard = Coro::guard { ... }	906	=item my $guard = Coro::guard { ... }
547		907
548	This function still exists, but is deprecated. Please use the	908	This function still exists, but is deprecated. Please use the
549	C<Guard::guard> function instead.	909	C<Guard::guard> function instead.
…		…
556		916
557	This utility function takes a BLOCK or code reference and "unblocks" it,	917	This utility function takes a BLOCK or code reference and "unblocks" it,
558	returning a new coderef. Unblocking means that calling the new coderef	918	returning a new coderef. Unblocking means that calling the new coderef
559	will return immediately without blocking, returning nothing, while the	919	will return immediately without blocking, returning nothing, while the
560	original code ref will be called (with parameters) from within another	920	original code ref will be called (with parameters) from within another
561	coroutine.	921	coro.
562		922
563	The reason this function exists is that many event libraries (such as the	923	The reason this function exists is that many event libraries (such as
564	venerable L<Event\|Event> module) are not coroutine-safe (a weaker form	924	the venerable L<Event\|Event> module) are not thread-safe (a weaker form
565	of reentrancy). This means you must not block within event callbacks,	925	of reentrancy). This means you must not block within event callbacks,
566	otherwise you might suffer from crashes or worse. The only event library	926	otherwise you might suffer from crashes or worse. The only event library
567	currently known that is safe to use without C<unblock_sub> is L<EV>.	927	currently known that is safe to use without C<unblock_sub> is L<EV> (but
		928	you might still run into deadlocks if all event loops are blocked).
		929
		930	Coro will try to catch you when you block in the event loop
		931	("FATAL:$Coro::IDLE blocked itself"), but this is just best effort and
		932	only works when you do not run your own event loop.
568		933
569	This function allows your callbacks to block by executing them in another	934	This function allows your callbacks to block by executing them in another
570	coroutine where it is safe to block. One example where blocking is handy	935	coro where it is safe to block. One example where blocking is handy
571	is when you use the L<Coro::AIO\|Coro::AIO> functions to save results to	936	is when you use the L<Coro::AIO\|Coro::AIO> functions to save results to
572	disk, for example.	937	disk, for example.
573		938
574	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when	939	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when
575	creating event callbacks that want to block.	940	creating event callbacks that want to block.
576		941
577	If your handler does not plan to block (e.g. simply sends a message to	942	If your handler does not plan to block (e.g. simply sends a message to
578	another coroutine, or puts some other coroutine into the ready queue),	943	another coro, or puts some other coro into the ready queue), there is
579	there is no reason to use C<unblock_sub>.	944	no reason to use C<unblock_sub>.
580		945
581	Note that you also need to use C<unblock_sub> for any other callbacks that	946	Note that you also need to use C<unblock_sub> for any other callbacks that
582	are indirectly executed by any C-based event loop. For example, when you	947	are indirectly executed by any C-based event loop. For example, when you
583	use a module that uses L<AnyEvent> (and you use L<Coro::AnyEvent>) and it	948	use a module that uses L<AnyEvent> (and you use L<Coro::AnyEvent>) and it
584	provides callbacks that are the result of some event callback, then you	949	provides callbacks that are the result of some event callback, then you
…		…
614	unshift @unblock_queue, [$cb, @_];	979	unshift @unblock_queue, [$cb, @_];
615	$unblock_scheduler->ready;	980	$unblock_scheduler->ready;
616	}	981	}
617	}	982	}
618		983
619	=item $cb = Coro::rouse_cb	984	=item $cb = rouse_cb
620		985
621	Create and return a "rouse callback". That's a code reference that,	986	Create and return a "rouse callback". That's a code reference that,
622	when called, will remember a copy of its arguments and notify the owner	987	when called, will remember a copy of its arguments and notify the owner
623	coroutine of the callback.	988	coro of the callback.
624		989
625	See the next function.	990	See the next function.
626		991
627	=item @args = Coro::rouse_wait [$cb]	992	=item @args = rouse_wait [$cb]
628		993
629	Wait for the specified rouse callback (or the last one that was created in	994	Wait for the specified rouse callback (or the last one that was created in
630	this coroutine).	995	this coro).
631		996
632	As soon as the callback is invoked (or when the callback was invoked	997	As soon as the callback is invoked (or when the callback was invoked
633	before C<rouse_wait>), it will return the arguments originally passed to	998	before C<rouse_wait>), it will return the arguments originally passed to
634	the rouse callback.	999	the rouse callback. In scalar context, that means you get the I<last>
		1000	argument, just as if C<rouse_wait> had a C<return ($a1, $a2, $a3...)>
		1001	statement at the end.
635		1002
636	See the section B<HOW TO WAIT FOR A CALLBACK> for an actual usage example.	1003	See the section B<HOW TO WAIT FOR A CALLBACK> for an actual usage example.
637		1004
638	=back	1005	=back
639		1006
640	=cut	1007	=cut
641		1008
		1009	for my $module (qw(Channel RWLock Semaphore SemaphoreSet Signal Specific)) {
		1010	my $old = defined &{"Coro::$module\::new"} && \&{"Coro::$module\::new"};
		1011
		1012	*{"Coro::$module\::new"} = sub {
		1013	require "Coro/$module.pm";
		1014
		1015	# some modules have their new predefined in State.xs, some don't
		1016	*{"Coro::$module\::new"} = $old
		1017	if $old;
		1018
		1019	goto &{"Coro::$module\::new"};
		1020	};
		1021	}
		1022
642	1;	1023	1;
643		1024
644	=head1 HOW TO WAIT FOR A CALLBACK	1025	=head1 HOW TO WAIT FOR A CALLBACK
645		1026
646	It is very common for a coroutine to wait for some callback to be	1027	It is very common for a coro to wait for some callback to be
647	called. This occurs naturally when you use coroutines in an otherwise	1028	called. This occurs naturally when you use coro in an otherwise
648	event-based program, or when you use event-based libraries.	1029	event-based program, or when you use event-based libraries.
649		1030
650	These typically register a callback for some event, and call that callback	1031	These typically register a callback for some event, and call that callback
651	when the event occured. In a coroutine, however, you typically want to	1032	when the event occured. In a coro, however, you typically want to
652	just wait for the event, simplyifying things.	1033	just wait for the event, simplyifying things.
653		1034
654	For example C<< AnyEvent->child >> registers a callback to be called when	1035	For example C<< AnyEvent->child >> registers a callback to be called when
655	a specific child has exited:	1036	a specific child has exited:
656		1037
657	my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... });	1038	my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... });
658		1039
659	But from withina coroutine, you often just want to write this:	1040	But from within a coro, you often just want to write this:
660		1041
661	my $status = wait_for_child $pid;	1042	my $status = wait_for_child $pid;
662		1043
663	Coro offers two functions specifically designed to make this easy,	1044	Coro offers two functions specifically designed to make this easy,
664	C<Coro::rouse_cb> and C<Coro::rouse_wait>.	1045	C<Coro::rouse_cb> and C<Coro::rouse_wait>.
665		1046
666	The first function, C<rouse_cb>, generates and returns a callback that,	1047	The first function, C<rouse_cb>, generates and returns a callback that,
667	when invoked, will save its arguments and notify the coroutine that	1048	when invoked, will save its arguments and notify the coro that
668	created the callback.	1049	created the callback.
669		1050
670	The second function, C<rouse_wait>, waits for the callback to be called	1051	The second function, C<rouse_wait>, waits for the callback to be called
671	(by calling C<schedule> to go to sleep) and returns the arguments	1052	(by calling C<schedule> to go to sleep) and returns the arguments
672	originally passed to the callback.	1053	originally passed to the callback.
…		…
687	you can roll your own, using C<schedule>:	1068	you can roll your own, using C<schedule>:
688		1069
689	sub wait_for_child($) {	1070	sub wait_for_child($) {
690	my ($pid) = @_;	1071	my ($pid) = @_;
691		1072
692	# store the current coroutine in $current,	1073	# store the current coro in $current,
693	# and provide result variables for the closure passed to ->child	1074	# and provide result variables for the closure passed to ->child
694	my $current = $Coro::current;	1075	my $current = $Coro::current;
695	my ($done, $rstatus);	1076	my ($done, $rstatus);
696		1077
697	# pass a closure to ->child	1078	# pass a closure to ->child
…		…
713		1094
714	=item fork with pthread backend	1095	=item fork with pthread backend
715		1096
716	When Coro is compiled using the pthread backend (which isn't recommended	1097	When Coro is compiled using the pthread backend (which isn't recommended
717	but required on many BSDs as their libcs are completely broken), then	1098	but required on many BSDs as their libcs are completely broken), then
718	coroutines will not survive a fork. There is no known workaround except to	1099	coro will not survive a fork. There is no known workaround except to
719	fix your libc and use a saner backend.	1100	fix your libc and use a saner backend.
720		1101
721	=item perl process emulation ("threads")	1102	=item perl process emulation ("threads")
722		1103
723	This module is not perl-pseudo-thread-safe. You should only ever use this	1104	This module is not perl-pseudo-thread-safe. You should only ever use this
…		…
725	future to allow per-thread schedulers, but Coro::State does not yet allow	1106	future to allow per-thread schedulers, but Coro::State does not yet allow
726	this). I recommend disabling thread support and using processes, as having	1107	this). I recommend disabling thread support and using processes, as having
727	the windows process emulation enabled under unix roughly halves perl	1108	the windows process emulation enabled under unix roughly halves perl
728	performance, even when not used.	1109	performance, even when not used.
729		1110
730	=item coroutine switching not signal safe	1111	=item coro switching is not signal safe
731		1112
732	You must not switch to another coroutine from within a signal handler	1113	You must not switch to another coro from within a signal handler (only
733	(only relevant with %SIG - most event libraries provide safe signals).	1114	relevant with %SIG - most event libraries provide safe signals), I<unless>
		1115	you are sure you are not interrupting a Coro function.
734		1116
735	That means you I<MUST NOT> call any function that might "block" the	1117	That means you I<MUST NOT> call any function that might "block" the
736	current coroutine - C<cede>, C<schedule> C<< Coro::Semaphore->down >> or	1118	current coro - C<cede>, C<schedule> C<< Coro::Semaphore->down >> or
737	anything that calls those. Everything else, including calling C<ready>,	1119	anything that calls those. Everything else, including calling C<ready>,
738	works.	1120	works.
739		1121
740	=back	1122	=back
741		1123
		1124
		1125	=head1 WINDOWS PROCESS EMULATION
		1126
		1127	A great many people seem to be confused about ithreads (for example, Chip
		1128	Salzenberg called me unintelligent, incapable, stupid and gullible,
		1129	while in the same mail making rather confused statements about perl
		1130	ithreads (for example, that memory or files would be shared), showing his
		1131	lack of understanding of this area - if it is hard to understand for Chip,
		1132	it is probably not obvious to everybody).
		1133
		1134	What follows is an ultra-condensed version of my talk about threads in
		1135	scripting languages given on the perl workshop 2009:
		1136
		1137	The so-called "ithreads" were originally implemented for two reasons:
		1138	first, to (badly) emulate unix processes on native win32 perls, and
		1139	secondly, to replace the older, real thread model ("5.005-threads").
		1140
		1141	It does that by using threads instead of OS processes. The difference
		1142	between processes and threads is that threads share memory (and other
		1143	state, such as files) between threads within a single process, while
		1144	processes do not share anything (at least not semantically). That
		1145	means that modifications done by one thread are seen by others, while
		1146	modifications by one process are not seen by other processes.
		1147
		1148	The "ithreads" work exactly like that: when creating a new ithreads
		1149	process, all state is copied (memory is copied physically, files and code
		1150	is copied logically). Afterwards, it isolates all modifications. On UNIX,
		1151	the same behaviour can be achieved by using operating system processes,
		1152	except that UNIX typically uses hardware built into the system to do this
		1153	efficiently, while the windows process emulation emulates this hardware in
		1154	software (rather efficiently, but of course it is still much slower than
		1155	dedicated hardware).
		1156
		1157	As mentioned before, loading code, modifying code, modifying data
		1158	structures and so on is only visible in the ithreads process doing the
		1159	modification, not in other ithread processes within the same OS process.
		1160
		1161	This is why "ithreads" do not implement threads for perl at all, only
		1162	processes. What makes it so bad is that on non-windows platforms, you can
		1163	actually take advantage of custom hardware for this purpose (as evidenced
		1164	by the forks module, which gives you the (i-) threads API, just much
		1165	faster).
		1166
		1167	Sharing data is in the i-threads model is done by transfering data
		1168	structures between threads using copying semantics, which is very slow -
		1169	shared data simply does not exist. Benchmarks using i-threads which are
		1170	communication-intensive show extremely bad behaviour with i-threads (in
		1171	fact, so bad that Coro, which cannot take direct advantage of multiple
		1172	CPUs, is often orders of magnitude faster because it shares data using
		1173	real threads, refer to my talk for details).
		1174
		1175	As summary, i-threads use threads to implement processes, while
		1176	the compatible forks module uses processes to emulate, uhm,
		1177	processes. I-threads slow down every perl program when enabled, and
		1178	outside of windows, serve no (or little) practical purpose, but
		1179	disadvantages every single-threaded Perl program.
		1180
		1181	This is the reason that I try to avoid the name "ithreads", as it is
		1182	misleading as it implies that it implements some kind of thread model for
		1183	perl, and prefer the name "windows process emulation", which describes the
		1184	actual use and behaviour of it much better.
742		1185
743	=head1 SEE ALSO	1186	=head1 SEE ALSO
744		1187
745	Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.	1188	Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.
746		1189

Diff Legend

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

Comparing Coro/Coro.pm (file contents): Revision 1.246 by root, Mon Dec 15 00:30:40 2008 UTC vs. Revision 1.292 by root, Sat Apr 30 05:17:42 2011 UTC

Diff Legend

Comparing Coro/Coro.pm (file contents):
Revision 1.246 by root, Mon Dec 15 00:30:40 2008 UTC vs.
Revision 1.292 by root, Sat Apr 30 05:17:42 2011 UTC