ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/Coro/Coro.pm

Comparing Coro/Coro.pm (file contents):
Revision 1.163 by root, Mon Dec 17 06:36:24 2007 UTC vs.
Revision 1.343 by root, Sun Jun 26 21:46:03 2016 UTC

1	=head1 NAME	1	=head1 NAME
2		2
3	Coro - coroutine process abstraction	3	Coro - the only real threads in perl
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use Coro;	7	use Coro;
8		8
9	async {	9	async {
10	# some asynchronous thread of execution	10	# some asynchronous thread of execution
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	my $lock = new Coro::Semaphore;	21	my $lock = new Coro::Semaphore;
22	my $locked;	22	my $locked;
23		23
24	$lock->down;	24	$lock->down;
25	$locked = 1;	25	$locked = 1;
26	$lock->up;	26	$lock->up;
27		27
28	=head1 DESCRIPTION	28	=head1 DESCRIPTION
29		29
30	This module collection manages coroutines. Coroutines are similar	30	For a tutorial-style introduction, please read the L<Coro::Intro>
31	to threads but don't run in parallel at the same time even on SMP	31	manpage. This manpage mainly contains reference information.
32	machines. The specific flavor of coroutine used in this module also
33	guarantees you that it will not switch between coroutines unless
34	necessary, at easily-identified points in your program, so locking and
35	parallel access are rarely an issue, making coroutine programming much
36	safer than threads programming.
37		32
38	(Perl, however, does not natively support real threads but instead does a	33	This module collection manages continuations in general, most often in
39	very slow and memory-intensive emulation of processes using threads. This	34	the form of cooperative threads (also called coros, or simply "coro"
40	is a performance win on Windows machines, and a loss everywhere else).	35	in the documentation). They are similar to kernel threads but don't (in
		36	general) run in parallel at the same time even on SMP machines. The
		37	specific flavor of thread offered by this module also guarantees you that
		38	it will not switch between threads unless necessary, at easily-identified
		39	points in your program, so locking and parallel access are rarely an
		40	issue, making thread programming much safer and easier than using other
		41	thread models.
41		42
		43	Unlike the so-called "Perl threads" (which are not actually real threads
		44	but only the windows process emulation (see section of same name for
		45	more details) ported to UNIX, and as such act as processes), Coro
		46	provides a full shared address space, which makes communication between
		47	threads very easy. And coro threads are fast, too: disabling the Windows
		48	process emulation code in your perl and using Coro can easily result in
		49	a two to four times speed increase for your programs. A parallel matrix
		50	multiplication benchmark (very communication-intensive) runs over 300
		51	times faster on a single core than perls pseudo-threads on a quad core
		52	using all four cores.
		53
		54	Coro achieves that by supporting multiple running interpreters that share
		55	data, which is especially useful to code pseudo-parallel processes and
		56	for event-based programming, such as multiple HTTP-GET requests running
		57	concurrently. See L<Coro::AnyEvent> to learn more on how to integrate Coro
		58	into an event-based environment.
		59
42	In this module, coroutines are defined as "callchain + lexical variables +	60	In this module, a thread is defined as "callchain + lexical variables +
43	@_ + $_ + $@ + $/ + C stack), that is, a coroutine has its own callchain,	61	some package variables + C stack), that is, a thread has its own callchain,
44	its own set of lexicals and its own set of perls most important global	62	its own set of lexicals and its own set of perls most important global
45	variables (see L<Coro::State> for more configuration).	63	variables (see L<Coro::State> for more configuration and background info).
		64
		65	See also the C<SEE ALSO> section at the end of this document - the Coro
		66	module family is quite large.
		67
		68	=head1 CORO THREAD LIFE CYCLE
		69
		70	During the long and exciting (or not) life of a coro thread, it goes
		71	through a number of states:
		72
		73	=over 4
		74
		75	=item 1. Creation
		76
		77	The first thing in the life of a coro thread is it's creation -
		78	obviously. The typical way to create a thread is to call the C<async
		79	BLOCK> function:
		80
		81	async {
		82	# thread code goes here
		83	};
		84
		85	You can also pass arguments, which are put in C<@_>:
		86
		87	async {
		88	print $_[1]; # prints 2
		89	} 1, 2, 3;
		90
		91	This creates a new coro thread and puts it into the ready queue, meaning
		92	it will run as soon as the CPU is free for it.
		93
		94	C<async> will return a Coro object - you can store this for future
		95	reference or ignore it - a thread that is running, ready to run or waiting
		96	for some event is alive on it's own.
		97
		98	Another way to create a thread is to call the C<new> constructor with a
		99	code-reference:
		100
		101	new Coro sub {
		102	# thread code goes here
		103	}, @optional_arguments;
		104
		105	This is quite similar to calling C<async>, but the important difference is
		106	that the new thread is not put into the ready queue, so the thread will
		107	not run until somebody puts it there. C<async> is, therefore, identical to
		108	this sequence:
		109
		110	my $coro = new Coro sub {
		111	# thread code goes here
		112	};
		113	$coro->ready;
		114	return $coro;
		115
		116	=item 2. Startup
		117
		118	When a new coro thread is created, only a copy of the code reference
		119	and the arguments are stored, no extra memory for stacks and so on is
		120	allocated, keeping the coro thread in a low-memory state.
		121
		122	Only when it actually starts executing will all the resources be finally
		123	allocated.
		124
		125	The optional arguments specified at coro creation are available in C<@_>,
		126	similar to function calls.
		127
		128	=item 3. Running / Blocking
		129
		130	A lot can happen after the coro thread has started running. Quite usually,
		131	it will not run to the end in one go (because you could use a function
		132	instead), but it will give up the CPU regularly because it waits for
		133	external events.
		134
		135	As long as a coro thread runs, its Coro object is available in the global
		136	variable C<$Coro::current>.
		137
		138	The low-level way to give up the CPU is to call the scheduler, which
		139	selects a new coro thread to run:
		140
		141	Coro::schedule;
		142
		143	Since running threads are not in the ready queue, calling the scheduler
		144	without doing anything else will block the coro thread forever - you need
		145	to arrange either for the coro to put woken up (readied) by some other
		146	event or some other thread, or you can put it into the ready queue before
		147	scheduling:
		148
		149	# this is exactly what Coro::cede does
		150	$Coro::current->ready;
		151	Coro::schedule;
		152
		153	All the higher-level synchronisation methods (Coro::Semaphore,
		154	Coro::rouse_*...) are actually implemented via C<< ->ready >> and C<<
		155	Coro::schedule >>.
		156
		157	While the coro thread is running it also might get assigned a C-level
		158	thread, or the C-level thread might be unassigned from it, as the Coro
		159	runtime wishes. A C-level thread needs to be assigned when your perl
		160	thread calls into some C-level function and that function in turn calls
		161	perl and perl then wants to switch coroutines. This happens most often
		162	when you run an event loop and block in the callback, or when perl
		163	itself calls some function such as C<AUTOLOAD> or methods via the C<tie>
		164	mechanism.
		165
		166	=item 4. Termination
		167
		168	Many threads actually terminate after some time. There are a number of
		169	ways to terminate a coro thread, the simplest is returning from the
		170	top-level code reference:
		171
		172	async {
		173	# after returning from here, the coro thread is terminated
		174	};
		175
		176	async {
		177	return if 0.5 < rand; # terminate a little earlier, maybe
		178	print "got a chance to print this\n";
		179	# or here
		180	};
		181
		182	Any values returned from the coroutine can be recovered using C<< ->join
		183	>>:
		184
		185	my $coro = async {
		186	"hello, world\n" # return a string
		187	};
		188
		189	my $hello_world = $coro->join;
		190
		191	print $hello_world;
		192
		193	Another way to terminate is to call C<< Coro::terminate >>, which at any
		194	subroutine call nesting level:
		195
		196	async {
		197	Coro::terminate "return value 1", "return value 2";
		198	};
		199
		200	Yet another way is to C<< ->cancel >> (or C<< ->safe_cancel >>) the coro
		201	thread from another thread:
		202
		203	my $coro = async {
		204	exit 1;
		205	};
		206
		207	$coro->cancel; # also accepts values for ->join to retrieve
		208
		209	Cancellation I<can> be dangerous - it's a bit like calling C<exit> without
		210	actually exiting, and might leave C libraries and XS modules in a weird
		211	state. Unlike other thread implementations, however, Coro is exceptionally
		212	safe with regards to cancellation, as perl will always be in a consistent
		213	state, and for those cases where you want to do truly marvellous things
		214	with your coro while it is being cancelled - that is, make sure all
		215	cleanup code is executed from the thread being cancelled - there is even a
		216	C<< ->safe_cancel >> method.
		217
		218	So, cancelling a thread that runs in an XS event loop might not be the
		219	best idea, but any other combination that deals with perl only (cancelling
		220	when a thread is in a C<tie> method or an C<AUTOLOAD> for example) is
		221	safe.
		222
		223	Last not least, a coro thread object that isn't referenced is C<<
		224	->cancel >>'ed automatically - just like other objects in Perl. This
		225	is not such a common case, however - a running thread is referencedy by
		226	C<$Coro::current>, a thread ready to run is referenced by the ready queue,
		227	a thread waiting on a lock or semaphore is referenced by being in some
		228	wait list and so on. But a thread that isn't in any of those queues gets
		229	cancelled:
		230
		231	async {
		232	schedule; # cede to other coros, don't go into the ready queue
		233	};
		234
		235	cede;
		236	# now the async above is destroyed, as it is not referenced by anything.
		237
		238	A slightly embellished example might make it clearer:
		239
		240	async {
		241	my $guard = Guard::guard { print "destroyed\n" };
		242	schedule while 1;
		243	};
		244
		245	cede;
		246
		247	Superficially one might not expect any output - since the C<async>
		248	implements an endless loop, the C<$guard> will not be cleaned up. However,
		249	since the thread object returned by C<async> is not stored anywhere, the
		250	thread is initially referenced because it is in the ready queue, when it
		251	runs it is referenced by C<$Coro::current>, but when it calls C<schedule>,
		252	it gets C<cancel>ed causing the guard object to be destroyed (see the next
		253	section), and printing it's message.
		254
		255	If this seems a bit drastic, remember that this only happens when nothing
		256	references the thread anymore, which means there is no way to further
		257	execute it, ever. The only options at this point are leaking the thread,
		258	or cleaning it up, which brings us to...
		259
		260	=item 5. Cleanup
		261
		262	Threads will allocate various resources. Most but not all will be returned
		263	when a thread terminates, during clean-up.
		264
		265	Cleanup is quite similar to throwing an uncaught exception: perl will
		266	work it's way up through all subroutine calls and blocks. On it's way, it
		267	will release all C<my> variables, undo all C<local>'s and free any other
		268	resources truly local to the thread.
		269
		270	So, a common way to free resources is to keep them referenced only by my
		271	variables:
		272
		273	async {
		274	my $big_cache = new Cache ...;
		275	};
		276
		277	If there are no other references, then the C<$big_cache> object will be
		278	freed when the thread terminates, regardless of how it does so.
		279
		280	What it does C<NOT> do is unlock any Coro::Semaphores or similar
		281	resources, but that's where the C<guard> methods come in handy:
		282
		283	my $sem = new Coro::Semaphore;
		284
		285	async {
		286	my $lock_guard = $sem->guard;
		287	# if we return, or die or get cancelled, here,
		288	# then the semaphore will be "up"ed.
		289	};
		290
		291	The C<Guard::guard> function comes in handy for any custom cleanup you
		292	might want to do (but you cannot switch to other coroutines from those
		293	code blocks):
		294
		295	async {
		296	my $window = new Gtk2::Window "toplevel";
		297	# The window will not be cleaned up automatically, even when $window
		298	# gets freed, so use a guard to ensure it's destruction
		299	# in case of an error:
		300	my $window_guard = Guard::guard { $window->destroy };
		301
		302	# we are safe here
		303	};
		304
		305	Last not least, C<local> can often be handy, too, e.g. when temporarily
		306	replacing the coro thread description:
		307
		308	sub myfunction {
		309	local $Coro::current->{desc} = "inside myfunction(@_)";
		310
		311	# if we return or die here, the description will be restored
		312	}
		313
		314	=item 6. Viva La Zombie Muerte
		315
		316	Even after a thread has terminated and cleaned up its resources, the Coro
		317	object still is there and stores the return values of the thread.
		318
		319	When there are no other references, it will simply be cleaned up and
		320	freed.
		321
		322	If there areany references, the Coro object will stay around, and you
		323	can call C<< ->join >> as many times as you wish to retrieve the result
		324	values:
		325
		326	async {
		327	print "hi\n";
		328	1
		329	};
		330
		331	# run the async above, and free everything before returning
		332	# from Coro::cede:
		333	Coro::cede;
		334
		335	{
		336	my $coro = async {
		337	print "hi\n";
		338	1
		339	};
		340
		341	# run the async above, and clean up, but do not free the coro
		342	# object:
		343	Coro::cede;
		344
		345	# optionally retrieve the result values
		346	my @results = $coro->join;
		347
		348	# now $coro goes out of scope, and presumably gets freed
		349	};
		350
		351	=back
46		352
47	=cut	353	=cut
48		354
49	package Coro;	355	package Coro;
50		356
51	use strict;	357	use common::sense;
52	no warnings "uninitialized";	358
		359	use Carp ();
		360
		361	use Guard ();
53		362
54	use Coro::State;	363	use Coro::State;
55		364
56	use base qw(Coro::State Exporter);	365	use base qw(Coro::State Exporter);
57		366
58	our $idle; # idle handler	367	our $idle; # idle handler
59	our $main; # main coroutine	368	our $main; # main coro
60	our $current; # current coroutine	369	our $current; # current coro
61		370
62	our $VERSION = '4.32';	371	our $VERSION = 6.511;
63		372
64	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub);	373	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub rouse_cb rouse_wait);
65	our %EXPORT_TAGS = (	374	our %EXPORT_TAGS = (
66	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],	375	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],
67	);	376	);
68	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));	377	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));
69		378
70	{	379	=head1 GLOBAL VARIABLES
71	my @async;
72	my $init;
73
74	# this way of handling attributes simply is NOT scalable ;()
75	sub import {
76	no strict 'refs';
77
78	Coro->export_to_level (1, @_);
79
80	my $old = *{(caller)[0]."::MODIFY_CODE_ATTRIBUTES"}{CODE};
81	*{(caller)[0]."::MODIFY_CODE_ATTRIBUTES"} = sub {
82	my ($package, $ref) = (shift, shift);
83	my @attrs;
84	for (@_) {
85	if ($_ eq "Coro") {
86	push @async, $ref;
87	unless ($init++) {
88	eval q{
89	sub INIT {
90	&async(pop @async) while @async;
91	}
92	};
93	}
94	} else {
95	push @attrs, $_;
96	}
97	}
98	return $old ? $old->($package, $ref, @attrs) : @attrs;
99	};
100	}
101
102	}
103		380
104	=over 4	381	=over 4
105		382
106	=item $main	383	=item $Coro::main
107		384
108	This coroutine represents the main program.	385	This variable stores the Coro object that represents the main
		386	program. While you can C<ready> it and do most other things you can do to
		387	coro, it is mainly useful to compare again C<$Coro::current>, to see
		388	whether you are running in the main program or not.
109		389
110	=cut	390	=cut
111		391
112	$main = new Coro;	392	# $main is now being initialised by Coro::State
113		393
114	=item $current (or as function: current)	394	=item $Coro::current
115		395
116	The current coroutine (the last coroutine switched to). The initial value	396	The Coro object representing the current coro (the last
		397	coro that the Coro scheduler switched to). The initial value is
117	is C<$main> (of course).	398	C<$Coro::main> (of course).
118		399
119	This variable is B<strictly> I<read-only>. It is provided for performance	400	This variable is B<strictly> I<read-only>. You can take copies of the
120	reasons. If performance is not essential you are encouraged to use the	401	value stored in it and use it as any other Coro object, but you must
121	C<Coro::current> function instead.	402	not otherwise modify the variable itself.
122		403
123	=cut	404	=cut
124		405
125	$main->{desc} = "[main::]";
126
127	# maybe some other module used Coro::Specific before...
128	$main->{_specific} = $current->{_specific}
129	if $current;
130
131	_set_current $main;
132
133	sub current() { $current }	406	sub current() { $current } # [DEPRECATED]
134		407
135	=item $idle	408	=item $Coro::idle
136		409
137	A callback that is called whenever the scheduler finds no ready coroutines	410	This variable is mainly useful to integrate Coro into event loops. It is
		411	usually better to rely on L<Coro::AnyEvent> or L<Coro::EV>, as this is
		412	pretty low-level functionality.
		413
		414	This variable stores a Coro object that is put into the ready queue when
		415	there are no other ready threads (without invoking any ready hooks).
		416
138	to run. The default implementation prints "FATAL: deadlock detected" and	417	The default implementation dies with "FATAL: deadlock detected.", followed
139	exits, because the program has no other way to continue.	418	by a thread listing, because the program has no other way to continue.
140		419
141	This hook is overwritten by modules such as C<Coro::Timer> and	420	This hook is overwritten by modules such as C<Coro::EV> and
142	C<Coro::Event> to wait on an external event that hopefully wake up a	421	C<Coro::AnyEvent> to wait on an external event that hopefully wakes up a
143	coroutine so the scheduler can run it.	422	coro so the scheduler can run it.
144		423
145	Please note that if your callback recursively invokes perl (e.g. for event	424	See L<Coro::EV> or L<Coro::AnyEvent> for examples of using this technique.
146	handlers), then it must be prepared to be called recursively itself.
147		425
148	=cut	426	=cut
149		427
150	$idle = sub {	428	# ||= because other modules could have provided their own by now
151	require Carp;	429	$idle ||= new Coro sub {
152	Carp::croak ("FATAL: deadlock detected");	430	require Coro::Debug;
		431	die "FATAL: deadlock detected.\n"
		432	. Coro::Debug::ps_listing ();
153	};	433	};
154		434
155	sub _cancel {
156	my ($self) = @_;
157
158	# free coroutine data and mark as destructed
159	$self->_destroy
160	or return;
161
162	# call all destruction callbacks
163	$_->(@{$self->{_status}})
164	for @{(delete $self->{_on_destroy}) || []};
165	}
166
167	# this coroutine is necessary because a coroutine	435	# this coro is necessary because a coro
168	# cannot destroy itself.	436	# cannot destroy itself.
169	my @destroy;	437	our @destroy;
170	my $manager;	438	our $manager;
171		439
172	$manager = new Coro sub {	440	$manager = new Coro sub {
173	while () {	441	while () {
174	(shift @destroy)->_cancel	442	_destroy shift @destroy
175	while @destroy;	443	while @destroy;
176		444
177	&schedule;	445	&schedule;
178	}	446	}
179	};	447	};
180	$manager->desc ("[coro manager]");	448	$manager->{desc} = "[coro manager]";
181	$manager->prio (PRIO_MAX);	449	$manager->prio (PRIO_MAX);
182		450
183	# static methods. not really.
184
185	=back	451	=back
186		452
187	=head2 STATIC METHODS	453	=head1 SIMPLE CORO CREATION
188
189	Static methods are actually functions that operate on the current coroutine only.
190		454
191	=over 4	455	=over 4
192		456
193	=item async { ... } [@args...]	457	=item async { ... } [@args...]
194		458
195	Create a new asynchronous coroutine and return it's coroutine object	459	Create a new coro and return its Coro object (usually
196	(usually unused). When the sub returns the new coroutine is automatically	460	unused). The coro will be put into the ready queue, so
		461	it will start running automatically on the next scheduler run.
		462
		463	The first argument is a codeblock/closure that should be executed in the
		464	coro. When it returns argument returns the coro is automatically
197	terminated.	465	terminated.
198		466
		467	The remaining arguments are passed as arguments to the closure.
		468
199	See the C<Coro::State::new> constructor for info about the coroutine	469	See the C<Coro::State::new> constructor for info about the coro
200	environment in which coroutines run.	470	environment in which coro are executed.
201		471
202	Calling C<exit> in a coroutine will do the same as calling exit outside	472	Calling C<exit> in a coro will do the same as calling exit outside
203	the coroutine. Likewise, when the coroutine dies, the program will exit,	473	the coro. Likewise, when the coro dies, the program will exit,
204	just as it would in the main program.	474	just as it would in the main program.
205		475
		476	If you do not want that, you can provide a default C<die> handler, or
		477	simply avoid dieing (by use of C<eval>).
		478
206	# create a new coroutine that just prints its arguments	479	Example: Create a new coro that just prints its arguments.
		480
207	async {	481	async {
208	print "@_\n";	482	print "@_\n";
209	} 1,2,3,4;	483	} 1,2,3,4;
210		484
211	=cut
212
213	sub async(&@) {
214	my $coro = new Coro @_;
215	$coro->ready;
216	$coro
217	}
218
219	=item async_pool { ... } [@args...]	485	=item async_pool { ... } [@args...]
220		486
221	Similar to C<async>, but uses a coroutine pool, so you should not call	487	Similar to C<async>, but uses a coro pool, so you should not call
222	terminate or join (although you are allowed to), and you get a coroutine	488	terminate or join on it (although you are allowed to), and you get a
223	that might have executed other code already (which can be good or bad :).	489	coro that might have executed other code already (which can be good
		490	or bad :).
224		491
		492	On the plus side, this function is about twice as fast as creating (and
		493	destroying) a completely new coro, so if you need a lot of generic
		494	coros in quick successsion, use C<async_pool>, not C<async>.
		495
225	Also, the block is executed in an C<eval> context and a warning will be	496	The code block is executed in an C<eval> context and a warning will be
226	issued in case of an exception instead of terminating the program, as	497	issued in case of an exception instead of terminating the program, as
227	C<async> does. As the coroutine is being reused, stuff like C<on_destroy>	498	C<async> does. As the coro is being reused, stuff like C<on_destroy>
228	will not work in the expected way, unless you call terminate or cancel,	499	will not work in the expected way, unless you call terminate or cancel,
229	which somehow defeats the purpose of pooling.	500	which somehow defeats the purpose of pooling (but is fine in the
		501	exceptional case).
230		502
231	The priority will be reset to C<0> after each job, tracing will be	503	The priority will be reset to C<0> after each run, all C<swap_sv> calls
232	disabled, the description will be reset and the default output filehandle	504	will be undone, tracing will be disabled, the description will be reset
233	gets restored, so you can change alkl these. Otherwise the coroutine will	505	and the default output filehandle gets restored, so you can change all
234	be re-used "as-is": most notably if you change other per-coroutine global	506	these. Otherwise the coro will be re-used "as-is": most notably if you
235	stuff such as C<$/> you need to revert that change, which is most simply	507	change other per-coro global stuff such as C<$/> you I<must needs> revert
236	done by using local as in C< local $/ >.	508	that change, which is most simply done by using local as in: C<< local $/
		509	>>.
237		510
238	The pool size is limited to 8 idle coroutines (this can be adjusted by	511	The idle pool size is limited to C<8> idle coros (this can be
239	changing $Coro::POOL_SIZE), and there can be as many non-idle coros as	512	adjusted by changing $Coro::POOL_SIZE), but there can be as many non-idle
240	required.	513	coros as required.
241		514
242	If you are concerned about pooled coroutines growing a lot because a	515	If you are concerned about pooled coros growing a lot because a
243	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool	516	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool
244	{ terminate }> once per second or so to slowly replenish the pool. In	517	{ terminate }> once per second or so to slowly replenish the pool. In
245	addition to that, when the stacks used by a handler grows larger than 16kb	518	addition to that, when the stacks used by a handler grows larger than 32kb
246	(adjustable with $Coro::POOL_RSS) it will also exit.	519	(adjustable via $Coro::POOL_RSS) it will also be destroyed.
247		520
248	=cut	521	=cut
249		522
250	our $POOL_SIZE = 8;	523	our $POOL_SIZE = 8;
251	our $POOL_RSS = 16 * 1024;	524	our $POOL_RSS = 32 * 1024;
252	our @async_pool;	525	our @async_pool;
253		526
254	sub pool_handler {	527	sub pool_handler {
255	my $cb;
256
257	while () {	528	while () {
258	eval {	529	eval {
259	while () {	530	&{&_pool_handler} while 1;
260	_pool_1 $cb;
261	&$cb;
262	_pool_2 $cb;
263	&schedule;
264	}
265	};	531	};
266		532
267	last if $@ eq "\3async_pool terminate\2\n";
268	warn $@ if $@;	533	warn $@ if $@;
269	}	534	}
270	}	535	}
271		536
272	sub async_pool(&@) {	537	=back
273	# this is also inlined into the unlock_scheduler
274	my $coro = (pop @async_pool) || new Coro \&pool_handler;
275		538
276	$coro->{_invoke} = [@_];	539	=head1 STATIC METHODS
277	$coro->ready;
278		540
279	$coro	541	Static methods are actually functions that implicitly operate on the
280	}	542	current coro.
		543
		544	=over 4
281		545
282	=item schedule	546	=item schedule
283		547
284	Calls the scheduler. Please note that the current coroutine will not be put	548	Calls the scheduler. The scheduler will find the next coro that is
		549	to be run from the ready queue and switches to it. The next coro
		550	to be run is simply the one with the highest priority that is longest
		551	in its ready queue. If there is no coro ready, it will call the
		552	C<$Coro::idle> hook.
		553
		554	Please note that the current coro will I<not> be put into the ready
285	into the ready queue, so calling this function usually means you will	555	queue, so calling this function usually means you will never be called
286	never be called again unless something else (e.g. an event handler) calls	556	again unless something else (e.g. an event handler) calls C<< ->ready >>,
287	ready.	557	thus waking you up.
288		558
289	The canonical way to wait on external events is this:	559	This makes C<schedule> I<the> generic method to use to block the current
		560	coro and wait for events: first you remember the current coro in
		561	a variable, then arrange for some callback of yours to call C<< ->ready
		562	>> on that once some event happens, and last you call C<schedule> to put
		563	yourself to sleep. Note that a lot of things can wake your coro up,
		564	so you need to check whether the event indeed happened, e.g. by storing the
		565	status in a variable.
290		566
291	{	567	See B<HOW TO WAIT FOR A CALLBACK>, below, for some ways to wait for callbacks.
292	# remember current coroutine
293	my $current = $Coro::current;
294		568
295	# register a hypothetical event handler	569	=item cede
296	on_event_invoke sub {	570
297	# wake up sleeping coroutine	571	"Cede" to other coros. This function puts the current coro into
298	$current->ready;	572	the ready queue and calls C<schedule>, which has the effect of giving
299	undef $current;	573	up the current "timeslice" to other coros of the same or higher
		574	priority. Once your coro gets its turn again it will automatically be
		575	resumed.
		576
		577	This function is often called C<yield> in other languages.
		578
		579	=item Coro::cede_notself
		580
		581	Works like cede, but is not exported by default and will cede to I<any>
		582	coro, regardless of priority. This is useful sometimes to ensure
		583	progress is made.
		584
		585	=item terminate [arg...]
		586
		587	Terminates the current coro with the given status values (see
		588	L<cancel>). The values will not be copied, but referenced directly.
		589
		590	=item Coro::on_enter BLOCK, Coro::on_leave BLOCK
		591
		592	These function install enter and leave winders in the current scope. The
		593	enter block will be executed when on_enter is called and whenever the
		594	current coro is re-entered by the scheduler, while the leave block is
		595	executed whenever the current coro is blocked by the scheduler, and
		596	also when the containing scope is exited (by whatever means, be it exit,
		597	die, last etc.).
		598
		599	I<Neither invoking the scheduler, nor exceptions, are allowed within those
		600	BLOCKs>. That means: do not even think about calling C<die> without an
		601	eval, and do not even think of entering the scheduler in any way.
		602
		603	Since both BLOCKs are tied to the current scope, they will automatically
		604	be removed when the current scope exits.
		605
		606	These functions implement the same concept as C<dynamic-wind> in scheme
		607	does, and are useful when you want to localise some resource to a specific
		608	coro.
		609
		610	They slow down thread switching considerably for coros that use them
		611	(about 40% for a BLOCK with a single assignment, so thread switching is
		612	still reasonably fast if the handlers are fast).
		613
		614	These functions are best understood by an example: The following function
		615	will change the current timezone to "Antarctica/South_Pole", which
		616	requires a call to C<tzset>, but by using C<on_enter> and C<on_leave>,
		617	which remember/change the current timezone and restore the previous
		618	value, respectively, the timezone is only changed for the coro that
		619	installed those handlers.
		620
		621	use POSIX qw(tzset);
		622
		623	async {
		624	my $old_tz; # store outside TZ value here
		625
		626	Coro::on_enter {
		627	$old_tz = $ENV{TZ}; # remember the old value
		628
		629	$ENV{TZ} = "Antarctica/South_Pole";
		630	tzset; # enable new value
300	};	631	};
301		632
302	# call schedule until event occurred.	633	Coro::on_leave {
303	# in case we are woken up for other reasons	634	$ENV{TZ} = $old_tz;
304	# (current still defined), loop.	635	tzset; # restore old value
305	Coro::schedule while $current;	636	};
		637
		638	# at this place, the timezone is Antarctica/South_Pole,
		639	# without disturbing the TZ of any other coro.
		640	};
		641
		642	This can be used to localise about any resource (locale, uid, current
		643	working directory etc.) to a block, despite the existance of other
		644	coros.
		645
		646	Another interesting example implements time-sliced multitasking using
		647	interval timers (this could obviously be optimised, but does the job):
		648
		649	# "timeslice" the given block
		650	sub timeslice(&) {
		651	use Time::HiRes ();
		652
		653	Coro::on_enter {
		654	# on entering the thread, we set an VTALRM handler to cede
		655	$SIG{VTALRM} = sub { cede };
		656	# and then start the interval timer
		657	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0.01, 0.01;
		658	};
		659	Coro::on_leave {
		660	# on leaving the thread, we stop the interval timer again
		661	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0, 0;
		662	};
		663
		664	&{+shift};
306	}	665	}
307		666
308	=item cede	667	# use like this:
		668	timeslice {
		669	# The following is an endless loop that would normally
		670	# monopolise the process. Since it runs in a timesliced
		671	# environment, it will regularly cede to other threads.
		672	while () { }
		673	};
309		674
310	"Cede" to other coroutines. This function puts the current coroutine into the
311	ready queue and calls C<schedule>, which has the effect of giving up the
312	current "timeslice" to other coroutines of the same or higher priority.
313
314	=item Coro::cede_notself
315
316	Works like cede, but is not exported by default and will cede to any
317	coroutine, regardless of priority, once.
318
319	=item terminate [arg...]
320
321	Terminates the current coroutine with the given status values (see L<cancel>).
322		675
323	=item killall	676	=item killall
324		677
325	Kills/terminates/cancels all coroutines except the currently running	678	Kills/terminates/cancels all coros except the currently running one.
326	one. This is useful after a fork, either in the child or the parent, as	679
327	usually only one of them should inherit the running coroutines.	680	Note that while this will try to free some of the main interpreter
		681	resources if the calling coro isn't the main coro, but one
		682	cannot free all of them, so if a coro that is not the main coro
		683	calls this function, there will be some one-time resource leak.
328		684
329	=cut	685	=cut
330
331	sub terminate {
332	$current->cancel (@_);
333	}
334		686
335	sub killall {	687	sub killall {
336	for (Coro::State::list) {	688	for (Coro::State::list) {
337	$_->cancel	689	$_->cancel
338	if $_ != $current && UNIVERSAL::isa $_, "Coro";	690	if $_ != $current && UNIVERSAL::isa $_, "Coro";
339	}	691	}
340	}	692	}
341		693
342	=back	694	=back
343		695
344	# dynamic methods
345
346	=head2 COROUTINE METHODS	696	=head1 CORO OBJECT METHODS
347		697
348	These are the methods you can call on coroutine objects.	698	These are the methods you can call on coro objects (or to create
		699	them).
349		700
350	=over 4	701	=over 4
351		702
352	=item new Coro \&sub [, @args...]	703	=item new Coro \&sub [, @args...]
353		704
354	Create a new coroutine and return it. When the sub returns the coroutine	705	Create a new coro and return it. When the sub returns, the coro
355	automatically terminates as if C<terminate> with the returned values were	706	automatically terminates as if C<terminate> with the returned values were
356	called. To make the coroutine run you must first put it into the ready queue	707	called. To make the coro run you must first put it into the ready
357	by calling the ready method.	708	queue by calling the ready method.
358		709
359	See C<async> and C<Coro::State::new> for additional info about the	710	See C<async> and C<Coro::State::new> for additional info about the
360	coroutine environment.	711	coro environment.
361		712
362	=cut	713	=cut
363		714
364	sub _run_coro {	715	sub _coro_run {
365	terminate &{+shift};	716	terminate &{+shift};
366	}	717	}
367		718
368	sub new {
369	my $class = shift;
370
371	$class->SUPER::new (\&_run_coro, @_)
372	}
373
374	=item $success = $coroutine->ready	719	=item $success = $coro->ready
375		720
376	Put the given coroutine into the ready queue (according to it's priority)	721	Put the given coro into the end of its ready queue (there is one
377	and return true. If the coroutine is already in the ready queue, do nothing	722	queue for each priority) and return true. If the coro is already in
378	and return false.	723	the ready queue, do nothing and return false.
379		724
		725	This ensures that the scheduler will resume this coro automatically
		726	once all the coro of higher priority and all coro of the same
		727	priority that were put into the ready queue earlier have been resumed.
		728
		729	=item $coro->suspend
		730
		731	Suspends the specified coro. A suspended coro works just like any other
		732	coro, except that the scheduler will not select a suspended coro for
		733	execution.
		734
		735	Suspending a coro can be useful when you want to keep the coro from
		736	running, but you don't want to destroy it, or when you want to temporarily
		737	freeze a coro (e.g. for debugging) to resume it later.
		738
		739	A scenario for the former would be to suspend all (other) coros after a
		740	fork and keep them alive, so their destructors aren't called, but new
		741	coros can be created.
		742
		743	=item $coro->resume
		744
		745	If the specified coro was suspended, it will be resumed. Note that when
		746	the coro was in the ready queue when it was suspended, it might have been
		747	unreadied by the scheduler, so an activation might have been lost.
		748
		749	To avoid this, it is best to put a suspended coro into the ready queue
		750	unconditionally, as every synchronisation mechanism must protect itself
		751	against spurious wakeups, and the one in the Coro family certainly do
		752	that.
		753
		754	=item $state->is_new
		755
		756	Returns true iff this Coro object is "new", i.e. has never been run
		757	yet. Those states basically consist of only the code reference to call and
		758	the arguments, but consumes very little other resources. New states will
		759	automatically get assigned a perl interpreter when they are transfered to.
		760
		761	=item $state->is_zombie
		762
		763	Returns true iff the Coro object has been cancelled, i.e.
		764	it's resources freed because they were C<cancel>'ed, C<terminate>'d,
		765	C<safe_cancel>'ed or simply went out of scope.
		766
		767	The name "zombie" stems from UNIX culture, where a process that has
		768	exited and only stores and exit status and no other resources is called a
		769	"zombie".
		770
380	=item $is_ready = $coroutine->is_ready	771	=item $is_ready = $coro->is_ready
381		772
382	Return wether the coroutine is currently the ready queue or not,	773	Returns true iff the Coro object is in the ready queue. Unless the Coro
		774	object gets destroyed, it will eventually be scheduled by the scheduler.
383		775
		776	=item $is_running = $coro->is_running
		777
		778	Returns true iff the Coro object is currently running. Only one Coro object
		779	can ever be in the running state (but it currently is possible to have
		780	multiple running Coro::States).
		781
		782	=item $is_suspended = $coro->is_suspended
		783
		784	Returns true iff this Coro object has been suspended. Suspended Coros will
		785	not ever be scheduled.
		786
384	=item $coroutine->cancel (arg...)	787	=item $coro->cancel (arg...)
385		788
386	Terminates the given coroutine and makes it return the given arguments as	789	Terminates the given Coro thread and makes it return the given arguments as
387	status (default: the empty list). Never returns if the coroutine is the	790	status (default: an empty list). Never returns if the Coro is the
388	current coroutine.	791	current Coro.
389		792
390	=cut	793	This is a rather brutal way to free a coro, with some limitations - if
		794	the thread is inside a C callback that doesn't expect to be canceled,
		795	bad things can happen, or if the cancelled thread insists on running
		796	complicated cleanup handlers that rely on its thread context, things will
		797	not work.
391		798
392	sub cancel {	799	Any cleanup code being run (e.g. from C<guard> blocks, destructors and so
393	my $self = shift;	800	on) will be run without a thread context, and is not allowed to switch
394	$self->{_status} = [@_];	801	to other threads. A common mistake is to call C<< ->cancel >> from a
		802	destructor called by die'ing inside the thread to be cancelled for
		803	example.
395		804
396	if ($current == $self) {	805	On the plus side, C<< ->cancel >> will always clean up the thread, no
397	push @destroy, $self;	806	matter what. If your cleanup code is complex or you want to avoid
398	$manager->ready;	807	cancelling a C-thread that doesn't know how to clean up itself, it can be
399	&schedule while 1;	808	better to C<< ->throw >> an exception, or use C<< ->safe_cancel >>.
400	} else {	809
401	$self->_cancel;	810	The arguments to C<< ->cancel >> are not copied, but instead will
		811	be referenced directly (e.g. if you pass C<$var> and after the call
		812	change that variable, then you might change the return values passed to
		813	e.g. C<join>, so don't do that).
		814
		815	The resources of the Coro are usually freed (or destructed) before this
		816	call returns, but this can be delayed for an indefinite amount of time, as
		817	in some cases the manager thread has to run first to actually destruct the
		818	Coro object.
		819
		820	=item $coro->safe_cancel ($arg...)
		821
		822	Works mostly like C<< ->cancel >>, but is inherently "safer", and
		823	consequently, can fail with an exception in cases the thread is not in a
		824	cancellable state. Essentially, C<< ->safe_cancel >> is a C<< ->cancel >>
		825	with extra checks before canceling.
		826
		827	It works a bit like throwing an exception that cannot be caught -
		828	specifically, it will clean up the thread from within itself, so all
		829	cleanup handlers (e.g. C<guard> blocks) are run with full thread
		830	context and can block if they wish. The downside is that there is no
		831	guarantee that the thread can be cancelled when you call this method, and
		832	therefore, it might fail. It is also considerably slower than C<cancel> or
		833	C<terminate>.
		834
		835	A thread is in a safe-cancellable state if it either hasn't been run yet,
		836	or it has no C context attached and is inside an SLF function.
		837
		838	The latter two basically mean that the thread isn't currently inside a
		839	perl callback called from some C function (usually via some XS modules)
		840	and isn't currently executing inside some C function itself (via Coro's XS
		841	API).
		842
		843	This call returns true when it could cancel the thread, or croaks with an
		844	error otherwise (i.e. it either returns true or doesn't return at all).
		845
		846	Why the weird interface? Well, there are two common models on how and
		847	when to cancel things. In the first, you have the expectation that your
		848	coro thread can be cancelled when you want to cancel it - if the thread
		849	isn't cancellable, this would be a bug somewhere, so C<< ->safe_cancel >>
		850	croaks to notify of the bug.
		851
		852	In the second model you sometimes want to ask nicely to cancel a thread,
		853	but if it's not a good time, well, then don't cancel. This can be done
		854	relatively easy like this:
		855
		856	if (! eval { $coro->safe_cancel }) {
		857	warn "unable to cancel thread: $@";
402	}	858	}
403	}
404		859
405	=item $coroutine->join	860	However, what you never should do is first try to cancel "safely" and
		861	if that fails, cancel the "hard" way with C<< ->cancel >>. That makes
		862	no sense: either you rely on being able to execute cleanup code in your
		863	thread context, or you don't. If you do, then C<< ->safe_cancel >> is the
		864	only way, and if you don't, then C<< ->cancel >> is always faster and more
		865	direct.
406		866
407	Wait until the coroutine terminates and return any values given to the	867	=item $coro->schedule_to
408	C<terminate> or C<cancel> functions. C<join> can be called concurrently
409	from multiple coroutines.
410		868
411	=cut	869	Puts the current coro to sleep (like C<Coro::schedule>), but instead
		870	of continuing with the next coro from the ready queue, always switch to
		871	the given coro object (regardless of priority etc.). The readyness
		872	state of that coro isn't changed.
412		873
413	sub join {	874	This is an advanced method for special cases - I'd love to hear about any
414	my $self = shift;	875	uses for this one.
415		876
416	unless ($self->{_status}) {	877	=item $coro->cede_to
417	my $current = $current;
418		878
419	push @{$self->{_on_destroy}}, sub {	879	Like C<schedule_to>, but puts the current coro into the ready
420	$current->ready;	880	queue. This has the effect of temporarily switching to the given
421	undef $current;	881	coro, and continuing some time later.
422	};
423		882
424	&schedule while $current;	883	This is an advanced method for special cases - I'd love to hear about any
425	}	884	uses for this one.
426		885
427	wantarray ? @{$self->{_status}} : $self->{_status}[0];
428	}
429
430	=item $coroutine->on_destroy (\&cb)
431
432	Registers a callback that is called when this coroutine gets destroyed,
433	but before it is joined. The callback gets passed the terminate arguments,
434	if any.
435
436	=cut
437
438	sub on_destroy {
439	my ($self, $cb) = @_;
440
441	push @{ $self->{_on_destroy} }, $cb;
442	}
443
444	=item $oldprio = $coroutine->prio ($newprio)
445
446	Sets (or gets, if the argument is missing) the priority of the
447	coroutine. Higher priority coroutines get run before lower priority
448	coroutines. Priorities are small signed integers (currently -4 .. +3),
449	that you can refer to using PRIO_xxx constants (use the import tag :prio
450	to get then):
451
452	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN
453	3 > 1 > 0 > -1 > -3 > -4
454
455	# set priority to HIGH
456	current->prio(PRIO_HIGH);
457
458	The idle coroutine ($Coro::idle) always has a lower priority than any
459	existing coroutine.
460
461	Changing the priority of the current coroutine will take effect immediately,
462	but changing the priority of coroutines in the ready queue (but not
463	running) will only take effect after the next schedule (of that
464	coroutine). This is a bug that will be fixed in some future version.
465
466	=item $newprio = $coroutine->nice ($change)
467
468	Similar to C<prio>, but subtract the given value from the priority (i.e.
469	higher values mean lower priority, just as in unix).
470
471	=item $olddesc = $coroutine->desc ($newdesc)
472
473	Sets (or gets in case the argument is missing) the description for this
474	coroutine. This is just a free-form string you can associate with a coroutine.
475
476	This method simply sets the C<< $coroutine->{desc} >> member to the given string. You
477	can modify this member directly if you wish.
478
479	=item $coroutine->throw ([$scalar])	886	=item $coro->throw ([$scalar])
480		887
481	If C<$throw> is specified and defined, it will be thrown as an exception	888	If C<$throw> is specified and defined, it will be thrown as an exception
482	inside the coroutine at the next convinient point in time (usually after	889	inside the coro at the next convenient point in time. Otherwise
483	it gains control at the next schedule/transfer/cede). Otherwise clears the
484	exception object.	890	clears the exception object.
		891
		892	Coro will check for the exception each time a schedule-like-function
		893	returns, i.e. after each C<schedule>, C<cede>, C<< Coro::Semaphore->down
		894	>>, C<< Coro::Handle->readable >> and so on. Most of those functions (all
		895	that are part of Coro itself) detect this case and return early in case an
		896	exception is pending.
485		897
486	The exception object will be thrown "as is" with the specified scalar in	898	The exception object will be thrown "as is" with the specified scalar in
487	C<$@>, i.e. if it is a string, no line number or newline will be appended	899	C<$@>, i.e. if it is a string, no line number or newline will be appended
488	(unlike with C<die>).	900	(unlike with C<die>).
489		901
490	This can be used as a softer means than C<cancel> to ask a coroutine to	902	This can be used as a softer means than either C<cancel> or C<safe_cancel
491	end itself, although there is no guarentee that the exception will lead to	903	>to ask a coro to end itself, although there is no guarantee that the
492	termination, and if the exception isn't caught it might well end the whole	904	exception will lead to termination, and if the exception isn't caught it
493	program.	905	might well end the whole program.
		906
		907	You might also think of C<throw> as being the moral equivalent of
		908	C<kill>ing a coro with a signal (in this case, a scalar).
		909
		910	=item $coro->join
		911
		912	Wait until the coro terminates and return any values given to the
		913	C<terminate> or C<cancel> functions. C<join> can be called concurrently
		914	from multiple threads, and all will be resumed and given the status
		915	return once the C<$coro> terminates.
		916
		917	=item $coro->on_destroy (\&cb)
		918
		919	Registers a callback that is called when this coro thread gets destroyed,
		920	that is, after it's resources have been freed but before it is joined. The
		921	callback gets passed the terminate/cancel arguments, if any, and I<must
		922	not> die, under any circumstances.
		923
		924	There can be any number of C<on_destroy> callbacks per coro, and there is
		925	currently no way to remove a callback once added.
		926
		927	=item $oldprio = $coro->prio ($newprio)
		928
		929	Sets (or gets, if the argument is missing) the priority of the
		930	coro thread. Higher priority coro get run before lower priority
		931	coros. Priorities are small signed integers (currently -4 .. +3),
		932	that you can refer to using PRIO_xxx constants (use the import tag :prio
		933	to get then):
		934
		935	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN
		936	3 > 1 > 0 > -1 > -3 > -4
		937
		938	# set priority to HIGH
		939	current->prio (PRIO_HIGH);
		940
		941	The idle coro thread ($Coro::idle) always has a lower priority than any
		942	existing coro.
		943
		944	Changing the priority of the current coro will take effect immediately,
		945	but changing the priority of a coro in the ready queue (but not running)
		946	will only take effect after the next schedule (of that coro). This is a
		947	bug that will be fixed in some future version.
		948
		949	=item $newprio = $coro->nice ($change)
		950
		951	Similar to C<prio>, but subtract the given value from the priority (i.e.
		952	higher values mean lower priority, just as in UNIX's nice command).
		953
		954	=item $olddesc = $coro->desc ($newdesc)
		955
		956	Sets (or gets in case the argument is missing) the description for this
		957	coro thread. This is just a free-form string you can associate with a
		958	coro.
		959
		960	This method simply sets the C<< $coro->{desc} >> member to the given
		961	string. You can modify this member directly if you wish, and in fact, this
		962	is often preferred to indicate major processing states that can then be
		963	seen for example in a L<Coro::Debug> session:
		964
		965	sub my_long_function {
		966	local $Coro::current->{desc} = "now in my_long_function";
		967	...
		968	$Coro::current->{desc} = "my_long_function: phase 1";
		969	...
		970	$Coro::current->{desc} = "my_long_function: phase 2";
		971	...
		972	}
494		973
495	=cut	974	=cut
496		975
497	sub desc {	976	sub desc {
498	my $old = $_[0]{desc};	977	my $old = $_[0]{desc};
499	$_[0]{desc} = $_[1] if @_ > 1;	978	$_[0]{desc} = $_[1] if @_ > 1;
500	$old;	979	$old;
501	}	980	}
502		981
		982	sub transfer {
		983	require Carp;
		984	Carp::croak ("You must not call ->transfer on Coro objects. Use Coro::State objects or the ->schedule_to method. Caught");
		985	}
		986
503	=back	987	=back
504		988
505	=head2 GLOBAL FUNCTIONS	989	=head1 GLOBAL FUNCTIONS
506		990
507	=over 4	991	=over 4
508		992
509	=item Coro::nready	993	=item Coro::nready
510		994
511	Returns the number of coroutines that are currently in the ready state,	995	Returns the number of coro that are currently in the ready state,
512	i.e. that can be switched to. The value C<0> means that the only runnable	996	i.e. that can be switched to by calling C<schedule> directory or
		997	indirectly. The value C<0> means that the only runnable coro is the
513	coroutine is the currently running one, so C<cede> would have no effect,	998	currently running one, so C<cede> would have no effect, and C<schedule>
514	and C<schedule> would cause a deadlock unless there is an idle handler	999	would cause a deadlock unless there is an idle handler that wakes up some
515	that wakes up some coroutines.	1000	coro.
516		1001
517	=item my $guard = Coro::guard { ... }	1002	=item my $guard = Coro::guard { ... }
518		1003
519	This creates and returns a guard object. Nothing happens until the object	1004	This function still exists, but is deprecated. Please use the
520	gets destroyed, in which case the codeblock given as argument will be	1005	C<Guard::guard> function instead.
521	executed. This is useful to free locks or other resources in case of a
522	runtime error or when the coroutine gets canceled, as in both cases the
523	guard block will be executed. The guard object supports only one method,
524	C<< ->cancel >>, which will keep the codeblock from being executed.
525
526	Example: set some flag and clear it again when the coroutine gets canceled
527	or the function returns:
528
529	sub do_something {
530	my $guard = Coro::guard { $busy = 0 };
531	$busy = 1;
532
533	# do something that requires $busy to be true
534	}
535		1006
536	=cut	1007	=cut
537		1008
538	sub guard(&) {	1009	BEGIN { *guard = \&Guard::guard }
539	bless \(my $cb = $_[0]), "Coro::guard"
540	}
541
542	sub Coro::guard::cancel {
543	${$_[0]} = sub { };
544	}
545
546	sub Coro::guard::DESTROY {
547	${$_[0]}->();
548	}
549
550		1010
551	=item unblock_sub { ... }	1011	=item unblock_sub { ... }
552		1012
553	This utility function takes a BLOCK or code reference and "unblocks" it,	1013	This utility function takes a BLOCK or code reference and "unblocks" it,
554	returning the new coderef. This means that the new coderef will return	1014	returning a new coderef. Unblocking means that calling the new coderef
555	immediately without blocking, returning nothing, while the original code	1015	will return immediately without blocking, returning nothing, while the
556	ref will be called (with parameters) from within its own coroutine.	1016	original code ref will be called (with parameters) from within another
		1017	coro.
557		1018
558	The reason this function exists is that many event libraries (such as the	1019	The reason this function exists is that many event libraries (such as
559	venerable L<Event|Event> module) are not coroutine-safe (a weaker form	1020	the venerable L<Event|Event> module) are not thread-safe (a weaker form
560	of thread-safety). This means you must not block within event callbacks,	1021	of reentrancy). This means you must not block within event callbacks,
561	otherwise you might suffer from crashes or worse.	1022	otherwise you might suffer from crashes or worse. The only event library
		1023	currently known that is safe to use without C<unblock_sub> is L<EV> (but
		1024	you might still run into deadlocks if all event loops are blocked).
		1025
		1026	Coro will try to catch you when you block in the event loop
		1027	("FATAL: $Coro::idle blocked itself"), but this is just best effort and
		1028	only works when you do not run your own event loop.
562		1029
563	This function allows your callbacks to block by executing them in another	1030	This function allows your callbacks to block by executing them in another
564	coroutine where it is safe to block. One example where blocking is handy	1031	coro where it is safe to block. One example where blocking is handy
565	is when you use the L<Coro::AIO|Coro::AIO> functions to save results to	1032	is when you use the L<Coro::AIO|Coro::AIO> functions to save results to
566	disk.	1033	disk, for example.
567		1034
568	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when	1035	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when
569	creating event callbacks that want to block.	1036	creating event callbacks that want to block.
		1037
		1038	If your handler does not plan to block (e.g. simply sends a message to
		1039	another coro, or puts some other coro into the ready queue), there is
		1040	no reason to use C<unblock_sub>.
		1041
		1042	Note that you also need to use C<unblock_sub> for any other callbacks that
		1043	are indirectly executed by any C-based event loop. For example, when you
		1044	use a module that uses L<AnyEvent> (and you use L<Coro::AnyEvent>) and it
		1045	provides callbacks that are the result of some event callback, then you
		1046	must not block either, or use C<unblock_sub>.
570		1047
571	=cut	1048	=cut
572		1049
573	our @unblock_queue;	1050	our @unblock_queue;
574		1051
…		…
577	# return immediately and can be reused) and because we cannot cede	1054	# return immediately and can be reused) and because we cannot cede
578	# inside an event callback.	1055	# inside an event callback.
579	our $unblock_scheduler = new Coro sub {	1056	our $unblock_scheduler = new Coro sub {
580	while () {	1057	while () {
581	while (my $cb = pop @unblock_queue) {	1058	while (my $cb = pop @unblock_queue) {
582	# this is an inlined copy of async_pool	1059	&async_pool (@$cb);
583	my $coro = (pop @async_pool) || new Coro \&pool_handler;
584		1060
585	$coro->{_invoke} = $cb;
586	$coro->ready;
587	cede; # for short-lived callbacks, this reduces pressure on the coro pool	1061	# for short-lived callbacks, this reduces pressure on the coro pool
		1062	# as the chance is very high that the async_poll coro will be back
		1063	# in the idle state when cede returns
		1064	cede;
588	}	1065	}
589	schedule; # sleep well	1066	schedule; # sleep well
590	}	1067	}
591	};	1068	};
592	$unblock_scheduler->desc ("[unblock_sub scheduler]");	1069	$unblock_scheduler->{desc} = "[unblock_sub scheduler]";
593		1070
594	sub unblock_sub(&) {	1071	sub unblock_sub(&) {
595	my $cb = shift;	1072	my $cb = shift;
596		1073
597	sub {	1074	sub {
598	unshift @unblock_queue, [$cb, @_];	1075	unshift @unblock_queue, [$cb, @_];
599	$unblock_scheduler->ready;	1076	$unblock_scheduler->ready;
600	}	1077	}
601	}	1078	}
602		1079
		1080	=item $cb = rouse_cb
		1081
		1082	Create and return a "rouse callback". That's a code reference that,
		1083	when called, will remember a copy of its arguments and notify the owner
		1084	coro of the callback.
		1085
		1086	See the next function.
		1087
		1088	=item @args = rouse_wait [$cb]
		1089
		1090	Wait for the specified rouse callback (or the last one that was created in
		1091	this coro).
		1092
		1093	As soon as the callback is invoked (or when the callback was invoked
		1094	before C<rouse_wait>), it will return the arguments originally passed to
		1095	the rouse callback. In scalar context, that means you get the I<last>
		1096	argument, just as if C<rouse_wait> had a C<return ($a1, $a2, $a3...)>
		1097	statement at the end.
		1098
		1099	See the section B<HOW TO WAIT FOR A CALLBACK> for an actual usage example.
		1100
603	=back	1101	=back
604		1102
605	=cut	1103	=cut
606		1104
		1105	for my $module (qw(Channel RWLock Semaphore SemaphoreSet Signal Specific)) {
		1106	my $old = defined &{"Coro::$module\::new"} && \&{"Coro::$module\::new"};
		1107
		1108	*{"Coro::$module\::new"} = sub {
		1109	require "Coro/$module.pm";
		1110
		1111	# some modules have their new predefined in State.xs, some don't
		1112	*{"Coro::$module\::new"} = $old
		1113	if $old;
		1114
		1115	goto &{"Coro::$module\::new"};
		1116	};
		1117	}
		1118
607	1;	1119	1;
608		1120
		1121	=head1 HOW TO WAIT FOR A CALLBACK
		1122
		1123	It is very common for a coro to wait for some callback to be
		1124	called. This occurs naturally when you use coro in an otherwise
		1125	event-based program, or when you use event-based libraries.
		1126
		1127	These typically register a callback for some event, and call that callback
		1128	when the event occured. In a coro, however, you typically want to
		1129	just wait for the event, simplyifying things.
		1130
		1131	For example C<< AnyEvent->child >> registers a callback to be called when
		1132	a specific child has exited:
		1133
		1134	my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... });
		1135
		1136	But from within a coro, you often just want to write this:
		1137
		1138	my $status = wait_for_child $pid;
		1139
		1140	Coro offers two functions specifically designed to make this easy,
		1141	C<rouse_cb> and C<rouse_wait>.
		1142
		1143	The first function, C<rouse_cb>, generates and returns a callback that,
		1144	when invoked, will save its arguments and notify the coro that
		1145	created the callback.
		1146
		1147	The second function, C<rouse_wait>, waits for the callback to be called
		1148	(by calling C<schedule> to go to sleep) and returns the arguments
		1149	originally passed to the callback.
		1150
		1151	Using these functions, it becomes easy to write the C<wait_for_child>
		1152	function mentioned above:
		1153
		1154	sub wait_for_child($) {
		1155	my ($pid) = @_;
		1156
		1157	my $watcher = AnyEvent->child (pid => $pid, cb => rouse_cb);
		1158
		1159	my ($rpid, $rstatus) = rouse_wait;
		1160	$rstatus
		1161	}
		1162
		1163	In the case where C<rouse_cb> and C<rouse_wait> are not flexible enough,
		1164	you can roll your own, using C<schedule> and C<ready>:
		1165
		1166	sub wait_for_child($) {
		1167	my ($pid) = @_;
		1168
		1169	# store the current coro in $current,
		1170	# and provide result variables for the closure passed to ->child
		1171	my $current = $Coro::current;
		1172	my ($done, $rstatus);
		1173
		1174	# pass a closure to ->child
		1175	my $watcher = AnyEvent->child (pid => $pid, cb => sub {
		1176	$rstatus = $_[1]; # remember rstatus
		1177	$done = 1; # mark $rstatus as valid
		1178	$current->ready; # wake up the waiting thread
		1179	});
		1180
		1181	# wait until the closure has been called
		1182	schedule while !$done;
		1183
		1184	$rstatus
		1185	}
		1186
		1187
609	=head1 BUGS/LIMITATIONS	1188	=head1 BUGS/LIMITATIONS
610		1189
611	- you must make very sure that no coro is still active on global	1190	=over 4
612	destruction. very bad things might happen otherwise (usually segfaults).
613		1191
		1192	=item fork with pthread backend
		1193
		1194	When Coro is compiled using the pthread backend (which isn't recommended
		1195	but required on many BSDs as their libcs are completely broken), then
		1196	coro will not survive a fork. There is no known workaround except to
		1197	fix your libc and use a saner backend.
		1198
		1199	=item perl process emulation ("threads")
		1200
614	- this module is not thread-safe. You should only ever use this module	1201	This module is not perl-pseudo-thread-safe. You should only ever use this
615	from the same thread (this requirement might be loosened in the future	1202	module from the first thread (this requirement might be removed in the
616	to allow per-thread schedulers, but Coro::State does not yet allow	1203	future to allow per-thread schedulers, but Coro::State does not yet allow
617	this).	1204	this). I recommend disabling thread support and using processes, as having
		1205	the windows process emulation enabled under unix roughly halves perl
		1206	performance, even when not used.
		1207
		1208	Attempts to use threads created in another emulated process will crash
		1209	("cleanly", with a null pointer exception).
		1210
		1211	=item coro switching is not signal safe
		1212
		1213	You must not switch to another coro from within a signal handler (only
		1214	relevant with %SIG - most event libraries provide safe signals), I<unless>
		1215	you are sure you are not interrupting a Coro function.
		1216
		1217	That means you I<MUST NOT> call any function that might "block" the
		1218	current coro - C<cede>, C<schedule> C<< Coro::Semaphore->down >> or
		1219	anything that calls those. Everything else, including calling C<ready>,
		1220	works.
		1221
		1222	=back
		1223
		1224
		1225	=head1 WINDOWS PROCESS EMULATION
		1226
		1227	A great many people seem to be confused about ithreads (for example, Chip
		1228	Salzenberg called me unintelligent, incapable, stupid and gullible,
		1229	while in the same mail making rather confused statements about perl
		1230	ithreads (for example, that memory or files would be shared), showing his
		1231	lack of understanding of this area - if it is hard to understand for Chip,
		1232	it is probably not obvious to everybody).
		1233
		1234	What follows is an ultra-condensed version of my talk about threads in
		1235	scripting languages given on the perl workshop 2009:
		1236
		1237	The so-called "ithreads" were originally implemented for two reasons:
		1238	first, to (badly) emulate unix processes on native win32 perls, and
		1239	secondly, to replace the older, real thread model ("5.005-threads").
		1240
		1241	It does that by using threads instead of OS processes. The difference
		1242	between processes and threads is that threads share memory (and other
		1243	state, such as files) between threads within a single process, while
		1244	processes do not share anything (at least not semantically). That
		1245	means that modifications done by one thread are seen by others, while
		1246	modifications by one process are not seen by other processes.
		1247
		1248	The "ithreads" work exactly like that: when creating a new ithreads
		1249	process, all state is copied (memory is copied physically, files and code
		1250	is copied logically). Afterwards, it isolates all modifications. On UNIX,
		1251	the same behaviour can be achieved by using operating system processes,
		1252	except that UNIX typically uses hardware built into the system to do this
		1253	efficiently, while the windows process emulation emulates this hardware in
		1254	software (rather efficiently, but of course it is still much slower than
		1255	dedicated hardware).
		1256
		1257	As mentioned before, loading code, modifying code, modifying data
		1258	structures and so on is only visible in the ithreads process doing the
		1259	modification, not in other ithread processes within the same OS process.
		1260
		1261	This is why "ithreads" do not implement threads for perl at all, only
		1262	processes. What makes it so bad is that on non-windows platforms, you can
		1263	actually take advantage of custom hardware for this purpose (as evidenced
		1264	by the forks module, which gives you the (i-) threads API, just much
		1265	faster).
		1266
		1267	Sharing data is in the i-threads model is done by transfering data
		1268	structures between threads using copying semantics, which is very slow -
		1269	shared data simply does not exist. Benchmarks using i-threads which are
		1270	communication-intensive show extremely bad behaviour with i-threads (in
		1271	fact, so bad that Coro, which cannot take direct advantage of multiple
		1272	CPUs, is often orders of magnitude faster because it shares data using
		1273	real threads, refer to my talk for details).
		1274
		1275	As summary, i-threads *use* threads to implement processes, while
		1276	the compatible forks module *uses* processes to emulate, uhm,
		1277	processes. I-threads slow down every perl program when enabled, and
		1278	outside of windows, serve no (or little) practical purpose, but
		1279	disadvantages every single-threaded Perl program.
		1280
		1281	This is the reason that I try to avoid the name "ithreads", as it is
		1282	misleading as it implies that it implements some kind of thread model for
		1283	perl, and prefer the name "windows process emulation", which describes the
		1284	actual use and behaviour of it much better.
618		1285
619	=head1 SEE ALSO	1286	=head1 SEE ALSO
620		1287
621	Lower level Configuration, Coroutine Environment: L<Coro::State>.	1288	Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.
622		1289
623	Debugging: L<Coro::Debug>.	1290	Debugging: L<Coro::Debug>.
624		1291
625	Support/Utility: L<Coro::Specific>, L<Coro::Util>.	1292	Support/Utility: L<Coro::Specific>, L<Coro::Util>.
626		1293
627	Locking/IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>, L<Coro::SemaphoreSet>, L<Coro::RWLock>.	1294	Locking and IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>,
		1295	L<Coro::SemaphoreSet>, L<Coro::RWLock>.
628		1296
629	Event/IO: L<Coro::Timer>, L<Coro::Event>, L<Coro::Handle>, L<Coro::Socket>.	1297	I/O and Timers: L<Coro::Timer>, L<Coro::Handle>, L<Coro::Socket>, L<Coro::AIO>.
630		1298
631	Compatibility: L<Coro::LWP>, L<Coro::Storable>, L<Coro::Select>.	1299	Compatibility with other modules: L<Coro::LWP> (but see also L<AnyEvent::HTTP> for
		1300	a better-working alternative), L<Coro::BDB>, L<Coro::Storable>,
		1301	L<Coro::Select>.
632		1302
633	Embedding: L<Coro::MakeMaker>.	1303	XS API: L<Coro::MakeMaker>.
634		1304
635	=head1 AUTHOR	1305	Low level Configuration, Thread Environment, Continuations: L<Coro::State>.
636		1306
		1307	=head1 AUTHOR/SUPPORT/CONTACT
		1308
637	Marc Lehmann <schmorp@schmorp.de>	1309	Marc A. Lehmann <schmorp@schmorp.de>
638	http://home.schmorp.de/	1310	http://software.schmorp.de/pkg/Coro.html
639		1311
640	=cut	1312	=cut
641		1313

Diff Legend

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