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

Comparing Coro/Coro.pm (file contents):
Revision 1.104 by root, Thu Jan 4 23:49:27 2007 UTC vs.
Revision 1.269 by root, Thu Oct 1 23:25:03 2009 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";
		12	cede; # yield back to main
		13	print "4\n";
11	};	14	};
12		15	print "1\n";
13	# alternatively create an async coroutine like this:	16	cede; # yield to coro
14		17	print "3\n";
15	sub some_func : Coro {	18	cede; # and again
16	# some more async code	19
17	}	20	# use locking
18		21	use Coro::Semaphore;
19	cede;	22	my $lock = new Coro::Semaphore;
		23	my $locked;
		24
		25	$lock->down;
		26	$locked = 1;
		27	$lock->up;
20		28
21	=head1 DESCRIPTION	29	=head1 DESCRIPTION
22		30
23	This module collection manages coroutines. Coroutines are similar	31	For a tutorial-style introduction, please read the L<Coro::Intro>
24	to threads but don't run in parallel at the same time even on SMP	32	manpage. This manpage mainly contains reference information.
25	machines. The specific flavor of coroutine use din this module also
26	guarentees you that it will not switch between coroutines unless
27	necessary, at easily-identified points in your program, so locking and
28	parallel access are rarely an issue, making coroutine programming much
29	safer than threads programming.
30		33
31	(Perl, however, does not natively support real threads but instead does a	34	This module collection manages continuations in general, most often in
32	very slow and memory-intensive emulation of processes using threads. This	35	the form of cooperative threads (also called coros, or simply "coro"
33	is a performance win on Windows machines, and a loss everywhere else).	36	in the documentation). They are similar to kernel threads but don't (in
		37	general) run in parallel at the same time even on SMP machines. The
		38	specific flavor of thread offered by this module also guarantees you that
		39	it will not switch between threads unless necessary, at easily-identified
		40	points in your program, so locking and parallel access are rarely an
		41	issue, making thread programming much safer and easier than using other
		42	thread models.
34		43
		44	Unlike the so-called "Perl threads" (which are not actually real threads
		45	but only the windows process emulation (see section of same name for more
		46	details) ported to unix, and as such act as processes), Coro provides
		47	a full shared address space, which makes communication between threads
		48	very easy. And Coro's threads are fast, too: disabling the Windows
		49	process emulation code in your perl and using Coro can easily result in
		50	a two to four times speed increase for your programs. A parallel matrix
		51	multiplication benchmark runs over 300 times faster on a single core than
		52	perl's pseudo-threads on a quad core 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
35	In this module, coroutines are defined as "callchain + lexical variables +	60	In this module, a thread is defined as "callchain + lexical variables +
36	@_ + $_ + $@ + $/ + C stack), that is, a coroutine has its own callchain,	61	some package variables + C stack), that is, a thread has its own callchain,
37	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
38	variables.	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.
39		67
40	=cut	68	=cut
41		69
42	package Coro;	70	package Coro;
43		71
44	use strict;	72	use common::sense;
45	no warnings "uninitialized";	73
		74	use Carp ();
		75
		76	use Guard ();
46		77
47	use Coro::State;	78	use Coro::State;
48		79
49	use base qw(Coro::State Exporter);	80	use base qw(Coro::State Exporter);
50		81
51	our $idle; # idle handler	82	our $idle; # idle handler
52	our $main; # main coroutine	83	our $main; # main coro
53	our $current; # current coroutine	84	our $current; # current coro
54		85
55	our $VERSION = '3.3';	86	our $VERSION = 5.17;
56		87
57	our @EXPORT = qw(async cede schedule terminate current unblock_sub);	88	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub);
58	our %EXPORT_TAGS = (	89	our %EXPORT_TAGS = (
59	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],	90	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],
60	);	91	);
61	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));	92	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));
62		93
63	{	94	=head1 GLOBAL VARIABLES
64	my @async;
65	my $init;
66
67	# this way of handling attributes simply is NOT scalable ;()
68	sub import {
69	no strict 'refs';
70
71	Coro->export_to_level (1, @_);
72
73	my $old = *{(caller)[0]."::MODIFY_CODE_ATTRIBUTES"}{CODE};
74	*{(caller)[0]."::MODIFY_CODE_ATTRIBUTES"} = sub {
75	my ($package, $ref) = (shift, shift);
76	my @attrs;
77	for (@_) {
78	if ($_ eq "Coro") {
79	push @async, $ref;
80	unless ($init++) {
81	eval q{
82	sub INIT {
83	&async(pop @async) while @async;
84	}
85	};
86	}
87	} else {
88	push @attrs, $_;
89	}
90	}
91	return $old ? $old->($package, $ref, @attrs) : @attrs;
92	};
93	}
94
95	}
96		95
97	=over 4	96	=over 4
98		97
99	=item $main	98	=item $Coro::main
100		99
101	This coroutine represents the main program.	100	This variable stores the Coro object that represents the main
		101	program. While you cna C<ready> it and do most other things you can do to
		102	coro, it is mainly useful to compare again C<$Coro::current>, to see
		103	whether you are running in the main program or not.
102		104
103	=cut	105	=cut
104		106
105	$main = new Coro;	107	# $main is now being initialised by Coro::State
106		108
107	=item $current (or as function: current)	109	=item $Coro::current
108		110
109	The current coroutine (the last coroutine switched to). The initial value	111	The Coro object representing the current coro (the last
		112	coro that the Coro scheduler switched to). The initial value is
110	is C<$main> (of course).	113	C<$Coro::main> (of course).
111		114
112	This variable is B<strictly> I<read-only>. It is provided for performance	115	This variable is B<strictly> I<read-only>. You can take copies of the
113	reasons. If performance is not essentiel you are encouraged to use the	116	value stored in it and use it as any other Coro object, but you must
114	C<Coro::current> function instead.	117	not otherwise modify the variable itself.
115		118
116	=cut	119	=cut
117		120
118	# maybe some other module used Coro::Specific before...
119	$main->{specific} = $current->{specific}
120	if $current;
121
122	_set_current $main;
123
124	sub current() { $current }	121	sub current() { $current } # [DEPRECATED]
125		122
126	=item $idle	123	=item $Coro::idle
127		124
		125	This variable is mainly useful to integrate Coro into event loops. It is
		126	usually better to rely on L<Coro::AnyEvent> or L<Coro::EV>, as this is
		127	pretty low-level functionality.
		128
		129	This variable stores either a Coro object or a callback.
		130
128	A callback that is called whenever the scheduler finds no ready coroutines	131	If it is a callback, the it is called whenever the scheduler finds no
129	to run. The default implementation prints "FATAL: deadlock detected" and	132	ready coros to run. The default implementation prints "FATAL:
130	exits, because the program has no other way to continue.	133	deadlock detected" and exits, because the program has no other way to
		134	continue.
131		135
		136	If it is a coro object, then this object will be readied (without
		137	invoking any ready hooks, however) when the scheduler finds no other ready
		138	coros to run.
		139
132	This hook is overwritten by modules such as C<Coro::Timer> and	140	This hook is overwritten by modules such as C<Coro::EV> and
133	C<Coro::Event> to wait on an external event that hopefully wake up a	141	C<Coro::AnyEvent> to wait on an external event that hopefully wake up a
134	coroutine so the scheduler can run it.	142	coro so the scheduler can run it.
		143
		144	Note that the callback I<must not>, under any circumstances, block
		145	the current coro. Normally, this is achieved by having an "idle
		146	coro" that calls the event loop and then blocks again, and then
		147	readying that coro in the idle handler, or by simply placing the idle
		148	coro in this variable.
		149
		150	See L<Coro::Event> or L<Coro::AnyEvent> for examples of using this
		151	technique.
135		152
136	Please note that if your callback recursively invokes perl (e.g. for event	153	Please note that if your callback recursively invokes perl (e.g. for event
137	handlers), then it must be prepared to be called recursively.	154	handlers), then it must be prepared to be called recursively itself.
138		155
139	=cut	156	=cut
140		157
141	$idle = sub {	158	$idle = sub {
142	require Carp;
143	Carp::croak ("FATAL: deadlock detected");	159	Carp::confess ("FATAL: deadlock detected");
144	};	160	};
145		161
146	sub _cancel {
147	my ($self) = @_;
148
149	# free coroutine data and mark as destructed
150	$self->_destroy
151	or return;
152
153	# call all destruction callbacks
154	$_->(@{$self->{status}})
155	for @{(delete $self->{destroy_cb}) \|\| []};
156	}
157
158	# this coroutine is necessary because a coroutine	162	# this coro is necessary because a coro
159	# cannot destroy itself.	163	# cannot destroy itself.
160	my @destroy;	164	our @destroy;
161	my $manager;	165	our $manager;
162		166
163	$manager = new Coro sub {	167	$manager = new Coro sub {
164	while () {	168	while () {
165	(shift @destroy)->_cancel	169	Coro::State::cancel shift @destroy
166	while @destroy;	170	while @destroy;
167		171
168	&schedule;	172	&schedule;
169	}	173	}
170	};	174	};
171		175	$manager->{desc} = "[coro manager]";
172	$manager->prio (PRIO_MAX);	176	$manager->prio (PRIO_MAX);
173		177
174	# static methods. not really.
175
176	=back	178	=back
177		179
178	=head2 STATIC METHODS	180	=head1 SIMPLE CORO CREATION
179
180	Static methods are actually functions that operate on the current coroutine only.
181		181
182	=over 4	182	=over 4
183		183
184	=item async { ... } [@args...]	184	=item async { ... } [@args...]
185		185
186	Create a new asynchronous coroutine and return it's coroutine object	186	Create a new coro and return its Coro object (usually
187	(usually unused). When the sub returns the new coroutine is automatically	187	unused). The coro will be put into the ready queue, so
		188	it will start running automatically on the next scheduler run.
		189
		190	The first argument is a codeblock/closure that should be executed in the
		191	coro. When it returns argument returns the coro is automatically
188	terminated.	192	terminated.
189		193
190	Calling C<exit> in a coroutine will not work correctly, so do not do that.	194	The remaining arguments are passed as arguments to the closure.
191		195
192	When the coroutine dies, the program will exit, just as in the main	196	See the C<Coro::State::new> constructor for info about the coro
193	program.	197	environment in which coro are executed.
194		198
		199	Calling C<exit> in a coro will do the same as calling exit outside
		200	the coro. Likewise, when the coro dies, the program will exit,
		201	just as it would in the main program.
		202
		203	If you do not want that, you can provide a default C<die> handler, or
		204	simply avoid dieing (by use of C<eval>).
		205
195	# create a new coroutine that just prints its arguments	206	Example: Create a new coro that just prints its arguments.
		207
196	async {	208	async {
197	print "@_\n";	209	print "@_\n";
198	} 1,2,3,4;	210	} 1,2,3,4;
199		211
200	=cut	212	=item async_pool { ... } [@args...]
201		213
202	sub async(&@) {	214	Similar to C<async>, but uses a coro pool, so you should not call
203	my $coro = new Coro @_;	215	terminate or join on it (although you are allowed to), and you get a
204	$coro->ready;	216	coro that might have executed other code already (which can be good
205	$coro	217	or bad :).
		218
		219	On the plus side, this function is about twice as fast as creating (and
		220	destroying) a completely new coro, so if you need a lot of generic
		221	coros in quick successsion, use C<async_pool>, not C<async>.
		222
		223	The code block is executed in an C<eval> context and a warning will be
		224	issued in case of an exception instead of terminating the program, as
		225	C<async> does. As the coro is being reused, stuff like C<on_destroy>
		226	will not work in the expected way, unless you call terminate or cancel,
		227	which somehow defeats the purpose of pooling (but is fine in the
		228	exceptional case).
		229
		230	The priority will be reset to C<0> after each run, tracing will be
		231	disabled, the description will be reset and the default output filehandle
		232	gets restored, so you can change all these. Otherwise the coro will
		233	be re-used "as-is": most notably if you change other per-coro global
		234	stuff such as C<$/> you I<must needs> revert that change, which is most
		235	simply done by using local as in: C<< local $/ >>.
		236
		237	The idle pool size is limited to C<8> idle coros (this can be
		238	adjusted by changing $Coro::POOL_SIZE), but there can be as many non-idle
		239	coros as required.
		240
		241	If you are concerned about pooled coros growing a lot because a
		242	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool
		243	{ terminate }> once per second or so to slowly replenish the pool. In
		244	addition to that, when the stacks used by a handler grows larger than 32kb
		245	(adjustable via $Coro::POOL_RSS) it will also be destroyed.
		246
		247	=cut
		248
		249	our $POOL_SIZE = 8;
		250	our $POOL_RSS = 32 * 1024;
		251	our @async_pool;
		252
		253	sub pool_handler {
		254	while () {
		255	eval {
		256	&{&_pool_handler} while 1;
		257	};
		258
		259	warn $@ if $@;
		260	}
206	}	261	}
207		262
		263	=back
		264
		265	=head1 STATIC METHODS
		266
		267	Static methods are actually functions that implicitly operate on the
		268	current coro.
		269
		270	=over 4
		271
208	=item schedule	272	=item schedule
209		273
210	Calls the scheduler. Please note that the current coroutine will not be put	274	Calls the scheduler. The scheduler will find the next coro that is
		275	to be run from the ready queue and switches to it. The next coro
		276	to be run is simply the one with the highest priority that is longest
		277	in its ready queue. If there is no coro ready, it will clal the
		278	C<$Coro::idle> hook.
		279
		280	Please note that the current coro will I<not> be put into the ready
211	into the ready queue, so calling this function usually means you will	281	queue, so calling this function usually means you will never be called
212	never be called again unless something else (e.g. an event handler) calls	282	again unless something else (e.g. an event handler) calls C<< ->ready >>,
213	ready.	283	thus waking you up.
214		284
215	The canonical way to wait on external events is this:	285	This makes C<schedule> I<the> generic method to use to block the current
		286	coro and wait for events: first you remember the current coro in
		287	a variable, then arrange for some callback of yours to call C<< ->ready
		288	>> on that once some event happens, and last you call C<schedule> to put
		289	yourself to sleep. Note that a lot of things can wake your coro up,
		290	so you need to check whether the event indeed happened, e.g. by storing the
		291	status in a variable.
216		292
		293	See B<HOW TO WAIT FOR A CALLBACK>, below, for some ways to wait for callbacks.
		294
		295	=item cede
		296
		297	"Cede" to other coros. This function puts the current coro into
		298	the ready queue and calls C<schedule>, which has the effect of giving
		299	up the current "timeslice" to other coros of the same or higher
		300	priority. Once your coro gets its turn again it will automatically be
		301	resumed.
		302
		303	This function is often called C<yield> in other languages.
		304
		305	=item Coro::cede_notself
		306
		307	Works like cede, but is not exported by default and will cede to I<any>
		308	coro, regardless of priority. This is useful sometimes to ensure
		309	progress is made.
		310
		311	=item terminate [arg...]
		312
		313	Terminates the current coro with the given status values (see L<cancel>).
		314
		315	=item Coro::on_enter BLOCK, Coro::on_leave BLOCK
		316
		317	These function install enter and leave winders in the current scope. The
		318	enter block will be executed when on_enter is called and whenever the
		319	current coro is re-entered by the scheduler, while the leave block is
		320	executed whenever the current coro is blocked by the scheduler, and
		321	also when the containing scope is exited (by whatever means, be it exit,
		322	die, last etc.).
		323
		324	I<Neither invoking the scheduler, nor exceptions, are allowed within those
		325	BLOCKs>. That means: do not even think about calling C<die> without an
		326	eval, and do not even think of entering the scheduler in any way.
		327
		328	Since both BLOCKs are tied to the current scope, they will automatically
		329	be removed when the current scope exits.
		330
		331	These functions implement the same concept as C<dynamic-wind> in scheme
		332	does, and are useful when you want to localise some resource to a specific
		333	coro.
		334
		335	They slow down thread switching considerably for coros that use them
		336	(about 40% for a BLOCK with a single assignment, so thread switching is
		337	still reasonably fast if the handlers are fast).
		338
		339	These functions are best understood by an example: The following function
		340	will change the current timezone to "Antarctica/South_Pole", which
		341	requires a call to C<tzset>, but by using C<on_enter> and C<on_leave>,
		342	which remember/change the current timezone and restore the previous
		343	value, respectively, the timezone is only changed for the coro that
		344	installed those handlers.
		345
		346	use POSIX qw(tzset);
		347
		348	async {
		349	my $old_tz; # store outside TZ value here
		350
		351	Coro::on_enter {
		352	$old_tz = $ENV{TZ}; # remember the old value
		353
		354	$ENV{TZ} = "Antarctica/South_Pole";
		355	tzset; # enable new value
		356	};
		357
		358	Coro::on_leave {
		359	$ENV{TZ} = $old_tz;
		360	tzset; # restore old value
		361	};
		362
		363	# at this place, the timezone is Antarctica/South_Pole,
		364	# without disturbing the TZ of any other coro.
		365	};
		366
		367	This can be used to localise about any resource (locale, uid, current
		368	working directory etc.) to a block, despite the existance of other
		369	coros.
		370
		371	Another interesting example implements time-sliced multitasking using
		372	interval timers (this could obviously be optimised, but does the job):
		373
		374	# "timeslice" the given block
		375	sub timeslice(&) {
		376	use Time::HiRes ();
		377
		378	Coro::on_enter {
		379	# on entering the thread, we set an VTALRM handler to cede
		380	$SIG{VTALRM} = sub { cede };
		381	# and then start the interval timer
		382	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0.01, 0.01;
		383	};
		384	Coro::on_leave {
		385	# on leaving the thread, we stop the interval timer again
		386	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0, 0;
		387	};
		388
		389	&{+shift};
		390	}
		391
		392	# use like this:
		393	timeslice {
		394	# The following is an endless loop that would normally
		395	# monopolise the process. Since it runs in a timesliced
		396	# environment, it will regularly cede to other threads.
		397	while () { }
		398	};
		399
		400
		401	=item killall
		402
		403	Kills/terminates/cancels all coros except the currently running one.
		404
		405	Note that while this will try to free some of the main interpreter
		406	resources if the calling coro isn't the main coro, but one
		407	cannot free all of them, so if a coro that is not the main coro
		408	calls this function, there will be some one-time resource leak.
		409
		410	=cut
		411
		412	sub killall {
		413	for (Coro::State::list) {
		414	$_->cancel
		415	if $_ != $current && UNIVERSAL::isa $_, "Coro";
217	{	416	}
218	# remember current coroutine	417	}
		418
		419	=back
		420
		421	=head1 CORO OBJECT METHODS
		422
		423	These are the methods you can call on coro objects (or to create
		424	them).
		425
		426	=over 4
		427
		428	=item new Coro \&sub [, @args...]
		429
		430	Create a new coro and return it. When the sub returns, the coro
		431	automatically terminates as if C<terminate> with the returned values were
		432	called. To make the coro run you must first put it into the ready
		433	queue by calling the ready method.
		434
		435	See C<async> and C<Coro::State::new> for additional info about the
		436	coro environment.
		437
		438	=cut
		439
		440	sub _coro_run {
		441	terminate &{+shift};
		442	}
		443
		444	=item $success = $coro->ready
		445
		446	Put the given coro into the end of its ready queue (there is one
		447	queue for each priority) and return true. If the coro is already in
		448	the ready queue, do nothing and return false.
		449
		450	This ensures that the scheduler will resume this coro automatically
		451	once all the coro of higher priority and all coro of the same
		452	priority that were put into the ready queue earlier have been resumed.
		453
		454	=item $coro->suspend
		455
		456	Suspends the specified coro. A suspended coro works just like any other
		457	coro, except that the scheduler will not select a suspended coro for
		458	execution.
		459
		460	Suspending a coro can be useful when you want to keep the coro from
		461	running, but you don't want to destroy it, or when you want to temporarily
		462	freeze a coro (e.g. for debugging) to resume it later.
		463
		464	A scenario for the former would be to suspend all (other) coros after a
		465	fork and keep them alive, so their destructors aren't called, but new
		466	coros can be created.
		467
		468	=item $coro->resume
		469
		470	If the specified coro was suspended, it will be resumed. Note that when
		471	the coro was in the ready queue when it was suspended, it might have been
		472	unreadied by the scheduler, so an activation might have been lost.
		473
		474	To avoid this, it is best to put a suspended coro into the ready queue
		475	unconditionally, as every synchronisation mechanism must protect itself
		476	against spurious wakeups, and the one in the Coro family certainly do
		477	that.
		478
		479	=item $is_ready = $coro->is_ready
		480
		481	Returns true iff the Coro object is in the ready queue. Unless the Coro
		482	object gets destroyed, it will eventually be scheduled by the scheduler.
		483
		484	=item $is_running = $coro->is_running
		485
		486	Returns true iff the Coro object is currently running. Only one Coro object
		487	can ever be in the running state (but it currently is possible to have
		488	multiple running Coro::States).
		489
		490	=item $is_suspended = $coro->is_suspended
		491
		492	Returns true iff this Coro object has been suspended. Suspended Coros will
		493	not ever be scheduled.
		494
		495	=item $coro->cancel (arg...)
		496
		497	Terminates the given Coro and makes it return the given arguments as
		498	status (default: the empty list). Never returns if the Coro is the
		499	current Coro.
		500
		501	=cut
		502
		503	sub cancel {
		504	my $self = shift;
		505
		506	if ($current == $self) {
		507	terminate @_;
		508	} else {
		509	$self->{_status} = [@_];
		510	Coro::State::cancel $self;
		511	}
		512	}
		513
		514	=item $coro->schedule_to
		515
		516	Puts the current coro to sleep (like C<Coro::schedule>), but instead
		517	of continuing with the next coro from the ready queue, always switch to
		518	the given coro object (regardless of priority etc.). The readyness
		519	state of that coro isn't changed.
		520
		521	This is an advanced method for special cases - I'd love to hear about any
		522	uses for this one.
		523
		524	=item $coro->cede_to
		525
		526	Like C<schedule_to>, but puts the current coro into the ready
		527	queue. This has the effect of temporarily switching to the given
		528	coro, and continuing some time later.
		529
		530	This is an advanced method for special cases - I'd love to hear about any
		531	uses for this one.
		532
		533	=item $coro->throw ([$scalar])
		534
		535	If C<$throw> is specified and defined, it will be thrown as an exception
		536	inside the coro at the next convenient point in time. Otherwise
		537	clears the exception object.
		538
		539	Coro will check for the exception each time a schedule-like-function
		540	returns, i.e. after each C<schedule>, C<cede>, C<< Coro::Semaphore->down
		541	>>, C<< Coro::Handle->readable >> and so on. Most of these functions
		542	detect this case and return early in case an exception is pending.
		543
		544	The exception object will be thrown "as is" with the specified scalar in
		545	C<$@>, i.e. if it is a string, no line number or newline will be appended
		546	(unlike with C<die>).
		547
		548	This can be used as a softer means than C<cancel> to ask a coro to
		549	end itself, although there is no guarantee that the exception will lead to
		550	termination, and if the exception isn't caught it might well end the whole
		551	program.
		552
		553	You might also think of C<throw> as being the moral equivalent of
		554	C<kill>ing a coro with a signal (in this case, a scalar).
		555
		556	=item $coro->join
		557
		558	Wait until the coro terminates and return any values given to the
		559	C<terminate> or C<cancel> functions. C<join> can be called concurrently
		560	from multiple coro, and all will be resumed and given the status
		561	return once the C<$coro> terminates.
		562
		563	=cut
		564
		565	sub join {
		566	my $self = shift;
		567
		568	unless ($self->{_status}) {
219	my $current = $Coro::current;	569	my $current = $current;
220		570
221	# register a hypothetical event handler	571	push @{$self->{_on_destroy}}, sub {
222	on_event_invoke sub {
223	# wake up sleeping coroutine
224	$current->ready;	572	$current->ready;
225	undef $current;	573	undef $current;
226	};	574	};
227		575
228	# call schedule until event occured.
229	# in case we are woken up for other reasons
230	# (current still defined), loop.
231	Coro::schedule while $current;
232	}
233
234	=item cede
235
236	"Cede" to other coroutines. This function puts the current coroutine into the
237	ready queue and calls C<schedule>, which has the effect of giving up the
238	current "timeslice" to other coroutines of the same or higher priority.
239
240	=item Coro::cede_notself
241
242	Works like cede, but is not exported by default and will cede to any
243	coroutine, regardless of priority, once.
244
245	=item terminate [arg...]
246
247	Terminates the current coroutine with the given status values (see L<cancel>).
248
249	=cut
250
251	sub terminate {
252	$current->cancel (@_);
253	}
254
255	=back
256
257	# dynamic methods
258
259	=head2 COROUTINE METHODS
260
261	These are the methods you can call on coroutine objects.
262
263	=over 4
264
265	=item new Coro \&sub [, @args...]
266
267	Create a new coroutine and return it. When the sub returns the coroutine
268	automatically terminates as if C<terminate> with the returned values were
269	called. To make the coroutine run you must first put it into the ready queue
270	by calling the ready method.
271
272	Calling C<exit> in a coroutine will not work correctly, so do not do that.
273
274	=cut
275
276	sub _run_coro {
277	terminate &{+shift};
278	}
279
280	sub new {
281	my $class = shift;
282
283	$class->SUPER::new (\&_run_coro, @_)
284	}
285
286	=item $success = $coroutine->ready
287
288	Put the given coroutine into the ready queue (according to it's priority)
289	and return true. If the coroutine is already in the ready queue, do nothing
290	and return false.
291
292	=item $is_ready = $coroutine->is_ready
293
294	Return wether the coroutine is currently the ready queue or not,
295
296	=item $coroutine->cancel (arg...)
297
298	Terminates the given coroutine and makes it return the given arguments as
299	status (default: the empty list). Never returns if the coroutine is the
300	current coroutine.
301
302	=cut
303
304	sub cancel {
305	my $self = shift;
306	$self->{status} = [@_];
307
308	if ($current == $self) {
309	push @destroy, $self;
310	$manager->ready;
311	&schedule while 1;
312	} else {
313	$self->_cancel;
314	}
315	}
316
317	=item $coroutine->join
318
319	Wait until the coroutine terminates and return any values given to the
320	C<terminate> or C<cancel> functions. C<join> can be called multiple times
321	from multiple coroutine.
322
323	=cut
324
325	sub join {
326	my $self = shift;
327
328	unless ($self->{status}) {
329	my $current = $current;
330
331	push @{$self->{destroy_cb}}, sub {
332	$current->ready;
333	undef $current;
334	};
335
336	&schedule while $current;	576	&schedule while $current;
337	}	577	}
338		578
339	wantarray ? @{$self->{status}} : $self->{status}[0];	579	wantarray ? @{$self->{_status}} : $self->{_status}[0];
340	}	580	}
341		581
342	=item $coroutine->on_destroy (\&cb)	582	=item $coro->on_destroy (\&cb)
343		583
344	Registers a callback that is called when this coroutine gets destroyed,	584	Registers a callback that is called when this coro gets destroyed,
345	but before it is joined. The callback gets passed the terminate arguments,	585	but before it is joined. The callback gets passed the terminate arguments,
346	if any.	586	if any, and I<must not> die, under any circumstances.
347		587
348	=cut	588	=cut
349		589
350	sub on_destroy {	590	sub on_destroy {
351	my ($self, $cb) = @_;	591	my ($self, $cb) = @_;
352		592
353	push @{ $self->{destroy_cb} }, $cb;	593	push @{ $self->{_on_destroy} }, $cb;
354	}	594	}
355		595
356	=item $oldprio = $coroutine->prio ($newprio)	596	=item $oldprio = $coro->prio ($newprio)
357		597
358	Sets (or gets, if the argument is missing) the priority of the	598	Sets (or gets, if the argument is missing) the priority of the
359	coroutine. Higher priority coroutines get run before lower priority	599	coro. Higher priority coro get run before lower priority
360	coroutines. Priorities are small signed integers (currently -4 .. +3),	600	coro. Priorities are small signed integers (currently -4 .. +3),
361	that you can refer to using PRIO_xxx constants (use the import tag :prio	601	that you can refer to using PRIO_xxx constants (use the import tag :prio
362	to get then):	602	to get then):
363		603
364	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN	604	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN
365	3 > 1 > 0 > -1 > -3 > -4	605	3 > 1 > 0 > -1 > -3 > -4
366		606
367	# set priority to HIGH	607	# set priority to HIGH
368	current->prio(PRIO_HIGH);	608	current->prio (PRIO_HIGH);
369		609
370	The idle coroutine ($Coro::idle) always has a lower priority than any	610	The idle coro ($Coro::idle) always has a lower priority than any
371	existing coroutine.	611	existing coro.
372		612
373	Changing the priority of the current coroutine will take effect immediately,	613	Changing the priority of the current coro will take effect immediately,
374	but changing the priority of coroutines in the ready queue (but not	614	but changing the priority of coro in the ready queue (but not
375	running) will only take effect after the next schedule (of that	615	running) will only take effect after the next schedule (of that
376	coroutine). This is a bug that will be fixed in some future version.	616	coro). This is a bug that will be fixed in some future version.
377		617
378	=item $newprio = $coroutine->nice ($change)	618	=item $newprio = $coro->nice ($change)
379		619
380	Similar to C<prio>, but subtract the given value from the priority (i.e.	620	Similar to C<prio>, but subtract the given value from the priority (i.e.
381	higher values mean lower priority, just as in unix).	621	higher values mean lower priority, just as in unix).
382		622
383	=item $olddesc = $coroutine->desc ($newdesc)	623	=item $olddesc = $coro->desc ($newdesc)
384		624
385	Sets (or gets in case the argument is missing) the description for this	625	Sets (or gets in case the argument is missing) the description for this
386	coroutine. This is just a free-form string you can associate with a coroutine.	626	coro. This is just a free-form string you can associate with a
		627	coro.
		628
		629	This method simply sets the C<< $coro->{desc} >> member to the given
		630	string. You can modify this member directly if you wish.
387		631
388	=cut	632	=cut
389		633
390	sub desc {	634	sub desc {
391	my $old = $_[0]{desc};	635	my $old = $_[0]{desc};
392	$_[0]{desc} = $_[1] if @_ > 1;	636	$_[0]{desc} = $_[1] if @_ > 1;
393	$old;	637	$old;
394	}	638	}
395		639
		640	sub transfer {
		641	require Carp;
		642	Carp::croak ("You must not call ->transfer on Coro objects. Use Coro::State objects or the ->schedule_to method. Caught");
		643	}
		644
396	=back	645	=back
397		646
398	=head2 GLOBAL FUNCTIONS	647	=head1 GLOBAL FUNCTIONS
399		648
400	=over 4	649	=over 4
401		650
402	=item Coro::nready	651	=item Coro::nready
403		652
404	Returns the number of coroutines that are currently in the ready state,	653	Returns the number of coro that are currently in the ready state,
405	i.e. that can be swicthed to. The value C<0> means that the only runnable	654	i.e. that can be switched to by calling C<schedule> directory or
		655	indirectly. The value C<0> means that the only runnable coro is the
406	coroutine is the currently running one, so C<cede> would have no effect,	656	currently running one, so C<cede> would have no effect, and C<schedule>
407	and C<schedule> would cause a deadlock unless there is an idle handler	657	would cause a deadlock unless there is an idle handler that wakes up some
408	that wakes up some coroutines.	658	coro.
409		659
410	=item my $guard = Coro::guard { ... }	660	=item my $guard = Coro::guard { ... }
411		661
412	This creates and returns a guard object. Nothing happens until the objetc	662	This function still exists, but is deprecated. Please use the
413	gets destroyed, in which case the codeblock given as argument will be	663	C<Guard::guard> function instead.
414	executed. This is useful to free locks or other resources in case of a
415	runtime error or when the coroutine gets canceled, as in both cases the
416	guard block will be executed. The guard object supports only one method,
417	C<< ->cancel >>, which will keep the codeblock from being executed.
418		664
419	Example: set some flag and clear it again when the coroutine gets canceled
420	or the function returns:
421
422	sub do_something {
423	my $guard = Coro::guard { $busy = 0 };
424	$busy = 1;
425
426	# do something that requires $busy to be true
427	}
428
429	=cut	665	=cut
430		666
431	sub guard(&) {	667	BEGIN { *guard = \&Guard::guard }
432	bless \(my $cb = $_[0]), "Coro::guard"
433	}
434
435	sub Coro::guard::cancel {
436	${$_[0]} = sub { };
437	}
438
439	sub Coro::guard::DESTROY {
440	${$_[0]}->();
441	}
442
443		668
444	=item unblock_sub { ... }	669	=item unblock_sub { ... }
445		670
446	This utility function takes a BLOCK or code reference and "unblocks" it,	671	This utility function takes a BLOCK or code reference and "unblocks" it,
447	returning the new coderef. This means that the new coderef will return	672	returning a new coderef. Unblocking means that calling the new coderef
448	immediately without blocking, returning nothing, while the original code	673	will return immediately without blocking, returning nothing, while the
449	ref will be called (with parameters) from within its own coroutine.	674	original code ref will be called (with parameters) from within another
		675	coro.
450		676
451	The reason this fucntion exists is that many event libraries (such as the	677	The reason this function exists is that many event libraries (such as the
452	venerable L<Event\|Event> module) are not coroutine-safe (a weaker form	678	venerable L<Event\|Event> module) are not thread-safe (a weaker form
453	of thread-safety). This means you must not block within event callbacks,	679	of reentrancy). This means you must not block within event callbacks,
454	otherwise you might suffer from crashes or worse.	680	otherwise you might suffer from crashes or worse. The only event library
		681	currently known that is safe to use without C<unblock_sub> is L<EV>.
455		682
456	This function allows your callbacks to block by executing them in another	683	This function allows your callbacks to block by executing them in another
457	coroutine where it is safe to block. One example where blocking is handy	684	coro where it is safe to block. One example where blocking is handy
458	is when you use the L<Coro::AIO\|Coro::AIO> functions to save results to	685	is when you use the L<Coro::AIO\|Coro::AIO> functions to save results to
459	disk.	686	disk, for example.
460		687
461	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when	688	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when
462	creating event callbacks that want to block.	689	creating event callbacks that want to block.
463		690
464	=cut	691	If your handler does not plan to block (e.g. simply sends a message to
		692	another coro, or puts some other coro into the ready queue), there is
		693	no reason to use C<unblock_sub>.
465		694
466	our @unblock_pool;	695	Note that you also need to use C<unblock_sub> for any other callbacks that
		696	are indirectly executed by any C-based event loop. For example, when you
		697	use a module that uses L<AnyEvent> (and you use L<Coro::AnyEvent>) and it
		698	provides callbacks that are the result of some event callback, then you
		699	must not block either, or use C<unblock_sub>.
		700
		701	=cut
		702
467	our @unblock_queue;	703	our @unblock_queue;
468	our $UNBLOCK_POOL_SIZE = 2;
469		704
470	sub unblock_handler_ {	705	# we create a special coro because we want to cede,
471	while () {	706	# to reduce pressure on the coro pool (because most callbacks
472	my ($cb, @arg) = @{ delete $Coro::current->{arg} };	707	# return immediately and can be reused) and because we cannot cede
473	$cb->(@arg);	708	# inside an event callback.
474
475	last if @unblock_pool >= $UNBLOCK_POOL_SIZE;
476	push @unblock_pool, $Coro::current;
477	schedule;
478	}
479	}
480
481	our $unblock_scheduler = async {	709	our $unblock_scheduler = new Coro sub {
482	while () {	710	while () {
483	while (my $cb = pop @unblock_queue) {	711	while (my $cb = pop @unblock_queue) {
484	my $handler = (pop @unblock_pool or new Coro \&unblock_handler_);	712	&async_pool (@$cb);
485	$handler->{arg} = $cb;	713
486	$handler->ready;	714	# for short-lived callbacks, this reduces pressure on the coro pool
		715	# as the chance is very high that the async_poll coro will be back
		716	# in the idle state when cede returns
487	cede;	717	cede;
488	}	718	}
489		719	schedule; # sleep well
490	schedule;
491	}	720	}
492	};	721	};
		722	$unblock_scheduler->{desc} = "[unblock_sub scheduler]";
493		723
494	sub unblock_sub(&) {	724	sub unblock_sub(&) {
495	my $cb = shift;	725	my $cb = shift;
496		726
497	sub {	727	sub {
498	push @unblock_queue, [$cb, @_];	728	unshift @unblock_queue, [$cb, @_];
499	$unblock_scheduler->ready;	729	$unblock_scheduler->ready;
500	}	730	}
501	}	731	}
502		732
		733	=item $cb = Coro::rouse_cb
		734
		735	Create and return a "rouse callback". That's a code reference that,
		736	when called, will remember a copy of its arguments and notify the owner
		737	coro of the callback.
		738
		739	See the next function.
		740
		741	=item @args = Coro::rouse_wait [$cb]
		742
		743	Wait for the specified rouse callback (or the last one that was created in
		744	this coro).
		745
		746	As soon as the callback is invoked (or when the callback was invoked
		747	before C<rouse_wait>), it will return the arguments originally passed to
		748	the rouse callback. In scalar context, that means you get the I<last>
		749	argument, just as if C<rouse_wait> had a C<return ($a1, $a2, $a3...)>
		750	statement at the end.
		751
		752	See the section B<HOW TO WAIT FOR A CALLBACK> for an actual usage example.
		753
503	=back	754	=back
504		755
505	=cut	756	=cut
506		757
507	1;	758	1;
508		759
		760	=head1 HOW TO WAIT FOR A CALLBACK
		761
		762	It is very common for a coro to wait for some callback to be
		763	called. This occurs naturally when you use coro in an otherwise
		764	event-based program, or when you use event-based libraries.
		765
		766	These typically register a callback for some event, and call that callback
		767	when the event occured. In a coro, however, you typically want to
		768	just wait for the event, simplyifying things.
		769
		770	For example C<< AnyEvent->child >> registers a callback to be called when
		771	a specific child has exited:
		772
		773	my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... });
		774
		775	But from within a coro, you often just want to write this:
		776
		777	my $status = wait_for_child $pid;
		778
		779	Coro offers two functions specifically designed to make this easy,
		780	C<Coro::rouse_cb> and C<Coro::rouse_wait>.
		781
		782	The first function, C<rouse_cb>, generates and returns a callback that,
		783	when invoked, will save its arguments and notify the coro that
		784	created the callback.
		785
		786	The second function, C<rouse_wait>, waits for the callback to be called
		787	(by calling C<schedule> to go to sleep) and returns the arguments
		788	originally passed to the callback.
		789
		790	Using these functions, it becomes easy to write the C<wait_for_child>
		791	function mentioned above:
		792
		793	sub wait_for_child($) {
		794	my ($pid) = @_;
		795
		796	my $watcher = AnyEvent->child (pid => $pid, cb => Coro::rouse_cb);
		797
		798	my ($rpid, $rstatus) = Coro::rouse_wait;
		799	$rstatus
		800	}
		801
		802	In the case where C<rouse_cb> and C<rouse_wait> are not flexible enough,
		803	you can roll your own, using C<schedule>:
		804
		805	sub wait_for_child($) {
		806	my ($pid) = @_;
		807
		808	# store the current coro in $current,
		809	# and provide result variables for the closure passed to ->child
		810	my $current = $Coro::current;
		811	my ($done, $rstatus);
		812
		813	# pass a closure to ->child
		814	my $watcher = AnyEvent->child (pid => $pid, cb => sub {
		815	$rstatus = $_[1]; # remember rstatus
		816	$done = 1; # mark $rstatus as valud
		817	});
		818
		819	# wait until the closure has been called
		820	schedule while !$done;
		821
		822	$rstatus
		823	}
		824
		825
509	=head1 BUGS/LIMITATIONS	826	=head1 BUGS/LIMITATIONS
510		827
511	- you must make very sure that no coro is still active on global	828	=over 4
512	destruction. very bad things might happen otherwise (usually segfaults).
513		829
		830	=item fork with pthread backend
		831
		832	When Coro is compiled using the pthread backend (which isn't recommended
		833	but required on many BSDs as their libcs are completely broken), then
		834	coro will not survive a fork. There is no known workaround except to
		835	fix your libc and use a saner backend.
		836
		837	=item perl process emulation ("threads")
		838
514	- this module is not thread-safe. You should only ever use this module	839	This module is not perl-pseudo-thread-safe. You should only ever use this
515	from the same thread (this requirement might be losened in the future	840	module from the first thread (this requirement might be removed in the
516	to allow per-thread schedulers, but Coro::State does not yet allow	841	future to allow per-thread schedulers, but Coro::State does not yet allow
517	this).	842	this). I recommend disabling thread support and using processes, as having
		843	the windows process emulation enabled under unix roughly halves perl
		844	performance, even when not used.
		845
		846	=item coro switching is not signal safe
		847
		848	You must not switch to another coro from within a signal handler
		849	(only relevant with %SIG - most event libraries provide safe signals).
		850
		851	That means you I<MUST NOT> call any function that might "block" the
		852	current coro - C<cede>, C<schedule> C<< Coro::Semaphore->down >> or
		853	anything that calls those. Everything else, including calling C<ready>,
		854	works.
		855
		856	=back
		857
		858
		859	=head1 WINDOWS PROCESS EMULATION
		860
		861	A great many people seem to be confused about ithreads (for example, Chip
		862	Salzenberg called me unintelligent, incapable, stupid and gullible,
		863	while in the same mail making rather confused statements about perl
		864	ithreads (for example, that memory or files would be shared), showing his
		865	lack of understanding of this area - if it is hard to understand for Chip,
		866	it is probably not obvious to everybody).
		867
		868	What follows is an ultra-condensed version of my talk about threads in
		869	scripting languages given onthe perl workshop 2009:
		870
		871	The so-called "ithreads" were originally implemented for two reasons:
		872	first, to (badly) emulate unix processes on native win32 perls, and
		873	secondly, to replace the older, real thread model ("5.005-threads").
		874
		875	It does that by using threads instead of OS processes. The difference
		876	between processes and threads is that threads share memory (and other
		877	state, such as files) between threads within a single process, while
		878	processes do not share anything (at least not semantically). That
		879	means that modifications done by one thread are seen by others, while
		880	modifications by one process are not seen by other processes.
		881
		882	The "ithreads" work exactly like that: when creating a new ithreads
		883	process, all state is copied (memory is copied physically, files and code
		884	is copied logically). Afterwards, it isolates all modifications. On UNIX,
		885	the same behaviour can be achieved by using operating system processes,
		886	except that UNIX typically uses hardware built into the system to do this
		887	efficiently, while the windows process emulation emulates this hardware in
		888	software (rather efficiently, but of course it is still much slower than
		889	dedicated hardware).
		890
		891	As mentioned before, loading code, modifying code, modifying data
		892	structures and so on is only visible in the ithreads process doing the
		893	modification, not in other ithread processes within the same OS process.
		894
		895	This is why "ithreads" do not implement threads for perl at all, only
		896	processes. What makes it so bad is that on non-windows platforms, you can
		897	actually take advantage of custom hardware for this purpose (as evidenced
		898	by the forks module, which gives you the (i-) threads API, just much
		899	faster).
		900
		901	Sharing data is in the i-threads model is done by transfering data
		902	structures between threads using copying semantics, which is very slow -
		903	shared data simply does not exist. Benchmarks using i-threads which are
		904	communication-intensive show extremely bad behaviour with i-threads (in
		905	fact, so bad that Coro, which cannot take direct advantage of multiple
		906	CPUs, is often orders of magnitude faster because it shares data using
		907	real threads, refer to my talk for details).
		908
		909	As summary, i-threads use threads to implement processes, while
		910	the compatible forks module uses processes to emulate, uhm,
		911	processes. I-threads slow down every perl program when enabled, and
		912	outside of windows, serve no (or little) practical purpose, but
		913	disadvantages every single-threaded Perl program.
		914
		915	This is the reason that I try to avoid the name "ithreads", as it is
		916	misleading as it implies that it implements some kind of thread model for
		917	perl, and prefer the name "windows process emulation", which describes the
		918	actual use and behaviour of it much better.
518		919
519	=head1 SEE ALSO	920	=head1 SEE ALSO
520		921
		922	Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.
		923
		924	Debugging: L<Coro::Debug>.
		925
521	Support/Utility: L<Coro::Cont>, L<Coro::Specific>, L<Coro::State>, L<Coro::Util>.	926	Support/Utility: L<Coro::Specific>, L<Coro::Util>.
522		927
523	Locking/IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>, L<Coro::SemaphoreSet>, L<Coro::RWLock>.	928	Locking and IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>,
		929	L<Coro::SemaphoreSet>, L<Coro::RWLock>.
524		930
525	Event/IO: L<Coro::Timer>, L<Coro::Event>, L<Coro::Handle>, L<Coro::Socket>, L<Coro::Select>.	931	I/O and Timers: L<Coro::Timer>, L<Coro::Handle>, L<Coro::Socket>, L<Coro::AIO>.
526		932
527	Embedding: L<Coro:MakeMaker>	933	Compatibility with other modules: L<Coro::LWP> (but see also L<AnyEvent::HTTP> for
		934	a better-working alternative), L<Coro::BDB>, L<Coro::Storable>,
		935	L<Coro::Select>.
		936
		937	XS API: L<Coro::MakeMaker>.
		938
		939	Low level Configuration, Thread Environment, Continuations: L<Coro::State>.
528		940
529	=head1 AUTHOR	941	=head1 AUTHOR
530		942
531	Marc Lehmann <schmorp@schmorp.de>	943	Marc Lehmann <schmorp@schmorp.de>
532	http://home.schmorp.de/	944	http://home.schmorp.de/

Diff Legend

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

Comparing Coro/Coro.pm (file contents): Revision 1.104 by root, Thu Jan 4 23:49:27 2007 UTC vs. Revision 1.269 by root, Thu Oct 1 23:25:03 2009 UTC

Diff Legend

Comparing Coro/Coro.pm (file contents):
Revision 1.104 by root, Thu Jan 4 23:49:27 2007 UTC vs.
Revision 1.269 by root, Thu Oct 1 23:25:03 2009 UTC