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

Comparing Coro/Coro.pm (file contents):
Revision 1.11 by root, Sun Jul 15 03:24:18 2001 UTC vs.
Revision 1.191 by root, Sat May 31 12:10:55 2008 UTC

…		…
2		2
3	Coro - coroutine process abstraction	3	Coro - coroutine process abstraction
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 process like this:	16	cede; # yield to coroutine
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	my $lock = new Coro::Semaphore;
19	yield;	22	my $locked;
		23
		24	$lock->down;
		25	$locked = 1;
		26	$lock->up;
20		27
21	=head1 DESCRIPTION	28	=head1 DESCRIPTION
22		29
		30	This module collection manages coroutines. Coroutines are similar to
		31	threads but don't (in general) run in parallel at the same time even
		32	on SMP machines. The specific flavor of coroutine used in this module
		33	also 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 and easier than threads programming.
		37
		38	Unlike a normal perl program, however, coroutines allow you to have
		39	multiple running interpreters that share data, which is especially useful
		40	to code pseudo-parallel processes and for event-based programming, such as
		41	multiple HTTP-GET requests running concurrently. See L<Coro::AnyEvent> to
		42	learn more.
		43
		44	Coroutines are also useful because Perl has no support for threads (the so
		45	called "threads" that perl offers are nothing more than the (bad) process
		46	emulation coming from the Windows platform: On standard operating systems
		47	they serve no purpose whatsoever, except by making your programs slow and
		48	making them use a lot of memory. Best disable them when building perl, or
		49	aks your software vendor/distributor to do it for you).
		50
		51	In this module, coroutines are defined as "callchain + lexical variables +
		52	@_ + $_ + $@ + $/ + C stack), that is, a coroutine has its own callchain,
		53	its own set of lexicals and its own set of perls most important global
		54	variables (see L<Coro::State> for more configuration).
		55
23	=cut	56	=cut
24		57
25	package Coro;	58	package Coro;
26		59
		60	use strict;
		61	no warnings "uninitialized";
		62
27	use Coro::State;	63	use Coro::State;
28		64
29	use base Exporter;	65	use base qw(Coro::State Exporter);
30		66
		67	our $idle; # idle handler
		68	our $main; # main coroutine
		69	our $current; # current coroutine
		70
31	$VERSION = 0.04;	71	our $VERSION = 4.742;
32		72
33	@EXPORT = qw(async yield schedule);	73	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub);
34	@EXPORT_OK = qw($current);	74	our %EXPORT_TAGS = (
		75	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],
		76	);
		77	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));
35		78
36	{	79	=over 4
37	use subs 'async';
38		80
39	my @async;	81	=item $Coro::main
40		82
41	# this way of handling attributes simply is NOT scalable ;()	83	This variable stores the coroutine object that represents the main
42	sub import {	84	program. While you cna C<ready> it and do most other things you can do to
43	Coro->export_to_level(1, @_);	85	coroutines, it is mainly useful to compare again C<$Coro::current>, to see
44	my $old = *{(caller)[0]."::MODIFY_CODE_ATTRIBUTES"}{CODE};	86	wether you are running in the main program or not.
45	*{(caller)[0]."::MODIFY_CODE_ATTRIBUTES"} = sub {	87
46	my ($package, $ref) = (shift, shift);	88	=cut
47	my @attrs;	89
48	for (@_) {	90	$main = new Coro;
49	if ($_ eq "Coro") {	91
50	push @async, $ref;	92	=item $Coro::current
51	} else {	93
52	push @attrs, @_;	94	The coroutine object representing the current coroutine (the last
		95	coroutine that the Coro scheduler switched to). The initial value is
		96	C<$main> (of course).
		97
		98	This variable is B<strictly> I<read-only>. You can take copies of the
		99	value stored in it and use it as any other coroutine object, but you must
		100	not otherwise modify the variable itself.
		101
		102	=cut
		103
		104	$main->{desc} = "[main::]";
		105
		106	# maybe some other module used Coro::Specific before...
		107	$main->{_specific} = $current->{_specific}
		108	if $current;
		109
		110	_set_current $main;
		111
		112	sub current() { $current } # [DEPRECATED]
		113
		114	=item $Coro::idle
		115
		116	This variable is mainly useful to integrate Coro into event loops. It is
		117	usually better to rely on L<Coro::AnyEvent> or LC<Coro::EV>, as this is
		118	pretty low-level functionality.
		119
		120	This variable stores a callback that is called whenever the scheduler
		121	finds no ready coroutines to run. The default implementation prints
		122	"FATAL: deadlock detected" and exits, because the program has no other way
		123	to continue.
		124
		125	This hook is overwritten by modules such as C<Coro::Timer> and
		126	C<Coro::AnyEvent> to wait on an external event that hopefully wake up a
		127	coroutine so the scheduler can run it.
		128
		129	Note that the callback I<must not>, under any circumstances, block
		130	the current coroutine. Normally, this is achieved by having an "idle
		131	coroutine" that calls the event loop and then blocks again, and then
		132	readying that coroutine in the idle handler.
		133
		134	See L<Coro::Event> or L<Coro::AnyEvent> for examples of using this
		135	technique.
		136
		137	Please note that if your callback recursively invokes perl (e.g. for event
		138	handlers), then it must be prepared to be called recursively itself.
		139
		140	=cut
		141
		142	$idle = sub {
		143	require Carp;
		144	Carp::croak ("FATAL: deadlock detected");
		145	};
		146
		147	sub _cancel {
		148	my ($self) = @_;
		149
		150	# free coroutine data and mark as destructed
		151	$self->_destroy
		152	or return;
		153
		154	# call all destruction callbacks
		155	$_->(@{$self->{_status}})
		156	for @{(delete $self->{_on_destroy}) \|\| []};
		157	}
		158
		159	# this coroutine is necessary because a coroutine
		160	# cannot destroy itself.
		161	my @destroy;
		162	my $manager;
		163
		164	$manager = new Coro sub {
		165	while () {
		166	(shift @destroy)->_cancel
		167	while @destroy;
		168
		169	&schedule;
		170	}
		171	};
		172	$manager->desc ("[coro manager]");
		173	$manager->prio (PRIO_MAX);
		174
		175	=back
		176
		177	=head2 SIMPLE COROUTINE CREATION
		178
		179	=over 4
		180
		181	=item async { ... } [@args...]
		182
		183	Create a new coroutine and return it's coroutine object (usually
		184	unused). The coroutine will be put into the ready queue, so
		185	it will start running automatically on the next scheduler run.
		186
		187	The first argument is a codeblock/closure that should be executed in the
		188	coroutine. When it returns argument returns the coroutine is automatically
		189	terminated.
		190
		191	The remaining arguments are passed as arguments to the closure.
		192
		193	See the C<Coro::State::new> constructor for info about the coroutine
		194	environment in which coroutines are executed.
		195
		196	Calling C<exit> in a coroutine will do the same as calling exit outside
		197	the coroutine. Likewise, when the coroutine dies, the program will exit,
		198	just as it would in the main program.
		199
		200	If you do not want that, you can provide a default C<die> handler, or
		201	simply avoid dieing (by use of C<eval>).
		202
		203	Example: Create a new coroutine that just prints its arguments.
		204
		205	async {
		206	print "@_\n";
		207	} 1,2,3,4;
		208
		209	=cut
		210
		211	sub async(&@) {
		212	my $coro = new Coro @_;
		213	$coro->ready;
		214	$coro
		215	}
		216
		217	=item async_pool { ... } [@args...]
		218
		219	Similar to C<async>, but uses a coroutine pool, so you should not call
		220	terminate or join on it (although you are allowed to), and you get a
		221	coroutine that might have executed other code already (which can be good
		222	or bad :).
		223
		224	On the plus side, this function is faster than creating (and destroying)
		225	a completely new coroutine, so if you need a lot of generic coroutines in
		226	quick successsion, use C<async_pool>, not C<async>.
		227
		228	The code block is executed in an C<eval> context and a warning will be
		229	issued in case of an exception instead of terminating the program, as
		230	C<async> does. As the coroutine is being reused, stuff like C<on_destroy>
		231	will not work in the expected way, unless you call terminate or cancel,
		232	which somehow defeats the purpose of pooling (but is fine in the
		233	exceptional case).
		234
		235	The priority will be reset to C<0> after each run, tracing will be
		236	disabled, the description will be reset and the default output filehandle
		237	gets restored, so you can change all these. Otherwise the coroutine will
		238	be re-used "as-is": most notably if you change other per-coroutine global
		239	stuff such as C<$/> you I<must needs> to revert that change, which is most
		240	simply done by using local as in: C< local $/ >.
		241
		242	The pool size is limited to C<8> idle coroutines (this can be adjusted by
		243	changing $Coro::POOL_SIZE), and there can be as many non-idle coros as
		244	required.
		245
		246	If you are concerned about pooled coroutines growing a lot because a
		247	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool
		248	{ terminate }> once per second or so to slowly replenish the pool. In
		249	addition to that, when the stacks used by a handler grows larger than 16kb
		250	(adjustable via $Coro::POOL_RSS) it will also be destroyed.
		251
		252	=cut
		253
		254	our $POOL_SIZE = 8;
		255	our $POOL_RSS = 16 * 1024;
		256	our @async_pool;
		257
		258	sub pool_handler {
		259	my $cb;
		260
		261	while () {
		262	eval {
		263	while () {
		264	_pool_1 $cb;
53	}	265	&$cb;
		266	_pool_2 $cb;
		267	&schedule;
54	}	268	}
55	return $old ? $old->($package, $name, @attrs) : @attrs;
56	};	269	};
57	}
58		270
59	sub INIT {	271	last if $@ eq "\3async_pool terminate\2\n";
60	async pop @async while @async;	272	warn $@ if $@;
61	}	273	}
62	}	274	}
63		275
64	=item $main	276	sub async_pool(&@) {
		277	# this is also inlined into the unlock_scheduler
		278	my $coro = (pop @async_pool) \|\| new Coro \&pool_handler;
65		279
66	This coroutine represents the main program.	280	$coro->{_invoke} = [@_];
		281	$coro->ready;
67		282
68	=cut	283	$coro
69
70	our $main = new Coro;
71
72	=item $current
73
74	The current coroutine (the last coroutine switched to). The initial value is C<$main> (of course).
75
76	=cut
77
78	# maybe some other module used Coro::Specific before...
79	if ($current) {
80	$main->{specific} = $current->{specific};
81	}	284	}
82		285
83	our $current = $main;	286	=back
84
85	=item $idle
86
87	The coroutine to switch to when no other coroutine is running. The default
88	implementation prints "FATAL: deadlock detected" and exits.
89
90	=cut
91
92	# should be done using priorities :(
93	our $idle = new Coro sub {
94	print STDERR "FATAL: deadlock detected\n";
95	exit(51);
96	};
97
98	# we really need priorities...
99	my @ready = (); # the ready queue. hehe, rather broken ;)
100
101	# static methods. not really.
102		287
103	=head2 STATIC METHODS	288	=head2 STATIC METHODS
104		289
105	Static methods are actually functions that operate on the current process only.	290	Static methods are actually functions that operate on the current coroutine.
106		291
107	=over 4	292	=over 4
108		293
109	=item async { ... };
110
111	Create a new asynchronous process and return it's process object
112	(usually unused). When the sub returns the new process is automatically
113	terminated.
114
115	=cut
116
117	sub async(&) {
118	my $pid = new Coro $_[0];
119	$pid->ready;
120	$pid;
121	}
122
123	=item schedule	294	=item schedule
124		295
125	Calls the scheduler. Please note that the current process will not be put	296	Calls the scheduler. The scheduler will find the next coroutine that is
		297	to be run from the ready queue and switches to it. The next coroutine
		298	to be run is simply the one with the highest priority that is longest
		299	in its ready queue. If there is no coroutine ready, it will clal the
		300	C<$Coro::idle> hook.
		301
		302	Please note that the current coroutine will I<not> be put into the ready
126	into the ready queue, so calling this function usually means you will	303	queue, so calling this function usually means you will never be called
127	never be called again.	304	again unless something else (e.g. an event handler) calls C<< ->ready >>,
		305	thus waking you up.
128		306
129	=cut	307	This makes C<schedule> I<the> generic method to use to block the current
		308	coroutine and wait for events: first you remember the current coroutine in
		309	a variable, then arrange for some callback of yours to call C<< ->ready
		310	>> on that once some event happens, and last you call C<schedule> to put
		311	yourself to sleep. Note that a lot of things can wake your coroutine up,
		312	so you need to check wether the event indeed happened, e.g. by storing the
		313	status in a variable.
130		314
131	my $prev;	315	The canonical way to wait on external events is this:
132		316
133	sub schedule {	317	{
134	# should be done using priorities :(	318	# remember current coroutine
135	($prev, $current) = ($current, shift @ready \|\| $idle);	319	my $current = $Coro::current;
136	Coro::State::transfer($prev, $current);
137	}
138		320
139	=item yield	321	# register a hypothetical event handler
140		322	on_event_invoke sub {
141	Yield to other processes. This function puts the current process into the	323	# wake up sleeping coroutine
142	ready queue and calls C<schedule>.
143
144	=cut
145
146	sub yield {
147	$current->ready;	324	$current->ready;
148	&schedule;	325	undef $current;
149	}	326	};
150		327
		328	# call schedule until event occurred.
		329	# in case we are woken up for other reasons
		330	# (current still defined), loop.
		331	Coro::schedule while $current;
		332	}
		333
		334	=item cede
		335
		336	"Cede" to other coroutines. This function puts the current coroutine into
		337	the ready queue and calls C<schedule>, which has the effect of giving
		338	up the current "timeslice" to other coroutines of the same or higher
		339	priority. Once your coroutine gets its turn again it will automatically be
		340	resumed.
		341
		342	This function is often called C<yield> in other languages.
		343
		344	=item Coro::cede_notself
		345
		346	Works like cede, but is not exported by default and will cede to I<any>
		347	coroutine, regardless of priority. This is useful sometimes to ensure
		348	progress is made.
		349
151	=item terminate	350	=item terminate [arg...]
152		351
153	Terminates the current process.	352	Terminates the current coroutine with the given status values (see L<cancel>).
		353
		354	=item killall
		355
		356	Kills/terminates/cancels all coroutines except the currently running
		357	one. This is useful after a fork, either in the child or the parent, as
		358	usually only one of them should inherit the running coroutines.
		359
		360	Note that while this will try to free some of the main programs resources,
		361	you cnanot free all of them, so if a coroutine that is not the main
		362	program calls this function, there will be some one-time resource leak.
154		363
155	=cut	364	=cut
156		365
157	sub terminate {	366	sub terminate {
158	&schedule;	367	$current->cancel (@_);
		368	}
		369
		370	sub killall {
		371	for (Coro::State::list) {
		372	$_->cancel
		373	if $_ != $current && UNIVERSAL::isa $_, "Coro";
		374	}
159	}	375	}
160		376
161	=back	377	=back
162		378
163	# dynamic methods
164
165	=head2 PROCESS METHODS	379	=head2 COROUTINE METHODS
166		380
167	These are the methods you can call on process objects.	381	These are the methods you can call on coroutine objects (or to create
		382	them).
168		383
169	=over 4	384	=over 4
170		385
171	=item new Coro \⊂	386	=item new Coro \&sub [, @args...]
172		387
173	Create a new process and return it. When the sub returns the process	388	Create a new coroutine and return it. When the sub returns, the coroutine
174	automatically terminates. To start the process you must first put it into	389	automatically terminates as if C<terminate> with the returned values were
		390	called. To make the coroutine run you must first put it into the ready
175	the ready queue by calling the ready method.	391	queue by calling the ready method.
176		392
		393	See C<async> and C<Coro::State::new> for additional info about the
		394	coroutine environment.
		395
177	=cut	396	=cut
		397
		398	sub _run_coro {
		399	terminate &{+shift};
		400	}
178		401
179	sub new {	402	sub new {
180	my $class = shift;	403	my $class = shift;
		404
		405	$class->SUPER::new (\&_run_coro, @_)
		406	}
		407
		408	=item $success = $coroutine->ready
		409
		410	Put the given coroutine into the end of its ready queue (there is one
		411	queue for each priority) and return true. If the coroutine is already in
		412	the ready queue, do nothing and return false.
		413
		414	This ensures that the scheduler will resume this coroutine automatically
		415	once all the coroutines of higher priority and all coroutines of the same
		416	priority that were put into the ready queue earlier have been resumed.
		417
		418	=item $is_ready = $coroutine->is_ready
		419
		420	Return wether the coroutine is currently the ready queue or not,
		421
		422	=item $coroutine->cancel (arg...)
		423
		424	Terminates the given coroutine and makes it return the given arguments as
		425	status (default: the empty list). Never returns if the coroutine is the
		426	current coroutine.
		427
		428	=cut
		429
		430	sub cancel {
		431	my $self = shift;
		432	$self->{_status} = [@_];
		433
		434	if ($current == $self) {
		435	push @destroy, $self;
		436	$manager->ready;
		437	&schedule while 1;
		438	} else {
		439	$self->_cancel;
		440	}
		441	}
		442
		443	=item $coroutine->join
		444
		445	Wait until the coroutine terminates and return any values given to the
		446	C<terminate> or C<cancel> functions. C<join> can be called concurrently
		447	from multiple coroutines, and all will be resumed and given the status
		448	return once the C<$coroutine> terminates.
		449
		450	=cut
		451
		452	sub join {
		453	my $self = shift;
		454
		455	unless ($self->{_status}) {
		456	my $current = $current;
		457
		458	push @{$self->{_on_destroy}}, sub {
		459	$current->ready;
		460	undef $current;
		461	};
		462
		463	&schedule while $current;
		464	}
		465
		466	wantarray ? @{$self->{_status}} : $self->{_status}[0];
		467	}
		468
		469	=item $coroutine->on_destroy (\&cb)
		470
		471	Registers a callback that is called when this coroutine gets destroyed,
		472	but before it is joined. The callback gets passed the terminate arguments,
		473	if any, and I<must not> die, under any circumstances.
		474
		475	=cut
		476
		477	sub on_destroy {
		478	my ($self, $cb) = @_;
		479
		480	push @{ $self->{_on_destroy} }, $cb;
		481	}
		482
		483	=item $oldprio = $coroutine->prio ($newprio)
		484
		485	Sets (or gets, if the argument is missing) the priority of the
		486	coroutine. Higher priority coroutines get run before lower priority
		487	coroutines. Priorities are small signed integers (currently -4 .. +3),
		488	that you can refer to using PRIO_xxx constants (use the import tag :prio
		489	to get then):
		490
		491	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN
		492	3 > 1 > 0 > -1 > -3 > -4
		493
		494	# set priority to HIGH
		495	current->prio(PRIO_HIGH);
		496
		497	The idle coroutine ($Coro::idle) always has a lower priority than any
		498	existing coroutine.
		499
		500	Changing the priority of the current coroutine will take effect immediately,
		501	but changing the priority of coroutines in the ready queue (but not
		502	running) will only take effect after the next schedule (of that
		503	coroutine). This is a bug that will be fixed in some future version.
		504
		505	=item $newprio = $coroutine->nice ($change)
		506
		507	Similar to C<prio>, but subtract the given value from the priority (i.e.
		508	higher values mean lower priority, just as in unix).
		509
		510	=item $olddesc = $coroutine->desc ($newdesc)
		511
		512	Sets (or gets in case the argument is missing) the description for this
		513	coroutine. This is just a free-form string you can associate with a coroutine.
		514
		515	This method simply sets the C<< $coroutine->{desc} >> member to the given string. You
		516	can modify this member directly if you wish.
		517
		518	=item $coroutine->throw ([$scalar])
		519
		520	If C<$throw> is specified and defined, it will be thrown as an exception
		521	inside the coroutine at the next convinient point in time (usually after
		522	it gains control at the next schedule/transfer/cede). Otherwise clears the
		523	exception object.
		524
		525	The exception object will be thrown "as is" with the specified scalar in
		526	C<$@>, i.e. if it is a string, no line number or newline will be appended
		527	(unlike with C<die>).
		528
		529	This can be used as a softer means than C<cancel> to ask a coroutine to
		530	end itself, although there is no guarentee that the exception will lead to
		531	termination, and if the exception isn't caught it might well end the whole
		532	program.
		533
		534	=cut
		535
		536	sub desc {
181	my $proc = $_[0];	537	my $old = $_[0]{desc};
182	bless {	538	$_[0]{desc} = $_[1] if @_ > 1;
183	_coro_state => new Coro::State ($proc ? sub { &$proc; &terminate } : $proc),	539	$old;
184	}, $class;
185	}
186
187	=item $process->ready
188
189	Put the current process into the ready queue.
190
191	=cut
192
193	sub ready {
194	push @ready, $_[0];
195	}	540	}
196		541
197	=back	542	=back
198		543
		544	=head2 GLOBAL FUNCTIONS
		545
		546	=over 4
		547
		548	=item Coro::nready
		549
		550	Returns the number of coroutines that are currently in the ready state,
		551	i.e. that can be switched to by calling C<schedule> directory or
		552	indirectly. The value C<0> means that the only runnable coroutine is the
		553	currently running one, so C<cede> would have no effect, and C<schedule>
		554	would cause a deadlock unless there is an idle handler that wakes up some
		555	coroutines.
		556
		557	=item my $guard = Coro::guard { ... }
		558
		559	This creates and returns a guard object. Nothing happens until the object
		560	gets destroyed, in which case the codeblock given as argument will be
		561	executed. This is useful to free locks or other resources in case of a
		562	runtime error or when the coroutine gets canceled, as in both cases the
		563	guard block will be executed. The guard object supports only one method,
		564	C<< ->cancel >>, which will keep the codeblock from being executed.
		565
		566	Example: set some flag and clear it again when the coroutine gets canceled
		567	or the function returns:
		568
		569	sub do_something {
		570	my $guard = Coro::guard { $busy = 0 };
		571	$busy = 1;
		572
		573	# do something that requires $busy to be true
		574	}
		575
		576	=cut
		577
		578	sub guard(&) {
		579	bless \(my $cb = $_[0]), "Coro::guard"
		580	}
		581
		582	sub Coro::guard::cancel {
		583	${$_[0]} = sub { };
		584	}
		585
		586	sub Coro::guard::DESTROY {
		587	${$_[0]}->();
		588	}
		589
		590
		591	=item unblock_sub { ... }
		592
		593	This utility function takes a BLOCK or code reference and "unblocks" it,
		594	returning a new coderef. Unblocking means that calling the new coderef
		595	will return immediately without blocking, returning nothing, while the
		596	original code ref will be called (with parameters) from within another
		597	coroutine.
		598
		599	The reason this function exists is that many event libraries (such as the
		600	venerable L<Event\|Event> module) are not coroutine-safe (a weaker form
		601	of thread-safety). This means you must not block within event callbacks,
		602	otherwise you might suffer from crashes or worse. The only event library
		603	currently known that is safe to use without C<unblock_sub> is L<EV>.
		604
		605	This function allows your callbacks to block by executing them in another
		606	coroutine where it is safe to block. One example where blocking is handy
		607	is when you use the L<Coro::AIO\|Coro::AIO> functions to save results to
		608	disk, for example.
		609
		610	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when
		611	creating event callbacks that want to block.
		612
		613	If your handler does not plan to block (e.g. simply sends a message to
		614	another coroutine, or puts some other coroutine into the ready queue),
		615	there is no reason to use C<unblock_sub>.
		616
		617	Note that you also need to use C<unblock_sub> for any other callbacks that
		618	are indirectly executed by any C-based event loop. For example, when you
		619	use a module that uses L<AnyEvent> (and you use L<Coro::AnyEvent>) and it
		620	provides callbacks that are the result of some event callback, then you
		621	must not block either, or use C<unblock_sub>.
		622
		623	=cut
		624
		625	our @unblock_queue;
		626
		627	# we create a special coro because we want to cede,
		628	# to reduce pressure on the coro pool (because most callbacks
		629	# return immediately and can be reused) and because we cannot cede
		630	# inside an event callback.
		631	our $unblock_scheduler = new Coro sub {
		632	while () {
		633	while (my $cb = pop @unblock_queue) {
		634	# this is an inlined copy of async_pool
		635	my $coro = (pop @async_pool) \|\| new Coro \&pool_handler;
		636
		637	$coro->{_invoke} = $cb;
		638	$coro->ready;
		639	cede; # for short-lived callbacks, this reduces pressure on the coro pool
		640	}
		641	schedule; # sleep well
		642	}
		643	};
		644	$unblock_scheduler->desc ("[unblock_sub scheduler]");
		645
		646	sub unblock_sub(&) {
		647	my $cb = shift;
		648
		649	sub {
		650	unshift @unblock_queue, [$cb, @_];
		651	$unblock_scheduler->ready;
		652	}
		653	}
		654
		655	=back
		656
199	=cut	657	=cut
200		658
201	1;	659	1;
202		660
		661	=head1 BUGS/LIMITATIONS
		662
		663	This module is not perl-pseudo-thread-safe. You should only ever use this
		664	module from the same thread (this requirement might be removed in the
		665	future to allow per-thread schedulers, but Coro::State does not yet allow
		666	this). I recommend disabling thread support and using processes, as this
		667	is much faster and uses less memory.
		668
203	=head1 SEE ALSO	669	=head1 SEE ALSO
204		670
205	L<Coro::Channel>, L<Coro::Cont>, L<Coro::Specific>, L<Coro::Semaphore>,	671	Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.
206	L<Coro::Signal>, L<Coro::State>, L<Coro::Event>.	672
		673	Debugging: L<Coro::Debug>.
		674
		675	Support/Utility: L<Coro::Specific>, L<Coro::Util>.
		676
		677	Locking/IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>, L<Coro::SemaphoreSet>, L<Coro::RWLock>.
		678
		679	IO/Timers: L<Coro::Timer>, L<Coro::Handle>, L<Coro::Socket>, L<Coro::AIO>.
		680
		681	Compatibility: L<Coro::LWP>, L<Coro::BDB>, L<Coro::Storable>, L<Coro::Select>.
		682
		683	XS API: L<Coro::MakeMaker>.
		684
		685	Low level Configuration, Coroutine Environment: L<Coro::State>.
207		686
208	=head1 AUTHOR	687	=head1 AUTHOR
209		688
210	Marc Lehmann <pcg@goof.com>	689	Marc Lehmann <schmorp@schmorp.de>
211	http://www.goof.com/pcg/marc/	690	http://home.schmorp.de/
212		691
213	=cut	692	=cut
214		693

Diff Legend

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

Comparing Coro/Coro.pm (file contents): Revision 1.11 by root, Sun Jul 15 03:24:18 2001 UTC vs. Revision 1.191 by root, Sat May 31 12:10:55 2008 UTC

Diff Legend

Comparing Coro/Coro.pm (file contents):
Revision 1.11 by root, Sun Jul 15 03:24:18 2001 UTC vs.
Revision 1.191 by root, Sat May 31 12:10:55 2008 UTC