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

Comparing cvsroot/Coro/Coro.pm (file contents):
Revision 1.181 by root, Fri May 9 22:04:37 2008 UTC vs.
Revision 1.283 by root, Sat Feb 5 21:20:47 2011 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
11	print "2\n";	11	print "2\n";
12	cede; # yield back to main	12	cede; # yield back to main
13	print "4\n";	13	print "4\n";
14	};	14	};
15	print "1\n";	15	print "1\n";
16	cede; # yield to coroutine	16	cede; # yield to coro
17	print "3\n";	17	print "3\n";
18	cede; # and again	18	cede; # and again
19		19
20	# use locking	20	# use locking
		21	use Coro::Semaphore;
21	my $lock = new Coro::Semaphore;	22	my $lock = new Coro::Semaphore;
22	my $locked;	23	my $locked;
23		24
24	$lock->down;	25	$lock->down;
25	$locked = 1;	26	$locked = 1;
26	$lock->up;	27	$lock->up;
27		28
28	=head1 DESCRIPTION	29	=head1 DESCRIPTION
29		30
30	This module collection manages coroutines. Coroutines are similar to	31	For a tutorial-style introduction, please read the L<Coro::Intro>
31	threads but don't (in general) run in parallel at the same time even	32	manpage. This manpage mainly contains reference information.
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		33
38	Unlike a normal perl program, however, coroutines allow you to have	34	This module collection manages continuations in general, most often in
39	multiple running interpreters that share data, which is especially useful	35	the form of cooperative threads (also called coros, or simply "coro"
40	to code pseudo-parallel processes, such as multiple HTTP-GET requests	36	in the documentation). They are similar to kernel threads but don't (in
41	running concurrently.	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.
42		43
43	Coroutines are also useful because Perl has no support for threads (the so	44	Unlike the so-called "Perl threads" (which are not actually real threads
44	called "threads" that perl offers are nothing more than the (bad) process	45	but only the windows process emulation (see section of same name for more
45	emulation coming from the Windows platform: On standard operating systems	46	details) ported to unix, and as such act as processes), Coro provides
46	they serve no purpose whatsoever, except by making your programs slow and	47	a full shared address space, which makes communication between threads
47	making them use a lot of memory. Best disable them when building perl, or	48	very easy. And Coro's threads are fast, too: disabling the Windows
48	aks your software vendor/distributor to do it for you).	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.
49		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
50	In this module, coroutines are defined as "callchain + lexical variables +	60	In this module, a thread is defined as "callchain + lexical variables +
51	@_ + $_ + $@ + $/ + C stack), that is, a coroutine has its own callchain,	61	some package variables + C stack), that is, a thread has its own callchain,
52	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
53	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.
54		67
55	=cut	68	=cut
56		69
57	package Coro;	70	package Coro;
58		71
59	use strict;	72	use common::sense;
60	no warnings "uninitialized";	73
		74	use Carp ();
		75
		76	use Guard ();
61		77
62	use Coro::State;	78	use Coro::State;
63		79
64	use base qw(Coro::State Exporter);	80	use base qw(Coro::State Exporter);
65		81
66	our $idle; # idle handler	82	our $idle; # idle handler
67	our $main; # main coroutine	83	our $main; # main coro
68	our $current; # current coroutine	84	our $current; # current coro
69		85
70	our $VERSION = 4.6;	86	our $VERSION = 5.25;
71		87
72	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub);	88	our @EXPORT = qw(async async_pool cede schedule terminate current unblock_sub rouse_cb rouse_wait);
73	our %EXPORT_TAGS = (	89	our %EXPORT_TAGS = (
74	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],	90	prio => [qw(PRIO_MAX PRIO_HIGH PRIO_NORMAL PRIO_LOW PRIO_IDLE PRIO_MIN)],
75	);	91	);
76	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));	92	our @EXPORT_OK = (@{$EXPORT_TAGS{prio}}, qw(nready));
77		93
		94	=head1 GLOBAL VARIABLES
		95
78	=over 4	96	=over 4
79		97
80	=item $Coro::main	98	=item $Coro::main
81		99
82	This variable stores the coroutine object that represents the main	100	This variable stores the Coro object that represents the main
83	program. While you cna C<ready> it and do most other things you can do to	101	program. While you cna C<ready> it and do most other things you can do to
84	coroutines, it is mainly useful to compare again C<$Coro::current>, to see	102	coro, it is mainly useful to compare again C<$Coro::current>, to see
85	wether you are running in the main program or not.	103	whether you are running in the main program or not.
86		104
87	=cut	105	=cut
88		106
89	$main = new Coro;	107	# $main is now being initialised by Coro::State
90		108
91	=item $Coro::current	109	=item $Coro::current
92		110
93	The coroutine object representing the current coroutine (the last	111	The Coro object representing the current coro (the last
94	coroutine that the Coro scheduler switched to). The initial value is	112	coro that the Coro scheduler switched to). The initial value is
95	C<$main> (of course).	113	C<$Coro::main> (of course).
96		114
97	This variable is B<strictly> I<read-only>. You can take copies of the	115	This variable is B<strictly> I<read-only>. You can take copies of the
98	value stored in it and use it as any other coroutine object, but you must	116	value stored in it and use it as any other Coro object, but you must
99	not otherwise modify the variable itself.	117	not otherwise modify the variable itself.
100		118
101	=cut	119	=cut
102		120
103	$main->{desc} = "[main::]";
104
105	# maybe some other module used Coro::Specific before...
106	$main->{_specific} = $current->{_specific}
107	if $current;
108
109	_set_current $main;
110
111	sub current() { $current } # [DEPRECATED]	121	sub current() { $current } # [DEPRECATED]
112		122
113	=item $Coro::idle	123	=item $Coro::idle
114		124
115	This variable is mainly useful to integrate Coro into event loops. It is	125	This variable is mainly useful to integrate Coro into event loops. It is
116	usually better to rely on L<Coro::AnyEvent> or LC<Coro::EV>, as this is	126	usually better to rely on L<Coro::AnyEvent> or L<Coro::EV>, as this is
117	pretty low-level functionality.	127	pretty low-level functionality.
118		128
119	This variable stores a callback that is called whenever the scheduler	129	This variable stores a Coro object that is put into the ready queue when
120	finds no ready coroutines to run. The default implementation prints	130	there are no other ready threads (without invoking any ready hooks).
121	"FATAL: deadlock detected" and exits, because the program has no other way
122	to continue.
123		131
		132	The default implementation dies with "FATAL: deadlock detected.", followed
		133	by a thread listing, because the program has no other way to continue.
		134
124	This hook is overwritten by modules such as C<Coro::Timer> and	135	This hook is overwritten by modules such as C<Coro::EV> and
125	C<Coro::AnyEvent> to wait on an external event that hopefully wake up a	136	C<Coro::AnyEvent> to wait on an external event that hopefully wake up a
126	coroutine so the scheduler can run it.	137	coro so the scheduler can run it.
127		138
128	Note that the callback I<must not>, under any circumstances, block
129	the current coroutine. Normally, this is achieved by having an "idle
130	coroutine" that calls the event loop and then blocks again, and then
131	readying that coroutine in the idle handler.
132
133	See L<Coro::Event> or L<Coro::AnyEvent> for examples of using this	139	See L<Coro::EV> or L<Coro::AnyEvent> for examples of using this technique.
134	technique.
135		140
136	Please note that if your callback recursively invokes perl (e.g. for event
137	handlers), then it must be prepared to be called recursively itself.
138
139	=cut	141	=cut
140		142
141	$idle = sub {	143	# \|\|= because other modules could have provided their own by now
142	require Carp;	144	$idle \|\|= new Coro sub {
143	Carp::croak ("FATAL: deadlock detected");	145	require Coro::Debug;
		146	die "FATAL: deadlock detected.\n"
		147	. Coro::Debug::ps_listing ();
144	};	148	};
145		149
146	sub _cancel {
147	my ($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->{_on_destroy}) \|\| []};
156	}
157
158	# this coroutine is necessary because a coroutine	150	# this coro is necessary because a coro
159	# cannot destroy itself.	151	# cannot destroy itself.
160	my @destroy;	152	our @destroy;
161	my $manager;	153	our $manager;
162		154
163	$manager = new Coro sub {	155	$manager = new Coro sub {
164	while () {	156	while () {
165	(shift @destroy)->_cancel	157	Coro::State::cancel shift @destroy
166	while @destroy;	158	while @destroy;
167		159
168	&schedule;	160	&schedule;
169	}	161	}
170	};	162	};
171	$manager->desc ("[coro manager]");	163	$manager->{desc} = "[coro manager]";
172	$manager->prio (PRIO_MAX);	164	$manager->prio (PRIO_MAX);
173		165
174	=back	166	=back
175		167
176	=head2 SIMPLE COROUTINE CREATION	168	=head1 SIMPLE CORO CREATION
177		169
178	=over 4	170	=over 4
179		171
180	=item async { ... } [@args...]	172	=item async { ... } [@args...]
181		173
182	Create a new coroutine and return it's coroutine object (usually	174	Create a new coro and return its Coro object (usually
183	unused). The coroutine will be put into the ready queue, so	175	unused). The coro will be put into the ready queue, so
184	it will start running automatically on the next scheduler run.	176	it will start running automatically on the next scheduler run.
185		177
186	The first argument is a codeblock/closure that should be executed in the	178	The first argument is a codeblock/closure that should be executed in the
187	coroutine. When it returns argument returns the coroutine is automatically	179	coro. When it returns argument returns the coro is automatically
188	terminated.	180	terminated.
189		181
190	The remaining arguments are passed as arguments to the closure.	182	The remaining arguments are passed as arguments to the closure.
191		183
192	See the C<Coro::State::new> constructor for info about the coroutine	184	See the C<Coro::State::new> constructor for info about the coro
193	environment in which coroutines are executed.	185	environment in which coro are executed.
194		186
195	Calling C<exit> in a coroutine will do the same as calling exit outside	187	Calling C<exit> in a coro will do the same as calling exit outside
196	the coroutine. Likewise, when the coroutine dies, the program will exit,	188	the coro. Likewise, when the coro dies, the program will exit,
197	just as it would in the main program.	189	just as it would in the main program.
198		190
199	If you do not want that, you can provide a default C<die> handler, or	191	If you do not want that, you can provide a default C<die> handler, or
200	simply avoid dieing (by use of C<eval>).	192	simply avoid dieing (by use of C<eval>).
201		193
202	Example: Create a new coroutine that just prints its arguments.	194	Example: Create a new coro that just prints its arguments.
203		195
204	async {	196	async {
205	print "@_\n";	197	print "@_\n";
206	} 1,2,3,4;	198	} 1,2,3,4;
207		199
208	=cut
209
210	sub async(&@) {
211	my $coro = new Coro @_;
212	$coro->ready;
213	$coro
214	}
215
216	=item async_pool { ... } [@args...]	200	=item async_pool { ... } [@args...]
217		201
218	Similar to C<async>, but uses a coroutine pool, so you should not call	202	Similar to C<async>, but uses a coro pool, so you should not call
219	terminate or join on it (although you are allowed to), and you get a	203	terminate or join on it (although you are allowed to), and you get a
220	coroutine that might have executed other code already (which can be good	204	coro that might have executed other code already (which can be good
221	or bad :).	205	or bad :).
222		206
223	On the plus side, this function is faster than creating (and destroying)	207	On the plus side, this function is about twice as fast as creating (and
224	a completely new coroutine, so if you need a lot of generic coroutines in	208	destroying) a completely new coro, so if you need a lot of generic
225	quick successsion, use C<async_pool>, not C<async>.	209	coros in quick successsion, use C<async_pool>, not C<async>.
226		210
227	The code block is executed in an C<eval> context and a warning will be	211	The code block is executed in an C<eval> context and a warning will be
228	issued in case of an exception instead of terminating the program, as	212	issued in case of an exception instead of terminating the program, as
229	C<async> does. As the coroutine is being reused, stuff like C<on_destroy>	213	C<async> does. As the coro is being reused, stuff like C<on_destroy>
230	will not work in the expected way, unless you call terminate or cancel,	214	will not work in the expected way, unless you call terminate or cancel,
231	which somehow defeats the purpose of pooling (but is fine in the	215	which somehow defeats the purpose of pooling (but is fine in the
232	exceptional case).	216	exceptional case).
233		217
234	The priority will be reset to C<0> after each run, tracing will be	218	The priority will be reset to C<0> after each run, tracing will be
235	disabled, the description will be reset and the default output filehandle	219	disabled, the description will be reset and the default output filehandle
236	gets restored, so you can change all these. Otherwise the coroutine will	220	gets restored, so you can change all these. Otherwise the coro will
237	be re-used "as-is": most notably if you change other per-coroutine global	221	be re-used "as-is": most notably if you change other per-coro global
238	stuff such as C<$/> you I<must needs> to revert that change, which is most	222	stuff such as C<$/> you I<must needs> revert that change, which is most
239	simply done by using local as in: C< local $/ >.	223	simply done by using local as in: C<< local $/ >>.
240		224
241	The pool size is limited to C<8> idle coroutines (this can be adjusted by	225	The idle pool size is limited to C<8> idle coros (this can be
242	changing $Coro::POOL_SIZE), and there can be as many non-idle coros as	226	adjusted by changing $Coro::POOL_SIZE), but there can be as many non-idle
243	required.	227	coros as required.
244		228
245	If you are concerned about pooled coroutines growing a lot because a	229	If you are concerned about pooled coros growing a lot because a
246	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool	230	single C<async_pool> used a lot of stackspace you can e.g. C<async_pool
247	{ terminate }> once per second or so to slowly replenish the pool. In	231	{ terminate }> once per second or so to slowly replenish the pool. In
248	addition to that, when the stacks used by a handler grows larger than 16kb	232	addition to that, when the stacks used by a handler grows larger than 32kb
249	(adjustable via $Coro::POOL_RSS) it will also be destroyed.	233	(adjustable via $Coro::POOL_RSS) it will also be destroyed.
250		234
251	=cut	235	=cut
252		236
253	our $POOL_SIZE = 8;	237	our $POOL_SIZE = 8;
254	our $POOL_RSS = 16 * 1024;	238	our $POOL_RSS = 32 * 1024;
255	our @async_pool;	239	our @async_pool;
256		240
257	sub pool_handler {	241	sub pool_handler {
258	my $cb;
259
260	while () {	242	while () {
261	eval {	243	eval {
262	while () {	244	&{&_pool_handler} while 1;
263	_pool_1 $cb;
264	&$cb;
265	_pool_2 $cb;
266	&schedule;
267	}
268	};	245	};
269		246
270	last if $@ eq "\3async_pool terminate\2\n";
271	warn $@ if $@;	247	warn $@ if $@;
272	}	248	}
273	}	249	}
274		250
275	sub async_pool(&@) {
276	# this is also inlined into the unlock_scheduler
277	my $coro = (pop @async_pool) \|\| new Coro \&pool_handler;
278
279	$coro->{_invoke} = [@_];
280	$coro->ready;
281
282	$coro
283	}
284
285	=back	251	=back
286		252
287	=head2 STATIC METHODS	253	=head1 STATIC METHODS
288		254
289	Static methods are actually functions that operate on the current coroutine.	255	Static methods are actually functions that implicitly operate on the
		256	current coro.
290		257
291	=over 4	258	=over 4
292		259
293	=item schedule	260	=item schedule
294		261
295	Calls the scheduler. The scheduler will find the next coroutine that is	262	Calls the scheduler. The scheduler will find the next coro that is
296	to be run from the ready queue and switches to it. The next coroutine	263	to be run from the ready queue and switches to it. The next coro
297	to be run is simply the one with the highest priority that is longest	264	to be run is simply the one with the highest priority that is longest
298	in its ready queue. If there is no coroutine ready, it will clal the	265	in its ready queue. If there is no coro ready, it will call the
299	C<$Coro::idle> hook.	266	C<$Coro::idle> hook.
300		267
301	Please note that the current coroutine will I<not> be put into the ready	268	Please note that the current coro will I<not> be put into the ready
302	queue, so calling this function usually means you will never be called	269	queue, so calling this function usually means you will never be called
303	again unless something else (e.g. an event handler) calls C<< ->ready >>,	270	again unless something else (e.g. an event handler) calls C<< ->ready >>,
304	thus waking you up.	271	thus waking you up.
305		272
306	This makes C<schedule> I<the> generic method to use to block the current	273	This makes C<schedule> I<the> generic method to use to block the current
307	coroutine and wait for events: first you remember the current coroutine in	274	coro and wait for events: first you remember the current coro in
308	a variable, then arrange for some callback of yours to call C<< ->ready	275	a variable, then arrange for some callback of yours to call C<< ->ready
309	>> on that once some event happens, and last you call C<schedule> to put	276	>> on that once some event happens, and last you call C<schedule> to put
310	yourself to sleep. Note that a lot of things can wake your coroutine up,	277	yourself to sleep. Note that a lot of things can wake your coro up,
311	so you need to check wether the event indeed happened, e.g. by storing the	278	so you need to check whether the event indeed happened, e.g. by storing the
312	status in a variable.	279	status in a variable.
313		280
314	The canonical way to wait on external events is this:	281	See B<HOW TO WAIT FOR A CALLBACK>, below, for some ways to wait for callbacks.
315		282
316	{	283	=item cede
317	# remember current coroutine
318	my $current = $Coro::current;
319		284
320	# register a hypothetical event handler	285	"Cede" to other coros. This function puts the current coro into
321	on_event_invoke sub {	286	the ready queue and calls C<schedule>, which has the effect of giving
322	# wake up sleeping coroutine	287	up the current "timeslice" to other coros of the same or higher
323	$current->ready;	288	priority. Once your coro gets its turn again it will automatically be
324	undef $current;	289	resumed.
		290
		291	This function is often called C<yield> in other languages.
		292
		293	=item Coro::cede_notself
		294
		295	Works like cede, but is not exported by default and will cede to I<any>
		296	coro, regardless of priority. This is useful sometimes to ensure
		297	progress is made.
		298
		299	=item terminate [arg...]
		300
		301	Terminates the current coro with the given status values (see L<cancel>).
		302
		303	=item Coro::on_enter BLOCK, Coro::on_leave BLOCK
		304
		305	These function install enter and leave winders in the current scope. The
		306	enter block will be executed when on_enter is called and whenever the
		307	current coro is re-entered by the scheduler, while the leave block is
		308	executed whenever the current coro is blocked by the scheduler, and
		309	also when the containing scope is exited (by whatever means, be it exit,
		310	die, last etc.).
		311
		312	I<Neither invoking the scheduler, nor exceptions, are allowed within those
		313	BLOCKs>. That means: do not even think about calling C<die> without an
		314	eval, and do not even think of entering the scheduler in any way.
		315
		316	Since both BLOCKs are tied to the current scope, they will automatically
		317	be removed when the current scope exits.
		318
		319	These functions implement the same concept as C<dynamic-wind> in scheme
		320	does, and are useful when you want to localise some resource to a specific
		321	coro.
		322
		323	They slow down thread switching considerably for coros that use them
		324	(about 40% for a BLOCK with a single assignment, so thread switching is
		325	still reasonably fast if the handlers are fast).
		326
		327	These functions are best understood by an example: The following function
		328	will change the current timezone to "Antarctica/South_Pole", which
		329	requires a call to C<tzset>, but by using C<on_enter> and C<on_leave>,
		330	which remember/change the current timezone and restore the previous
		331	value, respectively, the timezone is only changed for the coro that
		332	installed those handlers.
		333
		334	use POSIX qw(tzset);
		335
		336	async {
		337	my $old_tz; # store outside TZ value here
		338
		339	Coro::on_enter {
		340	$old_tz = $ENV{TZ}; # remember the old value
		341
		342	$ENV{TZ} = "Antarctica/South_Pole";
		343	tzset; # enable new value
325	};	344	};
326		345
327	# call schedule until event occurred.	346	Coro::on_leave {
328	# in case we are woken up for other reasons	347	$ENV{TZ} = $old_tz;
329	# (current still defined), loop.	348	tzset; # restore old value
330	Coro::schedule while $current;	349	};
		350
		351	# at this place, the timezone is Antarctica/South_Pole,
		352	# without disturbing the TZ of any other coro.
331	}	353	};
332		354
333	=item cede	355	This can be used to localise about any resource (locale, uid, current
		356	working directory etc.) to a block, despite the existance of other
		357	coros.
334		358
335	"Cede" to other coroutines. This function puts the current coroutine into	359	Another interesting example implements time-sliced multitasking using
336	the ready queue and calls C<schedule>, which has the effect of giving	360	interval timers (this could obviously be optimised, but does the job):
337	up the current "timeslice" to other coroutines of the same or higher
338	priority. Once your coroutine gets its turn again it will automatically be
339	resumed.
340		361
341	This function is often called C<yield> in other languages.	362	# "timeslice" the given block
		363	sub timeslice(&) {
		364	use Time::HiRes ();
342		365
343	=item Coro::cede_notself	366	Coro::on_enter {
		367	# on entering the thread, we set an VTALRM handler to cede
		368	$SIG{VTALRM} = sub { cede };
		369	# and then start the interval timer
		370	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0.01, 0.01;
		371	};
		372	Coro::on_leave {
		373	# on leaving the thread, we stop the interval timer again
		374	Time::HiRes::setitimer &Time::HiRes::ITIMER_VIRTUAL, 0, 0;
		375	};
344		376
345	Works like cede, but is not exported by default and will cede to I<any>	377	&{+shift};
346	coroutine, regardless of priority. This is useful sometimes to ensure	378	}
347	progress is made.
348		379
349	=item terminate [arg...]	380	# use like this:
		381	timeslice {
		382	# The following is an endless loop that would normally
		383	# monopolise the process. Since it runs in a timesliced
		384	# environment, it will regularly cede to other threads.
		385	while () { }
		386	};
350		387
351	Terminates the current coroutine with the given status values (see L<cancel>).
352		388
353	=item killall	389	=item killall
354		390
355	Kills/terminates/cancels all coroutines except the currently running	391	Kills/terminates/cancels all coros except the currently running one.
356	one. This is useful after a fork, either in the child or the parent, as
357	usually only one of them should inherit the running coroutines.
358		392
359	Note that while this will try to free some of the main programs resources,	393	Note that while this will try to free some of the main interpreter
		394	resources if the calling coro isn't the main coro, but one
360	you cnanot free all of them, so if a coroutine that is not the main	395	cannot free all of them, so if a coro that is not the main coro
361	program calls this function, there will be some one-time resource leak.	396	calls this function, there will be some one-time resource leak.
362		397
363	=cut	398	=cut
364
365	sub terminate {
366	$current->cancel (@_);
367	}
368		399
369	sub killall {	400	sub killall {
370	for (Coro::State::list) {	401	for (Coro::State::list) {
371	$_->cancel	402	$_->cancel
372	if $_ != $current && UNIVERSAL::isa $_, "Coro";	403	if $_ != $current && UNIVERSAL::isa $_, "Coro";
373	}	404	}
374	}	405	}
375		406
376	=back	407	=back
377		408
378	=head2 COROUTINE METHODS	409	=head1 CORO OBJECT METHODS
379		410
380	These are the methods you can call on coroutine objects (or to create	411	These are the methods you can call on coro objects (or to create
381	them).	412	them).
382		413
383	=over 4	414	=over 4
384		415
385	=item new Coro \&sub [, @args...]	416	=item new Coro \&sub [, @args...]
386		417
387	Create a new coroutine and return it. When the sub returns, the coroutine	418	Create a new coro and return it. When the sub returns, the coro
388	automatically terminates as if C<terminate> with the returned values were	419	automatically terminates as if C<terminate> with the returned values were
389	called. To make the coroutine run you must first put it into the ready	420	called. To make the coro run you must first put it into the ready
390	queue by calling the ready method.	421	queue by calling the ready method.
391		422
392	See C<async> and C<Coro::State::new> for additional info about the	423	See C<async> and C<Coro::State::new> for additional info about the
393	coroutine environment.	424	coro environment.
394		425
395	=cut	426	=cut
396		427
397	sub _run_coro {	428	sub _coro_run {
398	terminate &{+shift};	429	terminate &{+shift};
399	}	430	}
400		431
401	sub new {
402	my $class = shift;
403
404	$class->SUPER::new (\&_run_coro, @_)
405	}
406
407	=item $success = $coroutine->ready	432	=item $success = $coro->ready
408		433
409	Put the given coroutine into the end of its ready queue (there is one	434	Put the given coro into the end of its ready queue (there is one
410	queue for each priority) and return true. If the coroutine is already in	435	queue for each priority) and return true. If the coro is already in
411	the ready queue, do nothing and return false.	436	the ready queue, do nothing and return false.
412		437
413	This ensures that the scheduler will resume this coroutine automatically	438	This ensures that the scheduler will resume this coro automatically
414	once all the coroutines of higher priority and all coroutines of the same	439	once all the coro of higher priority and all coro of the same
415	priority that were put into the ready queue earlier have been resumed.	440	priority that were put into the ready queue earlier have been resumed.
416		441
		442	=item $coro->suspend
		443
		444	Suspends the specified coro. A suspended coro works just like any other
		445	coro, except that the scheduler will not select a suspended coro for
		446	execution.
		447
		448	Suspending a coro can be useful when you want to keep the coro from
		449	running, but you don't want to destroy it, or when you want to temporarily
		450	freeze a coro (e.g. for debugging) to resume it later.
		451
		452	A scenario for the former would be to suspend all (other) coros after a
		453	fork and keep them alive, so their destructors aren't called, but new
		454	coros can be created.
		455
		456	=item $coro->resume
		457
		458	If the specified coro was suspended, it will be resumed. Note that when
		459	the coro was in the ready queue when it was suspended, it might have been
		460	unreadied by the scheduler, so an activation might have been lost.
		461
		462	To avoid this, it is best to put a suspended coro into the ready queue
		463	unconditionally, as every synchronisation mechanism must protect itself
		464	against spurious wakeups, and the one in the Coro family certainly do
		465	that.
		466
417	=item $is_ready = $coroutine->is_ready	467	=item $is_ready = $coro->is_ready
418		468
419	Return wether the coroutine is currently the ready queue or not,	469	Returns true iff the Coro object is in the ready queue. Unless the Coro
		470	object gets destroyed, it will eventually be scheduled by the scheduler.
420		471
		472	=item $is_running = $coro->is_running
		473
		474	Returns true iff the Coro object is currently running. Only one Coro object
		475	can ever be in the running state (but it currently is possible to have
		476	multiple running Coro::States).
		477
		478	=item $is_suspended = $coro->is_suspended
		479
		480	Returns true iff this Coro object has been suspended. Suspended Coros will
		481	not ever be scheduled.
		482
421	=item $coroutine->cancel (arg...)	483	=item $coro->cancel (arg...)
422		484
423	Terminates the given coroutine and makes it return the given arguments as	485	Terminates the given Coro and makes it return the given arguments as
424	status (default: the empty list). Never returns if the coroutine is the	486	status (default: the empty list). Never returns if the Coro is the
425	current coroutine.	487	current Coro.
426		488
427	=cut	489	=cut
428		490
429	sub cancel {	491	sub cancel {
430	my $self = shift;	492	my $self = shift;
431	$self->{_status} = [@_];
432		493
433	if ($current == $self) {	494	if ($current == $self) {
434	push @destroy, $self;	495	terminate @_;
435	$manager->ready;
436	&schedule while 1;
437	} else {	496	} else {
438	$self->_cancel;	497	$self->{_status} = [@_];
		498	Coro::State::cancel $self;
439	}	499	}
440	}	500	}
441		501
		502	=item $coro->schedule_to
		503
		504	Puts the current coro to sleep (like C<Coro::schedule>), but instead
		505	of continuing with the next coro from the ready queue, always switch to
		506	the given coro object (regardless of priority etc.). The readyness
		507	state of that coro isn't changed.
		508
		509	This is an advanced method for special cases - I'd love to hear about any
		510	uses for this one.
		511
		512	=item $coro->cede_to
		513
		514	Like C<schedule_to>, but puts the current coro into the ready
		515	queue. This has the effect of temporarily switching to the given
		516	coro, and continuing some time later.
		517
		518	This is an advanced method for special cases - I'd love to hear about any
		519	uses for this one.
		520
		521	=item $coro->throw ([$scalar])
		522
		523	If C<$throw> is specified and defined, it will be thrown as an exception
		524	inside the coro at the next convenient point in time. Otherwise
		525	clears the exception object.
		526
		527	Coro will check for the exception each time a schedule-like-function
		528	returns, i.e. after each C<schedule>, C<cede>, C<< Coro::Semaphore->down
		529	>>, C<< Coro::Handle->readable >> and so on. Most of these functions
		530	detect this case and return early in case an exception is pending.
		531
		532	The exception object will be thrown "as is" with the specified scalar in
		533	C<$@>, i.e. if it is a string, no line number or newline will be appended
		534	(unlike with C<die>).
		535
		536	This can be used as a softer means than C<cancel> to ask a coro to
		537	end itself, although there is no guarantee that the exception will lead to
		538	termination, and if the exception isn't caught it might well end the whole
		539	program.
		540
		541	You might also think of C<throw> as being the moral equivalent of
		542	C<kill>ing a coro with a signal (in this case, a scalar).
		543
442	=item $coroutine->join	544	=item $coro->join
443		545
444	Wait until the coroutine terminates and return any values given to the	546	Wait until the coro terminates and return any values given to the
445	C<terminate> or C<cancel> functions. C<join> can be called concurrently	547	C<terminate> or C<cancel> functions. C<join> can be called concurrently
446	from multiple coroutines, and all will be resumed and given the status	548	from multiple coro, and all will be resumed and given the status
447	return once the C<$coroutine> terminates.	549	return once the C<$coro> terminates.
448		550
449	=cut	551	=cut
450		552
451	sub join {	553	sub join {
452	my $self = shift;	554	my $self = shift;
…		…
463	}	565	}
464		566
465	wantarray ? @{$self->{_status}} : $self->{_status}[0];	567	wantarray ? @{$self->{_status}} : $self->{_status}[0];
466	}	568	}
467		569
468	=item $coroutine->on_destroy (\&cb)	570	=item $coro->on_destroy (\&cb)
469		571
470	Registers a callback that is called when this coroutine gets destroyed,	572	Registers a callback that is called when this coro gets destroyed,
471	but before it is joined. The callback gets passed the terminate arguments,	573	but before it is joined. The callback gets passed the terminate arguments,
472	if any, and I<must not> die, under any circumstances.	574	if any, and I<must not> die, under any circumstances.
473		575
474	=cut	576	=cut
475		577
…		…
477	my ($self, $cb) = @_;	579	my ($self, $cb) = @_;
478		580
479	push @{ $self->{_on_destroy} }, $cb;	581	push @{ $self->{_on_destroy} }, $cb;
480	}	582	}
481		583
482	=item $oldprio = $coroutine->prio ($newprio)	584	=item $oldprio = $coro->prio ($newprio)
483		585
484	Sets (or gets, if the argument is missing) the priority of the	586	Sets (or gets, if the argument is missing) the priority of the
485	coroutine. Higher priority coroutines get run before lower priority	587	coro. Higher priority coro get run before lower priority
486	coroutines. Priorities are small signed integers (currently -4 .. +3),	588	coro. Priorities are small signed integers (currently -4 .. +3),
487	that you can refer to using PRIO_xxx constants (use the import tag :prio	589	that you can refer to using PRIO_xxx constants (use the import tag :prio
488	to get then):	590	to get then):
489		591
490	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN	592	PRIO_MAX > PRIO_HIGH > PRIO_NORMAL > PRIO_LOW > PRIO_IDLE > PRIO_MIN
491	3 > 1 > 0 > -1 > -3 > -4	593	3 > 1 > 0 > -1 > -3 > -4
492		594
493	# set priority to HIGH	595	# set priority to HIGH
494	current->prio(PRIO_HIGH);	596	current->prio (PRIO_HIGH);
495		597
496	The idle coroutine ($Coro::idle) always has a lower priority than any	598	The idle coro ($Coro::idle) always has a lower priority than any
497	existing coroutine.	599	existing coro.
498		600
499	Changing the priority of the current coroutine will take effect immediately,	601	Changing the priority of the current coro will take effect immediately,
500	but changing the priority of coroutines in the ready queue (but not	602	but changing the priority of coro in the ready queue (but not
501	running) will only take effect after the next schedule (of that	603	running) will only take effect after the next schedule (of that
502	coroutine). This is a bug that will be fixed in some future version.	604	coro). This is a bug that will be fixed in some future version.
503		605
504	=item $newprio = $coroutine->nice ($change)	606	=item $newprio = $coro->nice ($change)
505		607
506	Similar to C<prio>, but subtract the given value from the priority (i.e.	608	Similar to C<prio>, but subtract the given value from the priority (i.e.
507	higher values mean lower priority, just as in unix).	609	higher values mean lower priority, just as in unix).
508		610
509	=item $olddesc = $coroutine->desc ($newdesc)	611	=item $olddesc = $coro->desc ($newdesc)
510		612
511	Sets (or gets in case the argument is missing) the description for this	613	Sets (or gets in case the argument is missing) the description for this
512	coroutine. This is just a free-form string you can associate with a coroutine.	614	coro. This is just a free-form string you can associate with a
		615	coro.
513		616
514	This method simply sets the C<< $coroutine->{desc} >> member to the given string. You	617	This method simply sets the C<< $coro->{desc} >> member to the given
515	can modify this member directly if you wish.	618	string. You can modify this member directly if you wish, and in fact, this
		619	is often preferred to indicate major processing states that cna then be
		620	seen for example in a L<Coro::Debug> session:
516		621
517	=item $coroutine->throw ([$scalar])	622	sub my_long_function {
518		623	local $Coro::current->{desc} = "now in my_long_function";
519	If C<$throw> is specified and defined, it will be thrown as an exception	624	...
520	inside the coroutine at the next convinient point in time (usually after	625	$Coro::current->{desc} = "my_long_function: phase 1";
521	it gains control at the next schedule/transfer/cede). Otherwise clears the	626	...
522	exception object.	627	$Coro::current->{desc} = "my_long_function: phase 2";
523		628	...
524	The exception object will be thrown "as is" with the specified scalar in	629	}
525	C<$@>, i.e. if it is a string, no line number or newline will be appended
526	(unlike with C<die>).
527
528	This can be used as a softer means than C<cancel> to ask a coroutine to
529	end itself, although there is no guarentee that the exception will lead to
530	termination, and if the exception isn't caught it might well end the whole
531	program.
532		630
533	=cut	631	=cut
534		632
535	sub desc {	633	sub desc {
536	my $old = $_[0]{desc};	634	my $old = $_[0]{desc};
537	$_[0]{desc} = $_[1] if @_ > 1;	635	$_[0]{desc} = $_[1] if @_ > 1;
538	$old;	636	$old;
539	}	637	}
540		638
		639	sub transfer {
		640	require Carp;
		641	Carp::croak ("You must not call ->transfer on Coro objects. Use Coro::State objects or the ->schedule_to method. Caught");
		642	}
		643
541	=back	644	=back
542		645
543	=head2 GLOBAL FUNCTIONS	646	=head1 GLOBAL FUNCTIONS
544		647
545	=over 4	648	=over 4
546		649
547	=item Coro::nready	650	=item Coro::nready
548		651
549	Returns the number of coroutines that are currently in the ready state,	652	Returns the number of coro that are currently in the ready state,
550	i.e. that can be switched to by calling C<schedule> directory or	653	i.e. that can be switched to by calling C<schedule> directory or
551	indirectly. The value C<0> means that the only runnable coroutine is the	654	indirectly. The value C<0> means that the only runnable coro is the
552	currently running one, so C<cede> would have no effect, and C<schedule>	655	currently running one, so C<cede> would have no effect, and C<schedule>
553	would cause a deadlock unless there is an idle handler that wakes up some	656	would cause a deadlock unless there is an idle handler that wakes up some
554	coroutines.	657	coro.
555		658
556	=item my $guard = Coro::guard { ... }	659	=item my $guard = Coro::guard { ... }
557		660
558	This creates and returns a guard object. Nothing happens until the object	661	This function still exists, but is deprecated. Please use the
559	gets destroyed, in which case the codeblock given as argument will be	662	C<Guard::guard> function instead.
560	executed. This is useful to free locks or other resources in case of a
561	runtime error or when the coroutine gets canceled, as in both cases the
562	guard block will be executed. The guard object supports only one method,
563	C<< ->cancel >>, which will keep the codeblock from being executed.
564		663
565	Example: set some flag and clear it again when the coroutine gets canceled
566	or the function returns:
567
568	sub do_something {
569	my $guard = Coro::guard { $busy = 0 };
570	$busy = 1;
571
572	# do something that requires $busy to be true
573	}
574
575	=cut	664	=cut
576		665
577	sub guard(&) {	666	BEGIN { *guard = \&Guard::guard }
578	bless \(my $cb = $_[0]), "Coro::guard"
579	}
580
581	sub Coro::guard::cancel {
582	${$_[0]} = sub { };
583	}
584
585	sub Coro::guard::DESTROY {
586	${$_[0]}->();
587	}
588
589		667
590	=item unblock_sub { ... }	668	=item unblock_sub { ... }
591		669
592	This utility function takes a BLOCK or code reference and "unblocks" it,	670	This utility function takes a BLOCK or code reference and "unblocks" it,
593	returning a new coderef. Unblocking means that calling the new coderef	671	returning a new coderef. Unblocking means that calling the new coderef
594	will return immediately without blocking, returning nothing, while the	672	will return immediately without blocking, returning nothing, while the
595	original code ref will be called (with parameters) from within another	673	original code ref will be called (with parameters) from within another
596	coroutine.	674	coro.
597		675
598	The reason this function exists is that many event libraries (such as the	676	The reason this function exists is that many event libraries (such as
599	venerable L<Event\|Event> module) are not coroutine-safe (a weaker form	677	the venerable L<Event\|Event> module) are not thread-safe (a weaker form
600	of thread-safety). This means you must not block within event callbacks,	678	of reentrancy). This means you must not block within event callbacks,
601	otherwise you might suffer from crashes or worse. The only event library	679	otherwise you might suffer from crashes or worse. The only event library
602	currently known that is safe to use without C<unblock_sub> is L<EV>.	680	currently known that is safe to use without C<unblock_sub> is L<EV> (but
		681	you might still run into deadlocks if all event loops are blocked).
		682
		683	Coro will try to catch you when you block in the event loop
		684	("FATAL:$Coro::IDLE blocked itself"), but this is just best effort and
		685	only works when you do not run your own event loop.
603		686
604	This function allows your callbacks to block by executing them in another	687	This function allows your callbacks to block by executing them in another
605	coroutine where it is safe to block. One example where blocking is handy	688	coro where it is safe to block. One example where blocking is handy
606	is when you use the L<Coro::AIO\|Coro::AIO> functions to save results to	689	is when you use the L<Coro::AIO\|Coro::AIO> functions to save results to
607	disk, for example.	690	disk, for example.
608		691
609	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when	692	In short: simply use C<unblock_sub { ... }> instead of C<sub { ... }> when
610	creating event callbacks that want to block.	693	creating event callbacks that want to block.
611		694
612	If your handler does not plan to block (e.g. simply sends a message to	695	If your handler does not plan to block (e.g. simply sends a message to
613	another coroutine, or puts some other coroutine into the ready queue),	696	another coro, or puts some other coro into the ready queue), there is
614	there is no reason to use C<unblock_sub>.	697	no reason to use C<unblock_sub>.
		698
		699	Note that you also need to use C<unblock_sub> for any other callbacks that
		700	are indirectly executed by any C-based event loop. For example, when you
		701	use a module that uses L<AnyEvent> (and you use L<Coro::AnyEvent>) and it
		702	provides callbacks that are the result of some event callback, then you
		703	must not block either, or use C<unblock_sub>.
615		704
616	=cut	705	=cut
617		706
618	our @unblock_queue;	707	our @unblock_queue;
619		708
…		…
622	# return immediately and can be reused) and because we cannot cede	711	# return immediately and can be reused) and because we cannot cede
623	# inside an event callback.	712	# inside an event callback.
624	our $unblock_scheduler = new Coro sub {	713	our $unblock_scheduler = new Coro sub {
625	while () {	714	while () {
626	while (my $cb = pop @unblock_queue) {	715	while (my $cb = pop @unblock_queue) {
627	# this is an inlined copy of async_pool	716	&async_pool (@$cb);
628	my $coro = (pop @async_pool) \|\| new Coro \&pool_handler;
629		717
630	$coro->{_invoke} = $cb;
631	$coro->ready;
632	cede; # for short-lived callbacks, this reduces pressure on the coro pool	718	# for short-lived callbacks, this reduces pressure on the coro pool
		719	# as the chance is very high that the async_poll coro will be back
		720	# in the idle state when cede returns
		721	cede;
633	}	722	}
634	schedule; # sleep well	723	schedule; # sleep well
635	}	724	}
636	};	725	};
637	$unblock_scheduler->desc ("[unblock_sub scheduler]");	726	$unblock_scheduler->{desc} = "[unblock_sub scheduler]";
638		727
639	sub unblock_sub(&) {	728	sub unblock_sub(&) {
640	my $cb = shift;	729	my $cb = shift;
641		730
642	sub {	731	sub {
643	unshift @unblock_queue, [$cb, @_];	732	unshift @unblock_queue, [$cb, @_];
644	$unblock_scheduler->ready;	733	$unblock_scheduler->ready;
645	}	734	}
646	}	735	}
647		736
		737	=item $cb = rouse_cb
		738
		739	Create and return a "rouse callback". That's a code reference that,
		740	when called, will remember a copy of its arguments and notify the owner
		741	coro of the callback.
		742
		743	See the next function.
		744
		745	=item @args = rouse_wait [$cb]
		746
		747	Wait for the specified rouse callback (or the last one that was created in
		748	this coro).
		749
		750	As soon as the callback is invoked (or when the callback was invoked
		751	before C<rouse_wait>), it will return the arguments originally passed to
		752	the rouse callback. In scalar context, that means you get the I<last>
		753	argument, just as if C<rouse_wait> had a C<return ($a1, $a2, $a3...)>
		754	statement at the end.
		755
		756	See the section B<HOW TO WAIT FOR A CALLBACK> for an actual usage example.
		757
648	=back	758	=back
649		759
650	=cut	760	=cut
651		761
		762	for my $module (qw(Channel RWLock Semaphore SemaphoreSet Signal Specific)) {
		763	my $old = defined &{"Coro::$module\::new"} && \&{"Coro::$module\::new"};
		764
		765	*{"Coro::$module\::new"} = sub {
		766	require "Coro/$module.pm";
		767
		768	# some modules have their new predefined in State.xs, some don't
		769	*{"Coro::$module\::new"} = $old
		770	if $old;
		771
		772	goto &{"Coro::$module\::new"};
		773	};
		774	}
		775
652	1;	776	1;
653		777
		778	=head1 HOW TO WAIT FOR A CALLBACK
		779
		780	It is very common for a coro to wait for some callback to be
		781	called. This occurs naturally when you use coro in an otherwise
		782	event-based program, or when you use event-based libraries.
		783
		784	These typically register a callback for some event, and call that callback
		785	when the event occured. In a coro, however, you typically want to
		786	just wait for the event, simplyifying things.
		787
		788	For example C<< AnyEvent->child >> registers a callback to be called when
		789	a specific child has exited:
		790
		791	my $child_watcher = AnyEvent->child (pid => $pid, cb => sub { ... });
		792
		793	But from within a coro, you often just want to write this:
		794
		795	my $status = wait_for_child $pid;
		796
		797	Coro offers two functions specifically designed to make this easy,
		798	C<Coro::rouse_cb> and C<Coro::rouse_wait>.
		799
		800	The first function, C<rouse_cb>, generates and returns a callback that,
		801	when invoked, will save its arguments and notify the coro that
		802	created the callback.
		803
		804	The second function, C<rouse_wait>, waits for the callback to be called
		805	(by calling C<schedule> to go to sleep) and returns the arguments
		806	originally passed to the callback.
		807
		808	Using these functions, it becomes easy to write the C<wait_for_child>
		809	function mentioned above:
		810
		811	sub wait_for_child($) {
		812	my ($pid) = @_;
		813
		814	my $watcher = AnyEvent->child (pid => $pid, cb => Coro::rouse_cb);
		815
		816	my ($rpid, $rstatus) = Coro::rouse_wait;
		817	$rstatus
		818	}
		819
		820	In the case where C<rouse_cb> and C<rouse_wait> are not flexible enough,
		821	you can roll your own, using C<schedule>:
		822
		823	sub wait_for_child($) {
		824	my ($pid) = @_;
		825
		826	# store the current coro in $current,
		827	# and provide result variables for the closure passed to ->child
		828	my $current = $Coro::current;
		829	my ($done, $rstatus);
		830
		831	# pass a closure to ->child
		832	my $watcher = AnyEvent->child (pid => $pid, cb => sub {
		833	$rstatus = $_[1]; # remember rstatus
		834	$done = 1; # mark $rstatus as valud
		835	});
		836
		837	# wait until the closure has been called
		838	schedule while !$done;
		839
		840	$rstatus
		841	}
		842
		843
654	=head1 BUGS/LIMITATIONS	844	=head1 BUGS/LIMITATIONS
655		845
		846	=over 4
		847
		848	=item fork with pthread backend
		849
		850	When Coro is compiled using the pthread backend (which isn't recommended
		851	but required on many BSDs as their libcs are completely broken), then
		852	coro will not survive a fork. There is no known workaround except to
		853	fix your libc and use a saner backend.
		854
		855	=item perl process emulation ("threads")
		856
656	This module is not perl-pseudo-thread-safe. You should only ever use this	857	This module is not perl-pseudo-thread-safe. You should only ever use this
657	module from the same thread (this requirement might be removed in the	858	module from the first thread (this requirement might be removed in the
658	future to allow per-thread schedulers, but Coro::State does not yet allow	859	future to allow per-thread schedulers, but Coro::State does not yet allow
659	this). I recommend disabling thread support and using processes, as this	860	this). I recommend disabling thread support and using processes, as having
660	is much faster and uses less memory.	861	the windows process emulation enabled under unix roughly halves perl
		862	performance, even when not used.
		863
		864	=item coro switching is not signal safe
		865
		866	You must not switch to another coro from within a signal handler (only
		867	relevant with %SIG - most event libraries provide safe signals), I<unless>
		868	you are sure you are not interrupting a Coro function.
		869
		870	That means you I<MUST NOT> call any function that might "block" the
		871	current coro - C<cede>, C<schedule> C<< Coro::Semaphore->down >> or
		872	anything that calls those. Everything else, including calling C<ready>,
		873	works.
		874
		875	=back
		876
		877
		878	=head1 WINDOWS PROCESS EMULATION
		879
		880	A great many people seem to be confused about ithreads (for example, Chip
		881	Salzenberg called me unintelligent, incapable, stupid and gullible,
		882	while in the same mail making rather confused statements about perl
		883	ithreads (for example, that memory or files would be shared), showing his
		884	lack of understanding of this area - if it is hard to understand for Chip,
		885	it is probably not obvious to everybody).
		886
		887	What follows is an ultra-condensed version of my talk about threads in
		888	scripting languages given on the perl workshop 2009:
		889
		890	The so-called "ithreads" were originally implemented for two reasons:
		891	first, to (badly) emulate unix processes on native win32 perls, and
		892	secondly, to replace the older, real thread model ("5.005-threads").
		893
		894	It does that by using threads instead of OS processes. The difference
		895	between processes and threads is that threads share memory (and other
		896	state, such as files) between threads within a single process, while
		897	processes do not share anything (at least not semantically). That
		898	means that modifications done by one thread are seen by others, while
		899	modifications by one process are not seen by other processes.
		900
		901	The "ithreads" work exactly like that: when creating a new ithreads
		902	process, all state is copied (memory is copied physically, files and code
		903	is copied logically). Afterwards, it isolates all modifications. On UNIX,
		904	the same behaviour can be achieved by using operating system processes,
		905	except that UNIX typically uses hardware built into the system to do this
		906	efficiently, while the windows process emulation emulates this hardware in
		907	software (rather efficiently, but of course it is still much slower than
		908	dedicated hardware).
		909
		910	As mentioned before, loading code, modifying code, modifying data
		911	structures and so on is only visible in the ithreads process doing the
		912	modification, not in other ithread processes within the same OS process.
		913
		914	This is why "ithreads" do not implement threads for perl at all, only
		915	processes. What makes it so bad is that on non-windows platforms, you can
		916	actually take advantage of custom hardware for this purpose (as evidenced
		917	by the forks module, which gives you the (i-) threads API, just much
		918	faster).
		919
		920	Sharing data is in the i-threads model is done by transfering data
		921	structures between threads using copying semantics, which is very slow -
		922	shared data simply does not exist. Benchmarks using i-threads which are
		923	communication-intensive show extremely bad behaviour with i-threads (in
		924	fact, so bad that Coro, which cannot take direct advantage of multiple
		925	CPUs, is often orders of magnitude faster because it shares data using
		926	real threads, refer to my talk for details).
		927
		928	As summary, i-threads use threads to implement processes, while
		929	the compatible forks module uses processes to emulate, uhm,
		930	processes. I-threads slow down every perl program when enabled, and
		931	outside of windows, serve no (or little) practical purpose, but
		932	disadvantages every single-threaded Perl program.
		933
		934	This is the reason that I try to avoid the name "ithreads", as it is
		935	misleading as it implies that it implements some kind of thread model for
		936	perl, and prefer the name "windows process emulation", which describes the
		937	actual use and behaviour of it much better.
661		938
662	=head1 SEE ALSO	939	=head1 SEE ALSO
663		940
664	Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.	941	Event-Loop integration: L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.
665		942
666	Debugging: L<Coro::Debug>.	943	Debugging: L<Coro::Debug>.
667		944
668	Support/Utility: L<Coro::Specific>, L<Coro::Util>.	945	Support/Utility: L<Coro::Specific>, L<Coro::Util>.
669		946
670	Locking/IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>, L<Coro::SemaphoreSet>, L<Coro::RWLock>.	947	Locking and IPC: L<Coro::Signal>, L<Coro::Channel>, L<Coro::Semaphore>,
		948	L<Coro::SemaphoreSet>, L<Coro::RWLock>.
671		949
672	IO/Timers: L<Coro::Timer>, L<Coro::Handle>, L<Coro::Socket>, L<Coro::AIO>.	950	I/O and Timers: L<Coro::Timer>, L<Coro::Handle>, L<Coro::Socket>, L<Coro::AIO>.
673		951
674	Compatibility: L<Coro::LWP>, L<Coro::BDB>, L<Coro::Storable>, L<Coro::Select>.	952	Compatibility with other modules: L<Coro::LWP> (but see also L<AnyEvent::HTTP> for
		953	a better-working alternative), L<Coro::BDB>, L<Coro::Storable>,
		954	L<Coro::Select>.
675		955
676	XS API: L<Coro::MakeMaker>.	956	XS API: L<Coro::MakeMaker>.
677		957
678	Low level Configuration, Coroutine Environment: L<Coro::State>.	958	Low level Configuration, Thread Environment, Continuations: L<Coro::State>.
679		959
680	=head1 AUTHOR	960	=head1 AUTHOR
681		961
682	Marc Lehmann <schmorp@schmorp.de>	962	Marc Lehmann <schmorp@schmorp.de>
683	http://home.schmorp.de/	963	http://home.schmorp.de/

Diff Legend

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

Comparing cvsroot/Coro/Coro.pm (file contents): Revision 1.181 by root, Fri May 9 22:04:37 2008 UTC vs. Revision 1.283 by root, Sat Feb 5 21:20:47 2011 UTC

Diff Legend

Comparing cvsroot/Coro/Coro.pm (file contents):
Revision 1.181 by root, Fri May 9 22:04:37 2008 UTC vs.
Revision 1.283 by root, Sat Feb 5 21:20:47 2011 UTC