ext/threads/threads.pm

package threads;

use 5.008;
use strict;
use warnings;
use Config;

BEGIN {
    unless ($Config{useithreads}) {
        my @caller = caller(2);
        die <<EOF;
$caller[1] line $caller[2]:

This Perl hasn't been configured and built properly for the threads
module to work.  (The 'useithreads' configuration option hasn't been used.)

Having threads support requires all of Perl and all of the XS modules in
the Perl installation to be rebuilt, it is not just a question of adding
the threads module.  (In other words, threaded and non-threaded Perls
are binary incompatible.)

If you want to the use the threads module, please contact the people
who built your Perl.

Cannot continue, aborting.
EOF
    }
}

use overload
    '==' => \&equal,
    'fallback' => 1;

BEGIN {
    warn "Warning, threads::shared has already been loaded. ".
       "To enable shared variables for these modules 'use threads' ".
       "must be called before any of those modules are loaded\n"
               if($threads::shared::threads_shared);
}

require Exporter;
require DynaLoader;

our @ISA = qw(Exporter DynaLoader);

our %EXPORT_TAGS = ( all => [qw(yield)]);

our @EXPORT_OK = ( @{ $EXPORT_TAGS{'all'} } );

our @EXPORT = qw(
async   
);
our $VERSION = '1.05';


# || 0 to ensure compatibility with previous versions
sub equal { ($_[0]->tid == $_[1]->tid) || 0 }

# use "goto" trick to avoid pad problems from 5.8.1 (fixed in 5.8.2)
# should also be faster
sub async (&;@) { unshift @_,'threads'; goto &new }

sub object {
    return undef unless @_ > 1;
    foreach (threads->list) {
        return $_ if $_->tid == $_[1];
    }
    return undef;
}

$threads::threads = 1;

bootstrap threads $VERSION;

# why document 'new' then use 'create' in the tests!
*create = \&new;

# Preloaded methods go here.

1;
__END__

=head1 NAME

threads - Perl extension allowing use of interpreter based threads from perl

=head1 SYNOPSIS

    use threads;

    sub start_thread {
        print "Thread started\n";
    }

    my $thread  = threads->create("start_thread","argument");
    my $thread2 = $thread->create(sub { print "I am a thread"},"argument");
    my $thread3 = async { foreach (@files) { ... } };

    $thread->join();
    $thread->detach();

    $thread = threads->self();
    $thread = threads->object( $tid );

    $thread->tid();
    threads->tid();
    threads->self->tid();

    threads->yield();

    threads->list();

=head1 DESCRIPTION

Perl 5.6 introduced something called interpreter threads.  Interpreter
threads are different from "5005threads" (the thread model of Perl
5.005) by creating a new perl interpreter per thread and not sharing
any data or state between threads by default.

Prior to perl 5.8 this has only been available to people embedding
perl and for emulating fork() on windows.

The threads API is loosely based on the old Thread.pm API. It is very
important to note that variables are not shared between threads, all
variables are per default thread local.  To use shared variables one
must use threads::shared.

It is also important to note that you must enable threads by doing
C<use threads> as early as possible in the script itself and that it
is not possible to enable threading inside an C<eval "">, C<do>,
C<require>, or C<use>.  In particular, if you are intending to share
variables with threads::shared, you must C<use threads> before you
C<use threads::shared> and C<threads> will emit a warning if you do
it the other way around.

=over

=item $thread = threads->create(function, LIST)

This will create a new thread with the entry point function and give
it LIST as parameters.  It will return the corresponding threads
object, or C<undef> if thread creation failed. The new() method is an
alias for create().

=item $thread->join

This will wait for the corresponding thread to join. When the thread
finishes, join() will return the return values of the entry point
function. If the thread has been detached, an error will be thrown.

The context (scalar or list) of the thread creation is also the
context for join().  This means that if you intend to return an array
from a thread, you must use C<my ($thread) = threads->new(...)>, and
that if you intend to return a scalar, you must use C<my $thread = ...>.

If the program exits without all other threads having been either
joined or detached, then a warning will be issued. (A program exits
either because one of its threads explicitly calls exit(), or in the
case of the main thread, reaches the end of the main program file.)


=item $thread->detach

Will make the thread unjoinable, and cause any eventual return value
to be discarded.

=item threads->self

This will return the thread object for the current thread.

=item $thread->tid

This will return the id of the thread.  Thread IDs are integers, with
the main thread in a program being 0.  Currently Perl assigns a unique
tid to every thread ever created in your program, assigning the first
thread to be created a tid of 1, and increasing the tid by 1 for each
new thread that's created.

NB the class method C<< threads->tid() >> is a quick way to get the
current thread id if you don't have your thread object handy.

=item threads->object( tid )

This will return the thread object for the thread associated with the
specified tid.  Returns undef if there is no thread associated with the tid
or no tid is specified or the specified tid is undef.

=item threads->yield();

This is a suggestion to the OS to let this thread yield CPU time to other
threads.  What actually happens is highly dependent upon the underlying
thread implementation.

You may do C<use threads qw(yield)> then use just a bare C<yield> in your
code.

=item threads->list();

This will return a list of all non joined, non detached threads.

=item async BLOCK;

C<async> creates a thread to execute the block immediately following
it.  This block is treated as an anonymous sub, and so must have a
semi-colon after the closing brace. Like C<< threads->new >>, C<async>
returns a thread object.

=back

=head1 WARNINGS

=over 4

=item A thread exited while %d other threads were still running

A thread (not necessarily the main thread) exited while there were
still other threads running.  Usually it's a good idea to first collect
the return values of the created threads by joining them, and only then
exit from the main thread.

=back

=head1 TODO

The current implementation of threads has been an attempt to get
a correct threading system working that could be built on, 
and optimized, in newer versions of perl.

Currently the overhead of creating a thread is rather large,
also the cost of returning values can be large. These are areas
were there most likely will be work done to optimize what data
that needs to be cloned.

=head1 BUGS

=over

=item Parent-Child threads.

On some platforms it might not be possible to destroy "parent"
threads while there are still existing child "threads".

This will possibly be fixed in later versions of perl.
  
=item tid is I32

The thread id is a 32 bit integer, it can potentially overflow.
This might be fixed in a later version of perl.

=item Returning objects

When you return an object the entire stash that the object is blessed
as well.  This will lead to a large memory usage.  The ideal situation
would be to detect the original stash if it existed.

=item Creating threads inside BEGIN blocks

Creating threads inside BEGIN blocks (or during the compilation phase
in general) does not work.  (In Windows, trying to use fork() inside
BEGIN blocks is an equally losing proposition, since it has been
implemented in very much the same way as threads.)

=item PERL_OLD_SIGNALS are not threadsafe, will not be.

If your Perl has been built with PERL_OLD_SIGNALS (one has
to explicitly add that symbol to ccflags, see C<perl -V>),
signal handling is not threadsafe.

=back

=head1 AUTHOR and COPYRIGHT

Arthur Bergman E<lt>sky at nanisky.comE<gt>

threads is released under the same license as Perl.

Thanks to

Richard Soderberg E<lt>perl at crystalflame.netE<gt>
Helping me out tons, trying to find reasons for races and other weird bugs!

Simon Cozens E<lt>simon at brecon.co.ukE<gt>
Being there to answer zillions of annoying questions

Rocco Caputo E<lt>troc at netrus.netE<gt>

Vipul Ved Prakash E<lt>mail at vipul.netE<gt>
Helping with debugging.

please join perl-ithreads@perl.org for more information

=head1 SEE ALSO

L<threads::shared>, L<perlthrtut>, 
L<http://www.perl.com/pub/a/2002/06/11/threads.html>,
L<perlcall>, L<perlembed>, L<perlguts>

=cut
Revision:	1.1
Committed:	Thu Jun 30 14:27:06 2005 UTC (21 years, 3 months ago) by root
Branch:	MAIN
CVS Tags:	PERL-5-8-7, HEAD
Branch point for:	PERL
Log Message:	* empty log message *
#	Content
1	package threads;
2
3	use 5.008;
4	use strict;
5	use warnings;
6	use Config;
7
8	BEGIN {
9	unless ($Config{useithreads}) {
10	my @caller = caller(2);
11	die <<EOF;
12	$caller[1] line $caller[2]:
13
14	This Perl hasn't been configured and built properly for the threads
15	module to work. (The 'useithreads' configuration option hasn't been used.)
16
17	Having threads support requires all of Perl and all of the XS modules in
18	the Perl installation to be rebuilt, it is not just a question of adding
19	the threads module. (In other words, threaded and non-threaded Perls
20	are binary incompatible.)
21
22	If you want to the use the threads module, please contact the people
23	who built your Perl.
24
25	Cannot continue, aborting.
26	EOF
27	}
28	}
29
30	use overload
31	'==' => \&equal,
32	'fallback' => 1;
33
34	BEGIN {
35	warn "Warning, threads::shared has already been loaded. ".
36	"To enable shared variables for these modules 'use threads' ".
37	"must be called before any of those modules are loaded\n"
38	if($threads::shared::threads_shared);
39	}
40
41	require Exporter;
42	require DynaLoader;
43
44	our @ISA = qw(Exporter DynaLoader);
45
46	our %EXPORT_TAGS = ( all => [qw(yield)]);
47
48	our @EXPORT_OK = ( @{ $EXPORT_TAGS{'all'} } );
49
50	our @EXPORT = qw(
51	async
52	);
53	our $VERSION = '1.05';
54
55
56	# \|\| 0 to ensure compatibility with previous versions
57	sub equal { ($_[0]->tid == $_[1]->tid) \|\| 0 }
58
59	# use "goto" trick to avoid pad problems from 5.8.1 (fixed in 5.8.2)
60	# should also be faster
61	sub async (&;@) { unshift @_,'threads'; goto &new }
62
63	sub object {
64	return undef unless @_ > 1;
65	foreach (threads->list) {
66	return $_ if $_->tid == $_[1];
67	}
68	return undef;
69	}
70
71	$threads::threads = 1;
72
73	bootstrap threads $VERSION;
74
75	# why document 'new' then use 'create' in the tests!
76	*create = \&new;
77
78	# Preloaded methods go here.
79
80	1;
81	__END__
82
83	=head1 NAME
84
85	threads - Perl extension allowing use of interpreter based threads from perl
86
87	=head1 SYNOPSIS
88
89	use threads;
90
91	sub start_thread {
92	print "Thread started\n";
93	}
94
95	my $thread = threads->create("start_thread","argument");
96	my $thread2 = $thread->create(sub { print "I am a thread"},"argument");
97	my $thread3 = async { foreach (@files) { ... } };
98
99	$thread->join();
100	$thread->detach();
101
102	$thread = threads->self();
103	$thread = threads->object( $tid );
104
105	$thread->tid();
106	threads->tid();
107	threads->self->tid();
108
109	threads->yield();
110
111	threads->list();
112
113	=head1 DESCRIPTION
114
115	Perl 5.6 introduced something called interpreter threads. Interpreter
116	threads are different from "5005threads" (the thread model of Perl
117	5.005) by creating a new perl interpreter per thread and not sharing
118	any data or state between threads by default.
119
120	Prior to perl 5.8 this has only been available to people embedding
121	perl and for emulating fork() on windows.
122
123	The threads API is loosely based on the old Thread.pm API. It is very
124	important to note that variables are not shared between threads, all
125	variables are per default thread local. To use shared variables one
126	must use threads::shared.
127
128	It is also important to note that you must enable threads by doing
129	C<use threads> as early as possible in the script itself and that it
130	is not possible to enable threading inside an C<eval "">, C<do>,
131	C<require>, or C<use>. In particular, if you are intending to share
132	variables with threads::shared, you must C<use threads> before you
133	C<use threads::shared> and C<threads> will emit a warning if you do
134	it the other way around.
135
136	=over
137
138	=item $thread = threads->create(function, LIST)
139
140	This will create a new thread with the entry point function and give
141	it LIST as parameters. It will return the corresponding threads
142	object, or C<undef> if thread creation failed. The new() method is an
143	alias for create().
144
145	=item $thread->join
146
147	This will wait for the corresponding thread to join. When the thread
148	finishes, join() will return the return values of the entry point
149	function. If the thread has been detached, an error will be thrown.
150
151	The context (scalar or list) of the thread creation is also the
152	context for join(). This means that if you intend to return an array
153	from a thread, you must use C<my ($thread) = threads->new(...)>, and
154	that if you intend to return a scalar, you must use C<my $thread = ...>.
155
156	If the program exits without all other threads having been either
157	joined or detached, then a warning will be issued. (A program exits
158	either because one of its threads explicitly calls exit(), or in the
159	case of the main thread, reaches the end of the main program file.)
160
161
162	=item $thread->detach
163
164	Will make the thread unjoinable, and cause any eventual return value
165	to be discarded.
166
167	=item threads->self
168
169	This will return the thread object for the current thread.
170
171	=item $thread->tid
172
173	This will return the id of the thread. Thread IDs are integers, with
174	the main thread in a program being 0. Currently Perl assigns a unique
175	tid to every thread ever created in your program, assigning the first
176	thread to be created a tid of 1, and increasing the tid by 1 for each
177	new thread that's created.
178
179	NB the class method C<< threads->tid() >> is a quick way to get the
180	current thread id if you don't have your thread object handy.
181
182	=item threads->object( tid )
183
184	This will return the thread object for the thread associated with the
185	specified tid. Returns undef if there is no thread associated with the tid
186	or no tid is specified or the specified tid is undef.
187
188	=item threads->yield();
189
190	This is a suggestion to the OS to let this thread yield CPU time to other
191	threads. What actually happens is highly dependent upon the underlying
192	thread implementation.
193
194	You may do C<use threads qw(yield)> then use just a bare C<yield> in your
195	code.
196
197	=item threads->list();
198
199	This will return a list of all non joined, non detached threads.
200
201	=item async BLOCK;
202
203	C<async> creates a thread to execute the block immediately following
204	it. This block is treated as an anonymous sub, and so must have a
205	semi-colon after the closing brace. Like C<< threads->new >>, C<async>
206	returns a thread object.
207
208	=back
209
210	=head1 WARNINGS
211
212	=over 4
213
214	=item A thread exited while %d other threads were still running
215
216	A thread (not necessarily the main thread) exited while there were
217	still other threads running. Usually it's a good idea to first collect
218	the return values of the created threads by joining them, and only then
219	exit from the main thread.
220
221	=back
222
223	=head1 TODO
224
225	The current implementation of threads has been an attempt to get
226	a correct threading system working that could be built on,
227	and optimized, in newer versions of perl.
228
229	Currently the overhead of creating a thread is rather large,
230	also the cost of returning values can be large. These are areas
231	were there most likely will be work done to optimize what data
232	that needs to be cloned.
233
234	=head1 BUGS
235
236	=over
237
238	=item Parent-Child threads.
239
240	On some platforms it might not be possible to destroy "parent"
241	threads while there are still existing child "threads".
242
243	This will possibly be fixed in later versions of perl.
244
245	=item tid is I32
246
247	The thread id is a 32 bit integer, it can potentially overflow.
248	This might be fixed in a later version of perl.
249
250	=item Returning objects
251
252	When you return an object the entire stash that the object is blessed
253	as well. This will lead to a large memory usage. The ideal situation
254	would be to detect the original stash if it existed.
255
256	=item Creating threads inside BEGIN blocks
257
258	Creating threads inside BEGIN blocks (or during the compilation phase
259	in general) does not work. (In Windows, trying to use fork() inside
260	BEGIN blocks is an equally losing proposition, since it has been
261	implemented in very much the same way as threads.)
262
263	=item PERL_OLD_SIGNALS are not threadsafe, will not be.
264
265	If your Perl has been built with PERL_OLD_SIGNALS (one has
266	to explicitly add that symbol to ccflags, see C<perl -V>),
267	signal handling is not threadsafe.
268
269	=back
270
271	=head1 AUTHOR and COPYRIGHT
272
273	Arthur Bergman E<lt>sky at nanisky.comE<gt>
274
275	threads is released under the same license as Perl.
276
277	Thanks to
278
279	Richard Soderberg E<lt>perl at crystalflame.netE<gt>
280	Helping me out tons, trying to find reasons for races and other weird bugs!
281
282	Simon Cozens E<lt>simon at brecon.co.ukE<gt>
283	Being there to answer zillions of annoying questions
284
285	Rocco Caputo E<lt>troc at netrus.netE<gt>
286
287	Vipul Ved Prakash E<lt>mail at vipul.netE<gt>
288	Helping with debugging.
289
290	please join perl-ithreads@perl.org for more information
291
292	=head1 SEE ALSO
293
294	L<threads::shared>, L<perlthrtut>,
295	L<http://www.perl.com/pub/a/2002/06/11/threads.html>,
296	L<perlcall>, L<perlembed>, L<perlguts>
297
298	=cut