Commit | Line | Data |
d9bb3666 |
1 | package Thread; |
2 | require Exporter; |
3 | require DynaLoader; |
52e1cb5e |
4 | use vars qw($VERSION @ISA @EXPORT); |
5 | |
6 | $VERSION = "1.0"; |
7 | |
d9bb3666 |
8 | @ISA = qw(Exporter DynaLoader); |
8f4f90ac |
9 | @EXPORT_OK = qw(yield cond_signal cond_broadcast cond_wait async); |
734689b1 |
10 | |
d516a115 |
11 | =head1 NAME |
12 | |
13 | Thread - multithreading |
14 | |
15 | =head1 SYNOPSIS |
16 | |
17 | use Thread; |
18 | |
19 | my $t = new Thread \&start_sub, @start_args; |
20 | |
23a4f76c |
21 | $result = $t->join; |
22 | $result = $t->eval; |
23 | $t->detach; |
d516a115 |
24 | |
23a4f76c |
25 | if($t->equal($another_thread)) { |
26 | # ... |
27 | } |
d516a115 |
28 | |
23a4f76c |
29 | my $tid = Thread->self->tid; |
d516a115 |
30 | my $tlist = Thread->list; |
31 | |
32 | lock($scalar); |
23a4f76c |
33 | yield(); |
d516a115 |
34 | |
35 | use Thread 'async'; |
36 | |
d516a115 |
37 | =head1 DESCRIPTION |
38 | |
589fe9d5 |
39 | The C<Thread> module provides multithreading support for perl. |
40 | |
23a4f76c |
41 | WARNING: Threading is an experimental feature. Both the interface |
42 | and implementation are subject to change drastically. |
43 | |
589fe9d5 |
44 | =head1 FUNCTIONS |
45 | |
46 | =over 8 |
47 | |
48 | =item new \&start_sub |
49 | |
50 | =item new \&start_sub, LIST |
51 | |
52 | C<new> starts a new thread of execution in the referenced subroutine. The |
53 | optional list is passed as parameters to the subroutine. Execution |
54 | continues in both the subroutine and the code after the C<new> call. |
55 | |
56 | C<new Thread> returns a thread object representing the newly created |
57 | thread. |
58 | |
59 | =item lock VARIABLE |
60 | |
61 | C<lock> places a lock on a variable until the lock goes out of scope. If |
62 | the variable is locked by another thread, the C<lock> call will block until |
63 | it's available. C<lock> is recursive, so multiple calls to C<lock> are |
64 | safe--the variable will remain locked until the outermost lock on the |
65 | variable goes out of scope. |
66 | |
67 | Locks on variables only affect C<lock> calls--they do I<not> affect normal |
68 | access to a variable. (Locks on subs are different, and covered in a bit) |
69 | If you really, I<really> want locks to block access, then go ahead and tie |
70 | them to something and manage this yourself. This is done on purpose. While |
71 | managing access to variables is a good thing, perl doesn't force you out of |
72 | its living room... |
73 | |
74 | If a container object, such as a hash or array, is locked, all the elements |
75 | of that container are not locked. For example, if a thread does a C<lock |
76 | @a>, any other thread doing a C<lock($a[12])> won't block. |
77 | |
78 | You may also C<lock> a sub, using C<lock &sub>. Any calls to that sub from |
79 | another thread will block until the lock is released. This behaviour is not |
80 | equvalent to C<use attrs qw(locked)> in the sub. C<use attrs qw(locked)> |
81 | serializes access to a subroutine, but allows different threads |
82 | non-simultaneous access. C<lock &sub>, on the other hand, will not allow |
83 | I<any> other thread access for the duration of the lock. |
84 | |
85 | Finally, C<lock> will traverse up references exactly I<one> level. |
86 | C<lock(\$a)> is equivalent to C<lock($a)>, while C<lock(\\$a)> is not. |
87 | |
88 | =item async BLOCK; |
89 | |
90 | C<async> creates a thread to execute the block immediately following |
91 | it. This block is treated as an anonymous sub, and so must have a |
92 | semi-colon after the closing brace. Like C<new Thread>, C<async> returns a |
93 | thread object. |
94 | |
95 | =item Thread->self |
96 | |
97 | The C<Thread-E<gt>self> function returns a thread object that represents |
98 | the thread making the C<Thread-E<gt>self> call. |
99 | |
100 | =item Thread->list |
101 | |
102 | C<Thread-E<gt>list> returns a list of thread objects for all running and |
103 | finished but un-C<join>ed threads. |
104 | |
105 | =item cond_wait VARIABLE |
106 | |
107 | The C<cond_wait> function takes a B<locked> variable as a parameter, |
108 | unlocks the variable, and blocks until another thread does a C<cond_signal> |
109 | or C<cond_broadcast> for that same locked variable. The variable that |
110 | C<cond_wait> blocked on is relocked after the C<cond_wait> is satisfied. |
111 | If there are multiple threads C<cond_wait>ing on the same variable, all but |
112 | one will reblock waiting to reaquire the lock on the variable. (So if |
113 | you're only using C<cond_wait> for synchronization, give up the lock as |
114 | soon as possible) |
115 | |
116 | =item cond_signal VARIABLE |
117 | |
118 | The C<cond_signal> function takes a locked variable as a parameter and |
119 | unblocks one thread that's C<cond_wait>ing on that variable. If more than |
120 | one thread is blocked in a C<cond_wait> on that variable, only one (and |
121 | which one is indeterminate) will be unblocked. |
122 | |
123 | If there are no threads blocked in a C<cond_wait> on the variable, the |
124 | signal is discarded. |
125 | |
126 | =item cond_broadcast VARIABLE |
127 | |
128 | The C<cond_broadcast> function works similarly to C<cond_wait>. |
129 | C<cond_broadcast>, though, will unblock B<all> the threads that are blocked |
130 | in a C<cond_wait> on the locked variable, rather than only one. |
131 | |
23a4f76c |
132 | =item yield |
133 | |
134 | The C<yield> function allows another thread to take control of the |
135 | CPU. The exact results are implementation-dependent. |
136 | |
589fe9d5 |
137 | =back |
138 | |
139 | =head1 METHODS |
140 | |
141 | =over 8 |
142 | |
143 | =item join |
144 | |
145 | C<join> waits for a thread to end and returns any values the thread exited |
146 | with. C<join> will block until the thread has ended, though it won't block |
147 | if the thread has already terminated. |
148 | |
149 | If the thread being C<join>ed C<die>d, the error it died with will be |
150 | returned at this time. If you don't want the thread performing the C<join> |
151 | to die as well, you should either wrap the C<join> in an C<eval> or use the |
152 | C<eval> thread method instead of C<join>. |
153 | |
154 | =item eval |
155 | |
156 | The C<eval> method wraps an C<eval> around a C<join>, and so waits for a |
157 | thread to exit, passing along any values the thread might have returned. |
158 | Errors, of course, get placed into C<$@>. |
159 | |
23a4f76c |
160 | =item detach |
161 | |
162 | C<detach> tells a thread that it is never going to be joined i.e. |
163 | that all traces of its existence can be removed once it stops running. |
164 | Errors in detached threads will not be visible anywhere - if you want |
165 | to catch them, you should use $SIG{__DIE__} or something like that. |
166 | |
167 | =item equal |
168 | |
169 | C<equal> tests whether two thread objects represent the same thread and |
170 | returns true if they do. |
171 | |
589fe9d5 |
172 | =item tid |
173 | |
174 | The C<tid> method returns the tid of a thread. The tid is a monotonically |
175 | increasing integer assigned when a thread is created. The main thread of a |
176 | program will have a tid of zero, while subsequent threads will have tids |
177 | assigned starting with one. |
178 | |
179 | =head1 LIMITATIONS |
180 | |
181 | The sequence number used to assign tids is a simple integer, and no |
182 | checking is done to make sure the tid isn't currently in use. If a program |
183 | creates more than 2^32 - 1 threads in a single run, threads may be assigned |
184 | duplicate tids. This limitation may be lifted in a future version of Perl. |
d516a115 |
185 | |
186 | =head1 SEE ALSO |
187 | |
188 | L<attrs>, L<Thread::Queue>, L<Thread::Semaphore>, L<Thread::Specific>. |
189 | |
190 | =cut |
191 | |
734689b1 |
192 | # |
193 | # Methods |
194 | # |
195 | |
196 | # |
197 | # Exported functions |
198 | # |
199 | sub async (&) { |
200 | return new Thread $_[0]; |
201 | } |
d9bb3666 |
202 | |
458fb581 |
203 | sub eval { |
204 | return eval { shift->join; }; |
205 | } |
206 | |
d9bb3666 |
207 | bootstrap Thread; |
208 | |
d9bb3666 |
209 | 1; |