package threads; use 5.008; use strict; use warnings; our $VERSION = '1.33'; my $XS_VERSION = $VERSION; $VERSION = eval $VERSION; BEGIN { # Verify this Perl supports threads use Config; if (! $Config{useithreads}) { die("This Perl not built to support threads\n"); } # Declare that we have been loaded $threads::threads = 1; # Complain if 'threads' is loaded after 'threads::shared' if ($threads::shared::threads_shared) { warn <<'_MSG_'; Warning, threads::shared has already been loaded. To enable shared variables, 'use threads' must be called before threads::shared or any module that uses it. _MSG_ } } # Load the XS code require XSLoader; XSLoader::load('threads', $XS_VERSION); ### Export ### sub import { my $class = shift; # Not used # Exported subroutines my @EXPORT = qw(async); # Handle args while (my $sym = shift) { if ($sym =~ /^stack/) { threads->set_stack_size(shift); } elsif ($sym =~ /all/) { push(@EXPORT, qw(yield)); } else { push(@EXPORT, $sym); } } # Export subroutine names my $caller = caller(); foreach my $sym (@EXPORT) { no strict 'refs'; *{$caller.'::'.$sym} = \&{$sym}; } # Set stack size via environment variable if (exists($ENV{'PERL5_ITHREADS_STACK_SIZE'})) { threads->set_stack_size($ENV{'PERL5_ITHREADS_STACK_SIZE'}); } } ### Methods, etc. ### # Our own exit function/method sub exit { CORE::exit(0); } # 'new' is an alias for 'create' *new = \&create; # 'async' is a function alias for the 'threads->create()' method sub async (&;@) { unshift(@_, 'threads'); # Use "goto" trick to avoid pad problems from 5.8.1 (fixed in 5.8.2) goto &create; } # Thread object equality checking use overload ( '==' => \&equal, '!=' => sub { ! equal(@_) }, 'fallback' => 1 ); 1; __END__ =head1 NAME threads - Perl interpreter-based threads =head1 VERSION This document describes threads version 1.33 =head1 SYNOPSIS use threads ('yield', 'stack_size' => 64*4096); sub start_thread { my @args = @_; print('Thread started: ', join(' ', @args), "\n"); } my $thread = threads->create('start_thread', 'argument'); $thread->join(); threads->create(sub { print("I am a thread\n"); })->join(); my $thread3 = async { foreach (@files) { ... } }; $thread3->join(); # Invoke thread in list context (implicit) so it can return a list my ($thr) = threads->create(sub { return (qw/a b c/); }); # or specify list context explicitly my $thr = threads->create({'context' => 'list'}, sub { return (qw/a b c/); }); my @results = $thr->join(); $thread->detach(); $thread = threads->self(); $thread = threads->object($tid); $tid = threads->tid(); $tid = threads->self->tid(); $tid = $thread->tid(); threads->yield(); yield(); my @threads = threads->list(); my $thread_count = threads->list(); if ($thr1 == $thr2) { ... } $stack_size = threads->get_stack_size(); $old_size = threads->set_stack_size(32*4096); # Create a thread with a specific context and stack size my $thr = threads->create({ 'context' => 'list', 'stack_size' => 32*4096 }, \&foo); my @results = $thr->join(); $thr->kill('SIGUSR1'); threads->exit(); =head1 DESCRIPTION Perl 5.6 introduced something called interpreter threads. Interpreter threads are different from I<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 I 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 by default thread local. To use shared variables one must use L. It is also important to note that you must enable threads by doing C as early as possible in the script itself, and that it is not possible to enable threading inside an C, C, C, or C. In particular, if you are intending to share variables with L, you must C before you C. (C will emit a warning if you do it the other way around.) =over =item $thr = threads->create(FUNCTION, ARGS) This will create a new thread that will begin execution with the specified entry point function, and give it the I list as parameters. It will return the corresponding threads object, or C if thread creation failed. I may either be the name of a function, an anonymous subroutine, or a code ref. my $thr = threads->create('func_name', ...); # or my $thr = threads->create(sub { ... }, ...); # or my $thr = threads->create(\&func, ...); The C<-Enew()> method is an alias for C<-Ecreate()>. =item $thr->join() This will wait for the corresponding thread to complete its execution. When the thread finishes, C<-Ejoin()> will return the return value(s) of the entry point function. The context (void, scalar or list) for the return value(s) for C<-Ejoin()> is determined at the time of thread creation. # Create thread in list context (implicit) my ($thr1) = threads->create(sub { my @results = qw(a b c); return (@results); }); # or (explicit) my $thr1 = threads->create({'context' => 'list'}, sub { my @results = qw(a b c); return (@results); }); # Retrieve list results from thread my @res1 = $thr1->join(); # Create thread in scalar context (implicit) my $thr2 = threads->create(sub { my $result = 42; return ($result); }); # Retrieve scalar result from thread my $res2 = $thr2->join(); # Create a thread in void context (explicit) my $thr3 = threads->create({'void' => 1}, sub { print("Hello, world\n"); }); # Join the thread in void context (i.e., no return value) $thr3->join(); See L for more details. If the program exits without all threads having either been joined or detached, then a warning will be issued. Calling C<-Ejoin()> or C<-Edetach()> on an already joined thread will cause an error to be thrown. =item $thr->detach() Makes the thread unjoinable, and causes any eventual return value to be discarded. When the program exits, any detached threads that are still running are silently terminated. If the program exits without all threads having either been joined or detached, then a warning will be issued. Calling C<-Ejoin()> or C<-Edetach()> on an already detached thread will cause an error to be thrown. =item threads->detach() Class method that allows a thread to detach itself. =item threads->exit() The usual method for terminating a thread is to L from the entry point function with the appropriate return value(s). If needed, a thread can be exited at any time by calling Cexit()>. This will cause the thread to return C in a scalar context, or the empty list in a list context. Calling C in a thread indicates an abnormal exit for the thread. Any C<$SIG{__DIE__}> handler in the thread will be called first, and then the thread will exit with a warning message that will contain any arguments passed in the C call. Calling C in a thread is discouraged, but is equivalent to calling Cexit()>. If the desired affect is to truly terminate the application from a thread, then use L, if available. =item threads->self() Class method that allows a thread to obtain its own I object. =item $thr->tid() Returns the ID of the thread. Thread IDs are unique integers with the main thread in a program being 0, and incrementing by 1 for every thread created. =item threads->tid() Class method that allows a thread to obtain its own ID. =item threads->object($tid) This will return the I object for the I thread associated with the specified thread ID. Returns C if there is no thread associated with the TID, if the thread is joined or detached, if no TID is specified or if 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, and then just use C in your code. =item threads->list() In a list context, returns a list of all non-joined, non-detached I objects. In a scalar context, returns a count of the same. =item $thr1->equal($thr2) Tests if two threads objects are the same thread or not. This is overloaded to the more natural forms: if ($thr1 == $thr2) { print("Threads are the same\n"); } # or if ($thr1 != $thr2) { print("Threads differ\n"); } (Thread comparison is based on thread IDs.) =item async BLOCK; C creates a thread to execute the block immediately following it. This block is treated as an anonymous subroutine, and so must have a semi-colon after the closing brace. Like Ccreate()>, C returns a I object. =item $thr->_handle() This I method returns the memory location of the internal thread structure associated with a threads object. For Win32, this is a pointer to the C value returned by C (i.e., C); for other platforms, it is a pointer to the C structure used in the C call (i.e., C). This method is of no use for general Perl threads programming. Its intent is to provide other (XS-based) thread modules with the capability to access, and possibly manipulate, the underlying thread structure associated with a Perl thread. =item threads->_handle() Class method that allows a thread to obtain its own I. =back =head1 THREAD CONTEXT As with subroutines, the type of value returned from a thread's entry point function may be determined by the thread's I: list, scalar or void. The thread's context is determined at thread creation. This is necessary so that the context is available to the entry point function via L. The thread may then specify a value of the appropriate type to be returned from C<-Ejoin()>. =head2 Explicit context Because thread creation and thread joining may occur in different contexts, it may be desirable to state the context explicitly to the thread's entry point function. This may be done by calling C<-Ecreate()> with a parameter hash as the first argument: my $thr = threads->create({'context' => 'list'}, \&foo); ... my @results = $thr->join(); In the above, the threads object is returned to the parent thread in scalar context, and the thread's entry point function C will be called in list context such that the parent thread can receive a list from the C<-Ejoin()> call. Similarly, if you need the threads object, but your thread will not be returning a value (i.e., I context), you would do the following: my $thr = threads->create({'context' => 'void'}, \&foo); ... $thr->join(); The context type may also be used as the I in the parameter hash followed by a I value: threads->create({'scalar' => 1}, \&foo); ... my ($thr) = threads->list(); my $result = $thr->join(); =head2 Implicit context If not explicitly stated, the thread's context is implied from the context of the C<-Ecreate()> call: # Create thread in list context my ($thr) = threads->create(...); # Create thread in scalar context my $thr = threads->create(...); # Create thread in void context threads->create(...); =head1 THREAD STACK SIZE The default per-thread stack size for different platforms varies significantly, and is almost always far more than is needed for most applications. On Win32, Perl's makefile explicitly sets the default stack to 16 MB; on most other platforms, the system default is used, which again may be much larger than is needed. By tuning the stack size to more accurately reflect your application's needs, you may significantly reduce your application's memory usage, and increase the number of simultaneously running threads. N.B., on Windows, Address space allocation granularity is 64 KB, therefore, setting the stack smaller than that on Win32 Perl will not save any more memory. =over =item threads->get_stack_size(); Returns the current default per-thread stack size. The default is zero, which means the system default stack size is currently in use. =item $size = $thr->get_stack_size(); Returns the stack size for a particular thread. A return value of zero indicates the system default stack size was used for the thread. =item $old_size = threads->set_stack_size($new_size); Sets a new default per-thread stack size, and returns the previous setting. Some platforms have a minimum thread stack size. Trying to set the stack size below this value will result in a warning, and the minimum stack size will be used. Some Linux platforms have a maximum stack size. Setting too large of a stack size will cause thread creation to fail. If needed, C<$new_size> will be rounded up to the next multiple of the memory page size (usually 4096 or 8192). Threads created after the stack size is set will then either call C I<(for pthreads platforms)>, or supply the stack size to C I<(for Win32 Perl)>. (Obviously, this call does not affect any currently extant threads.) =item use threads ('stack_size' => VALUE); This sets the default per-thread stack size at the start of the application. =item $ENV{'PERL5_ITHREADS_STACK_SIZE'} The default per-thread stack size may be set at the start of the application through the use of the environment variable C: PERL5_ITHREADS_STACK_SIZE=1048576 export PERL5_ITHREADS_STACK_SIZE perl -e'use threads; print(threads->get_stack_size(), "\n")' This value overrides any C parameter given to C. Its primary purpose is to permit setting the per-thread stack size for legacy threaded applications. =item threads->create({'stack_size' => VALUE}, FUNCTION, ARGS) The stack size an individual threads may also be specified. This may be done by calling C<-Ecreate()> with a parameter hash as the first argument: my $thr = threads->create({'stack_size' => 32*4096}, \&foo, @args); =item $thr2 = $thr1->create(FUNCTION, ARGS) This creates a new thread (C<$thr2>) that inherits the stack size from an existing thread (C<$thr1>). This is shorthand for the following: my $stack_size = $thr1->get_stack_size(); my $thr2 = threads->create({'stack_size' => $stack_size}, FUNCTION, ARGS); =back =head1 THREAD SIGNALLING When safe signals is in effect (the default behavior - see L for more details), then signals may be sent and acted upon by individual threads. =over 4 =item $thr->kill('SIG...'); Sends the specified signal to the thread. Signal names and (positive) signal numbers are the same as those supported by L. For example, 'SIGTERM', 'TERM' and (depending on the OS) 15 are all valid arguments to C<-Ekill()>. Returns the thread object to allow for method chaining: $thr->kill('SIG...')->join(); =back Signal handlers need to be set up in the threads for the signals they are expected to act upon. Here's an example for I a thread: use threads; # Suppress warning message when thread is 'killed' no warnings 'threads'; sub thr_func { # Thread 'cancellation' signal handler $SIG{'KILL'} = sub { die("Thread killed\n"); }; ... } # Create a thread my $thr = threads->create('thr_func'); ... # Signal the thread to terminate, and then detach # it so that it will get cleaned up automatically $thr->kill('KILL')->detach(); Here's another simplistic example that illustrates the use of thread signalling in conjunction with a semaphore to provide rudimentary I and I capabilities: use threads; use Thread::Semaphore; sub thr_func { my $sema = shift; # Thread 'suspend/resume' signal handler $SIG{'STOP'} = sub { $sema->down(); # Thread suspended $sema->up(); # Thread resumes }; ... } # Create a semaphore and send it to a thread my $sema = Thread::Semaphore->new(); my $thr = threads->create('thr_func', $sema); # Suspend the thread $sema->down(); $thr->kill('STOP'); ... # Allow the thread to continue $sema->up(); CAVEAT: The thread signalling capability provided by this module does not actually send signals via the OS. It I signals at the Perl-level such that signal handlers are called in the appropriate thread. For example, sending C<$thr-Ekill('STOP')> does not actually suspend a thread (or the whole process), but does cause a C<$SIG{'STOP'}> handler to be called in that thread (as illustrated above). As such, signals that would normally not be appropriate to use in the C command (e.g., C) are okay to use with the C<-Ekill()> method (again, as illustrated above). Correspondingly, sending a signal to a thread does not disrupt the operation the thread is currently working on: The signal will be acted upon after the current operation has completed. For instance, if the thread is I on an I/O call, sending it a signal will not cause the I/O call to be interrupted such that the signal is acted up immediately. =head1 WARNINGS =over 4 =item Perl exited with active threads: If the program exits without all threads having either been joined or detached, then this warning will be issued. NOTE: This warning cannot be suppressed using C as suggested below. =item Thread creation failed: pthread_create returned # See the appropriate I page for C to determine the actual cause for the failure. =item Thread # terminated abnormally: ... A thread terminated in some manner other than just returning from its entry point function. For example, the thread may have terminated using C. =item Using minimum thread stack size of # Some platforms have a minimum thread stack size. Trying to set the stack size below this value will result in the above warning, and the stack size will be set to the minimum. =item Thread creation failed: pthread_attr_setstacksize(I) returned 22 The specified I exceeds the system's maximum stack size. Use a smaller value for the stack size. =back If needed, thread warnings can be suppressed by using: no warnings 'threads'; in the appropriate scope. =head1 ERRORS =over 4 =item This Perl not built to support threads The particular copy of Perl that you're trying to use was not built using the C configuration option. 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 L module (i.e., threaded and non-threaded Perls are binary incompatible.) =item Cannot change stack size of an existing thread The stack size of currently extant threads cannot be changed, therefore, the following results in the above error: $thr->set_stack_size($size); =item Cannot signal threads without safe signals Safe signals must be in effect to use the C<-Ekill()> signalling method. See L for more details. =item Unrecognized signal name: ... The particular copy of Perl that you're trying to use does not support the specified signal being used in a C<-Ekill()> call. =back =head1 BUGS =over =item Parent-child threads On some platforms, it might not be possible to destroy I threads while there are still existing I threads. =item Creating threads inside special blocks Creating threads inside C, C or C blocks should not be relied upon. Depending on the Perl version and the application code, results may range from success, to (apparently harmless) warnings of leaked scalar, or all the way up to crashing of the Perl interpreter. =item Unsafe signals Since Perl 5.8.0, signals have been made safer in Perl by postponing their handling until the interpreter is in a I state. See L and L for more details. Safe signals is the default behavior, and the old, immediate, unsafe signalling behavior is only in effect in the following situations: =over 4 =item * Perl was been built with C (see C). =item * The environment variable C is set to C (see L). =item * The module L is used. =back If unsafe signals is in effect, then signal handling is not thread-safe, and the C<-Ekill()> signalling method cannot be used. =item Returning closures from threads Returning closures from threads should not be relied upon. Depending of the Perl version and the application code, results may range from success, to (apparently harmless) warnings of leaked scalar, or all the way up to crashing of the Perl interpreter. =item Perl Bugs and the CPAN Version of L Support for threads extents beyond the code in this module (i.e., F and F), and into the Perl iterpreter itself. Older versions of Perl contain bugs that may manifest themselves despite using the latest version of L from CPAN. There is no workaround for this other than upgrading to the lastest version of Perl. (Before you consider posting a bug report, please consult, and possibly post a message to the discussion forum to see if what you've encountered is a known problem.) =back =head1 REQUIREMENTS Perl 5.8.0 or later =head1 SEE ALSO L Discussion Forum on CPAN: L Annotated POD for L: L L, L L and L Perl threads mailing list: L Stack size discussion: L =head1 AUTHOR Artur Bergman Esky AT crucially DOT netE threads is released under the same license as Perl. CPAN version produced by Jerry D. Hedden =head1 ACKNOWLEDGEMENTS Richard Soderberg Eperl AT crystalflame DOT netE - Helping me out tons, trying to find reasons for races and other weird bugs! Simon Cozens Esimon AT brecon DOT co DOT ukE - Being there to answer zillions of annoying questions Rocco Caputo Etroc AT netrus DOT netE Vipul Ved Prakash Email AT vipul DOT netE - Helping with debugging Dean Arnold Edarnold AT presicient DOT comE - Stack size API =cut