package IPC::Shareable; require 5.00503; use strict; use IPC::Semaphore; use IPC::Shareable::SharedMem; use IPC::SysV qw( IPC_PRIVATE IPC_CREAT IPC_EXCL IPC_NOWAIT SEM_UNDO ); use Storable 0.6 qw( freeze thaw ); use vars qw( $VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS ); $VERSION = 0.60; use constant LOCK_SH => 1; use constant LOCK_EX => 2; use constant LOCK_NB => 4; use constant LOCK_UN => 8; require Exporter; @ISA = 'Exporter'; @EXPORT = (); @EXPORT_OK = qw(LOCK_EX LOCK_SH LOCK_NB LOCK_UN); %EXPORT_TAGS = ( all => [qw( LOCK_EX LOCK_SH LOCK_NB LOCK_UN )], lock => [qw( LOCK_EX LOCK_SH LOCK_NB LOCK_UN )], 'flock' => [qw( LOCK_EX LOCK_SH LOCK_NB LOCK_UN )], ); Exporter::export_ok_tags('all', 'lock', 'flock'); use constant DEBUGGING => ($ENV{SHAREABLE_DEBUG} or 0); use constant SHM_BUFSIZ => 65536; use constant SEM_MARKER => 0; use constant SHM_EXISTS => 1; # Locking scheme copied from IPC::ShareLite -- ltl my %semop_args = ( (LOCK_EX), [ 1, 0, 0, # wait for readers to finish 2, 0, 0, # wait for writers to finish 2, 1, SEM_UNDO, # assert write lock ], (LOCK_EX|LOCK_NB), [ 1, 0, IPC_NOWAIT, # wait for readers to finish 2, 0, IPC_NOWAIT, # wait for writers to finish 2, 1, (SEM_UNDO | IPC_NOWAIT), # assert write lock ], (LOCK_EX|LOCK_UN), [ 2, -1, (SEM_UNDO | IPC_NOWAIT), ], (LOCK_SH), [ 2, 0, 0, # wait for writers to finish 1, 1, SEM_UNDO, # assert shared read lock ], (LOCK_SH|LOCK_NB), [ 2, 0, IPC_NOWAIT, # wait for writers to finish 1, 1, (SEM_UNDO | IPC_NOWAIT), # assert shared read lock ], (LOCK_SH|LOCK_UN), [ 1, -1, (SEM_UNDO | IPC_NOWAIT), # remove shared read lock ], ); my %Def_Opts = ( key => IPC_PRIVATE, create => '', exclusive => '', destroy => '', mode => 0666, size => SHM_BUFSIZ, ); # XXX Perl seems to garbage collect nested referents before we're done with them # XXX This cache holds a reference to things until END() is called my %Global_Reg; my %Proc_Reg; sub _trace; sub _debug; ############################################################################### # Debug mark # --- Public methods sub shlock { _trace @_ if DEBUGGING; my ($self, $typelock) = @_; ($typelock = LOCK_EX) unless defined $typelock; return $self->shunlock if ($typelock & LOCK_UN); return 1 if ($self->{_lock} & $typelock); # If they have a different lock than they want, release it first $self->shunlock if ($self->{_lock}); my $sem = $self->{_sem}; _debug "Attempting type=", $typelock, " lock on", $self->{_shm}, "via", $sem->id if DEBUGGING; my $return_val = $sem->op(@{ $semop_args{$typelock} }); if ($return_val) { $self->{_lock} = $typelock; _debug "Got lock on", $self->{_shm}, "via", $sem->id if DEBUGGING; $self->{_data} = _thaw($self->{_shm}), } else { _debug "Failed lock on", $self->{_shm}, "via", $sem->id if DEBUGGING; } return $return_val; } sub shunlock { _trace @_ if DEBUGGING; my $self = shift; return 1 unless $self->{_lock}; if ($self->{_was_changed}) { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!\n"; }; $self->{_was_changed} = 0; } my $sem = $self->{_sem}; _debug "Freeing lock on", $self->{_shm}, "via", $sem->id if DEBUGGING; my $typelock = $self->{_lock} | LOCK_UN; $typelock ^= LOCK_NB if ($typelock & LOCK_NB); $sem->op(@{ $semop_args{$typelock} }); $self->{_lock} = 0; _debug "Lock on", $self->{_shm}, "via", $sem->id, "freed" if DEBUGGING; 1; } # --- "Magic" methods sub TIESCALAR { _trace @_ if DEBUGGING; return _tie(SCALAR => @_); } sub TIEARRAY { _trace @_ if DEBUGGING; return _tie(ARRAY => @_); } sub TIEHASH { _trace @_ if DEBUGGING; return _tie(HASH => @_); } sub STORE { _trace @_ if DEBUGGING; my $self = shift; my $sid = $self->{_shm}->{_id}; $Global_Reg{$self->{_shm}->id} ||= $self; $self->{_data} = _thaw($self->{_shm}) unless ($self->{_lock}); TYPE: { if ($self->{_type} eq 'SCALAR') { my $val = shift; _mg_tie($self => $val) if _need_tie($val); $self->{_data} = \$val; last TYPE; } if ($self->{_type} eq 'ARRAY') { my $i = shift; my $val = shift; _mg_tie($self => $val) if _need_tie($val); $self->{_data}->[$i] = $val; last TYPE; } if ($self->{_type} eq 'HASH') { my $key = shift; my $val = shift; _mg_tie($self => $val) if _need_tie($val); $self->{_data}->{$key} = $val; last TYPE; } require Carp; Carp::croak "Variables of type $self->{_type} not supported"; } if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!\n"; }; } return 1; } sub FETCH { _trace @_ if DEBUGGING; my $self = shift; $Global_Reg{$self->{_shm}->id} ||= $self; my $data; if ($self->{_lock} || $self->{_iterating}) { $self->{_iterating} = ''; # In case we break out $data = $self->{_data}; } else { $data = _thaw($self->{_shm}); $self->{_data} = $data; } my $val; TYPE: { if ($self->{_type} eq 'SCALAR') { if (defined $data) { $val = $$data; last TYPE; } else { return; } } if ($self->{_type} eq 'ARRAY') { if (defined $data) { my $i = shift; $val = $data->[$i]; last TYPE; } else { return; } } if ($self->{_type} eq 'HASH') { if (defined $data) { my $key = shift; $val = $data->{$key}; last TYPE; } else { return; } } require Carp; Carp::croak "Variables of type $self->{_type} not supported"; } if (my $inner = _is_kid($val)) { my $s = $inner->{_shm}; $inner->{_data} = _thaw($s); } return $val; } sub CLEAR { _trace @_ if DEBUGGING; my $self = shift; if ($self->{_type} eq 'ARRAY') { $self->{_data} = [ ]; } elsif ($self->{_type} eq 'HASH') { $self->{_data} = { }; } else { require Carp; Carp::croak "Attempt to clear non-aggegrate"; } if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } } sub DELETE { _trace @_ if DEBUGGING; my $self = shift; my $key = shift; $self->{_data} = _thaw($self->{_shm}) unless $self->{_lock}; my $val = delete $self->{_data}->{$key}; if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } return $val; } sub EXISTS { _trace @_ if DEBUGGING; my $self = shift; my $key = shift; $self->{_data} = _thaw($self->{_shm}) unless $self->{_lock}; return exists $self->{_data}->{$key}; } sub FIRSTKEY { _trace @_ if DEBUGGING; my $self = shift; my $key = shift; _debug "setting hash iterator on", $self->{_shm}->id if DEBUGGING; $self->{_iterating} = 1; $self->{_data} = _thaw($self->{_shm}) unless $self->{_lock}; my $reset = keys %{$self->{_data}}; my $first = each %{$self->{_data}}; return $first; } sub NEXTKEY { _trace @_ if DEBUGGING; my $self = shift; # caveat emptor if hash was changed by another process my $next = each %{$self->{_data}}; if (not defined $next) { _debug "resetting hash iterator on", $self->{_shm}->id if DEBUGGING; $self->{_iterating} = ''; return; } else { $self->{_iterating} = 1; return $next; } } sub EXTEND { _trace @_ if DEBUGGING; #XXX Noop } sub PUSH { _trace @_ if DEBUGGING; my $self = shift; $Global_Reg{$self->{_shm}->id} ||= $self; $self->{_data} = _thaw($self->{_shm}, $self->{_data}) unless $self->{_lock}; push @{$self->{_data}} => @_; if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } } sub POP { _trace @_ if DEBUGGING; my $self = shift; $self->{_data} = _thaw($self->{_shm}, $self->{_data}) unless $self->{_lock}; my $val = pop @{$self->{_data}}; if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } return $val; } sub SHIFT { _trace @_ if DEBUGGING; my $self = shift; $self->{_data} = _thaw($self->{_shm}, $self->{_data}) unless $self->{_lock}; my $val = shift @{$self->{_data}}; if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } return $val; } sub UNSHIFT { _trace @_ if DEBUGGING; my $self = shift; $self->{_data} = _thaw($self->{_shm}, $self->{_data}) unless $self->{_lock}; my $val = unshift @{$self->{_data}} => @_; if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } return $val; } sub SPLICE { _trace @_ if DEBUGGING; my($self, $off, $n, @av) = @_; $self->{_data} = _thaw($self->{_shm}, $self->{_data}) unless $self->{_lock}; my @val = splice @{$self->{_data}}, $off, $n => @av; if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } return @val; } sub FETCHSIZE { _trace @_ if DEBUGGING; my $self = shift; $self->{_data} = _thaw($self->{_shm}) unless $self->{_lock}; return scalar(@{$self->{_data}}); } sub STORESIZE { _trace @_ if DEBUGGING; my $self = shift; my $n = shift; $self->{_data} = _thaw($self->{_shm}) unless $self->{_lock}; $#{@{$self->{_data}}} = $n - 1; if ($self->{_lock} & LOCK_EX) { $self->{_was_changed} = 1; } else { defined _freeze($self->{_shm} => $self->{_data}) or do { require Carp; Carp::croak "Could not write to shared memory: $!"; }; } return $n; } sub clean_up { _trace @_ if DEBUGGING; my $class = shift; for my $s (values %Proc_Reg) { next unless $s->{_opts}->{_owner} == $$; remove($s); } } sub clean_up_all { _trace @_ if DEBUGGING; my $class = shift; for my $s (values %Global_Reg) { remove($s); } } sub remove { _trace @_ if DEBUGGING; my $self = shift; my $s = $self->{_shm}; my $id = $s->id; $s->remove or do { require Carp; Carp::carp "Couldn't remove shared memory segment $id: $!"; }; $s = $self->{_sem}; $s->remove or do { require Carp; Carp::carp "Couldn't remove semaphore set $id: $!"; }; delete $Proc_Reg{$id}; delete $Global_Reg{$id}; } END { _trace @_ if DEBUGGING; for my $s (values %Proc_Reg) { shunlock($s); next unless $s->{_opts}->{destroy}; next unless $s->{_opts}->{_owner} == $$; remove($s); } } # --- Private methods below sub _freeze { _trace @_ if DEBUGGING; my $s = shift; my $water = shift; my $ice = freeze $water; # Could be a large string. No need to copy it. substr more efficient substr $ice, 0, 0, 'IPC::Shareable'; _debug "writing to shm segment ", $s->id, ": ", $ice if DEBUGGING; if (length($ice) > $s->size) { require Carp; Carp::croak "Length of shared data exceeds shared segment size"; }; $s->shmwrite($ice); } sub _thaw { _trace @_ if DEBUGGING; my $s = shift; my $ice = $s->shmread; _debug "read from shm segment ", $s->id, ": ", $ice if DEBUGGING; my $tag = substr $ice, 0, 14, ''; if ($tag eq 'IPC::Shareable') { my $water = thaw $ice; defined($water) or do { require Carp; Carp::croak "Munged shared memory segment (size exceeded?)"; }; return $water; } else { return; } } sub _tie { _trace @_ if DEBUGGING; my $type = shift; my $class = shift; my $opts = _parse_args(@_); my $key = _shm_key($opts); my $flags = _shm_flags($opts); my $shm_size = $opts->{size}; my $s = IPC::Shareable::SharedMem->new($key, $shm_size, $flags); defined $s or do { require Carp; Carp::croak "Could not create shared memory segment: $!\n"; }; _debug "shared memory id is", $s->id if DEBUGGING; my $sem = IPC::Semaphore->new($key, 3, $flags); defined $sem or do { require Carp; Carp::croak "Could not create semaphore set: $!\n"; }; _debug "semaphore id is", $sem->id if DEBUGGING; unless ( $sem->op(@{ $semop_args{(LOCK_SH)} }) ) { require Carp; Carp::croak "Could not obtain semaphore set lock: $!\n"; } my $sh = { _iterating => '', _key => $key, _lock => 0, _opts => $opts, _shm => $s, _sem => $sem, _type => $type, _was_changed => 0, }; $sh->{_data} = _thaw($s), my $there = $sem->getval(SEM_MARKER); if ($there == SHM_EXISTS) { _debug "binding to existing segment on ", $s->id if DEBUGGING; } else { _debug "brand new segment on ", $s->id if DEBUGGING; $Proc_Reg{$sh->{_shm}->id} ||= $sh; $sem->setval(SEM_MARKER, SHM_EXISTS) or do { require Carp; Carp::croak "Couldn't set semaphore during object creation: $!"; }; } $sem->op(@{ $semop_args{(LOCK_SH|LOCK_UN)} }); _debug "IPC::Shareable instance created:", $sh if DEBUGGING; return bless $sh => $class; } sub _parse_args { _trace @_ if DEBUGGING; my($proto, $opts) = @_; $proto = defined $proto ? $proto : 0; $opts = defined $opts ? $opts : { %Def_Opts }; if (ref $proto eq 'HASH') { $opts = $proto; } else { $opts->{key} = $proto; } for my $k (keys %Def_Opts) { if (not defined $opts->{$k}) { $opts->{$k} = $Def_Opts{$k}; } elsif ($opts->{$k} eq 'no') { if ($^W) { require Carp; Carp::carp("Use of `no' in IPC::Shareable args is obsolete"); } $opts->{$k} = ''; } } $opts->{_owner} = ($opts->{_owner} or $$); $opts->{_magic} = ($opts->{_magic} or ''); _debug "options are", $opts if DEBUGGING; return $opts; } sub _shm_key { _trace @_ if DEBUGGING; my $hv = shift; my $val = ($hv->{key} or ''); if ($val eq '') { return IPC_PRIVATE; } elsif ($val =~ /^\d+$/) { return $val; } else { # XXX This only uses the first four characters $val = pack A4 => $val; $val = unpack i => $val; return $val; } } sub _shm_flags { # --- Parses the anonymous hash passed to constructors; returns a list # --- of args suitable for passing to shmget _trace @_ if DEBUGGING; my $hv = shift; my $flags = 0; $flags |= IPC_CREAT if $hv->{create}; $flags |= IPC_EXCL if $hv->{exclusive}; $flags |= ($hv->{mode} or 0666); return $flags; } sub _mg_tie { _trace @_ if DEBUGGING; my $dad = shift; my $val = shift; # XXX How to generate a unique id ? my $key; if ($dad->{_key} == IPC_PRIVATE) { $key = IPC_PRIVATE; } else { $key = int(rand(1_000_000)); } my %opts = ( %{$dad->{_opts}}, key => $key, exclusive => 'yes', create => 'yes', _magic => 'yes' ); # XXX I wish I didn't have to take a copy of data here and copy it back in # XXX Also, have to peek inside potential objects to see their implementation my $kid; my $type = ref($val) ? "$val" : ''; if ($type =~ /SCALAR/) { my $copy = $$val; $kid = tie $$val => 'IPC::Shareable', $key, { %opts } or do { require Carp; Carp::croak "Could not create inner tie"; }; $$val = $copy; } elsif ($type =~ /ARRAY/) { my @copy = @$val; $kid = tie @$val => 'IPC::Shareable', $key, { %opts } or do { require Carp; Carp::croak "Could not create inner tie"; }; @$val = @copy; } elsif ($type =~ /HASH/) { my %copy = %$val; $kid = tie %$val => 'IPC::Shareable', $key, { %opts } or do { require Carp; Carp::croak "Could not create inner tie"; }; %$val = %copy; } else { require Carp; Carp::croak "Variables of type $type not implemented"; } return $kid; } sub _is_kid { my $data = shift; $data or return; my $type = "$data"; # XXX Have to peek inside potential objects to see their implementation my $obj; if ($type =~ /HASH/) { $obj = tied %$data; } elsif ($type =~ /ARRAY/) { $obj = tied @$data; } elsif ($type =~ /SCALAR/) { $obj = tied $$data; } if (ref $obj eq 'IPC::Shareable') { return $obj; } else { return; } } sub _need_tie { my $val = shift; # XXX Have to peek inside potential objects to see their implementation my $type = ref $val ? "$val" : ''; if ($type =~ /SCALAR/) { return !(tied $$val); } elsif ($type =~ /ARRAY/) { return !(tied @$val); } elsif ($type =~ /HASH/) { return !(tied %$val); } else { return; } } sub _trace { require Carp; require Data::Dumper; my $caller = ' ' . (caller(1))[3] . " called with:\n"; my $i = -1; my @msg = map { ++$i; my $obj; if (ref eq 'IPC::Shareable') { ' ' . "\$_[$i] = $_: shmid: $_->{_shm}->{_id}; " . Data::Dumper->Dump([ $_->{_opts} ], [ 'opts' ]); } else { ' ' . Data::Dumper->Dump( [ $_ ] => [ "\_[$i]" ]); } } @_; Carp::carp "IPC::Shareable ($$) debug:\n", $caller, @msg; } sub _debug { require Carp; require Data::Dumper; local $Data::Dumper::Terse = 1; my $caller = ' ' . (caller(1))[3] . " tells us that:\n"; my @msg = map { my $obj; if (ref eq 'IPC::Shareable') { ' ' . "$_: shmid: $_->{_shm}->{_id}; " . Data::Dumper->Dump([ $_->{_opts} ], [ 'opts' ]); } else { ' ' . Data::Dumper::Dumper($_); } } @_; Carp::carp "IPC::Shareable ($$) debug:\n", $caller, @msg; }; 1; __END__ =head1 NAME IPC::Shareable - share Perl variables between processes =head1 SYNOPSIS use IPC::Shareable (':lock'); tie SCALAR, 'IPC::Shareable', GLUE, OPTIONS; tie ARRAY, 'IPC::Shareable', GLUE, OPTIONS; tie HASH, 'IPC::Shareable', GLUE, OPTIONS; (tied VARIABLE)->shlock; (tied VARIABLE)->shunlock; (tied VARIABLE)->shlock(LOCK_SH|LOCK_NB) or print "resource unavailable\n"; (tied VARIABLE)->remove; IPC::Shareable->clean_up; IPC::Shareable->clean_up_all; =head1 CONVENTIONS The occurrence of a number in square brackets, as in [N], in the text of this document refers to a numbered note in the L. =head1 DESCRIPTION IPC::Shareable allows you to tie a variable to shared memory making it easy to share the contents of that variable with other Perl processes. Scalars, arrays, and hashes can be tied. The variable being tied may contain arbitrarily complex data structures - including references to arrays, hashes of hashes, etc. The association between variables in distinct processes is provided by GLUE. This is an integer number or 4 character string[1] that serves as a common identifier for data across process space. Hence the statement tie $scalar, 'IPC::Shareable', 'data'; in program one and the statement tie $variable, 'IPC::Shareable', 'data'; in program two will bind $scalar in program one and $variable in program two. There is no pre-set limit to the number of processes that can bind to data; nor is there a pre-set limit to the complexity of the underlying data of the tied variables[2]. The amount of data that can be shared within a single bound variable is limited by the system's maximum size for a shared memory segment (the exact value is system-dependent). The bound data structures are all linearized (using Raphael Manfredi's Storable module) before being slurped into shared memory. Upon retrieval, the original format of the data structure is recovered. Semaphore flags can be used for locking data between competing processes. =head1 OPTIONS Options are specified by passing a reference to a hash as the fourth argument to the tie() function that enchants a variable. Alternatively you can pass a reference to a hash as the third argument; IPC::Shareable will then look at the field named B in this hash for the value of GLUE. So, tie $variable, 'IPC::Shareable', 'data', \%options; is equivalent to tie $variable, 'IPC::Shareable', { key => 'data', ... }; Boolean option values can be specified using a value that evaluates to either true or false in the Perl sense. NOTE: Earlier versions allowed you to use the word B for true and the word B for false, but support for this "feature" is being removed. B will still act as true (since it is true, in the Perl sense), but use of the word B now emits an (optional) warning and then converts to a false value. This warning will become mandatory in a future release and then at some later date the use of B will stop working altogether. The following fields are recognized in the options hash. =over 4 =item B The B field is used to determine the GLUE when using the three-argument form of the call to tie(). This argument is then, in turn, used as the KEY argument in subsequent calls to shmget() and semget(). The default value is IPC_PRIVATE, meaning that your variables cannot be shared with other processes. =item B B is used to control whether calls to tie() create new shared memory segments or not. If B is set to a true value, IPC::Shareable will create a new binding associated with GLUE as needed. If B is false, IPC::Shareable will not attempt to create a new shared memory segment associated with GLUE. In this case, a shared memory segment associated with GLUE must already exist or the call to tie() will fail and return undef. The default is false. =item B If B field is set to a true value, calls to tie() will fail (returning undef) if a data binding associated with GLUE already exists. If set to a false value, calls to tie() will succeed even if a shared memory segment associated with GLUE already exists. The default is false =item B The I argument is an octal number specifying the access permissions when a new data binding is being created. These access permission are the same as file access permissions in that 0666 is world readable, 0600 is readable only by the effective UID of the process creating the shared variable, etc. The default is 0666 (world readable and writable). =item B If set to a true value, the shared memory segment underlying the data binding will be removed when the process calling tie() exits (gracefully)[3]. Use this option with care. In particular you should not use this option in a program that will fork after binding the data. On the other hand, shared memory is a finite resource and should be released if it is not needed. The default is false =item B This field may be used to specify the size of the shared memory segment allocated. The default is IPC::Shareable::SHM_BUFSIZ(). =back Default values for options are key => IPC_PRIVATE, create => 0, exclusive => 0, destroy => 0, mode => 0, size => IPC::Shareable::SHM_BUFSIZ(), =head1 LOCKING IPC::Shareable provides methods to implement application-level advisory locking of the shared data structures. These methods are called shlock() and shunlock(). To use them you must first get the object underlying the tied variable, either by saving the return value of the original call to tie() or by using the built-in tied() function. To lock a variable, do this: $knot = tie $sv, 'IPC::Shareable', $glue, { %options }; ... $knot->shlock; or equivalently tie($scalar, 'IPC::Shareable', $glue, { %options }); (tied $scalar)->shlock; This will place an exclusive lock on the data of $scalar. You can also get shared locks or attempt to get a lock without blocking. IPC::Shareable makes the constants LOCK_EX, LOCK_SH, LOCK_UN, and LOCK_NB exportable to your address space with the export tags C<:lock>, C<:flock>, or C<:all>. The values should be the same as the standard C option arguments. if ( (tied $scalar)->shlock(LOCK_SH|LOCK_NB) ) { print "The value is $scalar\n"; (tied $scalar)->shunlock; } else { print "Another process has an exlusive lock.\n"; } If no argument is provided to C, it defaults to LOCK_EX. To unlock a variable do this: $knot->shunlock; or (tied $scalar)->shunlock; or $knot->shlock(LOCK_UN); # Same as calling shunlock There are some pitfalls regarding locking and signals about which you should make yourself aware; these are discussed in L. If you use the advisory locking, IPC::Shareable assumes that you know what you are doing and attempts some optimizations. When you obtain a lock, either exclusive or shared, a fetch and thaw of the data is performed. No additional fetch/thaw operations are performed until you release the lock and access the bound variable again. During the time that the lock is kept, all accesses are perfomed on the copy in program memory. If other processes do not honor the lock, and update the shared memory region unfairly, the process with the lock will not be in sync. In other words, IPC::Shareable does not enforce the lock for you. A similar optimization is done if you obtain an exclusive lock. Updates to the shared memory region will be postponed until you release the lock (or downgrade to a shared lock). Use of locking can significantly improve performance for operations such as iterating over an array, retrieving a list from a slice or doing a slice assignment. =head1 REFERENCES When a reference to a non-tied scalar, hash, or array is assigned to a tie()d variable, IPC::Shareable will attempt to tie() the thingy being referenced[4]. This allows disparate processes to see changes to not only the top-level variable, but also changes to nested data. This feature is intended to be transparent to the application, but there are some caveats to be aware of. First of all, IPC::Shareable does not (yet) guarantee that the ids shared memory segments allocated automagically are unique. The more automagical tie()ing that happens, the greater the chance of a collision. Secondly, since a new shared memory segment is created for each thingy being referenced, the liberal use of references could cause the system to approach its limit for the total number of shared memory segments allowed. =head1 OBJECTS IPC::Shareable implements tie()ing objects to shared memory too. Since an object is just a reference, the same principles (and caveats) apply to tie()ing objects as other reference types. =head1 DESTRUCTION perl(1) will destroy the object underlying a tied variable when then tied variable goes out of scope. Unfortunately for IPC::Shareable, this may not be desirable: other processes may still need a handle on the relevant shared memory segment. IPC::Shareable therefore provides an interface to allow the application to control the timing of removal of shared memory segments. The interface consists of three methods - remove(), clean_up(), and clean_up_all() - and the B option to tie(). =over 4 =item B As described in L, specifying the B option when tie()ing a variable coerces IPC::Shareable to remove the underlying shared memory segment when the process calling tie() exits gracefully. Note that any related shared memory segments created automagically by the use of references will also be removed. =item B (tied $var)->remove; Calling remove() on the object underlying a tie()d variable removes the associated shared memory segment. The segment is removed irrespective of whether it has the B option set or not and irrespective of whether the calling process created the segment. =item B IPC::Shareable->clean_up; This is a class method that provokes IPC::Shareable to remove all shared memory segments created by the process. Segments not created by the calling process are not removed. =item B IPC::Shareable->clean_up_all; This is a class method that provokes IPC::Shareable to remove all shared memory segments encountered by the process. Segments are removed even if they were not created by the calling process. =back =head1 EXAMPLES In a file called B: #!/usr/bin/perl -w use strict; use IPC::Shareable; my $glue = 'data'; my %options = ( create => 'yes', exclusive => 0, mode => 0644, destroy => 'yes', ); my %colours; tie %colours, 'IPC::Shareable', $glue, { %options } or die "server: tie failed\n"; %colours = ( red => [ 'fire truck', 'leaves in the fall', ], blue => [ 'sky', 'police cars', ], ); ((print "server: there are 2 colours\n"), sleep 5) while scalar keys %colours == 2; print "server: here are all my colours:\n"; foreach my $c (keys %colours) { print "server: these are $c: ", join(', ', @{$colours{$c}}), "\n"; } exit; In a file called B #!/usr/bin/perl -w use strict; use IPC::Shareable; my $glue = 'data'; my %options = ( create => 0, exclusive => 0, mode => 0644, destroy => 0, ); my %colours; tie %colours, 'IPC::Shareable', $glue, { %options } or die "client: tie failed\n"; foreach my $c (keys %colours) { print "client: these are $c: ", join(', ', @{$colours{$c}}), "\n"; } delete $colours{'red'}; exit; And here is the output (the sleep commands in the command line prevent the output from being interrupted by shell prompts): bash$ ( ./server & ) ; sleep 10 ; ./client ; sleep 10 server: there are 2 colours server: there are 2 colours server: there are 2 colours client: these are blue: sky, police cars client: these are red: fire truck, leaves in the fall server: here are all my colours: server: these are blue: sky, police cars =head1 RETURN VALUES Calls to tie() that try to implement IPC::Shareable will return true if successful, I otherwise. The value returned is an instance of the IPC::Shareable class. =head1 AUTHOR Benjamin Sugars =head1 NOTES =head2 Footnotes from the above sections =over 4 =item 1 If GLUE is longer than 4 characters, only the 4 most significant characters are used. These characters are turned into integers by unpack()ing them. If GLUE is less than 4 characters, it is space padded. =item 2 IPC::Shareable provides no pre-set limits, but the system does. Namely, there are limits on the number of shared memory segments that can be allocated and the total amount of memory usable by shared memory. =item 3 If the process has been smoked by an untrapped signal, the binding will remain in shared memory. If you're cautious, you might try $SIG{INT} = \&catch_int; sub catch_int { die; } ... tie $variable, IPC::Shareable, 'data', { 'destroy' => 'Yes!' }; which will at least clean up after your user hits CTRL-C because IPC::Shareable's END method will be called. Or, maybe you'd like to leave the binding in shared memory, so subsequent process can recover the data... =item 4 This behaviour is markedly different from previous versions of IPC::Shareable. Older versions would sometimes tie() referenced thingies, and sometimes not. The new approach is more reliable (I think) and predictable (certainly) but uses more shared memory segments. =back =head2 General Notes =over 4 =item o When using shlock() to lock a variable, be careful to guard against signals. Under normal circumstances, IPC::Shareable's END method unlocks any locked variables when the process exits. However, if an untrapped signal is received while a process holds an exclusive lock, DESTROY will not be called and the lock may be maintained even though the process has exited. If this scares you, you might be better off implementing your own locking methods. One advantage of using C on some known file instead of the locking implemented with semaphores in IPC::Shareable is that when a process dies, it automatically releases any locks. This only happens with IPC::Shareable if the process dies gracefully. The alternative is to attempt to account for every possible calamitous ending for your process (robust signal handling in Perl is a source of much debate, though it usually works just fine) or to become familiar with your system's tools for removing shared memory and semaphores. This concern should be balanced against the significant performance improvements you can gain for larger data structures by using the locking mechanism implemented in IPC::Shareable. =item o There is a program called ipcs(1/8) (and ipcrm(1/8)) that is available on at least Solaris and Linux that might be useful for cleaning moribund shared memory segments or semaphore sets produced by bugs in either IPC::Shareable or applications using it. =item o This version of IPC::Shareable does not understand the format of shared memory segments created by versions prior to 0.60. If you try to tie to such segments, you will get an error. The only work around is to clear the shared memory segments and start with a fresh set. =item o Iterating over a hash causes a special optimization if you have not obtained a lock (it is better to obtain a read (or write) lock before iterating over a hash tied to Shareable, but we attempt this optimization if you do not). The fetch/thaw operation is performed when the first key is accessed. Subsequent key and and value accesses are done without accessing shared memory. Doing an assignment to the hash or fetching another value between key accesses causes the hash to be replaced from shared memory. The state of the iterator in this case is not defined by the Perl documentation. Caveat Emptor. =back =head1 CREDITS Thanks to all those with comments or bug fixes, especially Maurice Aubrey Stephane Bortzmeyer Doug MacEachern Robert Emmery Mohammed J. Kabir Terry Ewing Tim Fries Joe Thomas Paul Makepeace Raphael Manfredi Lee Lindley Dave Rolsky =head1 BUGS Certainly; this is beta software. When you discover an anomaly, send an email to me at bsugars@canoe.ca. =head1 SEE ALSO perl(1), perltie(1), Storable(3), shmget(2), ipcs(1), ipcrm(1) and other SysV IPC man pages. =cut