name
File::FcntlLock - File locking with fcntl(2)
SYNOPSIS
use File::FcntlLock;
my $fs = new Fcntl::FcntlLock;
$fs->l_type( F_RDLCK );
$fs->l_whence( SEEK_CUR );
$fs->l_start( 100 );
$fs->l_len( 123 );
my $fh;
open $fh, "<file_name" or die "Can't open file: $!\n";
$fs->lock( $fh, F_SETLK ) )
or print "Locking failed: " . $fs->error . "\n";
$fs->l_type( F_UNLCK );
$fs->lock( $fh, F_SETLK )
or print "Unlocking failed: " . $fs->error . "\n";
DESCRIPTION
File locking in Perl is usually done using the flock() function. Unfortunately, this only allows locks on whole files and is often implemented in terms of flock(2), which has some shortcomings.
Using this module file locking via fcntl(2) can be done (obviously, this restricts the use of the module to systems that have a fcntl(2) system call). Before a file (or parts of a file) can be locked, an object simulating a flock structure must be created and its properties set. Afterwards, by calling the lock()
method a lock can be set or it can be determined if and which process currently holds the lock.
To create a new object representing a flock structure call new()
:
$fs = new File::FcntlLock;
You also can pass the new()
method a set of key-value pairs to initialize the members of the flock structure, e.g.
$fs = new File::FcntlLock l_type => F_WRLCK,
l_whence => SEEK_SET,
l_start => 0,
l_len => 100;
if you plan to obtain a write lock for the first 100 bytes of a file.
Once you have created the object simulating the flock structure the following methods allow to query and in most cases also to modify the properties of the object.
l_type()
-
If called without an argument returns the current setting of the lock type, otherwise the lock type is set to the argument, which must be either
F_RDLCK
,F_WRLCK
orF_UNLCK
(for read lock, write lock or unlock). l_whence()
-
Queries or sets the
l_whence
member of the flock structure, determining if thel_start
value is relative to the start of the file, to the current position in the file or to the end of the file. The corresponding values areSEEK_SET
,SEEK_CUR
andSEEK_END
. See also the man page for lseek(2). l_start()
-
Queries or sets the start position (offset) of the lock in the file according to the mode selected by the
l_whence
member. See also the man page for lseek(2). l_len()
-
Queries or sets the length of the region (in bytes) in the file to be locked. A value of 0 means a lock (starting at
l_start
) to the very end of the file.According to SUSv3 negative values for
l_start
are allowed (resulting in a lock ranging froml_start + l_len
tol_start - 1
) Unfortunately, not all systems allow negative arguments and will return an error when you try to obtain the lock, so please read the fcntl(2) man page of your system carefully for details. l_pid()
-
This method allows to determine the PID of the process currently holding the lock after a call of
lock()
withF_GETLK
that indicated that another process is holding the lock.
When not initialized the flock structure member l_type
is set to F_RDLCK
by default, l_whence
to SEEK_SET
, and both l_start
and l_len
to 0, i.e. the settings for a read lock on the whole file.
After having set up the object representing a flock structure you can determine the current holder of a lock or try to obtain a lock by invoking the lock()
method with two arguments, a file handle (or a file descriptor, the module figures out automatically what it got) and a flag indicating the action to be taken, i.e.
$fs->lock( $fh, F_SETLK );
There are three values that can be used as the second argument:
F_GETLK
-
For
F_GETLK
thelock()
method determines if and who currently is holding the lock. If no other process is holding the lock thel_type
field is set toF_UNLCK
. Otherwise the flock structure object is set to the values that prevent us from obtaining a lock, with thel_pid
member set to the PID of the process holding the lock. F_SETLK
-
For
F_SETLK
thelock()
method tries to obtain the lock (whenl_type
is set to eitherF_WRLCK
orF_RDLCK
) or releases the lock (ifl_type
is set toF_UNLCK
). If a lock is held by some other proces the method call returnsundef
and errno is set toEACCESS
orEAGAIN
(please see the the man page for fcntl(2) for the details). F_SETLKW
-
is similar to
F_SETLK
but instead of returning an error if the lock can't be obtained immediately it blocks until the lock is obtained. If a signal is received while waiting for the lock the method returnsundef
and errno is set toEINTR
.
On success the method returns the string "0 but true". If the method fails (as indicated by an undef
return value) you can either immediately evaluate the error number (usingf $!, $ERRNO or $OS_ERROR) or check for it at some later time via the methods discussed below.
There are three methods for obtaining information about the reason the the last call of lock()
for the object failed:
lock_errno()
-
Returns the error number from the latest call of
lock()
. If the last call did not result in an error the method returnsundef
. error()
-
Returns a short description of the error that happened during the latest call of
lock()
with the object. Please take the messages with a grain of salt, they represent what SUSv3 (IEEE 1003.1-2001) and the Linux, TRUE64, OpenBSD3 and Solaris8 man pages tell what the error numbers mean, there could be differences (and additional error numbers) on other systems. If there was no error the method returnsundef
. system_error()
-
While the previous method,
error()
, tries to return a string with some relevance to the locking operation (i.e. "File or segment already locked by other process(es)" instead of "Permission denied") this method returns the "normal" system error message associated with errno. The method returnsundef
if there was no error.
EXPORT
F_GETLK F_SETLK F_SETLKW F_RDLCK F_WRLCK F_UNLCK SEEK_SET SEEK_CUR SEEK_END
CREDITS
Thanks to Mark-Jason Dominus (MJD) and Benjamin Goldberg (GOLDBB) for helpful discussions, code examples and encouragement.
AUTHOR
Jens Thoms Toerring <jt@toerring.de>