ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.12 by root, Mon Jul 11 01:03:17 2005 UTC vs.
Revision 1.187 by root, Fri Feb 11 00:05:17 2011 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use IO::AIO;	7	use IO::AIO;
8		8
9	aio_open "/etc/passwd", O_RDONLY, 0, sub {	9	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
10	my ($fh) = @_;	10	my $fh = shift
		11	or die "/etc/passwd: $!";
11	...	12	...
12	};	13	};
13		14
14	aio_unlink "/tmp/file", sub { };	15	aio_unlink "/tmp/file", sub { };
15		16
16	aio_read $fh, 30000, 1024, $buffer, 0, sub {	17	aio_read $fh, 30000, 1024, $buffer, 0, sub {
17	$_[0] > 0 or die "read error: $!";	18	$_[0] > 0 or die "read error: $!";
18	};	19	};
19		20
20	# Event	21	# version 2+ has request and group objects
21	Event->io (fd => IO::AIO::poll_fileno,	22	use IO::AIO 2;
22	poll => 'r',
23	cb => \&IO::AIO::poll_cb);
24		23
25	# Glib/Gtk2	24	aioreq_pri 4; # give next request a very high priority
26	add_watch Glib::IO IO::AIO::poll_fileno,	25	my $req = aio_unlink "/tmp/file", sub { };
27	\&IO::AIO::poll_cb;	26	$req->cancel; # cancel request if still in queue
28		27
29	# Tk	28	my $grp = aio_group sub { print "all stats done\n" };
30	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",	29	add $grp aio_stat "..." for ...;
31	readable => \&IO::AIO::poll_cb);
32
33	# Danga::Socket
34	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
35	\&IO::AIO::poll_cb);
36
37		30
38	=head1 DESCRIPTION	31	=head1 DESCRIPTION
39		32
40	This module implements asynchronous I/O using whatever means your	33	This module implements asynchronous I/O using whatever means your
41	operating system supports.	34	operating system supports. It is implemented as an interface to C<libeio>
		35	(L<http://software.schmorp.de/pkg/libeio.html>).
42		36
		37	Asynchronous means that operations that can normally block your program
		38	(e.g. reading from disk) will be done asynchronously: the operation
		39	will still block, but you can do something else in the meantime. This
		40	is extremely useful for programs that need to stay interactive even
		41	when doing heavy I/O (GUI programs, high performance network servers
		42	etc.), but can also be used to easily do operations in parallel that are
		43	normally done sequentially, e.g. stat'ing many files, which is much faster
		44	on a RAID volume or over NFS when you do a number of stat operations
		45	concurrently.
		46
		47	While most of this works on all types of file descriptors (for
		48	example sockets), using these functions on file descriptors that
		49	support nonblocking operation (again, sockets, pipes etc.) is
		50	very inefficient. Use an event loop for that (such as the L<EV>
		51	module): IO::AIO will naturally fit into such an event loop itself.
		52
43	Currently, a number of threads are started that execute your read/writes	53	In this version, a number of threads are started that execute your
44	and signal their completion. You don't need thread support in your libc or	54	requests and signal their completion. You don't need thread support
45	perl, and the threads created by this module will not be visible to the	55	in perl, and the threads created by this module will not be visible
46	pthreads library. In the future, this module might make use of the native	56	to perl. In the future, this module might make use of the native aio
47	aio functions available on many operating systems. However, they are often	57	functions available on many operating systems. However, they are often
48	not well-supported (Linux doesn't allow them on normal files currently,	58	not well-supported or restricted (GNU/Linux doesn't allow them on normal
49	for example), and they would only support aio_read and aio_write, so the	59	files currently, for example), and they would only support aio_read and
50	remaining functionality would have to be implemented using threads anyway.	60	aio_write, so the remaining functionality would have to be implemented
		61	using threads anyway.
51		62
52	Although the module will work with in the presence of other threads, it is	63	Although the module will work in the presence of other (Perl-) threads,
53	currently not reentrant, so use appropriate locking yourself.	64	it is currently not reentrant in any way, so use appropriate locking
		65	yourself, always call C<poll_cb> from within the same thread, or never
		66	call C<poll_cb> (or other C<aio_> functions) recursively.
		67
		68	=head2 EXAMPLE
		69
		70	This is a simple example that uses the EV module and loads
		71	F</etc/passwd> asynchronously:
		72
		73	use Fcntl;
		74	use EV;
		75	use IO::AIO;
		76
		77	# register the IO::AIO callback with EV
		78	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		79
		80	# queue the request to open /etc/passwd
		81	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
		82	my $fh = shift
		83	or die "error while opening: $!";
		84
		85	# stat'ing filehandles is generally non-blocking
		86	my $size = -s $fh;
		87
		88	# queue a request to read the file
		89	my $contents;
		90	aio_read $fh, 0, $size, $contents, 0, sub {
		91	$_[0] == $size
		92	or die "short read: $!";
		93
		94	close $fh;
		95
		96	# file contents now in $contents
		97	print $contents;
		98
		99	# exit event loop and program
		100	EV::unloop;
		101	};
		102	};
		103
		104	# possibly queue up other requests, or open GUI windows,
		105	# check for sockets etc. etc.
		106
		107	# process events as long as there are some:
		108	EV::loop;
		109
		110	=head1 REQUEST ANATOMY AND LIFETIME
		111
		112	Every C<aio_*> function creates a request. which is a C data structure not
		113	directly visible to Perl.
		114
		115	If called in non-void context, every request function returns a Perl
		116	object representing the request. In void context, nothing is returned,
		117	which saves a bit of memory.
		118
		119	The perl object is a fairly standard ref-to-hash object. The hash contents
		120	are not used by IO::AIO so you are free to store anything you like in it.
		121
		122	During their existance, aio requests travel through the following states,
		123	in order:
		124
		125	=over 4
		126
		127	=item ready
		128
		129	Immediately after a request is created it is put into the ready state,
		130	waiting for a thread to execute it.
		131
		132	=item execute
		133
		134	A thread has accepted the request for processing and is currently
		135	executing it (e.g. blocking in read).
		136
		137	=item pending
		138
		139	The request has been executed and is waiting for result processing.
		140
		141	While request submission and execution is fully asynchronous, result
		142	processing is not and relies on the perl interpreter calling C<poll_cb>
		143	(or another function with the same effect).
		144
		145	=item result
		146
		147	The request results are processed synchronously by C<poll_cb>.
		148
		149	The C<poll_cb> function will process all outstanding aio requests by
		150	calling their callbacks, freeing memory associated with them and managing
		151	any groups they are contained in.
		152
		153	=item done
		154
		155	Request has reached the end of its lifetime and holds no resources anymore
		156	(except possibly for the Perl object, but its connection to the actual
		157	aio request is severed and calling its methods will either do nothing or
		158	result in a runtime error).
		159
		160	=back
54		161
55	=cut	162	=cut
56		163
57	package IO::AIO;	164	package IO::AIO;
58		165
		166	use Carp ();
		167
		168	use common::sense;
		169
59	use base 'Exporter';	170	use base 'Exporter';
60		171
61	use Fcntl ();
62
63	BEGIN {	172	BEGIN {
64	$VERSION = 0.3;	173	our $VERSION = '3.72';
65		174
66	@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink	175	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close
67	aio_fsync aio_fdatasync aio_readahead);	176	aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
68	@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	177	aio_scandir aio_symlink aio_readlink aio_sync aio_fsync
		178	aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead
		179	aio_rename aio_link aio_move aio_copy aio_group
		180	aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
		181	aio_chmod aio_utime aio_truncate
		182	aio_msync aio_mtouch aio_mlock aio_mlockall
		183	aio_statvfs);
		184
		185	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		186	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		187	min_parallel max_parallel max_idle
		188	nreqs nready npending nthreads
		189	max_poll_time max_poll_reqs
		190	sendfile fadvise madvise
		191	mmap munmap munlock munlockall);
		192
		193	push @AIO_REQ, qw(aio_busy); # not exported
		194
		195	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
69		196
70	require XSLoader;	197	require XSLoader;
71	XSLoader::load IO::AIO, $VERSION;	198	XSLoader::load ("IO::AIO", $VERSION);
72	}	199	}
73		200
74	=head1 FUNCTIONS	201	=head1 FUNCTIONS
75		202
		203	=head2 QUICK OVERVIEW
		204
		205	This section simply lists the prototypes of the most important functions
		206	for quick reference. See the following sections for function-by-function
		207	documentation.
		208
		209	aio_open $pathname, $flags, $mode, $callback->($fh)
		210	aio_close $fh, $callback->($status)
		211	aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		212	aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		213	aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		214	aio_readahead $fh,$offset,$length, $callback->($retval)
		215	aio_stat $fh_or_path, $callback->($status)
		216	aio_lstat $fh, $callback->($status)
		217	aio_statvfs $fh_or_path, $callback->($statvfs)
		218	aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		219	aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		220	aio_truncate $fh_or_path, $offset, $callback->($status)
		221	aio_chmod $fh_or_path, $mode, $callback->($status)
		222	aio_unlink $pathname, $callback->($status)
		223	aio_mknod $path, $mode, $dev, $callback->($status)
		224	aio_link $srcpath, $dstpath, $callback->($status)
		225	aio_symlink $srcpath, $dstpath, $callback->($status)
		226	aio_readlink $path, $callback->($link)
		227	aio_rename $srcpath, $dstpath, $callback->($status)
		228	aio_mkdir $pathname, $mode, $callback->($status)
		229	aio_rmdir $pathname, $callback->($status)
		230	aio_readdir $pathname, $callback->($entries)
		231	aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		232	IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
		233	IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
		234	aio_load $path, $data, $callback->($status)
		235	aio_copy $srcpath, $dstpath, $callback->($status)
		236	aio_move $srcpath, $dstpath, $callback->($status)
		237	aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		238	aio_rmtree $path, $callback->($status)
		239	aio_sync $callback->($status)
		240	aio_fsync $fh, $callback->($status)
		241	aio_fdatasync $fh, $callback->($status)
		242	aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		243	aio_pathsync $path, $callback->($status)
		244	aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		245	aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		246	aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		247	aio_mlockall $flags, $callback->($status)
		248	aio_group $callback->(...)
		249	aio_nop $callback->()
		250
		251	$prev_pri = aioreq_pri [$pri]
		252	aioreq_nice $pri_adjust
		253
		254	IO::AIO::poll_wait
		255	IO::AIO::poll_cb
		256	IO::AIO::poll
		257	IO::AIO::flush
		258	IO::AIO::max_poll_reqs $nreqs
		259	IO::AIO::max_poll_time $seconds
		260	IO::AIO::min_parallel $nthreads
		261	IO::AIO::max_parallel $nthreads
		262	IO::AIO::max_idle $nthreads
		263	IO::AIO::max_outstanding $maxreqs
		264	IO::AIO::nreqs
		265	IO::AIO::nready
		266	IO::AIO::npending
		267
		268	IO::AIO::sendfile $ofh, $ifh, $offset, $count
		269	IO::AIO::fadvise $fh, $offset, $len, $advice
		270	IO::AIO::madvise $scalar, $offset, $length, $advice
		271	IO::AIO::mprotect $scalar, $offset, $length, $protect
		272	IO::AIO::munlock $scalar, $offset = 0, $length = undef
		273	IO::AIO::munlockall
		274
76	=head2 AIO FUNCTIONS	275	=head2 AIO REQUEST FUNCTIONS
77		276
78	All the C<aio_*> calls are more or less thin wrappers around the syscall	277	All the C<aio_*> calls are more or less thin wrappers around the syscall
79	with the same name (sans C<aio_>). The arguments are similar or identical,	278	with the same name (sans C<aio_>). The arguments are similar or identical,
80	and they all accept an additional C<$callback> argument which must be	279	and they all accept an additional (and optional) C<$callback> argument
81	a code reference. This code reference will get called with the syscall	280	which must be a code reference. This code reference will get called with
82	return code (e.g. most syscalls return C<-1> on error, unlike perl, which	281	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
83	usually delivers "false") as it's sole argument when the given syscall has	282	perl, which usually delivers "false") as its sole argument after the given
84	been executed asynchronously.	283	syscall has been executed asynchronously.
85		284
86	All functions that expect a filehandle will also accept a file descriptor.	285	All functions expecting a filehandle keep a copy of the filehandle
		286	internally until the request has finished.
87		287
		288	All functions return request objects of type L<IO::AIO::REQ> that allow
		289	further manipulation of those requests while they are in-flight.
		290
88	The filenames you pass to these routines I<must> be absolute. The reason	291	The pathnames you pass to these routines I<must> be absolute and
89	is that at the time the request is being executed, the current working	292	encoded as octets. The reason for the former is that at the time the
90	directory could have changed. Alternatively, you can make sure that you	293	request is being executed, the current working directory could have
91	never change the current working directory.	294	changed. Alternatively, you can make sure that you never change the
		295	current working directory anywhere in the program and then use relative
		296	paths.
		297
		298	To encode pathnames as octets, either make sure you either: a) always pass
		299	in filenames you got from outside (command line, readdir etc.) without
		300	tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
		301	your pathnames to the locale (or other) encoding in effect in the user
		302	environment, d) use Glib::filename_from_unicode on unicode filenames or e)
		303	use something else to ensure your scalar has the correct contents.
		304
		305	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		306	handles correctly whether it is set or not.
92		307
93	=over 4	308	=over 4
94		309
		310	=item $prev_pri = aioreq_pri [$pri]
		311
		312	Returns the priority value that would be used for the next request and, if
		313	C<$pri> is given, sets the priority for the next aio request.
		314
		315	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		316	and C<4>, respectively. Requests with higher priority will be serviced
		317	first.
		318
		319	The priority will be reset to C<0> after each call to one of the C<aio_*>
		320	functions.
		321
		322	Example: open a file with low priority, then read something from it with
		323	higher priority so the read request is serviced before other low priority
		324	open requests (potentially spamming the cache):
		325
		326	aioreq_pri -3;
		327	aio_open ..., sub {
		328	return unless $_[0];
		329
		330	aioreq_pri -2;
		331	aio_read $_[0], ..., sub {
		332	...
		333	};
		334	};
		335
		336
		337	=item aioreq_nice $pri_adjust
		338
		339	Similar to C<aioreq_pri>, but subtracts the given value from the current
		340	priority, so the effect is cumulative.
		341
		342
95	=item aio_open $pathname, $flags, $mode, $callback	343	=item aio_open $pathname, $flags, $mode, $callback->($fh)
96		344
97	Asynchronously open or create a file and call the callback with a newly	345	Asynchronously open or create a file and call the callback with a newly
98	created filehandle for the file.	346	created filehandle for the file.
99		347
100	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,	348	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
101	for an explanation.	349	for an explanation.
102		350
103	The C<$mode> argument is a bitmask. See the C<Fcntl> module for a	351	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
104	list. They are the same as used in C<sysopen>.	352	list. They are the same as used by C<sysopen>.
		353
		354	Likewise, C<$mode> specifies the mode of the newly created file, if it
		355	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
		356	except that it is mandatory (i.e. use C<0> if you don't create new files,
		357	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		358	by the umask in effect then the request is being executed, so better never
		359	change the umask.
105		360
106	Example:	361	Example:
107		362
108	aio_open "/etc/passwd", O_RDONLY, 0, sub {	363	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
109	if ($_[0]) {	364	if ($_[0]) {
110	print "open successful, fh is $_[0]\n";	365	print "open successful, fh is $_[0]\n";
111	...	366	...
112	} else {	367	} else {
113	die "open failed: $!\n";	368	die "open failed: $!\n";
114	}	369	}
115	};	370	};
116		371
		372
117	=item aio_close $fh, $callback	373	=item aio_close $fh, $callback->($status)
118		374
119	Asynchronously close a file and call the callback with the result	375	Asynchronously close a file and call the callback with the result
120	code. I<WARNING:> although accepted, you should not pass in a perl	376	code.
121	filehandle here, as perl will likely close the file descriptor itself when
122	the filehandle is destroyed. Normally, you can safely call perls C<close>
123	or just let filehandles go out of scope.
124		377
		378	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
		379	closing the file descriptor associated with the filehandle itself.
		380
		381	Therefore, C<aio_close> will not close the filehandle - instead it will
		382	use dup2 to overwrite the file descriptor with the write-end of a pipe
		383	(the pipe fd will be created on demand and will be cached).
		384
		385	Or in other words: the file descriptor will be closed, but it will not be
		386	free for reuse until the perl filehandle is closed.
		387
		388	=cut
		389
125	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	390	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
126		391
127	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	392	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
128		393
129	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	394	Reads or writes C<$length> bytes from or to the specified C<$fh> and
130	into the scalar given by C<data> and offset C<dataoffset> and calls the	395	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset>
131	callback without the actual number of bytes read (or -1 on error, just	396	and calls the callback without the actual number of bytes read (or -1 on
132	like the syscall).	397	error, just like the syscall).
133		398
		399	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		400	offset plus the actual number of bytes read.
		401
		402	If C<$offset> is undefined, then the current file descriptor offset will
		403	be used (and updated), otherwise the file descriptor offset will not be
		404	changed by these calls.
		405
		406	If C<$length> is undefined in C<aio_write>, use the remaining length of
		407	C<$data>.
		408
		409	If C<$dataoffset> is less than zero, it will be counted from the end of
		410	C<$data>.
		411
		412	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		413	is outstanding. Modifying it can result in segfaults or World War III (if
		414	the necessary/optional hardware is installed).
		415
134	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at	416	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
135	offset C<0> within the scalar:	417	offset C<0> within the scalar:
136		418
137	aio_read $fh, 7, 15, $buffer, 0, sub {	419	aio_read $fh, 7, 15, $buffer, 0, sub {
138	$_[0] > 0 or die "read error: $!";	420	$_[0] > 0 or die "read error: $!";
139	print "read $_[0] bytes: <$buffer>\n";	421	print "read $_[0] bytes: <$buffer>\n";
140	};	422	};
141		423
		424
		425	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		426
		427	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		428	reading at byte offset C<$in_offset>, and starts writing at the current
		429	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		430	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		431	other.
		432
		433	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		434	are written, and there is no way to find out how many bytes have been read
		435	from C<aio_sendfile> alone, as C<aio_sendfile> only provides the number of
		436	bytes written to C<$out_fh>. Only if the result value equals C<$length>
		437	one can assume that C<$length> bytes have been read.
		438
		439	Unlike with other C<aio_> functions, it makes a lot of sense to use
		440	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		441	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		442	the socket I/O will be non-blocking. Note, however, that you can run into
		443	a trap where C<aio_sendfile> reads some data with readahead, then fails
		444	to write all data, and when the socket is ready the next time, the data
		445	in the cache is already lost, forcing C<aio_sendfile> to again hit the
		446	disk. Explicit C<aio_read> + C<aio_write> let's you control resource usage
		447	much better.
		448
		449	This call tries to make use of a native C<sendfile> syscall to provide
		450	zero-copy operation. For this to work, C<$out_fh> should refer to a
		451	socket, and C<$in_fh> should refer to an mmap'able file.
		452
		453	If a native sendfile cannot be found or it fails with C<ENOSYS>,
		454	C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or C<ENOTSOCK>,
		455	it will be emulated, so you can call C<aio_sendfile> on any type of
		456	filehandle regardless of the limitations of the operating system.
		457
		458
142	=item aio_readahead $fh,$offset,$length, $callback	459	=item aio_readahead $fh,$offset,$length, $callback->($retval)
143		460
144	Asynchronously reads the specified byte range into the page cache, using
145	the C<readahead> syscall. If that syscall doesn't exist the status will be
146	C<-1> and C<$!> is set to ENOSYS.
147
148	readahead() populates the page cache with data from a file so that	461	C<aio_readahead> populates the page cache with data from a file so that
149	subsequent reads from that file will not block on disk I/O. The C<$offset>	462	subsequent reads from that file will not block on disk I/O. The C<$offset>
150	argument specifies the starting point from which data is to be read and	463	argument specifies the starting point from which data is to be read and
151	C<$length> specifies the number of bytes to be read. I/O is performed in	464	C<$length> specifies the number of bytes to be read. I/O is performed in
152	whole pages, so that offset is effectively rounded down to a page boundary	465	whole pages, so that offset is effectively rounded down to a page boundary
153	and bytes are read up to the next page boundary greater than or equal to	466	and bytes are read up to the next page boundary greater than or equal to
154	(off-set+length). aio_readahead() does not read beyond the end of the	467	(off-set+length). C<aio_readahead> does not read beyond the end of the
155	file. The current file offset of the file is left unchanged.	468	file. The current file offset of the file is left unchanged.
156		469
		470	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
		471	emulated by simply reading the data, which would have a similar effect.
		472
		473
157	=item aio_stat $fh_or_path, $callback	474	=item aio_stat $fh_or_path, $callback->($status)
158		475
159	=item aio_lstat $fh, $callback	476	=item aio_lstat $fh, $callback->($status)
160		477
161	Works like perl's C<stat> or C<lstat> in void context. The callback will	478	Works like perl's C<stat> or C<lstat> in void context. The callback will
162	be called after the stat and the results will be available using C<stat _>	479	be called after the stat and the results will be available using C<stat _>
163	or C<-s _> etc...	480	or C<-s _> etc...
164		481
…		…
166	for an explanation.	483	for an explanation.
167		484
168	Currently, the stats are always 64-bit-stats, i.e. instead of returning an	485	Currently, the stats are always 64-bit-stats, i.e. instead of returning an
169	error when stat'ing a large file, the results will be silently truncated	486	error when stat'ing a large file, the results will be silently truncated
170	unless perl itself is compiled with large file support.	487	unless perl itself is compiled with large file support.
		488
		489	To help interpret the mode and dev/rdev stat values, IO::AIO offers the
		490	following constants and functions (if not implemented, the constants will
		491	be C<0> and the functions will either C<croak> or fall back on traditional
		492	behaviour).
		493
		494	C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
		495	C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
		496	C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
171		497
172	Example: Print the length of F</etc/passwd>:	498	Example: Print the length of F</etc/passwd>:
173		499
174	aio_stat "/etc/passwd", sub {	500	aio_stat "/etc/passwd", sub {
175	$_[0] and die "stat failed: $!";	501	$_[0] and die "stat failed: $!";
176	print "size is ", -s _, "\n";	502	print "size is ", -s _, "\n";
177	};	503	};
178		504
		505
		506	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		507
		508	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		509	whether a file handle or path was passed.
		510
		511	On success, the callback is passed a hash reference with the following
		512	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		513	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		514	is passed.
		515
		516	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		517	C<ST_NOSUID>.
		518
		519	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		520	their correct value when available, or to C<0> on systems that do
		521	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		522	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		523	C<ST_NODIRATIME> and C<ST_RELATIME>.
		524
		525	Example: stat C</wd> and dump out the data if successful.
		526
		527	aio_statvfs "/wd", sub {
		528	my $f = $_[0]
		529	or die "statvfs: $!";
		530
		531	use Data::Dumper;
		532	say Dumper $f;
		533	};
		534
		535	# result:
		536	{
		537	bsize => 1024,
		538	bfree => 4333064312,
		539	blocks => 10253828096,
		540	files => 2050765568,
		541	flag => 4096,
		542	favail => 2042092649,
		543	bavail => 4333064312,
		544	ffree => 2042092649,
		545	namemax => 255,
		546	frsize => 1024,
		547	fsid => 1810
		548	}
		549
		550
		551	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		552
		553	Works like perl's C<utime> function (including the special case of $atime
		554	and $mtime being undef). Fractional times are supported if the underlying
		555	syscalls support them.
		556
		557	When called with a pathname, uses utimes(2) if available, otherwise
		558	utime(2). If called on a file descriptor, uses futimes(2) if available,
		559	otherwise returns ENOSYS, so this is not portable.
		560
		561	Examples:
		562
		563	# set atime and mtime to current time (basically touch(1)):
		564	aio_utime "path", undef, undef;
		565	# set atime to current time and mtime to beginning of the epoch:
		566	aio_utime "path", time, undef; # undef==0
		567
		568
		569	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		570
		571	Works like perl's C<chown> function, except that C<undef> for either $uid
		572	or $gid is being interpreted as "do not change" (but -1 can also be used).
		573
		574	Examples:
		575
		576	# same as "chown root path" in the shell:
		577	aio_chown "path", 0, -1;
		578	# same as above:
		579	aio_chown "path", 0, undef;
		580
		581
		582	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		583
		584	Works like truncate(2) or ftruncate(2).
		585
		586
		587	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		588
		589	Works like perl's C<chmod> function.
		590
		591
179	=item aio_unlink $pathname, $callback	592	=item aio_unlink $pathname, $callback->($status)
180		593
181	Asynchronously unlink (delete) a file and call the callback with the	594	Asynchronously unlink (delete) a file and call the callback with the
182	result code.	595	result code.
183		596
		597
		598	=item aio_mknod $path, $mode, $dev, $callback->($status)
		599
		600	[EXPERIMENTAL]
		601
		602	Asynchronously create a device node (or fifo). See mknod(2).
		603
		604	The only (POSIX-) portable way of calling this function is:
		605
		606	aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		607
		608	See C<aio_stat> for info about some potentially helpful extra constants
		609	and functions.
		610
		611	=item aio_link $srcpath, $dstpath, $callback->($status)
		612
		613	Asynchronously create a new link to the existing object at C<$srcpath> at
		614	the path C<$dstpath> and call the callback with the result code.
		615
		616
		617	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		618
		619	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		620	the path C<$dstpath> and call the callback with the result code.
		621
		622
		623	=item aio_readlink $path, $callback->($link)
		624
		625	Asynchronously read the symlink specified by C<$path> and pass it to
		626	the callback. If an error occurs, nothing or undef gets passed to the
		627	callback.
		628
		629
		630	=item aio_rename $srcpath, $dstpath, $callback->($status)
		631
		632	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		633	rename(2) and call the callback with the result code.
		634
		635
		636	=item aio_mkdir $pathname, $mode, $callback->($status)
		637
		638	Asynchronously mkdir (create) a directory and call the callback with
		639	the result code. C<$mode> will be modified by the umask at the time the
		640	request is executed, so do not change your umask.
		641
		642
		643	=item aio_rmdir $pathname, $callback->($status)
		644
		645	Asynchronously rmdir (delete) a directory and call the callback with the
		646	result code.
		647
		648
		649	=item aio_readdir $pathname, $callback->($entries)
		650
		651	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		652	directory (i.e. opendir + readdir + closedir). The entries will not be
		653	sorted, and will B<NOT> include the C<.> and C<..> entries.
		654
		655	The callback is passed a single argument which is either C<undef> or an
		656	array-ref with the filenames.
		657
		658
		659	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		660
		661	Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune
		662	behaviour and output format. In case of an error, C<$entries> will be
		663	C<undef>.
		664
		665	The flags are a combination of the following constants, ORed together (the
		666	flags will also be passed to the callback, possibly modified):
		667
		668	=over 4
		669
		670	=item IO::AIO::READDIR_DENTS
		671
		672	When this flag is off, then the callback gets an arrayref with of names
		673	only (as with C<aio_readdir>), otherwise it gets an arrayref with
		674	C<[$name, $type, $inode]> arrayrefs, each describing a single directory
		675	entry in more detail.
		676
		677	C<$name> is the name of the entry.
		678
		679	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		680
		681	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		682	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		683	C<IO::AIO::DT_WHT>.
		684
		685	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
		686	know, you have to run stat yourself. Also, for speed reasons, the C<$type>
		687	scalars are read-only: you can not modify them.
		688
		689	C<$inode> is the inode number (which might not be exact on systems with 64
		690	bit inode numbers and 32 bit perls). This field has unspecified content on
		691	systems that do not deliver the inode information.
		692
		693	=item IO::AIO::READDIR_DIRS_FIRST
		694
		695	When this flag is set, then the names will be returned in an order where
		696	likely directories come first. This is useful when you need to quickly
		697	find directories, or you want to find all directories while avoiding to
		698	stat() each entry.
		699
		700	If the system returns type information in readdir, then this is used
		701	to find directories directly. Otherwise, likely directories are files
		702	beginning with ".", or otherwise files with no dots, of which files with
		703	short names are tried first.
		704
		705	=item IO::AIO::READDIR_STAT_ORDER
		706
		707	When this flag is set, then the names will be returned in an order
		708	suitable for stat()'ing each one. That is, when you plan to stat()
		709	all files in the given directory, then the returned order will likely
		710	be fastest.
		711
		712	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
		713	the likely dirs come first, resulting in a less optimal stat order.
		714
		715	=item IO::AIO::READDIR_FOUND_UNKNOWN
		716
		717	This flag should not be set when calling C<aio_readdirx>. Instead, it
		718	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		719	C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all
		720	C<$type>'s are known, which can be used to speed up some algorithms.
		721
		722	=back
		723
		724
		725	=item aio_load $path, $data, $callback->($status)
		726
		727	This is a composite request that tries to fully load the given file into
		728	memory. Status is the same as with aio_read.
		729
		730	=cut
		731
		732	sub aio_load($$;$) {
		733	my ($path, undef, $cb) = @_;
		734	my $data = \$_[1];
		735
		736	my $pri = aioreq_pri;
		737	my $grp = aio_group $cb;
		738
		739	aioreq_pri $pri;
		740	add $grp aio_open $path, O_RDONLY, 0, sub {
		741	my $fh = shift
		742	or return $grp->result (-1);
		743
		744	aioreq_pri $pri;
		745	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		746	$grp->result ($_[0]);
		747	};
		748	};
		749
		750	$grp
		751	}
		752
		753	=item aio_copy $srcpath, $dstpath, $callback->($status)
		754
		755	Try to copy the I<file> (directories not supported as either source or
		756	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		757	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		758
		759	This is a composite request that creates the destination file with
		760	mode 0200 and copies the contents of the source file into it using
		761	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		762	uid/gid, in that order.
		763
		764	If an error occurs, the partial destination file will be unlinked, if
		765	possible, except when setting atime, mtime, access mode and uid/gid, where
		766	errors are being ignored.
		767
		768	=cut
		769
		770	sub aio_copy($$;$) {
		771	my ($src, $dst, $cb) = @_;
		772
		773	my $pri = aioreq_pri;
		774	my $grp = aio_group $cb;
		775
		776	aioreq_pri $pri;
		777	add $grp aio_open $src, O_RDONLY, 0, sub {
		778	if (my $src_fh = $_[0]) {
		779	my @stat = stat $src_fh; # hmm, might block over nfs?
		780
		781	aioreq_pri $pri;
		782	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		783	if (my $dst_fh = $_[0]) {
		784	aioreq_pri $pri;
		785	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		786	if ($_[0] == $stat[7]) {
		787	$grp->result (0);
		788	close $src_fh;
		789
		790	my $ch = sub {
		791	aioreq_pri $pri;
		792	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		793	aioreq_pri $pri;
		794	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		795	aioreq_pri $pri;
		796	add $grp aio_close $dst_fh;
		797	}
		798	};
		799	};
		800
		801	aioreq_pri $pri;
		802	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		803	if ($_[0] < 0 && $! == ENOSYS) {
		804	aioreq_pri $pri;
		805	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		806	} else {
		807	$ch->();
		808	}
		809	};
		810	} else {
		811	$grp->result (-1);
		812	close $src_fh;
		813	close $dst_fh;
		814
		815	aioreq $pri;
		816	add $grp aio_unlink $dst;
		817	}
		818	};
		819	} else {
		820	$grp->result (-1);
		821	}
		822	},
		823
		824	} else {
		825	$grp->result (-1);
		826	}
		827	};
		828
		829	$grp
		830	}
		831
		832	=item aio_move $srcpath, $dstpath, $callback->($status)
		833
		834	Try to move the I<file> (directories not supported as either source or
		835	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		836	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		837
		838	This is a composite request that tries to rename(2) the file first; if
		839	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		840	that is successful, unlinks the C<$srcpath>.
		841
		842	=cut
		843
		844	sub aio_move($$;$) {
		845	my ($src, $dst, $cb) = @_;
		846
		847	my $pri = aioreq_pri;
		848	my $grp = aio_group $cb;
		849
		850	aioreq_pri $pri;
		851	add $grp aio_rename $src, $dst, sub {
		852	if ($_[0] && $! == EXDEV) {
		853	aioreq_pri $pri;
		854	add $grp aio_copy $src, $dst, sub {
		855	$grp->result ($_[0]);
		856
		857	if (!$_[0]) {
		858	aioreq_pri $pri;
		859	add $grp aio_unlink $src;
		860	}
		861	};
		862	} else {
		863	$grp->result ($_[0]);
		864	}
		865	};
		866
		867	$grp
		868	}
		869
		870	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		871
		872	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		873	efficiently separate the entries of directory C<$path> into two sets of
		874	names, directories you can recurse into (directories), and ones you cannot
		875	recurse into (everything else, including symlinks to directories).
		876
		877	C<aio_scandir> is a composite request that creates of many sub requests_
		878	C<$maxreq> specifies the maximum number of outstanding aio requests that
		879	this function generates. If it is C<< <= 0 >>, then a suitable default
		880	will be chosen (currently 4).
		881
		882	On error, the callback is called without arguments, otherwise it receives
		883	two array-refs with path-relative entry names.
		884
		885	Example:
		886
		887	aio_scandir $dir, 0, sub {
		888	my ($dirs, $nondirs) = @_;
		889	print "real directories: @$dirs\n";
		890	print "everything else: @$nondirs\n";
		891	};
		892
		893	Implementation notes.
		894
		895	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		896
		897	If readdir returns file type information, then this is used directly to
		898	find directories.
		899
		900	Otherwise, after reading the directory, the modification time, size etc.
		901	of the directory before and after the readdir is checked, and if they
		902	match (and isn't the current time), the link count will be used to decide
		903	how many entries are directories (if >= 2). Otherwise, no knowledge of the
		904	number of subdirectories will be assumed.
		905
		906	Then entries will be sorted into likely directories a non-initial dot
		907	currently) and likely non-directories (see C<aio_readdirx>). Then every
		908	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
		909	in order of their inode numbers. If that succeeds, it assumes that the
		910	entry is a directory or a symlink to directory (which will be checked
		911	seperately). This is often faster than stat'ing the entry itself because
		912	filesystems might detect the type of the entry without reading the inode
		913	data (e.g. ext2fs filetype feature), even on systems that cannot return
		914	the filetype information on readdir.
		915
		916	If the known number of directories (link count - 2) has been reached, the
		917	rest of the entries is assumed to be non-directories.
		918
		919	This only works with certainty on POSIX (= UNIX) filesystems, which
		920	fortunately are the vast majority of filesystems around.
		921
		922	It will also likely work on non-POSIX filesystems with reduced efficiency
		923	as those tend to return 0 or 1 as link counts, which disables the
		924	directory counting heuristic.
		925
		926	=cut
		927
		928	sub aio_scandir($$;$) {
		929	my ($path, $maxreq, $cb) = @_;
		930
		931	my $pri = aioreq_pri;
		932
		933	my $grp = aio_group $cb;
		934
		935	$maxreq = 4 if $maxreq <= 0;
		936
		937	# stat once
		938	aioreq_pri $pri;
		939	add $grp aio_stat $path, sub {
		940	return $grp->result () if $_[0];
		941	my $now = time;
		942	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		943
		944	# read the directory entries
		945	aioreq_pri $pri;
		946	add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub {
		947	my $entries = shift
		948	or return $grp->result ();
		949
		950	# stat the dir another time
		951	aioreq_pri $pri;
		952	add $grp aio_stat $path, sub {
		953	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		954
		955	my $ndirs;
		956
		957	# take the slow route if anything looks fishy
		958	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		959	$ndirs = -1;
		960	} else {
		961	# if nlink == 2, we are finished
		962	# for non-posix-fs's, we rely on nlink < 2
		963	$ndirs = (stat _)[3] - 2
		964	or return $grp->result ([], $entries);
		965	}
		966
		967	my (@dirs, @nondirs);
		968
		969	my $statgrp = add $grp aio_group sub {
		970	$grp->result (\@dirs, \@nondirs);
		971	};
		972
		973	limit $statgrp $maxreq;
		974	feed $statgrp sub {
		975	return unless @$entries;
		976	my $entry = shift @$entries;
		977
		978	aioreq_pri $pri;
		979	add $statgrp aio_stat "$path/$entry/.", sub {
		980	if ($_[0] < 0) {
		981	push @nondirs, $entry;
		982	} else {
		983	# need to check for real directory
		984	aioreq_pri $pri;
		985	add $statgrp aio_lstat "$path/$entry", sub {
		986	if (-d _) {
		987	push @dirs, $entry;
		988
		989	unless (--$ndirs) {
		990	push @nondirs, @$entries;
		991	feed $statgrp;
		992	}
		993	} else {
		994	push @nondirs, $entry;
		995	}
		996	}
		997	}
		998	};
		999	};
		1000	};
		1001	};
		1002	};
		1003
		1004	$grp
		1005	}
		1006
		1007	=item aio_rmtree $path, $callback->($status)
		1008
		1009	Delete a directory tree starting (and including) C<$path>, return the
		1010	status of the final C<rmdir> only. This is a composite request that
		1011	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1012	everything else.
		1013
		1014	=cut
		1015
		1016	sub aio_rmtree;
		1017	sub aio_rmtree($;$) {
		1018	my ($path, $cb) = @_;
		1019
		1020	my $pri = aioreq_pri;
		1021	my $grp = aio_group $cb;
		1022
		1023	aioreq_pri $pri;
		1024	add $grp aio_scandir $path, 0, sub {
		1025	my ($dirs, $nondirs) = @_;
		1026
		1027	my $dirgrp = aio_group sub {
		1028	add $grp aio_rmdir $path, sub {
		1029	$grp->result ($_[0]);
		1030	};
		1031	};
		1032
		1033	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1034	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1035
		1036	add $grp $dirgrp;
		1037	};
		1038
		1039	$grp
		1040	}
		1041
		1042	=item aio_sync $callback->($status)
		1043
		1044	Asynchronously call sync and call the callback when finished.
		1045
184	=item aio_fsync $fh, $callback	1046	=item aio_fsync $fh, $callback->($status)
185		1047
186	Asynchronously call fsync on the given filehandle and call the callback	1048	Asynchronously call fsync on the given filehandle and call the callback
187	with the fsync result code.	1049	with the fsync result code.
188		1050
189	=item aio_fdatasync $fh, $callback	1051	=item aio_fdatasync $fh, $callback->($status)
190		1052
191	Asynchronously call fdatasync on the given filehandle and call the	1053	Asynchronously call fdatasync on the given filehandle and call the
192	callback with the fdatasync result code.	1054	callback with the fdatasync result code.
193		1055
		1056	If this call isn't available because your OS lacks it or it couldn't be
		1057	detected, it will be emulated by calling C<fsync> instead.
		1058
		1059	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1060
		1061	Sync the data portion of the file specified by C<$offset> and C<$length>
		1062	to disk (but NOT the metadata), by calling the Linux-specific
		1063	sync_file_range call. If sync_file_range is not available or it returns
		1064	ENOSYS, then fdatasync or fsync is being substituted.
		1065
		1066	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1067	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1068	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1069	manpage for details.
		1070
		1071	=item aio_pathsync $path, $callback->($status)
		1072
		1073	This request tries to open, fsync and close the given path. This is a
		1074	composite request intended to sync directories after directory operations
		1075	(E.g. rename). This might not work on all operating systems or have any
		1076	specific effect, but usually it makes sure that directory changes get
		1077	written to disc. It works for anything that can be opened for read-only,
		1078	not just directories.
		1079
		1080	Future versions of this function might fall back to other methods when
		1081	C<fsync> on the directory fails (such as calling C<sync>).
		1082
		1083	Passes C<0> when everything went ok, and C<-1> on error.
		1084
		1085	=cut
		1086
		1087	sub aio_pathsync($;$) {
		1088	my ($path, $cb) = @_;
		1089
		1090	my $pri = aioreq_pri;
		1091	my $grp = aio_group $cb;
		1092
		1093	aioreq_pri $pri;
		1094	add $grp aio_open $path, O_RDONLY, 0, sub {
		1095	my ($fh) = @_;
		1096	if ($fh) {
		1097	aioreq_pri $pri;
		1098	add $grp aio_fsync $fh, sub {
		1099	$grp->result ($_[0]);
		1100
		1101	aioreq_pri $pri;
		1102	add $grp aio_close $fh;
		1103	};
		1104	} else {
		1105	$grp->result (-1);
		1106	}
		1107	};
		1108
		1109	$grp
		1110	}
		1111
		1112	=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1113
		1114	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1115	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1116	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1117	scalar must only be modified in-place while an aio operation is pending on
		1118	it).
		1119
		1120	It calls the C<msync> function of your OS, if available, with the memory
		1121	area starting at C<$offset> in the string and ending C<$length> bytes
		1122	later. If C<$length> is negative, counts from the end, and if C<$length>
		1123	is C<undef>, then it goes till the end of the string. The flags can be
		1124	a combination of C<IO::AIO::MS_ASYNC>, C<IO::AIO::MS_INVALIDATE> and
		1125	C<IO::AIO::MS_SYNC>.
		1126
		1127	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1128
		1129	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1130	scalars.
		1131
		1132	It touches (reads or writes) all memory pages in the specified
		1133	range inside the scalar. All caveats and parameters are the same
		1134	as for C<aio_msync>, above, except for flags, which must be either
		1135	C<0> (which reads all pages and ensures they are instantiated) or
		1136	C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
		1137	writing an octet from it, which dirties the page).
		1138
		1139	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1140
		1141	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1142	scalars.
		1143
		1144	It reads in all the pages of the underlying storage into memory (if any)
		1145	and locks them, so they are not getting swapped/paged out or removed.
		1146
		1147	If C<$length> is undefined, then the scalar will be locked till the end.
		1148
		1149	On systems that do not implement C<mlock>, this function returns C<-1>
		1150	and sets errno to C<ENOSYS>.
		1151
		1152	Note that the corresponding C<munlock> is synchronous and is
		1153	documented under L<MISCELLANEOUS FUNCTIONS>.
		1154
		1155	Example: open a file, mmap and mlock it - both will be undone when
		1156	C<$data> gets destroyed.
		1157
		1158	open my $fh, "<", $path or die "$path: $!";
		1159	my $data;
		1160	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1161	aio_mlock $data; # mlock in background
		1162
		1163	=item aio_mlockall $flags, $callback->($status)
		1164
		1165	Calls the C<mlockall> function with the given C<$flags> (a combination of
		1166	C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
		1167
		1168	On systems that do not implement C<mlockall>, this function returns C<-1>
		1169	and sets errno to C<ENOSYS>.
		1170
		1171	Note that the corresponding C<munlockall> is synchronous and is
		1172	documented under L<MISCELLANEOUS FUNCTIONS>.
		1173
		1174	Example: asynchronously lock all current and future pages into memory.
		1175
		1176	aio_mlockall IO::AIO::MCL_FUTURE;
		1177
		1178	=item aio_group $callback->(...)
		1179
		1180	This is a very special aio request: Instead of doing something, it is a
		1181	container for other aio requests, which is useful if you want to bundle
		1182	many requests into a single, composite, request with a definite callback
		1183	and the ability to cancel the whole request with its subrequests.
		1184
		1185	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		1186	for more info.
		1187
		1188	Example:
		1189
		1190	my $grp = aio_group sub {
		1191	print "all stats done\n";
		1192	};
		1193
		1194	add $grp
		1195	(aio_stat ...),
		1196	(aio_stat ...),
		1197	...;
		1198
		1199	=item aio_nop $callback->()
		1200
		1201	This is a special request - it does nothing in itself and is only used for
		1202	side effects, such as when you want to add a dummy request to a group so
		1203	that finishing the requests in the group depends on executing the given
		1204	code.
		1205
		1206	While this request does nothing, it still goes through the execution
		1207	phase and still requires a worker thread. Thus, the callback will not
		1208	be executed immediately but only after other requests in the queue have
		1209	entered their execution phase. This can be used to measure request
		1210	latency.
		1211
		1212	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
		1213
		1214	Mainly used for debugging and benchmarking, this aio request puts one of
		1215	the request workers to sleep for the given time.
		1216
		1217	While it is theoretically handy to have simple I/O scheduling requests
		1218	like sleep and file handle readable/writable, the overhead this creates is
		1219	immense (it blocks a thread for a long time) so do not use this function
		1220	except to put your application under artificial I/O pressure.
		1221
194	=back	1222	=back
195		1223
		1224	=head2 IO::AIO::REQ CLASS
		1225
		1226	All non-aggregate C<aio_*> functions return an object of this class when
		1227	called in non-void context.
		1228
		1229	=over 4
		1230
		1231	=item cancel $req
		1232
		1233	Cancels the request, if possible. Has the effect of skipping execution
		1234	when entering the B<execute> state and skipping calling the callback when
		1235	entering the the B<result> state, but will leave the request otherwise
		1236	untouched (with the exception of readdir). That means that requests that
		1237	currently execute will not be stopped and resources held by the request
		1238	will not be freed prematurely.
		1239
		1240	=item cb $req $callback->(...)
		1241
		1242	Replace (or simply set) the callback registered to the request.
		1243
		1244	=back
		1245
		1246	=head2 IO::AIO::GRP CLASS
		1247
		1248	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		1249	objects of this class, too.
		1250
		1251	A IO::AIO::GRP object is a special request that can contain multiple other
		1252	aio requests.
		1253
		1254	You create one by calling the C<aio_group> constructing function with a
		1255	callback that will be called when all contained requests have entered the
		1256	C<done> state:
		1257
		1258	my $grp = aio_group sub {
		1259	print "all requests are done\n";
		1260	};
		1261
		1262	You add requests by calling the C<add> method with one or more
		1263	C<IO::AIO::REQ> objects:
		1264
		1265	$grp->add (aio_unlink "...");
		1266
		1267	add $grp aio_stat "...", sub {
		1268	$_[0] or return $grp->result ("error");
		1269
		1270	# add another request dynamically, if first succeeded
		1271	add $grp aio_open "...", sub {
		1272	$grp->result ("ok");
		1273	};
		1274	};
		1275
		1276	This makes it very easy to create composite requests (see the source of
		1277	C<aio_move> for an application) that work and feel like simple requests.
		1278
		1279	=over 4
		1280
		1281	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		1282	C<IO::AIO::poll_cb>, just like any other request.
		1283
		1284	=item * They can be canceled like any other request. Canceling will cancel not
		1285	only the request itself, but also all requests it contains.
		1286
		1287	=item * They can also can also be added to other IO::AIO::GRP objects.
		1288
		1289	=item * You must not add requests to a group from within the group callback (or
		1290	any later time).
		1291
		1292	=back
		1293
		1294	Their lifetime, simplified, looks like this: when they are empty, they
		1295	will finish very quickly. If they contain only requests that are in the
		1296	C<done> state, they will also finish. Otherwise they will continue to
		1297	exist.
		1298
		1299	That means after creating a group you have some time to add requests
		1300	(precisely before the callback has been invoked, which is only done within
		1301	the C<poll_cb>). And in the callbacks of those requests, you can add
		1302	further requests to the group. And only when all those requests have
		1303	finished will the the group itself finish.
		1304
		1305	=over 4
		1306
		1307	=item add $grp ...
		1308
		1309	=item $grp->add (...)
		1310
		1311	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		1312	be added, including other groups, as long as you do not create circular
		1313	dependencies.
		1314
		1315	Returns all its arguments.
		1316
		1317	=item $grp->cancel_subs
		1318
		1319	Cancel all subrequests and clears any feeder, but not the group request
		1320	itself. Useful when you queued a lot of events but got a result early.
		1321
		1322	The group request will finish normally (you cannot add requests to the
		1323	group).
		1324
		1325	=item $grp->result (...)
		1326
		1327	Set the result value(s) that will be passed to the group callback when all
		1328	subrequests have finished and set the groups errno to the current value
		1329	of errno (just like calling C<errno> without an error number). By default,
		1330	no argument will be passed and errno is zero.
		1331
		1332	=item $grp->errno ([$errno])
		1333
		1334	Sets the group errno value to C<$errno>, or the current value of errno
		1335	when the argument is missing.
		1336
		1337	Every aio request has an associated errno value that is restored when
		1338	the callback is invoked. This method lets you change this value from its
		1339	default (0).
		1340
		1341	Calling C<result> will also set errno, so make sure you either set C<$!>
		1342	before the call to C<result>, or call c<errno> after it.
		1343
		1344	=item feed $grp $callback->($grp)
		1345
		1346	Sets a feeder/generator on this group: every group can have an attached
		1347	generator that generates requests if idle. The idea behind this is that,
		1348	although you could just queue as many requests as you want in a group,
		1349	this might starve other requests for a potentially long time. For example,
		1350	C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests,
		1351	delaying any later requests for a long time.
		1352
		1353	To avoid this, and allow incremental generation of requests, you can
		1354	instead a group and set a feeder on it that generates those requests. The
		1355	feed callback will be called whenever there are few enough (see C<limit>,
		1356	below) requests active in the group itself and is expected to queue more
		1357	requests.
		1358
		1359	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1360	not impose any limits).
		1361
		1362	If the feed does not queue more requests when called, it will be
		1363	automatically removed from the group.
		1364
		1365	If the feed limit is C<0> when this method is called, it will be set to
		1366	C<2> automatically.
		1367
		1368	Example:
		1369
		1370	# stat all files in @files, but only ever use four aio requests concurrently:
		1371
		1372	my $grp = aio_group sub { print "finished\n" };
		1373	limit $grp 4;
		1374	feed $grp sub {
		1375	my $file = pop @files
		1376	or return;
		1377
		1378	add $grp aio_stat $file, sub { ... };
		1379	};
		1380
		1381	=item limit $grp $num
		1382
		1383	Sets the feeder limit for the group: The feeder will be called whenever
		1384	the group contains less than this many requests.
		1385
		1386	Setting the limit to C<0> will pause the feeding process.
		1387
		1388	The default value for the limit is C<0>, but note that setting a feeder
		1389	automatically bumps it up to C<2>.
		1390
		1391	=back
		1392
196	=head2 SUPPORT FUNCTIONS	1393	=head2 SUPPORT FUNCTIONS
197		1394
		1395	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1396
198	=over 4	1397	=over 4
199		1398
200	=item $fileno = IO::AIO::poll_fileno	1399	=item $fileno = IO::AIO::poll_fileno
201		1400
202	Return the I<request result pipe filehandle>. This filehandle must be	1401	Return the I<request result pipe file descriptor>. This filehandle must be
203	polled for reading by some mechanism outside this module (e.g. Event	1402	polled for reading by some mechanism outside this module (e.g. EV, Glib,
204	or select, see below). If the pipe becomes readable you have to call	1403	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
205	C<poll_cb> to check the results.	1404	you have to call C<poll_cb> to check the results.
206		1405
207	See C<poll_cb> for an example.	1406	See C<poll_cb> for an example.
208		1407
209	=item IO::AIO::poll_cb	1408	=item IO::AIO::poll_cb
210		1409
211	Process all outstanding events on the result pipe. You have to call this	1410	Process some outstanding events on the result pipe. You have to call this
212	regularly. Returns the number of events processed. Returns immediately	1411	regularly. Returns C<0> if all events could be processed, or C<-1> if it
213	when no events are outstanding.	1412	returned earlier for whatever reason. Returns immediately when no events
		1413	are outstanding. The amount of events processed depends on the settings of
		1414	C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
214		1415
215	You can use Event to multiplex, e.g.:	1416	If not all requests were processed for whatever reason, the filehandle
		1417	will still be ready when C<poll_cb> returns, so normally you don't have to
		1418	do anything special to have it called later.
		1419
		1420	Example: Install an Event watcher that automatically calls
		1421	IO::AIO::poll_cb with high priority (more examples can be found in the
		1422	SYNOPSIS section, at the top of this document):
216		1423
217	Event->io (fd => IO::AIO::poll_fileno,	1424	Event->io (fd => IO::AIO::poll_fileno,
218	poll => 'r', async => 1,	1425	poll => 'r', async => 1,
219	cb => \&IO::AIO::poll_cb);	1426	cb => \&IO::AIO::poll_cb);
220		1427
221	=item IO::AIO::poll_wait	1428	=item IO::AIO::poll_wait
222		1429
		1430	If there are any outstanding requests and none of them in the result
223	Wait till the result filehandle becomes ready for reading (simply does a	1431	phase, wait till the result filehandle becomes ready for reading (simply
224	select on the filehandle. This is useful if you want to synchronously wait	1432	does a C<select> on the filehandle. This is useful if you want to
225	for some requests to finish).	1433	synchronously wait for some requests to finish).
226		1434
227	See C<nreqs> for an example.	1435	See C<nreqs> for an example.
228		1436
		1437	=item IO::AIO::poll
		1438
		1439	Waits until some requests have been handled.
		1440
		1441	Returns the number of requests processed, but is otherwise strictly
		1442	equivalent to:
		1443
		1444	IO::AIO::poll_wait, IO::AIO::poll_cb
		1445
229	=item IO::AIO::nreqs	1446	=item IO::AIO::flush
230		1447
231	Returns the number of requests currently outstanding.	1448	Wait till all outstanding AIO requests have been handled.
232		1449
233	Example: wait till there are no outstanding requests anymore:	1450	Strictly equivalent to:
234		1451
235	IO::AIO::poll_wait, IO::AIO::poll_cb	1452	IO::AIO::poll_wait, IO::AIO::poll_cb
236	while IO::AIO::nreqs;	1453	while IO::AIO::nreqs;
237		1454
238	=item IO::AIO::flush	1455	=item IO::AIO::max_poll_reqs $nreqs
239		1456
240	Wait till all outstanding AIO requests have been handled.	1457	=item IO::AIO::max_poll_time $seconds
		1458
		1459	These set the maximum number of requests (default C<0>, meaning infinity)
		1460	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1461	the maximum amount of time (default C<0>, meaning infinity) spent in
		1462	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1463	of time C<poll_cb> is allowed to use).
		1464
		1465	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1466	syscall per request processed, which is not normally a problem unless your
		1467	callbacks are really really fast or your OS is really really slow (I am
		1468	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1469
		1470	Setting these is useful if you want to ensure some level of
		1471	interactiveness when perl is not fast enough to process all requests in
		1472	time.
		1473
		1474	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		1475
		1476	Example: Install an Event watcher that automatically calls
		1477	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1478	program get the CPU sometimes even under high AIO load.
		1479
		1480	# try not to spend much more than 0.1s in poll_cb
		1481	IO::AIO::max_poll_time 0.1;
		1482
		1483	# use a low priority so other tasks have priority
		1484	Event->io (fd => IO::AIO::poll_fileno,
		1485	poll => 'r', nice => 1,
		1486	cb => &IO::AIO::poll_cb);
		1487
		1488	=back
		1489
		1490	=head3 CONTROLLING THE NUMBER OF THREADS
		1491
		1492	=over
241		1493
242	=item IO::AIO::min_parallel $nthreads	1494	=item IO::AIO::min_parallel $nthreads
243		1495
244	Set the minimum number of AIO threads to C<$nthreads>. The default is	1496	Set the minimum number of AIO threads to C<$nthreads>. The current
245	C<1>, which means a single asynchronous operation can be done at one time	1497	default is C<8>, which means eight asynchronous operations can execute
246	(the number of outstanding operations, however, is unlimited).	1498	concurrently at any one time (the number of outstanding requests,
		1499	however, is unlimited).
247		1500
		1501	IO::AIO starts threads only on demand, when an AIO request is queued and
		1502	no free thread exists. Please note that queueing up a hundred requests can
		1503	create demand for a hundred threads, even if it turns out that everything
		1504	is in the cache and could have been processed faster by a single thread.
		1505
248	It is recommended to keep the number of threads low, as some Linux	1506	It is recommended to keep the number of threads relatively low, as some
249	kernel versions will scale negatively with the number of threads (higher	1507	Linux kernel versions will scale negatively with the number of threads
250	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	1508	(higher parallelity => MUCH higher latency). With current Linux 2.6
251	threads should be fine.	1509	versions, 4-32 threads should be fine.
252		1510
253	Under normal circumstances you don't need to call this function, as this	1511	Under most circumstances you don't need to call this function, as the
254	module automatically starts some threads (the exact number might change,	1512	module selects a default that is suitable for low to moderate load.
255	and is currently 4).
256		1513
257	=item IO::AIO::max_parallel $nthreads	1514	=item IO::AIO::max_parallel $nthreads
258		1515
259	Sets the maximum number of AIO threads to C<$nthreads>. If more than	1516	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
260	the specified number of threads are currently running, kill them. This	1517	specified number of threads are currently running, this function kills
261	function blocks until the limit is reached.	1518	them. This function blocks until the limit is reached.
		1519
		1520	While C<$nthreads> are zero, aio requests get queued but not executed
		1521	until the number of threads has been increased again.
262		1522
263	This module automatically runs C<max_parallel 0> at program end, to ensure	1523	This module automatically runs C<max_parallel 0> at program end, to ensure
264	that all threads are killed and that there are no outstanding requests.	1524	that all threads are killed and that there are no outstanding requests.
265		1525
266	Under normal circumstances you don't need to call this function.	1526	Under normal circumstances you don't need to call this function.
267		1527
		1528	=item IO::AIO::max_idle $nthreads
		1529
		1530	Limit the number of threads (default: 4) that are allowed to idle (i.e.,
		1531	threads that did not get a request to process within 10 seconds). That
		1532	means if a thread becomes idle while C<$nthreads> other threads are also
		1533	idle, it will free its resources and exit.
		1534
		1535	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1536	to allow for extremely high load situations, but want to free resources
		1537	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1538
		1539	The default is probably ok in most situations, especially if thread
		1540	creation is fast. If thread creation is very slow on your system you might
		1541	want to use larger values.
		1542
268	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	1543	=item IO::AIO::max_outstanding $maxreqs
		1544
		1545	This is a very bad function to use in interactive programs because it
		1546	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1547	use an C<aio_group> together with a feed callback.
269		1548
270	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1549	Sets the maximum number of outstanding requests to C<$nreqs>. If you
271	try to queue up more than this number of requests, the caller will block until	1550	do queue up more than this number of requests, the next call to the
272	some requests have been handled.	1551	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
		1552	function will block until the limit is no longer exceeded.
273		1553
274	The default is very large, so normally there is no practical limit. If you	1554	The default value is very large, so there is no practical limit on the
275	queue up many requests in a loop it it often improves speed if you set	1555	number of outstanding requests.
276	this to a relatively low number, such as C<100>.
277		1556
278	Under normal circumstances you don't need to call this function.	1557	You can still queue as many requests as you want. Therefore,
		1558	C<max_outstanding> is mainly useful in simple scripts (with low values) or
		1559	as a stop gap to shield against fatal memory overflow (with large values).
279		1560
280	=back	1561	=back
281		1562
		1563	=head3 STATISTICAL INFORMATION
		1564
		1565	=over
		1566
		1567	=item IO::AIO::nreqs
		1568
		1569	Returns the number of requests currently in the ready, execute or pending
		1570	states (i.e. for which their callback has not been invoked yet).
		1571
		1572	Example: wait till there are no outstanding requests anymore:
		1573
		1574	IO::AIO::poll_wait, IO::AIO::poll_cb
		1575	while IO::AIO::nreqs;
		1576
		1577	=item IO::AIO::nready
		1578
		1579	Returns the number of requests currently in the ready state (not yet
		1580	executed).
		1581
		1582	=item IO::AIO::npending
		1583
		1584	Returns the number of requests currently in the pending state (executed,
		1585	but not yet processed by poll_cb).
		1586
		1587	=back
		1588
		1589	=head3 MISCELLANEOUS FUNCTIONS
		1590
		1591	IO::AIO implements some functions that might be useful, but are not
		1592	asynchronous.
		1593
		1594	=over 4
		1595
		1596	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		1597
		1598	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		1599	but is blocking (this makes most sense if you know the input data is
		1600	likely cached already and the output filehandle is set to non-blocking
		1601	operations).
		1602
		1603	Returns the number of bytes copied, or C<-1> on error.
		1604
		1605	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		1606
		1607	Simply calls the C<posix_fadvise> function (see its
		1608	manpage for details). The following advice constants are
		1609	avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		1610	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		1611	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		1612
		1613	On systems that do not implement C<posix_fadvise>, this function returns
		1614	ENOSYS, otherwise the return value of C<posix_fadvise>.
		1615
		1616	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		1617
		1618	Simply calls the C<posix_madvise> function (see its
		1619	manpage for details). The following advice constants are
		1620	avaiable: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		1621	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
		1622
		1623	On systems that do not implement C<posix_madvise>, this function returns
		1624	ENOSYS, otherwise the return value of C<posix_madvise>.
		1625
		1626	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		1627
		1628	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		1629	$scalar (see its manpage for details). The following protect
		1630	constants are avaiable: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		1631	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		1632
		1633	On systems that do not implement C<mprotect>, this function returns
		1634	ENOSYS, otherwise the return value of C<mprotect>.
		1635
		1636	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		1637
		1638	Memory-maps a file (or anonymous memory range) and attaches it to the
		1639	given C<$scalar>, which will act like a string scalar.
		1640
		1641	The only operations allowed on the scalar are C<substr>/C<vec> that don't
		1642	change the string length, and most read-only operations such as copying it
		1643	or searching it with regexes and so on.
		1644
		1645	Anything else is unsafe and will, at best, result in memory leaks.
		1646
		1647	The memory map associated with the C<$scalar> is automatically removed
		1648	when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
		1649	C<IO::AIO::munmap> functions are called.
		1650
		1651	This calls the C<mmap>(2) function internally. See your system's manual
		1652	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		1653
		1654	The C<$length> must be larger than zero and smaller than the actual
		1655	filesize.
		1656
		1657	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		1658	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		1659
		1660	C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
		1661	C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
		1662	not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
		1663	(which is set to C<MAP_ANON> if your system only provides this
		1664	constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
		1665	C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
		1666	C<IO::AIO::MAP_NONBLOCK>
		1667
		1668	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		1669
		1670	C<$offset> is the offset from the start of the file - it generally must be
		1671	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		1672
		1673	Example:
		1674
		1675	use Digest::MD5;
		1676	use IO::AIO;
		1677
		1678	open my $fh, "<verybigfile"
		1679	or die "$!";
		1680
		1681	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		1682	or die "verybigfile: $!";
		1683
		1684	my $fast_md5 = md5 $data;
		1685
		1686	=item IO::AIO::munmap $scalar
		1687
		1688	Removes a previous mmap and undefines the C<$scalar>.
		1689
		1690	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		1691
		1692	Calls the C<munlock> function, undoing the effects of a previous
		1693	C<aio_mlock> call (see its description for details).
		1694
		1695	=item IO::AIO::munlockall
		1696
		1697	Calls the C<munlockall> function.
		1698
		1699	On systems that do not implement C<munlockall>, this function returns
		1700	ENOSYS, otherwise the return value of C<munlockall>.
		1701
		1702	=back
		1703
282	=cut	1704	=cut
283		1705
284	# support function to convert a fd into a perl filehandle
285	sub _fd2fh {
286	return undef if $_[0] < 0;
287
288	# try to be perl5.6-compatible
289	local *AIO_FH;
290	open AIO_FH, "+<&=$_[0]"
291	or return undef;
292
293	*AIO_FH
294	}
295
296	min_parallel 4;	1706	min_parallel 8;
297		1707
298	END {	1708	END { flush }
299	max_parallel 0;
300	}
301		1709
302	1;	1710	1;
303		1711
		1712	=head1 EVENT LOOP INTEGRATION
		1713
		1714	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		1715	automatically into many event loops:
		1716
		1717	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		1718	use AnyEvent::AIO;
		1719
		1720	You can also integrate IO::AIO manually into many event loops, here are
		1721	some examples of how to do this:
		1722
		1723	# EV integration
		1724	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		1725
		1726	# Event integration
		1727	Event->io (fd => IO::AIO::poll_fileno,
		1728	poll => 'r',
		1729	cb => \&IO::AIO::poll_cb);
		1730
		1731	# Glib/Gtk2 integration
		1732	add_watch Glib::IO IO::AIO::poll_fileno,
		1733	in => sub { IO::AIO::poll_cb; 1 };
		1734
		1735	# Tk integration
		1736	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		1737	readable => \&IO::AIO::poll_cb);
		1738
		1739	# Danga::Socket integration
		1740	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		1741	\&IO::AIO::poll_cb);
		1742
		1743	=head2 FORK BEHAVIOUR
		1744
		1745	This module should do "the right thing" when the process using it forks:
		1746
		1747	Before the fork, IO::AIO enters a quiescent state where no requests
		1748	can be added in other threads and no results will be processed. After
		1749	the fork the parent simply leaves the quiescent state and continues
		1750	request/result processing, while the child frees the request/result queue
		1751	(so that the requests started before the fork will only be handled in the
		1752	parent). Threads will be started on demand until the limit set in the
		1753	parent process has been reached again.
		1754
		1755	In short: the parent will, after a short pause, continue as if fork had
		1756	not been called, while the child will act as if IO::AIO has not been used
		1757	yet.
		1758
		1759	=head2 MEMORY USAGE
		1760
		1761	Per-request usage:
		1762
		1763	Each aio request uses - depending on your architecture - around 100-200
		1764	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1765	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1766	scalars and other data passed into aio requests will also be locked and
		1767	will consume memory till the request has entered the done state.
		1768
		1769	This is not awfully much, so queuing lots of requests is not usually a
		1770	problem.
		1771
		1772	Per-thread usage:
		1773
		1774	In the execution phase, some aio requests require more memory for
		1775	temporary buffers, and each thread requires a stack and other data
		1776	structures (usually around 16k-128k, depending on the OS).
		1777
		1778	=head1 KNOWN BUGS
		1779
		1780	Known bugs will be fixed in the next release.
		1781
304	=head1 SEE ALSO	1782	=head1 SEE ALSO
305		1783
306	L<Coro>, L<Linux::AIO>.	1784	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		1785	more natural syntax.
307		1786
308	=head1 AUTHOR	1787	=head1 AUTHOR
309		1788
310	Marc Lehmann <schmorp@schmorp.de>	1789	Marc Lehmann <schmorp@schmorp.de>
311	http://home.schmorp.de/	1790	http://home.schmorp.de/

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines