ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.9 by root, Sun Jul 10 22:20:55 2005 UTC vs.
Revision 1.279 by root, Sat Jan 6 01:04:42 2018 UTC

1	=head1 NAME	1	=head1 NAME
2		2
3	IO::AIO - Asynchronous Input/Output	3	IO::AIO - Asynchronous/Advanced Input/Output
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		30
33	=head1 DESCRIPTION	31	=head1 DESCRIPTION
34		32
35	This module implements asynchronous I/O using whatever means your	33	This module implements asynchronous I/O using whatever means your
36	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>).
37		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
38	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
39	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
40	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
41	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
42	aio functions available on many operating systems. However, they are often	57	functions available on many operating systems. However, they are often
43	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
44	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
45	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.
46		62
		63	In addition to asynchronous I/O, this module also exports some rather
		64	arcane interfaces, such as C<madvise> or linux's C<splice> system call,
		65	which is why the C<A> in C<AIO> can also mean I<advanced>.
		66
47	Although the module will work with in the presence of other threads, it is	67	Although the module will work in the presence of other (Perl-) threads,
48	currently not reentrant, so use appropriate locking yourself.	68	it is currently not reentrant in any way, so use appropriate locking
		69	yourself, always call C<poll_cb> from within the same thread, or never
		70	call C<poll_cb> (or other C<aio_> functions) recursively.
		71
		72	=head2 EXAMPLE
		73
		74	This is a simple example that uses the EV module and loads
		75	F</etc/passwd> asynchronously:
		76
		77	use EV;
		78	use IO::AIO;
		79
		80	# register the IO::AIO callback with EV
		81	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		82
		83	# queue the request to open /etc/passwd
		84	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
		85	my $fh = shift
		86	or die "error while opening: $!";
		87
		88	# stat'ing filehandles is generally non-blocking
		89	my $size = -s $fh;
		90
		91	# queue a request to read the file
		92	my $contents;
		93	aio_read $fh, 0, $size, $contents, 0, sub {
		94	$_[0] == $size
		95	or die "short read: $!";
		96
		97	close $fh;
		98
		99	# file contents now in $contents
		100	print $contents;
		101
		102	# exit event loop and program
		103	EV::break;
		104	};
		105	};
		106
		107	# possibly queue up other requests, or open GUI windows,
		108	# check for sockets etc. etc.
		109
		110	# process events as long as there are some:
		111	EV::run;
		112
		113	=head1 REQUEST ANATOMY AND LIFETIME
		114
		115	Every C<aio_*> function creates a request. which is a C data structure not
		116	directly visible to Perl.
		117
		118	If called in non-void context, every request function returns a Perl
		119	object representing the request. In void context, nothing is returned,
		120	which saves a bit of memory.
		121
		122	The perl object is a fairly standard ref-to-hash object. The hash contents
		123	are not used by IO::AIO so you are free to store anything you like in it.
		124
		125	During their existance, aio requests travel through the following states,
		126	in order:
		127
		128	=over 4
		129
		130	=item ready
		131
		132	Immediately after a request is created it is put into the ready state,
		133	waiting for a thread to execute it.
		134
		135	=item execute
		136
		137	A thread has accepted the request for processing and is currently
		138	executing it (e.g. blocking in read).
		139
		140	=item pending
		141
		142	The request has been executed and is waiting for result processing.
		143
		144	While request submission and execution is fully asynchronous, result
		145	processing is not and relies on the perl interpreter calling C<poll_cb>
		146	(or another function with the same effect).
		147
		148	=item result
		149
		150	The request results are processed synchronously by C<poll_cb>.
		151
		152	The C<poll_cb> function will process all outstanding aio requests by
		153	calling their callbacks, freeing memory associated with them and managing
		154	any groups they are contained in.
		155
		156	=item done
		157
		158	Request has reached the end of its lifetime and holds no resources anymore
		159	(except possibly for the Perl object, but its connection to the actual
		160	aio request is severed and calling its methods will either do nothing or
		161	result in a runtime error).
		162
		163	=back
49		164
50	=cut	165	=cut
51		166
52	package IO::AIO;	167	package IO::AIO;
53		168
		169	use Carp ();
		170
		171	use common::sense;
		172
54	use base 'Exporter';	173	use base 'Exporter';
55		174
56	use Fcntl ();
57
58	BEGIN {	175	BEGIN {
59	$VERSION = 0.2;	176	our $VERSION = 4.4;
60		177
61	@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink	178	our @AIO_REQ = qw(aio_sendfile aio_seek aio_read aio_write aio_open aio_close
62	aio_fsync aio_fdatasync aio_readahead);	179	aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
63	@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	180	aio_scandir aio_symlink aio_readlink aio_realpath aio_fcntl aio_ioctl
		181	aio_sync aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range
		182	aio_pathsync aio_readahead aio_fiemap aio_allocate
		183	aio_rename aio_rename2 aio_link aio_move aio_copy aio_group
		184	aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
		185	aio_chmod aio_utime aio_truncate
		186	aio_msync aio_mtouch aio_mlock aio_mlockall
		187	aio_statvfs
		188	aio_slurp
		189	aio_wd);
		190
		191	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		192	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		193	min_parallel max_parallel max_idle idle_timeout
		194	nreqs nready npending nthreads
		195	max_poll_time max_poll_reqs
		196	sendfile fadvise madvise
		197	mmap munmap munlock munlockall);
		198
		199	push @AIO_REQ, qw(aio_busy); # not exported
		200
		201	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
64		202
65	require XSLoader;	203	require XSLoader;
66	XSLoader::load IO::AIO, $VERSION;	204	XSLoader::load ("IO::AIO", $VERSION);
67	}	205	}
68		206
69	=head1 FUNCTIONS	207	=head1 FUNCTIONS
70		208
71	=head2 AIO FUNCTIONS	209	=head2 QUICK OVERVIEW
		210
		211	This section simply lists the prototypes most of the functions for
		212	quick reference. See the following sections for function-by-function
		213	documentation.
		214
		215	aio_wd $pathname, $callback->($wd)
		216	aio_open $pathname, $flags, $mode, $callback->($fh)
		217	aio_close $fh, $callback->($status)
		218	aio_seek $fh,$offset,$whence, $callback->($offs)
		219	aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		220	aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		221	aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		222	aio_readahead $fh,$offset,$length, $callback->($retval)
		223	aio_stat $fh_or_path, $callback->($status)
		224	aio_lstat $fh, $callback->($status)
		225	aio_statvfs $fh_or_path, $callback->($statvfs)
		226	aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		227	aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		228	aio_chmod $fh_or_path, $mode, $callback->($status)
		229	aio_truncate $fh_or_path, $offset, $callback->($status)
		230	aio_allocate $fh, $mode, $offset, $len, $callback->($status)
		231	aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
		232	aio_unlink $pathname, $callback->($status)
		233	aio_mknod $pathname, $mode, $dev, $callback->($status)
		234	aio_link $srcpath, $dstpath, $callback->($status)
		235	aio_symlink $srcpath, $dstpath, $callback->($status)
		236	aio_readlink $pathname, $callback->($link)
		237	aio_realpath $pathname, $callback->($path)
		238	aio_rename $srcpath, $dstpath, $callback->($status)
		239	aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
		240	aio_mkdir $pathname, $mode, $callback->($status)
		241	aio_rmdir $pathname, $callback->($status)
		242	aio_readdir $pathname, $callback->($entries)
		243	aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		244	IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
		245	IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
		246	aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		247	aio_load $pathname, $data, $callback->($status)
		248	aio_copy $srcpath, $dstpath, $callback->($status)
		249	aio_move $srcpath, $dstpath, $callback->($status)
		250	aio_rmtree $pathname, $callback->($status)
		251	aio_fcntl $fh, $cmd, $arg, $callback->($status)
		252	aio_ioctl $fh, $request, $buf, $callback->($status)
		253	aio_sync $callback->($status)
		254	aio_syncfs $fh, $callback->($status)
		255	aio_fsync $fh, $callback->($status)
		256	aio_fdatasync $fh, $callback->($status)
		257	aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		258	aio_pathsync $pathname, $callback->($status)
		259	aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
		260	aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		261	aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		262	aio_mlockall $flags, $callback->($status)
		263	aio_group $callback->(...)
		264	aio_nop $callback->()
		265
		266	$prev_pri = aioreq_pri [$pri]
		267	aioreq_nice $pri_adjust
		268
		269	IO::AIO::poll_wait
		270	IO::AIO::poll_cb
		271	IO::AIO::poll
		272	IO::AIO::flush
		273	IO::AIO::max_poll_reqs $nreqs
		274	IO::AIO::max_poll_time $seconds
		275	IO::AIO::min_parallel $nthreads
		276	IO::AIO::max_parallel $nthreads
		277	IO::AIO::max_idle $nthreads
		278	IO::AIO::idle_timeout $seconds
		279	IO::AIO::max_outstanding $maxreqs
		280	IO::AIO::nreqs
		281	IO::AIO::nready
		282	IO::AIO::npending
		283	$nfd = IO::AIO::get_fdlimit [EXPERIMENTAL]
		284	IO::AIO::min_fdlimit $nfd [EXPERIMENTAL]
		285
		286	IO::AIO::sendfile $ofh, $ifh, $offset, $count
		287	IO::AIO::fadvise $fh, $offset, $len, $advice
		288	IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
		289	IO::AIO::munmap $scalar
		290	IO::AIO::madvise $scalar, $offset, $length, $advice
		291	IO::AIO::mprotect $scalar, $offset, $length, $protect
		292	IO::AIO::munlock $scalar, $offset = 0, $length = undef
		293	IO::AIO::munlockall
		294
		295	=head2 API NOTES
72		296
73	All the C<aio_*> calls are more or less thin wrappers around the syscall	297	All the C<aio_*> calls are more or less thin wrappers around the syscall
74	with the same name (sans C<aio_>). The arguments are similar or identical,	298	with the same name (sans C<aio_>). The arguments are similar or identical,
75	and they all accept an additional C<$callback> argument which must be	299	and they all accept an additional (and optional) C<$callback> argument
76	a code reference. This code reference will get called with the syscall	300	which must be a code reference. This code reference will be called after
		301	the syscall has been executed in an asynchronous fashion. The results
		302	of the request will be passed as arguments to the callback (and, if an
		303	error occured, in C<$!>) - for most requests the syscall return code (e.g.
77	return code (e.g. most syscalls return C<-1> on error, unlike perl, which	304	most syscalls return C<-1> on error, unlike perl, which usually delivers
78	usually delivers "false") as it's sole argument when the given syscall has	305	"false").
79	been executed asynchronously.
80		306
81	All functions that expect a filehandle will also accept a file descriptor.	307	Some requests (such as C<aio_readdir>) pass the actual results and
		308	communicate failures by passing C<undef>.
82		309
		310	All functions expecting a filehandle keep a copy of the filehandle
		311	internally until the request has finished.
		312
		313	All functions return request objects of type L<IO::AIO::REQ> that allow
		314	further manipulation of those requests while they are in-flight.
		315
83	The filenames you pass to these routines I<must> be absolute. The reason	316	The pathnames you pass to these routines I<should> be absolute. The
84	is that at the time the request is being executed, the current working	317	reason for this is that at the time the request is being executed, the
85	directory could have changed. Alternatively, you can make sure that you	318	current working directory could have changed. Alternatively, you can
86	never change the current working directory.	319	make sure that you never change the current working directory anywhere
		320	in the program and then use relative paths. You can also take advantage
		321	of IO::AIOs working directory abstraction, that lets you specify paths
		322	relative to some previously-opened "working directory object" - see the
		323	description of the C<IO::AIO::WD> class later in this document.
		324
		325	To encode pathnames as octets, either make sure you either: a) always pass
		326	in filenames you got from outside (command line, readdir etc.) without
		327	tinkering, b) are in your native filesystem encoding, c) use the Encode
		328	module and encode your pathnames to the locale (or other) encoding in
		329	effect in the user environment, d) use Glib::filename_from_unicode on
		330	unicode filenames or e) use something else to ensure your scalar has the
		331	correct contents.
		332
		333	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		334	handles correctly whether it is set or not.
		335
		336	=head2 AIO REQUEST FUNCTIONS
87		337
88	=over 4	338	=over 4
89		339
		340	=item $prev_pri = aioreq_pri [$pri]
		341
		342	Returns the priority value that would be used for the next request and, if
		343	C<$pri> is given, sets the priority for the next aio request.
		344
		345	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		346	and C<4>, respectively. Requests with higher priority will be serviced
		347	first.
		348
		349	The priority will be reset to C<0> after each call to one of the C<aio_*>
		350	functions.
		351
		352	Example: open a file with low priority, then read something from it with
		353	higher priority so the read request is serviced before other low priority
		354	open requests (potentially spamming the cache):
		355
		356	aioreq_pri -3;
		357	aio_open ..., sub {
		358	return unless $_[0];
		359
		360	aioreq_pri -2;
		361	aio_read $_[0], ..., sub {
		362	...
		363	};
		364	};
		365
		366
		367	=item aioreq_nice $pri_adjust
		368
		369	Similar to C<aioreq_pri>, but subtracts the given value from the current
		370	priority, so the effect is cumulative.
		371
		372
90	=item aio_open $pathname, $flags, $mode, $callback	373	=item aio_open $pathname, $flags, $mode, $callback->($fh)
91		374
92	Asynchronously open or create a file and call the callback with a newly	375	Asynchronously open or create a file and call the callback with a newly
93	created filehandle for the file.	376	created filehandle for the file (or C<undef> in case of an error).
94		377
95	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,	378	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
96	for an explanation.	379	for an explanation.
97		380
98	The C<$mode> argument is a bitmask. See the C<Fcntl> module for a	381	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
99	list. They are the same as used in C<sysopen>.	382	list. They are the same as used by C<sysopen>.
		383
		384	Likewise, C<$mode> specifies the mode of the newly created file, if it
		385	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
		386	except that it is mandatory (i.e. use C<0> if you don't create new files,
		387	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		388	by the umask in effect then the request is being executed, so better never
		389	change the umask.
100		390
101	Example:	391	Example:
102		392
103	aio_open "/etc/passwd", O_RDONLY, 0, sub {	393	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
104	if ($_[0]) {	394	if ($_[0]) {
105	print "open successful, fh is $_[0]\n";	395	print "open successful, fh is $_[0]\n";
106	...	396	...
107	} else {	397	} else {
108	die "open failed: $!\n";	398	die "open failed: $!\n";
109	}	399	}
110	};	400	};
111		401
		402	In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
		403	C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
		404	following POSIX and non-POSIX constants are available (missing ones on
		405	your system are, as usual, C<0>):
		406
		407	C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
		408	C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
		409	C<O_RSYNC>, C<O_SYNC>, C<O_PATH>, C<O_TMPFILE>, and C<O_TTY_INIT>.
		410
		411
112	=item aio_close $fh, $callback	412	=item aio_close $fh, $callback->($status)
113		413
114	Asynchronously close a file and call the callback with the result	414	Asynchronously close a file and call the callback with the result
115	code. I<WARNING:> although accepted, you should not pass in a perl	415	code.
116	filehandle here, as perl will likely close the file descriptor itself when
117	the filehandle is destroyed. Normally, you can safely call perls C<close>
118	or just let filehandles go out of scope.
119		416
		417	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
		418	closing the file descriptor associated with the filehandle itself.
		419
		420	Therefore, C<aio_close> will not close the filehandle - instead it will
		421	use dup2 to overwrite the file descriptor with the write-end of a pipe
		422	(the pipe fd will be created on demand and will be cached).
		423
		424	Or in other words: the file descriptor will be closed, but it will not be
		425	free for reuse until the perl filehandle is closed.
		426
		427	=cut
		428
		429	=item aio_seek $fh, $offset, $whence, $callback->($offs)
		430
		431	Seeks the filehandle to the new C<$offset>, similarly to perl's
		432	C<sysseek>. The C<$whence> can use the traditional values (C<0> for
		433	C<IO::AIO::SEEK_SET>, C<1> for C<IO::AIO::SEEK_CUR> or C<2> for
		434	C<IO::AIO::SEEK_END>).
		435
		436	The resulting absolute offset will be passed to the callback, or C<-1> in
		437	case of an error.
		438
		439	In theory, the C<$whence> constants could be different than the
		440	corresponding values from L<Fcntl>, but perl guarantees they are the same,
		441	so don't panic.
		442
		443	As a GNU/Linux (and maybe Solaris) extension, also the constants
		444	C<IO::AIO::SEEK_DATA> and C<IO::AIO::SEEK_HOLE> are available, if they
		445	could be found. No guarantees about suitability for use in C<aio_seek> or
		446	Perl's C<sysseek> can be made though, although I would naively assume they
		447	"just work".
		448
120	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	449	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
121		450
122	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	451	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
123		452
124	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	453	Reads or writes C<$length> bytes from or to the specified C<$fh> and
125	into the scalar given by C<data> and offset C<dataoffset> and calls the	454	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> and
126	callback without the actual number of bytes read (or -1 on error, just	455	calls the callback with the actual number of bytes transferred (or -1 on
127	like the syscall).	456	error, just like the syscall).
128		457
		458	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		459	offset plus the actual number of bytes read.
		460
		461	If C<$offset> is undefined, then the current file descriptor offset will
		462	be used (and updated), otherwise the file descriptor offset will not be
		463	changed by these calls.
		464
		465	If C<$length> is undefined in C<aio_write>, use the remaining length of
		466	C<$data>.
		467
		468	If C<$dataoffset> is less than zero, it will be counted from the end of
		469	C<$data>.
		470
		471	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		472	is outstanding. Modifying it can result in segfaults or World War III (if
		473	the necessary/optional hardware is installed).
		474
129	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at	475	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
130	offset C<0> within the scalar:	476	offset C<0> within the scalar:
131		477
132	aio_read $fh, 7, 15, $buffer, 0, sub {	478	aio_read $fh, 7, 15, $buffer, 0, sub {
133	$_[0] > 0 or die "read error: $!";	479	$_[0] > 0 or die "read error: $!";
134	print "read $_[0] bytes: <$buffer>\n";	480	print "read $_[0] bytes: <$buffer>\n";
135	};	481	};
136		482
		483
		484	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		485
		486	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		487	reading at byte offset C<$in_offset>, and starts writing at the current
		488	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		489	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		490	other. The same C<$in_fh> works fine though, as this function does not
		491	move or use the file offset of C<$in_fh>.
		492
		493	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		494	are written, and there is no way to find out how many more bytes have been
		495	read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
		496	number of bytes written to C<$out_fh>. Only if the result value equals
		497	C<$length> one can assume that C<$length> bytes have been read.
		498
		499	Unlike with other C<aio_> functions, it makes a lot of sense to use
		500	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		501	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		502	the socket I/O will be non-blocking. Note, however, that you can run
		503	into a trap where C<aio_sendfile> reads some data with readahead, then
		504	fails to write all data, and when the socket is ready the next time, the
		505	data in the cache is already lost, forcing C<aio_sendfile> to again hit
		506	the disk. Explicit C<aio_read> + C<aio_write> let's you better control
		507	resource usage.
		508
		509	This call tries to make use of a native C<sendfile>-like syscall to
		510	provide zero-copy operation. For this to work, C<$out_fh> should refer to
		511	a socket, and C<$in_fh> should refer to an mmap'able file.
		512
		513	If a native sendfile cannot be found or it fails with C<ENOSYS>,
		514	C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
		515	C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
		516	type of filehandle regardless of the limitations of the operating system.
		517
		518	As native sendfile syscalls (as practically any non-POSIX interface hacked
		519	together in a hurry to improve benchmark numbers) tend to be rather buggy
		520	on many systems, this implementation tries to work around some known bugs
		521	in Linux and FreeBSD kernels (probably others, too), but that might fail,
		522	so you really really should check the return value of C<aio_sendfile> -
		523	fewer bytes than expected might have been transferred.
		524
		525
137	=item aio_readahead $fh,$offset,$length, $callback	526	=item aio_readahead $fh,$offset,$length, $callback->($retval)
138		527
139	Asynchronously reads the specified byte range into the page cache, using
140	the C<readahead> syscall. If that syscall doesn't exist the status will be
141	C<-1> and C<$!> is set to ENOSYS.
142
143	readahead() populates the page cache with data from a file so that	528	C<aio_readahead> populates the page cache with data from a file so that
144	subsequent reads from that file will not block on disk I/O. The C<$offset>	529	subsequent reads from that file will not block on disk I/O. The C<$offset>
145	argument specifies the starting point from which data is to be read and	530	argument specifies the starting point from which data is to be read and
146	C<$length> specifies the number of bytes to be read. I/O is performed in	531	C<$length> specifies the number of bytes to be read. I/O is performed in
147	whole pages, so that offset is effectively rounded down to a page boundary	532	whole pages, so that offset is effectively rounded down to a page boundary
148	and bytes are read up to the next page boundary greater than or equal to	533	and bytes are read up to the next page boundary greater than or equal to
149	(off-set+length). aio_readahead() does not read beyond the end of the	534	(off-set+length). C<aio_readahead> does not read beyond the end of the
150	file. The current file offset of the file is left unchanged.	535	file. The current file offset of the file is left unchanged.
151		536
		537	If that syscall doesn't exist (likely if your kernel isn't Linux) it will
		538	be emulated by simply reading the data, which would have a similar effect.
		539
		540
152	=item aio_stat $fh_or_path, $callback	541	=item aio_stat $fh_or_path, $callback->($status)
153		542
154	=item aio_lstat $fh, $callback	543	=item aio_lstat $fh, $callback->($status)
155		544
156	Works like perl's C<stat> or C<lstat> in void context. The callback will	545	Works like perl's C<stat> or C<lstat> in void context. The callback will
157	be called after the stat and the results will be available using C<stat _>	546	be called after the stat and the results will be available using C<stat _>
158	or C<-s _> etc...	547	or C<-s _> etc...
159		548
…		…
161	for an explanation.	550	for an explanation.
162		551
163	Currently, the stats are always 64-bit-stats, i.e. instead of returning an	552	Currently, the stats are always 64-bit-stats, i.e. instead of returning an
164	error when stat'ing a large file, the results will be silently truncated	553	error when stat'ing a large file, the results will be silently truncated
165	unless perl itself is compiled with large file support.	554	unless perl itself is compiled with large file support.
		555
		556	To help interpret the mode and dev/rdev stat values, IO::AIO offers the
		557	following constants and functions (if not implemented, the constants will
		558	be C<0> and the functions will either C<croak> or fall back on traditional
		559	behaviour).
		560
		561	C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
		562	C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
		563	C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
166		564
167	Example: Print the length of F</etc/passwd>:	565	Example: Print the length of F</etc/passwd>:
168		566
169	aio_stat "/etc/passwd", sub {	567	aio_stat "/etc/passwd", sub {
170	$_[0] and die "stat failed: $!";	568	$_[0] and die "stat failed: $!";
171	print "size is ", -s _, "\n";	569	print "size is ", -s _, "\n";
172	};	570	};
173		571
		572
		573	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		574
		575	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		576	whether a file handle or path was passed.
		577
		578	On success, the callback is passed a hash reference with the following
		579	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		580	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		581	is passed.
		582
		583	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		584	C<ST_NOSUID>.
		585
		586	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		587	their correct value when available, or to C<0> on systems that do
		588	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		589	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		590	C<ST_NODIRATIME> and C<ST_RELATIME>.
		591
		592	Example: stat C</wd> and dump out the data if successful.
		593
		594	aio_statvfs "/wd", sub {
		595	my $f = $_[0]
		596	or die "statvfs: $!";
		597
		598	use Data::Dumper;
		599	say Dumper $f;
		600	};
		601
		602	# result:
		603	{
		604	bsize => 1024,
		605	bfree => 4333064312,
		606	blocks => 10253828096,
		607	files => 2050765568,
		608	flag => 4096,
		609	favail => 2042092649,
		610	bavail => 4333064312,
		611	ffree => 2042092649,
		612	namemax => 255,
		613	frsize => 1024,
		614	fsid => 1810
		615	}
		616
		617	Here is a (likely partial - send me updates!) list of fsid values used by
		618	Linux - it is safe to hardcode these when C<$^O> is C<linux>:
		619
		620	0x0000adf5 adfs
		621	0x0000adff affs
		622	0x5346414f afs
		623	0x09041934 anon-inode filesystem
		624	0x00000187 autofs
		625	0x42465331 befs
		626	0x1badface bfs
		627	0x42494e4d binfmt_misc
		628	0x9123683e btrfs
		629	0x0027e0eb cgroupfs
		630	0xff534d42 cifs
		631	0x73757245 coda
		632	0x012ff7b7 coh
		633	0x28cd3d45 cramfs
		634	0x453dcd28 cramfs-wend (wrong endianness)
		635	0x64626720 debugfs
		636	0x00001373 devfs
		637	0x00001cd1 devpts
		638	0x0000f15f ecryptfs
		639	0x00414a53 efs
		640	0x0000137d ext
		641	0x0000ef53 ext2/ext3/ext4
		642	0x0000ef51 ext2
		643	0xf2f52010 f2fs
		644	0x00004006 fat
		645	0x65735546 fuseblk
		646	0x65735543 fusectl
		647	0x0bad1dea futexfs
		648	0x01161970 gfs2
		649	0x47504653 gpfs
		650	0x00004244 hfs
		651	0xf995e849 hpfs
		652	0x00c0ffee hostfs
		653	0x958458f6 hugetlbfs
		654	0x2bad1dea inotifyfs
		655	0x00009660 isofs
		656	0x000072b6 jffs2
		657	0x3153464a jfs
		658	0x6b414653 k-afs
		659	0x0bd00bd0 lustre
		660	0x0000137f minix
		661	0x0000138f minix 30 char names
		662	0x00002468 minix v2
		663	0x00002478 minix v2 30 char names
		664	0x00004d5a minix v3
		665	0x19800202 mqueue
		666	0x00004d44 msdos
		667	0x0000564c novell
		668	0x00006969 nfs
		669	0x6e667364 nfsd
		670	0x00003434 nilfs
		671	0x5346544e ntfs
		672	0x00009fa1 openprom
		673	0x7461636F ocfs2
		674	0x00009fa0 proc
		675	0x6165676c pstorefs
		676	0x0000002f qnx4
		677	0x68191122 qnx6
		678	0x858458f6 ramfs
		679	0x52654973 reiserfs
		680	0x00007275 romfs
		681	0x67596969 rpc_pipefs
		682	0x73636673 securityfs
		683	0xf97cff8c selinux
		684	0x0000517b smb
		685	0x534f434b sockfs
		686	0x73717368 squashfs
		687	0x62656572 sysfs
		688	0x012ff7b6 sysv2
		689	0x012ff7b5 sysv4
		690	0x01021994 tmpfs
		691	0x15013346 udf
		692	0x00011954 ufs
		693	0x54190100 ufs byteswapped
		694	0x00009fa2 usbdevfs
		695	0x01021997 v9fs
		696	0xa501fcf5 vxfs
		697	0xabba1974 xenfs
		698	0x012ff7b4 xenix
		699	0x58465342 xfs
		700	0x012fd16d xia
		701
		702	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		703
		704	Works like perl's C<utime> function (including the special case of $atime
		705	and $mtime being undef). Fractional times are supported if the underlying
		706	syscalls support them.
		707
		708	When called with a pathname, uses utimes(2) if available, otherwise
		709	utime(2). If called on a file descriptor, uses futimes(2) if available,
		710	otherwise returns ENOSYS, so this is not portable.
		711
		712	Examples:
		713
		714	# set atime and mtime to current time (basically touch(1)):
		715	aio_utime "path", undef, undef;
		716	# set atime to current time and mtime to beginning of the epoch:
		717	aio_utime "path", time, undef; # undef==0
		718
		719
		720	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		721
		722	Works like perl's C<chown> function, except that C<undef> for either $uid
		723	or $gid is being interpreted as "do not change" (but -1 can also be used).
		724
		725	Examples:
		726
		727	# same as "chown root path" in the shell:
		728	aio_chown "path", 0, -1;
		729	# same as above:
		730	aio_chown "path", 0, undef;
		731
		732
		733	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		734
		735	Works like truncate(2) or ftruncate(2).
		736
		737
		738	=item aio_allocate $fh, $mode, $offset, $len, $callback->($status)
		739
		740	Allocates or frees disk space according to the C<$mode> argument. See the
		741	linux C<fallocate> documentation for details.
		742
		743	C<$mode> is usually C<0> or C<IO::AIO::FALLOC_FL_KEEP_SIZE> to allocate
		744	space, or C<IO::AIO::FALLOC_FL_PUNCH_HOLE | IO::AIO::FALLOC_FL_KEEP_SIZE>,
		745	to deallocate a file range.
		746
		747	IO::AIO also supports C<FALLOC_FL_COLLAPSE_RANGE>, to remove a range
		748	(without leaving a hole), C<FALLOC_FL_ZERO_RANGE>, to zero a range,
		749	C<FALLOC_FL_INSERT_RANGE> to insert a range and C<FALLOC_FL_UNSHARE_RANGE>
		750	to unshare shared blocks (see your L<fallocate(2)> manpage).
		751
		752	The file system block size used by C<fallocate> is presumably the
		753	C<f_bsize> returned by C<statvfs>, but different filesystems and filetypes
		754	can dictate other limitations.
		755
		756	If C<fallocate> isn't available or cannot be emulated (currently no
		757	emulation will be attempted), passes C<-1> and sets C<$!> to C<ENOSYS>.
		758
		759
		760	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		761
		762	Works like perl's C<chmod> function.
		763
		764
174	=item aio_unlink $pathname, $callback	765	=item aio_unlink $pathname, $callback->($status)
175		766
176	Asynchronously unlink (delete) a file and call the callback with the	767	Asynchronously unlink (delete) a file and call the callback with the
177	result code.	768	result code.
178		769
		770
		771	=item aio_mknod $pathname, $mode, $dev, $callback->($status)
		772
		773	[EXPERIMENTAL]
		774
		775	Asynchronously create a device node (or fifo). See mknod(2).
		776
		777	The only (POSIX-) portable way of calling this function is:
		778
		779	aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		780
		781	See C<aio_stat> for info about some potentially helpful extra constants
		782	and functions.
		783
		784	=item aio_link $srcpath, $dstpath, $callback->($status)
		785
		786	Asynchronously create a new link to the existing object at C<$srcpath> at
		787	the path C<$dstpath> and call the callback with the result code.
		788
		789
		790	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		791
		792	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		793	the path C<$dstpath> and call the callback with the result code.
		794
		795
		796	=item aio_readlink $pathname, $callback->($link)
		797
		798	Asynchronously read the symlink specified by C<$path> and pass it to
		799	the callback. If an error occurs, nothing or undef gets passed to the
		800	callback.
		801
		802
		803	=item aio_realpath $pathname, $callback->($path)
		804
		805	Asynchronously make the path absolute and resolve any symlinks in
		806	C<$path>. The resulting path only consists of directories (same as
		807	L<Cwd::realpath>).
		808
		809	This request can be used to get the absolute path of the current working
		810	directory by passing it a path of F<.> (a single dot).
		811
		812
		813	=item aio_rename $srcpath, $dstpath, $callback->($status)
		814
		815	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		816	rename(2) and call the callback with the result code.
		817
		818	On systems that support the AIO::WD working directory abstraction
		819	natively, the case C<[$wd, "."]> as C<$srcpath> is specialcased - instead
		820	of failing, C<rename> is called on the absolute path of C<$wd>.
		821
		822
		823	=item aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
		824
		825	Basically a version of C<aio_rename> with an additional C<$flags>
		826	argument. Calling this with C<$flags=0> is the same as calling
		827	C<aio_rename>.
		828
		829	Non-zero flags are currently only supported on GNU/Linux systems that
		830	support renameat2. Other systems fail with C<ENOSYS> in this case.
		831
		832	The following constants are available (missing ones are, as usual C<0>),
		833	see renameat2(2) for details:
		834
		835	C<IO::AIO::RENAME_NOREPLACE>, C<IO::AIO::RENAME_EXCHANGE>
		836	and C<IO::AIO::RENAME_WHITEOUT>.
		837
		838
		839	=item aio_mkdir $pathname, $mode, $callback->($status)
		840
		841	Asynchronously mkdir (create) a directory and call the callback with
		842	the result code. C<$mode> will be modified by the umask at the time the
		843	request is executed, so do not change your umask.
		844
		845
		846	=item aio_rmdir $pathname, $callback->($status)
		847
		848	Asynchronously rmdir (delete) a directory and call the callback with the
		849	result code.
		850
		851	On systems that support the AIO::WD working directory abstraction
		852	natively, the case C<[$wd, "."]> is specialcased - instead of failing,
		853	C<rmdir> is called on the absolute path of C<$wd>.
		854
		855
		856	=item aio_readdir $pathname, $callback->($entries)
		857
		858	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		859	directory (i.e. opendir + readdir + closedir). The entries will not be
		860	sorted, and will B<NOT> include the C<.> and C<..> entries.
		861
		862	The callback is passed a single argument which is either C<undef> or an
		863	array-ref with the filenames.
		864
		865
		866	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		867
		868	Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
		869	tune behaviour and output format. In case of an error, C<$entries> will be
		870	C<undef>.
		871
		872	The flags are a combination of the following constants, ORed together (the
		873	flags will also be passed to the callback, possibly modified):
		874
		875	=over 4
		876
		877	=item IO::AIO::READDIR_DENTS
		878
		879	When this flag is off, then the callback gets an arrayref consisting of
		880	names only (as with C<aio_readdir>), otherwise it gets an arrayref with
		881	C<[$name, $type, $inode]> arrayrefs, each describing a single directory
		882	entry in more detail.
		883
		884	C<$name> is the name of the entry.
		885
		886	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		887
		888	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		889	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		890	C<IO::AIO::DT_WHT>.
		891
		892	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
		893	know, you have to run stat yourself. Also, for speed reasons, the C<$type>
		894	scalars are read-only: you can not modify them.
		895
		896	C<$inode> is the inode number (which might not be exact on systems with 64
		897	bit inode numbers and 32 bit perls). This field has unspecified content on
		898	systems that do not deliver the inode information.
		899
		900	=item IO::AIO::READDIR_DIRS_FIRST
		901
		902	When this flag is set, then the names will be returned in an order where
		903	likely directories come first, in optimal stat order. This is useful when
		904	you need to quickly find directories, or you want to find all directories
		905	while avoiding to stat() each entry.
		906
		907	If the system returns type information in readdir, then this is used
		908	to find directories directly. Otherwise, likely directories are names
		909	beginning with ".", or otherwise names with no dots, of which names with
		910	short names are tried first.
		911
		912	=item IO::AIO::READDIR_STAT_ORDER
		913
		914	When this flag is set, then the names will be returned in an order
		915	suitable for stat()'ing each one. That is, when you plan to stat()
		916	all files in the given directory, then the returned order will likely
		917	be fastest.
		918
		919	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
		920	the likely dirs come first, resulting in a less optimal stat order.
		921
		922	=item IO::AIO::READDIR_FOUND_UNKNOWN
		923
		924	This flag should not be set when calling C<aio_readdirx>. Instead, it
		925	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		926	C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
		927	C<$type>'s are known, which can be used to speed up some algorithms.
		928
		929	=back
		930
		931
		932	=item aio_slurp $pathname, $offset, $length, $data, $callback->($status)
		933
		934	Opens, reads and closes the given file. The data is put into C<$data>,
		935	which is resized as required.
		936
		937	If C<$offset> is negative, then it is counted from the end of the file.
		938
		939	If C<$length> is zero, then the remaining length of the file is
		940	used. Also, in this case, the same limitations to modifying C<$data> apply
		941	as when IO::AIO::mmap is used, i.e. it must only be modified in-place
		942	with C<substr>. If the size of the file is known, specifying a non-zero
		943	C<$length> results in a performance advantage.
		944
		945	This request is similar to the older C<aio_load> request, but since it is
		946	a single request, it might be more efficient to use.
		947
		948	Example: load F</etc/passwd> into C<$passwd>.
		949
		950	my $passwd;
		951	aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
		952	$_[0] >= 0
		953	or die "/etc/passwd: $!\n";
		954
		955	printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
		956	print $passwd;
		957	};
		958	IO::AIO::flush;
		959
		960
		961	=item aio_load $pathname, $data, $callback->($status)
		962
		963	This is a composite request that tries to fully load the given file into
		964	memory. Status is the same as with aio_read.
		965
		966	Using C<aio_slurp> might be more efficient, as it is a single request.
		967
		968	=cut
		969
		970	sub aio_load($$;$) {
		971	my ($path, undef, $cb) = @_;
		972	my $data = \$_[1];
		973
		974	my $pri = aioreq_pri;
		975	my $grp = aio_group $cb;
		976
		977	aioreq_pri $pri;
		978	add $grp aio_open $path, O_RDONLY, 0, sub {
		979	my $fh = shift
		980	or return $grp->result (-1);
		981
		982	aioreq_pri $pri;
		983	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		984	$grp->result ($_[0]);
		985	};
		986	};
		987
		988	$grp
		989	}
		990
		991	=item aio_copy $srcpath, $dstpath, $callback->($status)
		992
		993	Try to copy the I<file> (directories not supported as either source or
		994	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		995	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		996
		997	Existing destination files will be truncated.
		998
		999	This is a composite request that creates the destination file with
		1000	mode 0200 and copies the contents of the source file into it using
		1001	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		1002	uid/gid, in that order.
		1003
		1004	If an error occurs, the partial destination file will be unlinked, if
		1005	possible, except when setting atime, mtime, access mode and uid/gid, where
		1006	errors are being ignored.
		1007
		1008	=cut
		1009
		1010	sub aio_copy($$;$) {
		1011	my ($src, $dst, $cb) = @_;
		1012
		1013	my $pri = aioreq_pri;
		1014	my $grp = aio_group $cb;
		1015
		1016	aioreq_pri $pri;
		1017	add $grp aio_open $src, O_RDONLY, 0, sub {
		1018	if (my $src_fh = $_[0]) {
		1019	my @stat = stat $src_fh; # hmm, might block over nfs?
		1020
		1021	aioreq_pri $pri;
		1022	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		1023	if (my $dst_fh = $_[0]) {
		1024	aioreq_pri $pri;
		1025	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		1026	if ($_[0] == $stat[7]) {
		1027	$grp->result (0);
		1028	close $src_fh;
		1029
		1030	my $ch = sub {
		1031	aioreq_pri $pri;
		1032	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		1033	aioreq_pri $pri;
		1034	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		1035	aioreq_pri $pri;
		1036	add $grp aio_close $dst_fh;
		1037	}
		1038	};
		1039	};
		1040
		1041	aioreq_pri $pri;
		1042	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		1043	if ($_[0] < 0 && $! == ENOSYS) {
		1044	aioreq_pri $pri;
		1045	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		1046	} else {
		1047	$ch->();
		1048	}
		1049	};
		1050	} else {
		1051	$grp->result (-1);
		1052	close $src_fh;
		1053	close $dst_fh;
		1054
		1055	aioreq $pri;
		1056	add $grp aio_unlink $dst;
		1057	}
		1058	};
		1059	} else {
		1060	$grp->result (-1);
		1061	}
		1062	},
		1063
		1064	} else {
		1065	$grp->result (-1);
		1066	}
		1067	};
		1068
		1069	$grp
		1070	}
		1071
		1072	=item aio_move $srcpath, $dstpath, $callback->($status)
		1073
		1074	Try to move the I<file> (directories not supported as either source or
		1075	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		1076	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		1077
		1078	This is a composite request that tries to rename(2) the file first; if
		1079	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		1080	that is successful, unlinks the C<$srcpath>.
		1081
		1082	=cut
		1083
		1084	sub aio_move($$;$) {
		1085	my ($src, $dst, $cb) = @_;
		1086
		1087	my $pri = aioreq_pri;
		1088	my $grp = aio_group $cb;
		1089
		1090	aioreq_pri $pri;
		1091	add $grp aio_rename $src, $dst, sub {
		1092	if ($_[0] && $! == EXDEV) {
		1093	aioreq_pri $pri;
		1094	add $grp aio_copy $src, $dst, sub {
		1095	$grp->result ($_[0]);
		1096
		1097	unless ($_[0]) {
		1098	aioreq_pri $pri;
		1099	add $grp aio_unlink $src;
		1100	}
		1101	};
		1102	} else {
		1103	$grp->result ($_[0]);
		1104	}
		1105	};
		1106
		1107	$grp
		1108	}
		1109
		1110	=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		1111
		1112	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		1113	efficiently separate the entries of directory C<$path> into two sets of
		1114	names, directories you can recurse into (directories), and ones you cannot
		1115	recurse into (everything else, including symlinks to directories).
		1116
		1117	C<aio_scandir> is a composite request that generates many sub requests.
		1118	C<$maxreq> specifies the maximum number of outstanding aio requests that
		1119	this function generates. If it is C<< <= 0 >>, then a suitable default
		1120	will be chosen (currently 4).
		1121
		1122	On error, the callback is called without arguments, otherwise it receives
		1123	two array-refs with path-relative entry names.
		1124
		1125	Example:
		1126
		1127	aio_scandir $dir, 0, sub {
		1128	my ($dirs, $nondirs) = @_;
		1129	print "real directories: @$dirs\n";
		1130	print "everything else: @$nondirs\n";
		1131	};
		1132
		1133	Implementation notes.
		1134
		1135	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		1136
		1137	If readdir returns file type information, then this is used directly to
		1138	find directories.
		1139
		1140	Otherwise, after reading the directory, the modification time, size etc.
		1141	of the directory before and after the readdir is checked, and if they
		1142	match (and isn't the current time), the link count will be used to decide
		1143	how many entries are directories (if >= 2). Otherwise, no knowledge of the
		1144	number of subdirectories will be assumed.
		1145
		1146	Then entries will be sorted into likely directories a non-initial dot
		1147	currently) and likely non-directories (see C<aio_readdirx>). Then every
		1148	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
		1149	in order of their inode numbers. If that succeeds, it assumes that the
		1150	entry is a directory or a symlink to directory (which will be checked
		1151	separately). This is often faster than stat'ing the entry itself because
		1152	filesystems might detect the type of the entry without reading the inode
		1153	data (e.g. ext2fs filetype feature), even on systems that cannot return
		1154	the filetype information on readdir.
		1155
		1156	If the known number of directories (link count - 2) has been reached, the
		1157	rest of the entries is assumed to be non-directories.
		1158
		1159	This only works with certainty on POSIX (= UNIX) filesystems, which
		1160	fortunately are the vast majority of filesystems around.
		1161
		1162	It will also likely work on non-POSIX filesystems with reduced efficiency
		1163	as those tend to return 0 or 1 as link counts, which disables the
		1164	directory counting heuristic.
		1165
		1166	=cut
		1167
		1168	sub aio_scandir($$;$) {
		1169	my ($path, $maxreq, $cb) = @_;
		1170
		1171	my $pri = aioreq_pri;
		1172
		1173	my $grp = aio_group $cb;
		1174
		1175	$maxreq = 4 if $maxreq <= 0;
		1176
		1177	# get a wd object
		1178	aioreq_pri $pri;
		1179	add $grp aio_wd $path, sub {
		1180	$_[0]
		1181	or return $grp->result ();
		1182
		1183	my $wd = [shift, "."];
		1184
		1185	# stat once
		1186	aioreq_pri $pri;
		1187	add $grp aio_stat $wd, sub {
		1188	return $grp->result () if $_[0];
		1189	my $now = time;
		1190	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		1191
		1192	# read the directory entries
		1193	aioreq_pri $pri;
		1194	add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
		1195	my $entries = shift
		1196	or return $grp->result ();
		1197
		1198	# stat the dir another time
		1199	aioreq_pri $pri;
		1200	add $grp aio_stat $wd, sub {
		1201	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		1202
		1203	my $ndirs;
		1204
		1205	# take the slow route if anything looks fishy
		1206	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		1207	$ndirs = -1;
		1208	} else {
		1209	# if nlink == 2, we are finished
		1210	# for non-posix-fs's, we rely on nlink < 2
		1211	$ndirs = (stat _)[3] - 2
		1212	or return $grp->result ([], $entries);
		1213	}
		1214
		1215	my (@dirs, @nondirs);
		1216
		1217	my $statgrp = add $grp aio_group sub {
		1218	$grp->result (\@dirs, \@nondirs);
		1219	};
		1220
		1221	limit $statgrp $maxreq;
		1222	feed $statgrp sub {
		1223	return unless @$entries;
		1224	my $entry = shift @$entries;
		1225
		1226	aioreq_pri $pri;
		1227	$wd->[1] = "$entry/.";
		1228	add $statgrp aio_stat $wd, sub {
		1229	if ($_[0] < 0) {
		1230	push @nondirs, $entry;
		1231	} else {
		1232	# need to check for real directory
		1233	aioreq_pri $pri;
		1234	$wd->[1] = $entry;
		1235	add $statgrp aio_lstat $wd, sub {
		1236	if (-d _) {
		1237	push @dirs, $entry;
		1238
		1239	unless (--$ndirs) {
		1240	push @nondirs, @$entries;
		1241	feed $statgrp;
		1242	}
		1243	} else {
		1244	push @nondirs, $entry;
		1245	}
		1246	}
		1247	}
		1248	};
		1249	};
		1250	};
		1251	};
		1252	};
		1253	};
		1254
		1255	$grp
		1256	}
		1257
		1258	=item aio_rmtree $pathname, $callback->($status)
		1259
		1260	Delete a directory tree starting (and including) C<$path>, return the
		1261	status of the final C<rmdir> only. This is a composite request that
		1262	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1263	everything else.
		1264
		1265	=cut
		1266
		1267	sub aio_rmtree;
		1268	sub aio_rmtree($;$) {
		1269	my ($path, $cb) = @_;
		1270
		1271	my $pri = aioreq_pri;
		1272	my $grp = aio_group $cb;
		1273
		1274	aioreq_pri $pri;
		1275	add $grp aio_scandir $path, 0, sub {
		1276	my ($dirs, $nondirs) = @_;
		1277
		1278	my $dirgrp = aio_group sub {
		1279	add $grp aio_rmdir $path, sub {
		1280	$grp->result ($_[0]);
		1281	};
		1282	};
		1283
		1284	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1285	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1286
		1287	add $grp $dirgrp;
		1288	};
		1289
		1290	$grp
		1291	}
		1292
		1293	=item aio_fcntl $fh, $cmd, $arg, $callback->($status)
		1294
		1295	=item aio_ioctl $fh, $request, $buf, $callback->($status)
		1296
		1297	These work just like the C<fcntl> and C<ioctl> built-in functions, except
		1298	they execute asynchronously and pass the return value to the callback.
		1299
		1300	Both calls can be used for a lot of things, some of which make more sense
		1301	to run asynchronously in their own thread, while some others make less
		1302	sense. For example, calls that block waiting for external events, such
		1303	as locking, will also lock down an I/O thread while it is waiting, which
		1304	can deadlock the whole I/O system. At the same time, there might be no
		1305	alternative to using a thread to wait.
		1306
		1307	So in general, you should only use these calls for things that do
		1308	(filesystem) I/O, not for things that wait for other events (network,
		1309	other processes), although if you are careful and know what you are doing,
		1310	you still can.
		1311
		1312	The following constants are available (missing ones are, as usual C<0>):
		1313
		1314	C<F_DUPFD_CLOEXEC>,
		1315
		1316	C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>,
		1317
		1318	C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>.
		1319
		1320	C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>,
		1321	C<FS_IOC_FIEMAP>.
		1322
		1323	C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>,
		1324	C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>.
		1325
		1326	C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>,
		1327	C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>,
		1328	C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>,
		1329	C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>,
		1330	C<FS_FL_USER_MODIFIABLE>.
		1331
		1332	C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>,
		1333	C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>,
		1334	C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>,
		1335	C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>,
		1336
		1337	=item aio_sync $callback->($status)
		1338
		1339	Asynchronously call sync and call the callback when finished.
		1340
179	=item aio_fsync $fh, $callback	1341	=item aio_fsync $fh, $callback->($status)
180		1342
181	Asynchronously call fsync on the given filehandle and call the callback	1343	Asynchronously call fsync on the given filehandle and call the callback
182	with the fsync result code.	1344	with the fsync result code.
183		1345
184	=item aio_fdatasync $fh, $callback	1346	=item aio_fdatasync $fh, $callback->($status)
185		1347
186	Asynchronously call fdatasync on the given filehandle and call the	1348	Asynchronously call fdatasync on the given filehandle and call the
187	callback with the fdatasync result code.	1349	callback with the fdatasync result code.
188		1350
		1351	If this call isn't available because your OS lacks it or it couldn't be
		1352	detected, it will be emulated by calling C<fsync> instead.
		1353
		1354	=item aio_syncfs $fh, $callback->($status)
		1355
		1356	Asynchronously call the syncfs syscall to sync the filesystem associated
		1357	to the given filehandle and call the callback with the syncfs result
		1358	code. If syncfs is not available, calls sync(), but returns C<-1> and sets
		1359	errno to C<ENOSYS> nevertheless.
		1360
		1361	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1362
		1363	Sync the data portion of the file specified by C<$offset> and C<$length>
		1364	to disk (but NOT the metadata), by calling the Linux-specific
		1365	sync_file_range call. If sync_file_range is not available or it returns
		1366	ENOSYS, then fdatasync or fsync is being substituted.
		1367
		1368	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1369	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1370	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1371	manpage for details.
		1372
		1373	=item aio_pathsync $pathname, $callback->($status)
		1374
		1375	This request tries to open, fsync and close the given path. This is a
		1376	composite request intended to sync directories after directory operations
		1377	(E.g. rename). This might not work on all operating systems or have any
		1378	specific effect, but usually it makes sure that directory changes get
		1379	written to disc. It works for anything that can be opened for read-only,
		1380	not just directories.
		1381
		1382	Future versions of this function might fall back to other methods when
		1383	C<fsync> on the directory fails (such as calling C<sync>).
		1384
		1385	Passes C<0> when everything went ok, and C<-1> on error.
		1386
		1387	=cut
		1388
		1389	sub aio_pathsync($;$) {
		1390	my ($path, $cb) = @_;
		1391
		1392	my $pri = aioreq_pri;
		1393	my $grp = aio_group $cb;
		1394
		1395	aioreq_pri $pri;
		1396	add $grp aio_open $path, O_RDONLY, 0, sub {
		1397	my ($fh) = @_;
		1398	if ($fh) {
		1399	aioreq_pri $pri;
		1400	add $grp aio_fsync $fh, sub {
		1401	$grp->result ($_[0]);
		1402
		1403	aioreq_pri $pri;
		1404	add $grp aio_close $fh;
		1405	};
		1406	} else {
		1407	$grp->result (-1);
		1408	}
		1409	};
		1410
		1411	$grp
		1412	}
		1413
		1414	=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
		1415
		1416	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1417	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1418	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1419	scalar must only be modified in-place while an aio operation is pending on
		1420	it).
		1421
		1422	It calls the C<msync> function of your OS, if available, with the memory
		1423	area starting at C<$offset> in the string and ending C<$length> bytes
		1424	later. If C<$length> is negative, counts from the end, and if C<$length>
		1425	is C<undef>, then it goes till the end of the string. The flags can be
		1426	either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional
		1427	C<IO::AIO::MS_INVALIDATE>.
		1428
		1429	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1430
		1431	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1432	scalars.
		1433
		1434	It touches (reads or writes) all memory pages in the specified
		1435	range inside the scalar. All caveats and parameters are the same
		1436	as for C<aio_msync>, above, except for flags, which must be either
		1437	C<0> (which reads all pages and ensures they are instantiated) or
		1438	C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and
		1439	writing an octet from it, which dirties the page).
		1440
		1441	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1442
		1443	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1444	scalars.
		1445
		1446	It reads in all the pages of the underlying storage into memory (if any)
		1447	and locks them, so they are not getting swapped/paged out or removed.
		1448
		1449	If C<$length> is undefined, then the scalar will be locked till the end.
		1450
		1451	On systems that do not implement C<mlock>, this function returns C<-1>
		1452	and sets errno to C<ENOSYS>.
		1453
		1454	Note that the corresponding C<munlock> is synchronous and is
		1455	documented under L<MISCELLANEOUS FUNCTIONS>.
		1456
		1457	Example: open a file, mmap and mlock it - both will be undone when
		1458	C<$data> gets destroyed.
		1459
		1460	open my $fh, "<", $path or die "$path: $!";
		1461	my $data;
		1462	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1463	aio_mlock $data; # mlock in background
		1464
		1465	=item aio_mlockall $flags, $callback->($status)
		1466
		1467	Calls the C<mlockall> function with the given C<$flags> (a combination of
		1468	C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
		1469
		1470	On systems that do not implement C<mlockall>, this function returns C<-1>
		1471	and sets errno to C<ENOSYS>.
		1472
		1473	Note that the corresponding C<munlockall> is synchronous and is
		1474	documented under L<MISCELLANEOUS FUNCTIONS>.
		1475
		1476	Example: asynchronously lock all current and future pages into memory.
		1477
		1478	aio_mlockall IO::AIO::MCL_FUTURE;
		1479
		1480	=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
		1481
		1482	Queries the extents of the given file (by calling the Linux C<FIEMAP>
		1483	ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If
		1484	the ioctl is not available on your OS, then this request will fail with
		1485	C<ENOSYS>.
		1486
		1487	C<$start> is the starting offset to query extents for, C<$length> is the
		1488	size of the range to query - if it is C<undef>, then the whole file will
		1489	be queried.
		1490
		1491	C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or
		1492	C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also
		1493	exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query
		1494	the data portion.
		1495
		1496	C<$count> is the maximum number of extent records to return. If it is
		1497	C<undef>, then IO::AIO queries all extents of the range. As a very special
		1498	case, if it is C<0>, then the callback receives the number of extents
		1499	instead of the extents themselves (which is unreliable, see below).
		1500
		1501	If an error occurs, the callback receives no arguments. The special
		1502	C<errno> value C<IO::AIO::EBADR> is available to test for flag errors.
		1503
		1504	Otherwise, the callback receives an array reference with extent
		1505	structures. Each extent structure is an array reference itself, with the
		1506	following members:
		1507
		1508	[$logical, $physical, $length, $flags]
		1509
		1510	Flags is any combination of the following flag values (typically either C<0>
		1511	or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)):
		1512
		1513	C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>,
		1514	C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>,
		1515	C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>,
		1516	C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>,
		1517	C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or
		1518	C<IO::AIO::FIEMAP_EXTENT_SHARED>.
		1519
		1520	At the time of this writing (Linux 3.2), this request is unreliable unless
		1521	C<$count> is C<undef>, as the kernel has all sorts of bugs preventing
		1522	it to return all extents of a range for files with a large number of
		1523	extents. The code (only) works around all these issues if C<$count> is
		1524	C<undef>.
		1525
		1526	=item aio_group $callback->(...)
		1527
		1528	This is a very special aio request: Instead of doing something, it is a
		1529	container for other aio requests, which is useful if you want to bundle
		1530	many requests into a single, composite, request with a definite callback
		1531	and the ability to cancel the whole request with its subrequests.
		1532
		1533	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		1534	for more info.
		1535
		1536	Example:
		1537
		1538	my $grp = aio_group sub {
		1539	print "all stats done\n";
		1540	};
		1541
		1542	add $grp
		1543	(aio_stat ...),
		1544	(aio_stat ...),
		1545	...;
		1546
		1547	=item aio_nop $callback->()
		1548
		1549	This is a special request - it does nothing in itself and is only used for
		1550	side effects, such as when you want to add a dummy request to a group so
		1551	that finishing the requests in the group depends on executing the given
		1552	code.
		1553
		1554	While this request does nothing, it still goes through the execution
		1555	phase and still requires a worker thread. Thus, the callback will not
		1556	be executed immediately but only after other requests in the queue have
		1557	entered their execution phase. This can be used to measure request
		1558	latency.
		1559
		1560	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
		1561
		1562	Mainly used for debugging and benchmarking, this aio request puts one of
		1563	the request workers to sleep for the given time.
		1564
		1565	While it is theoretically handy to have simple I/O scheduling requests
		1566	like sleep and file handle readable/writable, the overhead this creates is
		1567	immense (it blocks a thread for a long time) so do not use this function
		1568	except to put your application under artificial I/O pressure.
		1569
189	=back	1570	=back
190		1571
		1572
		1573	=head2 IO::AIO::WD - multiple working directories
		1574
		1575	Your process only has one current working directory, which is used by all
		1576	threads. This makes it hard to use relative paths (some other component
		1577	could call C<chdir> at any time, and it is hard to control when the path
		1578	will be used by IO::AIO).
		1579
		1580	One solution for this is to always use absolute paths. This usually works,
		1581	but can be quite slow (the kernel has to walk the whole path on every
		1582	access), and can also be a hassle to implement.
		1583
		1584	Newer POSIX systems have a number of functions (openat, fdopendir,
		1585	futimensat and so on) that make it possible to specify working directories
		1586	per operation.
		1587
		1588	For portability, and because the clowns who "designed", or shall I write,
		1589	perpetrated this new interface were obviously half-drunk, this abstraction
		1590	cannot be perfect, though.
		1591
		1592	IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
		1593	object. This object stores the canonicalised, absolute version of the
		1594	path, and on systems that allow it, also a directory file descriptor.
		1595
		1596	Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
		1597	or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
		1598	object and a pathname instead (or the IO::AIO::WD object alone, which
		1599	gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
		1600	IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
		1601	to that IO::AIO::WD object.
		1602
		1603	For example, to get a wd object for F</etc> and then stat F<passwd>
		1604	inside, you would write:
		1605
		1606	aio_wd "/etc", sub {
		1607	my $etcdir = shift;
		1608
		1609	# although $etcdir can be undef on error, there is generally no reason
		1610	# to check for errors here, as aio_stat will fail with ENOENT
		1611	# when $etcdir is undef.
		1612
		1613	aio_stat [$etcdir, "passwd"], sub {
		1614	# yay
		1615	};
		1616	};
		1617
		1618	The fact that C<aio_wd> is a request and not a normal function shows that
		1619	creating an IO::AIO::WD object is itself a potentially blocking operation,
		1620	which is why it is done asynchronously.
		1621
		1622	To stat the directory obtained with C<aio_wd> above, one could write
		1623	either of the following three request calls:
		1624
		1625	aio_lstat "/etc" , sub { ... # pathname as normal string
		1626	aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
		1627	aio_lstat $wd , sub { ... # shorthand for the previous
		1628
		1629	As with normal pathnames, IO::AIO keeps a copy of the working directory
		1630	object and the pathname string, so you could write the following without
		1631	causing any issues due to C<$path> getting reused:
		1632
		1633	my $path = [$wd, undef];
		1634
		1635	for my $name (qw(abc def ghi)) {
		1636	$path->[1] = $name;
		1637	aio_stat $path, sub {
		1638	# ...
		1639	};
		1640	}
		1641
		1642	There are some caveats: when directories get renamed (or deleted), the
		1643	pathname string doesn't change, so will point to the new directory (or
		1644	nowhere at all), while the directory fd, if available on the system,
		1645	will still point to the original directory. Most functions accepting a
		1646	pathname will use the directory fd on newer systems, and the string on
		1647	older systems. Some functions (such as C<aio_realpath>) will always rely on
		1648	the string form of the pathname.
		1649
		1650	So this functionality is mainly useful to get some protection against
		1651	C<chdir>, to easily get an absolute path out of a relative path for future
		1652	reference, and to speed up doing many operations in the same directory
		1653	(e.g. when stat'ing all files in a directory).
		1654
		1655	The following functions implement this working directory abstraction:
		1656
		1657	=over 4
		1658
		1659	=item aio_wd $pathname, $callback->($wd)
		1660
		1661	Asynchonously canonicalise the given pathname and convert it to an
		1662	IO::AIO::WD object representing it. If possible and supported on the
		1663	system, also open a directory fd to speed up pathname resolution relative
		1664	to this working directory.
		1665
		1666	If something goes wrong, then C<undef> is passwd to the callback instead
		1667	of a working directory object and C<$!> is set appropriately. Since
		1668	passing C<undef> as working directory component of a pathname fails the
		1669	request with C<ENOENT>, there is often no need for error checking in the
		1670	C<aio_wd> callback, as future requests using the value will fail in the
		1671	expected way.
		1672
		1673	=item IO::AIO::CWD
		1674
		1675	This is a compiletime constant (object) that represents the process
		1676	current working directory.
		1677
		1678	Specifying this object as working directory object for a pathname is as if
		1679	the pathname would be specified directly, without a directory object. For
		1680	example, these calls are functionally identical:
		1681
		1682	aio_stat "somefile", sub { ... };
		1683	aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
		1684
		1685	=back
		1686
		1687	To recover the path associated with an IO::AIO::WD object, you can use
		1688	C<aio_realpath>:
		1689
		1690	aio_realpath $wd, sub {
		1691	warn "path is $_[0]\n";
		1692	};
		1693
		1694	Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir>
		1695	sometimes, fall back to using an absolue path.
		1696
		1697	=head2 IO::AIO::REQ CLASS
		1698
		1699	All non-aggregate C<aio_*> functions return an object of this class when
		1700	called in non-void context.
		1701
		1702	=over 4
		1703
		1704	=item cancel $req
		1705
		1706	Cancels the request, if possible. Has the effect of skipping execution
		1707	when entering the B<execute> state and skipping calling the callback when
		1708	entering the the B<result> state, but will leave the request otherwise
		1709	untouched (with the exception of readdir). That means that requests that
		1710	currently execute will not be stopped and resources held by the request
		1711	will not be freed prematurely.
		1712
		1713	=item cb $req $callback->(...)
		1714
		1715	Replace (or simply set) the callback registered to the request.
		1716
		1717	=back
		1718
		1719	=head2 IO::AIO::GRP CLASS
		1720
		1721	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		1722	objects of this class, too.
		1723
		1724	A IO::AIO::GRP object is a special request that can contain multiple other
		1725	aio requests.
		1726
		1727	You create one by calling the C<aio_group> constructing function with a
		1728	callback that will be called when all contained requests have entered the
		1729	C<done> state:
		1730
		1731	my $grp = aio_group sub {
		1732	print "all requests are done\n";
		1733	};
		1734
		1735	You add requests by calling the C<add> method with one or more
		1736	C<IO::AIO::REQ> objects:
		1737
		1738	$grp->add (aio_unlink "...");
		1739
		1740	add $grp aio_stat "...", sub {
		1741	$_[0] or return $grp->result ("error");
		1742
		1743	# add another request dynamically, if first succeeded
		1744	add $grp aio_open "...", sub {
		1745	$grp->result ("ok");
		1746	};
		1747	};
		1748
		1749	This makes it very easy to create composite requests (see the source of
		1750	C<aio_move> for an application) that work and feel like simple requests.
		1751
		1752	=over 4
		1753
		1754	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		1755	C<IO::AIO::poll_cb>, just like any other request.
		1756
		1757	=item * They can be canceled like any other request. Canceling will cancel not
		1758	only the request itself, but also all requests it contains.
		1759
		1760	=item * They can also can also be added to other IO::AIO::GRP objects.
		1761
		1762	=item * You must not add requests to a group from within the group callback (or
		1763	any later time).
		1764
		1765	=back
		1766
		1767	Their lifetime, simplified, looks like this: when they are empty, they
		1768	will finish very quickly. If they contain only requests that are in the
		1769	C<done> state, they will also finish. Otherwise they will continue to
		1770	exist.
		1771
		1772	That means after creating a group you have some time to add requests
		1773	(precisely before the callback has been invoked, which is only done within
		1774	the C<poll_cb>). And in the callbacks of those requests, you can add
		1775	further requests to the group. And only when all those requests have
		1776	finished will the the group itself finish.
		1777
		1778	=over 4
		1779
		1780	=item add $grp ...
		1781
		1782	=item $grp->add (...)
		1783
		1784	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		1785	be added, including other groups, as long as you do not create circular
		1786	dependencies.
		1787
		1788	Returns all its arguments.
		1789
		1790	=item $grp->cancel_subs
		1791
		1792	Cancel all subrequests and clears any feeder, but not the group request
		1793	itself. Useful when you queued a lot of events but got a result early.
		1794
		1795	The group request will finish normally (you cannot add requests to the
		1796	group).
		1797
		1798	=item $grp->result (...)
		1799
		1800	Set the result value(s) that will be passed to the group callback when all
		1801	subrequests have finished and set the groups errno to the current value
		1802	of errno (just like calling C<errno> without an error number). By default,
		1803	no argument will be passed and errno is zero.
		1804
		1805	=item $grp->errno ([$errno])
		1806
		1807	Sets the group errno value to C<$errno>, or the current value of errno
		1808	when the argument is missing.
		1809
		1810	Every aio request has an associated errno value that is restored when
		1811	the callback is invoked. This method lets you change this value from its
		1812	default (0).
		1813
		1814	Calling C<result> will also set errno, so make sure you either set C<$!>
		1815	before the call to C<result>, or call c<errno> after it.
		1816
		1817	=item feed $grp $callback->($grp)
		1818
		1819	Sets a feeder/generator on this group: every group can have an attached
		1820	generator that generates requests if idle. The idea behind this is that,
		1821	although you could just queue as many requests as you want in a group,
		1822	this might starve other requests for a potentially long time. For example,
		1823	C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
		1824	requests, delaying any later requests for a long time.
		1825
		1826	To avoid this, and allow incremental generation of requests, you can
		1827	instead a group and set a feeder on it that generates those requests. The
		1828	feed callback will be called whenever there are few enough (see C<limit>,
		1829	below) requests active in the group itself and is expected to queue more
		1830	requests.
		1831
		1832	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1833	not impose any limits).
		1834
		1835	If the feed does not queue more requests when called, it will be
		1836	automatically removed from the group.
		1837
		1838	If the feed limit is C<0> when this method is called, it will be set to
		1839	C<2> automatically.
		1840
		1841	Example:
		1842
		1843	# stat all files in @files, but only ever use four aio requests concurrently:
		1844
		1845	my $grp = aio_group sub { print "finished\n" };
		1846	limit $grp 4;
		1847	feed $grp sub {
		1848	my $file = pop @files
		1849	or return;
		1850
		1851	add $grp aio_stat $file, sub { ... };
		1852	};
		1853
		1854	=item limit $grp $num
		1855
		1856	Sets the feeder limit for the group: The feeder will be called whenever
		1857	the group contains less than this many requests.
		1858
		1859	Setting the limit to C<0> will pause the feeding process.
		1860
		1861	The default value for the limit is C<0>, but note that setting a feeder
		1862	automatically bumps it up to C<2>.
		1863
		1864	=back
		1865
191	=head2 SUPPORT FUNCTIONS	1866	=head2 SUPPORT FUNCTIONS
192		1867
		1868	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1869
193	=over 4	1870	=over 4
194		1871
195	=item $fileno = IO::AIO::poll_fileno	1872	=item $fileno = IO::AIO::poll_fileno
196		1873
197	Return the I<request result pipe filehandle>. This filehandle must be	1874	Return the I<request result pipe file descriptor>. This filehandle must be
198	polled for reading by some mechanism outside this module (e.g. Event	1875	polled for reading by some mechanism outside this module (e.g. EV, Glib,
199	or select, see below). If the pipe becomes readable you have to call	1876	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
200	C<poll_cb> to check the results.	1877	you have to call C<poll_cb> to check the results.
201		1878
202	See C<poll_cb> for an example.	1879	See C<poll_cb> for an example.
203		1880
204	=item IO::AIO::poll_cb	1881	=item IO::AIO::poll_cb
205		1882
206	Process all outstanding events on the result pipe. You have to call this	1883	Process some requests that have reached the result phase (i.e. they have
207	regularly. Returns the number of events processed. Returns immediately	1884	been executed but the results are not yet reported). You have to call
208	when no events are outstanding.	1885	this "regularly" to finish outstanding requests.
209		1886
210	You can use Event to multiplex, e.g.:	1887	Returns C<0> if all events could be processed (or there were no
		1888	events to process), or C<-1> if it returned earlier for whatever
		1889	reason. Returns immediately when no events are outstanding. The amount
		1890	of events processed depends on the settings of C<IO::AIO::max_poll_req>,
		1891	C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>.
		1892
		1893	If not all requests were processed for whatever reason, the poll file
		1894	descriptor will still be ready when C<poll_cb> returns, so normally you
		1895	don't have to do anything special to have it called later.
		1896
		1897	Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
		1898	ready, it can be beneficial to call this function from loops which submit
		1899	a lot of requests, to make sure the results get processed when they become
		1900	available and not just when the loop is finished and the event loop takes
		1901	over again. This function returns very fast when there are no outstanding
		1902	requests.
		1903
		1904	Example: Install an Event watcher that automatically calls
		1905	IO::AIO::poll_cb with high priority (more examples can be found in the
		1906	SYNOPSIS section, at the top of this document):
211		1907
212	Event->io (fd => IO::AIO::poll_fileno,	1908	Event->io (fd => IO::AIO::poll_fileno,
213	poll => 'r', async => 1,	1909	poll => 'r', async => 1,
214	cb => \&IO::AIO::poll_cb);	1910	cb => \&IO::AIO::poll_cb);
215		1911
216	=item IO::AIO::poll_wait	1912	=item IO::AIO::poll_wait
217		1913
218	Wait till the result filehandle becomes ready for reading (simply does a	1914	Wait until either at least one request is in the result phase or no
219	select on the filehandle. This is useful if you want to synchronously wait	1915	requests are outstanding anymore.
220	for some requests to finish).	1916
		1917	This is useful if you want to synchronously wait for some requests to
		1918	become ready, without actually handling them.
221		1919
222	See C<nreqs> for an example.	1920	See C<nreqs> for an example.
223		1921
		1922	=item IO::AIO::poll
		1923
		1924	Waits until some requests have been handled.
		1925
		1926	Returns the number of requests processed, but is otherwise strictly
		1927	equivalent to:
		1928
		1929	IO::AIO::poll_wait, IO::AIO::poll_cb
		1930
224	=item IO::AIO::nreqs	1931	=item IO::AIO::flush
225		1932
226	Returns the number of requests currently outstanding.	1933	Wait till all outstanding AIO requests have been handled.
227		1934
228	Example: wait till there are no outstanding requests anymore:	1935	Strictly equivalent to:
229		1936
230	IO::AIO::poll_wait, IO::AIO::poll_cb	1937	IO::AIO::poll_wait, IO::AIO::poll_cb
231	while IO::AIO::nreqs;	1938	while IO::AIO::nreqs;
232		1939
		1940	=item IO::AIO::max_poll_reqs $nreqs
		1941
		1942	=item IO::AIO::max_poll_time $seconds
		1943
		1944	These set the maximum number of requests (default C<0>, meaning infinity)
		1945	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1946	the maximum amount of time (default C<0>, meaning infinity) spent in
		1947	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1948	of time C<poll_cb> is allowed to use).
		1949
		1950	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1951	syscall per request processed, which is not normally a problem unless your
		1952	callbacks are really really fast or your OS is really really slow (I am
		1953	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1954
		1955	Setting these is useful if you want to ensure some level of
		1956	interactiveness when perl is not fast enough to process all requests in
		1957	time.
		1958
		1959	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		1960
		1961	Example: Install an Event watcher that automatically calls
		1962	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1963	program get the CPU sometimes even under high AIO load.
		1964
		1965	# try not to spend much more than 0.1s in poll_cb
		1966	IO::AIO::max_poll_time 0.1;
		1967
		1968	# use a low priority so other tasks have priority
		1969	Event->io (fd => IO::AIO::poll_fileno,
		1970	poll => 'r', nice => 1,
		1971	cb => &IO::AIO::poll_cb);
		1972
		1973	=back
		1974
		1975	=head3 CONTROLLING THE NUMBER OF THREADS
		1976
		1977	=over
		1978
233	=item IO::AIO::min_parallel $nthreads	1979	=item IO::AIO::min_parallel $nthreads
234		1980
235	Set the minimum number of AIO threads to C<$nthreads>. The default is	1981	Set the minimum number of AIO threads to C<$nthreads>. The current
236	C<1>, which means a single asynchronous operation can be done at one time	1982	default is C<8>, which means eight asynchronous operations can execute
237	(the number of outstanding operations, however, is unlimited).	1983	concurrently at any one time (the number of outstanding requests,
		1984	however, is unlimited).
238		1985
		1986	IO::AIO starts threads only on demand, when an AIO request is queued and
		1987	no free thread exists. Please note that queueing up a hundred requests can
		1988	create demand for a hundred threads, even if it turns out that everything
		1989	is in the cache and could have been processed faster by a single thread.
		1990
239	It is recommended to keep the number of threads low, as some Linux	1991	It is recommended to keep the number of threads relatively low, as some
240	kernel versions will scale negatively with the number of threads (higher	1992	Linux kernel versions will scale negatively with the number of threads
241	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	1993	(higher parallelity => MUCH higher latency). With current Linux 2.6
242	threads should be fine.	1994	versions, 4-32 threads should be fine.
243		1995
244	Under normal circumstances you don't need to call this function, as this	1996	Under most circumstances you don't need to call this function, as the
245	module automatically starts some threads (the exact number might change,	1997	module selects a default that is suitable for low to moderate load.
246	and is currently 4).
247		1998
248	=item IO::AIO::max_parallel $nthreads	1999	=item IO::AIO::max_parallel $nthreads
249		2000
250	Sets the maximum number of AIO threads to C<$nthreads>. If more than	2001	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
251	the specified number of threads are currently running, kill them. This	2002	specified number of threads are currently running, this function kills
252	function blocks until the limit is reached.	2003	them. This function blocks until the limit is reached.
		2004
		2005	While C<$nthreads> are zero, aio requests get queued but not executed
		2006	until the number of threads has been increased again.
253		2007
254	This module automatically runs C<max_parallel 0> at program end, to ensure	2008	This module automatically runs C<max_parallel 0> at program end, to ensure
255	that all threads are killed and that there are no outstanding requests.	2009	that all threads are killed and that there are no outstanding requests.
256		2010
257	Under normal circumstances you don't need to call this function.	2011	Under normal circumstances you don't need to call this function.
258		2012
		2013	=item IO::AIO::max_idle $nthreads
		2014
		2015	Limit the number of threads (default: 4) that are allowed to idle
		2016	(i.e., threads that did not get a request to process within the idle
		2017	timeout (default: 10 seconds). That means if a thread becomes idle while
		2018	C<$nthreads> other threads are also idle, it will free its resources and
		2019	exit.
		2020
		2021	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		2022	to allow for extremely high load situations, but want to free resources
		2023	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		2024
		2025	The default is probably ok in most situations, especially if thread
		2026	creation is fast. If thread creation is very slow on your system you might
		2027	want to use larger values.
		2028
		2029	=item IO::AIO::idle_timeout $seconds
		2030
		2031	Sets the minimum idle timeout (default 10) after which worker threads are
		2032	allowed to exit. SEe C<IO::AIO::max_idle>.
		2033
259	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	2034	=item IO::AIO::max_outstanding $maxreqs
260		2035
261	Sets the maximum number of outstanding requests to C<$nreqs>. If you	2036	Sets the maximum number of outstanding requests to C<$nreqs>. If
262	try to queue up more than this number of requests, the caller will block until	2037	you do queue up more than this number of requests, the next call to
263	some requests have been handled.	2038	C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
		2039	C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
		2040	longer exceeded.
264		2041
265	The default is very large, so normally there is no practical limit. If you	2042	In other words, this setting does not enforce a queue limit, but can be
266	queue up many requests in a loop it it often improves speed if you set	2043	used to make poll functions block if the limit is exceeded.
267	this to a relatively low number, such as C<100>.
268		2044
269	Under normal circumstances you don't need to call this function.	2045	This is a very bad function to use in interactive programs because it
		2046	blocks, and a bad way to reduce concurrency because it is inexact: Better
		2047	use an C<aio_group> together with a feed callback.
		2048
		2049	Its main use is in scripts without an event loop - when you want to stat
		2050	a lot of files, you can write something like this:
		2051
		2052	IO::AIO::max_outstanding 32;
		2053
		2054	for my $path (...) {
		2055	aio_stat $path , ...;
		2056	IO::AIO::poll_cb;
		2057	}
		2058
		2059	IO::AIO::flush;
		2060
		2061	The call to C<poll_cb> inside the loop will normally return instantly, but
		2062	as soon as more thna C<32> reqeusts are in-flight, it will block until
		2063	some requests have been handled. This keeps the loop from pushing a large
		2064	number of C<aio_stat> requests onto the queue.
		2065
		2066	The default value for C<max_outstanding> is very large, so there is no
		2067	practical limit on the number of outstanding requests.
270		2068
271	=back	2069	=back
272		2070
		2071	=head3 STATISTICAL INFORMATION
		2072
		2073	=over
		2074
		2075	=item IO::AIO::nreqs
		2076
		2077	Returns the number of requests currently in the ready, execute or pending
		2078	states (i.e. for which their callback has not been invoked yet).
		2079
		2080	Example: wait till there are no outstanding requests anymore:
		2081
		2082	IO::AIO::poll_wait, IO::AIO::poll_cb
		2083	while IO::AIO::nreqs;
		2084
		2085	=item IO::AIO::nready
		2086
		2087	Returns the number of requests currently in the ready state (not yet
		2088	executed).
		2089
		2090	=item IO::AIO::npending
		2091
		2092	Returns the number of requests currently in the pending state (executed,
		2093	but not yet processed by poll_cb).
		2094
		2095	=back
		2096
		2097	=head3 MISCELLANEOUS FUNCTIONS
		2098
		2099	IO::AIO implements some functions that are useful when you want to use
		2100	some "Advanced I/O" function not available to in Perl, without going the
		2101	"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*>
		2102	counterpart.
		2103
		2104	=over 4
		2105
		2106	=item $numfd = IO::AIO::get_fdlimit
		2107
		2108	This function is I<EXPERIMENTAL> and subject to change.
		2109
		2110	Tries to find the current file descriptor limit and returns it, or
		2111	C<undef> and sets C<$!> in case of an error. The limit is one larger than
		2112	the highest valid file descriptor number.
		2113
		2114	=item IO::AIO::min_fdlimit [$numfd]
		2115
		2116	This function is I<EXPERIMENTAL> and subject to change.
		2117
		2118	Try to increase the current file descriptor limit(s) to at least C<$numfd>
		2119	by changing the soft or hard file descriptor resource limit. If C<$numfd>
		2120	is missing, it will try to set a very high limit, although this is not
		2121	recommended when you know the actual minimum that you require.
		2122
		2123	If the limit cannot be raised enough, the function makes a best-effort
		2124	attempt to increase the limit as much as possible, using various
		2125	tricks, while still failing. You can query the resulting limit using
		2126	C<IO::AIO::get_fdlimit>.
		2127
		2128	If an error occurs, returns C<undef> and sets C<$!>, otherwise returns
		2129	true.
		2130
		2131	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		2132
		2133	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		2134	but is blocking (this makes most sense if you know the input data is
		2135	likely cached already and the output filehandle is set to non-blocking
		2136	operations).
		2137
		2138	Returns the number of bytes copied, or C<-1> on error.
		2139
		2140	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		2141
		2142	Simply calls the C<posix_fadvise> function (see its
		2143	manpage for details). The following advice constants are
		2144	available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		2145	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		2146	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		2147
		2148	On systems that do not implement C<posix_fadvise>, this function returns
		2149	ENOSYS, otherwise the return value of C<posix_fadvise>.
		2150
		2151	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		2152
		2153	Simply calls the C<posix_madvise> function (see its
		2154	manpage for details). The following advice constants are
		2155	available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		2156	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>,
		2157	C<IO::AIO::MADV_DONTNEED>.
		2158
		2159	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2160	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2161	will be reduced to fit into the C<$scalar>.
		2162
		2163	On systems that do not implement C<posix_madvise>, this function returns
		2164	ENOSYS, otherwise the return value of C<posix_madvise>.
		2165
		2166	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		2167
		2168	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		2169	$scalar (see its manpage for details). The following protect
		2170	constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		2171	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		2172
		2173	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2174	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2175	will be reduced to fit into the C<$scalar>.
		2176
		2177	On systems that do not implement C<mprotect>, this function returns
		2178	ENOSYS, otherwise the return value of C<mprotect>.
		2179
		2180	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		2181
		2182	Memory-maps a file (or anonymous memory range) and attaches it to the
		2183	given C<$scalar>, which will act like a string scalar. Returns true on
		2184	success, and false otherwise.
		2185
		2186	The scalar must exist, but its contents do not matter - this means you
		2187	cannot use a nonexistant array or hash element. When in doubt, C<undef>
		2188	the scalar first.
		2189
		2190	The only operations allowed on the mmapped scalar are C<substr>/C<vec>,
		2191	which don't change the string length, and most read-only operations such
		2192	as copying it or searching it with regexes and so on.
		2193
		2194	Anything else is unsafe and will, at best, result in memory leaks.
		2195
		2196	The memory map associated with the C<$scalar> is automatically removed
		2197	when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap>
		2198	or C<IO::AIO::munmap> functions are called on it.
		2199
		2200	This calls the C<mmap>(2) function internally. See your system's manual
		2201	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		2202
		2203	The C<$length> must be larger than zero and smaller than the actual
		2204	filesize.
		2205
		2206	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		2207	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		2208
		2209	C<$flags> can be a combination of
		2210	C<IO::AIO::MAP_SHARED> or
		2211	C<IO::AIO::MAP_PRIVATE>,
		2212	or a number of system-specific flags (when not available, the are C<0>):
		2213	C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant),
		2214	C<IO::AIO::MAP_LOCKED>,
		2215	C<IO::AIO::MAP_NORESERVE>,
		2216	C<IO::AIO::MAP_POPULATE>,
		2217	C<IO::AIO::MAP_NONBLOCK>,
		2218	C<IO::AIO::MAP_FIXED>,
		2219	C<IO::AIO::MAP_GROWSDOWN>,
		2220	C<IO::AIO::MAP_32BIT>,
		2221	C<IO::AIO::MAP_HUGETLB> or
		2222	C<IO::AIO::MAP_STACK>.
		2223
		2224	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		2225
		2226	C<$offset> is the offset from the start of the file - it generally must be
		2227	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		2228
		2229	Example:
		2230
		2231	use Digest::MD5;
		2232	use IO::AIO;
		2233
		2234	open my $fh, "<verybigfile"
		2235	or die "$!";
		2236
		2237	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		2238	or die "verybigfile: $!";
		2239
		2240	my $fast_md5 = md5 $data;
		2241
		2242	=item IO::AIO::munmap $scalar
		2243
		2244	Removes a previous mmap and undefines the C<$scalar>.
		2245
		2246	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		2247
		2248	Calls the C<munlock> function, undoing the effects of a previous
		2249	C<aio_mlock> call (see its description for details).
		2250
		2251	=item IO::AIO::munlockall
		2252
		2253	Calls the C<munlockall> function.
		2254
		2255	On systems that do not implement C<munlockall>, this function returns
		2256	ENOSYS, otherwise the return value of C<munlockall>.
		2257
		2258	=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
		2259
		2260	Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or
		2261	C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they
		2262	should be the file offset.
		2263
		2264	C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might
		2265	silently corrupt the data in this case.
		2266
		2267	The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>,
		2268	C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and
		2269	C<IO::AIO::SPLICE_F_GIFT>.
		2270
		2271	See the C<splice(2)> manpage for details.
		2272
		2273	=item IO::AIO::tee $r_fh, $w_fh, $length, $flags
		2274
		2275	Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the
		2276	description for C<IO::AIO::splice> above for details.
		2277
		2278	=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
		2279
		2280	Attempts to query or change the pipe buffer size. Obviously works only
		2281	on pipes, and currently works only on GNU/Linux systems, and fails with
		2282	C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer
		2283	size on other systems, drop me a note.
		2284
		2285	=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
		2286
		2287	This is a direct interface to the Linux L<pipe2(2)> system call. If
		2288	C<$flags> is missing or C<0>, then this should be the same as a call to
		2289	perl's built-in C<pipe> function and create a new pipe, and works on
		2290	systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe
		2291	(..., 4096, O_BINARY)>.
		2292
		2293	If C<$flags> is non-zero, it tries to invoke the pipe2 system call with
		2294	the given flags (Linux 2.6.27, glibc 2.9).
		2295
		2296	On success, the read and write file handles are returned.
		2297
		2298	On error, nothing will be returned. If the pipe2 syscall is missing and
		2299	C<$flags> is non-zero, fails with C<ENOSYS>.
		2300
		2301	Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the
		2302	time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and
		2303	C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported.
		2304
		2305	=back
		2306
273	=cut	2307	=cut
274		2308
275	# support function to convert a fd into a perl filehandle
276	sub _fd2fh {
277	return undef if $_[0] < 0;
278
279	# try to be perl5.6-compatible
280	local *AIO_FH;
281	open AIO_FH, "+<&=$_[0]"
282	or return undef;
283
284	*AIO_FH
285	}
286
287	min_parallel 4;	2309	min_parallel 8;
288		2310
289	END {	2311	END { flush }
290	max_parallel 0;
291	}
292		2312
293	1;	2313	1;
294		2314
		2315	=head1 EVENT LOOP INTEGRATION
		2316
		2317	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		2318	automatically into many event loops:
		2319
		2320	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		2321	use AnyEvent::AIO;
		2322
		2323	You can also integrate IO::AIO manually into many event loops, here are
		2324	some examples of how to do this:
		2325
		2326	# EV integration
		2327	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		2328
		2329	# Event integration
		2330	Event->io (fd => IO::AIO::poll_fileno,
		2331	poll => 'r',
		2332	cb => \&IO::AIO::poll_cb);
		2333
		2334	# Glib/Gtk2 integration
		2335	add_watch Glib::IO IO::AIO::poll_fileno,
		2336	in => sub { IO::AIO::poll_cb; 1 };
		2337
		2338	# Tk integration
		2339	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		2340	readable => \&IO::AIO::poll_cb);
		2341
		2342	# Danga::Socket integration
		2343	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		2344	\&IO::AIO::poll_cb);
		2345
		2346	=head2 FORK BEHAVIOUR
		2347
		2348	Usage of pthreads in a program changes the semantics of fork
		2349	considerably. Specifically, only async-safe functions can be called after
		2350	fork. Perl doesn't know about this, so in general, you cannot call fork
		2351	with defined behaviour in perl if pthreads are involved. IO::AIO uses
		2352	pthreads, so this applies, but many other extensions and (for inexplicable
		2353	reasons) perl itself often is linked against pthreads, so this limitation
		2354	applies to quite a lot of perls.
		2355
		2356	This module no longer tries to fight your OS, or POSIX. That means IO::AIO
		2357	only works in the process that loaded it. Forking is fully supported, but
		2358	using IO::AIO in the child is not.
		2359
		2360	You might get around by not I<using> IO::AIO before (or after)
		2361	forking. You could also try to call the L<IO::AIO::reinit> function in the
		2362	child:
		2363
		2364	=over 4
		2365
		2366	=item IO::AIO::reinit
		2367
		2368	Abandons all current requests and I/O threads and simply reinitialises all
		2369	data structures. This is not an operation supported by any standards, but
		2370	happens to work on GNU/Linux and some newer BSD systems.
		2371
		2372	The only reasonable use for this function is to call it after forking, if
		2373	C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
		2374	the process will result in undefined behaviour. Calling it at any time
		2375	will also result in any undefined (by POSIX) behaviour.
		2376
		2377	=back
		2378
		2379	=head2 MEMORY USAGE
		2380
		2381	Per-request usage:
		2382
		2383	Each aio request uses - depending on your architecture - around 100-200
		2384	bytes of memory. In addition, stat requests need a stat buffer (possibly
		2385	a few hundred bytes), readdir requires a result buffer and so on. Perl
		2386	scalars and other data passed into aio requests will also be locked and
		2387	will consume memory till the request has entered the done state.
		2388
		2389	This is not awfully much, so queuing lots of requests is not usually a
		2390	problem.
		2391
		2392	Per-thread usage:
		2393
		2394	In the execution phase, some aio requests require more memory for
		2395	temporary buffers, and each thread requires a stack and other data
		2396	structures (usually around 16k-128k, depending on the OS).
		2397
		2398	=head1 KNOWN BUGS
		2399
		2400	Known bugs will be fixed in the next release.
		2401
295	=head1 SEE ALSO	2402	=head1 SEE ALSO
296		2403
297	L<Coro>, L<Linux::AIO>.	2404	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		2405	more natural syntax.
298		2406
299	=head1 AUTHOR	2407	=head1 AUTHOR
300		2408
301	Marc Lehmann <schmorp@schmorp.de>	2409	Marc Lehmann <schmorp@schmorp.de>
302	http://home.schmorp.de/	2410	http://home.schmorp.de/

Diff Legend

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