ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.57 by root, Sun Oct 22 01:28:31 2006 UTC vs.
Revision 1.319 by root, Wed Sep 27 03:09:57 2023 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
…		…
18	};	19	};
19		20
20	# version 2+ has request and group objects	21	# version 2+ has request and group objects
21	use IO::AIO 2;	22	use IO::AIO 2;
22		23
		24	aioreq_pri 4; # give next request a very high priority
23	my $req = aio_unlink "/tmp/file", sub { };	25	my $req = aio_unlink "/tmp/file", sub { };
24	$req->cancel; # cancel request if still in queue	26	$req->cancel; # cancel request if still in queue
25		27
26	my $grp = aio_group sub { print "all stats done\n" };	28	my $grp = aio_group sub { print "all stats done\n" };
27	add $grp aio_stat "..." for ...;	29	add $grp aio_stat "..." for ...;
28		30
29	# AnyEvent integration
30	open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
31	my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
32
33	# Event integration
34	Event->io (fd => IO::AIO::poll_fileno,
35	poll => 'r',
36	cb => \&IO::AIO::poll_cb);
37
38	# Glib/Gtk2 integration
39	add_watch Glib::IO IO::AIO::poll_fileno,
40	in => sub { IO::AIO::poll_cb; 1 };
41
42	# Tk integration
43	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
44	readable => \&IO::AIO::poll_cb);
45
46	# Danga::Socket integration
47	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
48	\&IO::AIO::poll_cb);
49
50	=head1 DESCRIPTION	31	=head1 DESCRIPTION
51		32
52	This module implements asynchronous I/O using whatever means your	33	This module implements asynchronous I/O using whatever means your
53	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>).
54		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
55	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
56	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
57	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
58	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
59	aio functions available on many operating systems. However, they are often	57	functions available on many operating systems. However, they are often
60	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
61	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
62	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.
63		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
64	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,
65	currently not reentrant, so use appropriate locking yourself, always call	68	it is currently not reentrant in any way, so use appropriate locking
66	C<poll_cb> from within the same thread, or never call C<poll_cb> (or other	69	yourself, always call C<poll_cb> from within the same thread, or never
67	C<aio_> functions) recursively.	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
68		164
69	=cut	165	=cut
70		166
71	package IO::AIO;	167	package IO::AIO;
72		168
73	no warnings;	169	use Carp ();
74	use strict 'vars';	170
		171	use common::sense;
75		172
76	use base 'Exporter';	173	use base 'Exporter';
77		174
78	BEGIN {	175	BEGIN {
79	our $VERSION = '2.0';	176	our $VERSION = 4.80;
80		177
81	our @EXPORT = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat	178	our @AIO_REQ = qw(aio_sendfile aio_seek aio_read aio_write aio_open aio_close
82	aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink	179	aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
83	aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move	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
84	aio_group);	189	aio_wd);
85	our @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	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 mremap munlock munlockall
		198
		199	accept4 tee splice pipe2 pipesize
		200	fexecve mount umount memfd_create eventfd
		201	timerfd_create timerfd_settime timerfd_gettime
		202	pidfd_open pidfd_send_signal pidfd_getfd);
		203
		204	push @AIO_REQ, qw(aio_busy); # not exported
86		205
87	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';	206	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
88		207
89	require XSLoader;	208	require XSLoader;
90	XSLoader::load ("IO::AIO", $VERSION);	209	XSLoader::load ("IO::AIO", $VERSION);
91	}	210	}
92		211
93	=head1 FUNCTIONS	212	=head1 FUNCTIONS
94		213
95	=head2 AIO FUNCTIONS	214	=head2 QUICK OVERVIEW
		215
		216	This section simply lists the prototypes most of the functions for
		217	quick reference. See the following sections for function-by-function
		218	documentation.
		219
		220	aio_wd $pathname, $callback->($wd)
		221	aio_open $pathname, $flags, $mode, $callback->($fh)
		222	aio_close $fh, $callback->($status)
		223	aio_seek $fh,$offset,$whence, $callback->($offs)
		224	aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		225	aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		226	aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		227	aio_readahead $fh,$offset,$length, $callback->($retval)
		228	aio_stat $fh_or_path, $callback->($status)
		229	aio_lstat $fh, $callback->($status)
		230	aio_statvfs $fh_or_path, $callback->($statvfs)
		231	aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		232	aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		233	aio_chmod $fh_or_path, $mode, $callback->($status)
		234	aio_truncate $fh_or_path, $offset, $callback->($status)
		235	aio_allocate $fh, $mode, $offset, $len, $callback->($status)
		236	aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
		237	aio_unlink $pathname, $callback->($status)
		238	aio_mknod $pathname, $mode, $dev, $callback->($status)
		239	aio_link $srcpath, $dstpath, $callback->($status)
		240	aio_symlink $srcpath, $dstpath, $callback->($status)
		241	aio_readlink $pathname, $callback->($link)
		242	aio_realpath $pathname, $callback->($path)
		243	aio_rename $srcpath, $dstpath, $callback->($status)
		244	aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
		245	aio_mkdir $pathname, $mode, $callback->($status)
		246	aio_rmdir $pathname, $callback->($status)
		247	aio_readdir $pathname, $callback->($entries)
		248	aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		249	IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
		250	IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
		251	aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		252	aio_load $pathname, $data, $callback->($status)
		253	aio_copy $srcpath, $dstpath, $callback->($status)
		254	aio_move $srcpath, $dstpath, $callback->($status)
		255	aio_rmtree $pathname, $callback->($status)
		256	aio_fcntl $fh, $cmd, $arg, $callback->($status)
		257	aio_ioctl $fh, $request, $buf, $callback->($status)
		258	aio_sync $callback->($status)
		259	aio_syncfs $fh, $callback->($status)
		260	aio_fsync $fh, $callback->($status)
		261	aio_fdatasync $fh, $callback->($status)
		262	aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		263	aio_pathsync $pathname, $callback->($status)
		264	aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
		265	aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		266	aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		267	aio_mlockall $flags, $callback->($status)
		268	aio_group $callback->(...)
		269	aio_nop $callback->()
		270
		271	$prev_pri = aioreq_pri [$pri]
		272	aioreq_nice $pri_adjust
		273
		274	IO::AIO::poll_wait
		275	IO::AIO::poll_cb
		276	IO::AIO::poll
		277	IO::AIO::flush
		278	IO::AIO::max_poll_reqs $nreqs
		279	IO::AIO::max_poll_time $seconds
		280	IO::AIO::min_parallel $nthreads
		281	IO::AIO::max_parallel $nthreads
		282	IO::AIO::max_idle $nthreads
		283	IO::AIO::idle_timeout $seconds
		284	IO::AIO::max_outstanding $maxreqs
		285	IO::AIO::nreqs
		286	IO::AIO::nready
		287	IO::AIO::npending
		288	IO::AIO::reinit
		289
		290	$nfd = IO::AIO::get_fdlimit
		291	IO::AIO::min_fdlimit $nfd
		292
		293	IO::AIO::sendfile $ofh, $ifh, $offset, $count
		294	IO::AIO::fadvise $fh, $offset, $len, $advice
		295	IO::AIO::fexecve $fh, $argv, $envp
		296
		297	IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
		298	IO::AIO::munmap $scalar
		299	IO::AIO::mremap $scalar, $new_length, $flags[, $new_address]
		300	IO::AIO::madvise $scalar, $offset, $length, $advice
		301	IO::AIO::mprotect $scalar, $offset, $length, $protect
		302	IO::AIO::munlock $scalar, $offset = 0, $length = undef
		303	IO::AIO::munlockall
		304
		305	# stat extensions
		306	$counter = IO::AIO::st_gen
		307	$seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime
		308	($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
		309	$nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec
		310	$seconds = IO::AIO::st_btimesec
		311	($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
		312
		313	# very much unportable syscalls
		314	IO::AIO::accept4 $r_fh, $sockaddr, $sockaddr_len, $flags
		315	IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
		316	IO::AIO::tee $r_fh, $w_fh, $length, $flags
		317
		318	$actual_size = IO::AIO::pipesize $r_fh[, $new_size]
		319	($rfh, $wfh) = IO::AIO::pipe2 [$flags]
		320
		321	$fh = IO::AIO::eventfd [$initval, [$flags]]
		322	$fh = IO::AIO::memfd_create $pathname[, $flags]
		323
		324	$fh = IO::AIO::timerfd_create $clockid[, $flags]
		325	($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
		326	($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
		327
		328	$fh = IO::AIO::pidfd_open $pid[, $flags]
		329	$status = IO::AIO::pidfd_send_signal $pidfh, $signal[, $siginfo[, $flags]]
		330	$fh = IO::AIO::pidfd_getfd $pidfh, $targetfd[, $flags]
		331
		332	$retval = IO::AIO::mount $special, $path, $fstype, $flags = 0, $data = undef
		333	$retval = IO::AIO::umount $path, $flags = 0
		334
		335	=head2 API NOTES
96		336
97	All the C<aio_*> calls are more or less thin wrappers around the syscall	337	All the C<aio_*> calls are more or less thin wrappers around the syscall
98	with the same name (sans C<aio_>). The arguments are similar or identical,	338	with the same name (sans C<aio_>). The arguments are similar or identical,
99	and they all accept an additional (and optional) C<$callback> argument	339	and they all accept an additional (and optional) C<$callback> argument
100	which must be a code reference. This code reference will get called with	340	which must be a code reference. This code reference will be called after
101	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	341	the syscall has been executed in an asynchronous fashion. The results
102	perl, which usually delivers "false") as it's sole argument when the given	342	of the request will be passed as arguments to the callback (and, if an
103	syscall has been executed asynchronously.	343	error occured, in C<$!>) - for most requests the syscall return code (e.g.
		344	most syscalls return C<-1> on error, unlike perl, which usually delivers
		345	"false").
		346
		347	Some requests (such as C<aio_readdir>) pass the actual results and
		348	communicate failures by passing C<undef>.
104		349
105	All functions expecting a filehandle keep a copy of the filehandle	350	All functions expecting a filehandle keep a copy of the filehandle
106	internally until the request has finished.	351	internally until the request has finished.
107		352
108	All requests return objects of type L<IO::AIO::REQ> that allow further	353	All functions return request objects of type L<IO::AIO::REQ> that allow
109	manipulation of those requests while they are in-flight.	354	further manipulation of those requests while they are in-flight.
110		355
111	The pathnames you pass to these routines I<must> be absolute and	356	The pathnames you pass to these routines I<should> be absolute. The
112	encoded in byte form. The reason for the former is that at the time the	357	reason for this is that at the time the request is being executed, the
113	request is being executed, the current working directory could have	358	current working directory could have changed. Alternatively, you can
114	changed. Alternatively, you can make sure that you never change the	359	make sure that you never change the current working directory anywhere
115	current working directory.	360	in the program and then use relative paths. You can also take advantage
		361	of IO::AIOs working directory abstraction, that lets you specify paths
		362	relative to some previously-opened "working directory object" - see the
		363	description of the C<IO::AIO::WD> class later in this document.
116		364
117	To encode pathnames to byte form, either make sure you either: a)	365	To encode pathnames as octets, either make sure you either: a) always pass
118	always pass in filenames you got from outside (command line, readdir	366	in filenames you got from outside (command line, readdir etc.) without
119	etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode	367	tinkering, b) are in your native filesystem encoding, c) use the Encode
120	your pathnames to the locale (or other) encoding in effect in the user	368	module and encode your pathnames to the locale (or other) encoding in
121	environment, d) use Glib::filename_from_unicode on unicode filenames or e)	369	effect in the user environment, d) use Glib::filename_from_unicode on
122	use something else.	370	unicode filenames or e) use something else to ensure your scalar has the
		371	correct contents.
		372
		373	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		374	handles correctly whether it is set or not.
		375
		376	=head2 AIO REQUEST FUNCTIONS
123		377
124	=over 4	378	=over 4
125		379
		380	=item $prev_pri = aioreq_pri [$pri]
		381
		382	Returns the priority value that would be used for the next request and, if
		383	C<$pri> is given, sets the priority for the next aio request.
		384
		385	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		386	and C<4>, respectively. Requests with higher priority will be serviced
		387	first.
		388
		389	The priority will be reset to C<0> after each call to one of the C<aio_*>
		390	functions.
		391
		392	Example: open a file with low priority, then read something from it with
		393	higher priority so the read request is serviced before other low priority
		394	open requests (potentially spamming the cache):
		395
		396	aioreq_pri -3;
		397	aio_open ..., sub {
		398	return unless $_[0];
		399
		400	aioreq_pri -2;
		401	aio_read $_[0], ..., sub {
		402	...
		403	};
		404	};
		405
		406
		407	=item aioreq_nice $pri_adjust
		408
		409	Similar to C<aioreq_pri>, but subtracts the given value from the current
		410	priority, so the effect is cumulative.
		411
		412
126	=item aio_open $pathname, $flags, $mode, $callback->($fh)	413	=item aio_open $pathname, $flags, $mode, $callback->($fh)
127		414
128	Asynchronously open or create a file and call the callback with a newly	415	Asynchronously open or create a file and call the callback with a newly
129	created filehandle for the file.	416	created filehandle for the file (or C<undef> in case of an error).
130
131	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
132	for an explanation.
133		417
134	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a	418	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
135	list. They are the same as used by C<sysopen>.	419	list. They are the same as used by C<sysopen>.
136		420
137	Likewise, C<$mode> specifies the mode of the newly created file, if it	421	Likewise, C<$mode> specifies the mode of the newly created file, if it
138	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,	422	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
139	except that it is mandatory (i.e. use C<0> if you don't create new files,	423	except that it is mandatory (i.e. use C<0> if you don't create new files,
140	and C<0666> or C<0777> if you do).	424	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		425	by the umask in effect then the request is being executed, so better never
		426	change the umask.
141		427
142	Example:	428	Example:
143		429
144	aio_open "/etc/passwd", O_RDONLY, 0, sub {	430	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
145	if ($_[0]) {	431	if ($_[0]) {
146	print "open successful, fh is $_[0]\n";	432	print "open successful, fh is $_[0]\n";
147	...	433	...
148	} else {	434	} else {
149	die "open failed: $!\n";	435	die "open failed: $!\n";
150	}	436	}
151	};	437	};
152		438
		439	In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
		440	C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
		441	following POSIX and non-POSIX constants are available (missing ones on
		442	your system are, as usual, C<0>):
		443
		444	C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
		445	C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
		446	C<O_RSYNC>, C<O_SYNC>, C<O_PATH>, C<O_TMPFILE>, C<O_TTY_INIT> and C<O_ACCMODE>.
		447
		448
153	=item aio_close $fh, $callback->($status)	449	=item aio_close $fh, $callback->($status)
154		450
155	Asynchronously close a file and call the callback with the result	451	Asynchronously close a file and call the callback with the result
156	code. I<WARNING:> although accepted, you should not pass in a perl	452	code.
157	filehandle here, as perl will likely close the file descriptor another
158	time when the filehandle is destroyed. Normally, you can safely call perls
159	C<close> or just let filehandles go out of scope.
160		453
161	This is supposed to be a bug in the API, so that might change. It's	454	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
162	therefore best to avoid this function.	455	closing the file descriptor associated with the filehandle itself.
		456
		457	Therefore, C<aio_close> will not close the filehandle - instead it will
		458	use dup2 to overwrite the file descriptor with the write-end of a pipe
		459	(the pipe fd will be created on demand and will be cached).
		460
		461	Or in other words: the file descriptor will be closed, but it will not be
		462	free for reuse until the perl filehandle is closed.
		463
		464	=cut
		465
		466	=item aio_seek $fh, $offset, $whence, $callback->($offs)
		467
		468	Seeks the filehandle to the new C<$offset>, similarly to perl's
		469	C<sysseek>. The C<$whence> can use the traditional values (C<0> for
		470	C<IO::AIO::SEEK_SET>, C<1> for C<IO::AIO::SEEK_CUR> or C<2> for
		471	C<IO::AIO::SEEK_END>).
		472
		473	The resulting absolute offset will be passed to the callback, or C<-1> in
		474	case of an error.
		475
		476	In theory, the C<$whence> constants could be different than the
		477	corresponding values from L<Fcntl>, but perl guarantees they are the same,
		478	so don't panic.
		479
		480	As a GNU/Linux (and maybe Solaris) extension, also the constants
		481	C<IO::AIO::SEEK_DATA> and C<IO::AIO::SEEK_HOLE> are available, if they
		482	could be found. No guarantees about suitability for use in C<aio_seek> or
		483	Perl's C<sysseek> can be made though, although I would naively assume they
		484	"just work".
163		485
164	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	486	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
165		487
166	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	488	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
167		489
168	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	490	Reads or writes C<$length> bytes from or to the specified C<$fh> and
169	into the scalar given by C<data> and offset C<dataoffset> and calls the	491	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> and
170	callback without the actual number of bytes read (or -1 on error, just	492	calls the callback with the actual number of bytes transferred (or -1 on
171	like the syscall).	493	error, just like the syscall).
		494
		495	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		496	offset plus the actual number of bytes read.
		497
		498	If C<$offset> is undefined, then the current file descriptor offset will
		499	be used (and updated), otherwise the file descriptor offset will not be
		500	changed by these calls.
		501
		502	If C<$length> is undefined in C<aio_write>, use the remaining length of
		503	C<$data>.
		504
		505	If C<$dataoffset> is less than zero, it will be counted from the end of
		506	C<$data>.
172		507
173	The C<$data> scalar I<MUST NOT> be modified in any way while the request	508	The C<$data> scalar I<MUST NOT> be modified in any way while the request
174	is outstanding. Modifying it can result in segfaults or WW3 (if the	509	is outstanding. Modifying it can result in segfaults or World War III (if
175	necessary/optional hardware is installed).	510	the necessary/optional hardware is installed).
176		511
177	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	512	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
178	offset C<0> within the scalar:	513	offset C<0> within the scalar:
179		514
180	aio_read $fh, 7, 15, $buffer, 0, sub {	515	aio_read $fh, 7, 15, $buffer, 0, sub {
181	$_[0] > 0 or die "read error: $!";	516	$_[0] > 0 or die "read error: $!";
182	print "read $_[0] bytes: <$buffer>\n";	517	print "read $_[0] bytes: <$buffer>\n";
183	};	518	};
184		519
185	=item aio_move $srcpath, $dstpath, $callback->($status)
186
187	Try to move the I<file> (directories not supported as either source or
188	destination) from C<$srcpath> to C<$dstpath> and call the callback with
189	the C<0> (error) or C<-1> ok.
190
191	This is a composite request that tries to rename(2) the file first. If
192	rename files with C<EXDEV>, it creates the destination file with mode 0200
193	and copies the contents of the source file into it using C<aio_sendfile>,
194	followed by restoring atime, mtime, access mode and uid/gid, in that
195	order, and unlinking the C<$srcpath>.
196
197	If an error occurs, the partial destination file will be unlinked, if
198	possible, except when setting atime, mtime, access mode and uid/gid, where
199	errors are being ignored.
200
201	=cut
202
203	sub aio_move($$$) {
204	my ($src, $dst, $cb) = @_;
205
206	my $grp = aio_group;
207
208	add $grp aio_rename $src, $dst, sub {
209	if ($_[0] && $! == EXDEV) {
210	add $grp aio_open $src, O_RDONLY, 0, sub {
211	if (my $src_fh = $_[0]) {
212	my @stat = stat $src_fh;
213
214	add $grp aio_open $dst, O_WRONLY, 0200, sub {
215	if (my $dst_fh = $_[0]) {
216	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
217	close $src_fh;
218
219	if ($_[0] == $stat[7]) {
220	utime $stat[8], $stat[9], $dst;
221	chmod $stat[2] & 07777, $dst_fh;
222	chown $stat[4], $stat[5], $dst_fh;
223	close $dst_fh;
224
225	add $grp aio_unlink $src, sub {
226	$cb->($_[0]);
227	};
228	} else {
229	my $errno = $!;
230	add $grp aio_unlink $dst, sub {
231	$! = $errno;
232	$cb->(-1);
233	};
234	}
235	};
236	} else {
237	$cb->(-1);
238	}
239	},
240
241	} else {
242	$cb->(-1);
243	}
244	};
245	} else {
246	$cb->($_[0]);
247	}
248	};
249
250	$grp
251	}
252		520
253	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)	521	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
254		522
255	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts	523	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
256	reading at byte offset C<$in_offset>, and starts writing at the current	524	reading at byte offset C<$in_offset>, and starts writing at the current
257	file offset of C<$out_fh>. Because of that, it is not safe to issue more	525	file offset of C<$out_fh>. Because of that, it is not safe to issue more
258	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each	526	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
259	other.	527	other. The same C<$in_fh> works fine though, as this function does not
		528	move or use the file offset of C<$in_fh>.
260		529
		530	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		531	are written, and there is no way to find out how many more bytes have been
		532	read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
		533	number of bytes written to C<$out_fh>. Only if the result value equals
		534	C<$length> one can assume that C<$length> bytes have been read.
		535
		536	Unlike with other C<aio_> functions, it makes a lot of sense to use
		537	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		538	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		539	the socket I/O will be non-blocking. Note, however, that you can run
		540	into a trap where C<aio_sendfile> reads some data with readahead, then
		541	fails to write all data, and when the socket is ready the next time, the
		542	data in the cache is already lost, forcing C<aio_sendfile> to again hit
		543	the disk. Explicit C<aio_read> + C<aio_write> let's you better control
		544	resource usage.
		545
261	This call tries to make use of a native C<sendfile> syscall to provide	546	This call tries to make use of a native C<sendfile>-like syscall to
262	zero-copy operation. For this to work, C<$out_fh> should refer to a	547	provide zero-copy operation. For this to work, C<$out_fh> should refer to
263	socket, and C<$in_fh> should refer to mmap'able file.	548	a socket, and C<$in_fh> should refer to an mmap'able file.
264		549
265	If the native sendfile call fails or is not implemented, it will be	550	If a native sendfile cannot be found or it fails with C<ENOSYS>,
266	emulated, so you can call C<aio_sendfile> on any type of filehandle	551	C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
		552	C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
267	regardless of the limitations of the operating system.	553	type of filehandle regardless of the limitations of the operating system.
268		554
269	Please note, however, that C<aio_sendfile> can read more bytes from	555	As native sendfile syscalls (as practically any non-POSIX interface hacked
270	C<$in_fh> than are written, and there is no way to find out how many	556	together in a hurry to improve benchmark numbers) tend to be rather buggy
271	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only	557	on many systems, this implementation tries to work around some known bugs
272	provides the number of bytes written to C<$out_fh>. Only if the result	558	in Linux and FreeBSD kernels (probably others, too), but that might fail,
273	value equals C<$length> one can assume that C<$length> bytes have been	559	so you really really should check the return value of C<aio_sendfile> -
274	read.	560	fewer bytes than expected might have been transferred.
		561
275		562
276	=item aio_readahead $fh,$offset,$length, $callback->($retval)	563	=item aio_readahead $fh,$offset,$length, $callback->($retval)
277		564
278	C<aio_readahead> populates the page cache with data from a file so that	565	C<aio_readahead> populates the page cache with data from a file so that
279	subsequent reads from that file will not block on disk I/O. The C<$offset>	566	subsequent reads from that file will not block on disk I/O. The C<$offset>
…		…
282	whole pages, so that offset is effectively rounded down to a page boundary	569	whole pages, so that offset is effectively rounded down to a page boundary
283	and bytes are read up to the next page boundary greater than or equal to	570	and bytes are read up to the next page boundary greater than or equal to
284	(off-set+length). C<aio_readahead> does not read beyond the end of the	571	(off-set+length). C<aio_readahead> does not read beyond the end of the
285	file. The current file offset of the file is left unchanged.	572	file. The current file offset of the file is left unchanged.
286		573
287	If that syscall doesn't exist (likely if your OS isn't Linux) it will be	574	If that syscall doesn't exist (likely if your kernel isn't Linux) it will
288	emulated by simply reading the data, which would have a similar effect.	575	be emulated by simply reading the data, which would have a similar effect.
		576
289		577
290	=item aio_stat $fh_or_path, $callback->($status)	578	=item aio_stat $fh_or_path, $callback->($status)
291		579
292	=item aio_lstat $fh, $callback->($status)	580	=item aio_lstat $fh, $callback->($status)
293		581
294	Works like perl's C<stat> or C<lstat> in void context. The callback will	582	Works almost exactly like perl's C<stat> or C<lstat> in void context. The
295	be called after the stat and the results will be available using C<stat _>	583	callback will be called after the stat and the results will be available
296	or C<-s _> etc...	584	using C<stat _> or C<-s _> and other tests (with the exception of C<-B>
297		585	and C<-T>).
298	The pathname passed to C<aio_stat> must be absolute. See API NOTES, above,
299	for an explanation.
300		586
301	Currently, the stats are always 64-bit-stats, i.e. instead of returning an	587	Currently, the stats are always 64-bit-stats, i.e. instead of returning an
302	error when stat'ing a large file, the results will be silently truncated	588	error when stat'ing a large file, the results will be silently truncated
303	unless perl itself is compiled with large file support.	589	unless perl itself is compiled with large file support.
		590
		591	To help interpret the mode and dev/rdev stat values, IO::AIO offers the
		592	following constants and functions (if not implemented, the constants will
		593	be C<0> and the functions will either C<croak> or fall back on traditional
		594	behaviour).
		595
		596	C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
		597	C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
		598	C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
		599
		600	To access higher resolution stat timestamps, see L<SUBSECOND STAT TIME
		601	ACCESS>.
304		602
305	Example: Print the length of F</etc/passwd>:	603	Example: Print the length of F</etc/passwd>:
306		604
307	aio_stat "/etc/passwd", sub {	605	aio_stat "/etc/passwd", sub {
308	$_[0] and die "stat failed: $!";	606	$_[0] and die "stat failed: $!";
309	print "size is ", -s _, "\n";	607	print "size is ", -s _, "\n";
310	};	608	};
311		609
		610
		611	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		612
		613	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		614	whether a file handle or path was passed.
		615
		616	On success, the callback is passed a hash reference with the following
		617	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		618	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		619	is passed.
		620
		621	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		622	C<ST_NOSUID>.
		623
		624	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		625	their correct value when available, or to C<0> on systems that do
		626	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		627	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		628	C<ST_NODIRATIME> and C<ST_RELATIME>.
		629
		630	Example: stat C</wd> and dump out the data if successful.
		631
		632	aio_statvfs "/wd", sub {
		633	my $f = $_[0]
		634	or die "statvfs: $!";
		635
		636	use Data::Dumper;
		637	say Dumper $f;
		638	};
		639
		640	# result:
		641	{
		642	bsize => 1024,
		643	bfree => 4333064312,
		644	blocks => 10253828096,
		645	files => 2050765568,
		646	flag => 4096,
		647	favail => 2042092649,
		648	bavail => 4333064312,
		649	ffree => 2042092649,
		650	namemax => 255,
		651	frsize => 1024,
		652	fsid => 1810
		653	}
		654
		655	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		656
		657	Works like perl's C<utime> function (including the special case of $atime
		658	and $mtime being undef). Fractional times are supported if the underlying
		659	syscalls support them.
		660
		661	When called with a pathname, uses utimensat(2) or utimes(2) if available,
		662	otherwise utime(2). If called on a file descriptor, uses futimens(2)
		663	or futimes(2) if available, otherwise returns ENOSYS, so this is not
		664	portable.
		665
		666	Examples:
		667
		668	# set atime and mtime to current time (basically touch(1)):
		669	aio_utime "path", undef, undef;
		670	# set atime to current time and mtime to beginning of the epoch:
		671	aio_utime "path", time, undef; # undef==0
		672
		673
		674	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		675
		676	Works like perl's C<chown> function, except that C<undef> for either $uid
		677	or $gid is being interpreted as "do not change" (but -1 can also be used).
		678
		679	Examples:
		680
		681	# same as "chown root path" in the shell:
		682	aio_chown "path", 0, -1;
		683	# same as above:
		684	aio_chown "path", 0, undef;
		685
		686
		687	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		688
		689	Works like truncate(2) or ftruncate(2).
		690
		691
		692	=item aio_allocate $fh, $mode, $offset, $len, $callback->($status)
		693
		694	Allocates or frees disk space according to the C<$mode> argument. See the
		695	linux C<fallocate> documentation for details.
		696
		697	C<$mode> is usually C<0> or C<IO::AIO::FALLOC_FL_KEEP_SIZE> to allocate
		698	space, or C<IO::AIO::FALLOC_FL_PUNCH_HOLE | IO::AIO::FALLOC_FL_KEEP_SIZE>,
		699	to deallocate a file range.
		700
		701	IO::AIO also supports C<FALLOC_FL_COLLAPSE_RANGE>, to remove a range
		702	(without leaving a hole), C<FALLOC_FL_ZERO_RANGE>, to zero a range,
		703	C<FALLOC_FL_INSERT_RANGE> to insert a range and C<FALLOC_FL_UNSHARE_RANGE>
		704	to unshare shared blocks (see your L<fallocate(2)> manpage).
		705
		706	The file system block size used by C<fallocate> is presumably the
		707	C<f_bsize> returned by C<statvfs>, but different filesystems and filetypes
		708	can dictate other limitations.
		709
		710	If C<fallocate> isn't available or cannot be emulated (currently no
		711	emulation will be attempted), passes C<-1> and sets C<$!> to C<ENOSYS>.
		712
		713
		714	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		715
		716	Works like perl's C<chmod> function.
		717
		718
312	=item aio_unlink $pathname, $callback->($status)	719	=item aio_unlink $pathname, $callback->($status)
313		720
314	Asynchronously unlink (delete) a file and call the callback with the	721	Asynchronously unlink (delete) a file and call the callback with the
315	result code.	722	result code.
316		723
		724
		725	=item aio_mknod $pathname, $mode, $dev, $callback->($status)
		726
		727	[EXPERIMENTAL]
		728
		729	Asynchronously create a device node (or fifo). See mknod(2).
		730
		731	The only (POSIX-) portable way of calling this function is:
		732
		733	aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		734
		735	See C<aio_stat> for info about some potentially helpful extra constants
		736	and functions.
		737
317	=item aio_link $srcpath, $dstpath, $callback->($status)	738	=item aio_link $srcpath, $dstpath, $callback->($status)
318		739
319	Asynchronously create a new link to the existing object at C<$srcpath> at	740	Asynchronously create a new link to the existing object at C<$srcpath> at
320	the path C<$dstpath> and call the callback with the result code.	741	the path C<$dstpath> and call the callback with the result code.
321		742
		743
322	=item aio_symlink $srcpath, $dstpath, $callback->($status)	744	=item aio_symlink $srcpath, $dstpath, $callback->($status)
323		745
324	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at	746	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
325	the path C<$dstpath> and call the callback with the result code.	747	the path C<$dstpath> and call the callback with the result code.
326		748
		749
		750	=item aio_readlink $pathname, $callback->($link)
		751
		752	Asynchronously read the symlink specified by C<$path> and pass it to
		753	the callback. If an error occurs, nothing or undef gets passed to the
		754	callback.
		755
		756
		757	=item aio_realpath $pathname, $callback->($path)
		758
		759	Asynchronously make the path absolute and resolve any symlinks in
		760	C<$path>. The resulting path only consists of directories (same as
		761	L<Cwd::realpath>).
		762
		763	This request can be used to get the absolute path of the current working
		764	directory by passing it a path of F<.> (a single dot).
		765
		766
327	=item aio_rename $srcpath, $dstpath, $callback->($status)	767	=item aio_rename $srcpath, $dstpath, $callback->($status)
328		768
329	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as	769	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
330	rename(2) and call the callback with the result code.	770	rename(2) and call the callback with the result code.
331		771
		772	On systems that support the AIO::WD working directory abstraction
		773	natively, the case C<[$wd, "."]> as C<$srcpath> is specialcased - instead
		774	of failing, C<rename> is called on the absolute path of C<$wd>.
		775
		776
		777	=item aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
		778
		779	Basically a version of C<aio_rename> with an additional C<$flags>
		780	argument. Calling this with C<$flags=0> is the same as calling
		781	C<aio_rename>.
		782
		783	Non-zero flags are currently only supported on GNU/Linux systems that
		784	support renameat2. Other systems fail with C<ENOSYS> in this case.
		785
		786	The following constants are available (missing ones are, as usual C<0>),
		787	see renameat2(2) for details:
		788
		789	C<IO::AIO::RENAME_NOREPLACE>, C<IO::AIO::RENAME_EXCHANGE>
		790	and C<IO::AIO::RENAME_WHITEOUT>.
		791
		792
		793	=item aio_mkdir $pathname, $mode, $callback->($status)
		794
		795	Asynchronously mkdir (create) a directory and call the callback with
		796	the result code. C<$mode> will be modified by the umask at the time the
		797	request is executed, so do not change your umask.
		798
		799
332	=item aio_rmdir $pathname, $callback->($status)	800	=item aio_rmdir $pathname, $callback->($status)
333		801
334	Asynchronously rmdir (delete) a directory and call the callback with the	802	Asynchronously rmdir (delete) a directory and call the callback with the
335	result code.	803	result code.
		804
		805	On systems that support the AIO::WD working directory abstraction
		806	natively, the case C<[$wd, "."]> is specialcased - instead of failing,
		807	C<rmdir> is called on the absolute path of C<$wd>.
		808
336		809
337	=item aio_readdir $pathname, $callback->($entries)	810	=item aio_readdir $pathname, $callback->($entries)
338		811
339	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire	812	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
340	directory (i.e. opendir + readdir + closedir). The entries will not be	813	directory (i.e. opendir + readdir + closedir). The entries will not be
341	sorted, and will B<NOT> include the C<.> and C<..> entries.	814	sorted, and will B<NOT> include the C<.> and C<..> entries.
342		815
343	The callback a single argument which is either C<undef> or an array-ref	816	The callback is passed a single argument which is either C<undef> or an
344	with the filenames.	817	array-ref with the filenames.
345		818
		819
		820	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		821
		822	Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
		823	tune behaviour and output format. In case of an error, C<$entries> will be
		824	C<undef>.
		825
		826	The flags are a combination of the following constants, ORed together (the
		827	flags will also be passed to the callback, possibly modified):
		828
		829	=over 4
		830
		831	=item IO::AIO::READDIR_DENTS
		832
		833	Normally the callback gets an arrayref consisting of names only (as
		834	with C<aio_readdir>). If this flag is set, then the callback gets an
		835	arrayref with C<[$name, $type, $inode]> arrayrefs, each describing a
		836	single directory entry in more detail:
		837
		838	C<$name> is the name of the entry.
		839
		840	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		841
		842	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		843	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		844	C<IO::AIO::DT_WHT>.
		845
		846	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need
		847	to know, you have to run stat yourself. Also, for speed/memory reasons,
		848	the C<$type> scalars are read-only: you must not modify them.
		849
		850	C<$inode> is the inode number (which might not be exact on systems with 64
		851	bit inode numbers and 32 bit perls). This field has unspecified content on
		852	systems that do not deliver the inode information.
		853
		854	=item IO::AIO::READDIR_DIRS_FIRST
		855
		856	When this flag is set, then the names will be returned in an order where
		857	likely directories come first, in optimal stat order. This is useful when
		858	you need to quickly find directories, or you want to find all directories
		859	while avoiding to stat() each entry.
		860
		861	If the system returns type information in readdir, then this is used
		862	to find directories directly. Otherwise, likely directories are names
		863	beginning with ".", or otherwise names with no dots, of which names with
		864	short names are tried first.
		865
		866	=item IO::AIO::READDIR_STAT_ORDER
		867
		868	When this flag is set, then the names will be returned in an order
		869	suitable for stat()'ing each one. That is, when you plan to stat() most or
		870	all files in the given directory, then the returned order will likely be
		871	faster.
		872
		873	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified,
		874	then the likely dirs come first, resulting in a less optimal stat order
		875	for stat'ing all entries, but likely a more optimal order for finding
		876	subdirectories.
		877
		878	=item IO::AIO::READDIR_FOUND_UNKNOWN
		879
		880	This flag should not be set when calling C<aio_readdirx>. Instead, it
		881	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		882	C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
		883	C<$type>'s are known, which can be used to speed up some algorithms.
		884
		885	=back
		886
		887
		888	=item aio_slurp $pathname, $offset, $length, $data, $callback->($status)
		889
		890	Opens, reads and closes the given file. The data is put into C<$data>,
		891	which is resized as required.
		892
		893	If C<$offset> is negative, then it is counted from the end of the file.
		894
		895	If C<$length> is zero, then the remaining length of the file is
		896	used. Also, in this case, the same limitations to modifying C<$data> apply
		897	as when IO::AIO::mmap is used, i.e. it must only be modified in-place
		898	with C<substr>. If the size of the file is known, specifying a non-zero
		899	C<$length> results in a performance advantage.
		900
		901	This request is similar to the older C<aio_load> request, but since it is
		902	a single request, it might be more efficient to use.
		903
		904	Example: load F</etc/passwd> into C<$passwd>.
		905
		906	my $passwd;
		907	aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
		908	$_[0] >= 0
		909	or die "/etc/passwd: $!\n";
		910
		911	printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
		912	print $passwd;
		913	};
		914	IO::AIO::flush;
		915
		916
		917	=item aio_load $pathname, $data, $callback->($status)
		918
		919	This is a composite request that tries to fully load the given file into
		920	memory. Status is the same as with aio_read.
		921
		922	Using C<aio_slurp> might be more efficient, as it is a single request.
		923
		924	=cut
		925
		926	sub aio_load($$;$) {
		927	my ($path, undef, $cb) = @_;
		928	my $data = \$_[1];
		929
		930	my $pri = aioreq_pri;
		931	my $grp = aio_group $cb;
		932
		933	aioreq_pri $pri;
		934	add $grp aio_open $path, O_RDONLY, 0, sub {
		935	my $fh = shift
		936	or return $grp->result (-1);
		937
		938	aioreq_pri $pri;
		939	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		940	$grp->result ($_[0]);
		941	};
		942	};
		943
		944	$grp
		945	}
		946
		947	=item aio_copy $srcpath, $dstpath, $callback->($status)
		948
		949	Try to copy the I<file> (directories not supported as either source or
		950	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		951	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		952
		953	Existing destination files will be truncated.
		954
		955	This is a composite request that creates the destination file with
		956	mode 0200 and copies the contents of the source file into it using
		957	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		958	uid/gid, in that order.
		959
		960	If an error occurs, the partial destination file will be unlinked, if
		961	possible, except when setting atime, mtime, access mode and uid/gid, where
		962	errors are being ignored.
		963
		964	=cut
		965
		966	sub aio_copy($$;$) {
		967	my ($src, $dst, $cb) = @_;
		968
		969	my $pri = aioreq_pri;
		970	my $grp = aio_group $cb;
		971
		972	aioreq_pri $pri;
		973	add $grp aio_open $src, O_RDONLY, 0, sub {
		974	if (my $src_fh = $_[0]) {
		975	my @stat = stat $src_fh; # hmm, might block over nfs?
		976
		977	aioreq_pri $pri;
		978	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		979	if (my $dst_fh = $_[0]) {
		980
		981	# best-effort preallocate
		982	aioreq_pri $pri;
		983	add $grp aio_allocate $dst_fh, IO::AIO::FALLOC_FL_KEEP_SIZE, 0, $stat[7], sub { };
		984
		985	aioreq_pri $pri;
		986	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		987	if ($_[0] == $stat[7]) {
		988	$grp->result (0);
		989	close $src_fh;
		990
		991	my $ch = sub {
		992	aioreq_pri $pri;
		993	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		994	aioreq_pri $pri;
		995	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		996	aioreq_pri $pri;
		997	add $grp aio_close $dst_fh;
		998	}
		999	};
		1000	};
		1001
		1002	aioreq_pri $pri;
		1003	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		1004	if ($_[0] < 0 && $! == ENOSYS) {
		1005	aioreq_pri $pri;
		1006	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		1007	} else {
		1008	$ch->();
		1009	}
		1010	};
		1011	} else {
		1012	$grp->result (-1);
		1013	close $src_fh;
		1014	close $dst_fh;
		1015
		1016	aioreq $pri;
		1017	add $grp aio_unlink $dst;
		1018	}
		1019	};
		1020	} else {
		1021	$grp->result (-1);
		1022	}
		1023	},
		1024
		1025	} else {
		1026	$grp->result (-1);
		1027	}
		1028	};
		1029
		1030	$grp
		1031	}
		1032
		1033	=item aio_move $srcpath, $dstpath, $callback->($status)
		1034
		1035	Try to move the I<file> (directories not supported as either source or
		1036	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		1037	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		1038
		1039	This is a composite request that tries to rename(2) the file first; if
		1040	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		1041	that is successful, unlinks the C<$srcpath>.
		1042
		1043	=cut
		1044
		1045	sub aio_move($$;$) {
		1046	my ($src, $dst, $cb) = @_;
		1047
		1048	my $pri = aioreq_pri;
		1049	my $grp = aio_group $cb;
		1050
		1051	aioreq_pri $pri;
		1052	add $grp aio_rename $src, $dst, sub {
		1053	if ($_[0] && $! == EXDEV) {
		1054	aioreq_pri $pri;
		1055	add $grp aio_copy $src, $dst, sub {
		1056	$grp->result ($_[0]);
		1057
		1058	unless ($_[0]) {
		1059	aioreq_pri $pri;
		1060	add $grp aio_unlink $src;
		1061	}
		1062	};
		1063	} else {
		1064	$grp->result ($_[0]);
		1065	}
		1066	};
		1067
		1068	$grp
		1069	}
		1070
346	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)	1071	=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
347		1072
348	Scans a directory (similar to C<aio_readdir>) but additionally tries to	1073	Scans a directory (similar to C<aio_readdir>) but additionally tries to
349	separate the entries of directory C<$path> into two sets of names, ones	1074	efficiently separate the entries of directory C<$path> into two sets of
350	you can recurse into (directories or links to them), and ones you cannot	1075	names, directories you can recurse into (directories), and ones you cannot
351	recurse into (everything else).	1076	recurse into (everything else, including symlinks to directories).
352		1077
353	C<aio_scandir> is a composite request that consists of many sub	1078	C<aio_scandir> is a composite request that generates many sub requests.
354	requests. C<$maxreq> specifies the maximum number of outstanding aio	1079	C<$maxreq> specifies the maximum number of outstanding aio requests that
355	requests that this function generates. If it is C<< <= 0 >>, then a	1080	this function generates. If it is C<< <= 0 >>, then a suitable default
356	suitable default will be chosen (currently 8).	1081	will be chosen (currently 4).
357		1082
358	On error, the callback is called without arguments, otherwise it receives	1083	On error, the callback is called without arguments, otherwise it receives
359	two array-refs with path-relative entry names.	1084	two array-refs with path-relative entry names.
360		1085
361	Example:	1086	Example:
…		…
368		1093
369	Implementation notes.	1094	Implementation notes.
370		1095
371	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.	1096	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
372		1097
		1098	If readdir returns file type information, then this is used directly to
		1099	find directories.
		1100
373	After reading the directory, the modification time, size etc. of the	1101	Otherwise, after reading the directory, the modification time, size etc.
374	directory before and after the readdir is checked, and if they match (and	1102	of the directory before and after the readdir is checked, and if they
375	isn't the current time), the link count will be used to decide how many	1103	match (and isn't the current time), the link count will be used to decide
376	entries are directories (if >= 2). Otherwise, no knowledge of the number	1104	how many entries are directories (if >= 2). Otherwise, no knowledge of the
377	of subdirectories will be assumed.	1105	number of subdirectories will be assumed.
378		1106
379	Then entries will be sorted into likely directories (everything without	1107	Then entries will be sorted into likely directories a non-initial dot
380	a non-initial dot currently) and likely non-directories (everything	1108	currently) and likely non-directories (see C<aio_readdirx>). Then every
381	else). Then every entry plus an appended C</.> will be C<stat>'ed,	1109	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
382	likely directories first. If that succeeds, it assumes that the entry	1110	in order of their inode numbers. If that succeeds, it assumes that the
383	is a directory or a symlink to directory (which will be checked	1111	entry is a directory or a symlink to directory (which will be checked
384	seperately). This is often faster than stat'ing the entry itself because	1112	separately). This is often faster than stat'ing the entry itself because
385	filesystems might detect the type of the entry without reading the inode	1113	filesystems might detect the type of the entry without reading the inode
386	data (e.g. ext2fs filetype feature).	1114	data (e.g. ext2fs filetype feature), even on systems that cannot return
		1115	the filetype information on readdir.
387		1116
388	If the known number of directories (link count - 2) has been reached, the	1117	If the known number of directories (link count - 2) has been reached, the
389	rest of the entries is assumed to be non-directories.	1118	rest of the entries is assumed to be non-directories.
390		1119
391	This only works with certainty on POSIX (= UNIX) filesystems, which	1120	This only works with certainty on POSIX (= UNIX) filesystems, which
…		…
395	as those tend to return 0 or 1 as link counts, which disables the	1124	as those tend to return 0 or 1 as link counts, which disables the
396	directory counting heuristic.	1125	directory counting heuristic.
397		1126
398	=cut	1127	=cut
399		1128
400	sub aio_scandir($$$) {	1129	sub aio_scandir($$;$) {
401	my ($path, $maxreq, $cb) = @_;	1130	my ($path, $maxreq, $cb) = @_;
402		1131
		1132	my $pri = aioreq_pri;
		1133
403	my $grp = aio_group;	1134	my $grp = aio_group $cb;
404		1135
405	$maxreq = 8 if $maxreq <= 0;	1136	$maxreq = 4 if $maxreq <= 0;
406		1137
407	# stat once	1138	# get a wd object
		1139	aioreq_pri $pri;
408	add $grp aio_stat $path, sub {	1140	add $grp aio_wd $path, sub {
		1141	$_[0]
		1142	or return $grp->result ();
		1143
		1144	my $wd = [shift, "."];
		1145
		1146	# stat once
		1147	aioreq_pri $pri;
		1148	add $grp aio_stat $wd, sub {
409	return $cb->() if $_[0];	1149	return $grp->result () if $_[0];
410	my $now = time;	1150	my $now = time;
411	my $hash1 = join ":", (stat _)[0,1,3,7,9];	1151	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		1152	my $rdxflags = READDIR_DIRS_FIRST;
412		1153
		1154	if ((stat _)[3] < 2) {
		1155	# at least one non-POSIX filesystem exists
		1156	# that returns useful DT_type values: btrfs,
		1157	# so optimise for this here by requesting dents
		1158	$rdxflags |= READDIR_DENTS;
		1159	}
		1160
413	# read the directory entries	1161	# read the directory entries
		1162	aioreq_pri $pri;
414	add $grp aio_readdir $path, sub {	1163	add $grp aio_readdirx $wd, $rdxflags, sub {
415	my $entries = shift	1164	my ($entries, $flags) = @_
416	or return $cb->();	1165	or return $grp->result ();
417		1166
418	# stat the dir another time	1167	if ($rdxflags & READDIR_DENTS) {
419	add $grp aio_stat $path, sub {	1168	# if we requested type values, see if we can use them directly.
420	my $hash2 = join ":", (stat _)[0,1,3,7,9];
421		1169
		1170	# if there were any DT_UNKNOWN entries then we assume we
		1171	# don't know. alternatively, we could assume that if we get
		1172	# one DT_DIR, then all directories are indeed marked with
		1173	# DT_DIR, but this seems not required for btrfs, and this
		1174	# is basically the "btrfs can't get it's act together" code
		1175	# branch.
		1176	unless ($flags & READDIR_FOUND_UNKNOWN) {
		1177	# now we have valid DT_ information for all entries,
		1178	# so use it as an optimisation without further stat's.
		1179	# they must also all be at the beginning of @$entries
		1180	# by now.
		1181
422	my $ndirs;	1182	my $dirs;
423		1183
424	# take the slow route if anything looks fishy	1184	if (@$entries) {
425	if ($hash1 ne $hash2 or (stat _)[9] == $now) {	1185	for (0 .. $#$entries) {
426	$ndirs = -1;	1186	if ($entries->[$_][1] != DT_DIR) {
		1187	# splice out directories
		1188	$dirs = [splice @$entries, 0, $_];
		1189	last;
		1190	}
		1191	}
		1192
		1193	# if we didn't find any non-dir, then all entries are dirs
		1194	unless ($dirs) {
		1195	($dirs, $entries) = ($entries, []);
		1196	}
427	} else {	1197	} else {
428	# if nlink == 2, we are finished	1198	# directory is empty, so there are no sbdirs
429	# on non-posix-fs's, we rely on nlink < 2	1199	$dirs = [];
430	$ndirs = (stat _)[3] - 2	1200	}
		1201
		1202	# either splice'd the directories out or the dir was empty.
		1203	# convert dents to filenames
		1204	$_ = $_->[0] for @$dirs;
		1205	$_ = $_->[0] for @$entries;
		1206
431	or return $cb->([], $entries);	1207	return $grp->result ($dirs, $entries);
		1208	}
		1209
		1210	# cannot use, so return to our old ways
		1211	# by pretending we only scanned for names.
		1212	$_ = $_->[0] for @$entries;
432	}	1213	}
433		1214
434	# sort into likely dirs and likely nondirs	1215	# stat the dir another time
435	# dirs == files without ".", short entries first	1216	aioreq_pri $pri;
436	$entries = [map $_->[0],	1217	add $grp aio_stat $wd, sub {
437	sort { $b->[1] cmp $a->[1] }	1218	my $hash2 = join ":", (stat _)[0,1,3,7,9];
438	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
439	@$entries];
440		1219
441	my (@dirs, @nondirs);
442
443	my ($statcb, $schedcb);
444	my $nreq = 0;	1220	my $ndirs;
445		1221
446	$schedcb = sub {	1222	# take the slow route if anything looks fishy
447	if (@$entries) {	1223	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
448	if ($nreq < $maxreq) {	1224	$ndirs = -1;
449	my $ent = pop @$entries;
450	$nreq++;
451	add $grp aio_stat "$path/$ent/.", sub { $statcb->($_[0], $ent) };
452	}
453	} elsif (!$nreq) {	1225	} else {
454	# finished	1226	# if nlink == 2, we are finished
455	undef $statcb;	1227	# for non-posix-fs's, we rely on nlink < 2
456	undef $schedcb;	1228	$ndirs = (stat _)[3] - 2
457	$cb->(\@dirs, \@nondirs) if $cb;	1229	or return $grp->result ([], $entries);
458	undef $cb;
459	}	1230	}
		1231
		1232	my (@dirs, @nondirs);
		1233
		1234	my $statgrp = add $grp aio_group sub {
		1235	$grp->result (\@dirs, \@nondirs);
		1236	};
		1237
		1238	limit $statgrp $maxreq;
		1239	feed $statgrp sub {
		1240	return unless @$entries;
		1241	my $entry = shift @$entries;
		1242
		1243	aioreq_pri $pri;
		1244	$wd->[1] = "$entry/.";
		1245	add $statgrp aio_stat $wd, sub {
		1246	if ($_[0] < 0) {
		1247	push @nondirs, $entry;
		1248	} else {
		1249	# need to check for real directory
		1250	aioreq_pri $pri;
		1251	$wd->[1] = $entry;
		1252	add $statgrp aio_lstat $wd, sub {
		1253	if (-d _) {
		1254	push @dirs, $entry;
		1255
		1256	unless (--$ndirs) {
		1257	push @nondirs, @$entries;
		1258	feed $statgrp;
		1259	}
		1260	} else {
		1261	push @nondirs, $entry;
		1262	}
		1263	}
		1264	}
		1265	};
		1266	};
460	};	1267	};
461	$statcb = sub {
462	my ($status, $entry) = @_;
463
464	if ($status < 0) {
465	$nreq--;
466	push @nondirs, $entry;
467	&$schedcb;
468	} else {
469	# need to check for real directory
470	add $grp aio_lstat "$path/$entry", sub {
471	$nreq--;
472
473	if (-d _) {
474	push @dirs, $entry;
475
476	if (!--$ndirs) {
477	push @nondirs, @$entries;
478	$entries = [];
479	}
480	} else {
481	push @nondirs, $entry;
482	}
483
484	&$schedcb;
485	}
486	}
487	};
488
489	&$schedcb while @$entries && $nreq < $maxreq;
490	};	1268	};
491	};	1269	};
492	};	1270	};
493		1271
494	$grp	1272	$grp
495	}	1273	}
496		1274
		1275	=item aio_rmtree $pathname, $callback->($status)
		1276
		1277	Delete a directory tree starting (and including) C<$path>, return the
		1278	status of the final C<rmdir> only. This is a composite request that
		1279	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1280	everything else.
		1281
		1282	=cut
		1283
		1284	sub aio_rmtree;
		1285	sub aio_rmtree($;$) {
		1286	my ($path, $cb) = @_;
		1287
		1288	my $pri = aioreq_pri;
		1289	my $grp = aio_group $cb;
		1290
		1291	aioreq_pri $pri;
		1292	add $grp aio_scandir $path, 0, sub {
		1293	my ($dirs, $nondirs) = @_;
		1294
		1295	my $dirgrp = aio_group sub {
		1296	add $grp aio_rmdir $path, sub {
		1297	$grp->result ($_[0]);
		1298	};
		1299	};
		1300
		1301	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1302	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1303
		1304	add $grp $dirgrp;
		1305	};
		1306
		1307	$grp
		1308	}
		1309
		1310	=item aio_fcntl $fh, $cmd, $arg, $callback->($status)
		1311
		1312	=item aio_ioctl $fh, $request, $buf, $callback->($status)
		1313
		1314	These work just like the C<fcntl> and C<ioctl> built-in functions, except
		1315	they execute asynchronously and pass the return value to the callback.
		1316
		1317	Both calls can be used for a lot of things, some of which make more sense
		1318	to run asynchronously in their own thread, while some others make less
		1319	sense. For example, calls that block waiting for external events, such
		1320	as locking, will also lock down an I/O thread while it is waiting, which
		1321	can deadlock the whole I/O system. At the same time, there might be no
		1322	alternative to using a thread to wait.
		1323
		1324	So in general, you should only use these calls for things that do
		1325	(filesystem) I/O, not for things that wait for other events (network,
		1326	other processes), although if you are careful and know what you are doing,
		1327	you still can.
		1328
		1329	The following constants are available and can be used for normal C<ioctl>
		1330	and C<fcntl> as well (missing ones are, as usual C<0>):
		1331
		1332	C<F_DUPFD_CLOEXEC>,
		1333
		1334	C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>,
		1335
		1336	C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>.
		1337
		1338	C<F_ADD_SEALS>, C<F_GET_SEALS>, C<F_SEAL_SEAL>, C<F_SEAL_SHRINK>, C<F_SEAL_GROW> and
		1339	C<F_SEAL_WRITE>.
		1340
		1341	C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>,
		1342	C<FS_IOC_FIEMAP>.
		1343
		1344	C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>,
		1345	C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>.
		1346
		1347	C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>,
		1348	C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>,
		1349	C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>,
		1350	C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>,
		1351	C<FS_FL_USER_MODIFIABLE>.
		1352
		1353	C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>,
		1354	C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>,
		1355	C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>,
		1356	C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>,
		1357
		1358	C<BLKROSET>, C<BLKROGET>, C<BLKRRPART>, C<BLKGETSIZE>, C<BLKFLSBUF>, C<BLKRASET>,
		1359	C<BLKRAGET>, C<BLKFRASET>, C<BLKFRAGET>, C<BLKSECTSET>, C<BLKSECTGET>, C<BLKSSZGET>,
		1360	C<BLKBSZGET>, C<BLKBSZSET>, C<BLKGETSIZE64>,
		1361
		1362
		1363	=item aio_sync $callback->($status)
		1364
		1365	Asynchronously call sync and call the callback when finished.
		1366
497	=item aio_fsync $fh, $callback->($status)	1367	=item aio_fsync $fh, $callback->($status)
498		1368
499	Asynchronously call fsync on the given filehandle and call the callback	1369	Asynchronously call fsync on the given filehandle and call the callback
500	with the fsync result code.	1370	with the fsync result code.
501		1371
…		…
505	callback with the fdatasync result code.	1375	callback with the fdatasync result code.
506		1376
507	If this call isn't available because your OS lacks it or it couldn't be	1377	If this call isn't available because your OS lacks it or it couldn't be
508	detected, it will be emulated by calling C<fsync> instead.	1378	detected, it will be emulated by calling C<fsync> instead.
509		1379
		1380	=item aio_syncfs $fh, $callback->($status)
		1381
		1382	Asynchronously call the syncfs syscall to sync the filesystem associated
		1383	to the given filehandle and call the callback with the syncfs result
		1384	code. If syncfs is not available, calls sync(), but returns C<-1> and sets
		1385	errno to C<ENOSYS> nevertheless.
		1386
		1387	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1388
		1389	Sync the data portion of the file specified by C<$offset> and C<$length>
		1390	to disk (but NOT the metadata), by calling the Linux-specific
		1391	sync_file_range call. If sync_file_range is not available or it returns
		1392	ENOSYS, then fdatasync or fsync is being substituted.
		1393
		1394	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1395	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1396	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1397	manpage for details.
		1398
		1399	=item aio_pathsync $pathname, $callback->($status)
		1400
		1401	This request tries to open, fsync and close the given path. This is a
		1402	composite request intended to sync directories after directory operations
		1403	(E.g. rename). This might not work on all operating systems or have any
		1404	specific effect, but usually it makes sure that directory changes get
		1405	written to disc. It works for anything that can be opened for read-only,
		1406	not just directories.
		1407
		1408	Future versions of this function might fall back to other methods when
		1409	C<fsync> on the directory fails (such as calling C<sync>).
		1410
		1411	Passes C<0> when everything went ok, and C<-1> on error.
		1412
		1413	=cut
		1414
		1415	sub aio_pathsync($;$) {
		1416	my ($path, $cb) = @_;
		1417
		1418	my $pri = aioreq_pri;
		1419	my $grp = aio_group $cb;
		1420
		1421	aioreq_pri $pri;
		1422	add $grp aio_open $path, O_RDONLY, 0, sub {
		1423	my ($fh) = @_;
		1424	if ($fh) {
		1425	aioreq_pri $pri;
		1426	add $grp aio_fsync $fh, sub {
		1427	$grp->result ($_[0]);
		1428
		1429	aioreq_pri $pri;
		1430	add $grp aio_close $fh;
		1431	};
		1432	} else {
		1433	$grp->result (-1);
		1434	}
		1435	};
		1436
		1437	$grp
		1438	}
		1439
		1440	=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
		1441
		1442	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1443	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1444	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1445	scalar must only be modified in-place while an aio operation is pending on
		1446	it).
		1447
		1448	It calls the C<msync> function of your OS, if available, with the memory
		1449	area starting at C<$offset> in the string and ending C<$length> bytes
		1450	later. If C<$length> is negative, counts from the end, and if C<$length>
		1451	is C<undef>, then it goes till the end of the string. The flags can be
		1452	either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional
		1453	C<IO::AIO::MS_INVALIDATE>.
		1454
		1455	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1456
		1457	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1458	scalars.
		1459
		1460	It touches (reads or writes) all memory pages in the specified
		1461	range inside the scalar. All caveats and parameters are the same
		1462	as for C<aio_msync>, above, except for flags, which must be either
		1463	C<0> (which reads all pages and ensures they are instantiated) or
		1464	C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and
		1465	writing an octet from it, which dirties the page).
		1466
		1467	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1468
		1469	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1470	scalars.
		1471
		1472	It reads in all the pages of the underlying storage into memory (if any)
		1473	and locks them, so they are not getting swapped/paged out or removed.
		1474
		1475	If C<$length> is undefined, then the scalar will be locked till the end.
		1476
		1477	On systems that do not implement C<mlock>, this function returns C<-1>
		1478	and sets errno to C<ENOSYS>.
		1479
		1480	Note that the corresponding C<munlock> is synchronous and is
		1481	documented under L<MISCELLANEOUS FUNCTIONS>.
		1482
		1483	Example: open a file, mmap and mlock it - both will be undone when
		1484	C<$data> gets destroyed.
		1485
		1486	open my $fh, "<", $path or die "$path: $!";
		1487	my $data;
		1488	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1489	aio_mlock $data; # mlock in background
		1490
		1491	=item aio_mlockall $flags, $callback->($status)
		1492
		1493	Calls the C<mlockall> function with the given C<$flags> (a
		1494	combination of C<IO::AIO::MCL_CURRENT>, C<IO::AIO::MCL_FUTURE> and
		1495	C<IO::AIO::MCL_ONFAULT>).
		1496
		1497	On systems that do not implement C<mlockall>, this function returns C<-1>
		1498	and sets errno to C<ENOSYS>. Similarly, flag combinations not supported
		1499	by the system result in a return value of C<-1> with errno being set to
		1500	C<EINVAL>.
		1501
		1502	Note that the corresponding C<munlockall> is synchronous and is
		1503	documented under L<MISCELLANEOUS FUNCTIONS>.
		1504
		1505	Example: asynchronously lock all current and future pages into memory.
		1506
		1507	aio_mlockall IO::AIO::MCL_FUTURE;
		1508
		1509	=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
		1510
		1511	Queries the extents of the given file (by calling the Linux C<FIEMAP>
		1512	ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If
		1513	the ioctl is not available on your OS, then this request will fail with
		1514	C<ENOSYS>.
		1515
		1516	C<$start> is the starting offset to query extents for, C<$length> is the
		1517	size of the range to query - if it is C<undef>, then the whole file will
		1518	be queried.
		1519
		1520	C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or
		1521	C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also
		1522	exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query
		1523	the data portion.
		1524
		1525	C<$count> is the maximum number of extent records to return. If it is
		1526	C<undef>, then IO::AIO queries all extents of the range. As a very special
		1527	case, if it is C<0>, then the callback receives the number of extents
		1528	instead of the extents themselves (which is unreliable, see below).
		1529
		1530	If an error occurs, the callback receives no arguments. The special
		1531	C<errno> value C<IO::AIO::EBADR> is available to test for flag errors.
		1532
		1533	Otherwise, the callback receives an array reference with extent
		1534	structures. Each extent structure is an array reference itself, with the
		1535	following members:
		1536
		1537	[$logical, $physical, $length, $flags]
		1538
		1539	Flags is any combination of the following flag values (typically either C<0>
		1540	or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)):
		1541
		1542	C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>,
		1543	C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>,
		1544	C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>,
		1545	C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>,
		1546	C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or
		1547	C<IO::AIO::FIEMAP_EXTENT_SHARED>.
		1548
		1549	At the time of this writing (Linux 3.2), this request is unreliable unless
		1550	C<$count> is C<undef>, as the kernel has all sorts of bugs preventing
		1551	it to return all extents of a range for files with a large number of
		1552	extents. The code (only) works around all these issues if C<$count> is
		1553	C<undef>.
		1554
510	=item aio_group $callback->()	1555	=item aio_group $callback->(...)
511
512	[EXPERIMENTAL]
513		1556
514	This is a very special aio request: Instead of doing something, it is a	1557	This is a very special aio request: Instead of doing something, it is a
515	container for other aio requests, which is useful if you want to bundle	1558	container for other aio requests, which is useful if you want to bundle
516	many requests into a single, composite, request.	1559	many requests into a single, composite, request with a definite callback
		1560	and the ability to cancel the whole request with its subrequests.
517		1561
518	Returns an object of class L<IO::AIO::GRP>. See its documentation below	1562	Returns an object of class L<IO::AIO::GRP>. See its documentation below
519	for more info.	1563	for more info.
520		1564
521	Example:	1565	Example:
…		…
527	add $grp	1571	add $grp
528	(aio_stat ...),	1572	(aio_stat ...),
529	(aio_stat ...),	1573	(aio_stat ...),
530	...;	1574	...;
531		1575
		1576	=item aio_nop $callback->()
		1577
		1578	This is a special request - it does nothing in itself and is only used for
		1579	side effects, such as when you want to add a dummy request to a group so
		1580	that finishing the requests in the group depends on executing the given
		1581	code.
		1582
		1583	While this request does nothing, it still goes through the execution
		1584	phase and still requires a worker thread. Thus, the callback will not
		1585	be executed immediately but only after other requests in the queue have
		1586	entered their execution phase. This can be used to measure request
		1587	latency.
		1588
532	=item IO::AIO::aio_sleep $fractional_seconds, $callback->() *NOT EXPORTED*	1589	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
533		1590
534	Mainly used for debugging and benchmarking, this aio request puts one of	1591	Mainly used for debugging and benchmarking, this aio request puts one of
535	the request workers to sleep for the given time.	1592	the request workers to sleep for the given time.
536		1593
537	While it is theoretically handy to have simple I/O scheduling requests	1594	While it is theoretically handy to have simple I/O scheduling requests
538	like sleep and file handle readable/writable, the overhead this creates	1595	like sleep and file handle readable/writable, the overhead this creates is
539	is immense, so do not use this function except to put your application	1596	immense (it blocks a thread for a long time) so do not use this function
540	under artificial I/O pressure.	1597	except to put your application under artificial I/O pressure.
541		1598
542	=back	1599	=back
		1600
		1601
		1602	=head2 IO::AIO::WD - multiple working directories
		1603
		1604	Your process only has one current working directory, which is used by all
		1605	threads. This makes it hard to use relative paths (some other component
		1606	could call C<chdir> at any time, and it is hard to control when the path
		1607	will be used by IO::AIO).
		1608
		1609	One solution for this is to always use absolute paths. This usually works,
		1610	but can be quite slow (the kernel has to walk the whole path on every
		1611	access), and can also be a hassle to implement.
		1612
		1613	Newer POSIX systems have a number of functions (openat, fdopendir,
		1614	futimensat and so on) that make it possible to specify working directories
		1615	per operation.
		1616
		1617	For portability, and because the clowns who "designed", or shall I write,
		1618	perpetrated this new interface were obviously half-drunk, this abstraction
		1619	cannot be perfect, though.
		1620
		1621	IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
		1622	object. This object stores the canonicalised, absolute version of the
		1623	path, and on systems that allow it, also a directory file descriptor.
		1624
		1625	Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
		1626	or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
		1627	object and a pathname instead (or the IO::AIO::WD object alone, which
		1628	gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
		1629	IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
		1630	to that IO::AIO::WD object.
		1631
		1632	For example, to get a wd object for F</etc> and then stat F<passwd>
		1633	inside, you would write:
		1634
		1635	aio_wd "/etc", sub {
		1636	my $etcdir = shift;
		1637
		1638	# although $etcdir can be undef on error, there is generally no reason
		1639	# to check for errors here, as aio_stat will fail with ENOENT
		1640	# when $etcdir is undef.
		1641
		1642	aio_stat [$etcdir, "passwd"], sub {
		1643	# yay
		1644	};
		1645	};
		1646
		1647	The fact that C<aio_wd> is a request and not a normal function shows that
		1648	creating an IO::AIO::WD object is itself a potentially blocking operation,
		1649	which is why it is done asynchronously.
		1650
		1651	To stat the directory obtained with C<aio_wd> above, one could write
		1652	either of the following three request calls:
		1653
		1654	aio_lstat "/etc" , sub { ... # pathname as normal string
		1655	aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
		1656	aio_lstat $wd , sub { ... # shorthand for the previous
		1657
		1658	As with normal pathnames, IO::AIO keeps a copy of the working directory
		1659	object and the pathname string, so you could write the following without
		1660	causing any issues due to C<$path> getting reused:
		1661
		1662	my $path = [$wd, undef];
		1663
		1664	for my $name (qw(abc def ghi)) {
		1665	$path->[1] = $name;
		1666	aio_stat $path, sub {
		1667	# ...
		1668	};
		1669	}
		1670
		1671	There are some caveats: when directories get renamed (or deleted), the
		1672	pathname string doesn't change, so will point to the new directory (or
		1673	nowhere at all), while the directory fd, if available on the system,
		1674	will still point to the original directory. Most functions accepting a
		1675	pathname will use the directory fd on newer systems, and the string on
		1676	older systems. Some functions (such as C<aio_realpath>) will always rely on
		1677	the string form of the pathname.
		1678
		1679	So this functionality is mainly useful to get some protection against
		1680	C<chdir>, to easily get an absolute path out of a relative path for future
		1681	reference, and to speed up doing many operations in the same directory
		1682	(e.g. when stat'ing all files in a directory).
		1683
		1684	The following functions implement this working directory abstraction:
		1685
		1686	=over 4
		1687
		1688	=item aio_wd $pathname, $callback->($wd)
		1689
		1690	Asynchonously canonicalise the given pathname and convert it to an
		1691	IO::AIO::WD object representing it. If possible and supported on the
		1692	system, also open a directory fd to speed up pathname resolution relative
		1693	to this working directory.
		1694
		1695	If something goes wrong, then C<undef> is passwd to the callback instead
		1696	of a working directory object and C<$!> is set appropriately. Since
		1697	passing C<undef> as working directory component of a pathname fails the
		1698	request with C<ENOENT>, there is often no need for error checking in the
		1699	C<aio_wd> callback, as future requests using the value will fail in the
		1700	expected way.
		1701
		1702	=item IO::AIO::CWD
		1703
		1704	This is a compile time constant (object) that represents the process
		1705	current working directory.
		1706
		1707	Specifying this object as working directory object for a pathname is as if
		1708	the pathname would be specified directly, without a directory object. For
		1709	example, these calls are functionally identical:
		1710
		1711	aio_stat "somefile", sub { ... };
		1712	aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
		1713
		1714	=back
		1715
		1716	To recover the path associated with an IO::AIO::WD object, you can use
		1717	C<aio_realpath>:
		1718
		1719	aio_realpath $wd, sub {
		1720	warn "path is $_[0]\n";
		1721	};
		1722
		1723	Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir>
		1724	sometimes, fall back to using an absolue path.
543		1725
544	=head2 IO::AIO::REQ CLASS	1726	=head2 IO::AIO::REQ CLASS
545		1727
546	All non-aggregate C<aio_*> functions return an object of this class when	1728	All non-aggregate C<aio_*> functions return an object of this class when
547	called in non-void context.	1729	called in non-void context.
548		1730
549	A request always moves through the following five states in its lifetime,
550	in order: B<ready> (request has been created, but has not been executed
551	yet), B<execute> (request is currently being executed), B<pending>
552	(request has been executed but callback has not been called yet),
553	B<result> (results are being processed synchronously, includes calling the
554	callback) and B<done> (request has reached the end of its lifetime and
555	holds no resources anymore).
556
557	=over 4	1731	=over 4
558		1732
559	=item $req->cancel	1733	=item cancel $req
560		1734
561	Cancels the request, if possible. Has the effect of skipping execution	1735	Cancels the request, if possible. Has the effect of skipping execution
562	when entering the B<execute> state and skipping calling the callback when	1736	when entering the B<execute> state and skipping calling the callback when
563	entering the the B<result> state, but will leave the request otherwise	1737	entering the the B<result> state, but will leave the request otherwise
564	untouched. That means that requests that currently execute will not be	1738	untouched (with the exception of readdir). That means that requests that
565	stopped and resources held by the request will not be freed prematurely.	1739	currently execute will not be stopped and resources held by the request
		1740	will not be freed prematurely.
		1741
		1742	=item cb $req $callback->(...)
		1743
		1744	Replace (or simply set) the callback registered to the request.
566		1745
567	=back	1746	=back
568		1747
569	=head2 IO::AIO::GRP CLASS	1748	=head2 IO::AIO::GRP CLASS
570		1749
…		…
585	You add requests by calling the C<add> method with one or more	1764	You add requests by calling the C<add> method with one or more
586	C<IO::AIO::REQ> objects:	1765	C<IO::AIO::REQ> objects:
587		1766
588	$grp->add (aio_unlink "...");	1767	$grp->add (aio_unlink "...");
589		1768
590	add $grp aio_stat "...", sub { ... };	1769	add $grp aio_stat "...", sub {
		1770	$_[0] or return $grp->result ("error");
		1771
		1772	# add another request dynamically, if first succeeded
		1773	add $grp aio_open "...", sub {
		1774	$grp->result ("ok");
		1775	};
		1776	};
591		1777
592	This makes it very easy to create composite requests (see the source of	1778	This makes it very easy to create composite requests (see the source of
593	C<aio_move> for an application) that work and feel like simple requests.	1779	C<aio_move> for an application) that work and feel like simple requests.
594		1780
		1781	=over 4
		1782
595	The IO::AIO::GRP objects will be cleaned up during calls to	1783	=item * The IO::AIO::GRP objects will be cleaned up during calls to
596	C<IO::AIO::poll_cb>, just like any other request.	1784	C<IO::AIO::poll_cb>, just like any other request.
597		1785
598	They can be canceled like any other request. Canceling will cancel not	1786	=item * They can be canceled like any other request. Canceling will cancel not
599	just the request itself, but also all requests it contains.	1787	only the request itself, but also all requests it contains.
600		1788
601	They can also can also be added to other IO::AIO::GRP objects.	1789	=item * They can also can also be added to other IO::AIO::GRP objects.
		1790
		1791	=item * You must not add requests to a group from within the group callback (or
		1792	any later time).
		1793
		1794	=back
602		1795
603	Their lifetime, simplified, looks like this: when they are empty, they	1796	Their lifetime, simplified, looks like this: when they are empty, they
604	will finish very quickly. If they contain only requests that are in the	1797	will finish very quickly. If they contain only requests that are in the
605	C<done> state, they will also finish. Otherwise they will continue to	1798	C<done> state, they will also finish. Otherwise they will continue to
606	exist.	1799	exist.
607		1800
608	That means after creating a group you have some time to add requests. And	1801	That means after creating a group you have some time to add requests
609	in the callbacks of those requests, you can add further requests to the	1802	(precisely before the callback has been invoked, which is only done within
610	group. And only when all those requests have finished will the the group	1803	the C<poll_cb>). And in the callbacks of those requests, you can add
611	itself finish.	1804	further requests to the group. And only when all those requests have
		1805	finished will the the group itself finish.
612		1806
613	=over 4	1807	=over 4
614		1808
		1809	=item add $grp ...
		1810
615	=item $grp->add (...)	1811	=item $grp->add (...)
616
617	=item add $grp ...
618		1812
619	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can	1813	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
620	be added, including other groups, as long as you do not create circular	1814	be added, including other groups, as long as you do not create circular
621	dependencies.	1815	dependencies.
622		1816
623	Returns all its arguments.	1817	Returns all its arguments.
624		1818
		1819	=item $grp->cancel_subs
		1820
		1821	Cancel all subrequests and clears any feeder, but not the group request
		1822	itself. Useful when you queued a lot of events but got a result early.
		1823
		1824	The group request will finish normally (you cannot add requests to the
		1825	group).
		1826
		1827	=item $grp->result (...)
		1828
		1829	Set the result value(s) that will be passed to the group callback when all
		1830	subrequests have finished and set the groups errno to the current value
		1831	of errno (just like calling C<errno> without an error number). By default,
		1832	no argument will be passed and errno is zero.
		1833
		1834	=item $grp->errno ([$errno])
		1835
		1836	Sets the group errno value to C<$errno>, or the current value of errno
		1837	when the argument is missing.
		1838
		1839	Every aio request has an associated errno value that is restored when
		1840	the callback is invoked. This method lets you change this value from its
		1841	default (0).
		1842
		1843	Calling C<result> will also set errno, so make sure you either set C<$!>
		1844	before the call to C<result>, or call c<errno> after it.
		1845
		1846	=item feed $grp $callback->($grp)
		1847
		1848	Sets a feeder/generator on this group: every group can have an attached
		1849	generator that generates requests if idle. The idea behind this is that,
		1850	although you could just queue as many requests as you want in a group,
		1851	this might starve other requests for a potentially long time. For example,
		1852	C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
		1853	requests, delaying any later requests for a long time.
		1854
		1855	To avoid this, and allow incremental generation of requests, you can
		1856	instead a group and set a feeder on it that generates those requests. The
		1857	feed callback will be called whenever there are few enough (see C<limit>,
		1858	below) requests active in the group itself and is expected to queue more
		1859	requests.
		1860
		1861	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1862	not impose any limits).
		1863
		1864	If the feed does not queue more requests when called, it will be
		1865	automatically removed from the group.
		1866
		1867	If the feed limit is C<0> when this method is called, it will be set to
		1868	C<2> automatically.
		1869
		1870	Example:
		1871
		1872	# stat all files in @files, but only ever use four aio requests concurrently:
		1873
		1874	my $grp = aio_group sub { print "finished\n" };
		1875	limit $grp 4;
		1876	feed $grp sub {
		1877	my $file = pop @files
		1878	or return;
		1879
		1880	add $grp aio_stat $file, sub { ... };
		1881	};
		1882
		1883	=item limit $grp $num
		1884
		1885	Sets the feeder limit for the group: The feeder will be called whenever
		1886	the group contains less than this many requests.
		1887
		1888	Setting the limit to C<0> will pause the feeding process.
		1889
		1890	The default value for the limit is C<0>, but note that setting a feeder
		1891	automatically bumps it up to C<2>.
		1892
625	=back	1893	=back
626		1894
		1895
627	=head2 SUPPORT FUNCTIONS	1896	=head2 SUPPORT FUNCTIONS
628		1897
		1898	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1899
629	=over 4	1900	=over 4
630		1901
631	=item $fileno = IO::AIO::poll_fileno	1902	=item $fileno = IO::AIO::poll_fileno
632		1903
633	Return the I<request result pipe file descriptor>. This filehandle must be	1904	Return the I<request result pipe file descriptor>. This filehandle must be
634	polled for reading by some mechanism outside this module (e.g. Event or	1905	polled for reading by some mechanism outside this module (e.g. EV, Glib,
635	select, see below or the SYNOPSIS). If the pipe becomes readable you have	1906	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
636	to call C<poll_cb> to check the results.	1907	you have to call C<poll_cb> to check the results.
637		1908
638	See C<poll_cb> for an example.	1909	See C<poll_cb> for an example.
639		1910
640	=item IO::AIO::poll_cb	1911	=item IO::AIO::poll_cb
641		1912
642	Process all outstanding events on the result pipe. You have to call this	1913	Process some requests that have reached the result phase (i.e. they have
643	regularly. Returns the number of events processed. Returns immediately	1914	been executed but the results are not yet reported). You have to call
644	when no events are outstanding.	1915	this "regularly" to finish outstanding requests.
		1916
		1917	Returns C<0> if all events could be processed (or there were no
		1918	events to process), or C<-1> if it returned earlier for whatever
		1919	reason. Returns immediately when no events are outstanding. The amount
		1920	of events processed depends on the settings of C<IO::AIO::max_poll_req>,
		1921	C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>.
		1922
		1923	If not all requests were processed for whatever reason, the poll file
		1924	descriptor will still be ready when C<poll_cb> returns, so normally you
		1925	don't have to do anything special to have it called later.
		1926
		1927	Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
		1928	ready, it can be beneficial to call this function from loops which submit
		1929	a lot of requests, to make sure the results get processed when they become
		1930	available and not just when the loop is finished and the event loop takes
		1931	over again. This function returns very fast when there are no outstanding
		1932	requests.
645		1933
646	Example: Install an Event watcher that automatically calls	1934	Example: Install an Event watcher that automatically calls
647	IO::AIO::poll_cb with high priority:	1935	IO::AIO::poll_cb with high priority (more examples can be found in the
		1936	SYNOPSIS section, at the top of this document):
648		1937
649	Event->io (fd => IO::AIO::poll_fileno,	1938	Event->io (fd => IO::AIO::poll_fileno,
650	poll => 'r', async => 1,	1939	poll => 'r', async => 1,
651	cb => \&IO::AIO::poll_cb);	1940	cb => \&IO::AIO::poll_cb);
652		1941
653	=item IO::AIO::poll_wait	1942	=item IO::AIO::poll_wait
654		1943
655	Wait till the result filehandle becomes ready for reading (simply does a	1944	Wait until either at least one request is in the result phase or no
656	C<select> on the filehandle. This is useful if you want to synchronously wait	1945	requests are outstanding anymore.
657	for some requests to finish).	1946
		1947	This is useful if you want to synchronously wait for some requests to
		1948	become ready, without actually handling them.
658		1949
659	See C<nreqs> for an example.	1950	See C<nreqs> for an example.
660		1951
		1952	=item IO::AIO::poll
		1953
		1954	Waits until some requests have been handled.
		1955
		1956	Returns the number of requests processed, but is otherwise strictly
		1957	equivalent to:
		1958
		1959	IO::AIO::poll_wait, IO::AIO::poll_cb
		1960
661	=item IO::AIO::nreqs	1961	=item IO::AIO::flush
662		1962
663	Returns the number of requests currently outstanding (i.e. for which their	1963	Wait till all outstanding AIO requests have been handled.
664	callback has not been invoked yet).
665		1964
666	Example: wait till there are no outstanding requests anymore:	1965	Strictly equivalent to:
667		1966
668	IO::AIO::poll_wait, IO::AIO::poll_cb	1967	IO::AIO::poll_wait, IO::AIO::poll_cb
669	while IO::AIO::nreqs;	1968	while IO::AIO::nreqs;
670		1969
		1970	This function can be useful at program aborts, to make sure outstanding
		1971	I/O has been done (C<IO::AIO> uses an C<END> block which already calls
		1972	this function on normal exits), or when you are merely using C<IO::AIO>
		1973	for its more advanced functions, rather than for async I/O, e.g.:
		1974
		1975	my ($dirs, $nondirs);
		1976	IO::AIO::aio_scandir "/tmp", 0, sub { ($dirs, $nondirs) = @_ };
671	=item IO::AIO::flush	1977	IO::AIO::flush;
		1978	# $dirs, $nondirs are now set
672		1979
673	Wait till all outstanding AIO requests have been handled.	1980	=item IO::AIO::max_poll_reqs $nreqs
674		1981
675	Strictly equivalent to:	1982	=item IO::AIO::max_poll_time $seconds
676		1983
677	IO::AIO::poll_wait, IO::AIO::poll_cb	1984	These set the maximum number of requests (default C<0>, meaning infinity)
678	while IO::AIO::nreqs;	1985	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1986	the maximum amount of time (default C<0>, meaning infinity) spent in
		1987	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1988	of time C<poll_cb> is allowed to use).
679		1989
680	=item IO::AIO::poll	1990	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1991	syscall per request processed, which is not normally a problem unless your
		1992	callbacks are really really fast or your OS is really really slow (I am
		1993	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
681		1994
682	Waits until some requests have been handled.	1995	Setting these is useful if you want to ensure some level of
		1996	interactiveness when perl is not fast enough to process all requests in
		1997	time.
683		1998
684	Strictly equivalent to:	1999	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
685		2000
686	IO::AIO::poll_wait, IO::AIO::poll_cb	2001	Example: Install an Event watcher that automatically calls
687	if IO::AIO::nreqs;	2002	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		2003	program get the CPU sometimes even under high AIO load.
		2004
		2005	# try not to spend much more than 0.1s in poll_cb
		2006	IO::AIO::max_poll_time 0.1;
		2007
		2008	# use a low priority so other tasks have priority
		2009	Event->io (fd => IO::AIO::poll_fileno,
		2010	poll => 'r', nice => 1,
		2011	cb => &IO::AIO::poll_cb);
		2012
		2013	=back
		2014
		2015
		2016	=head3 CONTROLLING THE NUMBER OF THREADS
		2017
		2018	=over
688		2019
689	=item IO::AIO::min_parallel $nthreads	2020	=item IO::AIO::min_parallel $nthreads
690		2021
691	Set the minimum number of AIO threads to C<$nthreads>. The current default	2022	Set the minimum number of AIO threads to C<$nthreads>. The current
692	is C<4>, which means four asynchronous operations can be done at one time	2023	default is C<8>, which means eight asynchronous operations can execute
693	(the number of outstanding operations, however, is unlimited).	2024	concurrently at any one time (the number of outstanding requests,
		2025	however, is unlimited).
694		2026
695	IO::AIO starts threads only on demand, when an AIO request is queued and	2027	IO::AIO starts threads only on demand, when an AIO request is queued and
696	no free thread exists.	2028	no free thread exists. Please note that queueing up a hundred requests can
		2029	create demand for a hundred threads, even if it turns out that everything
		2030	is in the cache and could have been processed faster by a single thread.
697		2031
698	It is recommended to keep the number of threads low, as some Linux	2032	It is recommended to keep the number of threads relatively low, as some
699	kernel versions will scale negatively with the number of threads (higher	2033	Linux kernel versions will scale negatively with the number of threads
700	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	2034	(higher parallelity => MUCH higher latency). With current Linux 2.6
701	threads should be fine.	2035	versions, 4-32 threads should be fine.
702		2036
703	Under most circumstances you don't need to call this function, as the	2037	Under most circumstances you don't need to call this function, as the
704	module selects a default that is suitable for low to moderate load.	2038	module selects a default that is suitable for low to moderate load.
705		2039
706	=item IO::AIO::max_parallel $nthreads	2040	=item IO::AIO::max_parallel $nthreads
…		…
715	This module automatically runs C<max_parallel 0> at program end, to ensure	2049	This module automatically runs C<max_parallel 0> at program end, to ensure
716	that all threads are killed and that there are no outstanding requests.	2050	that all threads are killed and that there are no outstanding requests.
717		2051
718	Under normal circumstances you don't need to call this function.	2052	Under normal circumstances you don't need to call this function.
719		2053
		2054	=item IO::AIO::max_idle $nthreads
		2055
		2056	Limit the number of threads (default: 4) that are allowed to idle
		2057	(i.e., threads that did not get a request to process within the idle
		2058	timeout (default: 10 seconds). That means if a thread becomes idle while
		2059	C<$nthreads> other threads are also idle, it will free its resources and
		2060	exit.
		2061
		2062	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		2063	to allow for extremely high load situations, but want to free resources
		2064	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		2065
		2066	The default is probably ok in most situations, especially if thread
		2067	creation is fast. If thread creation is very slow on your system you might
		2068	want to use larger values.
		2069
		2070	=item IO::AIO::idle_timeout $seconds
		2071
		2072	Sets the minimum idle timeout (default 10) after which worker threads are
		2073	allowed to exit. SEe C<IO::AIO::max_idle>.
		2074
720	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	2075	=item IO::AIO::max_outstanding $maxreqs
721		2076
722	Sets the maximum number of outstanding requests to C<$nreqs>. If you	2077	Sets the maximum number of outstanding requests to C<$nreqs>. If
723	try to queue up more than this number of requests, the caller will block until	2078	you do queue up more than this number of requests, the next call to
724	some requests have been handled.	2079	C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
		2080	C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
		2081	longer exceeded.
725		2082
726	The default is very large, so normally there is no practical limit. If you	2083	In other words, this setting does not enforce a queue limit, but can be
727	queue up many requests in a loop it often improves speed if you set	2084	used to make poll functions block if the limit is exceeded.
728	this to a relatively low number, such as C<100>.
729		2085
730	Under normal circumstances you don't need to call this function.	2086	This is a bad function to use in interactive programs because it blocks,
		2087	and a bad way to reduce concurrency because it is inexact. If you need to
		2088	issue many requests without being able to call a poll function on demand,
		2089	it is better to use an C<aio_group> together with a feed callback.
		2090
		2091	Its main use is in scripts without an event loop - when you want to stat a
		2092	lot of files, you can write something like this:
		2093
		2094	IO::AIO::max_outstanding 32;
		2095
		2096	for my $path (...) {
		2097	aio_stat $path , ...;
		2098	IO::AIO::poll_cb;
		2099	}
		2100
		2101	IO::AIO::flush;
		2102
		2103	The call to C<poll_cb> inside the loop will normally return instantly,
		2104	allowing the loop to progress, but as soon as more than C<32> requests
		2105	are in-flight, it will block until some requests have been handled. This
		2106	keeps the loop from pushing a large number of C<aio_stat> requests onto
		2107	the queue (which, with many paths to stat, can use up a lot of memory).
		2108
		2109	The default value for C<max_outstanding> is very large, so there is no
		2110	practical limit on the number of outstanding requests.
731		2111
732	=back	2112	=back
733		2113
		2114
		2115	=head3 STATISTICAL INFORMATION
		2116
		2117	=over
		2118
		2119	=item IO::AIO::nreqs
		2120
		2121	Returns the number of requests currently in the ready, execute or pending
		2122	states (i.e. for which their callback has not been invoked yet).
		2123
		2124	Example: wait till there are no outstanding requests anymore:
		2125
		2126	IO::AIO::poll_wait, IO::AIO::poll_cb
		2127	while IO::AIO::nreqs;
		2128
		2129	=item IO::AIO::nready
		2130
		2131	Returns the number of requests currently in the ready state (not yet
		2132	executed).
		2133
		2134	=item IO::AIO::npending
		2135
		2136	Returns the number of requests currently in the pending state (executed,
		2137	but not yet processed by poll_cb).
		2138
		2139	=back
		2140
		2141
		2142	=head3 SUBSECOND STAT TIME ACCESS
		2143
		2144	Both C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> functions can
		2145	generally find access/modification and change times with subsecond time
		2146	accuracy of the system supports it, but perl's built-in functions only
		2147	return the integer part.
		2148
		2149	The following functions return the timestamps of the most recent
		2150	stat with subsecond precision on most systems and work both after
		2151	C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> calls. Their return
		2152	value is only meaningful after a successful C<stat>/C<lstat> call, or
		2153	during/after a successful C<aio_stat>/C<aio_lstat> callback.
		2154
		2155	This is similar to the L<Time::HiRes> C<stat> functions, but can return
		2156	full resolution without rounding and work with standard perl C<stat>,
		2157	alleviating the need to call the special C<Time::HiRes> functions, which
		2158	do not act like their perl counterparts.
		2159
		2160	On operating systems or file systems where subsecond time resolution is
		2161	not supported or could not be detected, a fractional part of C<0> is
		2162	returned, so it is always safe to call these functions.
		2163
		2164	=over 4
		2165
		2166	=item $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime
		2167
		2168	Return the access, modication, change or birth time, respectively,
		2169	including fractional part. Due to the limited precision of floating point,
		2170	the accuracy on most platforms is only a bit better than milliseconds
		2171	for times around now - see the I<nsec> function family, below, for full
		2172	accuracy.
		2173
		2174	File birth time is only available when the OS and perl support it (on
		2175	FreeBSD and NetBSD at the time of this writing, although support is
		2176	adaptive, so if your OS/perl gains support, IO::AIO can take advantage of
		2177	it). On systems where it isn't available, C<0> is currently returned, but
		2178	this might change to C<undef> in a future version.
		2179
		2180	=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
		2181
		2182	Returns access, modification, change and birth time all in one go, and
		2183	maybe more times in the future version.
		2184
		2185	=item $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec
		2186
		2187	Return the fractional access, modifcation, change or birth time, in nanoseconds,
		2188	as an integer in the range C<0> to C<999999999>.
		2189
		2190	Note that no accessors are provided for access, modification and
		2191	change times - you need to get those from C<stat _> if required (C<int
		2192	IO::AIO::st_atime> and so on will I<not> generally give you the correct
		2193	value).
		2194
		2195	=item $seconds = IO::AIO::st_btimesec
		2196
		2197	The (integral) seconds part of the file birth time, if available.
		2198
		2199	=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
		2200
		2201	Like the functions above, but returns all four times in one go (and maybe
		2202	more in future versions).
		2203
		2204	=item $counter = IO::AIO::st_gen
		2205
		2206	Returns the generation counter (in practice this is just a random number)
		2207	of the file. This is only available on platforms which have this member in
		2208	their C<struct stat> (most BSDs at the time of this writing) and generally
		2209	only to the root usert. If unsupported, C<0> is returned, but this might
		2210	change to C<undef> in a future version.
		2211
		2212	=back
		2213
		2214	Example: print the high resolution modification time of F</etc>, using
		2215	C<stat>, and C<IO::AIO::aio_stat>.
		2216
		2217	if (stat "/etc") {
		2218	printf "stat(/etc) mtime: %f\n", IO::AIO::st_mtime;
		2219	}
		2220
		2221	IO::AIO::aio_stat "/etc", sub {
		2222	$_[0]
		2223	and return;
		2224
		2225	printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::st_mtimensec;
		2226	};
		2227
		2228	IO::AIO::flush;
		2229
		2230	Output of the awbove on my system, showing reduced and full accuracy:
		2231
		2232	stat(/etc) mtime: 1534043702.020808
		2233	aio_stat(/etc) mtime: 1534043702.020807792
		2234
		2235
		2236	=head3 MISCELLANEOUS FUNCTIONS
		2237
		2238	IO::AIO implements some functions that are useful when you want to use
		2239	some "Advanced I/O" function not available to in Perl, without going the
		2240	"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*>
		2241	counterpart.
		2242
		2243	=over 4
		2244
		2245	=item $retval = IO::AIO::fexecve $fh, $argv, $envp
		2246
		2247	A more-or-less direct equivalent to the POSIX C<fexecve> functions, which
		2248	allows you to specify the program to be executed via a file descriptor (or
		2249	handle). Returns C<-1> and sets errno to C<ENOSYS> if not available.
		2250
		2251	=item $retval = IO::AIO::mount $special, $path, $fstype, $flags = 0, $data = undef
		2252
		2253	Calls the GNU/Linux mount syscall with the given arguments. All except
		2254	C<$flags> are strings, and if C<$data> is C<undef>, a C<NULL> will be
		2255	passed.
		2256
		2257	The following values for C<$flags> are available:
		2258
		2259	C<IO::AIO::MS_RDONLY>, C<IO::AIO::MS_NOSUID>, C<IO::AIO::MS_NODEV>, C<IO::AIO::MS_NOEXEC>, C<IO::AIO::MS_SYNCHRONOUS>,
		2260	C<IO::AIO::MS_REMOUNT>, C<IO::AIO::MS_MANDLOCK>, C<IO::AIO::MS_DIRSYNC>, C<IO::AIO::MS_NOATIME>,
		2261	C<IO::AIO::MS_NODIRATIME>, C<IO::AIO::MS_BIND>, C<IO::AIO::MS_MOVE>, C<IO::AIO::MS_REC>, C<IO::AIO::MS_SILENT>,
		2262	C<IO::AIO::MS_POSIXACL>, C<IO::AIO::MS_UNBINDABLE>, C<IO::AIO::MS_PRIVATE>, C<IO::AIO::MS_SLAVE>, C<IO::AIO::MS_SHARED>,
		2263	C<IO::AIO::MS_RELATIME>, C<IO::AIO::MS_KERNMOUNT>, C<IO::AIO::MS_I_VERSION>, C<IO::AIO::MS_STRICTATIME>,
		2264	C<IO::AIO::MS_LAZYTIME>, C<IO::AIO::MS_ACTIVE>, C<IO::AIO::MS_NOUSER>, C<IO::AIO::MS_RMT_MASK>, C<IO::AIO::MS_MGC_VAL> and
		2265	C<IO::AIO::MS_MGC_MSK>.
		2266
		2267	=item $retval = IO::AIO::umount $path, $flags = 0
		2268
		2269	Invokes the GNU/Linux C<umount> or C<umount2> syscalls. Always calls
		2270	C<umount> if C<$flags> is C<0>, otherwqise always tries to call
		2271	C<umount2>.
		2272
		2273	The following C<$flags> are available:
		2274
		2275	C<IO::AIO::MNT_FORCE>, C<IO::AIO::MNT_DETACH>, C<IO::AIO::MNT_EXPIRE> and C<IO::AIO::UMOUNT_NOFOLLOW>.
		2276
		2277	=item $numfd = IO::AIO::get_fdlimit
		2278
		2279	Tries to find the current file descriptor limit and returns it, or
		2280	C<undef> and sets C<$!> in case of an error. The limit is one larger than
		2281	the highest valid file descriptor number.
		2282
		2283	=item IO::AIO::min_fdlimit [$numfd]
		2284
		2285	Try to increase the current file descriptor limit(s) to at least C<$numfd>
		2286	by changing the soft or hard file descriptor resource limit. If C<$numfd>
		2287	is missing, it will try to set a very high limit, although this is not
		2288	recommended when you know the actual minimum that you require.
		2289
		2290	If the limit cannot be raised enough, the function makes a best-effort
		2291	attempt to increase the limit as much as possible, using various
		2292	tricks, while still failing. You can query the resulting limit using
		2293	C<IO::AIO::get_fdlimit>.
		2294
		2295	If an error occurs, returns C<undef> and sets C<$!>, otherwise returns
		2296	true.
		2297
		2298	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		2299
		2300	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		2301	but is blocking (this makes most sense if you know the input data is
		2302	likely cached already and the output filehandle is set to non-blocking
		2303	operations).
		2304
		2305	Returns the number of bytes copied, or C<-1> on error.
		2306
		2307	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		2308
		2309	Simply calls the C<posix_fadvise> function (see its
		2310	manpage for details). The following advice constants are
		2311	available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		2312	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		2313	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		2314
		2315	On systems that do not implement C<posix_fadvise>, this function returns
		2316	ENOSYS, otherwise the return value of C<posix_fadvise>.
		2317
		2318	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		2319
		2320	Simply calls the C<posix_madvise> function (see its
		2321	manpage for details). The following advice constants are
		2322	available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		2323	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>,
		2324	C<IO::AIO::MADV_DONTNEED>.
		2325
		2326	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2327	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2328	will be reduced to fit into the C<$scalar>.
		2329
		2330	On systems that do not implement C<posix_madvise>, this function returns
		2331	ENOSYS, otherwise the return value of C<posix_madvise>.
		2332
		2333	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		2334
		2335	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		2336	$scalar (see its manpage for details). The following protect
		2337	constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		2338	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		2339
		2340	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2341	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2342	will be reduced to fit into the C<$scalar>.
		2343
		2344	On systems that do not implement C<mprotect>, this function returns
		2345	ENOSYS, otherwise the return value of C<mprotect>.
		2346
		2347	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		2348
		2349	Memory-maps a file (or anonymous memory range) and attaches it to the
		2350	given C<$scalar>, which will act like a string scalar. Returns true on
		2351	success, and false otherwise.
		2352
		2353	The scalar must exist, but its contents do not matter - this means you
		2354	cannot use a nonexistant array or hash element. When in doubt, C<undef>
		2355	the scalar first.
		2356
		2357	The only operations allowed on the mmapped scalar are C<substr>/C<vec>,
		2358	which don't change the string length, and most read-only operations such
		2359	as copying it or searching it with regexes and so on.
		2360
		2361	Anything else is unsafe and will, at best, result in memory leaks.
		2362
		2363	The memory map associated with the C<$scalar> is automatically removed
		2364	when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap>
		2365	or C<IO::AIO::munmap> functions are called on it.
		2366
		2367	This calls the C<mmap>(2) function internally. See your system's manual
		2368	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		2369
		2370	The C<$length> must be larger than zero and smaller than the actual
		2371	filesize.
		2372
		2373	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		2374	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		2375
		2376	C<$flags> can be a combination of
		2377	C<IO::AIO::MAP_SHARED> or
		2378	C<IO::AIO::MAP_PRIVATE>,
		2379	or a number of system-specific flags (when not available, the are C<0>):
		2380	C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant),
		2381	C<IO::AIO::MAP_LOCKED>,
		2382	C<IO::AIO::MAP_NORESERVE>,
		2383	C<IO::AIO::MAP_POPULATE>,
		2384	C<IO::AIO::MAP_NONBLOCK>,
		2385	C<IO::AIO::MAP_FIXED>,
		2386	C<IO::AIO::MAP_GROWSDOWN>,
		2387	C<IO::AIO::MAP_32BIT>,
		2388	C<IO::AIO::MAP_HUGETLB>,
		2389	C<IO::AIO::MAP_STACK>,
		2390	C<IO::AIO::MAP_FIXED_NOREPLACE>,
		2391	C<IO::AIO::MAP_SHARED_VALIDATE>,
		2392	C<IO::AIO::MAP_SYNC> or
		2393	C<IO::AIO::MAP_UNINITIALIZED>.
		2394
		2395	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		2396
		2397	C<$offset> is the offset from the start of the file - it generally must be
		2398	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		2399
		2400	Example:
		2401
		2402	use Digest::MD5;
		2403	use IO::AIO;
		2404
		2405	open my $fh, "<verybigfile"
		2406	or die "$!";
		2407
		2408	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		2409	or die "verybigfile: $!";
		2410
		2411	my $fast_md5 = md5 $data;
		2412
		2413	=item IO::AIO::munmap $scalar
		2414
		2415	Removes a previous mmap and undefines the C<$scalar>.
		2416
		2417	=item IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[, $new_address = 0]
		2418
		2419	Calls the Linux-specific mremap(2) system call. The C<$scalar> must have
		2420	been mapped by C<IO::AIO::mmap>, and C<$flags> must currently either be
		2421	C<0> or C<IO::AIO::MREMAP_MAYMOVE>.
		2422
		2423	Returns true if successful, and false otherwise. If the underlying mmapped
		2424	region has changed address, then the true value has the numerical value
		2425	C<1>, otherwise it has the numerical value C<0>:
		2426
		2427	my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE
		2428	or die "mremap: $!";
		2429
		2430	if ($success*1) {
		2431	warn "scalar has chanegd address in memory\n";
		2432	}
		2433
		2434	C<IO::AIO::MREMAP_FIXED> and the C<$new_address> argument are currently
		2435	implemented, but not supported and might go away in a future version.
		2436
		2437	On systems where this call is not supported or is not emulated, this call
		2438	returns falls and sets C<$!> to C<ENOSYS>.
		2439
		2440	=item IO::AIO::mlockall $flags
		2441
		2442	Calls the C<eio_mlockall_sync> function, which is like C<aio_mlockall>,
		2443	but is blocking.
		2444
		2445	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		2446
		2447	Calls the C<munlock> function, undoing the effects of a previous
		2448	C<aio_mlock> call (see its description for details).
		2449
		2450	=item IO::AIO::munlockall
		2451
		2452	Calls the C<munlockall> function.
		2453
		2454	On systems that do not implement C<munlockall>, this function returns
		2455	ENOSYS, otherwise the return value of C<munlockall>.
		2456
		2457	=item $fh = IO::AIO::accept4 $r_fh, $sockaddr, $sockaddr_maxlen, $flags
		2458
		2459	Uses the GNU/Linux C<accept4(2)> syscall, if available, to accept a socket
		2460	and return the new file handle on success, or sets C<$!> and returns
		2461	C<undef> on error.
		2462
		2463	The remote name of the new socket will be stored in C<$sockaddr>, which
		2464	will be extended to allow for at least C<$sockaddr_maxlen> octets. If the
		2465	socket name does not fit into C<$sockaddr_maxlen> octets, this is signaled
		2466	by returning a longer string in C<$sockaddr>, which might or might not be
		2467	truncated.
		2468
		2469	To accept name-less sockets, use C<undef> for C<$sockaddr> and C<0> for
		2470	C<$sockaddr_maxlen>.
		2471
		2472	The main reasons to use this syscall rather than portable C<accept(2)>
		2473	are that you can specify C<SOCK_NONBLOCK> and/or C<SOCK_CLOEXEC>
		2474	flags and you can accept name-less sockets by specifying C<0> for
		2475	C<$sockaddr_maxlen>, which is sadly not possible with perl's interface to
		2476	C<accept>.
		2477
		2478	=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
		2479
		2480	Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or
		2481	C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they
		2482	should be the file offset.
		2483
		2484	C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might
		2485	silently corrupt the data in this case.
		2486
		2487	The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>,
		2488	C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and
		2489	C<IO::AIO::SPLICE_F_GIFT>.
		2490
		2491	See the C<splice(2)> manpage for details.
		2492
		2493	=item IO::AIO::tee $r_fh, $w_fh, $length, $flags
		2494
		2495	Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the
		2496	description for C<IO::AIO::splice> above for details.
		2497
		2498	=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
		2499
		2500	Attempts to query or change the pipe buffer size. Obviously works only
		2501	on pipes, and currently works only on GNU/Linux systems, and fails with
		2502	C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer
		2503	size on other systems, drop me a note.
		2504
		2505	=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
		2506
		2507	This is a direct interface to the Linux L<pipe2(2)> system call. If
		2508	C<$flags> is missing or C<0>, then this should be the same as a call to
		2509	perl's built-in C<pipe> function and create a new pipe, and works on
		2510	systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe
		2511	(..., 4096, O_BINARY)>.
		2512
		2513	If C<$flags> is non-zero, it tries to invoke the pipe2 system call with
		2514	the given flags (Linux 2.6.27, glibc 2.9).
		2515
		2516	On success, the read and write file handles are returned.
		2517
		2518	On error, nothing will be returned. If the pipe2 syscall is missing and
		2519	C<$flags> is non-zero, fails with C<ENOSYS>.
		2520
		2521	Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the
		2522	time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and
		2523	C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported.
		2524
		2525	Example: create a pipe race-free w.r.t. threads and fork:
		2526
		2527	my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
		2528	or die "pipe2: $!\n";
		2529
		2530	=item $fh = IO::AIO::memfd_create $pathname[, $flags]
		2531
		2532	This is a direct interface to the Linux L<memfd_create(2)> system
		2533	call. The (unhelpful) default for C<$flags> is C<0>, but your default
		2534	should be C<IO::AIO::MFD_CLOEXEC>.
		2535
		2536	On success, the new memfd filehandle is returned, otherwise returns
		2537	C<undef>. If the memfd_create syscall is missing, fails with C<ENOSYS>.
		2538
		2539	Please refer to L<memfd_create(2)> for more info on this call.
		2540
		2541	The following C<$flags> values are available: C<IO::AIO::MFD_CLOEXEC>,
		2542	C<IO::AIO::MFD_ALLOW_SEALING>, C<IO::AIO::MFD_HUGETLB>,
		2543	C<IO::AIO::MFD_HUGETLB_2MB> and C<IO::AIO::MFD_HUGETLB_1GB>.
		2544
		2545	Example: create a new memfd.
		2546
		2547	my $fh = IO::AIO::memfd_create "somenameforprocfd", IO::AIO::MFD_CLOEXEC
		2548	or die "memfd_create: $!\n";
		2549
		2550	=item $fh = IO::AIO::pidfd_open $pid[, $flags]
		2551
		2552	This is an interface to the Linux L<pidfd_open(2)> system call. The
		2553	default for C<$flags> is C<0>.
		2554
		2555	On success, a new pidfd filehandle is returned (that is already set to
		2556	close-on-exec), otherwise returns C<undef>. If the syscall is missing,
		2557	fails with C<ENOSYS>.
		2558
		2559	Example: open pid 6341 as pidfd.
		2560
		2561	my $fh = IO::AIO::pidfd_open 6341
		2562	or die "pidfd_open: $!\n";
		2563
		2564	=item $status = IO::AIO::pidfd_send_signal $pidfh, $signal[, $siginfo[, $flags]]
		2565
		2566	This is an interface to the Linux L<pidfd_send_signal> system call. The
		2567	default for C<$siginfo> is C<undef> and the default for C<$flags> is C<0>.
		2568
		2569	Returns the system call status. If the syscall is missing, fails with
		2570	C<ENOSYS>.
		2571
		2572	When specified, C<$siginfo> must be a reference to a hash with one or more
		2573	of the following members:
		2574
		2575	=over
		2576
		2577	=item code - the C<si_code> member
		2578
		2579	=item pid - the C<si_pid> member
		2580
		2581	=item uid - the C<si_uid> member
		2582
		2583	=item value_int - the C<si_value.sival_int> member
		2584
		2585	=item value_ptr - the C<si_value.sival_ptr> member, specified as an integer
		2586
		2587	=back
		2588
		2589	Example: send a SIGKILL to the specified process.
		2590
		2591	my $status = IO::AIO::pidfd_send_signal $pidfh, 9, undef
		2592	and die "pidfd_send_signal: $!\n";
		2593
		2594	Example: send a SIGKILL to the specified process with extra data.
		2595
		2596	my $status = IO::AIO::pidfd_send_signal $pidfh, 9, { code => -1, value_int => 7 }
		2597	and die "pidfd_send_signal: $!\n";
		2598
		2599	=item $fh = IO::AIO::pidfd_getfd $pidfh, $targetfd[, $flags]
		2600
		2601	This is an interface to the Linux L<pidfd_getfd> system call. The default
		2602	for C<$flags> is C<0>.
		2603
		2604	On success, returns a dup'ed copy of the target file descriptor (specified
		2605	as an integer) returned (that is already set to close-on-exec), otherwise
		2606	returns C<undef>. If the syscall is missing, fails with C<ENOSYS>.
		2607
		2608	Example: get a copy of standard error of another process and print soemthing to it.
		2609
		2610	my $errfh = IO::AIO::pidfd_getfd $pidfh, 2
		2611	or die "pidfd_getfd: $!\n";
		2612	print $errfh "stderr\n";
		2613
		2614	=item $fh = IO::AIO::eventfd [$initval, [$flags]]
		2615
		2616	This is a direct interface to the Linux L<eventfd(2)> system call. The
		2617	(unhelpful) defaults for C<$initval> and C<$flags> are C<0> for both.
		2618
		2619	On success, the new eventfd filehandle is returned, otherwise returns
		2620	C<undef>. If the eventfd syscall is missing, fails with C<ENOSYS>.
		2621
		2622	Please refer to L<eventfd(2)> for more info on this call.
		2623
		2624	The following symbol flag values are available: C<IO::AIO::EFD_CLOEXEC>,
		2625	C<IO::AIO::EFD_NONBLOCK> and C<IO::AIO::EFD_SEMAPHORE> (Linux 2.6.30).
		2626
		2627	Example: create a new eventfd filehandle:
		2628
		2629	$fh = IO::AIO::eventfd 0, IO::AIO::EFD_CLOEXEC
		2630	or die "eventfd: $!\n";
		2631
		2632	=item $fh = IO::AIO::timerfd_create $clockid[, $flags]
		2633
		2634	This is a direct interface to the Linux L<timerfd_create(2)> system
		2635	call. The (unhelpful) default for C<$flags> is C<0>, but your default
		2636	should be C<IO::AIO::TFD_CLOEXEC>.
		2637
		2638	On success, the new timerfd filehandle is returned, otherwise returns
		2639	C<undef>. If the timerfd_create syscall is missing, fails with C<ENOSYS>.
		2640
		2641	Please refer to L<timerfd_create(2)> for more info on this call.
		2642
		2643	The following C<$clockid> values are
		2644	available: C<IO::AIO::CLOCK_REALTIME>, C<IO::AIO::CLOCK_MONOTONIC>
		2645	C<IO::AIO::CLOCK_CLOCK_BOOTTIME> (Linux 3.15)
		2646	C<IO::AIO::CLOCK_CLOCK_REALTIME_ALARM> (Linux 3.11) and
		2647	C<IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM> (Linux 3.11).
		2648
		2649	The following C<$flags> values are available (Linux
		2650	2.6.27): C<IO::AIO::TFD_NONBLOCK> and C<IO::AIO::TFD_CLOEXEC>.
		2651
		2652	Example: create a new timerfd and set it to one-second repeated alarms,
		2653	then wait for two alarms:
		2654
		2655	my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC
		2656	or die "timerfd_create: $!\n";
		2657
		2658	defined IO::AIO::timerfd_settime $fh, 0, 1, 1
		2659	or die "timerfd_settime: $!\n";
		2660
		2661	for (1..2) {
		2662	8 == sysread $fh, my $buf, 8
		2663	or die "timerfd read failure\n";
		2664
		2665	printf "number of expirations (likely 1): %d\n",
		2666	unpack "Q", $buf;
		2667	}
		2668
		2669	=item ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
		2670
		2671	This is a direct interface to the Linux L<timerfd_settime(2)> system
		2672	call. Please refer to its manpage for more info on this call.
		2673
		2674	The new itimerspec is specified using two (possibly fractional) second
		2675	values, C<$new_interval> and C<$new_value>).
		2676
		2677	On success, the current interval and value are returned (as per
		2678	C<timerfd_gettime>). On failure, the empty list is returned.
		2679
		2680	The following C<$flags> values are
		2681	available: C<IO::AIO::TFD_TIMER_ABSTIME> and
		2682	C<IO::AIO::TFD_TIMER_CANCEL_ON_SET>.
		2683
		2684	See C<IO::AIO::timerfd_create> for a full example.
		2685
		2686	=item ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
		2687
		2688	This is a direct interface to the Linux L<timerfd_gettime(2)> system
		2689	call. Please refer to its manpage for more info on this call.
		2690
		2691	On success, returns the current values of interval and value for the given
		2692	timerfd (as potentially fractional second values). On failure, the empty
		2693	list is returned.
		2694
		2695	=back
		2696
734	=cut	2697	=cut
735		2698
736	# support function to convert a fd into a perl filehandle
737	sub _fd2fh {
738	return undef if $_[0] < 0;
739
740	# try to generate nice filehandles
741	my $sym = "IO::AIO::fd#$_[0]";
742	local *$sym;
743
744	open *$sym, "+<&=$_[0]" # usually works under any unix
745	or open *$sym, "<&=$_[0]" # cygwin needs this
746	or open *$sym, ">&=$_[0]" # or this
747	or return undef;
748
749	*$sym
750	}
751
752	min_parallel 4;	2699	min_parallel 8;
753		2700
754	END {	2701	END { flush }
755	max_parallel 0;
756	}
757		2702
758	1;	2703	1;
759		2704
		2705	=head1 EVENT LOOP INTEGRATION
		2706
		2707	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		2708	automatically into many event loops:
		2709
		2710	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		2711	use AnyEvent::AIO;
		2712
		2713	You can also integrate IO::AIO manually into many event loops, here are
		2714	some examples of how to do this:
		2715
		2716	# EV integration
		2717	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		2718
		2719	# Event integration
		2720	Event->io (fd => IO::AIO::poll_fileno,
		2721	poll => 'r',
		2722	cb => \&IO::AIO::poll_cb);
		2723
		2724	# Glib/Gtk2 integration
		2725	add_watch Glib::IO IO::AIO::poll_fileno,
		2726	in => sub { IO::AIO::poll_cb; 1 };
		2727
		2728	# Tk integration
		2729	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		2730	readable => \&IO::AIO::poll_cb);
		2731
		2732	# Danga::Socket integration
		2733	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		2734	\&IO::AIO::poll_cb);
		2735
760	=head2 FORK BEHAVIOUR	2736	=head2 FORK BEHAVIOUR
761		2737
762	This module should do "the right thing" when the process using it forks:	2738	Usage of pthreads in a program changes the semantics of fork
		2739	considerably. Specifically, only async-safe functions can be called after
		2740	fork. Perl doesn't know about this, so in general, you cannot call fork
		2741	with defined behaviour in perl if pthreads are involved. IO::AIO uses
		2742	pthreads, so this applies, but many other extensions and (for inexplicable
		2743	reasons) perl itself often is linked against pthreads, so this limitation
		2744	applies to quite a lot of perls.
763		2745
764	Before the fork, IO::AIO enters a quiescent state where no requests	2746	This module no longer tries to fight your OS, or POSIX. That means IO::AIO
765	can be added in other threads and no results will be processed. After	2747	only works in the process that loaded it. Forking is fully supported, but
766	the fork the parent simply leaves the quiescent state and continues	2748	using IO::AIO in the child is not.
767	request/result processing, while the child clears the request/result
768	queue (so the requests started before the fork will only be handled in
769	the parent). Threads will be started on demand until the limit ste in the
770	parent process has been reached again.
771		2749
772	In short: the parent will, after a short pause, continue as if fork had	2750	You might get around by not I<using> IO::AIO before (or after)
773	not been called, while the child will act as if IO::AIO has not been used	2751	forking. You could also try to call the L<IO::AIO::reinit> function in the
774	yet.	2752	child:
		2753
		2754	=over 4
		2755
		2756	=item IO::AIO::reinit
		2757
		2758	Abandons all current requests and I/O threads and simply reinitialises all
		2759	data structures. This is not an operation supported by any standards, but
		2760	happens to work on GNU/Linux and some newer BSD systems.
		2761
		2762	The only reasonable use for this function is to call it after forking, if
		2763	C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
		2764	the process will result in undefined behaviour. Calling it at any time
		2765	will also result in any undefined (by POSIX) behaviour.
		2766
		2767	=back
		2768
		2769	=head2 LINUX-SPECIFIC CALLS
		2770
		2771	When a call is documented as "linux-specific" then this means it
		2772	originated on GNU/Linux. C<IO::AIO> will usually try to autodetect the
		2773	availability and compatibility of such calls regardless of the platform
		2774	it is compiled on, so platforms such as FreeBSD which often implement
		2775	these calls will work. When in doubt, call them and see if they fail wth
		2776	C<ENOSYS>.
		2777
		2778	=head2 MEMORY USAGE
		2779
		2780	Per-request usage:
		2781
		2782	Each aio request uses - depending on your architecture - around 100-200
		2783	bytes of memory. In addition, stat requests need a stat buffer (possibly
		2784	a few hundred bytes), readdir requires a result buffer and so on. Perl
		2785	scalars and other data passed into aio requests will also be locked and
		2786	will consume memory till the request has entered the done state.
		2787
		2788	This is not awfully much, so queuing lots of requests is not usually a
		2789	problem.
		2790
		2791	Per-thread usage:
		2792
		2793	In the execution phase, some aio requests require more memory for
		2794	temporary buffers, and each thread requires a stack and other data
		2795	structures (usually around 16k-128k, depending on the OS).
		2796
		2797	=head1 KNOWN BUGS
		2798
		2799	Known bugs will be fixed in the next release :)
		2800
		2801	=head1 KNOWN ISSUES
		2802
		2803	Calls that try to "import" foreign memory areas (such as C<IO::AIO::mmap>
		2804	or C<IO::AIO::aio_slurp>) do not work with generic lvalues, such as
		2805	non-created hash slots or other scalars I didn't think of. It's best to
		2806	avoid such and either use scalar variables or making sure that the scalar
		2807	exists (e.g. by storing C<undef>) and isn't "funny" (e.g. tied).
		2808
		2809	I am not sure anything can be done about this, so this is considered a
		2810	known issue, rather than a bug.
775		2811
776	=head1 SEE ALSO	2812	=head1 SEE ALSO
777		2813
778	L<Coro>, L<Linux::AIO> (obsolete).	2814	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		2815	more natural syntax and L<IO::FDPass> for file descriptor passing.
779		2816
780	=head1 AUTHOR	2817	=head1 AUTHOR
781		2818
782	Marc Lehmann <schmorp@schmorp.de>	2819	Marc Lehmann <schmorp@schmorp.de>
783	http://home.schmorp.de/	2820	http://home.schmorp.de/

Diff Legend

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