ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.12 by root, Mon Jul 11 01:03:17 2005 UTC vs.
Revision 1.318 by root, Sat Apr 1 02:14:05 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
16	aio_read $fh, 30000, 1024, $buffer, 0, sub {	17	aio_read $fh, 30000, 1024, $buffer, 0, sub {
17	$_[0] > 0 or die "read error: $!";	18	$_[0] > 0 or die "read error: $!";
18	};	19	};
19		20
20	# Event	21	# version 2+ has request and group objects
21	Event->io (fd => IO::AIO::poll_fileno,	22	use IO::AIO 2;
22	poll => 'r',
23	cb => \&IO::AIO::poll_cb);
24		23
25	# Glib/Gtk2	24	aioreq_pri 4; # give next request a very high priority
26	add_watch Glib::IO IO::AIO::poll_fileno,	25	my $req = aio_unlink "/tmp/file", sub { };
27	\&IO::AIO::poll_cb;	26	$req->cancel; # cancel request if still in queue
28		27
29	# Tk	28	my $grp = aio_group sub { print "all stats done\n" };
30	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",	29	add $grp aio_stat "..." for ...;
31	readable => \&IO::AIO::poll_cb);
32
33	# Danga::Socket
34	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
35	\&IO::AIO::poll_cb);
36
37		30
38	=head1 DESCRIPTION	31	=head1 DESCRIPTION
39		32
40	This module implements asynchronous I/O using whatever means your	33	This module implements asynchronous I/O using whatever means your
41	operating system supports.	34	operating system supports. It is implemented as an interface to C<libeio>
		35	(L<http://software.schmorp.de/pkg/libeio.html>).
42		36
		37	Asynchronous means that operations that can normally block your program
		38	(e.g. reading from disk) will be done asynchronously: the operation
		39	will still block, but you can do something else in the meantime. This
		40	is extremely useful for programs that need to stay interactive even
		41	when doing heavy I/O (GUI programs, high performance network servers
		42	etc.), but can also be used to easily do operations in parallel that are
		43	normally done sequentially, e.g. stat'ing many files, which is much faster
		44	on a RAID volume or over NFS when you do a number of stat operations
		45	concurrently.
		46
		47	While most of this works on all types of file descriptors (for
		48	example sockets), using these functions on file descriptors that
		49	support nonblocking operation (again, sockets, pipes etc.) is
		50	very inefficient. Use an event loop for that (such as the L<EV>
		51	module): IO::AIO will naturally fit into such an event loop itself.
		52
43	Currently, a number of threads are started that execute your read/writes	53	In this version, a number of threads are started that execute your
44	and signal their completion. You don't need thread support in your libc or	54	requests and signal their completion. You don't need thread support
45	perl, and the threads created by this module will not be visible to the	55	in perl, and the threads created by this module will not be visible
46	pthreads library. In the future, this module might make use of the native	56	to perl. In the future, this module might make use of the native aio
47	aio functions available on many operating systems. However, they are often	57	functions available on many operating systems. However, they are often
48	not well-supported (Linux doesn't allow them on normal files currently,	58	not well-supported or restricted (GNU/Linux doesn't allow them on normal
49	for example), and they would only support aio_read and aio_write, so the	59	files currently, for example), and they would only support aio_read and
50	remaining functionality would have to be implemented using threads anyway.	60	aio_write, so the remaining functionality would have to be implemented
		61	using threads anyway.
51		62
		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
52	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,
53	currently not reentrant, so use appropriate locking yourself.	68	it is currently not reentrant in any way, so use appropriate locking
		69	yourself, always call C<poll_cb> from within the same thread, or never
		70	call C<poll_cb> (or other C<aio_> functions) recursively.
		71
		72	=head2 EXAMPLE
		73
		74	This is a simple example that uses the EV module and loads
		75	F</etc/passwd> asynchronously:
		76
		77	use EV;
		78	use IO::AIO;
		79
		80	# register the IO::AIO callback with EV
		81	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		82
		83	# queue the request to open /etc/passwd
		84	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
		85	my $fh = shift
		86	or die "error while opening: $!";
		87
		88	# stat'ing filehandles is generally non-blocking
		89	my $size = -s $fh;
		90
		91	# queue a request to read the file
		92	my $contents;
		93	aio_read $fh, 0, $size, $contents, 0, sub {
		94	$_[0] == $size
		95	or die "short read: $!";
		96
		97	close $fh;
		98
		99	# file contents now in $contents
		100	print $contents;
		101
		102	# exit event loop and program
		103	EV::break;
		104	};
		105	};
		106
		107	# possibly queue up other requests, or open GUI windows,
		108	# check for sockets etc. etc.
		109
		110	# process events as long as there are some:
		111	EV::run;
		112
		113	=head1 REQUEST ANATOMY AND LIFETIME
		114
		115	Every C<aio_*> function creates a request. which is a C data structure not
		116	directly visible to Perl.
		117
		118	If called in non-void context, every request function returns a Perl
		119	object representing the request. In void context, nothing is returned,
		120	which saves a bit of memory.
		121
		122	The perl object is a fairly standard ref-to-hash object. The hash contents
		123	are not used by IO::AIO so you are free to store anything you like in it.
		124
		125	During their existance, aio requests travel through the following states,
		126	in order:
		127
		128	=over 4
		129
		130	=item ready
		131
		132	Immediately after a request is created it is put into the ready state,
		133	waiting for a thread to execute it.
		134
		135	=item execute
		136
		137	A thread has accepted the request for processing and is currently
		138	executing it (e.g. blocking in read).
		139
		140	=item pending
		141
		142	The request has been executed and is waiting for result processing.
		143
		144	While request submission and execution is fully asynchronous, result
		145	processing is not and relies on the perl interpreter calling C<poll_cb>
		146	(or another function with the same effect).
		147
		148	=item result
		149
		150	The request results are processed synchronously by C<poll_cb>.
		151
		152	The C<poll_cb> function will process all outstanding aio requests by
		153	calling their callbacks, freeing memory associated with them and managing
		154	any groups they are contained in.
		155
		156	=item done
		157
		158	Request has reached the end of its lifetime and holds no resources anymore
		159	(except possibly for the Perl object, but its connection to the actual
		160	aio request is severed and calling its methods will either do nothing or
		161	result in a runtime error).
		162
		163	=back
54		164
55	=cut	165	=cut
56		166
57	package IO::AIO;	167	package IO::AIO;
58		168
		169	use Carp ();
		170
		171	use common::sense;
		172
59	use base 'Exporter';	173	use base 'Exporter';
60		174
61	use Fcntl ();
62
63	BEGIN {	175	BEGIN {
64	$VERSION = 0.3;	176	our $VERSION = 4.80;
65		177
66	@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink	178	our @AIO_REQ = qw(aio_sendfile aio_seek aio_read aio_write aio_open aio_close
67	aio_fsync aio_fdatasync aio_readahead);	179	aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
68	@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	180	aio_scandir aio_symlink aio_readlink aio_realpath aio_fcntl aio_ioctl
		181	aio_sync aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range
		182	aio_pathsync aio_readahead aio_fiemap aio_allocate
		183	aio_rename aio_rename2 aio_link aio_move aio_copy aio_group
		184	aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
		185	aio_chmod aio_utime aio_truncate
		186	aio_msync aio_mtouch aio_mlock aio_mlockall
		187	aio_statvfs
		188	aio_slurp
		189	aio_wd);
		190
		191	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		192	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		193	min_parallel max_parallel max_idle idle_timeout
		194	nreqs nready npending nthreads
		195	max_poll_time max_poll_reqs
		196	sendfile fadvise madvise
		197	mmap munmap 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
		205
		206	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
69		207
70	require XSLoader;	208	require XSLoader;
71	XSLoader::load IO::AIO, $VERSION;	209	XSLoader::load ("IO::AIO", $VERSION);
72	}	210	}
73		211
74	=head1 FUNCTIONS	212	=head1 FUNCTIONS
75		213
76	=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, $flags = 0
		334
		335	=head2 API NOTES
77		336
78	All the C<aio_*> calls are more or less thin wrappers around the syscall	337	All the C<aio_*> calls are more or less thin wrappers around the syscall
79	with the same name (sans C<aio_>). The arguments are similar or identical,	338	with the same name (sans C<aio_>). The arguments are similar or identical,
80	and they all accept an additional C<$callback> argument which must be	339	and they all accept an additional (and optional) C<$callback> argument
81	a code reference. This code reference will get called with the syscall	340	which must be a code reference. This code reference will be called after
		341	the syscall has been executed in an asynchronous fashion. The results
		342	of the request will be passed as arguments to the callback (and, if an
		343	error occured, in C<$!>) - for most requests the syscall return code (e.g.
82	return code (e.g. most syscalls return C<-1> on error, unlike perl, which	344	most syscalls return C<-1> on error, unlike perl, which usually delivers
83	usually delivers "false") as it's sole argument when the given syscall has	345	"false").
84	been executed asynchronously.
85		346
86	All functions that expect a filehandle will also accept a file descriptor.	347	Some requests (such as C<aio_readdir>) pass the actual results and
		348	communicate failures by passing C<undef>.
87		349
		350	All functions expecting a filehandle keep a copy of the filehandle
		351	internally until the request has finished.
		352
		353	All functions return request objects of type L<IO::AIO::REQ> that allow
		354	further manipulation of those requests while they are in-flight.
		355
88	The filenames you pass to these routines I<must> be absolute. The reason	356	The pathnames you pass to these routines I<should> be absolute. The
89	is that at the time the request is being executed, the current working	357	reason for this is that at the time the request is being executed, the
90	directory could have changed. Alternatively, you can make sure that you	358	current working directory could have changed. Alternatively, you can
91	never change the current working directory.	359	make sure that you never change the current working directory anywhere
		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.
		364
		365	To encode pathnames as octets, either make sure you either: a) always pass
		366	in filenames you got from outside (command line, readdir etc.) without
		367	tinkering, b) are in your native filesystem encoding, c) use the Encode
		368	module and encode your pathnames to the locale (or other) encoding in
		369	effect in the user environment, d) use Glib::filename_from_unicode on
		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
92		377
93	=over 4	378	=over 4
94		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
95	=item aio_open $pathname, $flags, $mode, $callback	413	=item aio_open $pathname, $flags, $mode, $callback->($fh)
96		414
97	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
98	created filehandle for the file.	416	created filehandle for the file (or C<undef> in case of an error).
99		417
100	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
101	for an explanation.
102
103	The C<$mode> argument is a bitmask. See the C<Fcntl> module for a	418	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
104	list. They are the same as used in C<sysopen>.	419	list. They are the same as used by C<sysopen>.
		420
		421	Likewise, C<$mode> specifies the mode of the newly created file, if it
		422	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
		423	except that it is mandatory (i.e. use C<0> if you don't create new files,
		424	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		425	by the umask in effect then the request is being executed, so better never
		426	change the umask.
105		427
106	Example:	428	Example:
107		429
108	aio_open "/etc/passwd", O_RDONLY, 0, sub {	430	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
109	if ($_[0]) {	431	if ($_[0]) {
110	print "open successful, fh is $_[0]\n";	432	print "open successful, fh is $_[0]\n";
111	...	433	...
112	} else {	434	} else {
113	die "open failed: $!\n";	435	die "open failed: $!\n";
114	}	436	}
115	};	437	};
116		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
117	=item aio_close $fh, $callback	449	=item aio_close $fh, $callback->($status)
118		450
119	Asynchronously close a file and call the callback with the result	451	Asynchronously close a file and call the callback with the result
120	code. I<WARNING:> although accepted, you should not pass in a perl	452	code.
121	filehandle here, as perl will likely close the file descriptor itself when
122	the filehandle is destroyed. Normally, you can safely call perls C<close>
123	or just let filehandles go out of scope.
124		453
		454	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
		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".
		485
125	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	486	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
126		487
127	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	488	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
128		489
129	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
130	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<$dataoffset> and
131	callback without the actual number of bytes read (or -1 on error, just	492	calls the callback with the actual number of bytes transferred (or -1 on
132	like the syscall).	493	error, just like the syscall).
133		494
		495	C<aio_read> will, like C<sysread>, shrink or grow the C<$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<$data>.
		507
		508	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		509	is outstanding. Modifying it can result in segfaults or World War III (if
		510	the necessary/optional hardware is installed).
		511
134	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at	512	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
135	offset C<0> within the scalar:	513	offset C<0> within the scalar:
136		514
137	aio_read $fh, 7, 15, $buffer, 0, sub {	515	aio_read $fh, 7, 15, $buffer, 0, sub {
138	$_[0] > 0 or die "read error: $!";	516	$_[0] > 0 or die "read error: $!";
139	print "read $_[0] bytes: <$buffer>\n";	517	print "read $_[0] bytes: <$buffer>\n";
140	};	518	};
141		519
		520
		521	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		522
		523	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		524	reading at byte offset C<$in_offset>, and starts writing at the current
		525	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		526	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		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>.
		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<$in_fh>) is a file - the file I/O will then be asynchronous, while
		539	the socket I/O will be non-blocking. Note, however, that you can run
		540	into a trap where C<aio_sendfile> reads some data with readahead, then
		541	fails to write all data, and when the socket is ready the next time, the
		542	data in the cache is already lost, forcing C<aio_sendfile> to again hit
		543	the disk. Explicit C<aio_read> + C<aio_write> let's you better control
		544	resource usage.
		545
		546	This call tries to make use of a native C<sendfile>-like syscall to
		547	provide zero-copy operation. For this to work, C<$out_fh> should refer to
		548	a socket, and C<$in_fh> should refer to an mmap'able file.
		549
		550	If a native sendfile cannot be found or it fails with C<ENOSYS>,
		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
		553	type of filehandle regardless of the limitations of the operating system.
		554
		555	As native sendfile syscalls (as practically any non-POSIX interface hacked
		556	together in a hurry to improve benchmark numbers) tend to be rather buggy
		557	on many systems, this implementation tries to work around some known bugs
		558	in Linux and FreeBSD kernels (probably others, too), but that might fail,
		559	so you really really should check the return value of C<aio_sendfile> -
		560	fewer bytes than expected might have been transferred.
		561
		562
142	=item aio_readahead $fh,$offset,$length, $callback	563	=item aio_readahead $fh,$offset,$length, $callback->($retval)
143		564
144	Asynchronously reads the specified byte range into the page cache, using
145	the C<readahead> syscall. If that syscall doesn't exist the status will be
146	C<-1> and C<$!> is set to ENOSYS.
147
148	readahead() populates the page cache with data from a file so that	565	C<aio_readahead> populates the page cache with data from a file so that
149	subsequent reads from that file will not block on disk I/O. The C<$offset>	566	subsequent reads from that file will not block on disk I/O. The C<$offset>
150	argument specifies the starting point from which data is to be read and	567	argument specifies the starting point from which data is to be read and
151	C<$length> specifies the number of bytes to be read. I/O is performed in	568	C<$length> specifies the number of bytes to be read. I/O is performed in
152	whole pages, so that offset is effectively rounded down to a page boundary	569	whole pages, so that offset is effectively rounded down to a page boundary
153	and bytes are read up to the next page boundary greater than or equal to	570	and bytes are read up to the next page boundary greater than or equal to
154	(off-set+length). aio_readahead() does not read beyond the end of the	571	(off-set+length). C<aio_readahead> does not read beyond the end of the
155	file. The current file offset of the file is left unchanged.	572	file. The current file offset of the file is left unchanged.
156		573
		574	If that syscall doesn't exist (likely if your kernel isn't Linux) it will
		575	be emulated by simply reading the data, which would have a similar effect.
		576
		577
157	=item aio_stat $fh_or_path, $callback	578	=item aio_stat $fh_or_path, $callback->($status)
158		579
159	=item aio_lstat $fh, $callback	580	=item aio_lstat $fh, $callback->($status)
160		581
161	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
162	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
163	or C<-s _> etc...	584	using C<stat _> or C<-s _> and other tests (with the exception of C<-B>
164		585	and C<-T>).
165	The pathname passed to C<aio_stat> must be absolute. See API NOTES, above,
166	for an explanation.
167		586
168	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
169	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
170	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>.
171		602
172	Example: Print the length of F</etc/passwd>:	603	Example: Print the length of F</etc/passwd>:
173		604
174	aio_stat "/etc/passwd", sub {	605	aio_stat "/etc/passwd", sub {
175	$_[0] and die "stat failed: $!";	606	$_[0] and die "stat failed: $!";
176	print "size is ", -s _, "\n";	607	print "size is ", -s _, "\n";
177	};	608	};
178		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
179	=item aio_unlink $pathname, $callback	719	=item aio_unlink $pathname, $callback->($status)
180		720
181	Asynchronously unlink (delete) a file and call the callback with the	721	Asynchronously unlink (delete) a file and call the callback with the
182	result code.	722	result code.
183		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
		738	=item aio_link $srcpath, $dstpath, $callback->($status)
		739
		740	Asynchronously create a new link to the existing object at C<$srcpath> at
		741	the path C<$dstpath> and call the callback with the result code.
		742
		743
		744	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		745
		746	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		747	the path C<$dstpath> and call the callback with the result code.
		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
		767	=item aio_rename $srcpath, $dstpath, $callback->($status)
		768
		769	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		770	rename(2) and call the callback with the result code.
		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
		800	=item aio_rmdir $pathname, $callback->($status)
		801
		802	Asynchronously rmdir (delete) a directory and call the callback with the
		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
		809
		810	=item aio_readdir $pathname, $callback->($entries)
		811
		812	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		813	directory (i.e. opendir + readdir + closedir). The entries will not be
		814	sorted, and will B<NOT> include the C<.> and C<..> entries.
		815
		816	The callback is passed a single argument which is either C<undef> or an
		817	array-ref with the filenames.
		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($$;$) {
		967	my ($src, $dst, $cb) = @_;
		968
		969	my $pri = aioreq_pri;
		970	my $grp = aio_group $cb;
		971
		972	aioreq_pri $pri;
		973	add $grp aio_open $src, O_RDONLY, 0, sub {
		974	if (my $src_fh = $_[0]) {
		975	my @stat = stat $src_fh; # hmm, might block over nfs?
		976
		977	aioreq_pri $pri;
		978	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		979	if (my $dst_fh = $_[0]) {
		980	aioreq_pri $pri;
		981	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		982	if ($_[0] == $stat[7]) {
		983	$grp->result (0);
		984	close $src_fh;
		985
		986	my $ch = sub {
		987	aioreq_pri $pri;
		988	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		989	aioreq_pri $pri;
		990	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		991	aioreq_pri $pri;
		992	add $grp aio_close $dst_fh;
		993	}
		994	};
		995	};
		996
		997	aioreq_pri $pri;
		998	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		999	if ($_[0] < 0 && $! == ENOSYS) {
		1000	aioreq_pri $pri;
		1001	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		1002	} else {
		1003	$ch->();
		1004	}
		1005	};
		1006	} else {
		1007	$grp->result (-1);
		1008	close $src_fh;
		1009	close $dst_fh;
		1010
		1011	aioreq $pri;
		1012	add $grp aio_unlink $dst;
		1013	}
		1014	};
		1015	} else {
		1016	$grp->result (-1);
		1017	}
		1018	},
		1019
		1020	} else {
		1021	$grp->result (-1);
		1022	}
		1023	};
		1024
		1025	$grp
		1026	}
		1027
		1028	=item aio_move $srcpath, $dstpath, $callback->($status)
		1029
		1030	Try to move the I<file> (directories not supported as either source or
		1031	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		1032	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		1033
		1034	This is a composite request that tries to rename(2) the file first; if
		1035	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		1036	that is successful, unlinks the C<$srcpath>.
		1037
		1038	=cut
		1039
		1040	sub aio_move($$;$) {
		1041	my ($src, $dst, $cb) = @_;
		1042
		1043	my $pri = aioreq_pri;
		1044	my $grp = aio_group $cb;
		1045
		1046	aioreq_pri $pri;
		1047	add $grp aio_rename $src, $dst, sub {
		1048	if ($_[0] && $! == EXDEV) {
		1049	aioreq_pri $pri;
		1050	add $grp aio_copy $src, $dst, sub {
		1051	$grp->result ($_[0]);
		1052
		1053	unless ($_[0]) {
		1054	aioreq_pri $pri;
		1055	add $grp aio_unlink $src;
		1056	}
		1057	};
		1058	} else {
		1059	$grp->result ($_[0]);
		1060	}
		1061	};
		1062
		1063	$grp
		1064	}
		1065
		1066	=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		1067
		1068	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		1069	efficiently separate the entries of directory C<$path> into two sets of
		1070	names, directories you can recurse into (directories), and ones you cannot
		1071	recurse into (everything else, including symlinks to directories).
		1072
		1073	C<aio_scandir> is a composite request that generates many sub requests.
		1074	C<$maxreq> specifies the maximum number of outstanding aio requests that
		1075	this function generates. If it is C<< <= 0 >>, then a suitable default
		1076	will be chosen (currently 4).
		1077
		1078	On error, the callback is called without arguments, otherwise it receives
		1079	two array-refs with path-relative entry names.
		1080
		1081	Example:
		1082
		1083	aio_scandir $dir, 0, sub {
		1084	my ($dirs, $nondirs) = @_;
		1085	print "real directories: @$dirs\n";
		1086	print "everything else: @$nondirs\n";
		1087	};
		1088
		1089	Implementation notes.
		1090
		1091	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		1092
		1093	If readdir returns file type information, then this is used directly to
		1094	find directories.
		1095
		1096	Otherwise, after reading the directory, the modification time, size etc.
		1097	of the directory before and after the readdir is checked, and if they
		1098	match (and isn't the current time), the link count will be used to decide
		1099	how many entries are directories (if >= 2). Otherwise, no knowledge of the
		1100	number of subdirectories will be assumed.
		1101
		1102	Then entries will be sorted into likely directories a non-initial dot
		1103	currently) and likely non-directories (see C<aio_readdirx>). Then every
		1104	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
		1105	in order of their inode numbers. If that succeeds, it assumes that the
		1106	entry is a directory or a symlink to directory (which will be checked
		1107	separately). This is often faster than stat'ing the entry itself because
		1108	filesystems might detect the type of the entry without reading the inode
		1109	data (e.g. ext2fs filetype feature), even on systems that cannot return
		1110	the filetype information on readdir.
		1111
		1112	If the known number of directories (link count - 2) has been reached, the
		1113	rest of the entries is assumed to be non-directories.
		1114
		1115	This only works with certainty on POSIX (= UNIX) filesystems, which
		1116	fortunately are the vast majority of filesystems around.
		1117
		1118	It will also likely work on non-POSIX filesystems with reduced efficiency
		1119	as those tend to return 0 or 1 as link counts, which disables the
		1120	directory counting heuristic.
		1121
		1122	=cut
		1123
		1124	sub aio_scandir($$;$) {
		1125	my ($path, $maxreq, $cb) = @_;
		1126
		1127	my $pri = aioreq_pri;
		1128
		1129	my $grp = aio_group $cb;
		1130
		1131	$maxreq = 4 if $maxreq <= 0;
		1132
		1133	# get a wd object
		1134	aioreq_pri $pri;
		1135	add $grp aio_wd $path, sub {
		1136	$_[0]
		1137	or return $grp->result ();
		1138
		1139	my $wd = [shift, "."];
		1140
		1141	# stat once
		1142	aioreq_pri $pri;
		1143	add $grp aio_stat $wd, sub {
		1144	return $grp->result () if $_[0];
		1145	my $now = time;
		1146	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		1147	my $rdxflags = READDIR_DIRS_FIRST;
		1148
		1149	if ((stat _)[3] < 2) {
		1150	# at least one non-POSIX filesystem exists
		1151	# that returns useful DT_type values: btrfs,
		1152	# so optimise for this here by requesting dents
		1153	$rdxflags |= READDIR_DENTS;
		1154	}
		1155
		1156	# read the directory entries
		1157	aioreq_pri $pri;
		1158	add $grp aio_readdirx $wd, $rdxflags, sub {
		1159	my ($entries, $flags) = @_
		1160	or return $grp->result ();
		1161
		1162	if ($rdxflags & READDIR_DENTS) {
		1163	# if we requested type values, see if we can use them directly.
		1164
		1165	# if there were any DT_UNKNOWN entries then we assume we
		1166	# don't know. alternatively, we could assume that if we get
		1167	# one DT_DIR, then all directories are indeed marked with
		1168	# DT_DIR, but this seems not required for btrfs, and this
		1169	# is basically the "btrfs can't get it's act together" code
		1170	# branch.
		1171	unless ($flags & READDIR_FOUND_UNKNOWN) {
		1172	# now we have valid DT_ information for all entries,
		1173	# so use it as an optimisation without further stat's.
		1174	# they must also all be at the beginning of @$entries
		1175	# by now.
		1176
		1177	my $dirs;
		1178
		1179	if (@$entries) {
		1180	for (0 .. $#$entries) {
		1181	if ($entries->[$_][1] != DT_DIR) {
		1182	# splice out directories
		1183	$dirs = [splice @$entries, 0, $_];
		1184	last;
		1185	}
		1186	}
		1187
		1188	# if we didn't find any non-dir, then all entries are dirs
		1189	unless ($dirs) {
		1190	($dirs, $entries) = ($entries, []);
		1191	}
		1192	} else {
		1193	# directory is empty, so there are no sbdirs
		1194	$dirs = [];
		1195	}
		1196
		1197	# either splice'd the directories out or the dir was empty.
		1198	# convert dents to filenames
		1199	$_ = $_->[0] for @$dirs;
		1200	$_ = $_->[0] for @$entries;
		1201
		1202	return $grp->result ($dirs, $entries);
		1203	}
		1204
		1205	# cannot use, so return to our old ways
		1206	# by pretending we only scanned for names.
		1207	$_ = $_->[0] for @$entries;
		1208	}
		1209
		1210	# stat the dir another time
		1211	aioreq_pri $pri;
		1212	add $grp aio_stat $wd, sub {
		1213	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		1214
		1215	my $ndirs;
		1216
		1217	# take the slow route if anything looks fishy
		1218	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		1219	$ndirs = -1;
		1220	} else {
		1221	# if nlink == 2, we are finished
		1222	# for non-posix-fs's, we rely on nlink < 2
		1223	$ndirs = (stat _)[3] - 2
		1224	or return $grp->result ([], $entries);
		1225	}
		1226
		1227	my (@dirs, @nondirs);
		1228
		1229	my $statgrp = add $grp aio_group sub {
		1230	$grp->result (\@dirs, \@nondirs);
		1231	};
		1232
		1233	limit $statgrp $maxreq;
		1234	feed $statgrp sub {
		1235	return unless @$entries;
		1236	my $entry = shift @$entries;
		1237
		1238	aioreq_pri $pri;
		1239	$wd->[1] = "$entry/.";
		1240	add $statgrp aio_stat $wd, sub {
		1241	if ($_[0] < 0) {
		1242	push @nondirs, $entry;
		1243	} else {
		1244	# need to check for real directory
		1245	aioreq_pri $pri;
		1246	$wd->[1] = $entry;
		1247	add $statgrp aio_lstat $wd, sub {
		1248	if (-d _) {
		1249	push @dirs, $entry;
		1250
		1251	unless (--$ndirs) {
		1252	push @nondirs, @$entries;
		1253	feed $statgrp;
		1254	}
		1255	} else {
		1256	push @nondirs, $entry;
		1257	}
		1258	}
		1259	}
		1260	};
		1261	};
		1262	};
		1263	};
		1264	};
		1265	};
		1266
		1267	$grp
		1268	}
		1269
		1270	=item aio_rmtree $pathname, $callback->($status)
		1271
		1272	Delete a directory tree starting (and including) C<$path>, return the
		1273	status of the final C<rmdir> only. This is a composite request that
		1274	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1275	everything else.
		1276
		1277	=cut
		1278
		1279	sub aio_rmtree;
		1280	sub aio_rmtree($;$) {
		1281	my ($path, $cb) = @_;
		1282
		1283	my $pri = aioreq_pri;
		1284	my $grp = aio_group $cb;
		1285
		1286	aioreq_pri $pri;
		1287	add $grp aio_scandir $path, 0, sub {
		1288	my ($dirs, $nondirs) = @_;
		1289
		1290	my $dirgrp = aio_group sub {
		1291	add $grp aio_rmdir $path, sub {
		1292	$grp->result ($_[0]);
		1293	};
		1294	};
		1295
		1296	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1297	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1298
		1299	add $grp $dirgrp;
		1300	};
		1301
		1302	$grp
		1303	}
		1304
		1305	=item aio_fcntl $fh, $cmd, $arg, $callback->($status)
		1306
		1307	=item aio_ioctl $fh, $request, $buf, $callback->($status)
		1308
		1309	These work just like the C<fcntl> and C<ioctl> built-in functions, except
		1310	they execute asynchronously and pass the return value to the callback.
		1311
		1312	Both calls can be used for a lot of things, some of which make more sense
		1313	to run asynchronously in their own thread, while some others make less
		1314	sense. For example, calls that block waiting for external events, such
		1315	as locking, will also lock down an I/O thread while it is waiting, which
		1316	can deadlock the whole I/O system. At the same time, there might be no
		1317	alternative to using a thread to wait.
		1318
		1319	So in general, you should only use these calls for things that do
		1320	(filesystem) I/O, not for things that wait for other events (network,
		1321	other processes), although if you are careful and know what you are doing,
		1322	you still can.
		1323
		1324	The following constants are available and can be used for normal C<ioctl>
		1325	and C<fcntl> as well (missing ones are, as usual C<0>):
		1326
		1327	C<F_DUPFD_CLOEXEC>,
		1328
		1329	C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>,
		1330
		1331	C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>.
		1332
		1333	C<F_ADD_SEALS>, C<F_GET_SEALS>, C<F_SEAL_SEAL>, C<F_SEAL_SHRINK>, C<F_SEAL_GROW> and
		1334	C<F_SEAL_WRITE>.
		1335
		1336	C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>,
		1337	C<FS_IOC_FIEMAP>.
		1338
		1339	C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>,
		1340	C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>.
		1341
		1342	C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>,
		1343	C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>,
		1344	C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>,
		1345	C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>,
		1346	C<FS_FL_USER_MODIFIABLE>.
		1347
		1348	C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>,
		1349	C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>,
		1350	C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>,
		1351	C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>,
		1352
		1353	C<BLKROSET>, C<BLKROGET>, C<BLKRRPART>, C<BLKGETSIZE>, C<BLKFLSBUF>, C<BLKRASET>,
		1354	C<BLKRAGET>, C<BLKFRASET>, C<BLKFRAGET>, C<BLKSECTSET>, C<BLKSECTGET>, C<BLKSSZGET>,
		1355	C<BLKBSZGET>, C<BLKBSZSET>, C<BLKGETSIZE64>,
		1356
		1357
		1358	=item aio_sync $callback->($status)
		1359
		1360	Asynchronously call sync and call the callback when finished.
		1361
184	=item aio_fsync $fh, $callback	1362	=item aio_fsync $fh, $callback->($status)
185		1363
186	Asynchronously call fsync on the given filehandle and call the callback	1364	Asynchronously call fsync on the given filehandle and call the callback
187	with the fsync result code.	1365	with the fsync result code.
188		1366
189	=item aio_fdatasync $fh, $callback	1367	=item aio_fdatasync $fh, $callback->($status)
190		1368
191	Asynchronously call fdatasync on the given filehandle and call the	1369	Asynchronously call fdatasync on the given filehandle and call the
192	callback with the fdatasync result code.	1370	callback with the fdatasync result code.
193		1371
		1372	If this call isn't available because your OS lacks it or it couldn't be
		1373	detected, it will be emulated by calling C<fsync> instead.
		1374
		1375	=item aio_syncfs $fh, $callback->($status)
		1376
		1377	Asynchronously call the syncfs syscall to sync the filesystem associated
		1378	to the given filehandle and call the callback with the syncfs result
		1379	code. If syncfs is not available, calls sync(), but returns C<-1> and sets
		1380	errno to C<ENOSYS> nevertheless.
		1381
		1382	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1383
		1384	Sync the data portion of the file specified by C<$offset> and C<$length>
		1385	to disk (but NOT the metadata), by calling the Linux-specific
		1386	sync_file_range call. If sync_file_range is not available or it returns
		1387	ENOSYS, then fdatasync or fsync is being substituted.
		1388
		1389	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1390	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1391	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1392	manpage for details.
		1393
		1394	=item aio_pathsync $pathname, $callback->($status)
		1395
		1396	This request tries to open, fsync and close the given path. This is a
		1397	composite request intended to sync directories after directory operations
		1398	(E.g. rename). This might not work on all operating systems or have any
		1399	specific effect, but usually it makes sure that directory changes get
		1400	written to disc. It works for anything that can be opened for read-only,
		1401	not just directories.
		1402
		1403	Future versions of this function might fall back to other methods when
		1404	C<fsync> on the directory fails (such as calling C<sync>).
		1405
		1406	Passes C<0> when everything went ok, and C<-1> on error.
		1407
		1408	=cut
		1409
		1410	sub aio_pathsync($;$) {
		1411	my ($path, $cb) = @_;
		1412
		1413	my $pri = aioreq_pri;
		1414	my $grp = aio_group $cb;
		1415
		1416	aioreq_pri $pri;
		1417	add $grp aio_open $path, O_RDONLY, 0, sub {
		1418	my ($fh) = @_;
		1419	if ($fh) {
		1420	aioreq_pri $pri;
		1421	add $grp aio_fsync $fh, sub {
		1422	$grp->result ($_[0]);
		1423
		1424	aioreq_pri $pri;
		1425	add $grp aio_close $fh;
		1426	};
		1427	} else {
		1428	$grp->result (-1);
		1429	}
		1430	};
		1431
		1432	$grp
		1433	}
		1434
		1435	=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
		1436
		1437	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1438	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1439	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1440	scalar must only be modified in-place while an aio operation is pending on
		1441	it).
		1442
		1443	It calls the C<msync> function of your OS, if available, with the memory
		1444	area starting at C<$offset> in the string and ending C<$length> bytes
		1445	later. If C<$length> is negative, counts from the end, and if C<$length>
		1446	is C<undef>, then it goes till the end of the string. The flags can be
		1447	either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional
		1448	C<IO::AIO::MS_INVALIDATE>.
		1449
		1450	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1451
		1452	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1453	scalars.
		1454
		1455	It touches (reads or writes) all memory pages in the specified
		1456	range inside the scalar. All caveats and parameters are the same
		1457	as for C<aio_msync>, above, except for flags, which must be either
		1458	C<0> (which reads all pages and ensures they are instantiated) or
		1459	C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and
		1460	writing an octet from it, which dirties the page).
		1461
		1462	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1463
		1464	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1465	scalars.
		1466
		1467	It reads in all the pages of the underlying storage into memory (if any)
		1468	and locks them, so they are not getting swapped/paged out or removed.
		1469
		1470	If C<$length> is undefined, then the scalar will be locked till the end.
		1471
		1472	On systems that do not implement C<mlock>, this function returns C<-1>
		1473	and sets errno to C<ENOSYS>.
		1474
		1475	Note that the corresponding C<munlock> is synchronous and is
		1476	documented under L<MISCELLANEOUS FUNCTIONS>.
		1477
		1478	Example: open a file, mmap and mlock it - both will be undone when
		1479	C<$data> gets destroyed.
		1480
		1481	open my $fh, "<", $path or die "$path: $!";
		1482	my $data;
		1483	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1484	aio_mlock $data; # mlock in background
		1485
		1486	=item aio_mlockall $flags, $callback->($status)
		1487
		1488	Calls the C<mlockall> function with the given C<$flags> (a
		1489	combination of C<IO::AIO::MCL_CURRENT>, C<IO::AIO::MCL_FUTURE> and
		1490	C<IO::AIO::MCL_ONFAULT>).
		1491
		1492	On systems that do not implement C<mlockall>, this function returns C<-1>
		1493	and sets errno to C<ENOSYS>. Similarly, flag combinations not supported
		1494	by the system result in a return value of C<-1> with errno being set to
		1495	C<EINVAL>.
		1496
		1497	Note that the corresponding C<munlockall> is synchronous and is
		1498	documented under L<MISCELLANEOUS FUNCTIONS>.
		1499
		1500	Example: asynchronously lock all current and future pages into memory.
		1501
		1502	aio_mlockall IO::AIO::MCL_FUTURE;
		1503
		1504	=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
		1505
		1506	Queries the extents of the given file (by calling the Linux C<FIEMAP>
		1507	ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If
		1508	the ioctl is not available on your OS, then this request will fail with
		1509	C<ENOSYS>.
		1510
		1511	C<$start> is the starting offset to query extents for, C<$length> is the
		1512	size of the range to query - if it is C<undef>, then the whole file will
		1513	be queried.
		1514
		1515	C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or
		1516	C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also
		1517	exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query
		1518	the data portion.
		1519
		1520	C<$count> is the maximum number of extent records to return. If it is
		1521	C<undef>, then IO::AIO queries all extents of the range. As a very special
		1522	case, if it is C<0>, then the callback receives the number of extents
		1523	instead of the extents themselves (which is unreliable, see below).
		1524
		1525	If an error occurs, the callback receives no arguments. The special
		1526	C<errno> value C<IO::AIO::EBADR> is available to test for flag errors.
		1527
		1528	Otherwise, the callback receives an array reference with extent
		1529	structures. Each extent structure is an array reference itself, with the
		1530	following members:
		1531
		1532	[$logical, $physical, $length, $flags]
		1533
		1534	Flags is any combination of the following flag values (typically either C<0>
		1535	or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)):
		1536
		1537	C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>,
		1538	C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>,
		1539	C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>,
		1540	C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>,
		1541	C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or
		1542	C<IO::AIO::FIEMAP_EXTENT_SHARED>.
		1543
		1544	At the time of this writing (Linux 3.2), this request is unreliable unless
		1545	C<$count> is C<undef>, as the kernel has all sorts of bugs preventing
		1546	it to return all extents of a range for files with a large number of
		1547	extents. The code (only) works around all these issues if C<$count> is
		1548	C<undef>.
		1549
		1550	=item aio_group $callback->(...)
		1551
		1552	This is a very special aio request: Instead of doing something, it is a
		1553	container for other aio requests, which is useful if you want to bundle
		1554	many requests into a single, composite, request with a definite callback
		1555	and the ability to cancel the whole request with its subrequests.
		1556
		1557	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		1558	for more info.
		1559
		1560	Example:
		1561
		1562	my $grp = aio_group sub {
		1563	print "all stats done\n";
		1564	};
		1565
		1566	add $grp
		1567	(aio_stat ...),
		1568	(aio_stat ...),
		1569	...;
		1570
		1571	=item aio_nop $callback->()
		1572
		1573	This is a special request - it does nothing in itself and is only used for
		1574	side effects, such as when you want to add a dummy request to a group so
		1575	that finishing the requests in the group depends on executing the given
		1576	code.
		1577
		1578	While this request does nothing, it still goes through the execution
		1579	phase and still requires a worker thread. Thus, the callback will not
		1580	be executed immediately but only after other requests in the queue have
		1581	entered their execution phase. This can be used to measure request
		1582	latency.
		1583
		1584	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
		1585
		1586	Mainly used for debugging and benchmarking, this aio request puts one of
		1587	the request workers to sleep for the given time.
		1588
		1589	While it is theoretically handy to have simple I/O scheduling requests
		1590	like sleep and file handle readable/writable, the overhead this creates is
		1591	immense (it blocks a thread for a long time) so do not use this function
		1592	except to put your application under artificial I/O pressure.
		1593
194	=back	1594	=back
195		1595
		1596
		1597	=head2 IO::AIO::WD - multiple working directories
		1598
		1599	Your process only has one current working directory, which is used by all
		1600	threads. This makes it hard to use relative paths (some other component
		1601	could call C<chdir> at any time, and it is hard to control when the path
		1602	will be used by IO::AIO).
		1603
		1604	One solution for this is to always use absolute paths. This usually works,
		1605	but can be quite slow (the kernel has to walk the whole path on every
		1606	access), and can also be a hassle to implement.
		1607
		1608	Newer POSIX systems have a number of functions (openat, fdopendir,
		1609	futimensat and so on) that make it possible to specify working directories
		1610	per operation.
		1611
		1612	For portability, and because the clowns who "designed", or shall I write,
		1613	perpetrated this new interface were obviously half-drunk, this abstraction
		1614	cannot be perfect, though.
		1615
		1616	IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
		1617	object. This object stores the canonicalised, absolute version of the
		1618	path, and on systems that allow it, also a directory file descriptor.
		1619
		1620	Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
		1621	or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
		1622	object and a pathname instead (or the IO::AIO::WD object alone, which
		1623	gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
		1624	IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
		1625	to that IO::AIO::WD object.
		1626
		1627	For example, to get a wd object for F</etc> and then stat F<passwd>
		1628	inside, you would write:
		1629
		1630	aio_wd "/etc", sub {
		1631	my $etcdir = shift;
		1632
		1633	# although $etcdir can be undef on error, there is generally no reason
		1634	# to check for errors here, as aio_stat will fail with ENOENT
		1635	# when $etcdir is undef.
		1636
		1637	aio_stat [$etcdir, "passwd"], sub {
		1638	# yay
		1639	};
		1640	};
		1641
		1642	The fact that C<aio_wd> is a request and not a normal function shows that
		1643	creating an IO::AIO::WD object is itself a potentially blocking operation,
		1644	which is why it is done asynchronously.
		1645
		1646	To stat the directory obtained with C<aio_wd> above, one could write
		1647	either of the following three request calls:
		1648
		1649	aio_lstat "/etc" , sub { ... # pathname as normal string
		1650	aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
		1651	aio_lstat $wd , sub { ... # shorthand for the previous
		1652
		1653	As with normal pathnames, IO::AIO keeps a copy of the working directory
		1654	object and the pathname string, so you could write the following without
		1655	causing any issues due to C<$path> getting reused:
		1656
		1657	my $path = [$wd, undef];
		1658
		1659	for my $name (qw(abc def ghi)) {
		1660	$path->[1] = $name;
		1661	aio_stat $path, sub {
		1662	# ...
		1663	};
		1664	}
		1665
		1666	There are some caveats: when directories get renamed (or deleted), the
		1667	pathname string doesn't change, so will point to the new directory (or
		1668	nowhere at all), while the directory fd, if available on the system,
		1669	will still point to the original directory. Most functions accepting a
		1670	pathname will use the directory fd on newer systems, and the string on
		1671	older systems. Some functions (such as C<aio_realpath>) will always rely on
		1672	the string form of the pathname.
		1673
		1674	So this functionality is mainly useful to get some protection against
		1675	C<chdir>, to easily get an absolute path out of a relative path for future
		1676	reference, and to speed up doing many operations in the same directory
		1677	(e.g. when stat'ing all files in a directory).
		1678
		1679	The following functions implement this working directory abstraction:
		1680
		1681	=over 4
		1682
		1683	=item aio_wd $pathname, $callback->($wd)
		1684
		1685	Asynchonously canonicalise the given pathname and convert it to an
		1686	IO::AIO::WD object representing it. If possible and supported on the
		1687	system, also open a directory fd to speed up pathname resolution relative
		1688	to this working directory.
		1689
		1690	If something goes wrong, then C<undef> is passwd to the callback instead
		1691	of a working directory object and C<$!> is set appropriately. Since
		1692	passing C<undef> as working directory component of a pathname fails the
		1693	request with C<ENOENT>, there is often no need for error checking in the
		1694	C<aio_wd> callback, as future requests using the value will fail in the
		1695	expected way.
		1696
		1697	=item IO::AIO::CWD
		1698
		1699	This is a compile time constant (object) that represents the process
		1700	current working directory.
		1701
		1702	Specifying this object as working directory object for a pathname is as if
		1703	the pathname would be specified directly, without a directory object. For
		1704	example, these calls are functionally identical:
		1705
		1706	aio_stat "somefile", sub { ... };
		1707	aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
		1708
		1709	=back
		1710
		1711	To recover the path associated with an IO::AIO::WD object, you can use
		1712	C<aio_realpath>:
		1713
		1714	aio_realpath $wd, sub {
		1715	warn "path is $_[0]\n";
		1716	};
		1717
		1718	Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir>
		1719	sometimes, fall back to using an absolue path.
		1720
		1721	=head2 IO::AIO::REQ CLASS
		1722
		1723	All non-aggregate C<aio_*> functions return an object of this class when
		1724	called in non-void context.
		1725
		1726	=over 4
		1727
		1728	=item cancel $req
		1729
		1730	Cancels the request, if possible. Has the effect of skipping execution
		1731	when entering the B<execute> state and skipping calling the callback when
		1732	entering the the B<result> state, but will leave the request otherwise
		1733	untouched (with the exception of readdir). That means that requests that
		1734	currently execute will not be stopped and resources held by the request
		1735	will not be freed prematurely.
		1736
		1737	=item cb $req $callback->(...)
		1738
		1739	Replace (or simply set) the callback registered to the request.
		1740
		1741	=back
		1742
		1743	=head2 IO::AIO::GRP CLASS
		1744
		1745	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		1746	objects of this class, too.
		1747
		1748	A IO::AIO::GRP object is a special request that can contain multiple other
		1749	aio requests.
		1750
		1751	You create one by calling the C<aio_group> constructing function with a
		1752	callback that will be called when all contained requests have entered the
		1753	C<done> state:
		1754
		1755	my $grp = aio_group sub {
		1756	print "all requests are done\n";
		1757	};
		1758
		1759	You add requests by calling the C<add> method with one or more
		1760	C<IO::AIO::REQ> objects:
		1761
		1762	$grp->add (aio_unlink "...");
		1763
		1764	add $grp aio_stat "...", sub {
		1765	$_[0] or return $grp->result ("error");
		1766
		1767	# add another request dynamically, if first succeeded
		1768	add $grp aio_open "...", sub {
		1769	$grp->result ("ok");
		1770	};
		1771	};
		1772
		1773	This makes it very easy to create composite requests (see the source of
		1774	C<aio_move> for an application) that work and feel like simple requests.
		1775
		1776	=over 4
		1777
		1778	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		1779	C<IO::AIO::poll_cb>, just like any other request.
		1780
		1781	=item * They can be canceled like any other request. Canceling will cancel not
		1782	only the request itself, but also all requests it contains.
		1783
		1784	=item * They can also can also be added to other IO::AIO::GRP objects.
		1785
		1786	=item * You must not add requests to a group from within the group callback (or
		1787	any later time).
		1788
		1789	=back
		1790
		1791	Their lifetime, simplified, looks like this: when they are empty, they
		1792	will finish very quickly. If they contain only requests that are in the
		1793	C<done> state, they will also finish. Otherwise they will continue to
		1794	exist.
		1795
		1796	That means after creating a group you have some time to add requests
		1797	(precisely before the callback has been invoked, which is only done within
		1798	the C<poll_cb>). And in the callbacks of those requests, you can add
		1799	further requests to the group. And only when all those requests have
		1800	finished will the the group itself finish.
		1801
		1802	=over 4
		1803
		1804	=item add $grp ...
		1805
		1806	=item $grp->add (...)
		1807
		1808	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		1809	be added, including other groups, as long as you do not create circular
		1810	dependencies.
		1811
		1812	Returns all its arguments.
		1813
		1814	=item $grp->cancel_subs
		1815
		1816	Cancel all subrequests and clears any feeder, but not the group request
		1817	itself. Useful when you queued a lot of events but got a result early.
		1818
		1819	The group request will finish normally (you cannot add requests to the
		1820	group).
		1821
		1822	=item $grp->result (...)
		1823
		1824	Set the result value(s) that will be passed to the group callback when all
		1825	subrequests have finished and set the groups errno to the current value
		1826	of errno (just like calling C<errno> without an error number). By default,
		1827	no argument will be passed and errno is zero.
		1828
		1829	=item $grp->errno ([$errno])
		1830
		1831	Sets the group errno value to C<$errno>, or the current value of errno
		1832	when the argument is missing.
		1833
		1834	Every aio request has an associated errno value that is restored when
		1835	the callback is invoked. This method lets you change this value from its
		1836	default (0).
		1837
		1838	Calling C<result> will also set errno, so make sure you either set C<$!>
		1839	before the call to C<result>, or call c<errno> after it.
		1840
		1841	=item feed $grp $callback->($grp)
		1842
		1843	Sets a feeder/generator on this group: every group can have an attached
		1844	generator that generates requests if idle. The idea behind this is that,
		1845	although you could just queue as many requests as you want in a group,
		1846	this might starve other requests for a potentially long time. For example,
		1847	C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
		1848	requests, delaying any later requests for a long time.
		1849
		1850	To avoid this, and allow incremental generation of requests, you can
		1851	instead a group and set a feeder on it that generates those requests. The
		1852	feed callback will be called whenever there are few enough (see C<limit>,
		1853	below) requests active in the group itself and is expected to queue more
		1854	requests.
		1855
		1856	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1857	not impose any limits).
		1858
		1859	If the feed does not queue more requests when called, it will be
		1860	automatically removed from the group.
		1861
		1862	If the feed limit is C<0> when this method is called, it will be set to
		1863	C<2> automatically.
		1864
		1865	Example:
		1866
		1867	# stat all files in @files, but only ever use four aio requests concurrently:
		1868
		1869	my $grp = aio_group sub { print "finished\n" };
		1870	limit $grp 4;
		1871	feed $grp sub {
		1872	my $file = pop @files
		1873	or return;
		1874
		1875	add $grp aio_stat $file, sub { ... };
		1876	};
		1877
		1878	=item limit $grp $num
		1879
		1880	Sets the feeder limit for the group: The feeder will be called whenever
		1881	the group contains less than this many requests.
		1882
		1883	Setting the limit to C<0> will pause the feeding process.
		1884
		1885	The default value for the limit is C<0>, but note that setting a feeder
		1886	automatically bumps it up to C<2>.
		1887
		1888	=back
		1889
		1890
196	=head2 SUPPORT FUNCTIONS	1891	=head2 SUPPORT FUNCTIONS
197		1892
		1893	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1894
198	=over 4	1895	=over 4
199		1896
200	=item $fileno = IO::AIO::poll_fileno	1897	=item $fileno = IO::AIO::poll_fileno
201		1898
202	Return the I<request result pipe filehandle>. This filehandle must be	1899	Return the I<request result pipe file descriptor>. This filehandle must be
203	polled for reading by some mechanism outside this module (e.g. Event	1900	polled for reading by some mechanism outside this module (e.g. EV, Glib,
204	or select, see below). If the pipe becomes readable you have to call	1901	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
205	C<poll_cb> to check the results.	1902	you have to call C<poll_cb> to check the results.
206		1903
207	See C<poll_cb> for an example.	1904	See C<poll_cb> for an example.
208		1905
209	=item IO::AIO::poll_cb	1906	=item IO::AIO::poll_cb
210		1907
211	Process all outstanding events on the result pipe. You have to call this	1908	Process some requests that have reached the result phase (i.e. they have
212	regularly. Returns the number of events processed. Returns immediately	1909	been executed but the results are not yet reported). You have to call
213	when no events are outstanding.	1910	this "regularly" to finish outstanding requests.
214		1911
215	You can use Event to multiplex, e.g.:	1912	Returns C<0> if all events could be processed (or there were no
		1913	events to process), or C<-1> if it returned earlier for whatever
		1914	reason. Returns immediately when no events are outstanding. The amount
		1915	of events processed depends on the settings of C<IO::AIO::max_poll_req>,
		1916	C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>.
		1917
		1918	If not all requests were processed for whatever reason, the poll file
		1919	descriptor will still be ready when C<poll_cb> returns, so normally you
		1920	don't have to do anything special to have it called later.
		1921
		1922	Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
		1923	ready, it can be beneficial to call this function from loops which submit
		1924	a lot of requests, to make sure the results get processed when they become
		1925	available and not just when the loop is finished and the event loop takes
		1926	over again. This function returns very fast when there are no outstanding
		1927	requests.
		1928
		1929	Example: Install an Event watcher that automatically calls
		1930	IO::AIO::poll_cb with high priority (more examples can be found in the
		1931	SYNOPSIS section, at the top of this document):
216		1932
217	Event->io (fd => IO::AIO::poll_fileno,	1933	Event->io (fd => IO::AIO::poll_fileno,
218	poll => 'r', async => 1,	1934	poll => 'r', async => 1,
219	cb => \&IO::AIO::poll_cb);	1935	cb => \&IO::AIO::poll_cb);
220		1936
221	=item IO::AIO::poll_wait	1937	=item IO::AIO::poll_wait
222		1938
223	Wait till the result filehandle becomes ready for reading (simply does a	1939	Wait until either at least one request is in the result phase or no
224	select on the filehandle. This is useful if you want to synchronously wait	1940	requests are outstanding anymore.
225	for some requests to finish).	1941
		1942	This is useful if you want to synchronously wait for some requests to
		1943	become ready, without actually handling them.
226		1944
227	See C<nreqs> for an example.	1945	See C<nreqs> for an example.
228		1946
		1947	=item IO::AIO::poll
		1948
		1949	Waits until some requests have been handled.
		1950
		1951	Returns the number of requests processed, but is otherwise strictly
		1952	equivalent to:
		1953
		1954	IO::AIO::poll_wait, IO::AIO::poll_cb
		1955
229	=item IO::AIO::nreqs	1956	=item IO::AIO::flush
230		1957
231	Returns the number of requests currently outstanding.	1958	Wait till all outstanding AIO requests have been handled.
232		1959
233	Example: wait till there are no outstanding requests anymore:	1960	Strictly equivalent to:
234		1961
235	IO::AIO::poll_wait, IO::AIO::poll_cb	1962	IO::AIO::poll_wait, IO::AIO::poll_cb
236	while IO::AIO::nreqs;	1963	while IO::AIO::nreqs;
237		1964
		1965	This function can be useful at program aborts, to make sure outstanding
		1966	I/O has been done (C<IO::AIO> uses an C<END> block which already calls
		1967	this function on normal exits), or when you are merely using C<IO::AIO>
		1968	for its more advanced functions, rather than for async I/O, e.g.:
		1969
		1970	my ($dirs, $nondirs);
		1971	IO::AIO::aio_scandir "/tmp", 0, sub { ($dirs, $nondirs) = @_ };
238	=item IO::AIO::flush	1972	IO::AIO::flush;
		1973	# $dirs, $nondirs are now set
239		1974
240	Wait till all outstanding AIO requests have been handled.	1975	=item IO::AIO::max_poll_reqs $nreqs
		1976
		1977	=item IO::AIO::max_poll_time $seconds
		1978
		1979	These set the maximum number of requests (default C<0>, meaning infinity)
		1980	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1981	the maximum amount of time (default C<0>, meaning infinity) spent in
		1982	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1983	of time C<poll_cb> is allowed to use).
		1984
		1985	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1986	syscall per request processed, which is not normally a problem unless your
		1987	callbacks are really really fast or your OS is really really slow (I am
		1988	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1989
		1990	Setting these is useful if you want to ensure some level of
		1991	interactiveness when perl is not fast enough to process all requests in
		1992	time.
		1993
		1994	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		1995
		1996	Example: Install an Event watcher that automatically calls
		1997	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1998	program get the CPU sometimes even under high AIO load.
		1999
		2000	# try not to spend much more than 0.1s in poll_cb
		2001	IO::AIO::max_poll_time 0.1;
		2002
		2003	# use a low priority so other tasks have priority
		2004	Event->io (fd => IO::AIO::poll_fileno,
		2005	poll => 'r', nice => 1,
		2006	cb => &IO::AIO::poll_cb);
		2007
		2008	=back
		2009
		2010
		2011	=head3 CONTROLLING THE NUMBER OF THREADS
		2012
		2013	=over
241		2014
242	=item IO::AIO::min_parallel $nthreads	2015	=item IO::AIO::min_parallel $nthreads
243		2016
244	Set the minimum number of AIO threads to C<$nthreads>. The default is	2017	Set the minimum number of AIO threads to C<$nthreads>. The current
245	C<1>, which means a single asynchronous operation can be done at one time	2018	default is C<8>, which means eight asynchronous operations can execute
246	(the number of outstanding operations, however, is unlimited).	2019	concurrently at any one time (the number of outstanding requests,
		2020	however, is unlimited).
247		2021
		2022	IO::AIO starts threads only on demand, when an AIO request is queued and
		2023	no free thread exists. Please note that queueing up a hundred requests can
		2024	create demand for a hundred threads, even if it turns out that everything
		2025	is in the cache and could have been processed faster by a single thread.
		2026
248	It is recommended to keep the number of threads low, as some Linux	2027	It is recommended to keep the number of threads relatively low, as some
249	kernel versions will scale negatively with the number of threads (higher	2028	Linux kernel versions will scale negatively with the number of threads
250	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	2029	(higher parallelity => MUCH higher latency). With current Linux 2.6
251	threads should be fine.	2030	versions, 4-32 threads should be fine.
252		2031
253	Under normal circumstances you don't need to call this function, as this	2032	Under most circumstances you don't need to call this function, as the
254	module automatically starts some threads (the exact number might change,	2033	module selects a default that is suitable for low to moderate load.
255	and is currently 4).
256		2034
257	=item IO::AIO::max_parallel $nthreads	2035	=item IO::AIO::max_parallel $nthreads
258		2036
259	Sets the maximum number of AIO threads to C<$nthreads>. If more than	2037	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
260	the specified number of threads are currently running, kill them. This	2038	specified number of threads are currently running, this function kills
261	function blocks until the limit is reached.	2039	them. This function blocks until the limit is reached.
		2040
		2041	While C<$nthreads> are zero, aio requests get queued but not executed
		2042	until the number of threads has been increased again.
262		2043
263	This module automatically runs C<max_parallel 0> at program end, to ensure	2044	This module automatically runs C<max_parallel 0> at program end, to ensure
264	that all threads are killed and that there are no outstanding requests.	2045	that all threads are killed and that there are no outstanding requests.
265		2046
266	Under normal circumstances you don't need to call this function.	2047	Under normal circumstances you don't need to call this function.
267		2048
		2049	=item IO::AIO::max_idle $nthreads
		2050
		2051	Limit the number of threads (default: 4) that are allowed to idle
		2052	(i.e., threads that did not get a request to process within the idle
		2053	timeout (default: 10 seconds). That means if a thread becomes idle while
		2054	C<$nthreads> other threads are also idle, it will free its resources and
		2055	exit.
		2056
		2057	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		2058	to allow for extremely high load situations, but want to free resources
		2059	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		2060
		2061	The default is probably ok in most situations, especially if thread
		2062	creation is fast. If thread creation is very slow on your system you might
		2063	want to use larger values.
		2064
		2065	=item IO::AIO::idle_timeout $seconds
		2066
		2067	Sets the minimum idle timeout (default 10) after which worker threads are
		2068	allowed to exit. SEe C<IO::AIO::max_idle>.
		2069
268	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	2070	=item IO::AIO::max_outstanding $maxreqs
269		2071
270	Sets the maximum number of outstanding requests to C<$nreqs>. If you	2072	Sets the maximum number of outstanding requests to C<$nreqs>. If
271	try to queue up more than this number of requests, the caller will block until	2073	you do queue up more than this number of requests, the next call to
272	some requests have been handled.	2074	C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
		2075	C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
		2076	longer exceeded.
273		2077
274	The default is very large, so normally there is no practical limit. If you	2078	In other words, this setting does not enforce a queue limit, but can be
275	queue up many requests in a loop it it often improves speed if you set	2079	used to make poll functions block if the limit is exceeded.
276	this to a relatively low number, such as C<100>.
277		2080
278	Under normal circumstances you don't need to call this function.	2081	This is a bad function to use in interactive programs because it blocks,
		2082	and a bad way to reduce concurrency because it is inexact. If you need to
		2083	issue many requests without being able to call a poll function on demand,
		2084	it is better to use an C<aio_group> together with a feed callback.
		2085
		2086	Its main use is in scripts without an event loop - when you want to stat a
		2087	lot of files, you can write something like this:
		2088
		2089	IO::AIO::max_outstanding 32;
		2090
		2091	for my $path (...) {
		2092	aio_stat $path , ...;
		2093	IO::AIO::poll_cb;
		2094	}
		2095
		2096	IO::AIO::flush;
		2097
		2098	The call to C<poll_cb> inside the loop will normally return instantly,
		2099	allowing the loop to progress, but as soon as more than C<32> requests
		2100	are in-flight, it will block until some requests have been handled. This
		2101	keeps the loop from pushing a large number of C<aio_stat> requests onto
		2102	the queue (which, with many paths to stat, can use up a lot of memory).
		2103
		2104	The default value for C<max_outstanding> is very large, so there is no
		2105	practical limit on the number of outstanding requests.
279		2106
280	=back	2107	=back
281		2108
		2109
		2110	=head3 STATISTICAL INFORMATION
		2111
		2112	=over
		2113
		2114	=item IO::AIO::nreqs
		2115
		2116	Returns the number of requests currently in the ready, execute or pending
		2117	states (i.e. for which their callback has not been invoked yet).
		2118
		2119	Example: wait till there are no outstanding requests anymore:
		2120
		2121	IO::AIO::poll_wait, IO::AIO::poll_cb
		2122	while IO::AIO::nreqs;
		2123
		2124	=item IO::AIO::nready
		2125
		2126	Returns the number of requests currently in the ready state (not yet
		2127	executed).
		2128
		2129	=item IO::AIO::npending
		2130
		2131	Returns the number of requests currently in the pending state (executed,
		2132	but not yet processed by poll_cb).
		2133
		2134	=back
		2135
		2136
		2137	=head3 SUBSECOND STAT TIME ACCESS
		2138
		2139	Both C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> functions can
		2140	generally find access/modification and change times with subsecond time
		2141	accuracy of the system supports it, but perl's built-in functions only
		2142	return the integer part.
		2143
		2144	The following functions return the timestamps of the most recent
		2145	stat with subsecond precision on most systems and work both after
		2146	C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> calls. Their return
		2147	value is only meaningful after a successful C<stat>/C<lstat> call, or
		2148	during/after a successful C<aio_stat>/C<aio_lstat> callback.
		2149
		2150	This is similar to the L<Time::HiRes> C<stat> functions, but can return
		2151	full resolution without rounding and work with standard perl C<stat>,
		2152	alleviating the need to call the special C<Time::HiRes> functions, which
		2153	do not act like their perl counterparts.
		2154
		2155	On operating systems or file systems where subsecond time resolution is
		2156	not supported or could not be detected, a fractional part of C<0> is
		2157	returned, so it is always safe to call these functions.
		2158
		2159	=over 4
		2160
		2161	=item $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime
		2162
		2163	Return the access, modication, change or birth time, respectively,
		2164	including fractional part. Due to the limited precision of floating point,
		2165	the accuracy on most platforms is only a bit better than milliseconds
		2166	for times around now - see the I<nsec> function family, below, for full
		2167	accuracy.
		2168
		2169	File birth time is only available when the OS and perl support it (on
		2170	FreeBSD and NetBSD at the time of this writing, although support is
		2171	adaptive, so if your OS/perl gains support, IO::AIO can take advantage of
		2172	it). On systems where it isn't available, C<0> is currently returned, but
		2173	this might change to C<undef> in a future version.
		2174
		2175	=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
		2176
		2177	Returns access, modification, change and birth time all in one go, and
		2178	maybe more times in the future version.
		2179
		2180	=item $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec
		2181
		2182	Return the fractional access, modifcation, change or birth time, in nanoseconds,
		2183	as an integer in the range C<0> to C<999999999>.
		2184
		2185	Note that no accessors are provided for access, modification and
		2186	change times - you need to get those from C<stat _> if required (C<int
		2187	IO::AIO::st_atime> and so on will I<not> generally give you the correct
		2188	value).
		2189
		2190	=item $seconds = IO::AIO::st_btimesec
		2191
		2192	The (integral) seconds part of the file birth time, if available.
		2193
		2194	=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
		2195
		2196	Like the functions above, but returns all four times in one go (and maybe
		2197	more in future versions).
		2198
		2199	=item $counter = IO::AIO::st_gen
		2200
		2201	Returns the generation counter (in practice this is just a random number)
		2202	of the file. This is only available on platforms which have this member in
		2203	their C<struct stat> (most BSDs at the time of this writing) and generally
		2204	only to the root usert. If unsupported, C<0> is returned, but this might
		2205	change to C<undef> in a future version.
		2206
		2207	=back
		2208
		2209	Example: print the high resolution modification time of F</etc>, using
		2210	C<stat>, and C<IO::AIO::aio_stat>.
		2211
		2212	if (stat "/etc") {
		2213	printf "stat(/etc) mtime: %f\n", IO::AIO::st_mtime;
		2214	}
		2215
		2216	IO::AIO::aio_stat "/etc", sub {
		2217	$_[0]
		2218	and return;
		2219
		2220	printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::st_mtimensec;
		2221	};
		2222
		2223	IO::AIO::flush;
		2224
		2225	Output of the awbove on my system, showing reduced and full accuracy:
		2226
		2227	stat(/etc) mtime: 1534043702.020808
		2228	aio_stat(/etc) mtime: 1534043702.020807792
		2229
		2230
		2231	=head3 MISCELLANEOUS FUNCTIONS
		2232
		2233	IO::AIO implements some functions that are useful when you want to use
		2234	some "Advanced I/O" function not available to in Perl, without going the
		2235	"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*>
		2236	counterpart.
		2237
		2238	=over 4
		2239
		2240	=item $retval = IO::AIO::fexecve $fh, $argv, $envp
		2241
		2242	A more-or-less direct equivalent to the POSIX C<fexecve> functions, which
		2243	allows you to specify the program to be executed via a file descriptor (or
		2244	handle). Returns C<-1> and sets errno to C<ENOSYS> if not available.
		2245
		2246	=item $retval = IO::AIO::mount $special, $path, $fstype, $flags = 0, $data = undef
		2247
		2248	Calls the GNU/Linux mount syscall with the given arguments. All except
		2249	C<$flags> are strings, and if C<$data> is C<undef>, a C<NULL> will be
		2250	passed.
		2251
		2252	The following values for C<$flags> are available:
		2253
		2254	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>,
		2255	C<IO::AIO::MS_REMOUNT>, C<IO::AIO::MS_MANDLOCK>, C<IO::AIO::MS_DIRSYNC>, C<IO::AIO::MS_NOATIME>,
		2256	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>,
		2257	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>,
		2258	C<IO::AIO::MS_RELATIME>, C<IO::AIO::MS_KERNMOUNT>, C<IO::AIO::MS_I_VERSION>, C<IO::AIO::MS_STRICTATIME>,
		2259	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
		2260	C<IO::AIO::MS_MGC_MSK>.
		2261
		2262	=item $retval = IO::AIO::umount $path, $flags = 0
		2263
		2264	Invokes the GNU/Linux C<umount> or C<umount2> syscalls. Always calls
		2265	C<umount> if C<$flags> is C<0>, otherwqise always tries to call
		2266	C<umount2>.
		2267
		2268	The following C<$flags> are available:
		2269
		2270	C<IO::AIO::MNT_FORCE>, C<IO::AIO::MNT_DETACH>, C<IO::AIO::MNT_EXPIRE> and C<IO::AIO::UMOUNT_NOFOLLOW>.
		2271
		2272	=item $numfd = IO::AIO::get_fdlimit
		2273
		2274	Tries to find the current file descriptor limit and returns it, or
		2275	C<undef> and sets C<$!> in case of an error. The limit is one larger than
		2276	the highest valid file descriptor number.
		2277
		2278	=item IO::AIO::min_fdlimit [$numfd]
		2279
		2280	Try to increase the current file descriptor limit(s) to at least C<$numfd>
		2281	by changing the soft or hard file descriptor resource limit. If C<$numfd>
		2282	is missing, it will try to set a very high limit, although this is not
		2283	recommended when you know the actual minimum that you require.
		2284
		2285	If the limit cannot be raised enough, the function makes a best-effort
		2286	attempt to increase the limit as much as possible, using various
		2287	tricks, while still failing. You can query the resulting limit using
		2288	C<IO::AIO::get_fdlimit>.
		2289
		2290	If an error occurs, returns C<undef> and sets C<$!>, otherwise returns
		2291	true.
		2292
		2293	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		2294
		2295	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		2296	but is blocking (this makes most sense if you know the input data is
		2297	likely cached already and the output filehandle is set to non-blocking
		2298	operations).
		2299
		2300	Returns the number of bytes copied, or C<-1> on error.
		2301
		2302	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		2303
		2304	Simply calls the C<posix_fadvise> function (see its
		2305	manpage for details). The following advice constants are
		2306	available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		2307	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		2308	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		2309
		2310	On systems that do not implement C<posix_fadvise>, this function returns
		2311	ENOSYS, otherwise the return value of C<posix_fadvise>.
		2312
		2313	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		2314
		2315	Simply calls the C<posix_madvise> function (see its
		2316	manpage for details). The following advice constants are
		2317	available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		2318	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>,
		2319	C<IO::AIO::MADV_DONTNEED>.
		2320
		2321	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2322	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2323	will be reduced to fit into the C<$scalar>.
		2324
		2325	On systems that do not implement C<posix_madvise>, this function returns
		2326	ENOSYS, otherwise the return value of C<posix_madvise>.
		2327
		2328	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		2329
		2330	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		2331	$scalar (see its manpage for details). The following protect
		2332	constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		2333	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		2334
		2335	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2336	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2337	will be reduced to fit into the C<$scalar>.
		2338
		2339	On systems that do not implement C<mprotect>, this function returns
		2340	ENOSYS, otherwise the return value of C<mprotect>.
		2341
		2342	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		2343
		2344	Memory-maps a file (or anonymous memory range) and attaches it to the
		2345	given C<$scalar>, which will act like a string scalar. Returns true on
		2346	success, and false otherwise.
		2347
		2348	The scalar must exist, but its contents do not matter - this means you
		2349	cannot use a nonexistant array or hash element. When in doubt, C<undef>
		2350	the scalar first.
		2351
		2352	The only operations allowed on the mmapped scalar are C<substr>/C<vec>,
		2353	which don't change the string length, and most read-only operations such
		2354	as copying it or searching it with regexes and so on.
		2355
		2356	Anything else is unsafe and will, at best, result in memory leaks.
		2357
		2358	The memory map associated with the C<$scalar> is automatically removed
		2359	when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap>
		2360	or C<IO::AIO::munmap> functions are called on it.
		2361
		2362	This calls the C<mmap>(2) function internally. See your system's manual
		2363	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		2364
		2365	The C<$length> must be larger than zero and smaller than the actual
		2366	filesize.
		2367
		2368	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		2369	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		2370
		2371	C<$flags> can be a combination of
		2372	C<IO::AIO::MAP_SHARED> or
		2373	C<IO::AIO::MAP_PRIVATE>,
		2374	or a number of system-specific flags (when not available, the are C<0>):
		2375	C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant),
		2376	C<IO::AIO::MAP_LOCKED>,
		2377	C<IO::AIO::MAP_NORESERVE>,
		2378	C<IO::AIO::MAP_POPULATE>,
		2379	C<IO::AIO::MAP_NONBLOCK>,
		2380	C<IO::AIO::MAP_FIXED>,
		2381	C<IO::AIO::MAP_GROWSDOWN>,
		2382	C<IO::AIO::MAP_32BIT>,
		2383	C<IO::AIO::MAP_HUGETLB>,
		2384	C<IO::AIO::MAP_STACK>,
		2385	C<IO::AIO::MAP_FIXED_NOREPLACE>,
		2386	C<IO::AIO::MAP_SHARED_VALIDATE>,
		2387	C<IO::AIO::MAP_SYNC> or
		2388	C<IO::AIO::MAP_UNINITIALIZED>.
		2389
		2390	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		2391
		2392	C<$offset> is the offset from the start of the file - it generally must be
		2393	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		2394
		2395	Example:
		2396
		2397	use Digest::MD5;
		2398	use IO::AIO;
		2399
		2400	open my $fh, "<verybigfile"
		2401	or die "$!";
		2402
		2403	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		2404	or die "verybigfile: $!";
		2405
		2406	my $fast_md5 = md5 $data;
		2407
		2408	=item IO::AIO::munmap $scalar
		2409
		2410	Removes a previous mmap and undefines the C<$scalar>.
		2411
		2412	=item IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[, $new_address = 0]
		2413
		2414	Calls the Linux-specific mremap(2) system call. The C<$scalar> must have
		2415	been mapped by C<IO::AIO::mmap>, and C<$flags> must currently either be
		2416	C<0> or C<IO::AIO::MREMAP_MAYMOVE>.
		2417
		2418	Returns true if successful, and false otherwise. If the underlying mmapped
		2419	region has changed address, then the true value has the numerical value
		2420	C<1>, otherwise it has the numerical value C<0>:
		2421
		2422	my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE
		2423	or die "mremap: $!";
		2424
		2425	if ($success*1) {
		2426	warn "scalar has chanegd address in memory\n";
		2427	}
		2428
		2429	C<IO::AIO::MREMAP_FIXED> and the C<$new_address> argument are currently
		2430	implemented, but not supported and might go away in a future version.
		2431
		2432	On systems where this call is not supported or is not emulated, this call
		2433	returns falls and sets C<$!> to C<ENOSYS>.
		2434
		2435	=item IO::AIO::mlockall $flags
		2436
		2437	Calls the C<eio_mlockall_sync> function, which is like C<aio_mlockall>,
		2438	but is blocking.
		2439
		2440	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		2441
		2442	Calls the C<munlock> function, undoing the effects of a previous
		2443	C<aio_mlock> call (see its description for details).
		2444
		2445	=item IO::AIO::munlockall
		2446
		2447	Calls the C<munlockall> function.
		2448
		2449	On systems that do not implement C<munlockall>, this function returns
		2450	ENOSYS, otherwise the return value of C<munlockall>.
		2451
		2452	=item $fh = IO::AIO::accept4 $r_fh, $sockaddr, $sockaddr_maxlen, $flags
		2453
		2454	Uses the GNU/Linux C<accept4(2)> syscall, if available, to accept a socket
		2455	and return the new file handle on success, or sets C<$!> and returns
		2456	C<undef> on error.
		2457
		2458	The remote name of the new socket will be stored in C<$sockaddr>, which
		2459	will be extended to allow for at least C<$sockaddr_maxlen> octets. If the
		2460	socket name does not fit into C<$sockaddr_maxlen> octets, this is signaled
		2461	by returning a longer string in C<$sockaddr>, which might or might not be
		2462	truncated.
		2463
		2464	To accept name-less sockets, use C<undef> for C<$sockaddr> and C<0> for
		2465	C<$sockaddr_maxlen>.
		2466
		2467	The main reasons to use this syscall rather than portable C<accept(2)>
		2468	are that you can specify C<SOCK_NONBLOCK> and/or C<SOCK_CLOEXEC>
		2469	flags and you can accept name-less sockets by specifying C<0> for
		2470	C<$sockaddr_maxlen>, which is sadly not possible with perl's interface to
		2471	C<accept>.
		2472
		2473	=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
		2474
		2475	Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or
		2476	C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they
		2477	should be the file offset.
		2478
		2479	C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might
		2480	silently corrupt the data in this case.
		2481
		2482	The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>,
		2483	C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and
		2484	C<IO::AIO::SPLICE_F_GIFT>.
		2485
		2486	See the C<splice(2)> manpage for details.
		2487
		2488	=item IO::AIO::tee $r_fh, $w_fh, $length, $flags
		2489
		2490	Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the
		2491	description for C<IO::AIO::splice> above for details.
		2492
		2493	=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
		2494
		2495	Attempts to query or change the pipe buffer size. Obviously works only
		2496	on pipes, and currently works only on GNU/Linux systems, and fails with
		2497	C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer
		2498	size on other systems, drop me a note.
		2499
		2500	=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
		2501
		2502	This is a direct interface to the Linux L<pipe2(2)> system call. If
		2503	C<$flags> is missing or C<0>, then this should be the same as a call to
		2504	perl's built-in C<pipe> function and create a new pipe, and works on
		2505	systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe
		2506	(..., 4096, O_BINARY)>.
		2507
		2508	If C<$flags> is non-zero, it tries to invoke the pipe2 system call with
		2509	the given flags (Linux 2.6.27, glibc 2.9).
		2510
		2511	On success, the read and write file handles are returned.
		2512
		2513	On error, nothing will be returned. If the pipe2 syscall is missing and
		2514	C<$flags> is non-zero, fails with C<ENOSYS>.
		2515
		2516	Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the
		2517	time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and
		2518	C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported.
		2519
		2520	Example: create a pipe race-free w.r.t. threads and fork:
		2521
		2522	my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
		2523	or die "pipe2: $!\n";
		2524
		2525	=item $fh = IO::AIO::memfd_create $pathname[, $flags]
		2526
		2527	This is a direct interface to the Linux L<memfd_create(2)> system
		2528	call. The (unhelpful) default for C<$flags> is C<0>, but your default
		2529	should be C<IO::AIO::MFD_CLOEXEC>.
		2530
		2531	On success, the new memfd filehandle is returned, otherwise returns
		2532	C<undef>. If the memfd_create syscall is missing, fails with C<ENOSYS>.
		2533
		2534	Please refer to L<memfd_create(2)> for more info on this call.
		2535
		2536	The following C<$flags> values are available: C<IO::AIO::MFD_CLOEXEC>,
		2537	C<IO::AIO::MFD_ALLOW_SEALING>, C<IO::AIO::MFD_HUGETLB>,
		2538	C<IO::AIO::MFD_HUGETLB_2MB> and C<IO::AIO::MFD_HUGETLB_1GB>.
		2539
		2540	Example: create a new memfd.
		2541
		2542	my $fh = IO::AIO::memfd_create "somenameforprocfd", IO::AIO::MFD_CLOEXEC
		2543	or die "memfd_create: $!\n";
		2544
		2545	=item $fh = IO::AIO::pidfd_open $pid[, $flags]
		2546
		2547	This is an interface to the Linux L<pidfd_open(2)> system call. The
		2548	default for C<$flags> is C<0>.
		2549
		2550	On success, a new pidfd filehandle is returned (that is already set to
		2551	close-on-exec), otherwise returns C<undef>. If the syscall is missing,
		2552	fails with C<ENOSYS>.
		2553
		2554	Example: open pid 6341 as pidfd.
		2555
		2556	my $fh = IO::AIO::pidfd_open 6341
		2557	or die "pidfd_open: $!\n";
		2558
		2559	=item $status = IO::AIO::pidfd_send_signal $pidfh, $signal[, $siginfo[, $flags]]
		2560
		2561	This is an interface to the Linux L<pidfd_send_signal> system call. The
		2562	default for C<$siginfo> is C<undef> and the default for C<$flags> is C<0>.
		2563
		2564	Returns the system call status. If the syscall is missing, fails with
		2565	C<ENOSYS>.
		2566
		2567	When specified, C<$siginfo> must be a reference to a hash with one or more
		2568	of the following members:
		2569
		2570	=over
		2571
		2572	=item code - the C<si_code> member
		2573
		2574	=item pid - the C<si_pid> member
		2575
		2576	=item uid - the C<si_uid> member
		2577
		2578	=item value_int - the C<si_value.sival_int> member
		2579
		2580	=item value_ptr - the C<si_value.sival_ptr> member, specified as an integer
		2581
		2582	=back
		2583
		2584	Example: send a SIGKILL to the specified process.
		2585
		2586	my $status = IO::AIO::pidfd_send_signal $pidfh, 9, undef
		2587	and die "pidfd_send_signal: $!\n";
		2588
		2589	Example: send a SIGKILL to the specified process with extra data.
		2590
		2591	my $status = IO::AIO::pidfd_send_signal $pidfh, 9, { code => -1, value_int => 7 }
		2592	and die "pidfd_send_signal: $!\n";
		2593
		2594	=item $fh = IO::AIO::pidfd_getfd $pidfh, $targetfd[, $flags]
		2595
		2596	This is an interface to the Linux L<pidfd_getfd> system call. The default
		2597	for C<$flags> is C<0>.
		2598
		2599	On success, returns a dup'ed copy of the target file descriptor (specified
		2600	as an integer) returned (that is already set to close-on-exec), otherwise
		2601	returns C<undef>. If the syscall is missing, fails with C<ENOSYS>.
		2602
		2603	Example: get a copy of standard error of another process and print soemthing to it.
		2604
		2605	my $errfh = IO::AIO::pidfd_getfd $pidfh, 2
		2606	or die "pidfd_getfd: $!\n";
		2607	print $errfh "stderr\n";
		2608
		2609	=item $fh = IO::AIO::eventfd [$initval, [$flags]]
		2610
		2611	This is a direct interface to the Linux L<eventfd(2)> system call. The
		2612	(unhelpful) defaults for C<$initval> and C<$flags> are C<0> for both.
		2613
		2614	On success, the new eventfd filehandle is returned, otherwise returns
		2615	C<undef>. If the eventfd syscall is missing, fails with C<ENOSYS>.
		2616
		2617	Please refer to L<eventfd(2)> for more info on this call.
		2618
		2619	The following symbol flag values are available: C<IO::AIO::EFD_CLOEXEC>,
		2620	C<IO::AIO::EFD_NONBLOCK> and C<IO::AIO::EFD_SEMAPHORE> (Linux 2.6.30).
		2621
		2622	Example: create a new eventfd filehandle:
		2623
		2624	$fh = IO::AIO::eventfd 0, IO::AIO::EFD_CLOEXEC
		2625	or die "eventfd: $!\n";
		2626
		2627	=item $fh = IO::AIO::timerfd_create $clockid[, $flags]
		2628
		2629	This is a direct interface to the Linux L<timerfd_create(2)> system
		2630	call. The (unhelpful) default for C<$flags> is C<0>, but your default
		2631	should be C<IO::AIO::TFD_CLOEXEC>.
		2632
		2633	On success, the new timerfd filehandle is returned, otherwise returns
		2634	C<undef>. If the timerfd_create syscall is missing, fails with C<ENOSYS>.
		2635
		2636	Please refer to L<timerfd_create(2)> for more info on this call.
		2637
		2638	The following C<$clockid> values are
		2639	available: C<IO::AIO::CLOCK_REALTIME>, C<IO::AIO::CLOCK_MONOTONIC>
		2640	C<IO::AIO::CLOCK_CLOCK_BOOTTIME> (Linux 3.15)
		2641	C<IO::AIO::CLOCK_CLOCK_REALTIME_ALARM> (Linux 3.11) and
		2642	C<IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM> (Linux 3.11).
		2643
		2644	The following C<$flags> values are available (Linux
		2645	2.6.27): C<IO::AIO::TFD_NONBLOCK> and C<IO::AIO::TFD_CLOEXEC>.
		2646
		2647	Example: create a new timerfd and set it to one-second repeated alarms,
		2648	then wait for two alarms:
		2649
		2650	my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC
		2651	or die "timerfd_create: $!\n";
		2652
		2653	defined IO::AIO::timerfd_settime $fh, 0, 1, 1
		2654	or die "timerfd_settime: $!\n";
		2655
		2656	for (1..2) {
		2657	8 == sysread $fh, my $buf, 8
		2658	or die "timerfd read failure\n";
		2659
		2660	printf "number of expirations (likely 1): %d\n",
		2661	unpack "Q", $buf;
		2662	}
		2663
		2664	=item ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
		2665
		2666	This is a direct interface to the Linux L<timerfd_settime(2)> system
		2667	call. Please refer to its manpage for more info on this call.
		2668
		2669	The new itimerspec is specified using two (possibly fractional) second
		2670	values, C<$new_interval> and C<$new_value>).
		2671
		2672	On success, the current interval and value are returned (as per
		2673	C<timerfd_gettime>). On failure, the empty list is returned.
		2674
		2675	The following C<$flags> values are
		2676	available: C<IO::AIO::TFD_TIMER_ABSTIME> and
		2677	C<IO::AIO::TFD_TIMER_CANCEL_ON_SET>.
		2678
		2679	See C<IO::AIO::timerfd_create> for a full example.
		2680
		2681	=item ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
		2682
		2683	This is a direct interface to the Linux L<timerfd_gettime(2)> system
		2684	call. Please refer to its manpage for more info on this call.
		2685
		2686	On success, returns the current values of interval and value for the given
		2687	timerfd (as potentially fractional second values). On failure, the empty
		2688	list is returned.
		2689
		2690	=back
		2691
282	=cut	2692	=cut
283		2693
284	# support function to convert a fd into a perl filehandle
285	sub _fd2fh {
286	return undef if $_[0] < 0;
287
288	# try to be perl5.6-compatible
289	local *AIO_FH;
290	open AIO_FH, "+<&=$_[0]"
291	or return undef;
292
293	*AIO_FH
294	}
295
296	min_parallel 4;	2694	min_parallel 8;
297		2695
298	END {	2696	END { flush }
299	max_parallel 0;
300	}
301		2697
302	1;	2698	1;
303		2699
		2700	=head1 EVENT LOOP INTEGRATION
		2701
		2702	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		2703	automatically into many event loops:
		2704
		2705	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		2706	use AnyEvent::AIO;
		2707
		2708	You can also integrate IO::AIO manually into many event loops, here are
		2709	some examples of how to do this:
		2710
		2711	# EV integration
		2712	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		2713
		2714	# Event integration
		2715	Event->io (fd => IO::AIO::poll_fileno,
		2716	poll => 'r',
		2717	cb => \&IO::AIO::poll_cb);
		2718
		2719	# Glib/Gtk2 integration
		2720	add_watch Glib::IO IO::AIO::poll_fileno,
		2721	in => sub { IO::AIO::poll_cb; 1 };
		2722
		2723	# Tk integration
		2724	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		2725	readable => \&IO::AIO::poll_cb);
		2726
		2727	# Danga::Socket integration
		2728	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		2729	\&IO::AIO::poll_cb);
		2730
		2731	=head2 FORK BEHAVIOUR
		2732
		2733	Usage of pthreads in a program changes the semantics of fork
		2734	considerably. Specifically, only async-safe functions can be called after
		2735	fork. Perl doesn't know about this, so in general, you cannot call fork
		2736	with defined behaviour in perl if pthreads are involved. IO::AIO uses
		2737	pthreads, so this applies, but many other extensions and (for inexplicable
		2738	reasons) perl itself often is linked against pthreads, so this limitation
		2739	applies to quite a lot of perls.
		2740
		2741	This module no longer tries to fight your OS, or POSIX. That means IO::AIO
		2742	only works in the process that loaded it. Forking is fully supported, but
		2743	using IO::AIO in the child is not.
		2744
		2745	You might get around by not I<using> IO::AIO before (or after)
		2746	forking. You could also try to call the L<IO::AIO::reinit> function in the
		2747	child:
		2748
		2749	=over 4
		2750
		2751	=item IO::AIO::reinit
		2752
		2753	Abandons all current requests and I/O threads and simply reinitialises all
		2754	data structures. This is not an operation supported by any standards, but
		2755	happens to work on GNU/Linux and some newer BSD systems.
		2756
		2757	The only reasonable use for this function is to call it after forking, if
		2758	C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
		2759	the process will result in undefined behaviour. Calling it at any time
		2760	will also result in any undefined (by POSIX) behaviour.
		2761
		2762	=back
		2763
		2764	=head2 LINUX-SPECIFIC CALLS
		2765
		2766	When a call is documented as "linux-specific" then this means it
		2767	originated on GNU/Linux. C<IO::AIO> will usually try to autodetect the
		2768	availability and compatibility of such calls regardless of the platform
		2769	it is compiled on, so platforms such as FreeBSD which often implement
		2770	these calls will work. When in doubt, call them and see if they fail wth
		2771	C<ENOSYS>.
		2772
		2773	=head2 MEMORY USAGE
		2774
		2775	Per-request usage:
		2776
		2777	Each aio request uses - depending on your architecture - around 100-200
		2778	bytes of memory. In addition, stat requests need a stat buffer (possibly
		2779	a few hundred bytes), readdir requires a result buffer and so on. Perl
		2780	scalars and other data passed into aio requests will also be locked and
		2781	will consume memory till the request has entered the done state.
		2782
		2783	This is not awfully much, so queuing lots of requests is not usually a
		2784	problem.
		2785
		2786	Per-thread usage:
		2787
		2788	In the execution phase, some aio requests require more memory for
		2789	temporary buffers, and each thread requires a stack and other data
		2790	structures (usually around 16k-128k, depending on the OS).
		2791
		2792	=head1 KNOWN BUGS
		2793
		2794	Known bugs will be fixed in the next release :)
		2795
		2796	=head1 KNOWN ISSUES
		2797
		2798	Calls that try to "import" foreign memory areas (such as C<IO::AIO::mmap>
		2799	or C<IO::AIO::aio_slurp>) do not work with generic lvalues, such as
		2800	non-created hash slots or other scalars I didn't think of. It's best to
		2801	avoid such and either use scalar variables or making sure that the scalar
		2802	exists (e.g. by storing C<undef>) and isn't "funny" (e.g. tied).
		2803
		2804	I am not sure anything can be done about this, so this is considered a
		2805	known issue, rather than a bug.
		2806
304	=head1 SEE ALSO	2807	=head1 SEE ALSO
305		2808
306	L<Coro>, L<Linux::AIO>.	2809	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		2810	more natural syntax and L<IO::FDPass> for file descriptor passing.
307		2811
308	=head1 AUTHOR	2812	=head1 AUTHOR
309		2813
310	Marc Lehmann <schmorp@schmorp.de>	2814	Marc Lehmann <schmorp@schmorp.de>
311	http://home.schmorp.de/	2815	http://home.schmorp.de/

Diff Legend

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