ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.19 by root, Tue Jul 12 11:02:54 2005 UTC vs.
Revision 1.289 by root, Sun Aug 12 05:21:35 2018 UTC

1	=head1 NAME	1	=head1 NAME
2		2
3	IO::AIO - Asynchronous Input/Output	3	IO::AIO - Asynchronous/Advanced Input/Output
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use IO::AIO;	7	use IO::AIO;
8		8
9	aio_open "/etc/passwd", O_RDONLY, 0, sub {	9	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
10	my ($fh) = @_;	10	my $fh = shift
		11	or die "/etc/passwd: $!";
11	...	12	...
12	};	13	};
13		14
14	aio_unlink "/tmp/file", sub { };	15	aio_unlink "/tmp/file", sub { };
15		16
16	aio_read $fh, 30000, 1024, $buffer, 0, sub {	17	aio_read $fh, 30000, 1024, $buffer, 0, sub {
17	$_[0] > 0 or die "read error: $!";	18	$_[0] > 0 or die "read error: $!";
18	};	19	};
19		20
20	# Event	21	# version 2+ has request and group objects
21	Event->io (fd => IO::AIO::poll_fileno,	22	use IO::AIO 2;
22	poll => 'r',
23	cb => \&IO::AIO::poll_cb);
24		23
25	# Glib/Gtk2	24	aioreq_pri 4; # give next request a very high priority
26	add_watch Glib::IO IO::AIO::poll_fileno,	25	my $req = aio_unlink "/tmp/file", sub { };
27	in => sub { IO::AIO::poll_cb, 1 };	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 = 0.5;	176	our $VERSION = 4.5;
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	push @AIO_REQ, qw(aio_busy); # not exported
		200
		201	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
69		202
70	require XSLoader;	203	require XSLoader;
71	XSLoader::load IO::AIO, $VERSION;	204	XSLoader::load ("IO::AIO", $VERSION);
72	}	205	}
73		206
74	=head1 FUNCTIONS	207	=head1 FUNCTIONS
75		208
76	=head2 AIO FUNCTIONS	209	=head2 QUICK OVERVIEW
		210
		211	This section simply lists the prototypes most of the functions for
		212	quick reference. See the following sections for function-by-function
		213	documentation.
		214
		215	aio_wd $pathname, $callback->($wd)
		216	aio_open $pathname, $flags, $mode, $callback->($fh)
		217	aio_close $fh, $callback->($status)
		218	aio_seek $fh,$offset,$whence, $callback->($offs)
		219	aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		220	aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		221	aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		222	aio_readahead $fh,$offset,$length, $callback->($retval)
		223	aio_stat $fh_or_path, $callback->($status)
		224	aio_lstat $fh, $callback->($status)
		225	aio_statvfs $fh_or_path, $callback->($statvfs)
		226	aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		227	aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		228	aio_chmod $fh_or_path, $mode, $callback->($status)
		229	aio_truncate $fh_or_path, $offset, $callback->($status)
		230	aio_allocate $fh, $mode, $offset, $len, $callback->($status)
		231	aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
		232	aio_unlink $pathname, $callback->($status)
		233	aio_mknod $pathname, $mode, $dev, $callback->($status)
		234	aio_link $srcpath, $dstpath, $callback->($status)
		235	aio_symlink $srcpath, $dstpath, $callback->($status)
		236	aio_readlink $pathname, $callback->($link)
		237	aio_realpath $pathname, $callback->($path)
		238	aio_rename $srcpath, $dstpath, $callback->($status)
		239	aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
		240	aio_mkdir $pathname, $mode, $callback->($status)
		241	aio_rmdir $pathname, $callback->($status)
		242	aio_readdir $pathname, $callback->($entries)
		243	aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		244	IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
		245	IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
		246	aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		247	aio_load $pathname, $data, $callback->($status)
		248	aio_copy $srcpath, $dstpath, $callback->($status)
		249	aio_move $srcpath, $dstpath, $callback->($status)
		250	aio_rmtree $pathname, $callback->($status)
		251	aio_fcntl $fh, $cmd, $arg, $callback->($status)
		252	aio_ioctl $fh, $request, $buf, $callback->($status)
		253	aio_sync $callback->($status)
		254	aio_syncfs $fh, $callback->($status)
		255	aio_fsync $fh, $callback->($status)
		256	aio_fdatasync $fh, $callback->($status)
		257	aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		258	aio_pathsync $pathname, $callback->($status)
		259	aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
		260	aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		261	aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		262	aio_mlockall $flags, $callback->($status)
		263	aio_group $callback->(...)
		264	aio_nop $callback->()
		265
		266	$prev_pri = aioreq_pri [$pri]
		267	aioreq_nice $pri_adjust
		268
		269	IO::AIO::poll_wait
		270	IO::AIO::poll_cb
		271	IO::AIO::poll
		272	IO::AIO::flush
		273	IO::AIO::max_poll_reqs $nreqs
		274	IO::AIO::max_poll_time $seconds
		275	IO::AIO::min_parallel $nthreads
		276	IO::AIO::max_parallel $nthreads
		277	IO::AIO::max_idle $nthreads
		278	IO::AIO::idle_timeout $seconds
		279	IO::AIO::max_outstanding $maxreqs
		280	IO::AIO::nreqs
		281	IO::AIO::nready
		282	IO::AIO::npending
		283	$nfd = IO::AIO::get_fdlimit [EXPERIMENTAL]
		284	IO::AIO::min_fdlimit $nfd [EXPERIMENTAL]
		285
		286	IO::AIO::sendfile $ofh, $ifh, $offset, $count
		287	IO::AIO::fadvise $fh, $offset, $len, $advice
		288	IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
		289	IO::AIO::munmap $scalar
		290	IO::AIO::mremap $scalar, $new_length, $flags[, $new_address]
		291	IO::AIO::madvise $scalar, $offset, $length, $advice
		292	IO::AIO::mprotect $scalar, $offset, $length, $protect
		293	IO::AIO::munlock $scalar, $offset = 0, $length = undef
		294	IO::AIO::munlockall
		295
		296	=head2 API NOTES
77		297
78	All the C<aio_*> calls are more or less thin wrappers around the syscall	298	All the C<aio_*> calls are more or less thin wrappers around the syscall
79	with the same name (sans C<aio_>). The arguments are similar or identical,	299	with the same name (sans C<aio_>). The arguments are similar or identical,
80	and they all accept an additional (and optional) C<$callback> argument	300	and they all accept an additional (and optional) C<$callback> argument
81	which must be a code reference. This code reference will get called with	301	which must be a code reference. This code reference will be called after
82	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	302	the syscall has been executed in an asynchronous fashion. The results
83	perl, which usually delivers "false") as it's sole argument when the given	303	of the request will be passed as arguments to the callback (and, if an
84	syscall has been executed asynchronously.	304	error occured, in C<$!>) - for most requests the syscall return code (e.g.
		305	most syscalls return C<-1> on error, unlike perl, which usually delivers
		306	"false").
85		307
86	All functions that expect a filehandle will also accept a file descriptor.	308	Some requests (such as C<aio_readdir>) pass the actual results and
		309	communicate failures by passing C<undef>.
87		310
		311	All functions expecting a filehandle keep a copy of the filehandle
		312	internally until the request has finished.
		313
		314	All functions return request objects of type L<IO::AIO::REQ> that allow
		315	further manipulation of those requests while they are in-flight.
		316
88	The filenames you pass to these routines I<must> be absolute. The reason	317	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	318	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	319	current working directory could have changed. Alternatively, you can
91	never change the current working directory.	320	make sure that you never change the current working directory anywhere
		321	in the program and then use relative paths. You can also take advantage
		322	of IO::AIOs working directory abstraction, that lets you specify paths
		323	relative to some previously-opened "working directory object" - see the
		324	description of the C<IO::AIO::WD> class later in this document.
		325
		326	To encode pathnames as octets, either make sure you either: a) always pass
		327	in filenames you got from outside (command line, readdir etc.) without
		328	tinkering, b) are in your native filesystem encoding, c) use the Encode
		329	module and encode your pathnames to the locale (or other) encoding in
		330	effect in the user environment, d) use Glib::filename_from_unicode on
		331	unicode filenames or e) use something else to ensure your scalar has the
		332	correct contents.
		333
		334	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		335	handles correctly whether it is set or not.
		336
		337	=head2 AIO REQUEST FUNCTIONS
92		338
93	=over 4	339	=over 4
94		340
		341	=item $prev_pri = aioreq_pri [$pri]
		342
		343	Returns the priority value that would be used for the next request and, if
		344	C<$pri> is given, sets the priority for the next aio request.
		345
		346	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		347	and C<4>, respectively. Requests with higher priority will be serviced
		348	first.
		349
		350	The priority will be reset to C<0> after each call to one of the C<aio_*>
		351	functions.
		352
		353	Example: open a file with low priority, then read something from it with
		354	higher priority so the read request is serviced before other low priority
		355	open requests (potentially spamming the cache):
		356
		357	aioreq_pri -3;
		358	aio_open ..., sub {
		359	return unless $_[0];
		360
		361	aioreq_pri -2;
		362	aio_read $_[0], ..., sub {
		363	...
		364	};
		365	};
		366
		367
		368	=item aioreq_nice $pri_adjust
		369
		370	Similar to C<aioreq_pri>, but subtracts the given value from the current
		371	priority, so the effect is cumulative.
		372
		373
95	=item aio_open $pathname, $flags, $mode, $callback	374	=item aio_open $pathname, $flags, $mode, $callback->($fh)
96		375
97	Asynchronously open or create a file and call the callback with a newly	376	Asynchronously open or create a file and call the callback with a newly
98	created filehandle for the file.	377	created filehandle for the file (or C<undef> in case of an error).
99		378
100	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,	379	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
101	for an explanation.	380	for an explanation.
102		381
103	The C<$mode> argument is a bitmask. See the C<Fcntl> module for a	382	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
104	list. They are the same as used in C<sysopen>.	383	list. They are the same as used by C<sysopen>.
		384
		385	Likewise, C<$mode> specifies the mode of the newly created file, if it
		386	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
		387	except that it is mandatory (i.e. use C<0> if you don't create new files,
		388	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		389	by the umask in effect then the request is being executed, so better never
		390	change the umask.
105		391
106	Example:	392	Example:
107		393
108	aio_open "/etc/passwd", O_RDONLY, 0, sub {	394	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
109	if ($_[0]) {	395	if ($_[0]) {
110	print "open successful, fh is $_[0]\n";	396	print "open successful, fh is $_[0]\n";
111	...	397	...
112	} else {	398	} else {
113	die "open failed: $!\n";	399	die "open failed: $!\n";
114	}	400	}
115	};	401	};
116		402
		403	In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
		404	C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
		405	following POSIX and non-POSIX constants are available (missing ones on
		406	your system are, as usual, C<0>):
		407
		408	C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
		409	C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
		410	C<O_RSYNC>, C<O_SYNC>, C<O_PATH>, C<O_TMPFILE>, C<O_TTY_INIT> and C<O_ACCMODE>.
		411
		412
117	=item aio_close $fh, $callback	413	=item aio_close $fh, $callback->($status)
118		414
119	Asynchronously close a file and call the callback with the result	415	Asynchronously close a file and call the callback with the result
120	code. I<WARNING:> although accepted, you should not pass in a perl	416	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		417
		418	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
		419	closing the file descriptor associated with the filehandle itself.
		420
		421	Therefore, C<aio_close> will not close the filehandle - instead it will
		422	use dup2 to overwrite the file descriptor with the write-end of a pipe
		423	(the pipe fd will be created on demand and will be cached).
		424
		425	Or in other words: the file descriptor will be closed, but it will not be
		426	free for reuse until the perl filehandle is closed.
		427
		428	=cut
		429
		430	=item aio_seek $fh, $offset, $whence, $callback->($offs)
		431
		432	Seeks the filehandle to the new C<$offset>, similarly to perl's
		433	C<sysseek>. The C<$whence> can use the traditional values (C<0> for
		434	C<IO::AIO::SEEK_SET>, C<1> for C<IO::AIO::SEEK_CUR> or C<2> for
		435	C<IO::AIO::SEEK_END>).
		436
		437	The resulting absolute offset will be passed to the callback, or C<-1> in
		438	case of an error.
		439
		440	In theory, the C<$whence> constants could be different than the
		441	corresponding values from L<Fcntl>, but perl guarantees they are the same,
		442	so don't panic.
		443
		444	As a GNU/Linux (and maybe Solaris) extension, also the constants
		445	C<IO::AIO::SEEK_DATA> and C<IO::AIO::SEEK_HOLE> are available, if they
		446	could be found. No guarantees about suitability for use in C<aio_seek> or
		447	Perl's C<sysseek> can be made though, although I would naively assume they
		448	"just work".
		449
125	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	450	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
126		451
127	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	452	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
128		453
129	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	454	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	455	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> and
131	callback without the actual number of bytes read (or -1 on error, just	456	calls the callback with the actual number of bytes transferred (or -1 on
132	like the syscall).	457	error, just like the syscall).
		458
		459	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		460	offset plus the actual number of bytes read.
		461
		462	If C<$offset> is undefined, then the current file descriptor offset will
		463	be used (and updated), otherwise the file descriptor offset will not be
		464	changed by these calls.
		465
		466	If C<$length> is undefined in C<aio_write>, use the remaining length of
		467	C<$data>.
		468
		469	If C<$dataoffset> is less than zero, it will be counted from the end of
		470	C<$data>.
		471
		472	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		473	is outstanding. Modifying it can result in segfaults or World War III (if
		474	the necessary/optional hardware is installed).
133		475
134	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	476	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
135	offset C<0> within the scalar:	477	offset C<0> within the scalar:
136		478
137	aio_read $fh, 7, 15, $buffer, 0, sub {	479	aio_read $fh, 7, 15, $buffer, 0, sub {
138	$_[0] > 0 or die "read error: $!";	480	$_[0] > 0 or die "read error: $!";
139	print "read $_[0] bytes: <$buffer>\n";	481	print "read $_[0] bytes: <$buffer>\n";
140	};	482	};
141		483
		484
		485	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		486
		487	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		488	reading at byte offset C<$in_offset>, and starts writing at the current
		489	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		490	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		491	other. The same C<$in_fh> works fine though, as this function does not
		492	move or use the file offset of C<$in_fh>.
		493
		494	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		495	are written, and there is no way to find out how many more bytes have been
		496	read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
		497	number of bytes written to C<$out_fh>. Only if the result value equals
		498	C<$length> one can assume that C<$length> bytes have been read.
		499
		500	Unlike with other C<aio_> functions, it makes a lot of sense to use
		501	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		502	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		503	the socket I/O will be non-blocking. Note, however, that you can run
		504	into a trap where C<aio_sendfile> reads some data with readahead, then
		505	fails to write all data, and when the socket is ready the next time, the
		506	data in the cache is already lost, forcing C<aio_sendfile> to again hit
		507	the disk. Explicit C<aio_read> + C<aio_write> let's you better control
		508	resource usage.
		509
		510	This call tries to make use of a native C<sendfile>-like syscall to
		511	provide zero-copy operation. For this to work, C<$out_fh> should refer to
		512	a socket, and C<$in_fh> should refer to an mmap'able file.
		513
		514	If a native sendfile cannot be found or it fails with C<ENOSYS>,
		515	C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
		516	C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
		517	type of filehandle regardless of the limitations of the operating system.
		518
		519	As native sendfile syscalls (as practically any non-POSIX interface hacked
		520	together in a hurry to improve benchmark numbers) tend to be rather buggy
		521	on many systems, this implementation tries to work around some known bugs
		522	in Linux and FreeBSD kernels (probably others, too), but that might fail,
		523	so you really really should check the return value of C<aio_sendfile> -
		524	fewer bytes than expected might have been transferred.
		525
		526
142	=item aio_readahead $fh,$offset,$length, $callback	527	=item aio_readahead $fh,$offset,$length, $callback->($retval)
143		528
144	Asynchronously reads the specified byte range into the page cache, using
145	the C<readahead> syscall. If that syscall doesn't exist (likely if your OS
146	isn't Linux) the status will be C<-1> and C<$!> is set to ENOSYS.
147
148	readahead() populates the page cache with data from a file so that	529	C<aio_readahead> populates the page cache with data from a file so that
149	subsequent reads from that file will not block on disk I/O. The C<$offset>	530	subsequent reads from that file will not block on disk I/O. The C<$offset>
150	argument specifies the starting point from which data is to be read and	531	argument specifies the starting point from which data is to be read and
151	C<$length> specifies the number of bytes to be read. I/O is performed in	532	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	533	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	534	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	535	(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.	536	file. The current file offset of the file is left unchanged.
156		537
		538	If that syscall doesn't exist (likely if your kernel isn't Linux) it will
		539	be emulated by simply reading the data, which would have a similar effect.
		540
		541
157	=item aio_stat $fh_or_path, $callback	542	=item aio_stat $fh_or_path, $callback->($status)
158		543
159	=item aio_lstat $fh, $callback	544	=item aio_lstat $fh, $callback->($status)
160		545
161	Works like perl's C<stat> or C<lstat> in void context. The callback will	546	Works like perl's C<stat> or C<lstat> in void context. The callback will
162	be called after the stat and the results will be available using C<stat _>	547	be called after the stat and the results will be available using C<stat _>
163	or C<-s _> etc...	548	or C<-s _> etc...
164		549
…		…
166	for an explanation.	551	for an explanation.
167		552
168	Currently, the stats are always 64-bit-stats, i.e. instead of returning an	553	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	554	error when stat'ing a large file, the results will be silently truncated
170	unless perl itself is compiled with large file support.	555	unless perl itself is compiled with large file support.
		556
		557	To help interpret the mode and dev/rdev stat values, IO::AIO offers the
		558	following constants and functions (if not implemented, the constants will
		559	be C<0> and the functions will either C<croak> or fall back on traditional
		560	behaviour).
		561
		562	C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
		563	C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
		564	C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
		565
		566	To access higher resolution stat timestamps, see L<SUBSECOND STAT TIME
		567	ACCESS>.
171		568
172	Example: Print the length of F</etc/passwd>:	569	Example: Print the length of F</etc/passwd>:
173		570
174	aio_stat "/etc/passwd", sub {	571	aio_stat "/etc/passwd", sub {
175	$_[0] and die "stat failed: $!";	572	$_[0] and die "stat failed: $!";
176	print "size is ", -s _, "\n";	573	print "size is ", -s _, "\n";
177	};	574	};
178		575
		576
		577	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		578
		579	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		580	whether a file handle or path was passed.
		581
		582	On success, the callback is passed a hash reference with the following
		583	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		584	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		585	is passed.
		586
		587	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		588	C<ST_NOSUID>.
		589
		590	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		591	their correct value when available, or to C<0> on systems that do
		592	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		593	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		594	C<ST_NODIRATIME> and C<ST_RELATIME>.
		595
		596	Example: stat C</wd> and dump out the data if successful.
		597
		598	aio_statvfs "/wd", sub {
		599	my $f = $_[0]
		600	or die "statvfs: $!";
		601
		602	use Data::Dumper;
		603	say Dumper $f;
		604	};
		605
		606	# result:
		607	{
		608	bsize => 1024,
		609	bfree => 4333064312,
		610	blocks => 10253828096,
		611	files => 2050765568,
		612	flag => 4096,
		613	favail => 2042092649,
		614	bavail => 4333064312,
		615	ffree => 2042092649,
		616	namemax => 255,
		617	frsize => 1024,
		618	fsid => 1810
		619	}
		620
		621	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		622
		623	Works like perl's C<utime> function (including the special case of $atime
		624	and $mtime being undef). Fractional times are supported if the underlying
		625	syscalls support them.
		626
		627	When called with a pathname, uses utimes(2) if available, otherwise
		628	utime(2). If called on a file descriptor, uses futimes(2) if available,
		629	otherwise returns ENOSYS, so this is not portable.
		630
		631	Examples:
		632
		633	# set atime and mtime to current time (basically touch(1)):
		634	aio_utime "path", undef, undef;
		635	# set atime to current time and mtime to beginning of the epoch:
		636	aio_utime "path", time, undef; # undef==0
		637
		638
		639	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		640
		641	Works like perl's C<chown> function, except that C<undef> for either $uid
		642	or $gid is being interpreted as "do not change" (but -1 can also be used).
		643
		644	Examples:
		645
		646	# same as "chown root path" in the shell:
		647	aio_chown "path", 0, -1;
		648	# same as above:
		649	aio_chown "path", 0, undef;
		650
		651
		652	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		653
		654	Works like truncate(2) or ftruncate(2).
		655
		656
		657	=item aio_allocate $fh, $mode, $offset, $len, $callback->($status)
		658
		659	Allocates or frees disk space according to the C<$mode> argument. See the
		660	linux C<fallocate> documentation for details.
		661
		662	C<$mode> is usually C<0> or C<IO::AIO::FALLOC_FL_KEEP_SIZE> to allocate
		663	space, or C<IO::AIO::FALLOC_FL_PUNCH_HOLE | IO::AIO::FALLOC_FL_KEEP_SIZE>,
		664	to deallocate a file range.
		665
		666	IO::AIO also supports C<FALLOC_FL_COLLAPSE_RANGE>, to remove a range
		667	(without leaving a hole), C<FALLOC_FL_ZERO_RANGE>, to zero a range,
		668	C<FALLOC_FL_INSERT_RANGE> to insert a range and C<FALLOC_FL_UNSHARE_RANGE>
		669	to unshare shared blocks (see your L<fallocate(2)> manpage).
		670
		671	The file system block size used by C<fallocate> is presumably the
		672	C<f_bsize> returned by C<statvfs>, but different filesystems and filetypes
		673	can dictate other limitations.
		674
		675	If C<fallocate> isn't available or cannot be emulated (currently no
		676	emulation will be attempted), passes C<-1> and sets C<$!> to C<ENOSYS>.
		677
		678
		679	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		680
		681	Works like perl's C<chmod> function.
		682
		683
179	=item aio_unlink $pathname, $callback	684	=item aio_unlink $pathname, $callback->($status)
180		685
181	Asynchronously unlink (delete) a file and call the callback with the	686	Asynchronously unlink (delete) a file and call the callback with the
182	result code.	687	result code.
183		688
		689
		690	=item aio_mknod $pathname, $mode, $dev, $callback->($status)
		691
		692	[EXPERIMENTAL]
		693
		694	Asynchronously create a device node (or fifo). See mknod(2).
		695
		696	The only (POSIX-) portable way of calling this function is:
		697
		698	aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		699
		700	See C<aio_stat> for info about some potentially helpful extra constants
		701	and functions.
		702
		703	=item aio_link $srcpath, $dstpath, $callback->($status)
		704
		705	Asynchronously create a new link to the existing object at C<$srcpath> at
		706	the path C<$dstpath> and call the callback with the result code.
		707
		708
		709	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		710
		711	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		712	the path C<$dstpath> and call the callback with the result code.
		713
		714
		715	=item aio_readlink $pathname, $callback->($link)
		716
		717	Asynchronously read the symlink specified by C<$path> and pass it to
		718	the callback. If an error occurs, nothing or undef gets passed to the
		719	callback.
		720
		721
		722	=item aio_realpath $pathname, $callback->($path)
		723
		724	Asynchronously make the path absolute and resolve any symlinks in
		725	C<$path>. The resulting path only consists of directories (same as
		726	L<Cwd::realpath>).
		727
		728	This request can be used to get the absolute path of the current working
		729	directory by passing it a path of F<.> (a single dot).
		730
		731
		732	=item aio_rename $srcpath, $dstpath, $callback->($status)
		733
		734	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		735	rename(2) and call the callback with the result code.
		736
		737	On systems that support the AIO::WD working directory abstraction
		738	natively, the case C<[$wd, "."]> as C<$srcpath> is specialcased - instead
		739	of failing, C<rename> is called on the absolute path of C<$wd>.
		740
		741
		742	=item aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
		743
		744	Basically a version of C<aio_rename> with an additional C<$flags>
		745	argument. Calling this with C<$flags=0> is the same as calling
		746	C<aio_rename>.
		747
		748	Non-zero flags are currently only supported on GNU/Linux systems that
		749	support renameat2. Other systems fail with C<ENOSYS> in this case.
		750
		751	The following constants are available (missing ones are, as usual C<0>),
		752	see renameat2(2) for details:
		753
		754	C<IO::AIO::RENAME_NOREPLACE>, C<IO::AIO::RENAME_EXCHANGE>
		755	and C<IO::AIO::RENAME_WHITEOUT>.
		756
		757
		758	=item aio_mkdir $pathname, $mode, $callback->($status)
		759
		760	Asynchronously mkdir (create) a directory and call the callback with
		761	the result code. C<$mode> will be modified by the umask at the time the
		762	request is executed, so do not change your umask.
		763
		764
		765	=item aio_rmdir $pathname, $callback->($status)
		766
		767	Asynchronously rmdir (delete) a directory and call the callback with the
		768	result code.
		769
		770	On systems that support the AIO::WD working directory abstraction
		771	natively, the case C<[$wd, "."]> is specialcased - instead of failing,
		772	C<rmdir> is called on the absolute path of C<$wd>.
		773
		774
		775	=item aio_readdir $pathname, $callback->($entries)
		776
		777	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		778	directory (i.e. opendir + readdir + closedir). The entries will not be
		779	sorted, and will B<NOT> include the C<.> and C<..> entries.
		780
		781	The callback is passed a single argument which is either C<undef> or an
		782	array-ref with the filenames.
		783
		784
		785	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		786
		787	Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
		788	tune behaviour and output format. In case of an error, C<$entries> will be
		789	C<undef>.
		790
		791	The flags are a combination of the following constants, ORed together (the
		792	flags will also be passed to the callback, possibly modified):
		793
		794	=over 4
		795
		796	=item IO::AIO::READDIR_DENTS
		797
		798	Normally the callback gets an arrayref consisting of names only (as
		799	with C<aio_readdir>). If this flag is set, then the callback gets an
		800	arrayref with C<[$name, $type, $inode]> arrayrefs, each describing a
		801	single directory entry in more detail:
		802
		803	C<$name> is the name of the entry.
		804
		805	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		806
		807	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		808	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		809	C<IO::AIO::DT_WHT>.
		810
		811	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need
		812	to know, you have to run stat yourself. Also, for speed/memory reasons,
		813	the C<$type> scalars are read-only: you must not modify them.
		814
		815	C<$inode> is the inode number (which might not be exact on systems with 64
		816	bit inode numbers and 32 bit perls). This field has unspecified content on
		817	systems that do not deliver the inode information.
		818
		819	=item IO::AIO::READDIR_DIRS_FIRST
		820
		821	When this flag is set, then the names will be returned in an order where
		822	likely directories come first, in optimal stat order. This is useful when
		823	you need to quickly find directories, or you want to find all directories
		824	while avoiding to stat() each entry.
		825
		826	If the system returns type information in readdir, then this is used
		827	to find directories directly. Otherwise, likely directories are names
		828	beginning with ".", or otherwise names with no dots, of which names with
		829	short names are tried first.
		830
		831	=item IO::AIO::READDIR_STAT_ORDER
		832
		833	When this flag is set, then the names will be returned in an order
		834	suitable for stat()'ing each one. That is, when you plan to stat() most or
		835	all files in the given directory, then the returned order will likely be
		836	faster.
		837
		838	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified,
		839	then the likely dirs come first, resulting in a less optimal stat order
		840	for stat'ing all entries, but likely a more optimal order for finding
		841	subdirectories.
		842
		843	=item IO::AIO::READDIR_FOUND_UNKNOWN
		844
		845	This flag should not be set when calling C<aio_readdirx>. Instead, it
		846	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		847	C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
		848	C<$type>'s are known, which can be used to speed up some algorithms.
		849
		850	=back
		851
		852
		853	=item aio_slurp $pathname, $offset, $length, $data, $callback->($status)
		854
		855	Opens, reads and closes the given file. The data is put into C<$data>,
		856	which is resized as required.
		857
		858	If C<$offset> is negative, then it is counted from the end of the file.
		859
		860	If C<$length> is zero, then the remaining length of the file is
		861	used. Also, in this case, the same limitations to modifying C<$data> apply
		862	as when IO::AIO::mmap is used, i.e. it must only be modified in-place
		863	with C<substr>. If the size of the file is known, specifying a non-zero
		864	C<$length> results in a performance advantage.
		865
		866	This request is similar to the older C<aio_load> request, but since it is
		867	a single request, it might be more efficient to use.
		868
		869	Example: load F</etc/passwd> into C<$passwd>.
		870
		871	my $passwd;
		872	aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
		873	$_[0] >= 0
		874	or die "/etc/passwd: $!\n";
		875
		876	printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
		877	print $passwd;
		878	};
		879	IO::AIO::flush;
		880
		881
		882	=item aio_load $pathname, $data, $callback->($status)
		883
		884	This is a composite request that tries to fully load the given file into
		885	memory. Status is the same as with aio_read.
		886
		887	Using C<aio_slurp> might be more efficient, as it is a single request.
		888
		889	=cut
		890
		891	sub aio_load($$;$) {
		892	my ($path, undef, $cb) = @_;
		893	my $data = \$_[1];
		894
		895	my $pri = aioreq_pri;
		896	my $grp = aio_group $cb;
		897
		898	aioreq_pri $pri;
		899	add $grp aio_open $path, O_RDONLY, 0, sub {
		900	my $fh = shift
		901	or return $grp->result (-1);
		902
		903	aioreq_pri $pri;
		904	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		905	$grp->result ($_[0]);
		906	};
		907	};
		908
		909	$grp
		910	}
		911
		912	=item aio_copy $srcpath, $dstpath, $callback->($status)
		913
		914	Try to copy the I<file> (directories not supported as either source or
		915	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		916	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		917
		918	Existing destination files will be truncated.
		919
		920	This is a composite request that creates the destination file with
		921	mode 0200 and copies the contents of the source file into it using
		922	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		923	uid/gid, in that order.
		924
		925	If an error occurs, the partial destination file will be unlinked, if
		926	possible, except when setting atime, mtime, access mode and uid/gid, where
		927	errors are being ignored.
		928
		929	=cut
		930
		931	sub aio_copy($$;$) {
		932	my ($src, $dst, $cb) = @_;
		933
		934	my $pri = aioreq_pri;
		935	my $grp = aio_group $cb;
		936
		937	aioreq_pri $pri;
		938	add $grp aio_open $src, O_RDONLY, 0, sub {
		939	if (my $src_fh = $_[0]) {
		940	my @stat = stat $src_fh; # hmm, might block over nfs?
		941
		942	aioreq_pri $pri;
		943	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		944	if (my $dst_fh = $_[0]) {
		945	aioreq_pri $pri;
		946	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		947	if ($_[0] == $stat[7]) {
		948	$grp->result (0);
		949	close $src_fh;
		950
		951	my $ch = sub {
		952	aioreq_pri $pri;
		953	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		954	aioreq_pri $pri;
		955	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		956	aioreq_pri $pri;
		957	add $grp aio_close $dst_fh;
		958	}
		959	};
		960	};
		961
		962	aioreq_pri $pri;
		963	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		964	if ($_[0] < 0 && $! == ENOSYS) {
		965	aioreq_pri $pri;
		966	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		967	} else {
		968	$ch->();
		969	}
		970	};
		971	} else {
		972	$grp->result (-1);
		973	close $src_fh;
		974	close $dst_fh;
		975
		976	aioreq $pri;
		977	add $grp aio_unlink $dst;
		978	}
		979	};
		980	} else {
		981	$grp->result (-1);
		982	}
		983	},
		984
		985	} else {
		986	$grp->result (-1);
		987	}
		988	};
		989
		990	$grp
		991	}
		992
		993	=item aio_move $srcpath, $dstpath, $callback->($status)
		994
		995	Try to move the I<file> (directories not supported as either source or
		996	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		997	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		998
		999	This is a composite request that tries to rename(2) the file first; if
		1000	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		1001	that is successful, unlinks the C<$srcpath>.
		1002
		1003	=cut
		1004
		1005	sub aio_move($$;$) {
		1006	my ($src, $dst, $cb) = @_;
		1007
		1008	my $pri = aioreq_pri;
		1009	my $grp = aio_group $cb;
		1010
		1011	aioreq_pri $pri;
		1012	add $grp aio_rename $src, $dst, sub {
		1013	if ($_[0] && $! == EXDEV) {
		1014	aioreq_pri $pri;
		1015	add $grp aio_copy $src, $dst, sub {
		1016	$grp->result ($_[0]);
		1017
		1018	unless ($_[0]) {
		1019	aioreq_pri $pri;
		1020	add $grp aio_unlink $src;
		1021	}
		1022	};
		1023	} else {
		1024	$grp->result ($_[0]);
		1025	}
		1026	};
		1027
		1028	$grp
		1029	}
		1030
		1031	=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		1032
		1033	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		1034	efficiently separate the entries of directory C<$path> into two sets of
		1035	names, directories you can recurse into (directories), and ones you cannot
		1036	recurse into (everything else, including symlinks to directories).
		1037
		1038	C<aio_scandir> is a composite request that generates many sub requests.
		1039	C<$maxreq> specifies the maximum number of outstanding aio requests that
		1040	this function generates. If it is C<< <= 0 >>, then a suitable default
		1041	will be chosen (currently 4).
		1042
		1043	On error, the callback is called without arguments, otherwise it receives
		1044	two array-refs with path-relative entry names.
		1045
		1046	Example:
		1047
		1048	aio_scandir $dir, 0, sub {
		1049	my ($dirs, $nondirs) = @_;
		1050	print "real directories: @$dirs\n";
		1051	print "everything else: @$nondirs\n";
		1052	};
		1053
		1054	Implementation notes.
		1055
		1056	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		1057
		1058	If readdir returns file type information, then this is used directly to
		1059	find directories.
		1060
		1061	Otherwise, after reading the directory, the modification time, size etc.
		1062	of the directory before and after the readdir is checked, and if they
		1063	match (and isn't the current time), the link count will be used to decide
		1064	how many entries are directories (if >= 2). Otherwise, no knowledge of the
		1065	number of subdirectories will be assumed.
		1066
		1067	Then entries will be sorted into likely directories a non-initial dot
		1068	currently) and likely non-directories (see C<aio_readdirx>). Then every
		1069	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
		1070	in order of their inode numbers. If that succeeds, it assumes that the
		1071	entry is a directory or a symlink to directory (which will be checked
		1072	separately). This is often faster than stat'ing the entry itself because
		1073	filesystems might detect the type of the entry without reading the inode
		1074	data (e.g. ext2fs filetype feature), even on systems that cannot return
		1075	the filetype information on readdir.
		1076
		1077	If the known number of directories (link count - 2) has been reached, the
		1078	rest of the entries is assumed to be non-directories.
		1079
		1080	This only works with certainty on POSIX (= UNIX) filesystems, which
		1081	fortunately are the vast majority of filesystems around.
		1082
		1083	It will also likely work on non-POSIX filesystems with reduced efficiency
		1084	as those tend to return 0 or 1 as link counts, which disables the
		1085	directory counting heuristic.
		1086
		1087	=cut
		1088
		1089	sub aio_scandir($$;$) {
		1090	my ($path, $maxreq, $cb) = @_;
		1091
		1092	my $pri = aioreq_pri;
		1093
		1094	my $grp = aio_group $cb;
		1095
		1096	$maxreq = 4 if $maxreq <= 0;
		1097
		1098	# get a wd object
		1099	aioreq_pri $pri;
		1100	add $grp aio_wd $path, sub {
		1101	$_[0]
		1102	or return $grp->result ();
		1103
		1104	my $wd = [shift, "."];
		1105
		1106	# stat once
		1107	aioreq_pri $pri;
		1108	add $grp aio_stat $wd, sub {
		1109	return $grp->result () if $_[0];
		1110	my $now = time;
		1111	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		1112
		1113	# read the directory entries
		1114	aioreq_pri $pri;
		1115	add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
		1116	my $entries = shift
		1117	or return $grp->result ();
		1118
		1119	# stat the dir another time
		1120	aioreq_pri $pri;
		1121	add $grp aio_stat $wd, sub {
		1122	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		1123
		1124	my $ndirs;
		1125
		1126	# take the slow route if anything looks fishy
		1127	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		1128	$ndirs = -1;
		1129	} else {
		1130	# if nlink == 2, we are finished
		1131	# for non-posix-fs's, we rely on nlink < 2
		1132	$ndirs = (stat _)[3] - 2
		1133	or return $grp->result ([], $entries);
		1134	}
		1135
		1136	my (@dirs, @nondirs);
		1137
		1138	my $statgrp = add $grp aio_group sub {
		1139	$grp->result (\@dirs, \@nondirs);
		1140	};
		1141
		1142	limit $statgrp $maxreq;
		1143	feed $statgrp sub {
		1144	return unless @$entries;
		1145	my $entry = shift @$entries;
		1146
		1147	aioreq_pri $pri;
		1148	$wd->[1] = "$entry/.";
		1149	add $statgrp aio_stat $wd, sub {
		1150	if ($_[0] < 0) {
		1151	push @nondirs, $entry;
		1152	} else {
		1153	# need to check for real directory
		1154	aioreq_pri $pri;
		1155	$wd->[1] = $entry;
		1156	add $statgrp aio_lstat $wd, sub {
		1157	if (-d _) {
		1158	push @dirs, $entry;
		1159
		1160	unless (--$ndirs) {
		1161	push @nondirs, @$entries;
		1162	feed $statgrp;
		1163	}
		1164	} else {
		1165	push @nondirs, $entry;
		1166	}
		1167	}
		1168	}
		1169	};
		1170	};
		1171	};
		1172	};
		1173	};
		1174	};
		1175
		1176	$grp
		1177	}
		1178
		1179	=item aio_rmtree $pathname, $callback->($status)
		1180
		1181	Delete a directory tree starting (and including) C<$path>, return the
		1182	status of the final C<rmdir> only. This is a composite request that
		1183	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1184	everything else.
		1185
		1186	=cut
		1187
		1188	sub aio_rmtree;
		1189	sub aio_rmtree($;$) {
		1190	my ($path, $cb) = @_;
		1191
		1192	my $pri = aioreq_pri;
		1193	my $grp = aio_group $cb;
		1194
		1195	aioreq_pri $pri;
		1196	add $grp aio_scandir $path, 0, sub {
		1197	my ($dirs, $nondirs) = @_;
		1198
		1199	my $dirgrp = aio_group sub {
		1200	add $grp aio_rmdir $path, sub {
		1201	$grp->result ($_[0]);
		1202	};
		1203	};
		1204
		1205	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1206	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1207
		1208	add $grp $dirgrp;
		1209	};
		1210
		1211	$grp
		1212	}
		1213
		1214	=item aio_fcntl $fh, $cmd, $arg, $callback->($status)
		1215
		1216	=item aio_ioctl $fh, $request, $buf, $callback->($status)
		1217
		1218	These work just like the C<fcntl> and C<ioctl> built-in functions, except
		1219	they execute asynchronously and pass the return value to the callback.
		1220
		1221	Both calls can be used for a lot of things, some of which make more sense
		1222	to run asynchronously in their own thread, while some others make less
		1223	sense. For example, calls that block waiting for external events, such
		1224	as locking, will also lock down an I/O thread while it is waiting, which
		1225	can deadlock the whole I/O system. At the same time, there might be no
		1226	alternative to using a thread to wait.
		1227
		1228	So in general, you should only use these calls for things that do
		1229	(filesystem) I/O, not for things that wait for other events (network,
		1230	other processes), although if you are careful and know what you are doing,
		1231	you still can.
		1232
		1233	The following constants are available (missing ones are, as usual C<0>):
		1234
		1235	C<F_DUPFD_CLOEXEC>,
		1236
		1237	C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>,
		1238
		1239	C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>.
		1240
		1241	C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>,
		1242	C<FS_IOC_FIEMAP>.
		1243
		1244	C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>,
		1245	C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>.
		1246
		1247	C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>,
		1248	C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>,
		1249	C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>,
		1250	C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>,
		1251	C<FS_FL_USER_MODIFIABLE>.
		1252
		1253	C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>,
		1254	C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>,
		1255	C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>,
		1256	C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>,
		1257
		1258	=item aio_sync $callback->($status)
		1259
		1260	Asynchronously call sync and call the callback when finished.
		1261
184	=item aio_fsync $fh, $callback	1262	=item aio_fsync $fh, $callback->($status)
185		1263
186	Asynchronously call fsync on the given filehandle and call the callback	1264	Asynchronously call fsync on the given filehandle and call the callback
187	with the fsync result code.	1265	with the fsync result code.
188		1266
189	=item aio_fdatasync $fh, $callback	1267	=item aio_fdatasync $fh, $callback->($status)
190		1268
191	Asynchronously call fdatasync on the given filehandle and call the	1269	Asynchronously call fdatasync on the given filehandle and call the
192	callback with the fdatasync result code.	1270	callback with the fdatasync result code.
193		1271
		1272	If this call isn't available because your OS lacks it or it couldn't be
		1273	detected, it will be emulated by calling C<fsync> instead.
		1274
		1275	=item aio_syncfs $fh, $callback->($status)
		1276
		1277	Asynchronously call the syncfs syscall to sync the filesystem associated
		1278	to the given filehandle and call the callback with the syncfs result
		1279	code. If syncfs is not available, calls sync(), but returns C<-1> and sets
		1280	errno to C<ENOSYS> nevertheless.
		1281
		1282	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1283
		1284	Sync the data portion of the file specified by C<$offset> and C<$length>
		1285	to disk (but NOT the metadata), by calling the Linux-specific
		1286	sync_file_range call. If sync_file_range is not available or it returns
		1287	ENOSYS, then fdatasync or fsync is being substituted.
		1288
		1289	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1290	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1291	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1292	manpage for details.
		1293
		1294	=item aio_pathsync $pathname, $callback->($status)
		1295
		1296	This request tries to open, fsync and close the given path. This is a
		1297	composite request intended to sync directories after directory operations
		1298	(E.g. rename). This might not work on all operating systems or have any
		1299	specific effect, but usually it makes sure that directory changes get
		1300	written to disc. It works for anything that can be opened for read-only,
		1301	not just directories.
		1302
		1303	Future versions of this function might fall back to other methods when
		1304	C<fsync> on the directory fails (such as calling C<sync>).
		1305
		1306	Passes C<0> when everything went ok, and C<-1> on error.
		1307
		1308	=cut
		1309
		1310	sub aio_pathsync($;$) {
		1311	my ($path, $cb) = @_;
		1312
		1313	my $pri = aioreq_pri;
		1314	my $grp = aio_group $cb;
		1315
		1316	aioreq_pri $pri;
		1317	add $grp aio_open $path, O_RDONLY, 0, sub {
		1318	my ($fh) = @_;
		1319	if ($fh) {
		1320	aioreq_pri $pri;
		1321	add $grp aio_fsync $fh, sub {
		1322	$grp->result ($_[0]);
		1323
		1324	aioreq_pri $pri;
		1325	add $grp aio_close $fh;
		1326	};
		1327	} else {
		1328	$grp->result (-1);
		1329	}
		1330	};
		1331
		1332	$grp
		1333	}
		1334
		1335	=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
		1336
		1337	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1338	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1339	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1340	scalar must only be modified in-place while an aio operation is pending on
		1341	it).
		1342
		1343	It calls the C<msync> function of your OS, if available, with the memory
		1344	area starting at C<$offset> in the string and ending C<$length> bytes
		1345	later. If C<$length> is negative, counts from the end, and if C<$length>
		1346	is C<undef>, then it goes till the end of the string. The flags can be
		1347	either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional
		1348	C<IO::AIO::MS_INVALIDATE>.
		1349
		1350	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1351
		1352	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1353	scalars.
		1354
		1355	It touches (reads or writes) all memory pages in the specified
		1356	range inside the scalar. All caveats and parameters are the same
		1357	as for C<aio_msync>, above, except for flags, which must be either
		1358	C<0> (which reads all pages and ensures they are instantiated) or
		1359	C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and
		1360	writing an octet from it, which dirties the page).
		1361
		1362	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1363
		1364	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1365	scalars.
		1366
		1367	It reads in all the pages of the underlying storage into memory (if any)
		1368	and locks them, so they are not getting swapped/paged out or removed.
		1369
		1370	If C<$length> is undefined, then the scalar will be locked till the end.
		1371
		1372	On systems that do not implement C<mlock>, this function returns C<-1>
		1373	and sets errno to C<ENOSYS>.
		1374
		1375	Note that the corresponding C<munlock> is synchronous and is
		1376	documented under L<MISCELLANEOUS FUNCTIONS>.
		1377
		1378	Example: open a file, mmap and mlock it - both will be undone when
		1379	C<$data> gets destroyed.
		1380
		1381	open my $fh, "<", $path or die "$path: $!";
		1382	my $data;
		1383	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1384	aio_mlock $data; # mlock in background
		1385
		1386	=item aio_mlockall $flags, $callback->($status)
		1387
		1388	Calls the C<mlockall> function with the given C<$flags> (a combination of
		1389	C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
		1390
		1391	On systems that do not implement C<mlockall>, this function returns C<-1>
		1392	and sets errno to C<ENOSYS>.
		1393
		1394	Note that the corresponding C<munlockall> is synchronous and is
		1395	documented under L<MISCELLANEOUS FUNCTIONS>.
		1396
		1397	Example: asynchronously lock all current and future pages into memory.
		1398
		1399	aio_mlockall IO::AIO::MCL_FUTURE;
		1400
		1401	=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
		1402
		1403	Queries the extents of the given file (by calling the Linux C<FIEMAP>
		1404	ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If
		1405	the ioctl is not available on your OS, then this request will fail with
		1406	C<ENOSYS>.
		1407
		1408	C<$start> is the starting offset to query extents for, C<$length> is the
		1409	size of the range to query - if it is C<undef>, then the whole file will
		1410	be queried.
		1411
		1412	C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or
		1413	C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also
		1414	exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query
		1415	the data portion.
		1416
		1417	C<$count> is the maximum number of extent records to return. If it is
		1418	C<undef>, then IO::AIO queries all extents of the range. As a very special
		1419	case, if it is C<0>, then the callback receives the number of extents
		1420	instead of the extents themselves (which is unreliable, see below).
		1421
		1422	If an error occurs, the callback receives no arguments. The special
		1423	C<errno> value C<IO::AIO::EBADR> is available to test for flag errors.
		1424
		1425	Otherwise, the callback receives an array reference with extent
		1426	structures. Each extent structure is an array reference itself, with the
		1427	following members:
		1428
		1429	[$logical, $physical, $length, $flags]
		1430
		1431	Flags is any combination of the following flag values (typically either C<0>
		1432	or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)):
		1433
		1434	C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>,
		1435	C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>,
		1436	C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>,
		1437	C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>,
		1438	C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or
		1439	C<IO::AIO::FIEMAP_EXTENT_SHARED>.
		1440
		1441	At the time of this writing (Linux 3.2), this request is unreliable unless
		1442	C<$count> is C<undef>, as the kernel has all sorts of bugs preventing
		1443	it to return all extents of a range for files with a large number of
		1444	extents. The code (only) works around all these issues if C<$count> is
		1445	C<undef>.
		1446
		1447	=item aio_group $callback->(...)
		1448
		1449	This is a very special aio request: Instead of doing something, it is a
		1450	container for other aio requests, which is useful if you want to bundle
		1451	many requests into a single, composite, request with a definite callback
		1452	and the ability to cancel the whole request with its subrequests.
		1453
		1454	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		1455	for more info.
		1456
		1457	Example:
		1458
		1459	my $grp = aio_group sub {
		1460	print "all stats done\n";
		1461	};
		1462
		1463	add $grp
		1464	(aio_stat ...),
		1465	(aio_stat ...),
		1466	...;
		1467
		1468	=item aio_nop $callback->()
		1469
		1470	This is a special request - it does nothing in itself and is only used for
		1471	side effects, such as when you want to add a dummy request to a group so
		1472	that finishing the requests in the group depends on executing the given
		1473	code.
		1474
		1475	While this request does nothing, it still goes through the execution
		1476	phase and still requires a worker thread. Thus, the callback will not
		1477	be executed immediately but only after other requests in the queue have
		1478	entered their execution phase. This can be used to measure request
		1479	latency.
		1480
		1481	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
		1482
		1483	Mainly used for debugging and benchmarking, this aio request puts one of
		1484	the request workers to sleep for the given time.
		1485
		1486	While it is theoretically handy to have simple I/O scheduling requests
		1487	like sleep and file handle readable/writable, the overhead this creates is
		1488	immense (it blocks a thread for a long time) so do not use this function
		1489	except to put your application under artificial I/O pressure.
		1490
194	=back	1491	=back
195		1492
		1493
		1494	=head2 IO::AIO::WD - multiple working directories
		1495
		1496	Your process only has one current working directory, which is used by all
		1497	threads. This makes it hard to use relative paths (some other component
		1498	could call C<chdir> at any time, and it is hard to control when the path
		1499	will be used by IO::AIO).
		1500
		1501	One solution for this is to always use absolute paths. This usually works,
		1502	but can be quite slow (the kernel has to walk the whole path on every
		1503	access), and can also be a hassle to implement.
		1504
		1505	Newer POSIX systems have a number of functions (openat, fdopendir,
		1506	futimensat and so on) that make it possible to specify working directories
		1507	per operation.
		1508
		1509	For portability, and because the clowns who "designed", or shall I write,
		1510	perpetrated this new interface were obviously half-drunk, this abstraction
		1511	cannot be perfect, though.
		1512
		1513	IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
		1514	object. This object stores the canonicalised, absolute version of the
		1515	path, and on systems that allow it, also a directory file descriptor.
		1516
		1517	Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
		1518	or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
		1519	object and a pathname instead (or the IO::AIO::WD object alone, which
		1520	gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
		1521	IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
		1522	to that IO::AIO::WD object.
		1523
		1524	For example, to get a wd object for F</etc> and then stat F<passwd>
		1525	inside, you would write:
		1526
		1527	aio_wd "/etc", sub {
		1528	my $etcdir = shift;
		1529
		1530	# although $etcdir can be undef on error, there is generally no reason
		1531	# to check for errors here, as aio_stat will fail with ENOENT
		1532	# when $etcdir is undef.
		1533
		1534	aio_stat [$etcdir, "passwd"], sub {
		1535	# yay
		1536	};
		1537	};
		1538
		1539	The fact that C<aio_wd> is a request and not a normal function shows that
		1540	creating an IO::AIO::WD object is itself a potentially blocking operation,
		1541	which is why it is done asynchronously.
		1542
		1543	To stat the directory obtained with C<aio_wd> above, one could write
		1544	either of the following three request calls:
		1545
		1546	aio_lstat "/etc" , sub { ... # pathname as normal string
		1547	aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
		1548	aio_lstat $wd , sub { ... # shorthand for the previous
		1549
		1550	As with normal pathnames, IO::AIO keeps a copy of the working directory
		1551	object and the pathname string, so you could write the following without
		1552	causing any issues due to C<$path> getting reused:
		1553
		1554	my $path = [$wd, undef];
		1555
		1556	for my $name (qw(abc def ghi)) {
		1557	$path->[1] = $name;
		1558	aio_stat $path, sub {
		1559	# ...
		1560	};
		1561	}
		1562
		1563	There are some caveats: when directories get renamed (or deleted), the
		1564	pathname string doesn't change, so will point to the new directory (or
		1565	nowhere at all), while the directory fd, if available on the system,
		1566	will still point to the original directory. Most functions accepting a
		1567	pathname will use the directory fd on newer systems, and the string on
		1568	older systems. Some functions (such as C<aio_realpath>) will always rely on
		1569	the string form of the pathname.
		1570
		1571	So this functionality is mainly useful to get some protection against
		1572	C<chdir>, to easily get an absolute path out of a relative path for future
		1573	reference, and to speed up doing many operations in the same directory
		1574	(e.g. when stat'ing all files in a directory).
		1575
		1576	The following functions implement this working directory abstraction:
		1577
		1578	=over 4
		1579
		1580	=item aio_wd $pathname, $callback->($wd)
		1581
		1582	Asynchonously canonicalise the given pathname and convert it to an
		1583	IO::AIO::WD object representing it. If possible and supported on the
		1584	system, also open a directory fd to speed up pathname resolution relative
		1585	to this working directory.
		1586
		1587	If something goes wrong, then C<undef> is passwd to the callback instead
		1588	of a working directory object and C<$!> is set appropriately. Since
		1589	passing C<undef> as working directory component of a pathname fails the
		1590	request with C<ENOENT>, there is often no need for error checking in the
		1591	C<aio_wd> callback, as future requests using the value will fail in the
		1592	expected way.
		1593
		1594	=item IO::AIO::CWD
		1595
		1596	This is a compiletime constant (object) that represents the process
		1597	current working directory.
		1598
		1599	Specifying this object as working directory object for a pathname is as if
		1600	the pathname would be specified directly, without a directory object. For
		1601	example, these calls are functionally identical:
		1602
		1603	aio_stat "somefile", sub { ... };
		1604	aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
		1605
		1606	=back
		1607
		1608	To recover the path associated with an IO::AIO::WD object, you can use
		1609	C<aio_realpath>:
		1610
		1611	aio_realpath $wd, sub {
		1612	warn "path is $_[0]\n";
		1613	};
		1614
		1615	Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir>
		1616	sometimes, fall back to using an absolue path.
		1617
		1618	=head2 IO::AIO::REQ CLASS
		1619
		1620	All non-aggregate C<aio_*> functions return an object of this class when
		1621	called in non-void context.
		1622
		1623	=over 4
		1624
		1625	=item cancel $req
		1626
		1627	Cancels the request, if possible. Has the effect of skipping execution
		1628	when entering the B<execute> state and skipping calling the callback when
		1629	entering the the B<result> state, but will leave the request otherwise
		1630	untouched (with the exception of readdir). That means that requests that
		1631	currently execute will not be stopped and resources held by the request
		1632	will not be freed prematurely.
		1633
		1634	=item cb $req $callback->(...)
		1635
		1636	Replace (or simply set) the callback registered to the request.
		1637
		1638	=back
		1639
		1640	=head2 IO::AIO::GRP CLASS
		1641
		1642	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		1643	objects of this class, too.
		1644
		1645	A IO::AIO::GRP object is a special request that can contain multiple other
		1646	aio requests.
		1647
		1648	You create one by calling the C<aio_group> constructing function with a
		1649	callback that will be called when all contained requests have entered the
		1650	C<done> state:
		1651
		1652	my $grp = aio_group sub {
		1653	print "all requests are done\n";
		1654	};
		1655
		1656	You add requests by calling the C<add> method with one or more
		1657	C<IO::AIO::REQ> objects:
		1658
		1659	$grp->add (aio_unlink "...");
		1660
		1661	add $grp aio_stat "...", sub {
		1662	$_[0] or return $grp->result ("error");
		1663
		1664	# add another request dynamically, if first succeeded
		1665	add $grp aio_open "...", sub {
		1666	$grp->result ("ok");
		1667	};
		1668	};
		1669
		1670	This makes it very easy to create composite requests (see the source of
		1671	C<aio_move> for an application) that work and feel like simple requests.
		1672
		1673	=over 4
		1674
		1675	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		1676	C<IO::AIO::poll_cb>, just like any other request.
		1677
		1678	=item * They can be canceled like any other request. Canceling will cancel not
		1679	only the request itself, but also all requests it contains.
		1680
		1681	=item * They can also can also be added to other IO::AIO::GRP objects.
		1682
		1683	=item * You must not add requests to a group from within the group callback (or
		1684	any later time).
		1685
		1686	=back
		1687
		1688	Their lifetime, simplified, looks like this: when they are empty, they
		1689	will finish very quickly. If they contain only requests that are in the
		1690	C<done> state, they will also finish. Otherwise they will continue to
		1691	exist.
		1692
		1693	That means after creating a group you have some time to add requests
		1694	(precisely before the callback has been invoked, which is only done within
		1695	the C<poll_cb>). And in the callbacks of those requests, you can add
		1696	further requests to the group. And only when all those requests have
		1697	finished will the the group itself finish.
		1698
		1699	=over 4
		1700
		1701	=item add $grp ...
		1702
		1703	=item $grp->add (...)
		1704
		1705	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		1706	be added, including other groups, as long as you do not create circular
		1707	dependencies.
		1708
		1709	Returns all its arguments.
		1710
		1711	=item $grp->cancel_subs
		1712
		1713	Cancel all subrequests and clears any feeder, but not the group request
		1714	itself. Useful when you queued a lot of events but got a result early.
		1715
		1716	The group request will finish normally (you cannot add requests to the
		1717	group).
		1718
		1719	=item $grp->result (...)
		1720
		1721	Set the result value(s) that will be passed to the group callback when all
		1722	subrequests have finished and set the groups errno to the current value
		1723	of errno (just like calling C<errno> without an error number). By default,
		1724	no argument will be passed and errno is zero.
		1725
		1726	=item $grp->errno ([$errno])
		1727
		1728	Sets the group errno value to C<$errno>, or the current value of errno
		1729	when the argument is missing.
		1730
		1731	Every aio request has an associated errno value that is restored when
		1732	the callback is invoked. This method lets you change this value from its
		1733	default (0).
		1734
		1735	Calling C<result> will also set errno, so make sure you either set C<$!>
		1736	before the call to C<result>, or call c<errno> after it.
		1737
		1738	=item feed $grp $callback->($grp)
		1739
		1740	Sets a feeder/generator on this group: every group can have an attached
		1741	generator that generates requests if idle. The idea behind this is that,
		1742	although you could just queue as many requests as you want in a group,
		1743	this might starve other requests for a potentially long time. For example,
		1744	C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
		1745	requests, delaying any later requests for a long time.
		1746
		1747	To avoid this, and allow incremental generation of requests, you can
		1748	instead a group and set a feeder on it that generates those requests. The
		1749	feed callback will be called whenever there are few enough (see C<limit>,
		1750	below) requests active in the group itself and is expected to queue more
		1751	requests.
		1752
		1753	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1754	not impose any limits).
		1755
		1756	If the feed does not queue more requests when called, it will be
		1757	automatically removed from the group.
		1758
		1759	If the feed limit is C<0> when this method is called, it will be set to
		1760	C<2> automatically.
		1761
		1762	Example:
		1763
		1764	# stat all files in @files, but only ever use four aio requests concurrently:
		1765
		1766	my $grp = aio_group sub { print "finished\n" };
		1767	limit $grp 4;
		1768	feed $grp sub {
		1769	my $file = pop @files
		1770	or return;
		1771
		1772	add $grp aio_stat $file, sub { ... };
		1773	};
		1774
		1775	=item limit $grp $num
		1776
		1777	Sets the feeder limit for the group: The feeder will be called whenever
		1778	the group contains less than this many requests.
		1779
		1780	Setting the limit to C<0> will pause the feeding process.
		1781
		1782	The default value for the limit is C<0>, but note that setting a feeder
		1783	automatically bumps it up to C<2>.
		1784
		1785	=back
		1786
196	=head2 SUPPORT FUNCTIONS	1787	=head2 SUPPORT FUNCTIONS
197		1788
		1789	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1790
198	=over 4	1791	=over 4
199		1792
200	=item $fileno = IO::AIO::poll_fileno	1793	=item $fileno = IO::AIO::poll_fileno
201		1794
202	Return the I<request result pipe filehandle>. This filehandle must be	1795	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	1796	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	1797	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
205	C<poll_cb> to check the results.	1798	you have to call C<poll_cb> to check the results.
206		1799
207	See C<poll_cb> for an example.	1800	See C<poll_cb> for an example.
208		1801
209	=item IO::AIO::poll_cb	1802	=item IO::AIO::poll_cb
210		1803
211	Process all outstanding events on the result pipe. You have to call this	1804	Process some requests that have reached the result phase (i.e. they have
212	regularly. Returns the number of events processed. Returns immediately	1805	been executed but the results are not yet reported). You have to call
213	when no events are outstanding.	1806	this "regularly" to finish outstanding requests.
214		1807
215	You can use Event to multiplex, e.g.:	1808	Returns C<0> if all events could be processed (or there were no
		1809	events to process), or C<-1> if it returned earlier for whatever
		1810	reason. Returns immediately when no events are outstanding. The amount
		1811	of events processed depends on the settings of C<IO::AIO::max_poll_req>,
		1812	C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>.
		1813
		1814	If not all requests were processed for whatever reason, the poll file
		1815	descriptor will still be ready when C<poll_cb> returns, so normally you
		1816	don't have to do anything special to have it called later.
		1817
		1818	Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
		1819	ready, it can be beneficial to call this function from loops which submit
		1820	a lot of requests, to make sure the results get processed when they become
		1821	available and not just when the loop is finished and the event loop takes
		1822	over again. This function returns very fast when there are no outstanding
		1823	requests.
		1824
		1825	Example: Install an Event watcher that automatically calls
		1826	IO::AIO::poll_cb with high priority (more examples can be found in the
		1827	SYNOPSIS section, at the top of this document):
216		1828
217	Event->io (fd => IO::AIO::poll_fileno,	1829	Event->io (fd => IO::AIO::poll_fileno,
218	poll => 'r', async => 1,	1830	poll => 'r', async => 1,
219	cb => \&IO::AIO::poll_cb);	1831	cb => \&IO::AIO::poll_cb);
220		1832
221	=item IO::AIO::poll_wait	1833	=item IO::AIO::poll_wait
222		1834
223	Wait till the result filehandle becomes ready for reading (simply does a	1835	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	1836	requests are outstanding anymore.
225	for some requests to finish).	1837
		1838	This is useful if you want to synchronously wait for some requests to
		1839	become ready, without actually handling them.
226		1840
227	See C<nreqs> for an example.	1841	See C<nreqs> for an example.
228		1842
		1843	=item IO::AIO::poll
		1844
		1845	Waits until some requests have been handled.
		1846
		1847	Returns the number of requests processed, but is otherwise strictly
		1848	equivalent to:
		1849
		1850	IO::AIO::poll_wait, IO::AIO::poll_cb
		1851
229	=item IO::AIO::nreqs	1852	=item IO::AIO::flush
230		1853
231	Returns the number of requests currently outstanding.	1854	Wait till all outstanding AIO requests have been handled.
232		1855
233	Example: wait till there are no outstanding requests anymore:	1856	Strictly equivalent to:
234		1857
235	IO::AIO::poll_wait, IO::AIO::poll_cb	1858	IO::AIO::poll_wait, IO::AIO::poll_cb
236	while IO::AIO::nreqs;	1859	while IO::AIO::nreqs;
237		1860
		1861	=item IO::AIO::max_poll_reqs $nreqs
		1862
		1863	=item IO::AIO::max_poll_time $seconds
		1864
		1865	These set the maximum number of requests (default C<0>, meaning infinity)
		1866	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1867	the maximum amount of time (default C<0>, meaning infinity) spent in
		1868	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1869	of time C<poll_cb> is allowed to use).
		1870
		1871	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1872	syscall per request processed, which is not normally a problem unless your
		1873	callbacks are really really fast or your OS is really really slow (I am
		1874	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1875
		1876	Setting these is useful if you want to ensure some level of
		1877	interactiveness when perl is not fast enough to process all requests in
		1878	time.
		1879
		1880	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		1881
		1882	Example: Install an Event watcher that automatically calls
		1883	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1884	program get the CPU sometimes even under high AIO load.
		1885
		1886	# try not to spend much more than 0.1s in poll_cb
		1887	IO::AIO::max_poll_time 0.1;
		1888
		1889	# use a low priority so other tasks have priority
		1890	Event->io (fd => IO::AIO::poll_fileno,
		1891	poll => 'r', nice => 1,
		1892	cb => &IO::AIO::poll_cb);
		1893
		1894	=back
		1895
		1896	=head3 CONTROLLING THE NUMBER OF THREADS
		1897
		1898	=over
		1899
		1900	=item IO::AIO::min_parallel $nthreads
		1901
		1902	Set the minimum number of AIO threads to C<$nthreads>. The current
		1903	default is C<8>, which means eight asynchronous operations can execute
		1904	concurrently at any one time (the number of outstanding requests,
		1905	however, is unlimited).
		1906
		1907	IO::AIO starts threads only on demand, when an AIO request is queued and
		1908	no free thread exists. Please note that queueing up a hundred requests can
		1909	create demand for a hundred threads, even if it turns out that everything
		1910	is in the cache and could have been processed faster by a single thread.
		1911
		1912	It is recommended to keep the number of threads relatively low, as some
		1913	Linux kernel versions will scale negatively with the number of threads
		1914	(higher parallelity => MUCH higher latency). With current Linux 2.6
		1915	versions, 4-32 threads should be fine.
		1916
		1917	Under most circumstances you don't need to call this function, as the
		1918	module selects a default that is suitable for low to moderate load.
		1919
		1920	=item IO::AIO::max_parallel $nthreads
		1921
		1922	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
		1923	specified number of threads are currently running, this function kills
		1924	them. This function blocks until the limit is reached.
		1925
		1926	While C<$nthreads> are zero, aio requests get queued but not executed
		1927	until the number of threads has been increased again.
		1928
		1929	This module automatically runs C<max_parallel 0> at program end, to ensure
		1930	that all threads are killed and that there are no outstanding requests.
		1931
		1932	Under normal circumstances you don't need to call this function.
		1933
		1934	=item IO::AIO::max_idle $nthreads
		1935
		1936	Limit the number of threads (default: 4) that are allowed to idle
		1937	(i.e., threads that did not get a request to process within the idle
		1938	timeout (default: 10 seconds). That means if a thread becomes idle while
		1939	C<$nthreads> other threads are also idle, it will free its resources and
		1940	exit.
		1941
		1942	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1943	to allow for extremely high load situations, but want to free resources
		1944	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1945
		1946	The default is probably ok in most situations, especially if thread
		1947	creation is fast. If thread creation is very slow on your system you might
		1948	want to use larger values.
		1949
		1950	=item IO::AIO::idle_timeout $seconds
		1951
		1952	Sets the minimum idle timeout (default 10) after which worker threads are
		1953	allowed to exit. SEe C<IO::AIO::max_idle>.
		1954
		1955	=item IO::AIO::max_outstanding $maxreqs
		1956
		1957	Sets the maximum number of outstanding requests to C<$nreqs>. If
		1958	you do queue up more than this number of requests, the next call to
		1959	C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
		1960	C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
		1961	longer exceeded.
		1962
		1963	In other words, this setting does not enforce a queue limit, but can be
		1964	used to make poll functions block if the limit is exceeded.
		1965
		1966	This is a very bad function to use in interactive programs because it
		1967	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1968	use an C<aio_group> together with a feed callback.
		1969
		1970	Its main use is in scripts without an event loop - when you want to stat
		1971	a lot of files, you can write something like this:
		1972
		1973	IO::AIO::max_outstanding 32;
		1974
		1975	for my $path (...) {
		1976	aio_stat $path , ...;
		1977	IO::AIO::poll_cb;
		1978	}
		1979
		1980	IO::AIO::flush;
		1981
		1982	The call to C<poll_cb> inside the loop will normally return instantly, but
		1983	as soon as more thna C<32> reqeusts are in-flight, it will block until
		1984	some requests have been handled. This keeps the loop from pushing a large
		1985	number of C<aio_stat> requests onto the queue.
		1986
		1987	The default value for C<max_outstanding> is very large, so there is no
		1988	practical limit on the number of outstanding requests.
		1989
		1990	=back
		1991
		1992	=head3 STATISTICAL INFORMATION
		1993
		1994	=over
		1995
238	=item IO::AIO::flush	1996	=item IO::AIO::nreqs
239		1997
240	Wait till all outstanding AIO requests have been handled.	1998	Returns the number of requests currently in the ready, execute or pending
		1999	states (i.e. for which their callback has not been invoked yet).
241		2000
242	Strictly equivalent to:	2001	Example: wait till there are no outstanding requests anymore:
243		2002
244	IO::AIO::poll_wait, IO::AIO::poll_cb	2003	IO::AIO::poll_wait, IO::AIO::poll_cb
245	while IO::AIO::nreqs;	2004	while IO::AIO::nreqs;
246		2005
		2006	=item IO::AIO::nready
		2007
		2008	Returns the number of requests currently in the ready state (not yet
		2009	executed).
		2010
247	=item IO::AIO::poll	2011	=item IO::AIO::npending
248		2012
249	Waits until some requests have been handled.	2013	Returns the number of requests currently in the pending state (executed,
250		2014	but not yet processed by poll_cb).
251	Strictly equivalent to:
252
253	IO::AIO::poll_wait, IO::AIO::poll_cb
254	if IO::AIO::nreqs;
255
256	=item IO::AIO::min_parallel $nthreads
257
258	Set the minimum number of AIO threads to C<$nthreads>. The default is
259	C<1>, which means a single asynchronous operation can be done at one time
260	(the number of outstanding operations, however, is unlimited).
261
262	It is recommended to keep the number of threads low, as some Linux
263	kernel versions will scale negatively with the number of threads (higher
264	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32
265	threads should be fine.
266
267	Under normal circumstances you don't need to call this function, as this
268	module automatically starts some threads (the exact number might change,
269	and is currently 4).
270
271	=item IO::AIO::max_parallel $nthreads
272
273	Sets the maximum number of AIO threads to C<$nthreads>. If more than
274	the specified number of threads are currently running, kill them. This
275	function blocks until the limit is reached.
276
277	This module automatically runs C<max_parallel 0> at program end, to ensure
278	that all threads are killed and that there are no outstanding requests.
279
280	Under normal circumstances you don't need to call this function.
281
282	=item $oldnreqs = IO::AIO::max_outstanding $nreqs
283
284	Sets the maximum number of outstanding requests to C<$nreqs>. If you
285	try to queue up more than this number of requests, the caller will block until
286	some requests have been handled.
287
288	The default is very large, so normally there is no practical limit. If you
289	queue up many requests in a loop it it often improves speed if you set
290	this to a relatively low number, such as C<100>.
291
292	Under normal circumstances you don't need to call this function.
293		2015
294	=back	2016	=back
295		2017
		2018	=head3 SUBSECOND STAT TIME ACCESS
		2019
		2020	Both C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> functions can
		2021	generally find access/modification and change times with subsecond time
		2022	accuracy of the system supports it, but perl's built-in functions only
		2023	return the integer part.
		2024
		2025	The following functions return the timestamps of the most recent
		2026	stat with subsecond precision on most systems and work both after
		2027	C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> calls. Their return
		2028	value is only meaningful after a successful C<stat>/C<lstat> call, or
		2029	during/after a successful C<aio_stat>/C<aio_lstat> callback.
		2030
		2031	This is similar to the L<Time::HiRes> C<stat> functions, but can return
		2032	full resolution without rounding and work with standard perl C<stat>,
		2033	alleviating the need to call the special C<Time::HiRes> functions, which
		2034	do not act like their perl counterparts.
		2035
		2036	On operating systems or file systems where subsecond time resolution is
		2037	not supported or could not be detected, a fractional part of C<0> is
		2038	returned, so it is always safe to call these functions.
		2039
		2040	=over 4
		2041
		2042	=item IO::AIO::stat_atime, IO::AIO::stat_mtime, IO::AIO::stat_ctime
		2043
		2044	Return the access, modication or change time, respectively, including
		2045	fractional part. Due to the limited precision of floating point, the
		2046	accuracy on most platforms is only a bit better than milliseconds for
		2047	times around now - see the I<nsec> function family, below, for full
		2048	accuracy.
		2049
		2050	=item ($atime, $mtime, $ctime, ...) = IO::AIO::stat_xtime
		2051
		2052	Returns access, modification and change time all in one go, and maybe more
		2053	times in the future version.
		2054
		2055	=item IO::AIO::stat_atimensec, IO::AIO::stat_mtimensec, IO::AIO::stat_ctimensec
		2056
		2057	Return the fractional access, modifcation or change time, in nanoseconds,
		2058	as an integer in the range C<0> to C<999999999>.
		2059
		2060	=back
		2061
		2062	Example: print the high resolution modification time of F</etc>, using
		2063	C<stat>, and C<IO::AIO::aio_stat>.
		2064
		2065	if (stat "/etc") {
		2066	printf "stat(/etc) mtime: %f\n", IO::AIO::stat_mtime;
		2067	}
		2068
		2069	IO::AIO::aio_stat "/etc", sub {
		2070	$_[0]
		2071	and return;
		2072
		2073	printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::stat_mtimensec;
		2074	};
		2075
		2076	IO::AIO::flush;
		2077
		2078	Output of the awbove on my system, showing reduced and full accuracy:
		2079
		2080	stat(/etc) mtime: 1534043702.020808
		2081	aio_stat(/etc) mtime: 1534043702.020807792
		2082
		2083	=head3 MISCELLANEOUS FUNCTIONS
		2084
		2085	IO::AIO implements some functions that are useful when you want to use
		2086	some "Advanced I/O" function not available to in Perl, without going the
		2087	"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*>
		2088	counterpart.
		2089
		2090	=over 4
		2091
		2092	=item $numfd = IO::AIO::get_fdlimit
		2093
		2094	This function is I<EXPERIMENTAL> and subject to change.
		2095
		2096	Tries to find the current file descriptor limit and returns it, or
		2097	C<undef> and sets C<$!> in case of an error. The limit is one larger than
		2098	the highest valid file descriptor number.
		2099
		2100	=item IO::AIO::min_fdlimit [$numfd]
		2101
		2102	This function is I<EXPERIMENTAL> and subject to change.
		2103
		2104	Try to increase the current file descriptor limit(s) to at least C<$numfd>
		2105	by changing the soft or hard file descriptor resource limit. If C<$numfd>
		2106	is missing, it will try to set a very high limit, although this is not
		2107	recommended when you know the actual minimum that you require.
		2108
		2109	If the limit cannot be raised enough, the function makes a best-effort
		2110	attempt to increase the limit as much as possible, using various
		2111	tricks, while still failing. You can query the resulting limit using
		2112	C<IO::AIO::get_fdlimit>.
		2113
		2114	If an error occurs, returns C<undef> and sets C<$!>, otherwise returns
		2115	true.
		2116
		2117	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		2118
		2119	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		2120	but is blocking (this makes most sense if you know the input data is
		2121	likely cached already and the output filehandle is set to non-blocking
		2122	operations).
		2123
		2124	Returns the number of bytes copied, or C<-1> on error.
		2125
		2126	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		2127
		2128	Simply calls the C<posix_fadvise> function (see its
		2129	manpage for details). The following advice constants are
		2130	available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		2131	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		2132	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		2133
		2134	On systems that do not implement C<posix_fadvise>, this function returns
		2135	ENOSYS, otherwise the return value of C<posix_fadvise>.
		2136
		2137	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		2138
		2139	Simply calls the C<posix_madvise> function (see its
		2140	manpage for details). The following advice constants are
		2141	available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		2142	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>,
		2143	C<IO::AIO::MADV_DONTNEED>.
		2144
		2145	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2146	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2147	will be reduced to fit into the C<$scalar>.
		2148
		2149	On systems that do not implement C<posix_madvise>, this function returns
		2150	ENOSYS, otherwise the return value of C<posix_madvise>.
		2151
		2152	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		2153
		2154	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		2155	$scalar (see its manpage for details). The following protect
		2156	constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		2157	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		2158
		2159	If C<$offset> is negative, counts from the end. If C<$length> is negative,
		2160	the remaining length of the C<$scalar> is used. If possible, C<$length>
		2161	will be reduced to fit into the C<$scalar>.
		2162
		2163	On systems that do not implement C<mprotect>, this function returns
		2164	ENOSYS, otherwise the return value of C<mprotect>.
		2165
		2166	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		2167
		2168	Memory-maps a file (or anonymous memory range) and attaches it to the
		2169	given C<$scalar>, which will act like a string scalar. Returns true on
		2170	success, and false otherwise.
		2171
		2172	The scalar must exist, but its contents do not matter - this means you
		2173	cannot use a nonexistant array or hash element. When in doubt, C<undef>
		2174	the scalar first.
		2175
		2176	The only operations allowed on the mmapped scalar are C<substr>/C<vec>,
		2177	which don't change the string length, and most read-only operations such
		2178	as copying it or searching it with regexes and so on.
		2179
		2180	Anything else is unsafe and will, at best, result in memory leaks.
		2181
		2182	The memory map associated with the C<$scalar> is automatically removed
		2183	when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap>
		2184	or C<IO::AIO::munmap> functions are called on it.
		2185
		2186	This calls the C<mmap>(2) function internally. See your system's manual
		2187	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		2188
		2189	The C<$length> must be larger than zero and smaller than the actual
		2190	filesize.
		2191
		2192	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		2193	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		2194
		2195	C<$flags> can be a combination of
		2196	C<IO::AIO::MAP_SHARED> or
		2197	C<IO::AIO::MAP_PRIVATE>,
		2198	or a number of system-specific flags (when not available, the are C<0>):
		2199	C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant),
		2200	C<IO::AIO::MAP_LOCKED>,
		2201	C<IO::AIO::MAP_NORESERVE>,
		2202	C<IO::AIO::MAP_POPULATE>,
		2203	C<IO::AIO::MAP_NONBLOCK>,
		2204	C<IO::AIO::MAP_FIXED>,
		2205	C<IO::AIO::MAP_GROWSDOWN>,
		2206	C<IO::AIO::MAP_32BIT>,
		2207	C<IO::AIO::MAP_HUGETLB> or
		2208	C<IO::AIO::MAP_STACK>.
		2209
		2210	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		2211
		2212	C<$offset> is the offset from the start of the file - it generally must be
		2213	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		2214
		2215	Example:
		2216
		2217	use Digest::MD5;
		2218	use IO::AIO;
		2219
		2220	open my $fh, "<verybigfile"
		2221	or die "$!";
		2222
		2223	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		2224	or die "verybigfile: $!";
		2225
		2226	my $fast_md5 = md5 $data;
		2227
		2228	=item IO::AIO::munmap $scalar
		2229
		2230	Removes a previous mmap and undefines the C<$scalar>.
		2231
		2232	=item IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[, $new_address = 0]
		2233
		2234	Calls the Linux-specific mremap(2) system call. The C<$scalar> must have
		2235	been mapped by C<IO::AIO::mmap>, and C<$flags> must currently either be
		2236	C<0> or C<IO::AIO::MREMAP_MAYMOVE>.
		2237
		2238	Returns true if successful, and false otherwise. If the underlying mmapped
		2239	region has changed address, then the true value has the numerical value
		2240	C<1>, otherwise it has the numerical value C<0>:
		2241
		2242	my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE
		2243	or die "mremap: $!";
		2244
		2245	if ($success*1) {
		2246	warn "scalar has chanegd address in memory\n";
		2247	}
		2248
		2249	C<IO::AIO::MREMAP_FIXED> and the C<$new_address> argument are currently
		2250	implemented, but not supported and might go away in a future version.
		2251
		2252	On systems where this call is not supported or is not emulated, this call
		2253	returns falls and sets C<$!> to C<ENOSYS>.
		2254
		2255	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		2256
		2257	Calls the C<munlock> function, undoing the effects of a previous
		2258	C<aio_mlock> call (see its description for details).
		2259
		2260	=item IO::AIO::munlockall
		2261
		2262	Calls the C<munlockall> function.
		2263
		2264	On systems that do not implement C<munlockall>, this function returns
		2265	ENOSYS, otherwise the return value of C<munlockall>.
		2266
		2267	=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
		2268
		2269	Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or
		2270	C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they
		2271	should be the file offset.
		2272
		2273	C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might
		2274	silently corrupt the data in this case.
		2275
		2276	The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>,
		2277	C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and
		2278	C<IO::AIO::SPLICE_F_GIFT>.
		2279
		2280	See the C<splice(2)> manpage for details.
		2281
		2282	=item IO::AIO::tee $r_fh, $w_fh, $length, $flags
		2283
		2284	Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the
		2285	description for C<IO::AIO::splice> above for details.
		2286
		2287	=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
		2288
		2289	Attempts to query or change the pipe buffer size. Obviously works only
		2290	on pipes, and currently works only on GNU/Linux systems, and fails with
		2291	C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer
		2292	size on other systems, drop me a note.
		2293
		2294	=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
		2295
		2296	This is a direct interface to the Linux L<pipe2(2)> system call. If
		2297	C<$flags> is missing or C<0>, then this should be the same as a call to
		2298	perl's built-in C<pipe> function and create a new pipe, and works on
		2299	systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe
		2300	(..., 4096, O_BINARY)>.
		2301
		2302	If C<$flags> is non-zero, it tries to invoke the pipe2 system call with
		2303	the given flags (Linux 2.6.27, glibc 2.9).
		2304
		2305	On success, the read and write file handles are returned.
		2306
		2307	On error, nothing will be returned. If the pipe2 syscall is missing and
		2308	C<$flags> is non-zero, fails with C<ENOSYS>.
		2309
		2310	Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the
		2311	time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and
		2312	C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported.
		2313
		2314	Example: create a pipe race-free w.r.t. threads and fork:
		2315
		2316	my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
		2317	or die "pipe2: $!\n";
		2318
		2319	=item $fh = IO::AIO::eventfd [$initval, [$flags]]
		2320
		2321	This is a direct interface to the Linux L<eventfd(2)> system call. The
		2322	(unhelpful) defaults for C<$initval> and C<$flags> are C<0> for both.
		2323
		2324	On success, the new eventfd filehandle is returned, otherwise returns
		2325	C<undef>. If the eventfd syscall is missing, fails with C<ENOSYS>.
		2326
		2327	Please refer to L<eventfd(2)> for more info on this call.
		2328
		2329	The following symbol flag values are available: C<IO::AIO::EFD_CLOEXEC>,
		2330	C<IO::AIO::EFD_NONBLOCK> and C<IO::AIO::EFD_SEMAPHORE> (Linux 2.6.30).
		2331
		2332	Example: create a new eventfd filehandle:
		2333
		2334	$fh = IO::AIO::eventfd 0, IO::AIO::O_CLOEXEC
		2335	or die "eventfd: $!\n";
		2336
		2337	=item $fh = IO::AIO::timerfd_create $clockid[, $flags]
		2338
		2339	This is a direct interface to the Linux L<timerfd_create(2)> system call. The
		2340	(unhelpful) default for C<$flags> is C<0>.
		2341
		2342	On success, the new timerfd filehandle is returned, otherwise returns
		2343	C<undef>. If the eventfd syscall is missing, fails with C<ENOSYS>.
		2344
		2345	Please refer to L<timerfd_create(2)> for more info on this call.
		2346
		2347	The following C<$clockid> values are
		2348	available: C<IO::AIO::CLOCK_REALTIME>, C<IO::AIO::CLOCK_MONOTONIC>
		2349	C<IO::AIO::CLOCK_CLOCK_BOOTTIME> (Linux 3.15)
		2350	C<IO::AIO::CLOCK_CLOCK_REALTIME_ALARM> (Linux 3.11) and
		2351	C<IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM> (Linux 3.11).
		2352
		2353	The following C<$flags> values are available (Linux
		2354	2.6.27): C<IO::AIO::TFD_NONBLOCK> and C<IO::AIO::TFD_CLOEXEC>.
		2355
		2356	Example: create a new timerfd and set it to one-second repeated alarms,
		2357	then wait for two alarms:
		2358
		2359	my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC
		2360	or die "timerfd_create: $!\n";
		2361
		2362	defined IO::AIO::timerfd_settime $fh, 0, 1, 1
		2363	or die "timerfd_settime: $!\n";
		2364
		2365	for (1..2) {
		2366	8 == sysread $fh, my $buf, 8
		2367	or die "timerfd read failure\n";
		2368
		2369	printf "number of expirations (likely 1): %d\n",
		2370	unpack "Q", $buf;
		2371	}
		2372
		2373	=item ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
		2374
		2375	This is a direct interface to the Linux L<timerfd_settime(2)> system
		2376	call. Please refer to its manpage for more info on this call.
		2377
		2378	The new itimerspec is specified using two (possibly fractional) second
		2379	values, C<$new_interval> and C<$new_value>).
		2380
		2381	On success, the current interval and value are returned (as per
		2382	C<timerfd_gettime>). On failure, the empty list is returned.
		2383
		2384	The following C<$flags> values are
		2385	available: C<IO::AIO::TFD_TIMER_ABSTIME> and
		2386	C<IO::AIO::TFD_TIMER_CANCEL_ON_SET>.
		2387
		2388	See C<IO::AIO::timerfd_create> for a full example.
		2389
		2390	=item ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
		2391
		2392	This is a direct interface to the Linux L<timerfd_gettime(2)> system
		2393	call. Please refer to its manpage for more info on this call.
		2394
		2395	On success, returns the current values of interval and value for the given
		2396	timerfd (as potentially fractional second values). On failure, the empty
		2397	list is returned.
		2398
		2399	=back
		2400
296	=cut	2401	=cut
297		2402
298	# support function to convert a fd into a perl filehandle
299	sub _fd2fh {
300	return undef if $_[0] < 0;
301
302	# try to be perl5.6-compatible
303	local *AIO_FH;
304	open AIO_FH, "+<&=$_[0]"
305	or return undef;
306
307	*AIO_FH
308	}
309
310	min_parallel 4;	2403	min_parallel 8;
311		2404
312	END {	2405	END { flush }
313	max_parallel 0;
314	}
315		2406
316	1;	2407	1;
317		2408
		2409	=head1 EVENT LOOP INTEGRATION
		2410
		2411	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		2412	automatically into many event loops:
		2413
		2414	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		2415	use AnyEvent::AIO;
		2416
		2417	You can also integrate IO::AIO manually into many event loops, here are
		2418	some examples of how to do this:
		2419
		2420	# EV integration
		2421	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		2422
		2423	# Event integration
		2424	Event->io (fd => IO::AIO::poll_fileno,
		2425	poll => 'r',
		2426	cb => \&IO::AIO::poll_cb);
		2427
		2428	# Glib/Gtk2 integration
		2429	add_watch Glib::IO IO::AIO::poll_fileno,
		2430	in => sub { IO::AIO::poll_cb; 1 };
		2431
		2432	# Tk integration
		2433	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		2434	readable => \&IO::AIO::poll_cb);
		2435
		2436	# Danga::Socket integration
		2437	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		2438	\&IO::AIO::poll_cb);
		2439
		2440	=head2 FORK BEHAVIOUR
		2441
		2442	Usage of pthreads in a program changes the semantics of fork
		2443	considerably. Specifically, only async-safe functions can be called after
		2444	fork. Perl doesn't know about this, so in general, you cannot call fork
		2445	with defined behaviour in perl if pthreads are involved. IO::AIO uses
		2446	pthreads, so this applies, but many other extensions and (for inexplicable
		2447	reasons) perl itself often is linked against pthreads, so this limitation
		2448	applies to quite a lot of perls.
		2449
		2450	This module no longer tries to fight your OS, or POSIX. That means IO::AIO
		2451	only works in the process that loaded it. Forking is fully supported, but
		2452	using IO::AIO in the child is not.
		2453
		2454	You might get around by not I<using> IO::AIO before (or after)
		2455	forking. You could also try to call the L<IO::AIO::reinit> function in the
		2456	child:
		2457
		2458	=over 4
		2459
		2460	=item IO::AIO::reinit
		2461
		2462	Abandons all current requests and I/O threads and simply reinitialises all
		2463	data structures. This is not an operation supported by any standards, but
		2464	happens to work on GNU/Linux and some newer BSD systems.
		2465
		2466	The only reasonable use for this function is to call it after forking, if
		2467	C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
		2468	the process will result in undefined behaviour. Calling it at any time
		2469	will also result in any undefined (by POSIX) behaviour.
		2470
		2471	=back
		2472
		2473	=head2 LINUX-SPECIFIC CALLS
		2474
		2475	When a call is documented as "linux-specific" then this means it
		2476	originated on GNU/Linux. C<IO::AIO> will usually try to autodetect the
		2477	availability and compatibility of such calls regardless of the platform
		2478	it is compiled on, so platforms such as FreeBSD which often implement
		2479	these calls will work. When in doubt, call them and see if they fail wth
		2480	C<ENOSYS>.
		2481
		2482	=head2 MEMORY USAGE
		2483
		2484	Per-request usage:
		2485
		2486	Each aio request uses - depending on your architecture - around 100-200
		2487	bytes of memory. In addition, stat requests need a stat buffer (possibly
		2488	a few hundred bytes), readdir requires a result buffer and so on. Perl
		2489	scalars and other data passed into aio requests will also be locked and
		2490	will consume memory till the request has entered the done state.
		2491
		2492	This is not awfully much, so queuing lots of requests is not usually a
		2493	problem.
		2494
		2495	Per-thread usage:
		2496
		2497	In the execution phase, some aio requests require more memory for
		2498	temporary buffers, and each thread requires a stack and other data
		2499	structures (usually around 16k-128k, depending on the OS).
		2500
		2501	=head1 KNOWN BUGS
		2502
		2503	Known bugs will be fixed in the next release :)
		2504
		2505	=head1 KNOWN ISSUES
		2506
		2507	Calls that try to "import" foreign memory areas (such as C<IO::AIO::mmap>
		2508	or C<IO::AIO::aio_slurp>) do not work with generic lvalues, such as
		2509	non-created hash slots or other scalars I didn't think of. It's best to
		2510	avoid such and either use scalar variables or making sure that the scalar
		2511	exists (e.g. by storing C<undef>) and isn't "funny" (e.g. tied).
		2512
		2513	I am not sure anything can be done about this, so this is considered a
		2514	known issue, rather than a bug.
		2515
318	=head1 SEE ALSO	2516	=head1 SEE ALSO
319		2517
320	L<Coro>, L<Linux::AIO>.	2518	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		2519	more natural syntax.
321		2520
322	=head1 AUTHOR	2521	=head1 AUTHOR
323		2522
324	Marc Lehmann <schmorp@schmorp.de>	2523	Marc Lehmann <schmorp@schmorp.de>
325	http://home.schmorp.de/	2524	http://home.schmorp.de/

Diff Legend

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