ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.51 by root, Sat Jun 24 19:14:04 2006 UTC vs.
Revision 1.207 by root, Mon Jul 25 16:50:33 2011 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use IO::AIO;	7	use IO::AIO;
8		8
9	aio_open "/etc/passwd", O_RDONLY, 0, sub {	9	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
10	my ($fh) = @_;	10	my $fh = shift
		11	or die "/etc/passwd: $!";
11	...	12	...
12	};	13	};
13		14
14	aio_unlink "/tmp/file", sub { };	15	aio_unlink "/tmp/file", sub { };
15		16
16	aio_read $fh, 30000, 1024, $buffer, 0, sub {	17	aio_read $fh, 30000, 1024, $buffer, 0, sub {
17	$_[0] > 0 or die "read error: $!";	18	$_[0] > 0 or die "read error: $!";
18	};	19	};
19		20
20	# AnyEvent	21	# version 2+ has request and group objects
21	open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";	22	use IO::AIO 2;
22	my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
23		23
24	# Event	24	aioreq_pri 4; # give next request a very high priority
25	Event->io (fd => IO::AIO::poll_fileno,	25	my $req = aio_unlink "/tmp/file", sub { };
26	poll => 'r',	26	$req->cancel; # cancel request if still in queue
27	cb => \&IO::AIO::poll_cb);
28		27
29	# Glib/Gtk2	28	my $grp = aio_group sub { print "all stats done\n" };
30	add_watch Glib::IO IO::AIO::poll_fileno,	29	add $grp aio_stat "..." for ...;
31	in => sub { IO::AIO::poll_cb; 1 };
32
33	# Tk
34	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
35	readable => \&IO::AIO::poll_cb);
36
37	# Danga::Socket
38	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
39	\&IO::AIO::poll_cb);
40
41		30
42	=head1 DESCRIPTION	31	=head1 DESCRIPTION
43		32
44	This module implements asynchronous I/O using whatever means your	33	This module implements asynchronous I/O using whatever means your
45	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>).
46		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
47	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
48	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
49	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
50	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
51	aio functions available on many operating systems. However, they are often	57	functions available on many operating systems. However, they are often
52	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
53	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
54	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.
55		62
56	Although the module will work with in the presence of other threads, it is	63	Although the module will work in the presence of other (Perl-) threads,
57	currently not reentrant, so use appropriate locking yourself, always call	64	it is currently not reentrant in any way, so use appropriate locking
58	C<poll_cb> from within the same thread, or never call C<poll_cb> (or other	65	yourself, always call C<poll_cb> from within the same thread, or never
59	C<aio_> functions) recursively.	66	call C<poll_cb> (or other C<aio_> functions) recursively.
		67
		68	=head2 EXAMPLE
		69
		70	This is a simple example that uses the EV module and loads
		71	F</etc/passwd> asynchronously:
		72
		73	use Fcntl;
		74	use EV;
		75	use IO::AIO;
		76
		77	# register the IO::AIO callback with EV
		78	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		79
		80	# queue the request to open /etc/passwd
		81	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
		82	my $fh = shift
		83	or die "error while opening: $!";
		84
		85	# stat'ing filehandles is generally non-blocking
		86	my $size = -s $fh;
		87
		88	# queue a request to read the file
		89	my $contents;
		90	aio_read $fh, 0, $size, $contents, 0, sub {
		91	$_[0] == $size
		92	or die "short read: $!";
		93
		94	close $fh;
		95
		96	# file contents now in $contents
		97	print $contents;
		98
		99	# exit event loop and program
		100	EV::unloop;
		101	};
		102	};
		103
		104	# possibly queue up other requests, or open GUI windows,
		105	# check for sockets etc. etc.
		106
		107	# process events as long as there are some:
		108	EV::loop;
		109
		110	=head1 REQUEST ANATOMY AND LIFETIME
		111
		112	Every C<aio_*> function creates a request. which is a C data structure not
		113	directly visible to Perl.
		114
		115	If called in non-void context, every request function returns a Perl
		116	object representing the request. In void context, nothing is returned,
		117	which saves a bit of memory.
		118
		119	The perl object is a fairly standard ref-to-hash object. The hash contents
		120	are not used by IO::AIO so you are free to store anything you like in it.
		121
		122	During their existance, aio requests travel through the following states,
		123	in order:
		124
		125	=over 4
		126
		127	=item ready
		128
		129	Immediately after a request is created it is put into the ready state,
		130	waiting for a thread to execute it.
		131
		132	=item execute
		133
		134	A thread has accepted the request for processing and is currently
		135	executing it (e.g. blocking in read).
		136
		137	=item pending
		138
		139	The request has been executed and is waiting for result processing.
		140
		141	While request submission and execution is fully asynchronous, result
		142	processing is not and relies on the perl interpreter calling C<poll_cb>
		143	(or another function with the same effect).
		144
		145	=item result
		146
		147	The request results are processed synchronously by C<poll_cb>.
		148
		149	The C<poll_cb> function will process all outstanding aio requests by
		150	calling their callbacks, freeing memory associated with them and managing
		151	any groups they are contained in.
		152
		153	=item done
		154
		155	Request has reached the end of its lifetime and holds no resources anymore
		156	(except possibly for the Perl object, but its connection to the actual
		157	aio request is severed and calling its methods will either do nothing or
		158	result in a runtime error).
		159
		160	=back
60		161
61	=cut	162	=cut
62		163
63	package IO::AIO;	164	package IO::AIO;
64		165
65	no warnings;	166	use Carp ();
66	use strict 'vars';	167
		168	use common::sense;
67		169
68	use base 'Exporter';	170	use base 'Exporter';
69		171
70	BEGIN {	172	BEGIN {
71	our $VERSION = '1.8';	173	our $VERSION = '4.0';
72		174
73	our @EXPORT = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat	175	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close
74	aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink	176	aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
75	aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move);	177	aio_scandir aio_symlink aio_readlink aio_realpath aio_sync
76	our @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	178	aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range aio_fallocate
		179	aio_pathsync aio_readahead
		180	aio_rename aio_link aio_move aio_copy aio_group
		181	aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
		182	aio_chmod aio_utime aio_truncate
		183	aio_msync aio_mtouch aio_mlock aio_mlockall
		184	aio_statvfs);
		185
		186	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		187	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		188	min_parallel max_parallel max_idle idle_timeout
		189	nreqs nready npending nthreads
		190	max_poll_time max_poll_reqs
		191	sendfile fadvise madvise
		192	mmap munmap munlock munlockall);
		193
		194	push @AIO_REQ, qw(aio_busy); # not exported
		195
		196	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
77		197
78	require XSLoader;	198	require XSLoader;
79	XSLoader::load ("IO::AIO", $VERSION);	199	XSLoader::load ("IO::AIO", $VERSION);
80	}	200	}
81		201
82	=head1 FUNCTIONS	202	=head1 FUNCTIONS
83		203
		204	=head2 QUICK OVERVIEW
		205
		206	This section simply lists the prototypes of the most important functions
		207	for quick reference. See the following sections for function-by-function
		208	documentation.
		209
		210	aio_open $pathname, $flags, $mode, $callback->($fh)
		211	aio_close $fh, $callback->($status)
		212	aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		213	aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		214	aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		215	aio_readahead $fh,$offset,$length, $callback->($retval)
		216	aio_stat $fh_or_path, $callback->($status)
		217	aio_lstat $fh, $callback->($status)
		218	aio_statvfs $fh_or_path, $callback->($statvfs)
		219	aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		220	aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		221	aio_truncate $fh_or_path, $offset, $callback->($status)
		222	aio_chmod $fh_or_path, $mode, $callback->($status)
		223	aio_unlink $pathname, $callback->($status)
		224	aio_mknod $path, $mode, $dev, $callback->($status)
		225	aio_link $srcpath, $dstpath, $callback->($status)
		226	aio_symlink $srcpath, $dstpath, $callback->($status)
		227	aio_readlink $path, $callback->($link)
		228	aio_realpath $path, $callback->($link)
		229	aio_rename $srcpath, $dstpath, $callback->($status)
		230	aio_mkdir $pathname, $mode, $callback->($status)
		231	aio_rmdir $pathname, $callback->($status)
		232	aio_readdir $pathname, $callback->($entries)
		233	aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		234	IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
		235	IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
		236	aio_load $path, $data, $callback->($status)
		237	aio_copy $srcpath, $dstpath, $callback->($status)
		238	aio_move $srcpath, $dstpath, $callback->($status)
		239	aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		240	aio_rmtree $path, $callback->($status)
		241	aio_sync $callback->($status)
		242	aio_syncfs $fh, $callback->($status)
		243	aio_fsync $fh, $callback->($status)
		244	aio_fdatasync $fh, $callback->($status)
		245	aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		246	aio_pathsync $path, $callback->($status)
		247	aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		248	aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		249	aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		250	aio_mlockall $flags, $callback->($status)
		251	aio_group $callback->(...)
		252	aio_nop $callback->()
		253
		254	$prev_pri = aioreq_pri [$pri]
		255	aioreq_nice $pri_adjust
		256
		257	IO::AIO::poll_wait
		258	IO::AIO::poll_cb
		259	IO::AIO::poll
		260	IO::AIO::flush
		261	IO::AIO::max_poll_reqs $nreqs
		262	IO::AIO::max_poll_time $seconds
		263	IO::AIO::min_parallel $nthreads
		264	IO::AIO::max_parallel $nthreads
		265	IO::AIO::max_idle $nthreads
		266	IO::AIO::idle_timeout $seconds
		267	IO::AIO::max_outstanding $maxreqs
		268	IO::AIO::nreqs
		269	IO::AIO::nready
		270	IO::AIO::npending
		271
		272	IO::AIO::sendfile $ofh, $ifh, $offset, $count
		273	IO::AIO::fadvise $fh, $offset, $len, $advice
		274	IO::AIO::madvise $scalar, $offset, $length, $advice
		275	IO::AIO::mprotect $scalar, $offset, $length, $protect
		276	IO::AIO::munlock $scalar, $offset = 0, $length = undef
		277	IO::AIO::munlockall
		278
84	=head2 AIO FUNCTIONS	279	=head2 AIO REQUEST FUNCTIONS
85		280
86	All the C<aio_*> calls are more or less thin wrappers around the syscall	281	All the C<aio_*> calls are more or less thin wrappers around the syscall
87	with the same name (sans C<aio_>). The arguments are similar or identical,	282	with the same name (sans C<aio_>). The arguments are similar or identical,
88	and they all accept an additional (and optional) C<$callback> argument	283	and they all accept an additional (and optional) C<$callback> argument
89	which must be a code reference. This code reference will get called with	284	which must be a code reference. This code reference will get called with
90	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	285	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
91	perl, which usually delivers "false") as it's sole argument when the given	286	perl, which usually delivers "false") as its sole argument after the given
92	syscall has been executed asynchronously.	287	syscall has been executed asynchronously.
93		288
94	All functions expecting a filehandle keep a copy of the filehandle	289	All functions expecting a filehandle keep a copy of the filehandle
95	internally until the request has finished.	290	internally until the request has finished.
96		291
		292	All functions return request objects of type L<IO::AIO::REQ> that allow
		293	further manipulation of those requests while they are in-flight.
		294
97	The pathnames you pass to these routines I<must> be absolute and	295	The pathnames you pass to these routines I<must> be absolute and
98	encoded in byte form. The reason for the former is that at the time the	296	encoded as octets. The reason for the former is that at the time the
99	request is being executed, the current working directory could have	297	request is being executed, the current working directory could have
100	changed. Alternatively, you can make sure that you never change the	298	changed. Alternatively, you can make sure that you never change the
101	current working directory.	299	current working directory anywhere in the program and then use relative
		300	paths.
102		301
103	To encode pathnames to byte form, either make sure you either: a)	302	To encode pathnames as octets, either make sure you either: a) always pass
104	always pass in filenames you got from outside (command line, readdir	303	in filenames you got from outside (command line, readdir etc.) without
105	etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode	304	tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
106	your pathnames to the locale (or other) encoding in effect in the user	305	your pathnames to the locale (or other) encoding in effect in the user
107	environment, d) use Glib::filename_from_unicode on unicode filenames or e)	306	environment, d) use Glib::filename_from_unicode on unicode filenames or e)
108	use something else.	307	use something else to ensure your scalar has the correct contents.
		308
		309	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		310	handles correctly whether it is set or not.
109		311
110	=over 4	312	=over 4
		313
		314	=item $prev_pri = aioreq_pri [$pri]
		315
		316	Returns the priority value that would be used for the next request and, if
		317	C<$pri> is given, sets the priority for the next aio request.
		318
		319	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		320	and C<4>, respectively. Requests with higher priority will be serviced
		321	first.
		322
		323	The priority will be reset to C<0> after each call to one of the C<aio_*>
		324	functions.
		325
		326	Example: open a file with low priority, then read something from it with
		327	higher priority so the read request is serviced before other low priority
		328	open requests (potentially spamming the cache):
		329
		330	aioreq_pri -3;
		331	aio_open ..., sub {
		332	return unless $_[0];
		333
		334	aioreq_pri -2;
		335	aio_read $_[0], ..., sub {
		336	...
		337	};
		338	};
		339
		340
		341	=item aioreq_nice $pri_adjust
		342
		343	Similar to C<aioreq_pri>, but subtracts the given value from the current
		344	priority, so the effect is cumulative.
		345
111		346
112	=item aio_open $pathname, $flags, $mode, $callback->($fh)	347	=item aio_open $pathname, $flags, $mode, $callback->($fh)
113		348
114	Asynchronously open or create a file and call the callback with a newly	349	Asynchronously open or create a file and call the callback with a newly
115	created filehandle for the file.	350	created filehandle for the file.
…		…
121	list. They are the same as used by C<sysopen>.	356	list. They are the same as used by C<sysopen>.
122		357
123	Likewise, C<$mode> specifies the mode of the newly created file, if it	358	Likewise, C<$mode> specifies the mode of the newly created file, if it
124	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,	359	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
125	except that it is mandatory (i.e. use C<0> if you don't create new files,	360	except that it is mandatory (i.e. use C<0> if you don't create new files,
126	and C<0666> or C<0777> if you do).	361	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		362	by the umask in effect then the request is being executed, so better never
		363	change the umask.
127		364
128	Example:	365	Example:
129		366
130	aio_open "/etc/passwd", O_RDONLY, 0, sub {	367	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
131	if ($_[0]) {	368	if ($_[0]) {
132	print "open successful, fh is $_[0]\n";	369	print "open successful, fh is $_[0]\n";
133	...	370	...
134	} else {	371	} else {
135	die "open failed: $!\n";	372	die "open failed: $!\n";
136	}	373	}
137	};	374	};
138		375
		376	In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
		377	C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
		378	following POSIX and non-POSIX constants are available (missing ones on
		379	your system are, as usual, C<0>):
		380
		381	C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
		382	C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
		383	C<O_RSYNC>, C<O_SYNC> and C<O_TTY_INIT>.
		384
		385
139	=item aio_close $fh, $callback->($status)	386	=item aio_close $fh, $callback->($status)
140		387
141	Asynchronously close a file and call the callback with the result	388	Asynchronously close a file and call the callback with the result
142	code. I<WARNING:> although accepted, you should not pass in a perl	389	code.
143	filehandle here, as perl will likely close the file descriptor another
144	time when the filehandle is destroyed. Normally, you can safely call perls
145	C<close> or just let filehandles go out of scope.
146		390
147	This is supposed to be a bug in the API, so that might change. It's	391	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
148	therefore best to avoid this function.	392	closing the file descriptor associated with the filehandle itself.
		393
		394	Therefore, C<aio_close> will not close the filehandle - instead it will
		395	use dup2 to overwrite the file descriptor with the write-end of a pipe
		396	(the pipe fd will be created on demand and will be cached).
		397
		398	Or in other words: the file descriptor will be closed, but it will not be
		399	free for reuse until the perl filehandle is closed.
		400
		401	=cut
149		402
150	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	403	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
151		404
152	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	405	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
153		406
154	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	407	Reads or writes C<$length> bytes from or to the specified C<$fh> and
155	into the scalar given by C<data> and offset C<dataoffset> and calls the	408	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset>
156	callback without the actual number of bytes read (or -1 on error, just	409	and calls the callback without the actual number of bytes read (or -1 on
157	like the syscall).	410	error, just like the syscall).
		411
		412	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		413	offset plus the actual number of bytes read.
		414
		415	If C<$offset> is undefined, then the current file descriptor offset will
		416	be used (and updated), otherwise the file descriptor offset will not be
		417	changed by these calls.
		418
		419	If C<$length> is undefined in C<aio_write>, use the remaining length of
		420	C<$data>.
		421
		422	If C<$dataoffset> is less than zero, it will be counted from the end of
		423	C<$data>.
158		424
159	The C<$data> scalar I<MUST NOT> be modified in any way while the request	425	The C<$data> scalar I<MUST NOT> be modified in any way while the request
160	is outstanding. Modifying it can result in segfaults or WW3 (if the	426	is outstanding. Modifying it can result in segfaults or World War III (if
161	necessary/optional hardware is installed).	427	the necessary/optional hardware is installed).
162		428
163	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	429	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
164	offset C<0> within the scalar:	430	offset C<0> within the scalar:
165		431
166	aio_read $fh, 7, 15, $buffer, 0, sub {	432	aio_read $fh, 7, 15, $buffer, 0, sub {
167	$_[0] > 0 or die "read error: $!";	433	$_[0] > 0 or die "read error: $!";
168	print "read $_[0] bytes: <$buffer>\n";	434	print "read $_[0] bytes: <$buffer>\n";
169	};	435	};
170		436
171	=item aio_move $srcpath, $dstpath, $callback->($status)
172
173	[EXPERIMENTAL]
174
175	Try to move the I<file> (directories not supported as either source or destination)
176	from C<$srcpath> to C<$dstpath> and call the callback with the C<0> (error) or C<-1> ok.
177
178	This is a composite request that tries to rename(2) the file first. If
179	rename files with C<EXDEV>, it creates the destination file with mode 0200
180	and copies the contents of the source file into it using C<aio_sendfile>,
181	followed by restoring atime, mtime, access mode and uid/gid, in that
182	order, and unlinking the C<$srcpath>.
183
184	If an error occurs, the partial destination file will be unlinked, if
185	possible, except when setting atime, mtime, access mode and uid/gid, where
186	errors are being ignored.
187
188	=cut
189
190	sub aio_move($$$) {
191	my ($src, $dst, $cb) = @_;
192
193	aio_rename $src, $dst, sub {
194	if ($_[0] && $! == EXDEV) {
195	aio_open $src, O_RDONLY, 0, sub {
196	if (my $src_fh = $_[0]) {
197	my @stat = stat $src_fh;
198
199	aio_open $dst, O_WRONLY, 0200, sub {
200	if (my $dst_fh = $_[0]) {
201	aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
202	close $src_fh;
203
204	if ($_[0] == $stat[7]) {
205	utime $stat[8], $stat[9], $dst;
206	chmod $stat[2] & 07777, $dst_fh;
207	chown $stat[4], $stat[5], $dst_fh;
208	close $dst_fh;
209
210	aio_unlink $src, sub {
211	$cb->($_[0]);
212	};
213	} else {
214	my $errno = $!;
215	aio_unlink $dst, sub {
216	$! = $errno;
217	$cb->(-1);
218	};
219	}
220	};
221	} else {
222	$cb->(-1);
223	}
224	},
225
226	} else {
227	$cb->(-1);
228	}
229	};
230	} else {
231	$cb->($_[0]);
232	}
233	};
234	}
235		437
236	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)	438	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
237		439
238	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts	440	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
239	reading at byte offset C<$in_offset>, and starts writing at the current	441	reading at byte offset C<$in_offset>, and starts writing at the current
240	file offset of C<$out_fh>. Because of that, it is not safe to issue more	442	file offset of C<$out_fh>. Because of that, it is not safe to issue more
241	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each	443	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
242	other.	444	other. The same C<$in_fh> works fine though, as this function does not
		445	move or use the file offset of C<$in_fh>.
243		446
		447	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		448	are written, and there is no way to find out how many more bytes have been
		449	read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
		450	number of bytes written to C<$out_fh>. Only if the result value equals
		451	C<$length> one can assume that C<$length> bytes have been read.
		452
		453	Unlike with other C<aio_> functions, it makes a lot of sense to use
		454	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		455	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		456	the socket I/O will be non-blocking. Note, however, that you can run
		457	into a trap where C<aio_sendfile> reads some data with readahead, then
		458	fails to write all data, and when the socket is ready the next time, the
		459	data in the cache is already lost, forcing C<aio_sendfile> to again hit
		460	the disk. Explicit C<aio_read> + C<aio_write> let's you better control
		461	resource usage.
		462
244	This call tries to make use of a native C<sendfile> syscall to provide	463	This call tries to make use of a native C<sendfile>-like syscall to
245	zero-copy operation. For this to work, C<$out_fh> should refer to a	464	provide zero-copy operation. For this to work, C<$out_fh> should refer to
246	socket, and C<$in_fh> should refer to mmap'able file.	465	a socket, and C<$in_fh> should refer to an mmap'able file.
247		466
248	If the native sendfile call fails or is not implemented, it will be	467	If a native sendfile cannot be found or it fails with C<ENOSYS>,
249	emulated, so you can call C<aio_sendfile> on any type of filehandle	468	C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
		469	C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
250	regardless of the limitations of the operating system.	470	type of filehandle regardless of the limitations of the operating system.
251		471
252	Please note, however, that C<aio_sendfile> can read more bytes from	472	As native sendfile syscalls (as practically any non-POSIX interface hacked
253	C<$in_fh> than are written, and there is no way to find out how many	473	together in a hurry to improve benchmark numbers) tend to be rather buggy
254	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only	474	on many systems, this implementation tries to work around some known bugs
255	provides the number of bytes written to C<$out_fh>. Only if the result	475	in Linux and FreeBSD kernels (probably others, too), but that might fail,
256	value equals C<$length> one can assume that C<$length> bytes have been	476	so you really really should check the return value of C<aio_sendfile> -
257	read.	477	fewre bytes than expected might have been transferred.
		478
258		479
259	=item aio_readahead $fh,$offset,$length, $callback->($retval)	480	=item aio_readahead $fh,$offset,$length, $callback->($retval)
260		481
261	C<aio_readahead> populates the page cache with data from a file so that	482	C<aio_readahead> populates the page cache with data from a file so that
262	subsequent reads from that file will not block on disk I/O. The C<$offset>	483	subsequent reads from that file will not block on disk I/O. The C<$offset>
…		…
268	file. The current file offset of the file is left unchanged.	489	file. The current file offset of the file is left unchanged.
269		490
270	If that syscall doesn't exist (likely if your OS isn't Linux) it will be	491	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
271	emulated by simply reading the data, which would have a similar effect.	492	emulated by simply reading the data, which would have a similar effect.
272		493
		494
273	=item aio_stat $fh_or_path, $callback->($status)	495	=item aio_stat $fh_or_path, $callback->($status)
274		496
275	=item aio_lstat $fh, $callback->($status)	497	=item aio_lstat $fh, $callback->($status)
276		498
277	Works like perl's C<stat> or C<lstat> in void context. The callback will	499	Works like perl's C<stat> or C<lstat> in void context. The callback will
…		…
282	for an explanation.	504	for an explanation.
283		505
284	Currently, the stats are always 64-bit-stats, i.e. instead of returning an	506	Currently, the stats are always 64-bit-stats, i.e. instead of returning an
285	error when stat'ing a large file, the results will be silently truncated	507	error when stat'ing a large file, the results will be silently truncated
286	unless perl itself is compiled with large file support.	508	unless perl itself is compiled with large file support.
		509
		510	To help interpret the mode and dev/rdev stat values, IO::AIO offers the
		511	following constants and functions (if not implemented, the constants will
		512	be C<0> and the functions will either C<croak> or fall back on traditional
		513	behaviour).
		514
		515	C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
		516	C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
		517	C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
287		518
288	Example: Print the length of F</etc/passwd>:	519	Example: Print the length of F</etc/passwd>:
289		520
290	aio_stat "/etc/passwd", sub {	521	aio_stat "/etc/passwd", sub {
291	$_[0] and die "stat failed: $!";	522	$_[0] and die "stat failed: $!";
292	print "size is ", -s _, "\n";	523	print "size is ", -s _, "\n";
293	};	524	};
294		525
		526
		527	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		528
		529	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		530	whether a file handle or path was passed.
		531
		532	On success, the callback is passed a hash reference with the following
		533	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		534	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		535	is passed.
		536
		537	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		538	C<ST_NOSUID>.
		539
		540	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		541	their correct value when available, or to C<0> on systems that do
		542	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		543	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		544	C<ST_NODIRATIME> and C<ST_RELATIME>.
		545
		546	Example: stat C</wd> and dump out the data if successful.
		547
		548	aio_statvfs "/wd", sub {
		549	my $f = $_[0]
		550	or die "statvfs: $!";
		551
		552	use Data::Dumper;
		553	say Dumper $f;
		554	};
		555
		556	# result:
		557	{
		558	bsize => 1024,
		559	bfree => 4333064312,
		560	blocks => 10253828096,
		561	files => 2050765568,
		562	flag => 4096,
		563	favail => 2042092649,
		564	bavail => 4333064312,
		565	ffree => 2042092649,
		566	namemax => 255,
		567	frsize => 1024,
		568	fsid => 1810
		569	}
		570
		571
		572	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		573
		574	Works like perl's C<utime> function (including the special case of $atime
		575	and $mtime being undef). Fractional times are supported if the underlying
		576	syscalls support them.
		577
		578	When called with a pathname, uses utimes(2) if available, otherwise
		579	utime(2). If called on a file descriptor, uses futimes(2) if available,
		580	otherwise returns ENOSYS, so this is not portable.
		581
		582	Examples:
		583
		584	# set atime and mtime to current time (basically touch(1)):
		585	aio_utime "path", undef, undef;
		586	# set atime to current time and mtime to beginning of the epoch:
		587	aio_utime "path", time, undef; # undef==0
		588
		589
		590	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		591
		592	Works like perl's C<chown> function, except that C<undef> for either $uid
		593	or $gid is being interpreted as "do not change" (but -1 can also be used).
		594
		595	Examples:
		596
		597	# same as "chown root path" in the shell:
		598	aio_chown "path", 0, -1;
		599	# same as above:
		600	aio_chown "path", 0, undef;
		601
		602
		603	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		604
		605	Works like truncate(2) or ftruncate(2).
		606
		607
		608	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		609
		610	Works like perl's C<chmod> function.
		611
		612
295	=item aio_unlink $pathname, $callback->($status)	613	=item aio_unlink $pathname, $callback->($status)
296		614
297	Asynchronously unlink (delete) a file and call the callback with the	615	Asynchronously unlink (delete) a file and call the callback with the
298	result code.	616	result code.
299		617
		618
		619	=item aio_mknod $path, $mode, $dev, $callback->($status)
		620
		621	[EXPERIMENTAL]
		622
		623	Asynchronously create a device node (or fifo). See mknod(2).
		624
		625	The only (POSIX-) portable way of calling this function is:
		626
		627	aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		628
		629	See C<aio_stat> for info about some potentially helpful extra constants
		630	and functions.
		631
300	=item aio_link $srcpath, $dstpath, $callback->($status)	632	=item aio_link $srcpath, $dstpath, $callback->($status)
301		633
302	Asynchronously create a new link to the existing object at C<$srcpath> at	634	Asynchronously create a new link to the existing object at C<$srcpath> at
303	the path C<$dstpath> and call the callback with the result code.	635	the path C<$dstpath> and call the callback with the result code.
304		636
		637
305	=item aio_symlink $srcpath, $dstpath, $callback->($status)	638	=item aio_symlink $srcpath, $dstpath, $callback->($status)
306		639
307	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at	640	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
308	the path C<$dstpath> and call the callback with the result code.	641	the path C<$dstpath> and call the callback with the result code.
309		642
		643
		644	=item aio_readlink $path, $callback->($link)
		645
		646	Asynchronously read the symlink specified by C<$path> and pass it to
		647	the callback. If an error occurs, nothing or undef gets passed to the
		648	callback.
		649
		650
		651	=item aio_realpath $path, $callback->($path)
		652
		653	Asynchronously make the path absolute and resolve any symlinks in
		654	C<$path>. The resulting path only consists of directories (Same as
		655	L<Cwd::realpath>).
		656
		657	This request can be used to get the absolute path of the current working
		658	directory by passing it a path of F<.> (a single dot).
		659
		660
310	=item aio_rename $srcpath, $dstpath, $callback->($status)	661	=item aio_rename $srcpath, $dstpath, $callback->($status)
311		662
312	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as	663	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
313	rename(2) and call the callback with the result code.	664	rename(2) and call the callback with the result code.
314		665
		666
		667	=item aio_mkdir $pathname, $mode, $callback->($status)
		668
		669	Asynchronously mkdir (create) a directory and call the callback with
		670	the result code. C<$mode> will be modified by the umask at the time the
		671	request is executed, so do not change your umask.
		672
		673
315	=item aio_rmdir $pathname, $callback->($status)	674	=item aio_rmdir $pathname, $callback->($status)
316		675
317	Asynchronously rmdir (delete) a directory and call the callback with the	676	Asynchronously rmdir (delete) a directory and call the callback with the
318	result code.	677	result code.
		678
319		679
320	=item aio_readdir $pathname, $callback->($entries)	680	=item aio_readdir $pathname, $callback->($entries)
321		681
322	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire	682	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
323	directory (i.e. opendir + readdir + closedir). The entries will not be	683	directory (i.e. opendir + readdir + closedir). The entries will not be
324	sorted, and will B<NOT> include the C<.> and C<..> entries.	684	sorted, and will B<NOT> include the C<.> and C<..> entries.
325		685
326	The callback a single argument which is either C<undef> or an array-ref	686	The callback is passed a single argument which is either C<undef> or an
327	with the filenames.	687	array-ref with the filenames.
		688
		689
		690	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		691
		692	Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
		693	tune behaviour and output format. In case of an error, C<$entries> will be
		694	C<undef>.
		695
		696	The flags are a combination of the following constants, ORed together (the
		697	flags will also be passed to the callback, possibly modified):
		698
		699	=over 4
		700
		701	=item IO::AIO::READDIR_DENTS
		702
		703	When this flag is off, then the callback gets an arrayref consisting of
		704	names only (as with C<aio_readdir>), otherwise it gets an arrayref with
		705	C<[$name, $type, $inode]> arrayrefs, each describing a single directory
		706	entry in more detail.
		707
		708	C<$name> is the name of the entry.
		709
		710	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		711
		712	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		713	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		714	C<IO::AIO::DT_WHT>.
		715
		716	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
		717	know, you have to run stat yourself. Also, for speed reasons, the C<$type>
		718	scalars are read-only: you can not modify them.
		719
		720	C<$inode> is the inode number (which might not be exact on systems with 64
		721	bit inode numbers and 32 bit perls). This field has unspecified content on
		722	systems that do not deliver the inode information.
		723
		724	=item IO::AIO::READDIR_DIRS_FIRST
		725
		726	When this flag is set, then the names will be returned in an order where
		727	likely directories come first, in optimal stat order. This is useful when
		728	you need to quickly find directories, or you want to find all directories
		729	while avoiding to stat() each entry.
		730
		731	If the system returns type information in readdir, then this is used
		732	to find directories directly. Otherwise, likely directories are names
		733	beginning with ".", or otherwise names with no dots, of which names with
		734	short names are tried first.
		735
		736	=item IO::AIO::READDIR_STAT_ORDER
		737
		738	When this flag is set, then the names will be returned in an order
		739	suitable for stat()'ing each one. That is, when you plan to stat()
		740	all files in the given directory, then the returned order will likely
		741	be fastest.
		742
		743	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
		744	the likely dirs come first, resulting in a less optimal stat order.
		745
		746	=item IO::AIO::READDIR_FOUND_UNKNOWN
		747
		748	This flag should not be set when calling C<aio_readdirx>. Instead, it
		749	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		750	C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
		751	C<$type>'s are known, which can be used to speed up some algorithms.
		752
		753	=back
		754
		755
		756	=item aio_load $path, $data, $callback->($status)
		757
		758	This is a composite request that tries to fully load the given file into
		759	memory. Status is the same as with aio_read.
		760
		761	=cut
		762
		763	sub aio_load($$;$) {
		764	my ($path, undef, $cb) = @_;
		765	my $data = \$_[1];
		766
		767	my $pri = aioreq_pri;
		768	my $grp = aio_group $cb;
		769
		770	aioreq_pri $pri;
		771	add $grp aio_open $path, O_RDONLY, 0, sub {
		772	my $fh = shift
		773	or return $grp->result (-1);
		774
		775	aioreq_pri $pri;
		776	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		777	$grp->result ($_[0]);
		778	};
		779	};
		780
		781	$grp
		782	}
		783
		784	=item aio_copy $srcpath, $dstpath, $callback->($status)
		785
		786	Try to copy the I<file> (directories not supported as either source or
		787	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		788	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		789
		790	This is a composite request that creates the destination file with
		791	mode 0200 and copies the contents of the source file into it using
		792	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		793	uid/gid, in that order.
		794
		795	If an error occurs, the partial destination file will be unlinked, if
		796	possible, except when setting atime, mtime, access mode and uid/gid, where
		797	errors are being ignored.
		798
		799	=cut
		800
		801	sub aio_copy($$;$) {
		802	my ($src, $dst, $cb) = @_;
		803
		804	my $pri = aioreq_pri;
		805	my $grp = aio_group $cb;
		806
		807	aioreq_pri $pri;
		808	add $grp aio_open $src, O_RDONLY, 0, sub {
		809	if (my $src_fh = $_[0]) {
		810	my @stat = stat $src_fh; # hmm, might block over nfs?
		811
		812	aioreq_pri $pri;
		813	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		814	if (my $dst_fh = $_[0]) {
		815	aioreq_pri $pri;
		816	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		817	if ($_[0] == $stat[7]) {
		818	$grp->result (0);
		819	close $src_fh;
		820
		821	my $ch = sub {
		822	aioreq_pri $pri;
		823	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		824	aioreq_pri $pri;
		825	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		826	aioreq_pri $pri;
		827	add $grp aio_close $dst_fh;
		828	}
		829	};
		830	};
		831
		832	aioreq_pri $pri;
		833	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		834	if ($_[0] < 0 && $! == ENOSYS) {
		835	aioreq_pri $pri;
		836	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		837	} else {
		838	$ch->();
		839	}
		840	};
		841	} else {
		842	$grp->result (-1);
		843	close $src_fh;
		844	close $dst_fh;
		845
		846	aioreq $pri;
		847	add $grp aio_unlink $dst;
		848	}
		849	};
		850	} else {
		851	$grp->result (-1);
		852	}
		853	},
		854
		855	} else {
		856	$grp->result (-1);
		857	}
		858	};
		859
		860	$grp
		861	}
		862
		863	=item aio_move $srcpath, $dstpath, $callback->($status)
		864
		865	Try to move the I<file> (directories not supported as either source or
		866	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		867	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		868
		869	This is a composite request that tries to rename(2) the file first; if
		870	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		871	that is successful, unlinks the C<$srcpath>.
		872
		873	=cut
		874
		875	sub aio_move($$;$) {
		876	my ($src, $dst, $cb) = @_;
		877
		878	my $pri = aioreq_pri;
		879	my $grp = aio_group $cb;
		880
		881	aioreq_pri $pri;
		882	add $grp aio_rename $src, $dst, sub {
		883	if ($_[0] && $! == EXDEV) {
		884	aioreq_pri $pri;
		885	add $grp aio_copy $src, $dst, sub {
		886	$grp->result ($_[0]);
		887
		888	unless ($_[0]) {
		889	aioreq_pri $pri;
		890	add $grp aio_unlink $src;
		891	}
		892	};
		893	} else {
		894	$grp->result ($_[0]);
		895	}
		896	};
		897
		898	$grp
		899	}
328		900
329	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)	901	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
330		902
331	Scans a directory (similar to C<aio_readdir>) and tries to separate the	903	Scans a directory (similar to C<aio_readdir>) but additionally tries to
332	entries of directory C<$path> into two sets of names, ones you can recurse	904	efficiently separate the entries of directory C<$path> into two sets of
333	into (directories), and ones you cannot recurse into (everything else).	905	names, directories you can recurse into (directories), and ones you cannot
		906	recurse into (everything else, including symlinks to directories).
334		907
335	C<aio_scandir> is a composite request that consists of many	908	C<aio_scandir> is a composite request that creates of many sub requests_
336	aio-primitives. C<$maxreq> specifies the maximum number of outstanding	909	C<$maxreq> specifies the maximum number of outstanding aio requests that
337	aio requests that this function generates. If it is C<< <= 0 >>, then a	910	this function generates. If it is C<< <= 0 >>, then a suitable default
338	suitable default will be chosen (currently 8).	911	will be chosen (currently 4).
339		912
340	On error, the callback is called without arguments, otherwise it receives	913	On error, the callback is called without arguments, otherwise it receives
341	two array-refs with path-relative entry names.	914	two array-refs with path-relative entry names.
342		915
343	Example:	916	Example:
…		…
350		923
351	Implementation notes.	924	Implementation notes.
352		925
353	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.	926	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
354		927
		928	If readdir returns file type information, then this is used directly to
		929	find directories.
		930
355	After reading the directory, the modification time, size etc. of the	931	Otherwise, after reading the directory, the modification time, size etc.
356	directory before and after the readdir is checked, and if they match, the	932	of the directory before and after the readdir is checked, and if they
357	link count will be used to decide how many entries are directories (if	933	match (and isn't the current time), the link count will be used to decide
358	>= 2). Otherwise, no knowledge of the number of subdirectories will be	934	how many entries are directories (if >= 2). Otherwise, no knowledge of the
359	assumed.	935	number of subdirectories will be assumed.
360		936
361	Then entires will be sorted into likely directories (everything without a	937	Then entries will be sorted into likely directories a non-initial dot
362	non-initial dot) and likely non-directories (everything else). Then every	938	currently) and likely non-directories (see C<aio_readdirx>). Then every
363	entry + C</.> will be C<stat>'ed, likely directories first. This is often	939	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
		940	in order of their inode numbers. If that succeeds, it assumes that the
		941	entry is a directory or a symlink to directory (which will be checked
		942	separately). This is often faster than stat'ing the entry itself because
364	faster because filesystems might detect the type of the entry without	943	filesystems might detect the type of the entry without reading the inode
365	reading the inode data (e.g. ext2fs filetype feature). If that succeeds,	944	data (e.g. ext2fs filetype feature), even on systems that cannot return
366	it assumes that the entry is a directory or a symlink to directory (which	945	the filetype information on readdir.
367	will be checked seperately).
368		946
369	If the known number of directories has been reached, the rest of the	947	If the known number of directories (link count - 2) has been reached, the
370	entries is assumed to be non-directories.	948	rest of the entries is assumed to be non-directories.
		949
		950	This only works with certainty on POSIX (= UNIX) filesystems, which
		951	fortunately are the vast majority of filesystems around.
		952
		953	It will also likely work on non-POSIX filesystems with reduced efficiency
		954	as those tend to return 0 or 1 as link counts, which disables the
		955	directory counting heuristic.
371		956
372	=cut	957	=cut
373		958
374	sub aio_scandir($$$) {	959	sub aio_scandir($$;$) {
375	my ($path, $maxreq, $cb) = @_;	960	my ($path, $maxreq, $cb) = @_;
376		961
		962	my $pri = aioreq_pri;
		963
		964	my $grp = aio_group $cb;
		965
377	$maxreq = 8 if $maxreq <= 0;	966	$maxreq = 4 if $maxreq <= 0;
378		967
379	# stat once	968	# stat once
		969	aioreq_pri $pri;
380	aio_stat $path, sub {	970	add $grp aio_stat $path, sub {
381	return $cb->() if $_[0];	971	return $grp->result () if $_[0];
		972	my $now = time;
382	my $hash1 = join ":", (stat _)[0,1,3,7,9];	973	my $hash1 = join ":", (stat _)[0,1,3,7,9];
383		974
384	# read the directory entries	975	# read the directory entries
385	aio_readdir $path, sub {	976	aioreq_pri $pri;
		977	add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub {
386	my $entries = shift	978	my $entries = shift
387	or return $cb->();	979	or return $grp->result ();
388		980
389	# stat the dir another time	981	# stat the dir another time
		982	aioreq_pri $pri;
390	aio_stat $path, sub {	983	add $grp aio_stat $path, sub {
391	my $hash2 = join ":", (stat _)[0,1,3,7,9];	984	my $hash2 = join ":", (stat _)[0,1,3,7,9];
392		985
393	my $ndirs;	986	my $ndirs;
394		987
395	# take the slow route if anything looks fishy	988	# take the slow route if anything looks fishy
396	if ($hash1 ne $hash2) {	989	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
397	$ndirs = -1;	990	$ndirs = -1;
398	} else {	991	} else {
399	# if nlink == 2, we are finished	992	# if nlink == 2, we are finished
400	# on non-posix-fs's, we rely on nlink < 2	993	# for non-posix-fs's, we rely on nlink < 2
401	$ndirs = (stat _)[3] - 2	994	$ndirs = (stat _)[3] - 2
402	or return $cb->([], $entries);	995	or return $grp->result ([], $entries);
403	}	996	}
404		997
405	# sort into likely dirs and likely nondirs
406	# dirs == files without ".", short entries first
407	$entries = [map $_->[0],
408	sort { $b->[1] cmp $a->[1] }
409	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
410	@$entries];
411
412	my (@dirs, @nondirs);	998	my (@dirs, @nondirs);
413		999
414	my ($statcb, $schedcb);	1000	my $statgrp = add $grp aio_group sub {
415	my $nreq = 0;	1001	$grp->result (\@dirs, \@nondirs);
		1002	};
416		1003
417	$schedcb = sub {	1004	limit $statgrp $maxreq;
418	if (@$entries) {	1005	feed $statgrp sub {
419	if ($nreq < $maxreq) {	1006	return unless @$entries;
420	my $ent = pop @$entries;	1007	my $entry = shift @$entries;
		1008
		1009	aioreq_pri $pri;
		1010	add $statgrp aio_stat "$path/$entry/.", sub {
		1011	if ($_[0] < 0) {
		1012	push @nondirs, $entry;
		1013	} else {
		1014	# need to check for real directory
		1015	aioreq_pri $pri;
		1016	add $statgrp aio_lstat "$path/$entry", sub {
		1017	if (-d _) {
		1018	push @dirs, $entry;
		1019
		1020	unless (--$ndirs) {
		1021	push @nondirs, @$entries;
		1022	feed $statgrp;
		1023	}
		1024	} else {
		1025	push @nondirs, $entry;
		1026	}
421	$nreq++;	1027	}
422	aio_stat "$path/$ent/.", sub { $statcb->($_[0], $ent) };
423	}	1028	}
424	} elsif (!$nreq) {
425	# finished
426	undef $statcb;
427	undef $schedcb;
428	$cb->(\@dirs, \@nondirs) if $cb;
429	undef $cb;
430	}	1029	};
431	};	1030	};
432	$statcb = sub {
433	my ($status, $entry) = @_;
434
435	if ($status < 0) {
436	$nreq--;
437	push @nondirs, $entry;
438	&$schedcb;
439	} else {
440	# need to check for real directory
441	aio_lstat "$path/$entry", sub {
442	$nreq--;
443
444	if (-d _) {
445	push @dirs, $entry;
446
447	if (!--$ndirs) {
448	push @nondirs, @$entries;
449	$entries = [];
450	}
451	} else {
452	push @nondirs, $entry;
453	}
454
455	&$schedcb;
456	}
457	}
458	};
459
460	&$schedcb while @$entries && $nreq < $maxreq;
461	};	1031	};
462	};	1032	};
463	};	1033	};
		1034
		1035	$grp
464	}	1036	}
		1037
		1038	=item aio_rmtree $path, $callback->($status)
		1039
		1040	Delete a directory tree starting (and including) C<$path>, return the
		1041	status of the final C<rmdir> only. This is a composite request that
		1042	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1043	everything else.
		1044
		1045	=cut
		1046
		1047	sub aio_rmtree;
		1048	sub aio_rmtree($;$) {
		1049	my ($path, $cb) = @_;
		1050
		1051	my $pri = aioreq_pri;
		1052	my $grp = aio_group $cb;
		1053
		1054	aioreq_pri $pri;
		1055	add $grp aio_scandir $path, 0, sub {
		1056	my ($dirs, $nondirs) = @_;
		1057
		1058	my $dirgrp = aio_group sub {
		1059	add $grp aio_rmdir $path, sub {
		1060	$grp->result ($_[0]);
		1061	};
		1062	};
		1063
		1064	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1065	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1066
		1067	add $grp $dirgrp;
		1068	};
		1069
		1070	$grp
		1071	}
		1072
		1073	=item aio_sync $callback->($status)
		1074
		1075	Asynchronously call sync and call the callback when finished.
465		1076
466	=item aio_fsync $fh, $callback->($status)	1077	=item aio_fsync $fh, $callback->($status)
467		1078
468	Asynchronously call fsync on the given filehandle and call the callback	1079	Asynchronously call fsync on the given filehandle and call the callback
469	with the fsync result code.	1080	with the fsync result code.
…		…
474	callback with the fdatasync result code.	1085	callback with the fdatasync result code.
475		1086
476	If this call isn't available because your OS lacks it or it couldn't be	1087	If this call isn't available because your OS lacks it or it couldn't be
477	detected, it will be emulated by calling C<fsync> instead.	1088	detected, it will be emulated by calling C<fsync> instead.
478		1089
		1090	=item aio_syncfs $fh, $callback->($status)
		1091
		1092	Asynchronously call the syncfs syscall to sync the filesystem associated
		1093	to the given filehandle and call the callback with the syncfs result
		1094	code. If syncfs is not available, calls sync(), but returns C<-1> and sets
		1095	errno to C<ENOSYS> nevertheless.
		1096
		1097	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1098
		1099	Sync the data portion of the file specified by C<$offset> and C<$length>
		1100	to disk (but NOT the metadata), by calling the Linux-specific
		1101	sync_file_range call. If sync_file_range is not available or it returns
		1102	ENOSYS, then fdatasync or fsync is being substituted.
		1103
		1104	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1105	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1106	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1107	manpage for details.
		1108
		1109	=item aio_pathsync $path, $callback->($status)
		1110
		1111	This request tries to open, fsync and close the given path. This is a
		1112	composite request intended to sync directories after directory operations
		1113	(E.g. rename). This might not work on all operating systems or have any
		1114	specific effect, but usually it makes sure that directory changes get
		1115	written to disc. It works for anything that can be opened for read-only,
		1116	not just directories.
		1117
		1118	Future versions of this function might fall back to other methods when
		1119	C<fsync> on the directory fails (such as calling C<sync>).
		1120
		1121	Passes C<0> when everything went ok, and C<-1> on error.
		1122
		1123	=cut
		1124
		1125	sub aio_pathsync($;$) {
		1126	my ($path, $cb) = @_;
		1127
		1128	my $pri = aioreq_pri;
		1129	my $grp = aio_group $cb;
		1130
		1131	aioreq_pri $pri;
		1132	add $grp aio_open $path, O_RDONLY, 0, sub {
		1133	my ($fh) = @_;
		1134	if ($fh) {
		1135	aioreq_pri $pri;
		1136	add $grp aio_fsync $fh, sub {
		1137	$grp->result ($_[0]);
		1138
		1139	aioreq_pri $pri;
		1140	add $grp aio_close $fh;
		1141	};
		1142	} else {
		1143	$grp->result (-1);
		1144	}
		1145	};
		1146
		1147	$grp
		1148	}
		1149
		1150	=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1151
		1152	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1153	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1154	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1155	scalar must only be modified in-place while an aio operation is pending on
		1156	it).
		1157
		1158	It calls the C<msync> function of your OS, if available, with the memory
		1159	area starting at C<$offset> in the string and ending C<$length> bytes
		1160	later. If C<$length> is negative, counts from the end, and if C<$length>
		1161	is C<undef>, then it goes till the end of the string. The flags can be
		1162	a combination of C<IO::AIO::MS_ASYNC>, C<IO::AIO::MS_INVALIDATE> and
		1163	C<IO::AIO::MS_SYNC>.
		1164
		1165	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1166
		1167	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1168	scalars.
		1169
		1170	It touches (reads or writes) all memory pages in the specified
		1171	range inside the scalar. All caveats and parameters are the same
		1172	as for C<aio_msync>, above, except for flags, which must be either
		1173	C<0> (which reads all pages and ensures they are instantiated) or
		1174	C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
		1175	writing an octet from it, which dirties the page).
		1176
		1177	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1178
		1179	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1180	scalars.
		1181
		1182	It reads in all the pages of the underlying storage into memory (if any)
		1183	and locks them, so they are not getting swapped/paged out or removed.
		1184
		1185	If C<$length> is undefined, then the scalar will be locked till the end.
		1186
		1187	On systems that do not implement C<mlock>, this function returns C<-1>
		1188	and sets errno to C<ENOSYS>.
		1189
		1190	Note that the corresponding C<munlock> is synchronous and is
		1191	documented under L<MISCELLANEOUS FUNCTIONS>.
		1192
		1193	Example: open a file, mmap and mlock it - both will be undone when
		1194	C<$data> gets destroyed.
		1195
		1196	open my $fh, "<", $path or die "$path: $!";
		1197	my $data;
		1198	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1199	aio_mlock $data; # mlock in background
		1200
		1201	=item aio_mlockall $flags, $callback->($status)
		1202
		1203	Calls the C<mlockall> function with the given C<$flags> (a combination of
		1204	C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
		1205
		1206	On systems that do not implement C<mlockall>, this function returns C<-1>
		1207	and sets errno to C<ENOSYS>.
		1208
		1209	Note that the corresponding C<munlockall> is synchronous and is
		1210	documented under L<MISCELLANEOUS FUNCTIONS>.
		1211
		1212	Example: asynchronously lock all current and future pages into memory.
		1213
		1214	aio_mlockall IO::AIO::MCL_FUTURE;
		1215
		1216	=item aio_group $callback->(...)
		1217
		1218	This is a very special aio request: Instead of doing something, it is a
		1219	container for other aio requests, which is useful if you want to bundle
		1220	many requests into a single, composite, request with a definite callback
		1221	and the ability to cancel the whole request with its subrequests.
		1222
		1223	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		1224	for more info.
		1225
		1226	Example:
		1227
		1228	my $grp = aio_group sub {
		1229	print "all stats done\n";
		1230	};
		1231
		1232	add $grp
		1233	(aio_stat ...),
		1234	(aio_stat ...),
		1235	...;
		1236
		1237	=item aio_nop $callback->()
		1238
		1239	This is a special request - it does nothing in itself and is only used for
		1240	side effects, such as when you want to add a dummy request to a group so
		1241	that finishing the requests in the group depends on executing the given
		1242	code.
		1243
		1244	While this request does nothing, it still goes through the execution
		1245	phase and still requires a worker thread. Thus, the callback will not
		1246	be executed immediately but only after other requests in the queue have
		1247	entered their execution phase. This can be used to measure request
		1248	latency.
		1249
		1250	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
		1251
		1252	Mainly used for debugging and benchmarking, this aio request puts one of
		1253	the request workers to sleep for the given time.
		1254
		1255	While it is theoretically handy to have simple I/O scheduling requests
		1256	like sleep and file handle readable/writable, the overhead this creates is
		1257	immense (it blocks a thread for a long time) so do not use this function
		1258	except to put your application under artificial I/O pressure.
		1259
479	=back	1260	=back
480		1261
		1262	=head2 IO::AIO::REQ CLASS
		1263
		1264	All non-aggregate C<aio_*> functions return an object of this class when
		1265	called in non-void context.
		1266
		1267	=over 4
		1268
		1269	=item cancel $req
		1270
		1271	Cancels the request, if possible. Has the effect of skipping execution
		1272	when entering the B<execute> state and skipping calling the callback when
		1273	entering the the B<result> state, but will leave the request otherwise
		1274	untouched (with the exception of readdir). That means that requests that
		1275	currently execute will not be stopped and resources held by the request
		1276	will not be freed prematurely.
		1277
		1278	=item cb $req $callback->(...)
		1279
		1280	Replace (or simply set) the callback registered to the request.
		1281
		1282	=back
		1283
		1284	=head2 IO::AIO::GRP CLASS
		1285
		1286	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		1287	objects of this class, too.
		1288
		1289	A IO::AIO::GRP object is a special request that can contain multiple other
		1290	aio requests.
		1291
		1292	You create one by calling the C<aio_group> constructing function with a
		1293	callback that will be called when all contained requests have entered the
		1294	C<done> state:
		1295
		1296	my $grp = aio_group sub {
		1297	print "all requests are done\n";
		1298	};
		1299
		1300	You add requests by calling the C<add> method with one or more
		1301	C<IO::AIO::REQ> objects:
		1302
		1303	$grp->add (aio_unlink "...");
		1304
		1305	add $grp aio_stat "...", sub {
		1306	$_[0] or return $grp->result ("error");
		1307
		1308	# add another request dynamically, if first succeeded
		1309	add $grp aio_open "...", sub {
		1310	$grp->result ("ok");
		1311	};
		1312	};
		1313
		1314	This makes it very easy to create composite requests (see the source of
		1315	C<aio_move> for an application) that work and feel like simple requests.
		1316
		1317	=over 4
		1318
		1319	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		1320	C<IO::AIO::poll_cb>, just like any other request.
		1321
		1322	=item * They can be canceled like any other request. Canceling will cancel not
		1323	only the request itself, but also all requests it contains.
		1324
		1325	=item * They can also can also be added to other IO::AIO::GRP objects.
		1326
		1327	=item * You must not add requests to a group from within the group callback (or
		1328	any later time).
		1329
		1330	=back
		1331
		1332	Their lifetime, simplified, looks like this: when they are empty, they
		1333	will finish very quickly. If they contain only requests that are in the
		1334	C<done> state, they will also finish. Otherwise they will continue to
		1335	exist.
		1336
		1337	That means after creating a group you have some time to add requests
		1338	(precisely before the callback has been invoked, which is only done within
		1339	the C<poll_cb>). And in the callbacks of those requests, you can add
		1340	further requests to the group. And only when all those requests have
		1341	finished will the the group itself finish.
		1342
		1343	=over 4
		1344
		1345	=item add $grp ...
		1346
		1347	=item $grp->add (...)
		1348
		1349	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		1350	be added, including other groups, as long as you do not create circular
		1351	dependencies.
		1352
		1353	Returns all its arguments.
		1354
		1355	=item $grp->cancel_subs
		1356
		1357	Cancel all subrequests and clears any feeder, but not the group request
		1358	itself. Useful when you queued a lot of events but got a result early.
		1359
		1360	The group request will finish normally (you cannot add requests to the
		1361	group).
		1362
		1363	=item $grp->result (...)
		1364
		1365	Set the result value(s) that will be passed to the group callback when all
		1366	subrequests have finished and set the groups errno to the current value
		1367	of errno (just like calling C<errno> without an error number). By default,
		1368	no argument will be passed and errno is zero.
		1369
		1370	=item $grp->errno ([$errno])
		1371
		1372	Sets the group errno value to C<$errno>, or the current value of errno
		1373	when the argument is missing.
		1374
		1375	Every aio request has an associated errno value that is restored when
		1376	the callback is invoked. This method lets you change this value from its
		1377	default (0).
		1378
		1379	Calling C<result> will also set errno, so make sure you either set C<$!>
		1380	before the call to C<result>, or call c<errno> after it.
		1381
		1382	=item feed $grp $callback->($grp)
		1383
		1384	Sets a feeder/generator on this group: every group can have an attached
		1385	generator that generates requests if idle. The idea behind this is that,
		1386	although you could just queue as many requests as you want in a group,
		1387	this might starve other requests for a potentially long time. For example,
		1388	C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests,
		1389	delaying any later requests for a long time.
		1390
		1391	To avoid this, and allow incremental generation of requests, you can
		1392	instead a group and set a feeder on it that generates those requests. The
		1393	feed callback will be called whenever there are few enough (see C<limit>,
		1394	below) requests active in the group itself and is expected to queue more
		1395	requests.
		1396
		1397	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1398	not impose any limits).
		1399
		1400	If the feed does not queue more requests when called, it will be
		1401	automatically removed from the group.
		1402
		1403	If the feed limit is C<0> when this method is called, it will be set to
		1404	C<2> automatically.
		1405
		1406	Example:
		1407
		1408	# stat all files in @files, but only ever use four aio requests concurrently:
		1409
		1410	my $grp = aio_group sub { print "finished\n" };
		1411	limit $grp 4;
		1412	feed $grp sub {
		1413	my $file = pop @files
		1414	or return;
		1415
		1416	add $grp aio_stat $file, sub { ... };
		1417	};
		1418
		1419	=item limit $grp $num
		1420
		1421	Sets the feeder limit for the group: The feeder will be called whenever
		1422	the group contains less than this many requests.
		1423
		1424	Setting the limit to C<0> will pause the feeding process.
		1425
		1426	The default value for the limit is C<0>, but note that setting a feeder
		1427	automatically bumps it up to C<2>.
		1428
		1429	=back
		1430
481	=head2 SUPPORT FUNCTIONS	1431	=head2 SUPPORT FUNCTIONS
482		1432
		1433	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1434
483	=over 4	1435	=over 4
484		1436
485	=item $fileno = IO::AIO::poll_fileno	1437	=item $fileno = IO::AIO::poll_fileno
486		1438
487	Return the I<request result pipe file descriptor>. This filehandle must be	1439	Return the I<request result pipe file descriptor>. This filehandle must be
488	polled for reading by some mechanism outside this module (e.g. Event or	1440	polled for reading by some mechanism outside this module (e.g. EV, Glib,
489	select, see below or the SYNOPSIS). If the pipe becomes readable you have	1441	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
490	to call C<poll_cb> to check the results.	1442	you have to call C<poll_cb> to check the results.
491		1443
492	See C<poll_cb> for an example.	1444	See C<poll_cb> for an example.
493		1445
494	=item IO::AIO::poll_cb	1446	=item IO::AIO::poll_cb
495		1447
496	Process all outstanding events on the result pipe. You have to call this	1448	Process some outstanding events on the result pipe. You have to call
497	regularly. Returns the number of events processed. Returns immediately	1449	this regularly. Returns C<0> if all events could be processed (or there
498	when no events are outstanding.	1450	were no events to process), or C<-1> if it returned earlier for whatever
		1451	reason. Returns immediately when no events are outstanding. The amount of
		1452	events processed depends on the settings of C<IO::AIO::max_poll_req> and
		1453	C<IO::AIO::max_poll_time>.
		1454
		1455	If not all requests were processed for whatever reason, the filehandle
		1456	will still be ready when C<poll_cb> returns, so normally you don't have to
		1457	do anything special to have it called later.
		1458
		1459	Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
		1460	ready, it can be beneficial to call this function from loops which submit
		1461	a lot of requests, to make sure the results get processed when they become
		1462	available and not just when the loop is finished and the event loop takes
		1463	over again. This function returns very fast when there are no outstanding
		1464	requests.
499		1465
500	Example: Install an Event watcher that automatically calls	1466	Example: Install an Event watcher that automatically calls
501	IO::AIO::poll_cb with high priority:	1467	IO::AIO::poll_cb with high priority (more examples can be found in the
		1468	SYNOPSIS section, at the top of this document):
502		1469
503	Event->io (fd => IO::AIO::poll_fileno,	1470	Event->io (fd => IO::AIO::poll_fileno,
504	poll => 'r', async => 1,	1471	poll => 'r', async => 1,
505	cb => \&IO::AIO::poll_cb);	1472	cb => \&IO::AIO::poll_cb);
506		1473
507	=item IO::AIO::poll_wait	1474	=item IO::AIO::poll_wait
508		1475
		1476	If there are any outstanding requests and none of them in the result
509	Wait till the result filehandle becomes ready for reading (simply does a	1477	phase, wait till the result filehandle becomes ready for reading (simply
510	C<select> on the filehandle. This is useful if you want to synchronously wait	1478	does a C<select> on the filehandle. This is useful if you want to
511	for some requests to finish).	1479	synchronously wait for some requests to finish).
512		1480
513	See C<nreqs> for an example.	1481	See C<nreqs> for an example.
514		1482
		1483	=item IO::AIO::poll
		1484
		1485	Waits until some requests have been handled.
		1486
		1487	Returns the number of requests processed, but is otherwise strictly
		1488	equivalent to:
		1489
		1490	IO::AIO::poll_wait, IO::AIO::poll_cb
		1491
515	=item IO::AIO::nreqs	1492	=item IO::AIO::flush
516		1493
517	Returns the number of requests currently outstanding (i.e. for which their	1494	Wait till all outstanding AIO requests have been handled.
518	callback has not been invoked yet).
519		1495
520	Example: wait till there are no outstanding requests anymore:	1496	Strictly equivalent to:
521		1497
522	IO::AIO::poll_wait, IO::AIO::poll_cb	1498	IO::AIO::poll_wait, IO::AIO::poll_cb
523	while IO::AIO::nreqs;	1499	while IO::AIO::nreqs;
524		1500
525	=item IO::AIO::flush	1501	=item IO::AIO::max_poll_reqs $nreqs
526		1502
527	Wait till all outstanding AIO requests have been handled.	1503	=item IO::AIO::max_poll_time $seconds
528		1504
529	Strictly equivalent to:	1505	These set the maximum number of requests (default C<0>, meaning infinity)
		1506	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1507	the maximum amount of time (default C<0>, meaning infinity) spent in
		1508	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1509	of time C<poll_cb> is allowed to use).
530		1510
531	IO::AIO::poll_wait, IO::AIO::poll_cb	1511	Setting C<max_poll_time> to a non-zero value creates an overhead of one
532	while IO::AIO::nreqs;	1512	syscall per request processed, which is not normally a problem unless your
		1513	callbacks are really really fast or your OS is really really slow (I am
		1514	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
533		1515
534	=item IO::AIO::poll	1516	Setting these is useful if you want to ensure some level of
		1517	interactiveness when perl is not fast enough to process all requests in
		1518	time.
535		1519
536	Waits until some requests have been handled.	1520	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
537		1521
538	Strictly equivalent to:	1522	Example: Install an Event watcher that automatically calls
		1523	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1524	program get the CPU sometimes even under high AIO load.
539		1525
540	IO::AIO::poll_wait, IO::AIO::poll_cb	1526	# try not to spend much more than 0.1s in poll_cb
541	if IO::AIO::nreqs;	1527	IO::AIO::max_poll_time 0.1;
		1528
		1529	# use a low priority so other tasks have priority
		1530	Event->io (fd => IO::AIO::poll_fileno,
		1531	poll => 'r', nice => 1,
		1532	cb => &IO::AIO::poll_cb);
		1533
		1534	=back
		1535
		1536	=head3 CONTROLLING THE NUMBER OF THREADS
		1537
		1538	=over
542		1539
543	=item IO::AIO::min_parallel $nthreads	1540	=item IO::AIO::min_parallel $nthreads
544		1541
545	Set the minimum number of AIO threads to C<$nthreads>. The current default	1542	Set the minimum number of AIO threads to C<$nthreads>. The current
546	is C<4>, which means four asynchronous operations can be done at one time	1543	default is C<8>, which means eight asynchronous operations can execute
547	(the number of outstanding operations, however, is unlimited).	1544	concurrently at any one time (the number of outstanding requests,
		1545	however, is unlimited).
548		1546
549	IO::AIO starts threads only on demand, when an AIO request is queued and	1547	IO::AIO starts threads only on demand, when an AIO request is queued and
550	no free thread exists.	1548	no free thread exists. Please note that queueing up a hundred requests can
		1549	create demand for a hundred threads, even if it turns out that everything
		1550	is in the cache and could have been processed faster by a single thread.
551		1551
552	It is recommended to keep the number of threads low, as some Linux	1552	It is recommended to keep the number of threads relatively low, as some
553	kernel versions will scale negatively with the number of threads (higher	1553	Linux kernel versions will scale negatively with the number of threads
554	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	1554	(higher parallelity => MUCH higher latency). With current Linux 2.6
555	threads should be fine.	1555	versions, 4-32 threads should be fine.
556		1556
557	Under most circumstances you don't need to call this function, as the	1557	Under most circumstances you don't need to call this function, as the
558	module selects a default that is suitable for low to moderate load.	1558	module selects a default that is suitable for low to moderate load.
559		1559
560	=item IO::AIO::max_parallel $nthreads	1560	=item IO::AIO::max_parallel $nthreads
…		…
569	This module automatically runs C<max_parallel 0> at program end, to ensure	1569	This module automatically runs C<max_parallel 0> at program end, to ensure
570	that all threads are killed and that there are no outstanding requests.	1570	that all threads are killed and that there are no outstanding requests.
571		1571
572	Under normal circumstances you don't need to call this function.	1572	Under normal circumstances you don't need to call this function.
573		1573
		1574	=item IO::AIO::max_idle $nthreads
		1575
		1576	Limit the number of threads (default: 4) that are allowed to idle
		1577	(i.e., threads that did not get a request to process within the idle
		1578	timeout (default: 10 seconds). That means if a thread becomes idle while
		1579	C<$nthreads> other threads are also idle, it will free its resources and
		1580	exit.
		1581
		1582	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1583	to allow for extremely high load situations, but want to free resources
		1584	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1585
		1586	The default is probably ok in most situations, especially if thread
		1587	creation is fast. If thread creation is very slow on your system you might
		1588	want to use larger values.
		1589
		1590	=item IO::AIO::idle_timeout $seconds
		1591
		1592	Sets the minimum idle timeout (default 10) after which worker threads are
		1593	allowed to exit. SEe C<IO::AIO::max_idle>.
		1594
574	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	1595	=item IO::AIO::max_outstanding $maxreqs
575		1596
576	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1597	Sets the maximum number of outstanding requests to C<$nreqs>. If
577	try to queue up more than this number of requests, the caller will block until	1598	you do queue up more than this number of requests, the next call to
578	some requests have been handled.	1599	C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
		1600	C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
		1601	longer exceeded.
579		1602
580	The default is very large, so normally there is no practical limit. If you	1603	In other words, this setting does not enforce a queue limit, but can be
581	queue up many requests in a loop it often improves speed if you set	1604	used to make poll functions block if the limit is exceeded.
582	this to a relatively low number, such as C<100>.
583		1605
584	Under normal circumstances you don't need to call this function.	1606	This is a very bad function to use in interactive programs because it
		1607	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1608	use an C<aio_group> together with a feed callback.
		1609
		1610	It's main use is in scripts without an event loop - when you want to stat
		1611	a lot of files, you can write somehting like this:
		1612
		1613	IO::AIO::max_outstanding 32;
		1614
		1615	for my $path (...) {
		1616	aio_stat $path , ...;
		1617	IO::AIO::poll_cb;
		1618	}
		1619
		1620	IO::AIO::flush;
		1621
		1622	The call to C<poll_cb> inside the loop will normally return instantly, but
		1623	as soon as more thna C<32> reqeusts are in-flight, it will block until
		1624	some requests have been handled. This keeps the loop from pushing a large
		1625	number of C<aio_stat> requests onto the queue.
		1626
		1627	The default value for C<max_outstanding> is very large, so there is no
		1628	practical limit on the number of outstanding requests.
585		1629
586	=back	1630	=back
587		1631
		1632	=head3 STATISTICAL INFORMATION
		1633
		1634	=over
		1635
		1636	=item IO::AIO::nreqs
		1637
		1638	Returns the number of requests currently in the ready, execute or pending
		1639	states (i.e. for which their callback has not been invoked yet).
		1640
		1641	Example: wait till there are no outstanding requests anymore:
		1642
		1643	IO::AIO::poll_wait, IO::AIO::poll_cb
		1644	while IO::AIO::nreqs;
		1645
		1646	=item IO::AIO::nready
		1647
		1648	Returns the number of requests currently in the ready state (not yet
		1649	executed).
		1650
		1651	=item IO::AIO::npending
		1652
		1653	Returns the number of requests currently in the pending state (executed,
		1654	but not yet processed by poll_cb).
		1655
		1656	=back
		1657
		1658	=head3 MISCELLANEOUS FUNCTIONS
		1659
		1660	IO::AIO implements some functions that might be useful, but are not
		1661	asynchronous.
		1662
		1663	=over 4
		1664
		1665	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		1666
		1667	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		1668	but is blocking (this makes most sense if you know the input data is
		1669	likely cached already and the output filehandle is set to non-blocking
		1670	operations).
		1671
		1672	Returns the number of bytes copied, or C<-1> on error.
		1673
		1674	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		1675
		1676	Simply calls the C<posix_fadvise> function (see its
		1677	manpage for details). The following advice constants are
		1678	available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		1679	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		1680	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		1681
		1682	On systems that do not implement C<posix_fadvise>, this function returns
		1683	ENOSYS, otherwise the return value of C<posix_fadvise>.
		1684
		1685	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		1686
		1687	Simply calls the C<posix_madvise> function (see its
		1688	manpage for details). The following advice constants are
		1689	available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		1690	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
		1691
		1692	On systems that do not implement C<posix_madvise>, this function returns
		1693	ENOSYS, otherwise the return value of C<posix_madvise>.
		1694
		1695	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		1696
		1697	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		1698	$scalar (see its manpage for details). The following protect
		1699	constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		1700	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		1701
		1702	On systems that do not implement C<mprotect>, this function returns
		1703	ENOSYS, otherwise the return value of C<mprotect>.
		1704
		1705	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		1706
		1707	Memory-maps a file (or anonymous memory range) and attaches it to the
		1708	given C<$scalar>, which will act like a string scalar.
		1709
		1710	The only operations allowed on the scalar are C<substr>/C<vec> that don't
		1711	change the string length, and most read-only operations such as copying it
		1712	or searching it with regexes and so on.
		1713
		1714	Anything else is unsafe and will, at best, result in memory leaks.
		1715
		1716	The memory map associated with the C<$scalar> is automatically removed
		1717	when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
		1718	C<IO::AIO::munmap> functions are called.
		1719
		1720	This calls the C<mmap>(2) function internally. See your system's manual
		1721	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		1722
		1723	The C<$length> must be larger than zero and smaller than the actual
		1724	filesize.
		1725
		1726	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		1727	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		1728
		1729	C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
		1730	C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
		1731	not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
		1732	(which is set to C<MAP_ANON> if your system only provides this
		1733	constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
		1734	C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
		1735	C<IO::AIO::MAP_NONBLOCK>
		1736
		1737	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		1738
		1739	C<$offset> is the offset from the start of the file - it generally must be
		1740	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		1741
		1742	Example:
		1743
		1744	use Digest::MD5;
		1745	use IO::AIO;
		1746
		1747	open my $fh, "<verybigfile"
		1748	or die "$!";
		1749
		1750	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		1751	or die "verybigfile: $!";
		1752
		1753	my $fast_md5 = md5 $data;
		1754
		1755	=item IO::AIO::munmap $scalar
		1756
		1757	Removes a previous mmap and undefines the C<$scalar>.
		1758
		1759	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		1760
		1761	Calls the C<munlock> function, undoing the effects of a previous
		1762	C<aio_mlock> call (see its description for details).
		1763
		1764	=item IO::AIO::munlockall
		1765
		1766	Calls the C<munlockall> function.
		1767
		1768	On systems that do not implement C<munlockall>, this function returns
		1769	ENOSYS, otherwise the return value of C<munlockall>.
		1770
		1771	=back
		1772
588	=cut	1773	=cut
589		1774
590	# support function to convert a fd into a perl filehandle
591	sub _fd2fh {
592	return undef if $_[0] < 0;
593
594	# try to generate nice filehandles
595	my $sym = "IO::AIO::fd#$_[0]";
596	local *$sym;
597
598	open *$sym, "+<&=$_[0]" # usually works under any unix
599	or open *$sym, "<&=$_[0]" # cygwin needs this
600	or open *$sym, ">&=$_[0]" # or this
601	or return undef;
602
603	*$sym
604	}
605
606	min_parallel 4;	1775	min_parallel 8;
607		1776
608	END {	1777	END { flush }
609	max_parallel 0;
610	}
611		1778
612	1;	1779	1;
613		1780
		1781	=head1 EVENT LOOP INTEGRATION
		1782
		1783	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		1784	automatically into many event loops:
		1785
		1786	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		1787	use AnyEvent::AIO;
		1788
		1789	You can also integrate IO::AIO manually into many event loops, here are
		1790	some examples of how to do this:
		1791
		1792	# EV integration
		1793	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		1794
		1795	# Event integration
		1796	Event->io (fd => IO::AIO::poll_fileno,
		1797	poll => 'r',
		1798	cb => \&IO::AIO::poll_cb);
		1799
		1800	# Glib/Gtk2 integration
		1801	add_watch Glib::IO IO::AIO::poll_fileno,
		1802	in => sub { IO::AIO::poll_cb; 1 };
		1803
		1804	# Tk integration
		1805	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		1806	readable => \&IO::AIO::poll_cb);
		1807
		1808	# Danga::Socket integration
		1809	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		1810	\&IO::AIO::poll_cb);
		1811
614	=head2 FORK BEHAVIOUR	1812	=head2 FORK BEHAVIOUR
615		1813
616	Before the fork, IO::AIO enters a quiescent state where no requests	1814	Usage of pthreads in a program changes the semantics of fork
617	can be added in other threads and no results will be processed. After	1815	considerably. Specifically, only async-safe functions can be called after
618	the fork the parent simply leaves the quiescent state and continues	1816	fork. Perl doesn't know about this, so in general, you cannot call fork
619	request/result processing, while the child clears the request/result	1817	with defined behaviour in perl if pthreads are involved. IO::AIO uses
620	queue (so the requests started before the fork will only be handled in	1818	pthreads, so this applies, but many other extensions and (for inexplicable
621	the parent). Threats will be started on demand until the limit ste in the	1819	reasons) perl itself often is linked against pthreads, so this limitation
622	parent process has been reached again.	1820	applies to quite a lot of perls.
		1821
		1822	This module no longer tries to fight your OS, or POSIX. That means IO::AIO
		1823	only works in the process that loaded it. Forking is fully supported, but
		1824	using IO::AIO in the child is not.
		1825
		1826	You might get around by not I<using> IO::AIO before (or after)
		1827	forking. You could also try to call the L<IO::AIO::reinit> function in the
		1828	child:
		1829
		1830	=over 4
		1831
		1832	=item IO::AIO::reinit
		1833
		1834	Abandons all current requests and I/O threads and simply reinitialises all
		1835	data structures. This is not an operation supported by any standards, but
		1836	happens to work on GNU/Linux and some newer BSD systems.
		1837
		1838	The only reasonable use for this function is to call it after forking, if
		1839	C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
		1840	the process will result in undefined behaviour. Calling it at any time
		1841	will also result in any undefined (by POSIX) behaviour.
		1842
		1843	=back
		1844
		1845	=head2 MEMORY USAGE
		1846
		1847	Per-request usage:
		1848
		1849	Each aio request uses - depending on your architecture - around 100-200
		1850	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1851	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1852	scalars and other data passed into aio requests will also be locked and
		1853	will consume memory till the request has entered the done state.
		1854
		1855	This is not awfully much, so queuing lots of requests is not usually a
		1856	problem.
		1857
		1858	Per-thread usage:
		1859
		1860	In the execution phase, some aio requests require more memory for
		1861	temporary buffers, and each thread requires a stack and other data
		1862	structures (usually around 16k-128k, depending on the OS).
		1863
		1864	=head1 KNOWN BUGS
		1865
		1866	Known bugs will be fixed in the next release.
623		1867
624	=head1 SEE ALSO	1868	=head1 SEE ALSO
625		1869
626	L<Coro>, L<Linux::AIO>.	1870	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		1871	more natural syntax.
627		1872
628	=head1 AUTHOR	1873	=head1 AUTHOR
629		1874
630	Marc Lehmann <schmorp@schmorp.de>	1875	Marc Lehmann <schmorp@schmorp.de>
631	http://home.schmorp.de/	1876	http://home.schmorp.de/

Diff Legend

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