ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.54 by root, Sun Oct 22 00:19:05 2006 UTC vs.
Revision 1.159 by root, Tue Jun 30 07:33:58 2009 UTC

…		…
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", 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	use IO::AIO 2; # version has aio objects	21	# version 2+ has request and group objects
		22	use IO::AIO 2;
21		23
		24	aioreq_pri 4; # give next request a very high priority
22	my $req = aio_unlink "/tmp/file", sub { };	25	my $req = aio_unlink "/tmp/file", sub { };
23	$req->cancel; # cancel request if still in queue	26	$req->cancel; # cancel request if still in queue
24		27
25	# AnyEvent	28	my $grp = aio_group sub { print "all stats done\n" };
26	open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";	29	add $grp aio_stat "..." for ...;
27	my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
28		30
29	# Event	31	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		32	use AnyEvent::AIO;
		33
		34	# EV integration
		35	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		36
		37	# Event integration
30	Event->io (fd => IO::AIO::poll_fileno,	38	Event->io (fd => IO::AIO::poll_fileno,
31	poll => 'r',	39	poll => 'r',
32	cb => \&IO::AIO::poll_cb);	40	cb => \&IO::AIO::poll_cb);
33		41
34	# Glib/Gtk2	42	# Glib/Gtk2 integration
35	add_watch Glib::IO IO::AIO::poll_fileno,	43	add_watch Glib::IO IO::AIO::poll_fileno,
36	in => sub { IO::AIO::poll_cb; 1 };	44	in => sub { IO::AIO::poll_cb; 1 };
37		45
38	# Tk	46	# Tk integration
39	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",	47	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
40	readable => \&IO::AIO::poll_cb);	48	readable => \&IO::AIO::poll_cb);
41		49
42	# Danga::Socket	50	# Danga::Socket integration
43	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>	51	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
44	\&IO::AIO::poll_cb);	52	\&IO::AIO::poll_cb);
45		53
46
47	=head1 DESCRIPTION	54	=head1 DESCRIPTION
48		55
49	This module implements asynchronous I/O using whatever means your	56	This module implements asynchronous I/O using whatever means your
50	operating system supports.	57	operating system supports. It is implemented as an interface to C<libeio>
		58	(L<http://software.schmorp.de/pkg/libeio.html>).
51		59
		60	Asynchronous means that operations that can normally block your program
		61	(e.g. reading from disk) will be done asynchronously: the operation
		62	will still block, but you can do something else in the meantime. This
		63	is extremely useful for programs that need to stay interactive even
		64	when doing heavy I/O (GUI programs, high performance network servers
		65	etc.), but can also be used to easily do operations in parallel that are
		66	normally done sequentially, e.g. stat'ing many files, which is much faster
		67	on a RAID volume or over NFS when you do a number of stat operations
		68	concurrently.
		69
		70	While most of this works on all types of file descriptors (for
		71	example sockets), using these functions on file descriptors that
		72	support nonblocking operation (again, sockets, pipes etc.) is
		73	very inefficient. Use an event loop for that (such as the L<EV>
		74	module): IO::AIO will naturally fit into such an event loop itself.
		75
52	Currently, a number of threads are started that execute your read/writes	76	In this version, a number of threads are started that execute your
53	and signal their completion. You don't need thread support in your libc or	77	requests and signal their completion. You don't need thread support
54	perl, and the threads created by this module will not be visible to the	78	in perl, and the threads created by this module will not be visible
55	pthreads library. In the future, this module might make use of the native	79	to perl. In the future, this module might make use of the native aio
56	aio functions available on many operating systems. However, they are often	80	functions available on many operating systems. However, they are often
57	not well-supported (Linux doesn't allow them on normal files currently,	81	not well-supported or restricted (GNU/Linux doesn't allow them on normal
58	for example), and they would only support aio_read and aio_write, so the	82	files currently, for example), and they would only support aio_read and
59	remaining functionality would have to be implemented using threads anyway.	83	aio_write, so the remaining functionality would have to be implemented
		84	using threads anyway.
60		85
61	Although the module will work with in the presence of other threads, it is	86	Although the module will work in the presence of other (Perl-) threads,
62	currently not reentrant, so use appropriate locking yourself, always call	87	it is currently not reentrant in any way, so use appropriate locking
63	C<poll_cb> from within the same thread, or never call C<poll_cb> (or other	88	yourself, always call C<poll_cb> from within the same thread, or never
64	C<aio_> functions) recursively.	89	call C<poll_cb> (or other C<aio_> functions) recursively.
		90
		91	=head2 EXAMPLE
		92
		93	This is a simple example that uses the EV module and loads
		94	F</etc/passwd> asynchronously:
		95
		96	use Fcntl;
		97	use EV;
		98	use IO::AIO;
		99
		100	# register the IO::AIO callback with EV
		101	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		102
		103	# queue the request to open /etc/passwd
		104	aio_open "/etc/passwd", O_RDONLY, 0, sub {
		105	my $fh = shift
		106	or die "error while opening: $!";
		107
		108	# stat'ing filehandles is generally non-blocking
		109	my $size = -s $fh;
		110
		111	# queue a request to read the file
		112	my $contents;
		113	aio_read $fh, 0, $size, $contents, 0, sub {
		114	$_[0] == $size
		115	or die "short read: $!";
		116
		117	close $fh;
		118
		119	# file contents now in $contents
		120	print $contents;
		121
		122	# exit event loop and program
		123	EV::unloop;
		124	};
		125	};
		126
		127	# possibly queue up other requests, or open GUI windows,
		128	# check for sockets etc. etc.
		129
		130	# process events as long as there are some:
		131	EV::loop;
		132
		133	=head1 REQUEST ANATOMY AND LIFETIME
		134
		135	Every C<aio_*> function creates a request. which is a C data structure not
		136	directly visible to Perl.
		137
		138	If called in non-void context, every request function returns a Perl
		139	object representing the request. In void context, nothing is returned,
		140	which saves a bit of memory.
		141
		142	The perl object is a fairly standard ref-to-hash object. The hash contents
		143	are not used by IO::AIO so you are free to store anything you like in it.
		144
		145	During their existance, aio requests travel through the following states,
		146	in order:
		147
		148	=over 4
		149
		150	=item ready
		151
		152	Immediately after a request is created it is put into the ready state,
		153	waiting for a thread to execute it.
		154
		155	=item execute
		156
		157	A thread has accepted the request for processing and is currently
		158	executing it (e.g. blocking in read).
		159
		160	=item pending
		161
		162	The request has been executed and is waiting for result processing.
		163
		164	While request submission and execution is fully asynchronous, result
		165	processing is not and relies on the perl interpreter calling C<poll_cb>
		166	(or another function with the same effect).
		167
		168	=item result
		169
		170	The request results are processed synchronously by C<poll_cb>.
		171
		172	The C<poll_cb> function will process all outstanding aio requests by
		173	calling their callbacks, freeing memory associated with them and managing
		174	any groups they are contained in.
		175
		176	=item done
		177
		178	Request has reached the end of its lifetime and holds no resources anymore
		179	(except possibly for the Perl object, but its connection to the actual
		180	aio request is severed and calling its methods will either do nothing or
		181	result in a runtime error).
		182
		183	=back
65		184
66	=cut	185	=cut
67		186
68	package IO::AIO;	187	package IO::AIO;
		188
		189	use Carp ();
69		190
70	no warnings;	191	no warnings;
71	use strict 'vars';	192	use strict 'vars';
72		193
73	use base 'Exporter';	194	use base 'Exporter';
74		195
75	BEGIN {	196	BEGIN {
76	our $VERSION = '1.99';	197	our $VERSION = '3.26';
77		198
78	our @EXPORT = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat	199	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close
79	aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink	200	aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
80	aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move	201	aio_scandir aio_symlink aio_readlink aio_sync aio_fsync
81	aio_group);	202	aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead
82	our @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	203	aio_rename aio_link aio_move aio_copy aio_group
		204	aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
		205	aio_chmod aio_utime aio_truncate);
		206
		207	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		208	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		209	min_parallel max_parallel max_idle
		210	nreqs nready npending nthreads
		211	max_poll_time max_poll_reqs
		212	sendfile fadvise);
		213
		214	push @AIO_REQ, qw(aio_busy); # not exported
83		215
84	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';	216	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
85		217
86	require XSLoader;	218	require XSLoader;
87	XSLoader::load ("IO::AIO", $VERSION);	219	XSLoader::load ("IO::AIO", $VERSION);
88	}	220	}
89		221
90	=head1 FUNCTIONS	222	=head1 FUNCTIONS
91		223
92	=head2 AIO FUNCTIONS	224	=head2 AIO REQUEST FUNCTIONS
93		225
94	All the C<aio_*> calls are more or less thin wrappers around the syscall	226	All the C<aio_*> calls are more or less thin wrappers around the syscall
95	with the same name (sans C<aio_>). The arguments are similar or identical,	227	with the same name (sans C<aio_>). The arguments are similar or identical,
96	and they all accept an additional (and optional) C<$callback> argument	228	and they all accept an additional (and optional) C<$callback> argument
97	which must be a code reference. This code reference will get called with	229	which must be a code reference. This code reference will get called with
98	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	230	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
99	perl, which usually delivers "false") as it's sole argument when the given	231	perl, which usually delivers "false") as its sole argument after the given
100	syscall has been executed asynchronously.	232	syscall has been executed asynchronously.
101		233
102	All functions expecting a filehandle keep a copy of the filehandle	234	All functions expecting a filehandle keep a copy of the filehandle
103	internally until the request has finished.	235	internally until the request has finished.
104		236
105	All non-composite requests (requests that are not broken down into
106	multiple requests) return objects of type L<IO::AIO::REQ> that allow	237	All functions return request objects of type L<IO::AIO::REQ> that allow
107	further manipulation of running requests.	238	further manipulation of those requests while they are in-flight.
108		239
109	The pathnames you pass to these routines I<must> be absolute and	240	The pathnames you pass to these routines I<must> be absolute and
110	encoded in byte form. The reason for the former is that at the time the	241	encoded as octets. The reason for the former is that at the time the
111	request is being executed, the current working directory could have	242	request is being executed, the current working directory could have
112	changed. Alternatively, you can make sure that you never change the	243	changed. Alternatively, you can make sure that you never change the
113	current working directory.	244	current working directory anywhere in the program and then use relative
		245	paths.
114		246
115	To encode pathnames to byte form, either make sure you either: a)	247	To encode pathnames as octets, either make sure you either: a) always pass
116	always pass in filenames you got from outside (command line, readdir	248	in filenames you got from outside (command line, readdir etc.) without
117	etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode	249	tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
118	your pathnames to the locale (or other) encoding in effect in the user	250	your pathnames to the locale (or other) encoding in effect in the user
119	environment, d) use Glib::filename_from_unicode on unicode filenames or e)	251	environment, d) use Glib::filename_from_unicode on unicode filenames or e)
120	use something else.	252	use something else to ensure your scalar has the correct contents.
		253
		254	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		255	handles correctly whether it is set or not.
121		256
122	=over 4	257	=over 4
		258
		259	=item $prev_pri = aioreq_pri [$pri]
		260
		261	Returns the priority value that would be used for the next request and, if
		262	C<$pri> is given, sets the priority for the next aio request.
		263
		264	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		265	and C<4>, respectively. Requests with higher priority will be serviced
		266	first.
		267
		268	The priority will be reset to C<0> after each call to one of the C<aio_*>
		269	functions.
		270
		271	Example: open a file with low priority, then read something from it with
		272	higher priority so the read request is serviced before other low priority
		273	open requests (potentially spamming the cache):
		274
		275	aioreq_pri -3;
		276	aio_open ..., sub {
		277	return unless $_[0];
		278
		279	aioreq_pri -2;
		280	aio_read $_[0], ..., sub {
		281	...
		282	};
		283	};
		284
		285
		286	=item aioreq_nice $pri_adjust
		287
		288	Similar to C<aioreq_pri>, but subtracts the given value from the current
		289	priority, so the effect is cumulative.
		290
123		291
124	=item aio_open $pathname, $flags, $mode, $callback->($fh)	292	=item aio_open $pathname, $flags, $mode, $callback->($fh)
125		293
126	Asynchronously open or create a file and call the callback with a newly	294	Asynchronously open or create a file and call the callback with a newly
127	created filehandle for the file.	295	created filehandle for the file.
…		…
133	list. They are the same as used by C<sysopen>.	301	list. They are the same as used by C<sysopen>.
134		302
135	Likewise, C<$mode> specifies the mode of the newly created file, if it	303	Likewise, C<$mode> specifies the mode of the newly created file, if it
136	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,	304	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
137	except that it is mandatory (i.e. use C<0> if you don't create new files,	305	except that it is mandatory (i.e. use C<0> if you don't create new files,
138	and C<0666> or C<0777> if you do).	306	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		307	by the umask in effect then the request is being executed, so better never
		308	change the umask.
139		309
140	Example:	310	Example:
141		311
142	aio_open "/etc/passwd", O_RDONLY, 0, sub {	312	aio_open "/etc/passwd", O_RDONLY, 0, sub {
143	if ($_[0]) {	313	if ($_[0]) {
…		…
146	} else {	316	} else {
147	die "open failed: $!\n";	317	die "open failed: $!\n";
148	}	318	}
149	};	319	};
150		320
		321
151	=item aio_close $fh, $callback->($status)	322	=item aio_close $fh, $callback->($status)
152		323
153	Asynchronously close a file and call the callback with the result	324	Asynchronously close a file and call the callback with the result
154	code. I<WARNING:> although accepted, you should not pass in a perl	325	code.
155	filehandle here, as perl will likely close the file descriptor another
156	time when the filehandle is destroyed. Normally, you can safely call perls
157	C<close> or just let filehandles go out of scope.
158		326
159	This is supposed to be a bug in the API, so that might change. It's	327	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
160	therefore best to avoid this function.	328	closing the file descriptor associated with the filehandle itself.
		329
		330	Therefore, C<aio_close> will not close the filehandle - instead it will
		331	use dup2 to overwrite the file descriptor with the write-end of a pipe
		332	(the pipe fd will be created on demand and will be cached).
		333
		334	Or in other words: the file descriptor will be closed, but it will not be
		335	free for reuse until the perl filehandle is closed.
		336
		337	=cut
161		338
162	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	339	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
163		340
164	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	341	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
165		342
166	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	343	Reads or writes C<$length> bytes from or to the specified C<$fh> and
167	into the scalar given by C<data> and offset C<dataoffset> and calls the	344	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset>
168	callback without the actual number of bytes read (or -1 on error, just	345	and calls the callback without the actual number of bytes read (or -1 on
169	like the syscall).	346	error, just like the syscall).
		347
		348	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		349	offset plus the actual number of bytes read.
		350
		351	If C<$offset> is undefined, then the current file descriptor offset will
		352	be used (and updated), otherwise the file descriptor offset will not be
		353	changed by these calls.
		354
		355	If C<$length> is undefined in C<aio_write>, use the remaining length of
		356	C<$data>.
		357
		358	If C<$dataoffset> is less than zero, it will be counted from the end of
		359	C<$data>.
170		360
171	The C<$data> scalar I<MUST NOT> be modified in any way while the request	361	The C<$data> scalar I<MUST NOT> be modified in any way while the request
172	is outstanding. Modifying it can result in segfaults or WW3 (if the	362	is outstanding. Modifying it can result in segfaults or World War III (if
173	necessary/optional hardware is installed).	363	the necessary/optional hardware is installed).
174		364
175	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	365	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
176	offset C<0> within the scalar:	366	offset C<0> within the scalar:
177		367
178	aio_read $fh, 7, 15, $buffer, 0, sub {	368	aio_read $fh, 7, 15, $buffer, 0, sub {
179	$_[0] > 0 or die "read error: $!";	369	$_[0] > 0 or die "read error: $!";
180	print "read $_[0] bytes: <$buffer>\n";	370	print "read $_[0] bytes: <$buffer>\n";
181	};	371	};
182		372
183	=item aio_move $srcpath, $dstpath, $callback->($status)
184
185	Try to move the I<file> (directories not supported as either source or
186	destination) from C<$srcpath> to C<$dstpath> and call the callback with
187	the C<0> (error) or C<-1> ok.
188
189	This is a composite request that tries to rename(2) the file first. If
190	rename files with C<EXDEV>, it creates the destination file with mode 0200
191	and copies the contents of the source file into it using C<aio_sendfile>,
192	followed by restoring atime, mtime, access mode and uid/gid, in that
193	order, and unlinking the C<$srcpath>.
194
195	If an error occurs, the partial destination file will be unlinked, if
196	possible, except when setting atime, mtime, access mode and uid/gid, where
197	errors are being ignored.
198
199	=cut
200
201	sub aio_move($$$) {
202	my ($src, $dst, $cb) = @_;
203
204	aio_rename $src, $dst, sub {
205	if ($_[0] && $! == EXDEV) {
206	aio_open $src, O_RDONLY, 0, sub {
207	if (my $src_fh = $_[0]) {
208	my @stat = stat $src_fh;
209
210	aio_open $dst, O_WRONLY, 0200, sub {
211	if (my $dst_fh = $_[0]) {
212	aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
213	close $src_fh;
214
215	if ($_[0] == $stat[7]) {
216	utime $stat[8], $stat[9], $dst;
217	chmod $stat[2] & 07777, $dst_fh;
218	chown $stat[4], $stat[5], $dst_fh;
219	close $dst_fh;
220
221	aio_unlink $src, sub {
222	$cb->($_[0]);
223	};
224	} else {
225	my $errno = $!;
226	aio_unlink $dst, sub {
227	$! = $errno;
228	$cb->(-1);
229	};
230	}
231	};
232	} else {
233	$cb->(-1);
234	}
235	},
236
237	} else {
238	$cb->(-1);
239	}
240	};
241	} else {
242	$cb->($_[0]);
243	}
244	};
245	}
246		373
247	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)	374	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
248		375
249	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts	376	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
250	reading at byte offset C<$in_offset>, and starts writing at the current	377	reading at byte offset C<$in_offset>, and starts writing at the current
…		…
264	C<$in_fh> than are written, and there is no way to find out how many	391	C<$in_fh> than are written, and there is no way to find out how many
265	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only	392	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only
266	provides the number of bytes written to C<$out_fh>. Only if the result	393	provides the number of bytes written to C<$out_fh>. Only if the result
267	value equals C<$length> one can assume that C<$length> bytes have been	394	value equals C<$length> one can assume that C<$length> bytes have been
268	read.	395	read.
		396
269		397
270	=item aio_readahead $fh,$offset,$length, $callback->($retval)	398	=item aio_readahead $fh,$offset,$length, $callback->($retval)
271		399
272	C<aio_readahead> populates the page cache with data from a file so that	400	C<aio_readahead> populates the page cache with data from a file so that
273	subsequent reads from that file will not block on disk I/O. The C<$offset>	401	subsequent reads from that file will not block on disk I/O. The C<$offset>
…		…
279	file. The current file offset of the file is left unchanged.	407	file. The current file offset of the file is left unchanged.
280		408
281	If that syscall doesn't exist (likely if your OS isn't Linux) it will be	409	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
282	emulated by simply reading the data, which would have a similar effect.	410	emulated by simply reading the data, which would have a similar effect.
283		411
		412
284	=item aio_stat $fh_or_path, $callback->($status)	413	=item aio_stat $fh_or_path, $callback->($status)
285		414
286	=item aio_lstat $fh, $callback->($status)	415	=item aio_lstat $fh, $callback->($status)
287		416
288	Works like perl's C<stat> or C<lstat> in void context. The callback will	417	Works like perl's C<stat> or C<lstat> in void context. The callback will
…		…
301	aio_stat "/etc/passwd", sub {	430	aio_stat "/etc/passwd", sub {
302	$_[0] and die "stat failed: $!";	431	$_[0] and die "stat failed: $!";
303	print "size is ", -s _, "\n";	432	print "size is ", -s _, "\n";
304	};	433	};
305		434
		435
		436	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		437
		438	Works like perl's C<utime> function (including the special case of $atime
		439	and $mtime being undef). Fractional times are supported if the underlying
		440	syscalls support them.
		441
		442	When called with a pathname, uses utimes(2) if available, otherwise
		443	utime(2). If called on a file descriptor, uses futimes(2) if available,
		444	otherwise returns ENOSYS, so this is not portable.
		445
		446	Examples:
		447
		448	# set atime and mtime to current time (basically touch(1)):
		449	aio_utime "path", undef, undef;
		450	# set atime to current time and mtime to beginning of the epoch:
		451	aio_utime "path", time, undef; # undef==0
		452
		453
		454	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		455
		456	Works like perl's C<chown> function, except that C<undef> for either $uid
		457	or $gid is being interpreted as "do not change" (but -1 can also be used).
		458
		459	Examples:
		460
		461	# same as "chown root path" in the shell:
		462	aio_chown "path", 0, -1;
		463	# same as above:
		464	aio_chown "path", 0, undef;
		465
		466
		467	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		468
		469	Works like truncate(2) or ftruncate(2).
		470
		471
		472	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		473
		474	Works like perl's C<chmod> function.
		475
		476
306	=item aio_unlink $pathname, $callback->($status)	477	=item aio_unlink $pathname, $callback->($status)
307		478
308	Asynchronously unlink (delete) a file and call the callback with the	479	Asynchronously unlink (delete) a file and call the callback with the
309	result code.	480	result code.
310		481
		482
		483	=item aio_mknod $path, $mode, $dev, $callback->($status)
		484
		485	[EXPERIMENTAL]
		486
		487	Asynchronously create a device node (or fifo). See mknod(2).
		488
		489	The only (POSIX-) portable way of calling this function is:
		490
		491	aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		492
		493
311	=item aio_link $srcpath, $dstpath, $callback->($status)	494	=item aio_link $srcpath, $dstpath, $callback->($status)
312		495
313	Asynchronously create a new link to the existing object at C<$srcpath> at	496	Asynchronously create a new link to the existing object at C<$srcpath> at
314	the path C<$dstpath> and call the callback with the result code.	497	the path C<$dstpath> and call the callback with the result code.
315		498
		499
316	=item aio_symlink $srcpath, $dstpath, $callback->($status)	500	=item aio_symlink $srcpath, $dstpath, $callback->($status)
317		501
318	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at	502	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
319	the path C<$dstpath> and call the callback with the result code.	503	the path C<$dstpath> and call the callback with the result code.
320		504
		505
		506	=item aio_readlink $path, $callback->($link)
		507
		508	Asynchronously read the symlink specified by C<$path> and pass it to
		509	the callback. If an error occurs, nothing or undef gets passed to the
		510	callback.
		511
		512
321	=item aio_rename $srcpath, $dstpath, $callback->($status)	513	=item aio_rename $srcpath, $dstpath, $callback->($status)
322		514
323	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as	515	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
324	rename(2) and call the callback with the result code.	516	rename(2) and call the callback with the result code.
325		517
		518
		519	=item aio_mkdir $pathname, $mode, $callback->($status)
		520
		521	Asynchronously mkdir (create) a directory and call the callback with
		522	the result code. C<$mode> will be modified by the umask at the time the
		523	request is executed, so do not change your umask.
		524
		525
326	=item aio_rmdir $pathname, $callback->($status)	526	=item aio_rmdir $pathname, $callback->($status)
327		527
328	Asynchronously rmdir (delete) a directory and call the callback with the	528	Asynchronously rmdir (delete) a directory and call the callback with the
329	result code.	529	result code.
		530
330		531
331	=item aio_readdir $pathname, $callback->($entries)	532	=item aio_readdir $pathname, $callback->($entries)
332		533
333	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire	534	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
334	directory (i.e. opendir + readdir + closedir). The entries will not be	535	directory (i.e. opendir + readdir + closedir). The entries will not be
335	sorted, and will B<NOT> include the C<.> and C<..> entries.	536	sorted, and will B<NOT> include the C<.> and C<..> entries.
336		537
337	The callback a single argument which is either C<undef> or an array-ref	538	The callback is passed a single argument which is either C<undef> or an
338	with the filenames.	539	array-ref with the filenames.
		540
		541
		542	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		543
		544	Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune
		545	behaviour and output format. In case of an error, C<$entries> will be
		546	C<undef>.
		547
		548	The flags are a combination of the following constants, ORed together (the
		549	flags will also be passed to the callback, possibly modified):
		550
		551	=over 4
		552
		553	=item IO::AIO::READDIR_DENTS
		554
		555	When this flag is off, then the callback gets an arrayref with of names
		556	only (as with C<aio_readdir>), otherwise it gets an arrayref with
		557	C<[$name, $type, $inode]> arrayrefs, each describing a single directory
		558	entry in more detail.
		559
		560	C<$name> is the name of the entry.
		561
		562	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		563
		564	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		565	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		566	C<IO::AIO::DT_WHT>.
		567
		568	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
		569	know, you have to run stat yourself. Also, for speed reasons, the C<$type>
		570	scalars are read-only: you can not modify them.
		571
		572	C<$inode> is the inode number (which might not be exact on systems with 64
		573	bit inode numbers and 32 bit perls). This field has unspecified content on
		574	systems that do not deliver the inode information.
		575
		576	=item IO::AIO::READDIR_DIRS_FIRST
		577
		578	When this flag is set, then the names will be returned in an order where
		579	likely directories come first. This is useful when you need to quickly
		580	find directories, or you want to find all directories while avoiding to
		581	stat() each entry.
		582
		583	If the system returns type information in readdir, then this is used
		584	to find directories directly. Otherwise, likely directories are files
		585	beginning with ".", or otherwise files with no dots, of which files with
		586	short names are tried first.
		587
		588	=item IO::AIO::READDIR_STAT_ORDER
		589
		590	When this flag is set, then the names will be returned in an order
		591	suitable for stat()'ing each one. That is, when you plan to stat()
		592	all files in the given directory, then the returned order will likely
		593	be fastest.
		594
		595	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
		596	the likely dirs come first, resulting in a less optimal stat order.
		597
		598	=item IO::AIO::READDIR_FOUND_UNKNOWN
		599
		600	This flag should not be set when calling C<aio_readdirx>. Instead, it
		601	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		602	C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all
		603	C<$type>'s are known, which can be used to speed up some algorithms.
		604
		605	=back
		606
		607
		608	=item aio_load $path, $data, $callback->($status)
		609
		610	This is a composite request that tries to fully load the given file into
		611	memory. Status is the same as with aio_read.
		612
		613	=cut
		614
		615	sub aio_load($$;$) {
		616	my ($path, undef, $cb) = @_;
		617	my $data = \$_[1];
		618
		619	my $pri = aioreq_pri;
		620	my $grp = aio_group $cb;
		621
		622	aioreq_pri $pri;
		623	add $grp aio_open $path, O_RDONLY, 0, sub {
		624	my $fh = shift
		625	or return $grp->result (-1);
		626
		627	aioreq_pri $pri;
		628	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		629	$grp->result ($_[0]);
		630	};
		631	};
		632
		633	$grp
		634	}
		635
		636	=item aio_copy $srcpath, $dstpath, $callback->($status)
		637
		638	Try to copy the I<file> (directories not supported as either source or
		639	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		640	the C<0> (error) or C<-1> ok.
		641
		642	This is a composite request that creates the destination file with
		643	mode 0200 and copies the contents of the source file into it using
		644	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		645	uid/gid, in that order.
		646
		647	If an error occurs, the partial destination file will be unlinked, if
		648	possible, except when setting atime, mtime, access mode and uid/gid, where
		649	errors are being ignored.
		650
		651	=cut
		652
		653	sub aio_copy($$;$) {
		654	my ($src, $dst, $cb) = @_;
		655
		656	my $pri = aioreq_pri;
		657	my $grp = aio_group $cb;
		658
		659	aioreq_pri $pri;
		660	add $grp aio_open $src, O_RDONLY, 0, sub {
		661	if (my $src_fh = $_[0]) {
		662	my @stat = stat $src_fh; # hmm, might bock over nfs?
		663
		664	aioreq_pri $pri;
		665	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		666	if (my $dst_fh = $_[0]) {
		667	aioreq_pri $pri;
		668	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		669	if ($_[0] == $stat[7]) {
		670	$grp->result (0);
		671	close $src_fh;
		672
		673	my $ch = sub {
		674	aioreq_pri $pri;
		675	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		676	aioreq_pri $pri;
		677	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		678	aioreq_pri $pri;
		679	add $grp aio_close $dst_fh;
		680	}
		681	};
		682	};
		683
		684	aioreq_pri $pri;
		685	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		686	if ($_[0] < 0 && $! == ENOSYS) {
		687	aioreq_pri $pri;
		688	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		689	} else {
		690	$ch->();
		691	}
		692	};
		693	} else {
		694	$grp->result (-1);
		695	close $src_fh;
		696	close $dst_fh;
		697
		698	aioreq $pri;
		699	add $grp aio_unlink $dst;
		700	}
		701	};
		702	} else {
		703	$grp->result (-1);
		704	}
		705	},
		706
		707	} else {
		708	$grp->result (-1);
		709	}
		710	};
		711
		712	$grp
		713	}
		714
		715	=item aio_move $srcpath, $dstpath, $callback->($status)
		716
		717	Try to move the I<file> (directories not supported as either source or
		718	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		719	the C<0> (error) or C<-1> ok.
		720
		721	This is a composite request that tries to rename(2) the file first; if
		722	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		723	that is successful, unlinks the C<$srcpath>.
		724
		725	=cut
		726
		727	sub aio_move($$;$) {
		728	my ($src, $dst, $cb) = @_;
		729
		730	my $pri = aioreq_pri;
		731	my $grp = aio_group $cb;
		732
		733	aioreq_pri $pri;
		734	add $grp aio_rename $src, $dst, sub {
		735	if ($_[0] && $! == EXDEV) {
		736	aioreq_pri $pri;
		737	add $grp aio_copy $src, $dst, sub {
		738	$grp->result ($_[0]);
		739
		740	if (!$_[0]) {
		741	aioreq_pri $pri;
		742	add $grp aio_unlink $src;
		743	}
		744	};
		745	} else {
		746	$grp->result ($_[0]);
		747	}
		748	};
		749
		750	$grp
		751	}
339		752
340	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)	753	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
341		754
342	Scans a directory (similar to C<aio_readdir>) but additionally tries to	755	Scans a directory (similar to C<aio_readdir>) but additionally tries to
343	separate the entries of directory C<$path> into two sets of names, ones	756	efficiently separate the entries of directory C<$path> into two sets of
344	you can recurse into (directories or links to them), and ones you cannot	757	names, directories you can recurse into (directories), and ones you cannot
345	recurse into (everything else).	758	recurse into (everything else, including symlinks to directories).
346		759
347	C<aio_scandir> is a composite request that consists of many sub	760	C<aio_scandir> is a composite request that creates of many sub requests_
348	requests. C<$maxreq> specifies the maximum number of outstanding aio	761	C<$maxreq> specifies the maximum number of outstanding aio requests that
349	requests that this function generates. If it is C<< <= 0 >>, then a	762	this function generates. If it is C<< <= 0 >>, then a suitable default
350	suitable default will be chosen (currently 8).	763	will be chosen (currently 4).
351		764
352	On error, the callback is called without arguments, otherwise it receives	765	On error, the callback is called without arguments, otherwise it receives
353	two array-refs with path-relative entry names.	766	two array-refs with path-relative entry names.
354		767
355	Example:	768	Example:
…		…
362		775
363	Implementation notes.	776	Implementation notes.
364		777
365	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.	778	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
366		779
		780	If readdir returns file type information, then this is used directly to
		781	find directories.
		782
367	After reading the directory, the modification time, size etc. of the	783	Otherwise, after reading the directory, the modification time, size etc.
368	directory before and after the readdir is checked, and if they match (and	784	of the directory before and after the readdir is checked, and if they
369	isn't the current time), the link count will be used to decide how many	785	match (and isn't the current time), the link count will be used to decide
370	entries are directories (if >= 2). Otherwise, no knowledge of the number	786	how many entries are directories (if >= 2). Otherwise, no knowledge of the
371	of subdirectories will be assumed.	787	number of subdirectories will be assumed.
372		788
373	Then entries will be sorted into likely directories (everything without	789	Then entries will be sorted into likely directories a non-initial dot
374	a non-initial dot currently) and likely non-directories (everything	790	currently) and likely non-directories (see C<aio_readdirx>). Then every
375	else). Then every entry plus an appended C</.> will be C<stat>'ed,	791	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
376	likely directories first. If that succeeds, it assumes that the entry	792	in order of their inode numbers. If that succeeds, it assumes that the
377	is a directory or a symlink to directory (which will be checked	793	entry is a directory or a symlink to directory (which will be checked
378	seperately). This is often faster than stat'ing the entry itself because	794	seperately). This is often faster than stat'ing the entry itself because
379	filesystems might detect the type of the entry without reading the inode	795	filesystems might detect the type of the entry without reading the inode
380	data (e.g. ext2fs filetype feature).	796	data (e.g. ext2fs filetype feature), even on systems that cannot return
		797	the filetype information on readdir.
381		798
382	If the known number of directories (link count - 2) has been reached, the	799	If the known number of directories (link count - 2) has been reached, the
383	rest of the entries is assumed to be non-directories.	800	rest of the entries is assumed to be non-directories.
384		801
385	This only works with certainty on POSIX (= UNIX) filesystems, which	802	This only works with certainty on POSIX (= UNIX) filesystems, which
…		…
389	as those tend to return 0 or 1 as link counts, which disables the	806	as those tend to return 0 or 1 as link counts, which disables the
390	directory counting heuristic.	807	directory counting heuristic.
391		808
392	=cut	809	=cut
393		810
394	sub aio_scandir($$$) {	811	sub aio_scandir($$;$) {
395	my ($path, $maxreq, $cb) = @_;	812	my ($path, $maxreq, $cb) = @_;
396		813
		814	my $pri = aioreq_pri;
		815
		816	my $grp = aio_group $cb;
		817
397	$maxreq = 8 if $maxreq <= 0;	818	$maxreq = 4 if $maxreq <= 0;
398		819
399	# stat once	820	# stat once
		821	aioreq_pri $pri;
400	aio_stat $path, sub {	822	add $grp aio_stat $path, sub {
401	return $cb->() if $_[0];	823	return $grp->result () if $_[0];
402	my $now = time;	824	my $now = time;
403	my $hash1 = join ":", (stat _)[0,1,3,7,9];	825	my $hash1 = join ":", (stat _)[0,1,3,7,9];
404		826
405	# read the directory entries	827	# read the directory entries
406	aio_readdir $path, sub {	828	aioreq_pri $pri;
		829	add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub {
407	my $entries = shift	830	my $entries = shift
408	or return $cb->();	831	or return $grp->result ();
409		832
410	# stat the dir another time	833	# stat the dir another time
		834	aioreq_pri $pri;
411	aio_stat $path, sub {	835	add $grp aio_stat $path, sub {
412	my $hash2 = join ":", (stat _)[0,1,3,7,9];	836	my $hash2 = join ":", (stat _)[0,1,3,7,9];
413		837
414	my $ndirs;	838	my $ndirs;
415		839
416	# take the slow route if anything looks fishy	840	# take the slow route if anything looks fishy
417	if ($hash1 ne $hash2 or (stat _)[9] == $now) {	841	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
418	$ndirs = -1;	842	$ndirs = -1;
419	} else {	843	} else {
420	# if nlink == 2, we are finished	844	# if nlink == 2, we are finished
421	# on non-posix-fs's, we rely on nlink < 2	845	# for non-posix-fs's, we rely on nlink < 2
422	$ndirs = (stat _)[3] - 2	846	$ndirs = (stat _)[3] - 2
423	or return $cb->([], $entries);	847	or return $grp->result ([], $entries);
424	}	848	}
425		849
426	# sort into likely dirs and likely nondirs
427	# dirs == files without ".", short entries first
428	$entries = [map $_->[0],
429	sort { $b->[1] cmp $a->[1] }
430	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
431	@$entries];
432
433	my (@dirs, @nondirs);	850	my (@dirs, @nondirs);
434		851
435	my ($statcb, $schedcb);	852	my $statgrp = add $grp aio_group sub {
436	my $nreq = 0;	853	$grp->result (\@dirs, \@nondirs);
		854	};
437		855
438	$schedcb = sub {	856	limit $statgrp $maxreq;
439	if (@$entries) {	857	feed $statgrp sub {
440	if ($nreq < $maxreq) {	858	return unless @$entries;
441	my $ent = pop @$entries;	859	my $entry = shift @$entries;
		860
		861	aioreq_pri $pri;
		862	add $statgrp aio_stat "$path/$entry/.", sub {
		863	if ($_[0] < 0) {
		864	push @nondirs, $entry;
		865	} else {
		866	# need to check for real directory
		867	aioreq_pri $pri;
		868	add $statgrp aio_lstat "$path/$entry", sub {
		869	if (-d _) {
		870	push @dirs, $entry;
		871
		872	unless (--$ndirs) {
		873	push @nondirs, @$entries;
		874	feed $statgrp;
		875	}
		876	} else {
		877	push @nondirs, $entry;
		878	}
442	$nreq++;	879	}
443	aio_stat "$path/$ent/.", sub { $statcb->($_[0], $ent) };
444	}	880	}
445	} elsif (!$nreq) {
446	# finished
447	undef $statcb;
448	undef $schedcb;
449	$cb->(\@dirs, \@nondirs) if $cb;
450	undef $cb;
451	}	881	};
452	};	882	};
453	$statcb = sub {
454	my ($status, $entry) = @_;
455
456	if ($status < 0) {
457	$nreq--;
458	push @nondirs, $entry;
459	&$schedcb;
460	} else {
461	# need to check for real directory
462	aio_lstat "$path/$entry", sub {
463	$nreq--;
464
465	if (-d _) {
466	push @dirs, $entry;
467
468	if (!--$ndirs) {
469	push @nondirs, @$entries;
470	$entries = [];
471	}
472	} else {
473	push @nondirs, $entry;
474	}
475
476	&$schedcb;
477	}
478	}
479	};
480
481	&$schedcb while @$entries && $nreq < $maxreq;
482	};	883	};
483	};	884	};
484	};	885	};
		886
		887	$grp
485	}	888	}
		889
		890	=item aio_rmtree $path, $callback->($status)
		891
		892	Delete a directory tree starting (and including) C<$path>, return the
		893	status of the final C<rmdir> only. This is a composite request that
		894	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		895	everything else.
		896
		897	=cut
		898
		899	sub aio_rmtree;
		900	sub aio_rmtree($;$) {
		901	my ($path, $cb) = @_;
		902
		903	my $pri = aioreq_pri;
		904	my $grp = aio_group $cb;
		905
		906	aioreq_pri $pri;
		907	add $grp aio_scandir $path, 0, sub {
		908	my ($dirs, $nondirs) = @_;
		909
		910	my $dirgrp = aio_group sub {
		911	add $grp aio_rmdir $path, sub {
		912	$grp->result ($_[0]);
		913	};
		914	};
		915
		916	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		917	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		918
		919	add $grp $dirgrp;
		920	};
		921
		922	$grp
		923	}
		924
		925	=item aio_sync $callback->($status)
		926
		927	Asynchronously call sync and call the callback when finished.
486		928
487	=item aio_fsync $fh, $callback->($status)	929	=item aio_fsync $fh, $callback->($status)
488		930
489	Asynchronously call fsync on the given filehandle and call the callback	931	Asynchronously call fsync on the given filehandle and call the callback
490	with the fsync result code.	932	with the fsync result code.
…		…
495	callback with the fdatasync result code.	937	callback with the fdatasync result code.
496		938
497	If this call isn't available because your OS lacks it or it couldn't be	939	If this call isn't available because your OS lacks it or it couldn't be
498	detected, it will be emulated by calling C<fsync> instead.	940	detected, it will be emulated by calling C<fsync> instead.
499		941
		942	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		943
		944	Sync the data portion of the file specified by C<$offset> and C<$length>
		945	to disk (but NOT the metadata), by calling the Linux-specific
		946	sync_file_range call. If sync_file_range is not available or it returns
		947	ENOSYS, then fdatasync or fsync is being substituted.
		948
		949	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		950	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		951	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		952	manpage for details.
		953
		954	=item aio_pathsync $path, $callback->($status)
		955
		956	This request tries to open, fsync and close the given path. This is a
		957	composite request intended to sync directories after directory operations
		958	(E.g. rename). This might not work on all operating systems or have any
		959	specific effect, but usually it makes sure that directory changes get
		960	written to disc. It works for anything that can be opened for read-only,
		961	not just directories.
		962
		963	Passes C<0> when everything went ok, and C<-1> on error.
		964
		965	=cut
		966
		967	sub aio_pathsync($;$) {
		968	my ($path, $cb) = @_;
		969
		970	my $pri = aioreq_pri;
		971	my $grp = aio_group $cb;
		972
		973	aioreq_pri $pri;
		974	add $grp aio_open $path, O_RDONLY, 0, sub {
		975	my ($fh) = @_;
		976	if ($fh) {
		977	aioreq_pri $pri;
		978	add $grp aio_fsync $fh, sub {
		979	$grp->result ($_[0]);
		980
		981	aioreq_pri $pri;
		982	add $grp aio_close $fh;
		983	};
		984	} else {
		985	$grp->result (-1);
		986	}
		987	};
		988
		989	$grp
		990	}
		991
500	=item aio_group $callback->()	992	=item aio_group $callback->(...)
501		993
		994	This is a very special aio request: Instead of doing something, it is a
		995	container for other aio requests, which is useful if you want to bundle
		996	many requests into a single, composite, request with a definite callback
		997	and the ability to cancel the whole request with its subrequests.
		998
		999	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		1000	for more info.
		1001
		1002	Example:
		1003
		1004	my $grp = aio_group sub {
		1005	print "all stats done\n";
		1006	};
		1007
		1008	add $grp
		1009	(aio_stat ...),
		1010	(aio_stat ...),
		1011	...;
		1012
		1013	=item aio_nop $callback->()
		1014
		1015	This is a special request - it does nothing in itself and is only used for
		1016	side effects, such as when you want to add a dummy request to a group so
		1017	that finishing the requests in the group depends on executing the given
		1018	code.
		1019
		1020	While this request does nothing, it still goes through the execution
		1021	phase and still requires a worker thread. Thus, the callback will not
		1022	be executed immediately but only after other requests in the queue have
		1023	entered their execution phase. This can be used to measure request
		1024	latency.
		1025
502	=item aio_sleep $fractional_seconds, $callback->() *NOT EXPORTED*	1026	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
503		1027
504	Mainly used for debugging and benchmarking, this aio request puts one of	1028	Mainly used for debugging and benchmarking, this aio request puts one of
505	the request workers to sleep for the given time.	1029	the request workers to sleep for the given time.
506		1030
		1031	While it is theoretically handy to have simple I/O scheduling requests
		1032	like sleep and file handle readable/writable, the overhead this creates is
		1033	immense (it blocks a thread for a long time) so do not use this function
		1034	except to put your application under artificial I/O pressure.
		1035
507	=back	1036	=back
508		1037
509	=head2 IO::AIO::REQ CLASS	1038	=head2 IO::AIO::REQ CLASS
510		1039
511	All non-aggregate C<aio_*> functions return an object of this class when	1040	All non-aggregate C<aio_*> functions return an object of this class when
512	called in non-void context.	1041	called in non-void context.
513		1042
514	A request always moves through the following five states in its lifetime,
515	in order: B<ready> (request has been created, but has not been executed
516	yet), B<execute> (request is currently being executed), B<pending>
517	(request has been executed but callback has not been called yet),
518	B<result> (results are being processed synchronously, includes calling the
519	callback) and B<done> (request has reached the end of its lifetime and
520	holds no resources anymore).
521
522	=over 4	1043	=over 4
523		1044
524	=item $req->cancel	1045	=item cancel $req
525		1046
526	Cancels the request, if possible. Has the effect of skipping execution	1047	Cancels the request, if possible. Has the effect of skipping execution
527	when entering the B<execute> state and skipping calling the callback when	1048	when entering the B<execute> state and skipping calling the callback when
528	entering the the B<result> state, but will leave the request otherwise	1049	entering the the B<result> state, but will leave the request otherwise
529	untouched. That means that requests that currently execute will not be	1050	untouched (with the exception of readdir). That means that requests that
530	stopped and resources held by the request will not be freed prematurely.	1051	currently execute will not be stopped and resources held by the request
		1052	will not be freed prematurely.
		1053
		1054	=item cb $req $callback->(...)
		1055
		1056	Replace (or simply set) the callback registered to the request.
531		1057
532	=back	1058	=back
533		1059
		1060	=head2 IO::AIO::GRP CLASS
		1061
		1062	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		1063	objects of this class, too.
		1064
		1065	A IO::AIO::GRP object is a special request that can contain multiple other
		1066	aio requests.
		1067
		1068	You create one by calling the C<aio_group> constructing function with a
		1069	callback that will be called when all contained requests have entered the
		1070	C<done> state:
		1071
		1072	my $grp = aio_group sub {
		1073	print "all requests are done\n";
		1074	};
		1075
		1076	You add requests by calling the C<add> method with one or more
		1077	C<IO::AIO::REQ> objects:
		1078
		1079	$grp->add (aio_unlink "...");
		1080
		1081	add $grp aio_stat "...", sub {
		1082	$_[0] or return $grp->result ("error");
		1083
		1084	# add another request dynamically, if first succeeded
		1085	add $grp aio_open "...", sub {
		1086	$grp->result ("ok");
		1087	};
		1088	};
		1089
		1090	This makes it very easy to create composite requests (see the source of
		1091	C<aio_move> for an application) that work and feel like simple requests.
		1092
		1093	=over 4
		1094
		1095	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		1096	C<IO::AIO::poll_cb>, just like any other request.
		1097
		1098	=item * They can be canceled like any other request. Canceling will cancel not
		1099	only the request itself, but also all requests it contains.
		1100
		1101	=item * They can also can also be added to other IO::AIO::GRP objects.
		1102
		1103	=item * You must not add requests to a group from within the group callback (or
		1104	any later time).
		1105
		1106	=back
		1107
		1108	Their lifetime, simplified, looks like this: when they are empty, they
		1109	will finish very quickly. If they contain only requests that are in the
		1110	C<done> state, they will also finish. Otherwise they will continue to
		1111	exist.
		1112
		1113	That means after creating a group you have some time to add requests
		1114	(precisely before the callback has been invoked, which is only done within
		1115	the C<poll_cb>). And in the callbacks of those requests, you can add
		1116	further requests to the group. And only when all those requests have
		1117	finished will the the group itself finish.
		1118
		1119	=over 4
		1120
		1121	=item add $grp ...
		1122
		1123	=item $grp->add (...)
		1124
		1125	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		1126	be added, including other groups, as long as you do not create circular
		1127	dependencies.
		1128
		1129	Returns all its arguments.
		1130
		1131	=item $grp->cancel_subs
		1132
		1133	Cancel all subrequests and clears any feeder, but not the group request
		1134	itself. Useful when you queued a lot of events but got a result early.
		1135
		1136	=item $grp->result (...)
		1137
		1138	Set the result value(s) that will be passed to the group callback when all
		1139	subrequests have finished and set the groups errno to the current value
		1140	of errno (just like calling C<errno> without an error number). By default,
		1141	no argument will be passed and errno is zero.
		1142
		1143	=item $grp->errno ([$errno])
		1144
		1145	Sets the group errno value to C<$errno>, or the current value of errno
		1146	when the argument is missing.
		1147
		1148	Every aio request has an associated errno value that is restored when
		1149	the callback is invoked. This method lets you change this value from its
		1150	default (0).
		1151
		1152	Calling C<result> will also set errno, so make sure you either set C<$!>
		1153	before the call to C<result>, or call c<errno> after it.
		1154
		1155	=item feed $grp $callback->($grp)
		1156
		1157	Sets a feeder/generator on this group: every group can have an attached
		1158	generator that generates requests if idle. The idea behind this is that,
		1159	although you could just queue as many requests as you want in a group,
		1160	this might starve other requests for a potentially long time. For example,
		1161	C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests,
		1162	delaying any later requests for a long time.
		1163
		1164	To avoid this, and allow incremental generation of requests, you can
		1165	instead a group and set a feeder on it that generates those requests. The
		1166	feed callback will be called whenever there are few enough (see C<limit>,
		1167	below) requests active in the group itself and is expected to queue more
		1168	requests.
		1169
		1170	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1171	not impose any limits).
		1172
		1173	If the feed does not queue more requests when called, it will be
		1174	automatically removed from the group.
		1175
		1176	If the feed limit is C<0> when this method is called, it will be set to
		1177	C<2> automatically.
		1178
		1179	Example:
		1180
		1181	# stat all files in @files, but only ever use four aio requests concurrently:
		1182
		1183	my $grp = aio_group sub { print "finished\n" };
		1184	limit $grp 4;
		1185	feed $grp sub {
		1186	my $file = pop @files
		1187	or return;
		1188
		1189	add $grp aio_stat $file, sub { ... };
		1190	};
		1191
		1192	=item limit $grp $num
		1193
		1194	Sets the feeder limit for the group: The feeder will be called whenever
		1195	the group contains less than this many requests.
		1196
		1197	Setting the limit to C<0> will pause the feeding process.
		1198
		1199	The default value for the limit is C<0>, but note that setting a feeder
		1200	automatically bumps it up to C<2>.
		1201
		1202	=back
		1203
534	=head2 SUPPORT FUNCTIONS	1204	=head2 SUPPORT FUNCTIONS
535		1205
		1206	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1207
536	=over 4	1208	=over 4
537		1209
538	=item $fileno = IO::AIO::poll_fileno	1210	=item $fileno = IO::AIO::poll_fileno
539		1211
540	Return the I<request result pipe file descriptor>. This filehandle must be	1212	Return the I<request result pipe file descriptor>. This filehandle must be
541	polled for reading by some mechanism outside this module (e.g. Event or	1213	polled for reading by some mechanism outside this module (e.g. EV, Glib,
542	select, see below or the SYNOPSIS). If the pipe becomes readable you have	1214	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
543	to call C<poll_cb> to check the results.	1215	you have to call C<poll_cb> to check the results.
544		1216
545	See C<poll_cb> for an example.	1217	See C<poll_cb> for an example.
546		1218
547	=item IO::AIO::poll_cb	1219	=item IO::AIO::poll_cb
548		1220
549	Process all outstanding events on the result pipe. You have to call this	1221	Process some outstanding events on the result pipe. You have to call this
550	regularly. Returns the number of events processed. Returns immediately	1222	regularly. Returns C<0> if all events could be processed, or C<-1> if it
551	when no events are outstanding.	1223	returned earlier for whatever reason. Returns immediately when no events
		1224	are outstanding. The amount of events processed depends on the settings of
		1225	C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
		1226
		1227	If not all requests were processed for whatever reason, the filehandle
		1228	will still be ready when C<poll_cb> returns, so normally you don't have to
		1229	do anything special to have it called later.
552		1230
553	Example: Install an Event watcher that automatically calls	1231	Example: Install an Event watcher that automatically calls
554	IO::AIO::poll_cb with high priority:	1232	IO::AIO::poll_cb with high priority (more examples can be found in the
		1233	SYNOPSIS section, at the top of this document):
555		1234
556	Event->io (fd => IO::AIO::poll_fileno,	1235	Event->io (fd => IO::AIO::poll_fileno,
557	poll => 'r', async => 1,	1236	poll => 'r', async => 1,
558	cb => \&IO::AIO::poll_cb);	1237	cb => \&IO::AIO::poll_cb);
559		1238
		1239	=item IO::AIO::max_poll_reqs $nreqs
		1240
		1241	=item IO::AIO::max_poll_time $seconds
		1242
		1243	These set the maximum number of requests (default C<0>, meaning infinity)
		1244	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1245	the maximum amount of time (default C<0>, meaning infinity) spent in
		1246	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1247	of time C<poll_cb> is allowed to use).
		1248
		1249	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1250	syscall per request processed, which is not normally a problem unless your
		1251	callbacks are really really fast or your OS is really really slow (I am
		1252	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1253
		1254	Setting these is useful if you want to ensure some level of
		1255	interactiveness when perl is not fast enough to process all requests in
		1256	time.
		1257
		1258	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		1259
		1260	Example: Install an Event watcher that automatically calls
		1261	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1262	program get the CPU sometimes even under high AIO load.
		1263
		1264	# try not to spend much more than 0.1s in poll_cb
		1265	IO::AIO::max_poll_time 0.1;
		1266
		1267	# use a low priority so other tasks have priority
		1268	Event->io (fd => IO::AIO::poll_fileno,
		1269	poll => 'r', nice => 1,
		1270	cb => &IO::AIO::poll_cb);
		1271
560	=item IO::AIO::poll_wait	1272	=item IO::AIO::poll_wait
561		1273
		1274	If there are any outstanding requests and none of them in the result
562	Wait till the result filehandle becomes ready for reading (simply does a	1275	phase, wait till the result filehandle becomes ready for reading (simply
563	C<select> on the filehandle. This is useful if you want to synchronously wait	1276	does a C<select> on the filehandle. This is useful if you want to
564	for some requests to finish).	1277	synchronously wait for some requests to finish).
565		1278
566	See C<nreqs> for an example.	1279	See C<nreqs> for an example.
567		1280
		1281	=item IO::AIO::poll
		1282
		1283	Waits until some requests have been handled.
		1284
		1285	Returns the number of requests processed, but is otherwise strictly
		1286	equivalent to:
		1287
		1288	IO::AIO::poll_wait, IO::AIO::poll_cb
		1289
568	=item IO::AIO::nreqs	1290	=item IO::AIO::flush
569		1291
570	Returns the number of requests currently outstanding (i.e. for which their	1292	Wait till all outstanding AIO requests have been handled.
571	callback has not been invoked yet).
572		1293
573	Example: wait till there are no outstanding requests anymore:	1294	Strictly equivalent to:
574		1295
575	IO::AIO::poll_wait, IO::AIO::poll_cb	1296	IO::AIO::poll_wait, IO::AIO::poll_cb
576	while IO::AIO::nreqs;	1297	while IO::AIO::nreqs;
577		1298
578	=item IO::AIO::flush	1299	=back
579		1300
580	Wait till all outstanding AIO requests have been handled.	1301	=head3 CONTROLLING THE NUMBER OF THREADS
581		1302
582	Strictly equivalent to:	1303	=over
583
584	IO::AIO::poll_wait, IO::AIO::poll_cb
585	while IO::AIO::nreqs;
586
587	=item IO::AIO::poll
588
589	Waits until some requests have been handled.
590
591	Strictly equivalent to:
592
593	IO::AIO::poll_wait, IO::AIO::poll_cb
594	if IO::AIO::nreqs;
595		1304
596	=item IO::AIO::min_parallel $nthreads	1305	=item IO::AIO::min_parallel $nthreads
597		1306
598	Set the minimum number of AIO threads to C<$nthreads>. The current default	1307	Set the minimum number of AIO threads to C<$nthreads>. The current
599	is C<4>, which means four asynchronous operations can be done at one time	1308	default is C<8>, which means eight asynchronous operations can execute
600	(the number of outstanding operations, however, is unlimited).	1309	concurrently at any one time (the number of outstanding requests,
		1310	however, is unlimited).
601		1311
602	IO::AIO starts threads only on demand, when an AIO request is queued and	1312	IO::AIO starts threads only on demand, when an AIO request is queued and
603	no free thread exists.	1313	no free thread exists. Please note that queueing up a hundred requests can
		1314	create demand for a hundred threads, even if it turns out that everything
		1315	is in the cache and could have been processed faster by a single thread.
604		1316
605	It is recommended to keep the number of threads low, as some Linux	1317	It is recommended to keep the number of threads relatively low, as some
606	kernel versions will scale negatively with the number of threads (higher	1318	Linux kernel versions will scale negatively with the number of threads
607	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	1319	(higher parallelity => MUCH higher latency). With current Linux 2.6
608	threads should be fine.	1320	versions, 4-32 threads should be fine.
609		1321
610	Under most circumstances you don't need to call this function, as the	1322	Under most circumstances you don't need to call this function, as the
611	module selects a default that is suitable for low to moderate load.	1323	module selects a default that is suitable for low to moderate load.
612		1324
613	=item IO::AIO::max_parallel $nthreads	1325	=item IO::AIO::max_parallel $nthreads
…		…
622	This module automatically runs C<max_parallel 0> at program end, to ensure	1334	This module automatically runs C<max_parallel 0> at program end, to ensure
623	that all threads are killed and that there are no outstanding requests.	1335	that all threads are killed and that there are no outstanding requests.
624		1336
625	Under normal circumstances you don't need to call this function.	1337	Under normal circumstances you don't need to call this function.
626		1338
		1339	=item IO::AIO::max_idle $nthreads
		1340
		1341	Limit the number of threads (default: 4) that are allowed to idle (i.e.,
		1342	threads that did not get a request to process within 10 seconds). That
		1343	means if a thread becomes idle while C<$nthreads> other threads are also
		1344	idle, it will free its resources and exit.
		1345
		1346	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1347	to allow for extremely high load situations, but want to free resources
		1348	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1349
		1350	The default is probably ok in most situations, especially if thread
		1351	creation is fast. If thread creation is very slow on your system you might
		1352	want to use larger values.
		1353
627	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	1354	=item IO::AIO::max_outstanding $maxreqs
		1355
		1356	This is a very bad function to use in interactive programs because it
		1357	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1358	use an C<aio_group> together with a feed callback.
628		1359
629	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1360	Sets the maximum number of outstanding requests to C<$nreqs>. If you
630	try to queue up more than this number of requests, the caller will block until	1361	do queue up more than this number of requests, the next call to the
631	some requests have been handled.	1362	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
		1363	function will block until the limit is no longer exceeded.
632		1364
633	The default is very large, so normally there is no practical limit. If you	1365	The default value is very large, so there is no practical limit on the
634	queue up many requests in a loop it often improves speed if you set	1366	number of outstanding requests.
635	this to a relatively low number, such as C<100>.
636		1367
637	Under normal circumstances you don't need to call this function.	1368	You can still queue as many requests as you want. Therefore,
		1369	C<max_outstanding> is mainly useful in simple scripts (with low values) or
		1370	as a stop gap to shield against fatal memory overflow (with large values).
638		1371
639	=back	1372	=back
640		1373
		1374	=head3 STATISTICAL INFORMATION
		1375
		1376	=over
		1377
		1378	=item IO::AIO::nreqs
		1379
		1380	Returns the number of requests currently in the ready, execute or pending
		1381	states (i.e. for which their callback has not been invoked yet).
		1382
		1383	Example: wait till there are no outstanding requests anymore:
		1384
		1385	IO::AIO::poll_wait, IO::AIO::poll_cb
		1386	while IO::AIO::nreqs;
		1387
		1388	=item IO::AIO::nready
		1389
		1390	Returns the number of requests currently in the ready state (not yet
		1391	executed).
		1392
		1393	=item IO::AIO::npending
		1394
		1395	Returns the number of requests currently in the pending state (executed,
		1396	but not yet processed by poll_cb).
		1397
		1398	=back
		1399
		1400	=head3 MISCELLANEOUS FUNCTIONS
		1401
		1402	IO::AIO implements some functions that might be useful, but are not
		1403	asynchronous.
		1404
		1405	=over 4
		1406
		1407	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		1408
		1409	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		1410	but is blocking (this makes most sense if you know the input data is
		1411	likely cached already and the output filehandle is set to non-blocking
		1412	operations).
		1413
		1414	Returns the number of bytes copied, or C<-1> on error.
		1415
		1416	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		1417
		1418	Simply calls the C<posix_fadvise> function (see it's
		1419	manpage for details). The following advice constants are
		1420	avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		1421	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		1422	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		1423
		1424	On systems that do not implement C<posix_fadvise>, this function returns
		1425	ENOSYS, otherwise the return value of C<posix_fadvise>.
		1426
		1427	=back
		1428
641	=cut	1429	=cut
642		1430
643	# support function to convert a fd into a perl filehandle
644	sub _fd2fh {
645	return undef if $_[0] < 0;
646
647	# try to generate nice filehandles
648	my $sym = "IO::AIO::fd#$_[0]";
649	local *$sym;
650
651	open *$sym, "+<&=$_[0]" # usually works under any unix
652	or open *$sym, "<&=$_[0]" # cygwin needs this
653	or open *$sym, ">&=$_[0]" # or this
654	or return undef;
655
656	*$sym
657	}
658
659	min_parallel 4;	1431	min_parallel 8;
660		1432
661	END {	1433	END { flush }
662	max_parallel 0;
663	}
664		1434
665	1;	1435	1;
666		1436
667	=head2 FORK BEHAVIOUR	1437	=head2 FORK BEHAVIOUR
668		1438
669	This module should do "the right thing" when the process using it forks:	1439	This module should do "the right thing" when the process using it forks:
670		1440
671	Before the fork, IO::AIO enters a quiescent state where no requests	1441	Before the fork, IO::AIO enters a quiescent state where no requests
672	can be added in other threads and no results will be processed. After	1442	can be added in other threads and no results will be processed. After
673	the fork the parent simply leaves the quiescent state and continues	1443	the fork the parent simply leaves the quiescent state and continues
674	request/result processing, while the child clears the request/result	1444	request/result processing, while the child frees the request/result queue
675	queue (so the requests started before the fork will only be handled in	1445	(so that the requests started before the fork will only be handled in the
676	the parent). Threads will be started on demand until the limit ste in the	1446	parent). Threads will be started on demand until the limit set in the
677	parent process has been reached again.	1447	parent process has been reached again.
678		1448
679	In short: the parent will, after a short pause, continue as if fork had	1449	In short: the parent will, after a short pause, continue as if fork had
680	not been called, while the child will act as if IO::AIO has not been used	1450	not been called, while the child will act as if IO::AIO has not been used
681	yet.	1451	yet.
682		1452
		1453	=head2 MEMORY USAGE
		1454
		1455	Per-request usage:
		1456
		1457	Each aio request uses - depending on your architecture - around 100-200
		1458	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1459	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1460	scalars and other data passed into aio requests will also be locked and
		1461	will consume memory till the request has entered the done state.
		1462
		1463	This is not awfully much, so queuing lots of requests is not usually a
		1464	problem.
		1465
		1466	Per-thread usage:
		1467
		1468	In the execution phase, some aio requests require more memory for
		1469	temporary buffers, and each thread requires a stack and other data
		1470	structures (usually around 16k-128k, depending on the OS).
		1471
		1472	=head1 KNOWN BUGS
		1473
		1474	Known bugs will be fixed in the next release.
		1475
683	=head1 SEE ALSO	1476	=head1 SEE ALSO
684		1477
685	L<Coro>, L<Linux::AIO> (obsolete).	1478	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		1479	more natural syntax.
686		1480
687	=head1 AUTHOR	1481	=head1 AUTHOR
688		1482
689	Marc Lehmann <schmorp@schmorp.de>	1483	Marc Lehmann <schmorp@schmorp.de>
690	http://home.schmorp.de/	1484	http://home.schmorp.de/

Diff Legend

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