ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.56 by root, Sun Oct 22 00:53:47 2006 UTC vs.
Revision 1.185 by root, Sat Dec 11 19:06:07 2010 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
…		…
18	};	19	};
19		20
20	# version 2+ has request and group objects	21	# version 2+ has request and group objects
21	use IO::AIO 2;	22	use IO::AIO 2;
22		23
		24	aioreq_pri 4; # give next request a very high priority
23	my $req = aio_unlink "/tmp/file", sub { };	25	my $req = aio_unlink "/tmp/file", sub { };
24	$req->cancel; # cancel request if still in queue	26	$req->cancel; # cancel request if still in queue
25		27
26	my $grp = aio_group sub { print "all stats done\n" };	28	my $grp = aio_group sub { print "all stats done\n" };
27	add $grp aio_stat "..." for ...;	29	add $grp aio_stat "..." for ...;
28		30
29	# AnyEvent integration
30	open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
31	my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
32
33	# Event integration
34	Event->io (fd => IO::AIO::poll_fileno,
35	poll => 'r',
36	cb => \&IO::AIO::poll_cb);
37
38	# Glib/Gtk2 integration
39	add_watch Glib::IO IO::AIO::poll_fileno,
40	in => sub { IO::AIO::poll_cb; 1 };
41
42	# Tk integration
43	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
44	readable => \&IO::AIO::poll_cb);
45
46	# Danga::Socket integration
47	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
48	\&IO::AIO::poll_cb);
49
50	=head1 DESCRIPTION	31	=head1 DESCRIPTION
51		32
52	This module implements asynchronous I/O using whatever means your	33	This module implements asynchronous I/O using whatever means your
53	operating system supports.	34	operating system supports. It is implemented as an interface to C<libeio>
		35	(L<http://software.schmorp.de/pkg/libeio.html>).
54		36
		37	Asynchronous means that operations that can normally block your program
		38	(e.g. reading from disk) will be done asynchronously: the operation
		39	will still block, but you can do something else in the meantime. This
		40	is extremely useful for programs that need to stay interactive even
		41	when doing heavy I/O (GUI programs, high performance network servers
		42	etc.), but can also be used to easily do operations in parallel that are
		43	normally done sequentially, e.g. stat'ing many files, which is much faster
		44	on a RAID volume or over NFS when you do a number of stat operations
		45	concurrently.
		46
		47	While most of this works on all types of file descriptors (for
		48	example sockets), using these functions on file descriptors that
		49	support nonblocking operation (again, sockets, pipes etc.) is
		50	very inefficient. Use an event loop for that (such as the L<EV>
		51	module): IO::AIO will naturally fit into such an event loop itself.
		52
55	Currently, a number of threads are started that execute your read/writes	53	In this version, a number of threads are started that execute your
56	and signal their completion. You don't need thread support in your libc or	54	requests and signal their completion. You don't need thread support
57	perl, and the threads created by this module will not be visible to the	55	in perl, and the threads created by this module will not be visible
58	pthreads library. In the future, this module might make use of the native	56	to perl. In the future, this module might make use of the native aio
59	aio functions available on many operating systems. However, they are often	57	functions available on many operating systems. However, they are often
60	not well-supported (Linux doesn't allow them on normal files currently,	58	not well-supported or restricted (GNU/Linux doesn't allow them on normal
61	for example), and they would only support aio_read and aio_write, so the	59	files currently, for example), and they would only support aio_read and
62	remaining functionality would have to be implemented using threads anyway.	60	aio_write, so the remaining functionality would have to be implemented
		61	using threads anyway.
63		62
64	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,
65	currently not reentrant, so use appropriate locking yourself, always call	64	it is currently not reentrant in any way, so use appropriate locking
66	C<poll_cb> from within the same thread, or never call C<poll_cb> (or other	65	yourself, always call C<poll_cb> from within the same thread, or never
67	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
68		161
69	=cut	162	=cut
70		163
71	package IO::AIO;	164	package IO::AIO;
72		165
73	no warnings;	166	use Carp ();
74	use strict 'vars';	167
		168	use common::sense;
75		169
76	use base 'Exporter';	170	use base 'Exporter';
77		171
78	BEGIN {	172	BEGIN {
79	our $VERSION = '2.0';	173	our $VERSION = '3.7';
80		174
81	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
82	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
83	aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move	177	aio_scandir aio_symlink aio_readlink aio_sync aio_fsync
		178	aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead
		179	aio_rename aio_link aio_move aio_copy aio_group
		180	aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
		181	aio_chmod aio_utime aio_truncate
		182	aio_msync aio_mtouch aio_mlock aio_mlockall
84	aio_group);	183	aio_statvfs);
85	our @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	184
		185	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		186	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		187	min_parallel max_parallel max_idle
		188	nreqs nready npending nthreads
		189	max_poll_time max_poll_reqs
		190	sendfile fadvise madvise
		191	mmap munmap munlock munlockall);
		192
		193	push @AIO_REQ, qw(aio_busy); # not exported
86		194
87	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';	195	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
88		196
89	require XSLoader;	197	require XSLoader;
90	XSLoader::load ("IO::AIO", $VERSION);	198	XSLoader::load ("IO::AIO", $VERSION);
91	}	199	}
92		200
93	=head1 FUNCTIONS	201	=head1 FUNCTIONS
94		202
		203	=head2 QUICK OVERVIEW
		204
		205	This section simply lists the prototypes of the most important functions
		206	for quick reference. See the following sections for function-by-function
		207	documentation.
		208
		209	aio_open $pathname, $flags, $mode, $callback->($fh)
		210	aio_close $fh, $callback->($status)
		211	aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		212	aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		213	aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		214	aio_readahead $fh,$offset,$length, $callback->($retval)
		215	aio_stat $fh_or_path, $callback->($status)
		216	aio_lstat $fh, $callback->($status)
		217	aio_statvfs $fh_or_path, $callback->($statvfs)
		218	aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		219	aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		220	aio_truncate $fh_or_path, $offset, $callback->($status)
		221	aio_chmod $fh_or_path, $mode, $callback->($status)
		222	aio_unlink $pathname, $callback->($status)
		223	aio_mknod $path, $mode, $dev, $callback->($status)
		224	aio_link $srcpath, $dstpath, $callback->($status)
		225	aio_symlink $srcpath, $dstpath, $callback->($status)
		226	aio_readlink $path, $callback->($link)
		227	aio_rename $srcpath, $dstpath, $callback->($status)
		228	aio_mkdir $pathname, $mode, $callback->($status)
		229	aio_rmdir $pathname, $callback->($status)
		230	aio_readdir $pathname, $callback->($entries)
		231	aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		232	IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
		233	IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
		234	aio_load $path, $data, $callback->($status)
		235	aio_copy $srcpath, $dstpath, $callback->($status)
		236	aio_move $srcpath, $dstpath, $callback->($status)
		237	aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		238	aio_rmtree $path, $callback->($status)
		239	aio_sync $callback->($status)
		240	aio_fsync $fh, $callback->($status)
		241	aio_fdatasync $fh, $callback->($status)
		242	aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		243	aio_pathsync $path, $callback->($status)
		244	aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		245	aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		246	aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		247	aio_mlockall $flags, $callback->($status)
		248	aio_group $callback->(...)
		249	aio_nop $callback->()
		250
		251	$prev_pri = aioreq_pri [$pri]
		252	aioreq_nice $pri_adjust
		253
		254	IO::AIO::poll_wait
		255	IO::AIO::poll_cb
		256	IO::AIO::poll
		257	IO::AIO::flush
		258	IO::AIO::max_poll_reqs $nreqs
		259	IO::AIO::max_poll_time $seconds
		260	IO::AIO::min_parallel $nthreads
		261	IO::AIO::max_parallel $nthreads
		262	IO::AIO::max_idle $nthreads
		263	IO::AIO::max_outstanding $maxreqs
		264	IO::AIO::nreqs
		265	IO::AIO::nready
		266	IO::AIO::npending
		267
		268	IO::AIO::sendfile $ofh, $ifh, $offset, $count
		269	IO::AIO::fadvise $fh, $offset, $len, $advice
		270	IO::AIO::madvise $scalar, $offset, $length, $advice
		271	IO::AIO::mprotect $scalar, $offset, $length, $protect
		272	IO::AIO::munlock $scalar, $offset = 0, $length = undef
		273	IO::AIO::munlockall
		274
95	=head2 AIO FUNCTIONS	275	=head2 AIO REQUEST FUNCTIONS
96		276
97	All the C<aio_*> calls are more or less thin wrappers around the syscall	277	All the C<aio_*> calls are more or less thin wrappers around the syscall
98	with the same name (sans C<aio_>). The arguments are similar or identical,	278	with the same name (sans C<aio_>). The arguments are similar or identical,
99	and they all accept an additional (and optional) C<$callback> argument	279	and they all accept an additional (and optional) C<$callback> argument
100	which must be a code reference. This code reference will get called with	280	which must be a code reference. This code reference will get called with
101	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	281	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
102	perl, which usually delivers "false") as it's sole argument when the given	282	perl, which usually delivers "false") as its sole argument after the given
103	syscall has been executed asynchronously.	283	syscall has been executed asynchronously.
104		284
105	All functions expecting a filehandle keep a copy of the filehandle	285	All functions expecting a filehandle keep a copy of the filehandle
106	internally until the request has finished.	286	internally until the request has finished.
107		287
108	All requests return objects of type L<IO::AIO::REQ> that allow further	288	All functions return request objects of type L<IO::AIO::REQ> that allow
109	manipulation of those requests while they are in-flight.	289	further manipulation of those requests while they are in-flight.
110		290
111	The pathnames you pass to these routines I<must> be absolute and	291	The pathnames you pass to these routines I<must> be absolute and
112	encoded in byte form. The reason for the former is that at the time the	292	encoded as octets. The reason for the former is that at the time the
113	request is being executed, the current working directory could have	293	request is being executed, the current working directory could have
114	changed. Alternatively, you can make sure that you never change the	294	changed. Alternatively, you can make sure that you never change the
115	current working directory.	295	current working directory anywhere in the program and then use relative
		296	paths.
116		297
117	To encode pathnames to byte form, either make sure you either: a)	298	To encode pathnames as octets, either make sure you either: a) always pass
118	always pass in filenames you got from outside (command line, readdir	299	in filenames you got from outside (command line, readdir etc.) without
119	etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode	300	tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
120	your pathnames to the locale (or other) encoding in effect in the user	301	your pathnames to the locale (or other) encoding in effect in the user
121	environment, d) use Glib::filename_from_unicode on unicode filenames or e)	302	environment, d) use Glib::filename_from_unicode on unicode filenames or e)
122	use something else.	303	use something else to ensure your scalar has the correct contents.
		304
		305	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		306	handles correctly whether it is set or not.
123		307
124	=over 4	308	=over 4
		309
		310	=item $prev_pri = aioreq_pri [$pri]
		311
		312	Returns the priority value that would be used for the next request and, if
		313	C<$pri> is given, sets the priority for the next aio request.
		314
		315	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		316	and C<4>, respectively. Requests with higher priority will be serviced
		317	first.
		318
		319	The priority will be reset to C<0> after each call to one of the C<aio_*>
		320	functions.
		321
		322	Example: open a file with low priority, then read something from it with
		323	higher priority so the read request is serviced before other low priority
		324	open requests (potentially spamming the cache):
		325
		326	aioreq_pri -3;
		327	aio_open ..., sub {
		328	return unless $_[0];
		329
		330	aioreq_pri -2;
		331	aio_read $_[0], ..., sub {
		332	...
		333	};
		334	};
		335
		336
		337	=item aioreq_nice $pri_adjust
		338
		339	Similar to C<aioreq_pri>, but subtracts the given value from the current
		340	priority, so the effect is cumulative.
		341
125		342
126	=item aio_open $pathname, $flags, $mode, $callback->($fh)	343	=item aio_open $pathname, $flags, $mode, $callback->($fh)
127		344
128	Asynchronously open or create a file and call the callback with a newly	345	Asynchronously open or create a file and call the callback with a newly
129	created filehandle for the file.	346	created filehandle for the file.
…		…
135	list. They are the same as used by C<sysopen>.	352	list. They are the same as used by C<sysopen>.
136		353
137	Likewise, C<$mode> specifies the mode of the newly created file, if it	354	Likewise, C<$mode> specifies the mode of the newly created file, if it
138	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,	355	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
139	except that it is mandatory (i.e. use C<0> if you don't create new files,	356	except that it is mandatory (i.e. use C<0> if you don't create new files,
140	and C<0666> or C<0777> if you do).	357	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		358	by the umask in effect then the request is being executed, so better never
		359	change the umask.
141		360
142	Example:	361	Example:
143		362
144	aio_open "/etc/passwd", O_RDONLY, 0, sub {	363	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
145	if ($_[0]) {	364	if ($_[0]) {
146	print "open successful, fh is $_[0]\n";	365	print "open successful, fh is $_[0]\n";
147	...	366	...
148	} else {	367	} else {
149	die "open failed: $!\n";	368	die "open failed: $!\n";
150	}	369	}
151	};	370	};
152		371
		372
153	=item aio_close $fh, $callback->($status)	373	=item aio_close $fh, $callback->($status)
154		374
155	Asynchronously close a file and call the callback with the result	375	Asynchronously close a file and call the callback with the result
156	code. I<WARNING:> although accepted, you should not pass in a perl	376	code.
157	filehandle here, as perl will likely close the file descriptor another
158	time when the filehandle is destroyed. Normally, you can safely call perls
159	C<close> or just let filehandles go out of scope.
160		377
161	This is supposed to be a bug in the API, so that might change. It's	378	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
162	therefore best to avoid this function.	379	closing the file descriptor associated with the filehandle itself.
		380
		381	Therefore, C<aio_close> will not close the filehandle - instead it will
		382	use dup2 to overwrite the file descriptor with the write-end of a pipe
		383	(the pipe fd will be created on demand and will be cached).
		384
		385	Or in other words: the file descriptor will be closed, but it will not be
		386	free for reuse until the perl filehandle is closed.
		387
		388	=cut
163		389
164	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	390	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
165		391
166	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	392	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
167		393
168	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	394	Reads or writes C<$length> bytes from or to the specified C<$fh> and
169	into the scalar given by C<data> and offset C<dataoffset> and calls the	395	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset>
170	callback without the actual number of bytes read (or -1 on error, just	396	and calls the callback without the actual number of bytes read (or -1 on
171	like the syscall).	397	error, just like the syscall).
		398
		399	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		400	offset plus the actual number of bytes read.
		401
		402	If C<$offset> is undefined, then the current file descriptor offset will
		403	be used (and updated), otherwise the file descriptor offset will not be
		404	changed by these calls.
		405
		406	If C<$length> is undefined in C<aio_write>, use the remaining length of
		407	C<$data>.
		408
		409	If C<$dataoffset> is less than zero, it will be counted from the end of
		410	C<$data>.
172		411
173	The C<$data> scalar I<MUST NOT> be modified in any way while the request	412	The C<$data> scalar I<MUST NOT> be modified in any way while the request
174	is outstanding. Modifying it can result in segfaults or WW3 (if the	413	is outstanding. Modifying it can result in segfaults or World War III (if
175	necessary/optional hardware is installed).	414	the necessary/optional hardware is installed).
176		415
177	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	416	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
178	offset C<0> within the scalar:	417	offset C<0> within the scalar:
179		418
180	aio_read $fh, 7, 15, $buffer, 0, sub {	419	aio_read $fh, 7, 15, $buffer, 0, sub {
181	$_[0] > 0 or die "read error: $!";	420	$_[0] > 0 or die "read error: $!";
182	print "read $_[0] bytes: <$buffer>\n";	421	print "read $_[0] bytes: <$buffer>\n";
183	};	422	};
184		423
185	=item aio_move $srcpath, $dstpath, $callback->($status)
186
187	Try to move the I<file> (directories not supported as either source or
188	destination) from C<$srcpath> to C<$dstpath> and call the callback with
189	the C<0> (error) or C<-1> ok.
190
191	This is a composite request that tries to rename(2) the file first. If
192	rename files with C<EXDEV>, it creates the destination file with mode 0200
193	and copies the contents of the source file into it using C<aio_sendfile>,
194	followed by restoring atime, mtime, access mode and uid/gid, in that
195	order, and unlinking the C<$srcpath>.
196
197	If an error occurs, the partial destination file will be unlinked, if
198	possible, except when setting atime, mtime, access mode and uid/gid, where
199	errors are being ignored.
200
201	=cut
202
203	sub aio_move($$$) {
204	my ($src, $dst, $cb) = @_;
205
206	my $grp = aio_group;
207
208	add $grp aio_rename $src, $dst, sub {
209	if ($_[0] && $! == EXDEV) {
210	add $grp aio_open $src, O_RDONLY, 0, sub {
211	if (my $src_fh = $_[0]) {
212	my @stat = stat $src_fh;
213
214	add $grp aio_open $dst, O_WRONLY, 0200, sub {
215	if (my $dst_fh = $_[0]) {
216	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
217	close $src_fh;
218
219	if ($_[0] == $stat[7]) {
220	utime $stat[8], $stat[9], $dst;
221	chmod $stat[2] & 07777, $dst_fh;
222	chown $stat[4], $stat[5], $dst_fh;
223	close $dst_fh;
224
225	add $grp aio_unlink $src, sub {
226	$cb->($_[0]);
227	};
228	} else {
229	my $errno = $!;
230	add $grp aio_unlink $dst, sub {
231	$! = $errno;
232	$cb->(-1);
233	};
234	}
235	};
236	} else {
237	$cb->(-1);
238	}
239	},
240
241	} else {
242	$cb->(-1);
243	}
244	};
245	} else {
246	$cb->($_[0]);
247	}
248	};
249
250	$grp
251	}
252		424
253	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)	425	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
254		426
255	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts	427	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
256	reading at byte offset C<$in_offset>, and starts writing at the current	428	reading at byte offset C<$in_offset>, and starts writing at the current
257	file offset of C<$out_fh>. Because of that, it is not safe to issue more	429	file offset of C<$out_fh>. Because of that, it is not safe to issue more
258	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each	430	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
259	other.	431	other.
260		432
		433	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		434	are written, and there is no way to find out how many bytes have been read
		435	from C<aio_sendfile> alone, as C<aio_sendfile> only provides the number of
		436	bytes written to C<$out_fh>. Only if the result value equals C<$length>
		437	one can assume that C<$length> bytes have been read.
		438
		439	Unlike with other C<aio_> functions, it makes a lot of sense to use
		440	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		441	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		442	the socket I/O will be non-blocking. Note, however, that you can run into
		443	a trap where C<aio_sendfile> reads some data with readahead, then fails
		444	to write all data, and when the socket is ready the next time, the data
		445	in the cache is already lost, forcing C<aio_sendfile> to again hit the
		446	disk. Explicit C<aio_read> + C<aio_write> let's you control resource usage
		447	much better.
		448
261	This call tries to make use of a native C<sendfile> syscall to provide	449	This call tries to make use of a native C<sendfile> syscall to provide
262	zero-copy operation. For this to work, C<$out_fh> should refer to a	450	zero-copy operation. For this to work, C<$out_fh> should refer to a
263	socket, and C<$in_fh> should refer to mmap'able file.	451	socket, and C<$in_fh> should refer to an mmap'able file.
264		452
265	If the native sendfile call fails or is not implemented, it will be	453	If a native sendfile cannot be found or it fails with C<ENOSYS>,
		454	C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or C<ENOTSOCK>,
266	emulated, so you can call C<aio_sendfile> on any type of filehandle	455	it will be emulated, so you can call C<aio_sendfile> on any type of
267	regardless of the limitations of the operating system.	456	filehandle regardless of the limitations of the operating system.
268		457
269	Please note, however, that C<aio_sendfile> can read more bytes from
270	C<$in_fh> than are written, and there is no way to find out how many
271	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only
272	provides the number of bytes written to C<$out_fh>. Only if the result
273	value equals C<$length> one can assume that C<$length> bytes have been
274	read.
275		458
276	=item aio_readahead $fh,$offset,$length, $callback->($retval)	459	=item aio_readahead $fh,$offset,$length, $callback->($retval)
277		460
278	C<aio_readahead> populates the page cache with data from a file so that	461	C<aio_readahead> populates the page cache with data from a file so that
279	subsequent reads from that file will not block on disk I/O. The C<$offset>	462	subsequent reads from that file will not block on disk I/O. The C<$offset>
…		…
285	file. The current file offset of the file is left unchanged.	468	file. The current file offset of the file is left unchanged.
286		469
287	If that syscall doesn't exist (likely if your OS isn't Linux) it will be	470	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
288	emulated by simply reading the data, which would have a similar effect.	471	emulated by simply reading the data, which would have a similar effect.
289		472
		473
290	=item aio_stat $fh_or_path, $callback->($status)	474	=item aio_stat $fh_or_path, $callback->($status)
291		475
292	=item aio_lstat $fh, $callback->($status)	476	=item aio_lstat $fh, $callback->($status)
293		477
294	Works like perl's C<stat> or C<lstat> in void context. The callback will	478	Works like perl's C<stat> or C<lstat> in void context. The callback will
…		…
307	aio_stat "/etc/passwd", sub {	491	aio_stat "/etc/passwd", sub {
308	$_[0] and die "stat failed: $!";	492	$_[0] and die "stat failed: $!";
309	print "size is ", -s _, "\n";	493	print "size is ", -s _, "\n";
310	};	494	};
311		495
		496
		497	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		498
		499	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		500	whether a file handle or path was passed.
		501
		502	On success, the callback is passed a hash reference with the following
		503	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		504	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		505	is passed.
		506
		507	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		508	C<ST_NOSUID>.
		509
		510	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		511	their correct value when available, or to C<0> on systems that do
		512	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		513	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		514	C<ST_NODIRATIME> and C<ST_RELATIME>.
		515
		516	Example: stat C</wd> and dump out the data if successful.
		517
		518	aio_statvfs "/wd", sub {
		519	my $f = $_[0]
		520	or die "statvfs: $!";
		521
		522	use Data::Dumper;
		523	say Dumper $f;
		524	};
		525
		526	# result:
		527	{
		528	bsize => 1024,
		529	bfree => 4333064312,
		530	blocks => 10253828096,
		531	files => 2050765568,
		532	flag => 4096,
		533	favail => 2042092649,
		534	bavail => 4333064312,
		535	ffree => 2042092649,
		536	namemax => 255,
		537	frsize => 1024,
		538	fsid => 1810
		539	}
		540
		541
		542	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		543
		544	Works like perl's C<utime> function (including the special case of $atime
		545	and $mtime being undef). Fractional times are supported if the underlying
		546	syscalls support them.
		547
		548	When called with a pathname, uses utimes(2) if available, otherwise
		549	utime(2). If called on a file descriptor, uses futimes(2) if available,
		550	otherwise returns ENOSYS, so this is not portable.
		551
		552	Examples:
		553
		554	# set atime and mtime to current time (basically touch(1)):
		555	aio_utime "path", undef, undef;
		556	# set atime to current time and mtime to beginning of the epoch:
		557	aio_utime "path", time, undef; # undef==0
		558
		559
		560	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		561
		562	Works like perl's C<chown> function, except that C<undef> for either $uid
		563	or $gid is being interpreted as "do not change" (but -1 can also be used).
		564
		565	Examples:
		566
		567	# same as "chown root path" in the shell:
		568	aio_chown "path", 0, -1;
		569	# same as above:
		570	aio_chown "path", 0, undef;
		571
		572
		573	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		574
		575	Works like truncate(2) or ftruncate(2).
		576
		577
		578	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		579
		580	Works like perl's C<chmod> function.
		581
		582
312	=item aio_unlink $pathname, $callback->($status)	583	=item aio_unlink $pathname, $callback->($status)
313		584
314	Asynchronously unlink (delete) a file and call the callback with the	585	Asynchronously unlink (delete) a file and call the callback with the
315	result code.	586	result code.
316		587
		588
		589	=item aio_mknod $path, $mode, $dev, $callback->($status)
		590
		591	[EXPERIMENTAL]
		592
		593	Asynchronously create a device node (or fifo). See mknod(2).
		594
		595	The only (POSIX-) portable way of calling this function is:
		596
		597	aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		598
		599
317	=item aio_link $srcpath, $dstpath, $callback->($status)	600	=item aio_link $srcpath, $dstpath, $callback->($status)
318		601
319	Asynchronously create a new link to the existing object at C<$srcpath> at	602	Asynchronously create a new link to the existing object at C<$srcpath> at
320	the path C<$dstpath> and call the callback with the result code.	603	the path C<$dstpath> and call the callback with the result code.
321		604
		605
322	=item aio_symlink $srcpath, $dstpath, $callback->($status)	606	=item aio_symlink $srcpath, $dstpath, $callback->($status)
323		607
324	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at	608	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
325	the path C<$dstpath> and call the callback with the result code.	609	the path C<$dstpath> and call the callback with the result code.
326		610
		611
		612	=item aio_readlink $path, $callback->($link)
		613
		614	Asynchronously read the symlink specified by C<$path> and pass it to
		615	the callback. If an error occurs, nothing or undef gets passed to the
		616	callback.
		617
		618
327	=item aio_rename $srcpath, $dstpath, $callback->($status)	619	=item aio_rename $srcpath, $dstpath, $callback->($status)
328		620
329	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as	621	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
330	rename(2) and call the callback with the result code.	622	rename(2) and call the callback with the result code.
331		623
		624
		625	=item aio_mkdir $pathname, $mode, $callback->($status)
		626
		627	Asynchronously mkdir (create) a directory and call the callback with
		628	the result code. C<$mode> will be modified by the umask at the time the
		629	request is executed, so do not change your umask.
		630
		631
332	=item aio_rmdir $pathname, $callback->($status)	632	=item aio_rmdir $pathname, $callback->($status)
333		633
334	Asynchronously rmdir (delete) a directory and call the callback with the	634	Asynchronously rmdir (delete) a directory and call the callback with the
335	result code.	635	result code.
		636
336		637
337	=item aio_readdir $pathname, $callback->($entries)	638	=item aio_readdir $pathname, $callback->($entries)
338		639
339	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire	640	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
340	directory (i.e. opendir + readdir + closedir). The entries will not be	641	directory (i.e. opendir + readdir + closedir). The entries will not be
341	sorted, and will B<NOT> include the C<.> and C<..> entries.	642	sorted, and will B<NOT> include the C<.> and C<..> entries.
342		643
343	The callback a single argument which is either C<undef> or an array-ref	644	The callback is passed a single argument which is either C<undef> or an
344	with the filenames.	645	array-ref with the filenames.
		646
		647
		648	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		649
		650	Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune
		651	behaviour and output format. In case of an error, C<$entries> will be
		652	C<undef>.
		653
		654	The flags are a combination of the following constants, ORed together (the
		655	flags will also be passed to the callback, possibly modified):
		656
		657	=over 4
		658
		659	=item IO::AIO::READDIR_DENTS
		660
		661	When this flag is off, then the callback gets an arrayref with of names
		662	only (as with C<aio_readdir>), otherwise it gets an arrayref with
		663	C<[$name, $type, $inode]> arrayrefs, each describing a single directory
		664	entry in more detail.
		665
		666	C<$name> is the name of the entry.
		667
		668	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		669
		670	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		671	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		672	C<IO::AIO::DT_WHT>.
		673
		674	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
		675	know, you have to run stat yourself. Also, for speed reasons, the C<$type>
		676	scalars are read-only: you can not modify them.
		677
		678	C<$inode> is the inode number (which might not be exact on systems with 64
		679	bit inode numbers and 32 bit perls). This field has unspecified content on
		680	systems that do not deliver the inode information.
		681
		682	=item IO::AIO::READDIR_DIRS_FIRST
		683
		684	When this flag is set, then the names will be returned in an order where
		685	likely directories come first. This is useful when you need to quickly
		686	find directories, or you want to find all directories while avoiding to
		687	stat() each entry.
		688
		689	If the system returns type information in readdir, then this is used
		690	to find directories directly. Otherwise, likely directories are files
		691	beginning with ".", or otherwise files with no dots, of which files with
		692	short names are tried first.
		693
		694	=item IO::AIO::READDIR_STAT_ORDER
		695
		696	When this flag is set, then the names will be returned in an order
		697	suitable for stat()'ing each one. That is, when you plan to stat()
		698	all files in the given directory, then the returned order will likely
		699	be fastest.
		700
		701	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
		702	the likely dirs come first, resulting in a less optimal stat order.
		703
		704	=item IO::AIO::READDIR_FOUND_UNKNOWN
		705
		706	This flag should not be set when calling C<aio_readdirx>. Instead, it
		707	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		708	C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all
		709	C<$type>'s are known, which can be used to speed up some algorithms.
		710
		711	=back
		712
		713
		714	=item aio_load $path, $data, $callback->($status)
		715
		716	This is a composite request that tries to fully load the given file into
		717	memory. Status is the same as with aio_read.
		718
		719	=cut
		720
		721	sub aio_load($$;$) {
		722	my ($path, undef, $cb) = @_;
		723	my $data = \$_[1];
		724
		725	my $pri = aioreq_pri;
		726	my $grp = aio_group $cb;
		727
		728	aioreq_pri $pri;
		729	add $grp aio_open $path, O_RDONLY, 0, sub {
		730	my $fh = shift
		731	or return $grp->result (-1);
		732
		733	aioreq_pri $pri;
		734	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		735	$grp->result ($_[0]);
		736	};
		737	};
		738
		739	$grp
		740	}
		741
		742	=item aio_copy $srcpath, $dstpath, $callback->($status)
		743
		744	Try to copy the I<file> (directories not supported as either source or
		745	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		746	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		747
		748	This is a composite request that creates the destination file with
		749	mode 0200 and copies the contents of the source file into it using
		750	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		751	uid/gid, in that order.
		752
		753	If an error occurs, the partial destination file will be unlinked, if
		754	possible, except when setting atime, mtime, access mode and uid/gid, where
		755	errors are being ignored.
		756
		757	=cut
		758
		759	sub aio_copy($$;$) {
		760	my ($src, $dst, $cb) = @_;
		761
		762	my $pri = aioreq_pri;
		763	my $grp = aio_group $cb;
		764
		765	aioreq_pri $pri;
		766	add $grp aio_open $src, O_RDONLY, 0, sub {
		767	if (my $src_fh = $_[0]) {
		768	my @stat = stat $src_fh; # hmm, might block over nfs?
		769
		770	aioreq_pri $pri;
		771	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		772	if (my $dst_fh = $_[0]) {
		773	aioreq_pri $pri;
		774	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		775	if ($_[0] == $stat[7]) {
		776	$grp->result (0);
		777	close $src_fh;
		778
		779	my $ch = sub {
		780	aioreq_pri $pri;
		781	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		782	aioreq_pri $pri;
		783	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		784	aioreq_pri $pri;
		785	add $grp aio_close $dst_fh;
		786	}
		787	};
		788	};
		789
		790	aioreq_pri $pri;
		791	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		792	if ($_[0] < 0 && $! == ENOSYS) {
		793	aioreq_pri $pri;
		794	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		795	} else {
		796	$ch->();
		797	}
		798	};
		799	} else {
		800	$grp->result (-1);
		801	close $src_fh;
		802	close $dst_fh;
		803
		804	aioreq $pri;
		805	add $grp aio_unlink $dst;
		806	}
		807	};
		808	} else {
		809	$grp->result (-1);
		810	}
		811	},
		812
		813	} else {
		814	$grp->result (-1);
		815	}
		816	};
		817
		818	$grp
		819	}
		820
		821	=item aio_move $srcpath, $dstpath, $callback->($status)
		822
		823	Try to move the I<file> (directories not supported as either source or
		824	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		825	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		826
		827	This is a composite request that tries to rename(2) the file first; if
		828	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		829	that is successful, unlinks the C<$srcpath>.
		830
		831	=cut
		832
		833	sub aio_move($$;$) {
		834	my ($src, $dst, $cb) = @_;
		835
		836	my $pri = aioreq_pri;
		837	my $grp = aio_group $cb;
		838
		839	aioreq_pri $pri;
		840	add $grp aio_rename $src, $dst, sub {
		841	if ($_[0] && $! == EXDEV) {
		842	aioreq_pri $pri;
		843	add $grp aio_copy $src, $dst, sub {
		844	$grp->result ($_[0]);
		845
		846	if (!$_[0]) {
		847	aioreq_pri $pri;
		848	add $grp aio_unlink $src;
		849	}
		850	};
		851	} else {
		852	$grp->result ($_[0]);
		853	}
		854	};
		855
		856	$grp
		857	}
345		858
346	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)	859	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
347		860
348	Scans a directory (similar to C<aio_readdir>) but additionally tries to	861	Scans a directory (similar to C<aio_readdir>) but additionally tries to
349	separate the entries of directory C<$path> into two sets of names, ones	862	efficiently separate the entries of directory C<$path> into two sets of
350	you can recurse into (directories or links to them), and ones you cannot	863	names, directories you can recurse into (directories), and ones you cannot
351	recurse into (everything else).	864	recurse into (everything else, including symlinks to directories).
352		865
353	C<aio_scandir> is a composite request that consists of many sub	866	C<aio_scandir> is a composite request that creates of many sub requests_
354	requests. C<$maxreq> specifies the maximum number of outstanding aio	867	C<$maxreq> specifies the maximum number of outstanding aio requests that
355	requests that this function generates. If it is C<< <= 0 >>, then a	868	this function generates. If it is C<< <= 0 >>, then a suitable default
356	suitable default will be chosen (currently 8).	869	will be chosen (currently 4).
357		870
358	On error, the callback is called without arguments, otherwise it receives	871	On error, the callback is called without arguments, otherwise it receives
359	two array-refs with path-relative entry names.	872	two array-refs with path-relative entry names.
360		873
361	Example:	874	Example:
…		…
368		881
369	Implementation notes.	882	Implementation notes.
370		883
371	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.	884	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
372		885
		886	If readdir returns file type information, then this is used directly to
		887	find directories.
		888
373	After reading the directory, the modification time, size etc. of the	889	Otherwise, after reading the directory, the modification time, size etc.
374	directory before and after the readdir is checked, and if they match (and	890	of the directory before and after the readdir is checked, and if they
375	isn't the current time), the link count will be used to decide how many	891	match (and isn't the current time), the link count will be used to decide
376	entries are directories (if >= 2). Otherwise, no knowledge of the number	892	how many entries are directories (if >= 2). Otherwise, no knowledge of the
377	of subdirectories will be assumed.	893	number of subdirectories will be assumed.
378		894
379	Then entries will be sorted into likely directories (everything without	895	Then entries will be sorted into likely directories a non-initial dot
380	a non-initial dot currently) and likely non-directories (everything	896	currently) and likely non-directories (see C<aio_readdirx>). Then every
381	else). Then every entry plus an appended C</.> will be C<stat>'ed,	897	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
382	likely directories first. If that succeeds, it assumes that the entry	898	in order of their inode numbers. If that succeeds, it assumes that the
383	is a directory or a symlink to directory (which will be checked	899	entry is a directory or a symlink to directory (which will be checked
384	seperately). This is often faster than stat'ing the entry itself because	900	seperately). This is often faster than stat'ing the entry itself because
385	filesystems might detect the type of the entry without reading the inode	901	filesystems might detect the type of the entry without reading the inode
386	data (e.g. ext2fs filetype feature).	902	data (e.g. ext2fs filetype feature), even on systems that cannot return
		903	the filetype information on readdir.
387		904
388	If the known number of directories (link count - 2) has been reached, the	905	If the known number of directories (link count - 2) has been reached, the
389	rest of the entries is assumed to be non-directories.	906	rest of the entries is assumed to be non-directories.
390		907
391	This only works with certainty on POSIX (= UNIX) filesystems, which	908	This only works with certainty on POSIX (= UNIX) filesystems, which
…		…
395	as those tend to return 0 or 1 as link counts, which disables the	912	as those tend to return 0 or 1 as link counts, which disables the
396	directory counting heuristic.	913	directory counting heuristic.
397		914
398	=cut	915	=cut
399		916
400	sub aio_scandir($$$) {	917	sub aio_scandir($$;$) {
401	my ($path, $maxreq, $cb) = @_;	918	my ($path, $maxreq, $cb) = @_;
402		919
		920	my $pri = aioreq_pri;
		921
403	my $grp = aio_group;	922	my $grp = aio_group $cb;
404		923
405	$maxreq = 8 if $maxreq <= 0;	924	$maxreq = 4 if $maxreq <= 0;
406		925
407	# stat once	926	# stat once
		927	aioreq_pri $pri;
408	add $grp aio_stat $path, sub {	928	add $grp aio_stat $path, sub {
409	return $cb->() if $_[0];	929	return $grp->result () if $_[0];
410	my $now = time;	930	my $now = time;
411	my $hash1 = join ":", (stat _)[0,1,3,7,9];	931	my $hash1 = join ":", (stat _)[0,1,3,7,9];
412		932
413	# read the directory entries	933	# read the directory entries
		934	aioreq_pri $pri;
414	add $grp aio_readdir $path, sub {	935	add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub {
415	my $entries = shift	936	my $entries = shift
416	or return $cb->();	937	or return $grp->result ();
417		938
418	# stat the dir another time	939	# stat the dir another time
		940	aioreq_pri $pri;
419	add $grp aio_stat $path, sub {	941	add $grp aio_stat $path, sub {
420	my $hash2 = join ":", (stat _)[0,1,3,7,9];	942	my $hash2 = join ":", (stat _)[0,1,3,7,9];
421		943
422	my $ndirs;	944	my $ndirs;
423		945
424	# take the slow route if anything looks fishy	946	# take the slow route if anything looks fishy
425	if ($hash1 ne $hash2 or (stat _)[9] == $now) {	947	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
426	$ndirs = -1;	948	$ndirs = -1;
427	} else {	949	} else {
428	# if nlink == 2, we are finished	950	# if nlink == 2, we are finished
429	# on non-posix-fs's, we rely on nlink < 2	951	# for non-posix-fs's, we rely on nlink < 2
430	$ndirs = (stat _)[3] - 2	952	$ndirs = (stat _)[3] - 2
431	or return $cb->([], $entries);	953	or return $grp->result ([], $entries);
432	}	954	}
433		955
434	# sort into likely dirs and likely nondirs
435	# dirs == files without ".", short entries first
436	$entries = [map $_->[0],
437	sort { $b->[1] cmp $a->[1] }
438	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
439	@$entries];
440
441	my (@dirs, @nondirs);	956	my (@dirs, @nondirs);
442		957
443	my ($statcb, $schedcb);	958	my $statgrp = add $grp aio_group sub {
444	my $nreq = 0;	959	$grp->result (\@dirs, \@nondirs);
		960	};
445		961
446	$schedcb = sub {	962	limit $statgrp $maxreq;
447	if (@$entries) {	963	feed $statgrp sub {
448	if ($nreq < $maxreq) {	964	return unless @$entries;
449	my $ent = pop @$entries;	965	my $entry = shift @$entries;
		966
		967	aioreq_pri $pri;
		968	add $statgrp aio_stat "$path/$entry/.", sub {
		969	if ($_[0] < 0) {
		970	push @nondirs, $entry;
		971	} else {
		972	# need to check for real directory
		973	aioreq_pri $pri;
		974	add $statgrp aio_lstat "$path/$entry", sub {
		975	if (-d _) {
		976	push @dirs, $entry;
		977
		978	unless (--$ndirs) {
		979	push @nondirs, @$entries;
		980	feed $statgrp;
		981	}
		982	} else {
		983	push @nondirs, $entry;
		984	}
450	$nreq++;	985	}
451	add $grp aio_stat "$path/$ent/.", sub { $statcb->($_[0], $ent) };
452	}	986	}
453	} elsif (!$nreq) {
454	# finished
455	undef $statcb;
456	undef $schedcb;
457	$cb->(\@dirs, \@nondirs) if $cb;
458	undef $cb;
459	}	987	};
460	};	988	};
461	$statcb = sub {
462	my ($status, $entry) = @_;
463
464	if ($status < 0) {
465	$nreq--;
466	push @nondirs, $entry;
467	&$schedcb;
468	} else {
469	# need to check for real directory
470	add $grp aio_lstat "$path/$entry", sub {
471	$nreq--;
472
473	if (-d _) {
474	push @dirs, $entry;
475
476	if (!--$ndirs) {
477	push @nondirs, @$entries;
478	$entries = [];
479	}
480	} else {
481	push @nondirs, $entry;
482	}
483
484	&$schedcb;
485	}
486	}
487	};
488
489	&$schedcb while @$entries && $nreq < $maxreq;
490	};	989	};
491	};	990	};
492	};	991	};
493		992
494	$grp	993	$grp
495	}	994	}
496		995
		996	=item aio_rmtree $path, $callback->($status)
		997
		998	Delete a directory tree starting (and including) C<$path>, return the
		999	status of the final C<rmdir> only. This is a composite request that
		1000	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1001	everything else.
		1002
		1003	=cut
		1004
		1005	sub aio_rmtree;
		1006	sub aio_rmtree($;$) {
		1007	my ($path, $cb) = @_;
		1008
		1009	my $pri = aioreq_pri;
		1010	my $grp = aio_group $cb;
		1011
		1012	aioreq_pri $pri;
		1013	add $grp aio_scandir $path, 0, sub {
		1014	my ($dirs, $nondirs) = @_;
		1015
		1016	my $dirgrp = aio_group sub {
		1017	add $grp aio_rmdir $path, sub {
		1018	$grp->result ($_[0]);
		1019	};
		1020	};
		1021
		1022	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1023	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1024
		1025	add $grp $dirgrp;
		1026	};
		1027
		1028	$grp
		1029	}
		1030
		1031	=item aio_sync $callback->($status)
		1032
		1033	Asynchronously call sync and call the callback when finished.
		1034
497	=item aio_fsync $fh, $callback->($status)	1035	=item aio_fsync $fh, $callback->($status)
498		1036
499	Asynchronously call fsync on the given filehandle and call the callback	1037	Asynchronously call fsync on the given filehandle and call the callback
500	with the fsync result code.	1038	with the fsync result code.
501		1039
…		…
505	callback with the fdatasync result code.	1043	callback with the fdatasync result code.
506		1044
507	If this call isn't available because your OS lacks it or it couldn't be	1045	If this call isn't available because your OS lacks it or it couldn't be
508	detected, it will be emulated by calling C<fsync> instead.	1046	detected, it will be emulated by calling C<fsync> instead.
509		1047
		1048	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1049
		1050	Sync the data portion of the file specified by C<$offset> and C<$length>
		1051	to disk (but NOT the metadata), by calling the Linux-specific
		1052	sync_file_range call. If sync_file_range is not available or it returns
		1053	ENOSYS, then fdatasync or fsync is being substituted.
		1054
		1055	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1056	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1057	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1058	manpage for details.
		1059
		1060	=item aio_pathsync $path, $callback->($status)
		1061
		1062	This request tries to open, fsync and close the given path. This is a
		1063	composite request intended to sync directories after directory operations
		1064	(E.g. rename). This might not work on all operating systems or have any
		1065	specific effect, but usually it makes sure that directory changes get
		1066	written to disc. It works for anything that can be opened for read-only,
		1067	not just directories.
		1068
		1069	Future versions of this function might fall back to other methods when
		1070	C<fsync> on the directory fails (such as calling C<sync>).
		1071
		1072	Passes C<0> when everything went ok, and C<-1> on error.
		1073
		1074	=cut
		1075
		1076	sub aio_pathsync($;$) {
		1077	my ($path, $cb) = @_;
		1078
		1079	my $pri = aioreq_pri;
		1080	my $grp = aio_group $cb;
		1081
		1082	aioreq_pri $pri;
		1083	add $grp aio_open $path, O_RDONLY, 0, sub {
		1084	my ($fh) = @_;
		1085	if ($fh) {
		1086	aioreq_pri $pri;
		1087	add $grp aio_fsync $fh, sub {
		1088	$grp->result ($_[0]);
		1089
		1090	aioreq_pri $pri;
		1091	add $grp aio_close $fh;
		1092	};
		1093	} else {
		1094	$grp->result (-1);
		1095	}
		1096	};
		1097
		1098	$grp
		1099	}
		1100
		1101	=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1102
		1103	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1104	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1105	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1106	scalar must only be modified in-place while an aio operation is pending on
		1107	it).
		1108
		1109	It calls the C<msync> function of your OS, if available, with the memory
		1110	area starting at C<$offset> in the string and ending C<$length> bytes
		1111	later. If C<$length> is negative, counts from the end, and if C<$length>
		1112	is C<undef>, then it goes till the end of the string. The flags can be
		1113	a combination of C<IO::AIO::MS_ASYNC>, C<IO::AIO::MS_INVALIDATE> and
		1114	C<IO::AIO::MS_SYNC>.
		1115
		1116	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1117
		1118	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1119	scalars.
		1120
		1121	It touches (reads or writes) all memory pages in the specified
		1122	range inside the scalar. All caveats and parameters are the same
		1123	as for C<aio_msync>, above, except for flags, which must be either
		1124	C<0> (which reads all pages and ensures they are instantiated) or
		1125	C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
		1126	writing an octet from it, which dirties the page).
		1127
		1128	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1129
		1130	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1131	scalars.
		1132
		1133	It reads in all the pages of the underlying storage into memory (if any)
		1134	and locks them, so they are not getting swapped/paged out or removed.
		1135
		1136	If C<$length> is undefined, then the scalar will be locked till the end.
		1137
		1138	On systems that do not implement C<mlock>, this function returns C<-1>
		1139	and sets errno to C<ENOSYS>.
		1140
		1141	Note that the corresponding C<munlock> is synchronous and is
		1142	documented under L<MISCELLANEOUS FUNCTIONS>.
		1143
		1144	Example: open a file, mmap and mlock it - both will be undone when
		1145	C<$data> gets destroyed.
		1146
		1147	open my $fh, "<", $path or die "$path: $!";
		1148	my $data;
		1149	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1150	aio_mlock $data; # mlock in background
		1151
		1152	=item aio_mlockall $flags, $callback->($status)
		1153
		1154	Calls the C<mlockall> function with the given C<$flags> (a combination of
		1155	C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
		1156
		1157	On systems that do not implement C<mlockall>, this function returns C<-1>
		1158	and sets errno to C<ENOSYS>.
		1159
		1160	Note that the corresponding C<munlockall> is synchronous and is
		1161	documented under L<MISCELLANEOUS FUNCTIONS>.
		1162
		1163	Example: asynchronously lock all current and future pages into memory.
		1164
		1165	aio_mlockall IO::AIO::MCL_FUTURE;
		1166
510	=item aio_group $callback->()	1167	=item aio_group $callback->(...)
511
512	[EXPERIMENTAL]
513		1168
514	This is a very special aio request: Instead of doing something, it is a	1169	This is a very special aio request: Instead of doing something, it is a
515	container for other aio requests, which is useful if you want to bundle	1170	container for other aio requests, which is useful if you want to bundle
516	many requests into a single, composite, request.	1171	many requests into a single, composite, request with a definite callback
		1172	and the ability to cancel the whole request with its subrequests.
517		1173
518	Returns an object of class L<IO::AIO::GRP>. See its documentation below	1174	Returns an object of class L<IO::AIO::GRP>. See its documentation below
519	for more info.	1175	for more info.
520		1176
521	Example:	1177	Example:
…		…
527	add $grp	1183	add $grp
528	(aio_stat ...),	1184	(aio_stat ...),
529	(aio_stat ...),	1185	(aio_stat ...),
530	...;	1186	...;
531		1187
		1188	=item aio_nop $callback->()
		1189
		1190	This is a special request - it does nothing in itself and is only used for
		1191	side effects, such as when you want to add a dummy request to a group so
		1192	that finishing the requests in the group depends on executing the given
		1193	code.
		1194
		1195	While this request does nothing, it still goes through the execution
		1196	phase and still requires a worker thread. Thus, the callback will not
		1197	be executed immediately but only after other requests in the queue have
		1198	entered their execution phase. This can be used to measure request
		1199	latency.
		1200
532	=item IO::AIO::aio_sleep $fractional_seconds, $callback->() *NOT EXPORTED*	1201	=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
533		1202
534	Mainly used for debugging and benchmarking, this aio request puts one of	1203	Mainly used for debugging and benchmarking, this aio request puts one of
535	the request workers to sleep for the given time.	1204	the request workers to sleep for the given time.
536		1205
537	While it is theoretically handy to have simple I/O scheduling requests	1206	While it is theoretically handy to have simple I/O scheduling requests
538	like sleep and file handle readable/writable, the overhead this creates	1207	like sleep and file handle readable/writable, the overhead this creates is
539	is immense, so do not use this function except to put your application	1208	immense (it blocks a thread for a long time) so do not use this function
540	under artificial I/O pressure.	1209	except to put your application under artificial I/O pressure.
541		1210
542	=back	1211	=back
543		1212
544	=head2 IO::AIO::REQ CLASS	1213	=head2 IO::AIO::REQ CLASS
545		1214
546	All non-aggregate C<aio_*> functions return an object of this class when	1215	All non-aggregate C<aio_*> functions return an object of this class when
547	called in non-void context.	1216	called in non-void context.
548		1217
549	A request always moves through the following five states in its lifetime,
550	in order: B<ready> (request has been created, but has not been executed
551	yet), B<execute> (request is currently being executed), B<pending>
552	(request has been executed but callback has not been called yet),
553	B<result> (results are being processed synchronously, includes calling the
554	callback) and B<done> (request has reached the end of its lifetime and
555	holds no resources anymore).
556
557	=over 4	1218	=over 4
558		1219
559	=item $req->cancel	1220	=item cancel $req
560		1221
561	Cancels the request, if possible. Has the effect of skipping execution	1222	Cancels the request, if possible. Has the effect of skipping execution
562	when entering the B<execute> state and skipping calling the callback when	1223	when entering the B<execute> state and skipping calling the callback when
563	entering the the B<result> state, but will leave the request otherwise	1224	entering the the B<result> state, but will leave the request otherwise
564	untouched. That means that requests that currently execute will not be	1225	untouched (with the exception of readdir). That means that requests that
565	stopped and resources held by the request will not be freed prematurely.	1226	currently execute will not be stopped and resources held by the request
		1227	will not be freed prematurely.
		1228
		1229	=item cb $req $callback->(...)
		1230
		1231	Replace (or simply set) the callback registered to the request.
566		1232
567	=back	1233	=back
568		1234
569	=head2 IO::AIO::GRP CLASS	1235	=head2 IO::AIO::GRP CLASS
570		1236
…		…
585	You add requests by calling the C<add> method with one or more	1251	You add requests by calling the C<add> method with one or more
586	C<IO::AIO::REQ> objects:	1252	C<IO::AIO::REQ> objects:
587		1253
588	$grp->add (aio_unlink "...");	1254	$grp->add (aio_unlink "...");
589		1255
590	add $grp aio_stat "...", sub { ... };	1256	add $grp aio_stat "...", sub {
		1257	$_[0] or return $grp->result ("error");
		1258
		1259	# add another request dynamically, if first succeeded
		1260	add $grp aio_open "...", sub {
		1261	$grp->result ("ok");
		1262	};
		1263	};
591		1264
592	This makes it very easy to create composite requests (see the source of	1265	This makes it very easy to create composite requests (see the source of
593	C<aio_move> for an application) that work and feel like simple requests.	1266	C<aio_move> for an application) that work and feel like simple requests.
594		1267
		1268	=over 4
		1269
595	The IO::AIO::GRP objects will be cleaned up during calls to	1270	=item * The IO::AIO::GRP objects will be cleaned up during calls to
596	C<IO::AIO::poll_cb>, just like any other request.	1271	C<IO::AIO::poll_cb>, just like any other request.
597		1272
598	They can be canceled like any other request. Canceling will cancel not	1273	=item * They can be canceled like any other request. Canceling will cancel not
599	just the request itself, but also all requests it contains.	1274	only the request itself, but also all requests it contains.
600		1275
601	They can also can also be added to other IO::AIO::GRP objects.	1276	=item * They can also can also be added to other IO::AIO::GRP objects.
		1277
		1278	=item * You must not add requests to a group from within the group callback (or
		1279	any later time).
		1280
		1281	=back
602		1282
603	Their lifetime, simplified, looks like this: when they are empty, they	1283	Their lifetime, simplified, looks like this: when they are empty, they
604	will finish very quickly. If they contain only requests that are in the	1284	will finish very quickly. If they contain only requests that are in the
605	C<done> state, they will also finish. Otherwise they will continue to	1285	C<done> state, they will also finish. Otherwise they will continue to
606	exist.	1286	exist.
607		1287
		1288	That means after creating a group you have some time to add requests
		1289	(precisely before the callback has been invoked, which is only done within
		1290	the C<poll_cb>). And in the callbacks of those requests, you can add
		1291	further requests to the group. And only when all those requests have
		1292	finished will the the group itself finish.
		1293
608	=over 4	1294	=over 4
609		1295
		1296	=item add $grp ...
		1297
610	=item $grp->add (...)	1298	=item $grp->add (...)
611		1299
612	=item add $grp ...	1300	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		1301	be added, including other groups, as long as you do not create circular
		1302	dependencies.
613		1303
614	Add one or more	1304	Returns all its arguments.
615	Cancels the request, if possible. Has the effect of skipping execution	1305
616	when entering the B<execute> state and skipping calling the callback when	1306	=item $grp->cancel_subs
617	entering the the B<result> state, but will leave the request otherwise	1307
618	untouched. That means that requests that currently execute will not be	1308	Cancel all subrequests and clears any feeder, but not the group request
619	stopped and resources held by the request will not be freed prematurely.	1309	itself. Useful when you queued a lot of events but got a result early.
		1310
		1311	The group request will finish normally (you cannot add requests to the
		1312	group).
		1313
		1314	=item $grp->result (...)
		1315
		1316	Set the result value(s) that will be passed to the group callback when all
		1317	subrequests have finished and set the groups errno to the current value
		1318	of errno (just like calling C<errno> without an error number). By default,
		1319	no argument will be passed and errno is zero.
		1320
		1321	=item $grp->errno ([$errno])
		1322
		1323	Sets the group errno value to C<$errno>, or the current value of errno
		1324	when the argument is missing.
		1325
		1326	Every aio request has an associated errno value that is restored when
		1327	the callback is invoked. This method lets you change this value from its
		1328	default (0).
		1329
		1330	Calling C<result> will also set errno, so make sure you either set C<$!>
		1331	before the call to C<result>, or call c<errno> after it.
		1332
		1333	=item feed $grp $callback->($grp)
		1334
		1335	Sets a feeder/generator on this group: every group can have an attached
		1336	generator that generates requests if idle. The idea behind this is that,
		1337	although you could just queue as many requests as you want in a group,
		1338	this might starve other requests for a potentially long time. For example,
		1339	C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests,
		1340	delaying any later requests for a long time.
		1341
		1342	To avoid this, and allow incremental generation of requests, you can
		1343	instead a group and set a feeder on it that generates those requests. The
		1344	feed callback will be called whenever there are few enough (see C<limit>,
		1345	below) requests active in the group itself and is expected to queue more
		1346	requests.
		1347
		1348	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1349	not impose any limits).
		1350
		1351	If the feed does not queue more requests when called, it will be
		1352	automatically removed from the group.
		1353
		1354	If the feed limit is C<0> when this method is called, it will be set to
		1355	C<2> automatically.
		1356
		1357	Example:
		1358
		1359	# stat all files in @files, but only ever use four aio requests concurrently:
		1360
		1361	my $grp = aio_group sub { print "finished\n" };
		1362	limit $grp 4;
		1363	feed $grp sub {
		1364	my $file = pop @files
		1365	or return;
		1366
		1367	add $grp aio_stat $file, sub { ... };
		1368	};
		1369
		1370	=item limit $grp $num
		1371
		1372	Sets the feeder limit for the group: The feeder will be called whenever
		1373	the group contains less than this many requests.
		1374
		1375	Setting the limit to C<0> will pause the feeding process.
		1376
		1377	The default value for the limit is C<0>, but note that setting a feeder
		1378	automatically bumps it up to C<2>.
620		1379
621	=back	1380	=back
622		1381
623	=head2 SUPPORT FUNCTIONS	1382	=head2 SUPPORT FUNCTIONS
624		1383
		1384	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1385
625	=over 4	1386	=over 4
626		1387
627	=item $fileno = IO::AIO::poll_fileno	1388	=item $fileno = IO::AIO::poll_fileno
628		1389
629	Return the I<request result pipe file descriptor>. This filehandle must be	1390	Return the I<request result pipe file descriptor>. This filehandle must be
630	polled for reading by some mechanism outside this module (e.g. Event or	1391	polled for reading by some mechanism outside this module (e.g. EV, Glib,
631	select, see below or the SYNOPSIS). If the pipe becomes readable you have	1392	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
632	to call C<poll_cb> to check the results.	1393	you have to call C<poll_cb> to check the results.
633		1394
634	See C<poll_cb> for an example.	1395	See C<poll_cb> for an example.
635		1396
636	=item IO::AIO::poll_cb	1397	=item IO::AIO::poll_cb
637		1398
638	Process all outstanding events on the result pipe. You have to call this	1399	Process some outstanding events on the result pipe. You have to call this
639	regularly. Returns the number of events processed. Returns immediately	1400	regularly. Returns C<0> if all events could be processed, or C<-1> if it
640	when no events are outstanding.	1401	returned earlier for whatever reason. Returns immediately when no events
		1402	are outstanding. The amount of events processed depends on the settings of
		1403	C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
		1404
		1405	If not all requests were processed for whatever reason, the filehandle
		1406	will still be ready when C<poll_cb> returns, so normally you don't have to
		1407	do anything special to have it called later.
641		1408
642	Example: Install an Event watcher that automatically calls	1409	Example: Install an Event watcher that automatically calls
643	IO::AIO::poll_cb with high priority:	1410	IO::AIO::poll_cb with high priority (more examples can be found in the
		1411	SYNOPSIS section, at the top of this document):
644		1412
645	Event->io (fd => IO::AIO::poll_fileno,	1413	Event->io (fd => IO::AIO::poll_fileno,
646	poll => 'r', async => 1,	1414	poll => 'r', async => 1,
647	cb => \&IO::AIO::poll_cb);	1415	cb => \&IO::AIO::poll_cb);
648		1416
649	=item IO::AIO::poll_wait	1417	=item IO::AIO::poll_wait
650		1418
		1419	If there are any outstanding requests and none of them in the result
651	Wait till the result filehandle becomes ready for reading (simply does a	1420	phase, wait till the result filehandle becomes ready for reading (simply
652	C<select> on the filehandle. This is useful if you want to synchronously wait	1421	does a C<select> on the filehandle. This is useful if you want to
653	for some requests to finish).	1422	synchronously wait for some requests to finish).
654		1423
655	See C<nreqs> for an example.	1424	See C<nreqs> for an example.
656		1425
		1426	=item IO::AIO::poll
		1427
		1428	Waits until some requests have been handled.
		1429
		1430	Returns the number of requests processed, but is otherwise strictly
		1431	equivalent to:
		1432
		1433	IO::AIO::poll_wait, IO::AIO::poll_cb
		1434
657	=item IO::AIO::nreqs	1435	=item IO::AIO::flush
658		1436
659	Returns the number of requests currently outstanding (i.e. for which their	1437	Wait till all outstanding AIO requests have been handled.
660	callback has not been invoked yet).
661		1438
662	Example: wait till there are no outstanding requests anymore:	1439	Strictly equivalent to:
663		1440
664	IO::AIO::poll_wait, IO::AIO::poll_cb	1441	IO::AIO::poll_wait, IO::AIO::poll_cb
665	while IO::AIO::nreqs;	1442	while IO::AIO::nreqs;
666		1443
667	=item IO::AIO::flush	1444	=item IO::AIO::max_poll_reqs $nreqs
668		1445
669	Wait till all outstanding AIO requests have been handled.	1446	=item IO::AIO::max_poll_time $seconds
670		1447
671	Strictly equivalent to:	1448	These set the maximum number of requests (default C<0>, meaning infinity)
		1449	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1450	the maximum amount of time (default C<0>, meaning infinity) spent in
		1451	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1452	of time C<poll_cb> is allowed to use).
672		1453
673	IO::AIO::poll_wait, IO::AIO::poll_cb	1454	Setting C<max_poll_time> to a non-zero value creates an overhead of one
674	while IO::AIO::nreqs;	1455	syscall per request processed, which is not normally a problem unless your
		1456	callbacks are really really fast or your OS is really really slow (I am
		1457	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
675		1458
676	=item IO::AIO::poll	1459	Setting these is useful if you want to ensure some level of
		1460	interactiveness when perl is not fast enough to process all requests in
		1461	time.
677		1462
678	Waits until some requests have been handled.	1463	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
679		1464
680	Strictly equivalent to:	1465	Example: Install an Event watcher that automatically calls
		1466	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1467	program get the CPU sometimes even under high AIO load.
681		1468
682	IO::AIO::poll_wait, IO::AIO::poll_cb	1469	# try not to spend much more than 0.1s in poll_cb
683	if IO::AIO::nreqs;	1470	IO::AIO::max_poll_time 0.1;
		1471
		1472	# use a low priority so other tasks have priority
		1473	Event->io (fd => IO::AIO::poll_fileno,
		1474	poll => 'r', nice => 1,
		1475	cb => &IO::AIO::poll_cb);
		1476
		1477	=back
		1478
		1479	=head3 CONTROLLING THE NUMBER OF THREADS
		1480
		1481	=over
684		1482
685	=item IO::AIO::min_parallel $nthreads	1483	=item IO::AIO::min_parallel $nthreads
686		1484
687	Set the minimum number of AIO threads to C<$nthreads>. The current default	1485	Set the minimum number of AIO threads to C<$nthreads>. The current
688	is C<4>, which means four asynchronous operations can be done at one time	1486	default is C<8>, which means eight asynchronous operations can execute
689	(the number of outstanding operations, however, is unlimited).	1487	concurrently at any one time (the number of outstanding requests,
		1488	however, is unlimited).
690		1489
691	IO::AIO starts threads only on demand, when an AIO request is queued and	1490	IO::AIO starts threads only on demand, when an AIO request is queued and
692	no free thread exists.	1491	no free thread exists. Please note that queueing up a hundred requests can
		1492	create demand for a hundred threads, even if it turns out that everything
		1493	is in the cache and could have been processed faster by a single thread.
693		1494
694	It is recommended to keep the number of threads low, as some Linux	1495	It is recommended to keep the number of threads relatively low, as some
695	kernel versions will scale negatively with the number of threads (higher	1496	Linux kernel versions will scale negatively with the number of threads
696	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	1497	(higher parallelity => MUCH higher latency). With current Linux 2.6
697	threads should be fine.	1498	versions, 4-32 threads should be fine.
698		1499
699	Under most circumstances you don't need to call this function, as the	1500	Under most circumstances you don't need to call this function, as the
700	module selects a default that is suitable for low to moderate load.	1501	module selects a default that is suitable for low to moderate load.
701		1502
702	=item IO::AIO::max_parallel $nthreads	1503	=item IO::AIO::max_parallel $nthreads
…		…
711	This module automatically runs C<max_parallel 0> at program end, to ensure	1512	This module automatically runs C<max_parallel 0> at program end, to ensure
712	that all threads are killed and that there are no outstanding requests.	1513	that all threads are killed and that there are no outstanding requests.
713		1514
714	Under normal circumstances you don't need to call this function.	1515	Under normal circumstances you don't need to call this function.
715		1516
		1517	=item IO::AIO::max_idle $nthreads
		1518
		1519	Limit the number of threads (default: 4) that are allowed to idle (i.e.,
		1520	threads that did not get a request to process within 10 seconds). That
		1521	means if a thread becomes idle while C<$nthreads> other threads are also
		1522	idle, it will free its resources and exit.
		1523
		1524	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1525	to allow for extremely high load situations, but want to free resources
		1526	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1527
		1528	The default is probably ok in most situations, especially if thread
		1529	creation is fast. If thread creation is very slow on your system you might
		1530	want to use larger values.
		1531
716	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	1532	=item IO::AIO::max_outstanding $maxreqs
		1533
		1534	This is a very bad function to use in interactive programs because it
		1535	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1536	use an C<aio_group> together with a feed callback.
717		1537
718	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1538	Sets the maximum number of outstanding requests to C<$nreqs>. If you
719	try to queue up more than this number of requests, the caller will block until	1539	do queue up more than this number of requests, the next call to the
720	some requests have been handled.	1540	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
		1541	function will block until the limit is no longer exceeded.
721		1542
722	The default is very large, so normally there is no practical limit. If you	1543	The default value is very large, so there is no practical limit on the
723	queue up many requests in a loop it often improves speed if you set	1544	number of outstanding requests.
724	this to a relatively low number, such as C<100>.
725		1545
726	Under normal circumstances you don't need to call this function.	1546	You can still queue as many requests as you want. Therefore,
		1547	C<max_outstanding> is mainly useful in simple scripts (with low values) or
		1548	as a stop gap to shield against fatal memory overflow (with large values).
727		1549
728	=back	1550	=back
729		1551
		1552	=head3 STATISTICAL INFORMATION
		1553
		1554	=over
		1555
		1556	=item IO::AIO::nreqs
		1557
		1558	Returns the number of requests currently in the ready, execute or pending
		1559	states (i.e. for which their callback has not been invoked yet).
		1560
		1561	Example: wait till there are no outstanding requests anymore:
		1562
		1563	IO::AIO::poll_wait, IO::AIO::poll_cb
		1564	while IO::AIO::nreqs;
		1565
		1566	=item IO::AIO::nready
		1567
		1568	Returns the number of requests currently in the ready state (not yet
		1569	executed).
		1570
		1571	=item IO::AIO::npending
		1572
		1573	Returns the number of requests currently in the pending state (executed,
		1574	but not yet processed by poll_cb).
		1575
		1576	=back
		1577
		1578	=head3 MISCELLANEOUS FUNCTIONS
		1579
		1580	IO::AIO implements some functions that might be useful, but are not
		1581	asynchronous.
		1582
		1583	=over 4
		1584
		1585	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		1586
		1587	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		1588	but is blocking (this makes most sense if you know the input data is
		1589	likely cached already and the output filehandle is set to non-blocking
		1590	operations).
		1591
		1592	Returns the number of bytes copied, or C<-1> on error.
		1593
		1594	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		1595
		1596	Simply calls the C<posix_fadvise> function (see its
		1597	manpage for details). The following advice constants are
		1598	avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		1599	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		1600	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		1601
		1602	On systems that do not implement C<posix_fadvise>, this function returns
		1603	ENOSYS, otherwise the return value of C<posix_fadvise>.
		1604
		1605	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		1606
		1607	Simply calls the C<posix_madvise> function (see its
		1608	manpage for details). The following advice constants are
		1609	avaiable: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		1610	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
		1611
		1612	On systems that do not implement C<posix_madvise>, this function returns
		1613	ENOSYS, otherwise the return value of C<posix_madvise>.
		1614
		1615	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		1616
		1617	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		1618	$scalar (see its manpage for details). The following protect
		1619	constants are avaiable: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		1620	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		1621
		1622	On systems that do not implement C<mprotect>, this function returns
		1623	ENOSYS, otherwise the return value of C<mprotect>.
		1624
		1625	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		1626
		1627	Memory-maps a file (or anonymous memory range) and attaches it to the
		1628	given C<$scalar>, which will act like a string scalar.
		1629
		1630	The only operations allowed on the scalar are C<substr>/C<vec> that don't
		1631	change the string length, and most read-only operations such as copying it
		1632	or searching it with regexes and so on.
		1633
		1634	Anything else is unsafe and will, at best, result in memory leaks.
		1635
		1636	The memory map associated with the C<$scalar> is automatically removed
		1637	when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
		1638	C<IO::AIO::munmap> functions are called.
		1639
		1640	This calls the C<mmap>(2) function internally. See your system's manual
		1641	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		1642
		1643	The C<$length> must be larger than zero and smaller than the actual
		1644	filesize.
		1645
		1646	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		1647	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		1648
		1649	C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
		1650	C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
		1651	not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
		1652	(which is set to C<MAP_ANON> if your system only provides this
		1653	constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
		1654	C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
		1655	C<IO::AIO::MAP_NONBLOCK>
		1656
		1657	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		1658
		1659	C<$offset> is the offset from the start of the file - it generally must be
		1660	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		1661
		1662	Example:
		1663
		1664	use Digest::MD5;
		1665	use IO::AIO;
		1666
		1667	open my $fh, "<verybigfile"
		1668	or die "$!";
		1669
		1670	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		1671	or die "verybigfile: $!";
		1672
		1673	my $fast_md5 = md5 $data;
		1674
		1675	=item IO::AIO::munmap $scalar
		1676
		1677	Removes a previous mmap and undefines the C<$scalar>.
		1678
		1679	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		1680
		1681	Calls the C<munlock> function, undoing the effects of a previous
		1682	C<aio_mlock> call (see its description for details).
		1683
		1684	=item IO::AIO::munlockall
		1685
		1686	Calls the C<munlockall> function.
		1687
		1688	On systems that do not implement C<munlockall>, this function returns
		1689	ENOSYS, otherwise the return value of C<munlockall>.
		1690
		1691	=back
		1692
730	=cut	1693	=cut
731		1694
732	# support function to convert a fd into a perl filehandle
733	sub _fd2fh {
734	return undef if $_[0] < 0;
735
736	# try to generate nice filehandles
737	my $sym = "IO::AIO::fd#$_[0]";
738	local *$sym;
739
740	open *$sym, "+<&=$_[0]" # usually works under any unix
741	or open *$sym, "<&=$_[0]" # cygwin needs this
742	or open *$sym, ">&=$_[0]" # or this
743	or return undef;
744
745	*$sym
746	}
747
748	min_parallel 4;	1695	min_parallel 8;
749		1696
750	END {	1697	END { flush }
751	max_parallel 0;
752	}
753		1698
754	1;	1699	1;
		1700
		1701	=head1 EVENT LOOP INTEGRATION
		1702
		1703	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		1704	automatically into many event loops:
		1705
		1706	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		1707	use AnyEvent::AIO;
		1708
		1709	You can also integrate IO::AIO manually into many event loops, here are
		1710	some examples of how to do this:
		1711
		1712	# EV integration
		1713	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		1714
		1715	# Event integration
		1716	Event->io (fd => IO::AIO::poll_fileno,
		1717	poll => 'r',
		1718	cb => \&IO::AIO::poll_cb);
		1719
		1720	# Glib/Gtk2 integration
		1721	add_watch Glib::IO IO::AIO::poll_fileno,
		1722	in => sub { IO::AIO::poll_cb; 1 };
		1723
		1724	# Tk integration
		1725	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		1726	readable => \&IO::AIO::poll_cb);
		1727
		1728	# Danga::Socket integration
		1729	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		1730	\&IO::AIO::poll_cb);
755		1731
756	=head2 FORK BEHAVIOUR	1732	=head2 FORK BEHAVIOUR
757		1733
758	This module should do "the right thing" when the process using it forks:	1734	This module should do "the right thing" when the process using it forks:
759		1735
760	Before the fork, IO::AIO enters a quiescent state where no requests	1736	Before the fork, IO::AIO enters a quiescent state where no requests
761	can be added in other threads and no results will be processed. After	1737	can be added in other threads and no results will be processed. After
762	the fork the parent simply leaves the quiescent state and continues	1738	the fork the parent simply leaves the quiescent state and continues
763	request/result processing, while the child clears the request/result	1739	request/result processing, while the child frees the request/result queue
764	queue (so the requests started before the fork will only be handled in	1740	(so that the requests started before the fork will only be handled in the
765	the parent). Threads will be started on demand until the limit ste in the	1741	parent). Threads will be started on demand until the limit set in the
766	parent process has been reached again.	1742	parent process has been reached again.
767		1743
768	In short: the parent will, after a short pause, continue as if fork had	1744	In short: the parent will, after a short pause, continue as if fork had
769	not been called, while the child will act as if IO::AIO has not been used	1745	not been called, while the child will act as if IO::AIO has not been used
770	yet.	1746	yet.
771		1747
		1748	=head2 MEMORY USAGE
		1749
		1750	Per-request usage:
		1751
		1752	Each aio request uses - depending on your architecture - around 100-200
		1753	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1754	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1755	scalars and other data passed into aio requests will also be locked and
		1756	will consume memory till the request has entered the done state.
		1757
		1758	This is not awfully much, so queuing lots of requests is not usually a
		1759	problem.
		1760
		1761	Per-thread usage:
		1762
		1763	In the execution phase, some aio requests require more memory for
		1764	temporary buffers, and each thread requires a stack and other data
		1765	structures (usually around 16k-128k, depending on the OS).
		1766
		1767	=head1 KNOWN BUGS
		1768
		1769	Known bugs will be fixed in the next release.
		1770
772	=head1 SEE ALSO	1771	=head1 SEE ALSO
773		1772
774	L<Coro>, L<Linux::AIO> (obsolete).	1773	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		1774	more natural syntax.
775		1775
776	=head1 AUTHOR	1776	=head1 AUTHOR
777		1777
778	Marc Lehmann <schmorp@schmorp.de>	1778	Marc Lehmann <schmorp@schmorp.de>
779	http://home.schmorp.de/	1779	http://home.schmorp.de/

Diff Legend

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