[ViewVC] Diff of: cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.6 by root, Sun Jul 10 22:19:48 2005 UTC vs.
Revision 1.94 by root, Wed Nov 8 02:01:02 2006 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	# Event	21	# version 2+ has request and group objects
		22	use IO::AIO 2;
		23
		24	aioreq_pri 4; # give next request a very high priority
		25	my $req = aio_unlink "/tmp/file", sub { };
		26	$req->cancel; # cancel request if still in queue
		27
		28	my $grp = aio_group sub { print "all stats done\n" };
		29	add $grp aio_stat "..." for ...;
		30
		31	# AnyEvent integration
		32	open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
		33	my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
		34
		35	# Event integration
21	Event->io (fd => IO::AIO::poll_fileno,	36	Event->io (fd => IO::AIO::poll_fileno,
22	poll => 'r', async => 1,	37	poll => 'r',
23	cb => \&IO::AIO::poll_cb);	38	cb => \&IO::AIO::poll_cb);
24		39
25	# Glib/Gtk2	40	# Glib/Gtk2 integration
26	add_watch Glib::IO IO::AIO::poll_fileno,	41	add_watch Glib::IO IO::AIO::poll_fileno,
27	\&IO::AIO::poll_cb;	42	in => sub { IO::AIO::poll_cb; 1 };
28		43
29	# Tk	44	# Tk integration
30	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",	45	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
31	readable => \&IO::AIO::poll_cb);	46	readable => \&IO::AIO::poll_cb);
32		47
		48	# Danga::Socket integration
		49	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		50	\&IO::AIO::poll_cb);
		51
33	=head1 DESCRIPTION	52	=head1 DESCRIPTION
34		53
35	This module implements asynchronous I/O using whatever means your	54	This module implements asynchronous I/O using whatever means your
36	operating system supports.	55	operating system supports.
37		56
		57	Asynchronous means that operations that can normally block your program
		58	(e.g. reading from disk) will be done asynchronously: the operation
		59	will still block, but you can do something else in the meantime. This
		60	is extremely useful for programs that need to stay interactive even
		61	when doing heavy I/O (GUI programs, high performance network servers
		62	etc.), but can also be used to easily do operations in parallel that are
		63	normally done sequentially, e.g. stat'ing many files, which is much faster
		64	on a RAID volume or over NFS when you do a number of stat operations
		65	concurrently.
		66
		67	While most of this works on all types of file descriptors (for example
		68	sockets), using these functions on file descriptors that support
		69	nonblocking operation (again, sockets, pipes etc.) is very inefficient or
		70	might not work (aio_read fails on sockets/pipes/fifos). Use an event loop
		71	for that (such as the L<Event\|Event> module): IO::AIO will naturally fit
		72	into such an event loop itself.
		73
38	Currently, a number of threads are started that execute your read/writes	74	In this version, a number of threads are started that execute your
39	and signal their completion. You don't need thread support in your libc or	75	requests and signal their completion. You don't need thread support
40	perl, and the threads created by this module will not be visible to the	76	in perl, and the threads created by this module will not be visible
41	pthreads library. In the future, this module might make use of the native	77	to perl. In the future, this module might make use of the native aio
42	aio functions available on many operating systems. However, they are often	78	functions available on many operating systems. However, they are often
43	not well-supported (Linux doesn't allow them on normal files currently,	79	not well-supported or restricted (GNU/Linux doesn't allow them on normal
44	for example), and they would only support aio_read and aio_write, so the	80	files currently, for example), and they would only support aio_read and
45	remaining functionality would have to be implemented using threads anyway.	81	aio_write, so the remaining functionality would have to be implemented
		82	using threads anyway.
46		83
47	Although the module will work with in the presence of other threads, it is	84	Although the module will work with in the presence of other (Perl-)
48	currently not reentrant, so use appropriate locking yourself.	85	threads, it is currently not reentrant in any way, so use appropriate
		86	locking yourself, always call C<poll_cb> from within the same thread, or
		87	never call C<poll_cb> (or other C<aio_> functions) recursively.
		88
		89	=head2 EXAMPLE
		90
		91	This is a simple example that uses the Event module and loads
		92	F</etc/passwd> asynchronously:
		93
		94	use Fcntl;
		95	use Event;
		96	use IO::AIO;
		97
		98	# register the IO::AIO callback with Event
		99	Event->io (fd => IO::AIO::poll_fileno,
		100	poll => 'r',
		101	cb => \&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	Event::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	Event::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
49		184
50	=cut	185	=cut
51		186
52	package IO::AIO;	187	package IO::AIO;
53		188
		189	no warnings;
		190	use strict 'vars';
		191
54	use base 'Exporter';	192	use base 'Exporter';
55		193
56	use Fcntl ();
57
58	BEGIN {	194	BEGIN {
59	$VERSION = 0.2;	195	our $VERSION = '2.2';
60		196
61	@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink	197	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat
62	aio_fsync aio_fdatasync aio_readahead);	198	aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink
63	@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);	199	aio_readlink aio_fsync aio_fdatasync aio_readahead aio_rename aio_link
		200	aio_move aio_copy aio_group aio_nop aio_mknod);
		201	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		202	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		203	min_parallel max_parallel max_idle
		204	nreqs nready npending nthreads
		205	max_poll_time max_poll_reqs);
		206
		207	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
64		208
65	require XSLoader;	209	require XSLoader;
66	XSLoader::load IO::AIO, $VERSION;	210	XSLoader::load ("IO::AIO", $VERSION);
67	}	211	}
68		212
69	=head1 FUNCTIONS	213	=head1 FUNCTIONS
70		214
71	=head2 AIO FUNCTIONS	215	=head2 AIO REQUEST FUNCTIONS
72		216
73	All the C<aio_*> calls are more or less thin wrappers around the syscall	217	All the C<aio_*> calls are more or less thin wrappers around the syscall
74	with the same name (sans C<aio_>). The arguments are similar or identical,	218	with the same name (sans C<aio_>). The arguments are similar or identical,
75	and they all accept an additional C<$callback> argument which must be	219	and they all accept an additional (and optional) C<$callback> argument
76	a code reference. This code reference will get called with the syscall	220	which must be a code reference. This code reference will get called with
77	return code (e.g. most syscalls return C<-1> on error, unlike perl, which	221	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
78	usually delivers "false") as it's sole argument when the given syscall has	222	perl, which usually delivers "false") as it's sole argument when the given
79	been executed asynchronously.	223	syscall has been executed asynchronously.
80		224
81	All functions that expect a filehandle will also accept a file descriptor.	225	All functions expecting a filehandle keep a copy of the filehandle
		226	internally until the request has finished.
82		227
		228	All functions return request objects of type L<IO::AIO::REQ> that allow
		229	further manipulation of those requests while they are in-flight.
		230
83	The filenames you pass to these routines I<must> be absolute. The reason	231	The pathnames you pass to these routines I<must> be absolute and
84	is that at the time the request is being executed, the current working	232	encoded as octets. The reason for the former is that at the time the
85	directory could have changed. Alternatively, you can make sure that you	233	request is being executed, the current working directory could have
86	never change the current working directory.	234	changed. Alternatively, you can make sure that you never change the
		235	current working directory anywhere in the program and then use relative
		236	paths.
		237
		238	To encode pathnames as octets, either make sure you either: a) always pass
		239	in filenames you got from outside (command line, readdir etc.) without
		240	tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
		241	your pathnames to the locale (or other) encoding in effect in the user
		242	environment, d) use Glib::filename_from_unicode on unicode filenames or e)
		243	use something else to ensure your scalar has the correct contents.
		244
		245	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		246	handles correctly wether it is set or not.
87		247
88	=over 4	248	=over 4
89		249
		250	=item $prev_pri = aioreq_pri [$pri]
		251
		252	Returns the priority value that would be used for the next request and, if
		253	C<$pri> is given, sets the priority for the next aio request.
		254
		255	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		256	and C<4>, respectively. Requests with higher priority will be serviced
		257	first.
		258
		259	The priority will be reset to C<0> after each call to one of the C<aio_*>
		260	functions.
		261
		262	Example: open a file with low priority, then read something from it with
		263	higher priority so the read request is serviced before other low priority
		264	open requests (potentially spamming the cache):
		265
		266	aioreq_pri -3;
		267	aio_open ..., sub {
		268	return unless $_[0];
		269
		270	aioreq_pri -2;
		271	aio_read $_[0], ..., sub {
		272	...
		273	};
		274	};
		275
		276	=item aioreq_nice $pri_adjust
		277
		278	Similar to C<aioreq_pri>, but subtracts the given value from the current
		279	priority, so the effect is cumulative.
		280
90	=item aio_open $pathname, $flags, $mode, $callback	281	=item aio_open $pathname, $flags, $mode, $callback->($fh)
91		282
92	Asynchronously open or create a file and call the callback with a newly	283	Asynchronously open or create a file and call the callback with a newly
93	created filehandle for the file.	284	created filehandle for the file.
94		285
95	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,	286	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
96	for an explanation.	287	for an explanation.
97		288
98	The C<$mode> argument is a bitmask. See the C<Fcntl> module for a	289	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
99	list. They are the same as used in C<sysopen>.	290	list. They are the same as used by C<sysopen>.
		291
		292	Likewise, C<$mode> specifies the mode of the newly created file, if it
		293	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
		294	except that it is mandatory (i.e. use C<0> if you don't create new files,
		295	and C<0666> or C<0777> if you do).
100		296
101	Example:	297	Example:
102		298
103	aio_open "/etc/passwd", O_RDONLY, 0, sub {	299	aio_open "/etc/passwd", O_RDONLY, 0, sub {
104	if ($_[0]) {	300	if ($_[0]) {
…		…
107	} else {	303	} else {
108	die "open failed: $!\n";	304	die "open failed: $!\n";
109	}	305	}
110	};	306	};
111		307
112	=item aio_close $fh, $callback	308	=item aio_close $fh, $callback->($status)
113		309
114	Asynchronously close a file and call the callback with the result	310	Asynchronously close a file and call the callback with the result
115	code. I<WARNING:> although accepted, you should not pass in a perl	311	code. I<WARNING:> although accepted, you should not pass in a perl
116	filehandle here, as perl will likely close the file descriptor itself when	312	filehandle here, as perl will likely close the file descriptor another
117	the filehandle is destroyed. Normally, you can safely call perls C<close>	313	time when the filehandle is destroyed. Normally, you can safely call perls
118	or just let filehandles go out of scope.	314	C<close> or just let filehandles go out of scope.
119		315
		316	This is supposed to be a bug in the API, so that might change. It's
		317	therefore best to avoid this function.
		318
120	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	319	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
121		320
122	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	321	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
123		322
124	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	323	Reads or writes C<length> bytes from the specified C<fh> and C<offset>
125	into the scalar given by C<data> and offset C<dataoffset> and calls the	324	into the scalar given by C<data> and offset C<dataoffset> and calls the
126	callback without the actual number of bytes read (or -1 on error, just	325	callback without the actual number of bytes read (or -1 on error, just
127	like the syscall).	326	like the syscall).
128		327
		328	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		329	is outstanding. Modifying it can result in segfaults or WW3 (if the
		330	necessary/optional hardware is installed).
		331
129	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at	332	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
130	offset C<0> within the scalar:	333	offset C<0> within the scalar:
131		334
132	aio_read $fh, 7, 15, $buffer, 0, sub {	335	aio_read $fh, 7, 15, $buffer, 0, sub {
133	$_[0] >= 0 or die "read error: $!";	336	$_[0] > 0 or die "read error: $!";
134	print "read <$buffer>\n";	337	print "read $_[0] bytes: <$buffer>\n";
135	};	338	};
136		339
		340	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		341
		342	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		343	reading at byte offset C<$in_offset>, and starts writing at the current
		344	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		345	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		346	other.
		347
		348	This call tries to make use of a native C<sendfile> syscall to provide
		349	zero-copy operation. For this to work, C<$out_fh> should refer to a
		350	socket, and C<$in_fh> should refer to mmap'able file.
		351
		352	If the native sendfile call fails or is not implemented, it will be
		353	emulated, so you can call C<aio_sendfile> on any type of filehandle
		354	regardless of the limitations of the operating system.
		355
		356	Please note, however, that C<aio_sendfile> can read more bytes from
		357	C<$in_fh> than are written, and there is no way to find out how many
		358	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only
		359	provides the number of bytes written to C<$out_fh>. Only if the result
		360	value equals C<$length> one can assume that C<$length> bytes have been
		361	read.
		362
137	=item aio_readahead $fh,$offset,$length, $callback	363	=item aio_readahead $fh,$offset,$length, $callback->($retval)
138		364
139	Asynchronously reads the specified byte range into the page cache, using
140	the C<readahead> syscall. If that syscall doesn't exist the status will be
141	C<-1> and C<$!> is set to ENOSYS.
142
143	readahead() populates the page cache with data from a file so that	365	C<aio_readahead> populates the page cache with data from a file so that
144	subsequent reads from that file will not block on disk I/O. The C<$offset>	366	subsequent reads from that file will not block on disk I/O. The C<$offset>
145	argument specifies the starting point from which data is to be read and	367	argument specifies the starting point from which data is to be read and
146	C<$length> specifies the number of bytes to be read. I/O is performed in	368	C<$length> specifies the number of bytes to be read. I/O is performed in
147	whole pages, so that offset is effectively rounded down to a page boundary	369	whole pages, so that offset is effectively rounded down to a page boundary
148	and bytes are read up to the next page boundary greater than or equal to	370	and bytes are read up to the next page boundary greater than or equal to
149	(off-set+length). aio_readahead() does not read beyond the end of the	371	(off-set+length). C<aio_readahead> does not read beyond the end of the
150	file. The current file offset of the file is left unchanged.	372	file. The current file offset of the file is left unchanged.
151		373
		374	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
		375	emulated by simply reading the data, which would have a similar effect.
		376
152	=item aio_stat $fh_or_path, $callback	377	=item aio_stat $fh_or_path, $callback->($status)
153		378
154	=item aio_lstat $fh, $callback	379	=item aio_lstat $fh, $callback->($status)
155		380
156	Works like perl's C<stat> or C<lstat> in void context. The callback will	381	Works like perl's C<stat> or C<lstat> in void context. The callback will
157	be called after the stat and the results will be available using C<stat _>	382	be called after the stat and the results will be available using C<stat _>
158	or C<-s _> etc...	383	or C<-s _> etc...
159		384
…		…
169	aio_stat "/etc/passwd", sub {	394	aio_stat "/etc/passwd", sub {
170	$_[0] and die "stat failed: $!";	395	$_[0] and die "stat failed: $!";
171	print "size is ", -s _, "\n";	396	print "size is ", -s _, "\n";
172	};	397	};
173		398
174	=item aio_unlink $pathname, $callback	399	=item aio_unlink $pathname, $callback->($status)
175		400
176	Asynchronously unlink (delete) a file and call the callback with the	401	Asynchronously unlink (delete) a file and call the callback with the
177	result code.	402	result code.
178		403
		404	=item aio_mknod $path, $mode, $dev, $callback->($status)
		405
		406	[EXPERIMENTAL]
		407
		408	Asynchronously create a device node (or fifo). See mknod(2).
		409
		410	The only (POSIX-) portable way of calling this function is:
		411
		412	aio_mknod $path, IO::AIO::S_IFIFO \| $mode, 0, sub { ...
		413
		414	=item aio_link $srcpath, $dstpath, $callback->($status)
		415
		416	Asynchronously create a new link to the existing object at C<$srcpath> at
		417	the path C<$dstpath> and call the callback with the result code.
		418
		419	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		420
		421	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		422	the path C<$dstpath> and call the callback with the result code.
		423
		424	=item aio_readlink $path, $callback->($link)
		425
		426	Asynchronously read the symlink specified by C<$path> and pass it to
		427	the callback. If an error occurs, nothing or undef gets passed to the
		428	callback.
		429
		430	=item aio_rename $srcpath, $dstpath, $callback->($status)
		431
		432	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		433	rename(2) and call the callback with the result code.
		434
		435	=item aio_rmdir $pathname, $callback->($status)
		436
		437	Asynchronously rmdir (delete) a directory and call the callback with the
		438	result code.
		439
		440	=item aio_readdir $pathname, $callback->($entries)
		441
		442	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		443	directory (i.e. opendir + readdir + closedir). The entries will not be
		444	sorted, and will B<NOT> include the C<.> and C<..> entries.
		445
		446	The callback a single argument which is either C<undef> or an array-ref
		447	with the filenames.
		448
		449	=item aio_copy $srcpath, $dstpath, $callback->($status)
		450
		451	Try to copy the I<file> (directories not supported as either source or
		452	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		453	the C<0> (error) or C<-1> ok.
		454
		455	This is a composite request that it creates the destination file with
		456	mode 0200 and copies the contents of the source file into it using
		457	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		458	uid/gid, in that order.
		459
		460	If an error occurs, the partial destination file will be unlinked, if
		461	possible, except when setting atime, mtime, access mode and uid/gid, where
		462	errors are being ignored.
		463
		464	=cut
		465
		466	sub aio_copy($$;$) {
		467	my ($src, $dst, $cb) = @_;
		468
		469	my $pri = aioreq_pri;
		470	my $grp = aio_group $cb;
		471
		472	aioreq_pri $pri;
		473	add $grp aio_open $src, O_RDONLY, 0, sub {
		474	if (my $src_fh = $_[0]) {
		475	my @stat = stat $src_fh;
		476
		477	aioreq_pri $pri;
		478	add $grp aio_open $dst, O_CREAT \| O_WRONLY \| O_TRUNC, 0200, sub {
		479	if (my $dst_fh = $_[0]) {
		480	aioreq_pri $pri;
		481	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		482	if ($_[0] == $stat[7]) {
		483	$grp->result (0);
		484	close $src_fh;
		485
		486	# those should not normally block. should. should.
		487	utime $stat[8], $stat[9], $dst;
		488	chmod $stat[2] & 07777, $dst_fh;
		489	chown $stat[4], $stat[5], $dst_fh;
		490	close $dst_fh;
		491	} else {
		492	$grp->result (-1);
		493	close $src_fh;
		494	close $dst_fh;
		495
		496	aioreq $pri;
		497	add $grp aio_unlink $dst;
		498	}
		499	};
		500	} else {
		501	$grp->result (-1);
		502	}
		503	},
		504
		505	} else {
		506	$grp->result (-1);
		507	}
		508	};
		509
		510	$grp
		511	}
		512
		513	=item aio_move $srcpath, $dstpath, $callback->($status)
		514
		515	Try to move the I<file> (directories not supported as either source or
		516	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		517	the C<0> (error) or C<-1> ok.
		518
		519	This is a composite request that tries to rename(2) the file first. If
		520	rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if
		521	that is successful, unlinking the C<$srcpath>.
		522
		523	=cut
		524
		525	sub aio_move($$;$) {
		526	my ($src, $dst, $cb) = @_;
		527
		528	my $pri = aioreq_pri;
		529	my $grp = aio_group $cb;
		530
		531	aioreq_pri $pri;
		532	add $grp aio_rename $src, $dst, sub {
		533	if ($_[0] && $! == EXDEV) {
		534	aioreq_pri $pri;
		535	add $grp aio_copy $src, $dst, sub {
		536	$grp->result ($_[0]);
		537
		538	if (!$_[0]) {
		539	aioreq_pri $pri;
		540	add $grp aio_unlink $src;
		541	}
		542	};
		543	} else {
		544	$grp->result ($_[0]);
		545	}
		546	};
		547
		548	$grp
		549	}
		550
		551	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		552
		553	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		554	efficiently separate the entries of directory C<$path> into two sets of
		555	names, directories you can recurse into (directories), and ones you cannot
		556	recurse into (everything else, including symlinks to directories).
		557
		558	C<aio_scandir> is a composite request that creates of many sub requests_
		559	C<$maxreq> specifies the maximum number of outstanding aio requests that
		560	this function generates. If it is C<< <= 0 >>, then a suitable default
		561	will be chosen (currently 4).
		562
		563	On error, the callback is called without arguments, otherwise it receives
		564	two array-refs with path-relative entry names.
		565
		566	Example:
		567
		568	aio_scandir $dir, 0, sub {
		569	my ($dirs, $nondirs) = @_;
		570	print "real directories: @$dirs\n";
		571	print "everything else: @$nondirs\n";
		572	};
		573
		574	Implementation notes.
		575
		576	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		577
		578	After reading the directory, the modification time, size etc. of the
		579	directory before and after the readdir is checked, and if they match (and
		580	isn't the current time), the link count will be used to decide how many
		581	entries are directories (if >= 2). Otherwise, no knowledge of the number
		582	of subdirectories will be assumed.
		583
		584	Then entries will be sorted into likely directories (everything without
		585	a non-initial dot currently) and likely non-directories (everything
		586	else). Then every entry plus an appended C</.> will be C<stat>'ed,
		587	likely directories first. If that succeeds, it assumes that the entry
		588	is a directory or a symlink to directory (which will be checked
		589	seperately). This is often faster than stat'ing the entry itself because
		590	filesystems might detect the type of the entry without reading the inode
		591	data (e.g. ext2fs filetype feature).
		592
		593	If the known number of directories (link count - 2) has been reached, the
		594	rest of the entries is assumed to be non-directories.
		595
		596	This only works with certainty on POSIX (= UNIX) filesystems, which
		597	fortunately are the vast majority of filesystems around.
		598
		599	It will also likely work on non-POSIX filesystems with reduced efficiency
		600	as those tend to return 0 or 1 as link counts, which disables the
		601	directory counting heuristic.
		602
		603	=cut
		604
		605	sub aio_scandir($$$) {
		606	my ($path, $maxreq, $cb) = @_;
		607
		608	my $pri = aioreq_pri;
		609
		610	my $grp = aio_group $cb;
		611
		612	$maxreq = 4 if $maxreq <= 0;
		613
		614	# stat once
		615	aioreq_pri $pri;
		616	add $grp aio_stat $path, sub {
		617	return $grp->result () if $_[0];
		618	my $now = time;
		619	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		620
		621	# read the directory entries
		622	aioreq_pri $pri;
		623	add $grp aio_readdir $path, sub {
		624	my $entries = shift
		625	or return $grp->result ();
		626
		627	# stat the dir another time
		628	aioreq_pri $pri;
		629	add $grp aio_stat $path, sub {
		630	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		631
		632	my $ndirs;
		633
		634	# take the slow route if anything looks fishy
		635	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		636	$ndirs = -1;
		637	} else {
		638	# if nlink == 2, we are finished
		639	# on non-posix-fs's, we rely on nlink < 2
		640	$ndirs = (stat _)[3] - 2
		641	or return $grp->result ([], $entries);
		642	}
		643
		644	# sort into likely dirs and likely nondirs
		645	# dirs == files without ".", short entries first
		646	$entries = [map $_->[0],
		647	sort { $b->[1] cmp $a->[1] }
		648	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
		649	@$entries];
		650
		651	my (@dirs, @nondirs);
		652
		653	my $statgrp = add $grp aio_group sub {
		654	$grp->result (\@dirs, \@nondirs);
		655	};
		656
		657	limit $statgrp $maxreq;
		658	feed $statgrp sub {
		659	return unless @$entries;
		660	my $entry = pop @$entries;
		661
		662	aioreq_pri $pri;
		663	add $statgrp aio_stat "$path/$entry/.", sub {
		664	if ($_[0] < 0) {
		665	push @nondirs, $entry;
		666	} else {
		667	# need to check for real directory
		668	aioreq_pri $pri;
		669	add $statgrp aio_lstat "$path/$entry", sub {
		670	if (-d _) {
		671	push @dirs, $entry;
		672
		673	unless (--$ndirs) {
		674	push @nondirs, @$entries;
		675	feed $statgrp;
		676	}
		677	} else {
		678	push @nondirs, $entry;
		679	}
		680	}
		681	}
		682	};
		683	};
		684	};
		685	};
		686	};
		687
		688	$grp
		689	}
		690
179	=item aio_fsync $fh, $callback	691	=item aio_fsync $fh, $callback->($status)
180		692
181	Asynchronously call fsync on the given filehandle and call the callback	693	Asynchronously call fsync on the given filehandle and call the callback
182	with the fsync result code.	694	with the fsync result code.
183		695
184	=item aio_fdatasync $fh, $callback	696	=item aio_fdatasync $fh, $callback->($status)
185		697
186	Asynchronously call fdatasync on the given filehandle and call the	698	Asynchronously call fdatasync on the given filehandle and call the
187	callback with the fdatasync result code.	699	callback with the fdatasync result code.
188		700
		701	If this call isn't available because your OS lacks it or it couldn't be
		702	detected, it will be emulated by calling C<fsync> instead.
		703
		704	=item aio_group $callback->(...)
		705
		706	This is a very special aio request: Instead of doing something, it is a
		707	container for other aio requests, which is useful if you want to bundle
		708	many requests into a single, composite, request with a definite callback
		709	and the ability to cancel the whole request with its subrequests.
		710
		711	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		712	for more info.
		713
		714	Example:
		715
		716	my $grp = aio_group sub {
		717	print "all stats done\n";
		718	};
		719
		720	add $grp
		721	(aio_stat ...),
		722	(aio_stat ...),
		723	...;
		724
		725	=item aio_nop $callback->()
		726
		727	This is a special request - it does nothing in itself and is only used for
		728	side effects, such as when you want to add a dummy request to a group so
		729	that finishing the requests in the group depends on executing the given
		730	code.
		731
		732	While this request does nothing, it still goes through the execution
		733	phase and still requires a worker thread. Thus, the callback will not
		734	be executed immediately but only after other requests in the queue have
		735	entered their execution phase. This can be used to measure request
		736	latency.
		737
		738	=item IO::AIO::aio_busy $fractional_seconds, $callback->() NOT EXPORTED
		739
		740	Mainly used for debugging and benchmarking, this aio request puts one of
		741	the request workers to sleep for the given time.
		742
		743	While it is theoretically handy to have simple I/O scheduling requests
		744	like sleep and file handle readable/writable, the overhead this creates is
		745	immense (it blocks a thread for a long time) so do not use this function
		746	except to put your application under artificial I/O pressure.
		747
189	=back	748	=back
190		749
		750	=head2 IO::AIO::REQ CLASS
		751
		752	All non-aggregate C<aio_*> functions return an object of this class when
		753	called in non-void context.
		754
		755	=over 4
		756
		757	=item cancel $req
		758
		759	Cancels the request, if possible. Has the effect of skipping execution
		760	when entering the B<execute> state and skipping calling the callback when
		761	entering the the B<result> state, but will leave the request otherwise
		762	untouched. That means that requests that currently execute will not be
		763	stopped and resources held by the request will not be freed prematurely.
		764
		765	=item cb $req $callback->(...)
		766
		767	Replace (or simply set) the callback registered to the request.
		768
		769	=back
		770
		771	=head2 IO::AIO::GRP CLASS
		772
		773	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		774	objects of this class, too.
		775
		776	A IO::AIO::GRP object is a special request that can contain multiple other
		777	aio requests.
		778
		779	You create one by calling the C<aio_group> constructing function with a
		780	callback that will be called when all contained requests have entered the
		781	C<done> state:
		782
		783	my $grp = aio_group sub {
		784	print "all requests are done\n";
		785	};
		786
		787	You add requests by calling the C<add> method with one or more
		788	C<IO::AIO::REQ> objects:
		789
		790	$grp->add (aio_unlink "...");
		791
		792	add $grp aio_stat "...", sub {
		793	$_[0] or return $grp->result ("error");
		794
		795	# add another request dynamically, if first succeeded
		796	add $grp aio_open "...", sub {
		797	$grp->result ("ok");
		798	};
		799	};
		800
		801	This makes it very easy to create composite requests (see the source of
		802	C<aio_move> for an application) that work and feel like simple requests.
		803
		804	=over 4
		805
		806	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		807	C<IO::AIO::poll_cb>, just like any other request.
		808
		809	=item * They can be canceled like any other request. Canceling will cancel not
		810	only the request itself, but also all requests it contains.
		811
		812	=item * They can also can also be added to other IO::AIO::GRP objects.
		813
		814	=item * You must not add requests to a group from within the group callback (or
		815	any later time).
		816
		817	=back
		818
		819	Their lifetime, simplified, looks like this: when they are empty, they
		820	will finish very quickly. If they contain only requests that are in the
		821	C<done> state, they will also finish. Otherwise they will continue to
		822	exist.
		823
		824	That means after creating a group you have some time to add requests. And
		825	in the callbacks of those requests, you can add further requests to the
		826	group. And only when all those requests have finished will the the group
		827	itself finish.
		828
		829	=over 4
		830
		831	=item add $grp ...
		832
		833	=item $grp->add (...)
		834
		835	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		836	be added, including other groups, as long as you do not create circular
		837	dependencies.
		838
		839	Returns all its arguments.
		840
		841	=item $grp->cancel_subs
		842
		843	Cancel all subrequests and clears any feeder, but not the group request
		844	itself. Useful when you queued a lot of events but got a result early.
		845
		846	=item $grp->result (...)
		847
		848	Set the result value(s) that will be passed to the group callback when all
		849	subrequests have finished and set thre groups errno to the current value
		850	of errno (just like calling C<errno> without an error number). By default,
		851	no argument will be passed and errno is zero.
		852
		853	=item $grp->errno ([$errno])
		854
		855	Sets the group errno value to C<$errno>, or the current value of errno
		856	when the argument is missing.
		857
		858	Every aio request has an associated errno value that is restored when
		859	the callback is invoked. This method lets you change this value from its
		860	default (0).
		861
		862	Calling C<result> will also set errno, so make sure you either set C<$!>
		863	before the call to C<result>, or call c<errno> after it.
		864
		865	=item feed $grp $callback->($grp)
		866
		867	Sets a feeder/generator on this group: every group can have an attached
		868	generator that generates requests if idle. The idea behind this is that,
		869	although you could just queue as many requests as you want in a group,
		870	this might starve other requests for a potentially long time. For
		871	example, C<aio_scandir> might generate hundreds of thousands C<aio_stat>
		872	requests, delaying any later requests for a long time.
		873
		874	To avoid this, and allow incremental generation of requests, you can
		875	instead a group and set a feeder on it that generates those requests. The
		876	feed callback will be called whenever there are few enough (see C<limit>,
		877	below) requests active in the group itself and is expected to queue more
		878	requests.
		879
		880	The feed callback can queue as many requests as it likes (i.e. C<add> does
		881	not impose any limits).
		882
		883	If the feed does not queue more requests when called, it will be
		884	automatically removed from the group.
		885
		886	If the feed limit is C<0>, it will be set to C<2> automatically.
		887
		888	Example:
		889
		890	# stat all files in @files, but only ever use four aio requests concurrently:
		891
		892	my $grp = aio_group sub { print "finished\n" };
		893	limit $grp 4;
		894	feed $grp sub {
		895	my $file = pop @files
		896	or return;
		897
		898	add $grp aio_stat $file, sub { ... };
		899	};
		900
		901	=item limit $grp $num
		902
		903	Sets the feeder limit for the group: The feeder will be called whenever
		904	the group contains less than this many requests.
		905
		906	Setting the limit to C<0> will pause the feeding process.
		907
		908	=back
		909
191	=head2 SUPPORT FUNCTIONS	910	=head2 SUPPORT FUNCTIONS
192		911
		912	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		913
193	=over 4	914	=over 4
194		915
195	=item $fileno = IO::AIO::poll_fileno	916	=item $fileno = IO::AIO::poll_fileno
196		917
197	Return the I<request result pipe filehandle>. This filehandle must be	918	Return the I<request result pipe file descriptor>. This filehandle must be
198	polled for reading by some mechanism outside this module (e.g. Event	919	polled for reading by some mechanism outside this module (e.g. Event or
199	or select, see below). If the pipe becomes readable you have to call	920	select, see below or the SYNOPSIS). If the pipe becomes readable you have
200	C<poll_cb> to check the results.	921	to call C<poll_cb> to check the results.
201		922
202	See C<poll_cb> for an example.	923	See C<poll_cb> for an example.
203		924
204	=item IO::AIO::poll_cb	925	=item IO::AIO::poll_cb
205		926
206	Process all outstanding events on the result pipe. You have to call this	927	Process some outstanding events on the result pipe. You have to call this
207	regularly. Returns the number of events processed. Returns immediately	928	regularly. Returns the number of events processed. Returns immediately
208	when no events are outstanding.	929	when no events are outstanding. The amount of events processed depends on
		930	the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
209		931
210	You can use Event to multiplex, e.g.:	932	If not all requests were processed for whatever reason, the filehandle
		933	will still be ready when C<poll_cb> returns.
		934
		935	Example: Install an Event watcher that automatically calls
		936	IO::AIO::poll_cb with high priority:
211		937
212	Event->io (fd => IO::AIO::poll_fileno,	938	Event->io (fd => IO::AIO::poll_fileno,
213	poll => 'r', async => 1,	939	poll => 'r', async => 1,
214	cb => \&IO::AIO::poll_cb);	940	cb => \&IO::AIO::poll_cb);
215		941
		942	=item IO::AIO::max_poll_reqs $nreqs
		943
		944	=item IO::AIO::max_poll_time $seconds
		945
		946	These set the maximum number of requests (default C<0>, meaning infinity)
		947	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		948	the maximum amount of time (default C<0>, meaning infinity) spent in
		949	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		950	of time C<poll_cb> is allowed to use).
		951
		952	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		953	syscall per request processed, which is not normally a problem unless your
		954	callbacks are really really fast or your OS is really really slow (I am
		955	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		956
		957	Setting these is useful if you want to ensure some level of
		958	interactiveness when perl is not fast enough to process all requests in
		959	time.
		960
		961	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		962
		963	Example: Install an Event watcher that automatically calls
		964	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		965	program get the CPU sometimes even under high AIO load.
		966
		967	# try not to spend much more than 0.1s in poll_cb
		968	IO::AIO::max_poll_time 0.1;
		969
		970	# use a low priority so other tasks have priority
		971	Event->io (fd => IO::AIO::poll_fileno,
		972	poll => 'r', nice => 1,
		973	cb => &IO::AIO::poll_cb);
		974
216	=item IO::AIO::poll_wait	975	=item IO::AIO::poll_wait
217		976
		977	If there are any outstanding requests and none of them in the result
218	Wait till the result filehandle becomes ready for reading (simply does a	978	phase, wait till the result filehandle becomes ready for reading (simply
219	select on the filehandle. This is useful if you want to synchronously wait	979	does a C<select> on the filehandle. This is useful if you want to
220	for some requests to finish).	980	synchronously wait for some requests to finish).
221		981
222	See C<nreqs> for an example.	982	See C<nreqs> for an example.
223		983
		984	=item IO::AIO::poll
		985
		986	Waits until some requests have been handled.
		987
		988	Returns the number of requests processed, but is otherwise strictly
		989	equivalent to:
		990
		991	IO::AIO::poll_wait, IO::AIO::poll_cb
		992
224	=item IO::AIO::nreqs	993	=item IO::AIO::flush
225		994
226	Returns the number of requests currently outstanding.	995	Wait till all outstanding AIO requests have been handled.
227		996
228	Example: wait till there are no outstanding requests anymore:	997	Strictly equivalent to:
229		998
230	IO::AIO::poll_wait, IO::AIO::poll_cb	999	IO::AIO::poll_wait, IO::AIO::poll_cb
231	while IO::AIO::nreqs;	1000	while IO::AIO::nreqs;
232		1001
		1002	=head3 CONTROLLING THE NUMBER OF THREADS
		1003
233	=item IO::AIO::min_parallel $nthreads	1004	=item IO::AIO::min_parallel $nthreads
234		1005
235	Set the minimum number of AIO threads to C<$nthreads>. The default is	1006	Set the minimum number of AIO threads to C<$nthreads>. The current
236	C<1>, which means a single asynchronous operation can be done at one time	1007	default is C<8>, which means eight asynchronous operations can execute
237	(the number of outstanding operations, however, is unlimited).	1008	concurrently at any one time (the number of outstanding requests,
		1009	however, is unlimited).
238		1010
		1011	IO::AIO starts threads only on demand, when an AIO request is queued and
		1012	no free thread exists. Please note that queueing up a hundred requests can
		1013	create demand for a hundred threads, even if it turns out that everything
		1014	is in the cache and could have been processed faster by a single thread.
		1015
239	It is recommended to keep the number of threads low, as some Linux	1016	It is recommended to keep the number of threads relatively low, as some
240	kernel versions will scale negatively with the number of threads (higher	1017	Linux kernel versions will scale negatively with the number of threads
241	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	1018	(higher parallelity => MUCH higher latency). With current Linux 2.6
242	threads should be fine.	1019	versions, 4-32 threads should be fine.
243		1020
244	Under normal circumstances you don't need to call this function, as this	1021	Under most circumstances you don't need to call this function, as the
245	module automatically starts some threads (the exact number might change,	1022	module selects a default that is suitable for low to moderate load.
246	and is currently 4).
247		1023
248	=item IO::AIO::max_parallel $nthreads	1024	=item IO::AIO::max_parallel $nthreads
249		1025
250	Sets the maximum number of AIO threads to C<$nthreads>. If more than	1026	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
251	the specified number of threads are currently running, kill them. This	1027	specified number of threads are currently running, this function kills
252	function blocks until the limit is reached.	1028	them. This function blocks until the limit is reached.
		1029
		1030	While C<$nthreads> are zero, aio requests get queued but not executed
		1031	until the number of threads has been increased again.
253		1032
254	This module automatically runs C<max_parallel 0> at program end, to ensure	1033	This module automatically runs C<max_parallel 0> at program end, to ensure
255	that all threads are killed and that there are no outstanding requests.	1034	that all threads are killed and that there are no outstanding requests.
256		1035
257	Under normal circumstances you don't need to call this function.	1036	Under normal circumstances you don't need to call this function.
258		1037
		1038	=item IO::AIO::max_idle $nthreads
		1039
		1040	Limit the number of threads (default: 4) that are allowed to idle (i.e.,
		1041	threads that did not get a request to process within 10 seconds). That
		1042	means if a thread becomes idle while C<$nthreads> other threads are also
		1043	idle, it will free its resources and exit.
		1044
		1045	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1046	to allow for extremely high load situations, but want to free resources
		1047	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1048
		1049	The default is probably ok in most situations, especially if thread
		1050	creation is fast. If thread creation is very slow on your system you might
		1051	want to use larger values.
		1052
259	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	1053	=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs
		1054
		1055	This is a very bad function to use in interactive programs because it
		1056	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1057	use an C<aio_group> together with a feed callback.
260		1058
261	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1059	Sets the maximum number of outstanding requests to C<$nreqs>. If you
262	try to queue up more than this number of requests, the caller will block until	1060	to queue up more than this number of requests, the next call to the
263	some requests have been handled.	1061	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
		1062	function will block until the limit is no longer exceeded.
264		1063
265	The default is very large, so normally there is no practical limit. If you	1064	The default value is very large, so there is no practical limit on the
266	queue up many requests in a loop it it often improves speed if you set	1065	number of outstanding requests.
267	this to a relatively low number, such as C<100>.
268		1066
269	Under normal circumstances you don't need to call this function.	1067	You can still queue as many requests as you want. Therefore,
		1068	C<max_oustsanding> is mainly useful in simple scripts (with low values) or
		1069	as a stop gap to shield against fatal memory overflow (with large values).
		1070
		1071	=head3 STATISTICAL INFORMATION
		1072
		1073	=item IO::AIO::nreqs
		1074
		1075	Returns the number of requests currently in the ready, execute or pending
		1076	states (i.e. for which their callback has not been invoked yet).
		1077
		1078	Example: wait till there are no outstanding requests anymore:
		1079
		1080	IO::AIO::poll_wait, IO::AIO::poll_cb
		1081	while IO::AIO::nreqs;
		1082
		1083	=item IO::AIO::nready
		1084
		1085	Returns the number of requests currently in the ready state (not yet
		1086	executed).
		1087
		1088	=item IO::AIO::npending
		1089
		1090	Returns the number of requests currently in the pending state (executed,
		1091	but not yet processed by poll_cb).
270		1092
271	=back	1093	=back
272		1094
273	=cut	1095	=cut
274		1096
275	# support function to convert a fd into a perl filehandle	1097	# support function to convert a fd into a perl filehandle
276	sub _fd2fh {	1098	sub _fd2fh {
277	return undef if $_[0] < 0;	1099	return undef if $_[0] < 0;
278		1100
279	# try to be perl5.6-compatible	1101	# try to generate nice filehandles
280	local *AIO_FH;	1102	my $sym = "IO::AIO::fd#$_[0]";
281	open AIO_FH, "+<&=$_[0]"	1103	local *$sym;
		1104
		1105	open *$sym, "+<&=$_[0]" # usually works under any unix
		1106	or open *$sym, "<&=$_[0]" # cygwin needs this
		1107	or open *$sym, ">&=$_[0]" # or this
282	or return undef;	1108	or return undef;
283		1109
284	*AIO_FH	1110	*$sym
285	}	1111	}
286		1112
287	min_parallel 4;	1113	min_parallel 8;
288		1114
289	END {	1115	END {
290	max_parallel 0;	1116	min_parallel 1;
291	}	1117	flush;
		1118	};
292		1119
293	1;	1120	1;
294		1121
		1122	=head2 FORK BEHAVIOUR
		1123
		1124	This module should do "the right thing" when the process using it forks:
		1125
		1126	Before the fork, IO::AIO enters a quiescent state where no requests
		1127	can be added in other threads and no results will be processed. After
		1128	the fork the parent simply leaves the quiescent state and continues
		1129	request/result processing, while the child frees the request/result queue
		1130	(so that the requests started before the fork will only be handled in the
		1131	parent). Threads will be started on demand until the limit set in the
		1132	parent process has been reached again.
		1133
		1134	In short: the parent will, after a short pause, continue as if fork had
		1135	not been called, while the child will act as if IO::AIO has not been used
		1136	yet.
		1137
		1138	=head2 MEMORY USAGE
		1139
		1140	Per-request usage:
		1141
		1142	Each aio request uses - depending on your architecture - around 100-200
		1143	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1144	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1145	scalars and other data passed into aio requests will also be locked and
		1146	will consume memory till the request has entered the done state.
		1147
		1148	This is now awfully much, so queuing lots of requests is not usually a
		1149	problem.
		1150
		1151	Per-thread usage:
		1152
		1153	In the execution phase, some aio requests require more memory for
		1154	temporary buffers, and each thread requires a stack and other data
		1155	structures (usually around 16k-128k, depending on the OS).
		1156
		1157	=head1 KNOWN BUGS
		1158
		1159	Known bugs will be fixed in the next release.
		1160
295	=head1 SEE ALSO	1161	=head1 SEE ALSO
296		1162
297	L<Coro>, L<Linux::AIO>.	1163	L<Coro::AIO>.
298		1164
299	=head1 AUTHOR	1165	=head1 AUTHOR
300		1166
301	Marc Lehmann <schmorp@schmorp.de>	1167	Marc Lehmann <schmorp@schmorp.de>
302	http://home.schmorp.de/	1168	http://home.schmorp.de/

Diff Legend

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

Comparing IO-AIO/AIO.pm (file contents): Revision 1.6 by root, Sun Jul 10 22:19:48 2005 UTC vs. Revision 1.94 by root, Wed Nov 8 02:01:02 2006 UTC

Diff Legend

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.6 by root, Sun Jul 10 22:19:48 2005 UTC vs.
Revision 1.94 by root, Wed Nov 8 02:01:02 2006 UTC