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

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.21 by root, Wed Jul 13 00:13:09 2005 UTC vs.
Revision 1.103 by root, Tue Jan 23 22:57:34 2007 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',	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	in => sub { IO::AIO::poll_cb, 1 };	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
33	# Danga::Socket	48	# Danga::Socket integration
34	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>	49	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
35	\&IO::AIO::poll_cb);	50	\&IO::AIO::poll_cb);
36		51
37
38	=head1 DESCRIPTION	52	=head1 DESCRIPTION
39		53
40	This module implements asynchronous I/O using whatever means your	54	This module implements asynchronous I/O using whatever means your
41	operating system supports.	55	operating system supports.
42		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
43	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
44	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
45	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
46	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
47	aio functions available on many operating systems. However, they are often	78	functions available on many operating systems. However, they are often
48	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
49	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
50	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.
51		83
52	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-)
53	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
54		184
55	=cut	185	=cut
56		186
57	package IO::AIO;	187	package IO::AIO;
58		188
		189	no warnings;
		190	use strict 'vars';
		191
59	use base 'Exporter';	192	use base 'Exporter';
60		193
61	use Fcntl ();
62
63	BEGIN {	194	BEGIN {
64	$VERSION = 0.9;	195	our $VERSION = '2.33';
65		196
66	@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
67	aio_fsync aio_fdatasync aio_readahead);	198	aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink
68	@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 aio_load aio_rmtree aio_mkdir);
		201	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice aio_block));
		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';
69		208
70	require XSLoader;	209	require XSLoader;
71	XSLoader::load IO::AIO, $VERSION;	210	XSLoader::load ("IO::AIO", $VERSION);
72	}	211	}
73		212
74	=head1 FUNCTIONS	213	=head1 FUNCTIONS
75		214
76	=head2 AIO FUNCTIONS	215	=head2 AIO REQUEST FUNCTIONS
77		216
78	All the C<aio_*> calls are more or less thin wrappers around the syscall	217	All the C<aio_*> calls are more or less thin wrappers around the syscall
79	with the same name (sans C<aio_>). The arguments are similar or identical,	218	with the same name (sans C<aio_>). The arguments are similar or identical,
80	and they all accept an additional (and optional) C<$callback> argument	219	and they all accept an additional (and optional) C<$callback> argument
81	which must be a code reference. This code reference will get called with	220	which must be a code reference. This code reference will get called with
82	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	221	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
83	perl, which usually delivers "false") as it's sole argument when the given	222	perl, which usually delivers "false") as it's sole argument when the given
84	syscall has been executed asynchronously.	223	syscall has been executed asynchronously.
85		224
86	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.
87		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
88	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
89	for this is that at the time the request is being executed, the current	232	encoded as octets. The reason for the former is that at the time the
90	working directory could have changed. Alternatively, you can make sure	233	request is being executed, the current working directory could have
91	that you 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.
92		247
93	=over 4	248	=over 4
94		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
95	=item aio_open $pathname, $flags, $mode, $callback	281	=item aio_open $pathname, $flags, $mode, $callback->($fh)
96		282
97	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
98	created filehandle for the file.	284	created filehandle for the file.
99		285
100	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,
…		…
104	list. They are the same as used by C<sysopen>.	290	list. They are the same as used by C<sysopen>.
105		291
106	Likewise, C<$mode> specifies the mode of the newly created file, if it	292	Likewise, C<$mode> specifies the mode of the newly created file, if it
107	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,	293	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
108	except that it is mandatory (i.e. use C<0> if you don't create new files,	294	except that it is mandatory (i.e. use C<0> if you don't create new files,
109	and C<0666> or C<0777> if you do).	295	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		296	by the umask in effect then the request is being executed, so better never
		297	change the umask.
110		298
111	Example:	299	Example:
112		300
113	aio_open "/etc/passwd", O_RDONLY, 0, sub {	301	aio_open "/etc/passwd", O_RDONLY, 0, sub {
114	if ($_[0]) {	302	if ($_[0]) {
…		…
117	} else {	305	} else {
118	die "open failed: $!\n";	306	die "open failed: $!\n";
119	}	307	}
120	};	308	};
121		309
122	=item aio_close $fh, $callback	310	=item aio_close $fh, $callback->($status)
123		311
124	Asynchronously close a file and call the callback with the result	312	Asynchronously close a file and call the callback with the result
125	code. I<WARNING:> although accepted, you should not pass in a perl	313	code. I<WARNING:> although accepted, you should not pass in a perl
126	filehandle here, as perl will likely close the file descriptor another	314	filehandle here, as perl will likely close the file descriptor another
127	time when the filehandle is destroyed. Normally, you can safely call perls	315	time when the filehandle is destroyed. Normally, you can safely call perls
128	C<close> or just let filehandles go out of scope.	316	C<close> or just let filehandles go out of scope.
129		317
130	This is supposed to be a bug in the API, so that might change. It's	318	This is supposed to be a bug in the API, so that might change. It's
131	therefore best to avoid this function.	319	therefore best to avoid this function.
132		320
133	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	321	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
134		322
135	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	323	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
136		324
137	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	325	Reads or writes C<length> bytes from the specified C<fh> and C<offset>
138	into the scalar given by C<data> and offset C<dataoffset> and calls the	326	into the scalar given by C<data> and offset C<dataoffset> and calls the
139	callback without the actual number of bytes read (or -1 on error, just	327	callback without the actual number of bytes read (or -1 on error, just
140	like the syscall).	328	like the syscall).
141		329
		330	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		331	is outstanding. Modifying it can result in segfaults or WW3 (if the
		332	necessary/optional hardware is installed).
		333
142	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	334	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
143	offset C<0> within the scalar:	335	offset C<0> within the scalar:
144		336
145	aio_read $fh, 7, 15, $buffer, 0, sub {	337	aio_read $fh, 7, 15, $buffer, 0, sub {
146	$_[0] > 0 or die "read error: $!";	338	$_[0] > 0 or die "read error: $!";
147	print "read $_[0] bytes: <$buffer>\n";	339	print "read $_[0] bytes: <$buffer>\n";
148	};	340	};
149		341
		342	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		343
		344	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		345	reading at byte offset C<$in_offset>, and starts writing at the current
		346	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		347	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		348	other.
		349
		350	This call tries to make use of a native C<sendfile> syscall to provide
		351	zero-copy operation. For this to work, C<$out_fh> should refer to a
		352	socket, and C<$in_fh> should refer to mmap'able file.
		353
		354	If the native sendfile call fails or is not implemented, it will be
		355	emulated, so you can call C<aio_sendfile> on any type of filehandle
		356	regardless of the limitations of the operating system.
		357
		358	Please note, however, that C<aio_sendfile> can read more bytes from
		359	C<$in_fh> than are written, and there is no way to find out how many
		360	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only
		361	provides the number of bytes written to C<$out_fh>. Only if the result
		362	value equals C<$length> one can assume that C<$length> bytes have been
		363	read.
		364
150	=item aio_readahead $fh,$offset,$length, $callback	365	=item aio_readahead $fh,$offset,$length, $callback->($retval)
151
152	Asynchronously reads the specified byte range into the page cache, using
153	the C<readahead> syscall. If that syscall doesn't exist (likely if your OS
154	isn't Linux) the status will be C<-1> and C<$!> is set to C<ENOSYS>.
155		366
156	C<aio_readahead> populates the page cache with data from a file so that	367	C<aio_readahead> populates the page cache with data from a file so that
157	subsequent reads from that file will not block on disk I/O. The C<$offset>	368	subsequent reads from that file will not block on disk I/O. The C<$offset>
158	argument specifies the starting point from which data is to be read and	369	argument specifies the starting point from which data is to be read and
159	C<$length> specifies the number of bytes to be read. I/O is performed in	370	C<$length> specifies the number of bytes to be read. I/O is performed in
160	whole pages, so that offset is effectively rounded down to a page boundary	371	whole pages, so that offset is effectively rounded down to a page boundary
161	and bytes are read up to the next page boundary greater than or equal to	372	and bytes are read up to the next page boundary greater than or equal to
162	(off-set+length). C<aio_readahead> does not read beyond the end of the	373	(off-set+length). C<aio_readahead> does not read beyond the end of the
163	file. The current file offset of the file is left unchanged.	374	file. The current file offset of the file is left unchanged.
164		375
		376	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
		377	emulated by simply reading the data, which would have a similar effect.
		378
165	=item aio_stat $fh_or_path, $callback	379	=item aio_stat $fh_or_path, $callback->($status)
166		380
167	=item aio_lstat $fh, $callback	381	=item aio_lstat $fh, $callback->($status)
168		382
169	Works like perl's C<stat> or C<lstat> in void context. The callback will	383	Works like perl's C<stat> or C<lstat> in void context. The callback will
170	be called after the stat and the results will be available using C<stat _>	384	be called after the stat and the results will be available using C<stat _>
171	or C<-s _> etc...	385	or C<-s _> etc...
172		386
…		…
182	aio_stat "/etc/passwd", sub {	396	aio_stat "/etc/passwd", sub {
183	$_[0] and die "stat failed: $!";	397	$_[0] and die "stat failed: $!";
184	print "size is ", -s _, "\n";	398	print "size is ", -s _, "\n";
185	};	399	};
186		400
187	=item aio_unlink $pathname, $callback	401	=item aio_unlink $pathname, $callback->($status)
188		402
189	Asynchronously unlink (delete) a file and call the callback with the	403	Asynchronously unlink (delete) a file and call the callback with the
190	result code.	404	result code.
191		405
		406	=item aio_mknod $path, $mode, $dev, $callback->($status)
		407
		408	[EXPERIMENTAL]
		409
		410	Asynchronously create a device node (or fifo). See mknod(2).
		411
		412	The only (POSIX-) portable way of calling this function is:
		413
		414	aio_mknod $path, IO::AIO::S_IFIFO \| $mode, 0, sub { ...
		415
		416	=item aio_link $srcpath, $dstpath, $callback->($status)
		417
		418	Asynchronously create a new link to the existing object at C<$srcpath> at
		419	the path C<$dstpath> and call the callback with the result code.
		420
		421	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		422
		423	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		424	the path C<$dstpath> and call the callback with the result code.
		425
		426	=item aio_readlink $path, $callback->($link)
		427
		428	Asynchronously read the symlink specified by C<$path> and pass it to
		429	the callback. If an error occurs, nothing or undef gets passed to the
		430	callback.
		431
		432	=item aio_rename $srcpath, $dstpath, $callback->($status)
		433
		434	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		435	rename(2) and call the callback with the result code.
		436
		437	=item aio_mkdir $pathname, $mode, $callback->($status)
		438
		439	Asynchronously mkdir (create) a directory and call the callback with
		440	the result code. C<$mode> will be modified by the umask at the time the
		441	request is executed, so do not change your umask.
		442
		443	=item aio_rmdir $pathname, $callback->($status)
		444
		445	Asynchronously rmdir (delete) a directory and call the callback with the
		446	result code.
		447
		448	=item aio_readdir $pathname, $callback->($entries)
		449
		450	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		451	directory (i.e. opendir + readdir + closedir). The entries will not be
		452	sorted, and will B<NOT> include the C<.> and C<..> entries.
		453
		454	The callback a single argument which is either C<undef> or an array-ref
		455	with the filenames.
		456
		457	=item aio_load $path, $data, $callback->($status)
		458
		459	This is a composite request that tries to fully load the given file into
		460	memory. Status is the same as with aio_read.
		461
		462	=cut
		463
		464	sub aio_load($$;$) {
		465	aio_block {
		466	my ($path, undef, $cb) = @_;
		467	my $data = \$_[1];
		468
		469	my $pri = aioreq_pri;
		470	my $grp = aio_group $cb;
		471
		472	aioreq_pri $pri;
		473	add $grp aio_open $path, O_RDONLY, 0, sub {
		474	my $fh = shift
		475	or return $grp->result (-1);
		476
		477	aioreq_pri $pri;
		478	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		479	$grp->result ($_[0]);
		480	};
		481	};
		482
		483	$grp
		484	}
		485	}
		486
		487	=item aio_copy $srcpath, $dstpath, $callback->($status)
		488
		489	Try to copy the I<file> (directories not supported as either source or
		490	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		491	the C<0> (error) or C<-1> ok.
		492
		493	This is a composite request that it creates the destination file with
		494	mode 0200 and copies the contents of the source file into it using
		495	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		496	uid/gid, in that order.
		497
		498	If an error occurs, the partial destination file will be unlinked, if
		499	possible, except when setting atime, mtime, access mode and uid/gid, where
		500	errors are being ignored.
		501
		502	=cut
		503
		504	sub aio_copy($$;$) {
		505	aio_block {
		506	my ($src, $dst, $cb) = @_;
		507
		508	my $pri = aioreq_pri;
		509	my $grp = aio_group $cb;
		510
		511	aioreq_pri $pri;
		512	add $grp aio_open $src, O_RDONLY, 0, sub {
		513	if (my $src_fh = $_[0]) {
		514	my @stat = stat $src_fh;
		515
		516	aioreq_pri $pri;
		517	add $grp aio_open $dst, O_CREAT \| O_WRONLY \| O_TRUNC, 0200, sub {
		518	if (my $dst_fh = $_[0]) {
		519	aioreq_pri $pri;
		520	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		521	if ($_[0] == $stat[7]) {
		522	$grp->result (0);
		523	close $src_fh;
		524
		525	# those should not normally block. should. should.
		526	utime $stat[8], $stat[9], $dst;
		527	chmod $stat[2] & 07777, $dst_fh;
		528	chown $stat[4], $stat[5], $dst_fh;
		529	close $dst_fh;
		530	} else {
		531	$grp->result (-1);
		532	close $src_fh;
		533	close $dst_fh;
		534
		535	aioreq $pri;
		536	add $grp aio_unlink $dst;
		537	}
		538	};
		539	} else {
		540	$grp->result (-1);
		541	}
		542	},
		543
		544	} else {
		545	$grp->result (-1);
		546	}
		547	};
		548
		549	$grp
		550	}
		551	}
		552
		553	=item aio_move $srcpath, $dstpath, $callback->($status)
		554
		555	Try to move the I<file> (directories not supported as either source or
		556	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		557	the C<0> (error) or C<-1> ok.
		558
		559	This is a composite request that tries to rename(2) the file first. If
		560	rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if
		561	that is successful, unlinking the C<$srcpath>.
		562
		563	=cut
		564
		565	sub aio_move($$;$) {
		566	aio_block {
		567	my ($src, $dst, $cb) = @_;
		568
		569	my $pri = aioreq_pri;
		570	my $grp = aio_group $cb;
		571
		572	aioreq_pri $pri;
		573	add $grp aio_rename $src, $dst, sub {
		574	if ($_[0] && $! == EXDEV) {
		575	aioreq_pri $pri;
		576	add $grp aio_copy $src, $dst, sub {
		577	$grp->result ($_[0]);
		578
		579	if (!$_[0]) {
		580	aioreq_pri $pri;
		581	add $grp aio_unlink $src;
		582	}
		583	};
		584	} else {
		585	$grp->result ($_[0]);
		586	}
		587	};
		588
		589	$grp
		590	}
		591	}
		592
		593	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		594
		595	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		596	efficiently separate the entries of directory C<$path> into two sets of
		597	names, directories you can recurse into (directories), and ones you cannot
		598	recurse into (everything else, including symlinks to directories).
		599
		600	C<aio_scandir> is a composite request that creates of many sub requests_
		601	C<$maxreq> specifies the maximum number of outstanding aio requests that
		602	this function generates. If it is C<< <= 0 >>, then a suitable default
		603	will be chosen (currently 4).
		604
		605	On error, the callback is called without arguments, otherwise it receives
		606	two array-refs with path-relative entry names.
		607
		608	Example:
		609
		610	aio_scandir $dir, 0, sub {
		611	my ($dirs, $nondirs) = @_;
		612	print "real directories: @$dirs\n";
		613	print "everything else: @$nondirs\n";
		614	};
		615
		616	Implementation notes.
		617
		618	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		619
		620	After reading the directory, the modification time, size etc. of the
		621	directory before and after the readdir is checked, and if they match (and
		622	isn't the current time), the link count will be used to decide how many
		623	entries are directories (if >= 2). Otherwise, no knowledge of the number
		624	of subdirectories will be assumed.
		625
		626	Then entries will be sorted into likely directories (everything without
		627	a non-initial dot currently) and likely non-directories (everything
		628	else). Then every entry plus an appended C</.> will be C<stat>'ed,
		629	likely directories first. If that succeeds, it assumes that the entry
		630	is a directory or a symlink to directory (which will be checked
		631	seperately). This is often faster than stat'ing the entry itself because
		632	filesystems might detect the type of the entry without reading the inode
		633	data (e.g. ext2fs filetype feature).
		634
		635	If the known number of directories (link count - 2) has been reached, the
		636	rest of the entries is assumed to be non-directories.
		637
		638	This only works with certainty on POSIX (= UNIX) filesystems, which
		639	fortunately are the vast majority of filesystems around.
		640
		641	It will also likely work on non-POSIX filesystems with reduced efficiency
		642	as those tend to return 0 or 1 as link counts, which disables the
		643	directory counting heuristic.
		644
		645	=cut
		646
		647	sub aio_scandir($$;$) {
		648	aio_block {
		649	my ($path, $maxreq, $cb) = @_;
		650
		651	my $pri = aioreq_pri;
		652
		653	my $grp = aio_group $cb;
		654
		655	$maxreq = 4 if $maxreq <= 0;
		656
		657	# stat once
		658	aioreq_pri $pri;
		659	add $grp aio_stat $path, sub {
		660	return $grp->result () if $_[0];
		661	my $now = time;
		662	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		663
		664	# read the directory entries
		665	aioreq_pri $pri;
		666	add $grp aio_readdir $path, sub {
		667	my $entries = shift
		668	or return $grp->result ();
		669
		670	# stat the dir another time
		671	aioreq_pri $pri;
		672	add $grp aio_stat $path, sub {
		673	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		674
		675	my $ndirs;
		676
		677	# take the slow route if anything looks fishy
		678	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		679	$ndirs = -1;
		680	} else {
		681	# if nlink == 2, we are finished
		682	# on non-posix-fs's, we rely on nlink < 2
		683	$ndirs = (stat _)[3] - 2
		684	or return $grp->result ([], $entries);
		685	}
		686
		687	# sort into likely dirs and likely nondirs
		688	# dirs == files without ".", short entries first
		689	$entries = [map $_->[0],
		690	sort { $b->[1] cmp $a->[1] }
		691	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
		692	@$entries];
		693
		694	my (@dirs, @nondirs);
		695
		696	my $statgrp = add $grp aio_group sub {
		697	$grp->result (\@dirs, \@nondirs);
		698	};
		699
		700	limit $statgrp $maxreq;
		701	feed $statgrp sub {
		702	return unless @$entries;
		703	my $entry = pop @$entries;
		704
		705	aioreq_pri $pri;
		706	add $statgrp aio_stat "$path/$entry/.", sub {
		707	if ($_[0] < 0) {
		708	push @nondirs, $entry;
		709	} else {
		710	# need to check for real directory
		711	aioreq_pri $pri;
		712	add $statgrp aio_lstat "$path/$entry", sub {
		713	if (-d _) {
		714	push @dirs, $entry;
		715
		716	unless (--$ndirs) {
		717	push @nondirs, @$entries;
		718	feed $statgrp;
		719	}
		720	} else {
		721	push @nondirs, $entry;
		722	}
		723	}
		724	}
		725	};
		726	};
		727	};
		728	};
		729	};
		730
		731	$grp
		732	}
		733	}
		734
		735	=item aio_rmtree $path, $callback->($status)
		736
		737	Delete a directory tree starting (and including) C<$path>, return the
		738	status of the final C<rmdir> only. This is a composite request that
		739	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		740	everything else.
		741
		742	=cut
		743
		744	sub aio_rmtree;
		745	sub aio_rmtree($;$) {
		746	aio_block {
		747	my ($path, $cb) = @_;
		748
		749	my $pri = aioreq_pri;
		750	my $grp = aio_group $cb;
		751
		752	aioreq_pri $pri;
		753	add $grp aio_scandir $path, 0, sub {
		754	my ($dirs, $nondirs) = @_;
		755
		756	my $dirgrp = aio_group sub {
		757	add $grp aio_rmdir $path, sub {
		758	$grp->result ($_[0]);
		759	};
		760	};
		761
		762	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		763	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		764
		765	add $grp $dirgrp;
		766	};
		767
		768	$grp
		769	}
		770	}
		771
192	=item aio_fsync $fh, $callback	772	=item aio_fsync $fh, $callback->($status)
193		773
194	Asynchronously call fsync on the given filehandle and call the callback	774	Asynchronously call fsync on the given filehandle and call the callback
195	with the fsync result code.	775	with the fsync result code.
196		776
197	=item aio_fdatasync $fh, $callback	777	=item aio_fdatasync $fh, $callback->($status)
198		778
199	Asynchronously call fdatasync on the given filehandle and call the	779	Asynchronously call fdatasync on the given filehandle and call the
200	callback with the fdatasync result code. Might set C<$!> to C<ENOSYS> if	780	callback with the fdatasync result code.
201	C<fdatasync> is not available.	781
		782	If this call isn't available because your OS lacks it or it couldn't be
		783	detected, it will be emulated by calling C<fsync> instead.
		784
		785	=item aio_group $callback->(...)
		786
		787	This is a very special aio request: Instead of doing something, it is a
		788	container for other aio requests, which is useful if you want to bundle
		789	many requests into a single, composite, request with a definite callback
		790	and the ability to cancel the whole request with its subrequests.
		791
		792	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		793	for more info.
		794
		795	Example:
		796
		797	my $grp = aio_group sub {
		798	print "all stats done\n";
		799	};
		800
		801	add $grp
		802	(aio_stat ...),
		803	(aio_stat ...),
		804	...;
		805
		806	=item aio_nop $callback->()
		807
		808	This is a special request - it does nothing in itself and is only used for
		809	side effects, such as when you want to add a dummy request to a group so
		810	that finishing the requests in the group depends on executing the given
		811	code.
		812
		813	While this request does nothing, it still goes through the execution
		814	phase and still requires a worker thread. Thus, the callback will not
		815	be executed immediately but only after other requests in the queue have
		816	entered their execution phase. This can be used to measure request
		817	latency.
		818
		819	=item IO::AIO::aio_busy $fractional_seconds, $callback->() NOT EXPORTED
		820
		821	Mainly used for debugging and benchmarking, this aio request puts one of
		822	the request workers to sleep for the given time.
		823
		824	While it is theoretically handy to have simple I/O scheduling requests
		825	like sleep and file handle readable/writable, the overhead this creates is
		826	immense (it blocks a thread for a long time) so do not use this function
		827	except to put your application under artificial I/O pressure.
202		828
203	=back	829	=back
204		830
		831	=head2 IO::AIO::REQ CLASS
		832
		833	All non-aggregate C<aio_*> functions return an object of this class when
		834	called in non-void context.
		835
		836	=over 4
		837
		838	=item cancel $req
		839
		840	Cancels the request, if possible. Has the effect of skipping execution
		841	when entering the B<execute> state and skipping calling the callback when
		842	entering the the B<result> state, but will leave the request otherwise
		843	untouched. That means that requests that currently execute will not be
		844	stopped and resources held by the request will not be freed prematurely.
		845
		846	=item cb $req $callback->(...)
		847
		848	Replace (or simply set) the callback registered to the request.
		849
		850	=back
		851
		852	=head2 IO::AIO::GRP CLASS
		853
		854	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		855	objects of this class, too.
		856
		857	A IO::AIO::GRP object is a special request that can contain multiple other
		858	aio requests.
		859
		860	You create one by calling the C<aio_group> constructing function with a
		861	callback that will be called when all contained requests have entered the
		862	C<done> state:
		863
		864	my $grp = aio_group sub {
		865	print "all requests are done\n";
		866	};
		867
		868	You add requests by calling the C<add> method with one or more
		869	C<IO::AIO::REQ> objects:
		870
		871	$grp->add (aio_unlink "...");
		872
		873	add $grp aio_stat "...", sub {
		874	$_[0] or return $grp->result ("error");
		875
		876	# add another request dynamically, if first succeeded
		877	add $grp aio_open "...", sub {
		878	$grp->result ("ok");
		879	};
		880	};
		881
		882	This makes it very easy to create composite requests (see the source of
		883	C<aio_move> for an application) that work and feel like simple requests.
		884
		885	=over 4
		886
		887	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		888	C<IO::AIO::poll_cb>, just like any other request.
		889
		890	=item * They can be canceled like any other request. Canceling will cancel not
		891	only the request itself, but also all requests it contains.
		892
		893	=item * They can also can also be added to other IO::AIO::GRP objects.
		894
		895	=item * You must not add requests to a group from within the group callback (or
		896	any later time).
		897
		898	=back
		899
		900	Their lifetime, simplified, looks like this: when they are empty, they
		901	will finish very quickly. If they contain only requests that are in the
		902	C<done> state, they will also finish. Otherwise they will continue to
		903	exist.
		904
		905	That means after creating a group you have some time to add requests. And
		906	in the callbacks of those requests, you can add further requests to the
		907	group. And only when all those requests have finished will the the group
		908	itself finish.
		909
		910	=over 4
		911
		912	=item add $grp ...
		913
		914	=item $grp->add (...)
		915
		916	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		917	be added, including other groups, as long as you do not create circular
		918	dependencies.
		919
		920	Returns all its arguments.
		921
		922	=item $grp->cancel_subs
		923
		924	Cancel all subrequests and clears any feeder, but not the group request
		925	itself. Useful when you queued a lot of events but got a result early.
		926
		927	=item $grp->result (...)
		928
		929	Set the result value(s) that will be passed to the group callback when all
		930	subrequests have finished and set thre groups errno to the current value
		931	of errno (just like calling C<errno> without an error number). By default,
		932	no argument will be passed and errno is zero.
		933
		934	=item $grp->errno ([$errno])
		935
		936	Sets the group errno value to C<$errno>, or the current value of errno
		937	when the argument is missing.
		938
		939	Every aio request has an associated errno value that is restored when
		940	the callback is invoked. This method lets you change this value from its
		941	default (0).
		942
		943	Calling C<result> will also set errno, so make sure you either set C<$!>
		944	before the call to C<result>, or call c<errno> after it.
		945
		946	=item feed $grp $callback->($grp)
		947
		948	Sets a feeder/generator on this group: every group can have an attached
		949	generator that generates requests if idle. The idea behind this is that,
		950	although you could just queue as many requests as you want in a group,
		951	this might starve other requests for a potentially long time. For
		952	example, C<aio_scandir> might generate hundreds of thousands C<aio_stat>
		953	requests, delaying any later requests for a long time.
		954
		955	To avoid this, and allow incremental generation of requests, you can
		956	instead a group and set a feeder on it that generates those requests. The
		957	feed callback will be called whenever there are few enough (see C<limit>,
		958	below) requests active in the group itself and is expected to queue more
		959	requests.
		960
		961	The feed callback can queue as many requests as it likes (i.e. C<add> does
		962	not impose any limits).
		963
		964	If the feed does not queue more requests when called, it will be
		965	automatically removed from the group.
		966
		967	If the feed limit is C<0>, it will be set to C<2> automatically.
		968
		969	Example:
		970
		971	# stat all files in @files, but only ever use four aio requests concurrently:
		972
		973	my $grp = aio_group sub { print "finished\n" };
		974	limit $grp 4;
		975	feed $grp sub {
		976	my $file = pop @files
		977	or return;
		978
		979	add $grp aio_stat $file, sub { ... };
		980	};
		981
		982	=item limit $grp $num
		983
		984	Sets the feeder limit for the group: The feeder will be called whenever
		985	the group contains less than this many requests.
		986
		987	Setting the limit to C<0> will pause the feeding process.
		988
		989	=back
		990
205	=head2 SUPPORT FUNCTIONS	991	=head2 SUPPORT FUNCTIONS
		992
		993	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
206		994
207	=over 4	995	=over 4
208		996
209	=item $fileno = IO::AIO::poll_fileno	997	=item $fileno = IO::AIO::poll_fileno
210		998
…		…
215		1003
216	See C<poll_cb> for an example.	1004	See C<poll_cb> for an example.
217		1005
218	=item IO::AIO::poll_cb	1006	=item IO::AIO::poll_cb
219		1007
220	Process all outstanding events on the result pipe. You have to call this	1008	Process some outstanding events on the result pipe. You have to call this
221	regularly. Returns the number of events processed. Returns immediately	1009	regularly. Returns the number of events processed. Returns immediately
222	when no events are outstanding.	1010	when no events are outstanding. The amount of events processed depends on
		1011	the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
		1012
		1013	If not all requests were processed for whatever reason, the filehandle
		1014	will still be ready when C<poll_cb> returns.
223		1015
224	Example: Install an Event watcher that automatically calls	1016	Example: Install an Event watcher that automatically calls
225	IO::AIO::poll_cb with high priority:	1017	IO::AIO::poll_cb with high priority:
226		1018
227	Event->io (fd => IO::AIO::poll_fileno,	1019	Event->io (fd => IO::AIO::poll_fileno,
228	poll => 'r', async => 1,	1020	poll => 'r', async => 1,
229	cb => \&IO::AIO::poll_cb);	1021	cb => \&IO::AIO::poll_cb);
230		1022
		1023	=item IO::AIO::max_poll_reqs $nreqs
		1024
		1025	=item IO::AIO::max_poll_time $seconds
		1026
		1027	These set the maximum number of requests (default C<0>, meaning infinity)
		1028	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1029	the maximum amount of time (default C<0>, meaning infinity) spent in
		1030	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1031	of time C<poll_cb> is allowed to use).
		1032
		1033	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1034	syscall per request processed, which is not normally a problem unless your
		1035	callbacks are really really fast or your OS is really really slow (I am
		1036	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1037
		1038	Setting these is useful if you want to ensure some level of
		1039	interactiveness when perl is not fast enough to process all requests in
		1040	time.
		1041
		1042	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		1043
		1044	Example: Install an Event watcher that automatically calls
		1045	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1046	program get the CPU sometimes even under high AIO load.
		1047
		1048	# try not to spend much more than 0.1s in poll_cb
		1049	IO::AIO::max_poll_time 0.1;
		1050
		1051	# use a low priority so other tasks have priority
		1052	Event->io (fd => IO::AIO::poll_fileno,
		1053	poll => 'r', nice => 1,
		1054	cb => &IO::AIO::poll_cb);
		1055
231	=item IO::AIO::poll_wait	1056	=item IO::AIO::poll_wait
232		1057
		1058	If there are any outstanding requests and none of them in the result
233	Wait till the result filehandle becomes ready for reading (simply does a	1059	phase, wait till the result filehandle becomes ready for reading (simply
234	C<select> on the filehandle. This is useful if you want to synchronously wait	1060	does a C<select> on the filehandle. This is useful if you want to
235	for some requests to finish).	1061	synchronously wait for some requests to finish).
236		1062
237	See C<nreqs> for an example.	1063	See C<nreqs> for an example.
238		1064
		1065	=item IO::AIO::poll
		1066
		1067	Waits until some requests have been handled.
		1068
		1069	Returns the number of requests processed, but is otherwise strictly
		1070	equivalent to:
		1071
		1072	IO::AIO::poll_wait, IO::AIO::poll_cb
		1073
239	=item IO::AIO::nreqs	1074	=item IO::AIO::flush
240		1075
241	Returns the number of requests currently outstanding (i.e. for which their	1076	Wait till all outstanding AIO requests have been handled.
242	callback has not been invoked yet).
243		1077
244	Example: wait till there are no outstanding requests anymore:	1078	Strictly equivalent to:
245		1079
246	IO::AIO::poll_wait, IO::AIO::poll_cb	1080	IO::AIO::poll_wait, IO::AIO::poll_cb
247	while IO::AIO::nreqs;	1081	while IO::AIO::nreqs;
248		1082
		1083	=head3 CONTROLLING THE NUMBER OF THREADS
		1084
		1085	=item IO::AIO::min_parallel $nthreads
		1086
		1087	Set the minimum number of AIO threads to C<$nthreads>. The current
		1088	default is C<8>, which means eight asynchronous operations can execute
		1089	concurrently at any one time (the number of outstanding requests,
		1090	however, is unlimited).
		1091
		1092	IO::AIO starts threads only on demand, when an AIO request is queued and
		1093	no free thread exists. Please note that queueing up a hundred requests can
		1094	create demand for a hundred threads, even if it turns out that everything
		1095	is in the cache and could have been processed faster by a single thread.
		1096
		1097	It is recommended to keep the number of threads relatively low, as some
		1098	Linux kernel versions will scale negatively with the number of threads
		1099	(higher parallelity => MUCH higher latency). With current Linux 2.6
		1100	versions, 4-32 threads should be fine.
		1101
		1102	Under most circumstances you don't need to call this function, as the
		1103	module selects a default that is suitable for low to moderate load.
		1104
		1105	=item IO::AIO::max_parallel $nthreads
		1106
		1107	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
		1108	specified number of threads are currently running, this function kills
		1109	them. This function blocks until the limit is reached.
		1110
		1111	While C<$nthreads> are zero, aio requests get queued but not executed
		1112	until the number of threads has been increased again.
		1113
		1114	This module automatically runs C<max_parallel 0> at program end, to ensure
		1115	that all threads are killed and that there are no outstanding requests.
		1116
		1117	Under normal circumstances you don't need to call this function.
		1118
		1119	=item IO::AIO::max_idle $nthreads
		1120
		1121	Limit the number of threads (default: 4) that are allowed to idle (i.e.,
		1122	threads that did not get a request to process within 10 seconds). That
		1123	means if a thread becomes idle while C<$nthreads> other threads are also
		1124	idle, it will free its resources and exit.
		1125
		1126	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1127	to allow for extremely high load situations, but want to free resources
		1128	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1129
		1130	The default is probably ok in most situations, especially if thread
		1131	creation is fast. If thread creation is very slow on your system you might
		1132	want to use larger values.
		1133
		1134	=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs
		1135
		1136	This is a very bad function to use in interactive programs because it
		1137	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1138	use an C<aio_group> together with a feed callback.
		1139
		1140	Sets the maximum number of outstanding requests to C<$nreqs>. If you
		1141	to queue up more than this number of requests, the next call to the
		1142	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
		1143	function will block until the limit is no longer exceeded.
		1144
		1145	The default value is very large, so there is no practical limit on the
		1146	number of outstanding requests.
		1147
		1148	You can still queue as many requests as you want. Therefore,
		1149	C<max_oustsanding> is mainly useful in simple scripts (with low values) or
		1150	as a stop gap to shield against fatal memory overflow (with large values).
		1151
		1152	=head3 STATISTICAL INFORMATION
		1153
249	=item IO::AIO::flush	1154	=item IO::AIO::nreqs
250		1155
251	Wait till all outstanding AIO requests have been handled.	1156	Returns the number of requests currently in the ready, execute or pending
		1157	states (i.e. for which their callback has not been invoked yet).
252		1158
253	Strictly equivalent to:	1159	Example: wait till there are no outstanding requests anymore:
254		1160
255	IO::AIO::poll_wait, IO::AIO::poll_cb	1161	IO::AIO::poll_wait, IO::AIO::poll_cb
256	while IO::AIO::nreqs;	1162	while IO::AIO::nreqs;
257		1163
		1164	=item IO::AIO::nready
		1165
		1166	Returns the number of requests currently in the ready state (not yet
		1167	executed).
		1168
258	=item IO::AIO::poll	1169	=item IO::AIO::npending
259		1170
260	Waits until some requests have been handled.	1171	Returns the number of requests currently in the pending state (executed,
261		1172	but not yet processed by poll_cb).
262	Strictly equivalent to:
263
264	IO::AIO::poll_wait, IO::AIO::poll_cb
265	if IO::AIO::nreqs;
266
267	=item IO::AIO::min_parallel $nthreads
268
269	Set the minimum number of AIO threads to C<$nthreads>. The default is
270	C<1>, which means a single asynchronous operation can be done at one time
271	(the number of outstanding operations, however, is unlimited).
272
273	It is recommended to keep the number of threads low, as some Linux
274	kernel versions will scale negatively with the number of threads (higher
275	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32
276	threads should be fine.
277
278	Under normal circumstances you don't need to call this function, as this
279	module automatically starts some threads (the exact number might change,
280	and is currently 4).
281
282	=item IO::AIO::max_parallel $nthreads
283
284	Sets the maximum number of AIO threads to C<$nthreads>. If more than
285	the specified number of threads are currently running, kill them. This
286	function blocks until the limit is reached.
287
288	This module automatically runs C<max_parallel 0> at program end, to ensure
289	that all threads are killed and that there are no outstanding requests.
290
291	Under normal circumstances you don't need to call this function.
292
293	=item $oldnreqs = IO::AIO::max_outstanding $nreqs
294
295	Sets the maximum number of outstanding requests to C<$nreqs>. If you
296	try to queue up more than this number of requests, the caller will block until
297	some requests have been handled.
298
299	The default is very large, so normally there is no practical limit. If you
300	queue up many requests in a loop it it often improves speed if you set
301	this to a relatively low number, such as C<100>.
302
303	Under normal circumstances you don't need to call this function.
304		1173
305	=back	1174	=back
306		1175
307	=cut	1176	=cut
308		1177
309	# support function to convert a fd into a perl filehandle	1178	# support function to convert a fd into a perl filehandle
310	sub _fd2fh {	1179	sub _fd2fh {
311	return undef if $_[0] < 0;	1180	return undef if $_[0] < 0;
312		1181
313	# try to be perl5.6-compatible	1182	# try to generate nice filehandles
314	local *AIO_FH;	1183	my $sym = "IO::AIO::fd#$_[0]";
315	open AIO_FH, "+<&=$_[0]"	1184	local *$sym;
		1185
		1186	open *$sym, "+<&=$_[0]" # usually works under any unix
		1187	or open *$sym, "<&=$_[0]" # cygwin needs this
		1188	or open *$sym, ">&=$_[0]" # or this
316	or return undef;	1189	or return undef;
317		1190
318	*AIO_FH	1191	*$sym
319	}	1192	}
320		1193
321	min_parallel 4;	1194	min_parallel 8;
322		1195
323	END {	1196	END { flush }
324	max_parallel 0;
325	}
326		1197
327	1;	1198	1;
328		1199
		1200	=head2 FORK BEHAVIOUR
		1201
		1202	This module should do "the right thing" when the process using it forks:
		1203
		1204	Before the fork, IO::AIO enters a quiescent state where no requests
		1205	can be added in other threads and no results will be processed. After
		1206	the fork the parent simply leaves the quiescent state and continues
		1207	request/result processing, while the child frees the request/result queue
		1208	(so that the requests started before the fork will only be handled in the
		1209	parent). Threads will be started on demand until the limit set in the
		1210	parent process has been reached again.
		1211
		1212	In short: the parent will, after a short pause, continue as if fork had
		1213	not been called, while the child will act as if IO::AIO has not been used
		1214	yet.
		1215
		1216	=head2 MEMORY USAGE
		1217
		1218	Per-request usage:
		1219
		1220	Each aio request uses - depending on your architecture - around 100-200
		1221	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1222	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1223	scalars and other data passed into aio requests will also be locked and
		1224	will consume memory till the request has entered the done state.
		1225
		1226	This is now awfully much, so queuing lots of requests is not usually a
		1227	problem.
		1228
		1229	Per-thread usage:
		1230
		1231	In the execution phase, some aio requests require more memory for
		1232	temporary buffers, and each thread requires a stack and other data
		1233	structures (usually around 16k-128k, depending on the OS).
		1234
		1235	=head1 KNOWN BUGS
		1236
		1237	Known bugs will be fixed in the next release.
		1238
329	=head1 SEE ALSO	1239	=head1 SEE ALSO
330		1240
331	L<Coro>, L<Linux::AIO>.	1241	L<Coro::AIO>.
332		1242
333	=head1 AUTHOR	1243	=head1 AUTHOR
334		1244
335	Marc Lehmann <schmorp@schmorp.de>	1245	Marc Lehmann <schmorp@schmorp.de>
336	http://home.schmorp.de/	1246	http://home.schmorp.de/

Diff Legend

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

Comparing IO-AIO/AIO.pm (file contents): Revision 1.21 by root, Wed Jul 13 00:13:09 2005 UTC vs. Revision 1.103 by root, Tue Jan 23 22:57:34 2007 UTC

Diff Legend

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.21 by root, Wed Jul 13 00:13:09 2005 UTC vs.
Revision 1.103 by root, Tue Jan 23 22:57:34 2007 UTC