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

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.12 by root, Mon Jul 11 01:03:17 2005 UTC vs.
Revision 1.108 by root, Fri Jun 1 13:25:50 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	\&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
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
		68	example sockets), using these functions on file descriptors that
		69	support nonblocking operation (again, sockets, pipes etc.) is very
		70	inefficient. Use an event loop for that (such as the L<Event\|Event>
		71	module): IO::AIO will naturally fit into such an event loop itself.
		72
43	Currently, a number of threads are started that execute your read/writes	73	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	74	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	75	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	76	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	77	functions available on many operating systems. However, they are often
48	not well-supported (Linux doesn't allow them on normal files currently,	78	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	79	files currently, for example), and they would only support aio_read and
50	remaining functionality would have to be implemented using threads anyway.	80	aio_write, so the remaining functionality would have to be implemented
		81	using threads anyway.
51		82
52	Although the module will work with in the presence of other threads, it is	83	Although the module will work in the presence of other (Perl-) threads,
53	currently not reentrant, so use appropriate locking yourself.	84	it is currently not reentrant in any way, so use appropriate locking
		85	yourself, always call C<poll_cb> from within the same thread, or never
		86	call C<poll_cb> (or other C<aio_> functions) recursively.
		87
		88	=head2 EXAMPLE
		89
		90	This is a simple example that uses the Event module and loads
		91	F</etc/passwd> asynchronously:
		92
		93	use Fcntl;
		94	use Event;
		95	use IO::AIO;
		96
		97	# register the IO::AIO callback with Event
		98	Event->io (fd => IO::AIO::poll_fileno,
		99	poll => 'r',
		100	cb => \&IO::AIO::poll_cb);
		101
		102	# queue the request to open /etc/passwd
		103	aio_open "/etc/passwd", O_RDONLY, 0, sub {
		104	my $fh = shift
		105	or die "error while opening: $!";
		106
		107	# stat'ing filehandles is generally non-blocking
		108	my $size = -s $fh;
		109
		110	# queue a request to read the file
		111	my $contents;
		112	aio_read $fh, 0, $size, $contents, 0, sub {
		113	$_[0] == $size
		114	or die "short read: $!";
		115
		116	close $fh;
		117
		118	# file contents now in $contents
		119	print $contents;
		120
		121	# exit event loop and program
		122	Event::unloop;
		123	};
		124	};
		125
		126	# possibly queue up other requests, or open GUI windows,
		127	# check for sockets etc. etc.
		128
		129	# process events as long as there are some:
		130	Event::loop;
		131
		132	=head1 REQUEST ANATOMY AND LIFETIME
		133
		134	Every C<aio_*> function creates a request. which is a C data structure not
		135	directly visible to Perl.
		136
		137	If called in non-void context, every request function returns a Perl
		138	object representing the request. In void context, nothing is returned,
		139	which saves a bit of memory.
		140
		141	The perl object is a fairly standard ref-to-hash object. The hash contents
		142	are not used by IO::AIO so you are free to store anything you like in it.
		143
		144	During their existance, aio requests travel through the following states,
		145	in order:
		146
		147	=over 4
		148
		149	=item ready
		150
		151	Immediately after a request is created it is put into the ready state,
		152	waiting for a thread to execute it.
		153
		154	=item execute
		155
		156	A thread has accepted the request for processing and is currently
		157	executing it (e.g. blocking in read).
		158
		159	=item pending
		160
		161	The request has been executed and is waiting for result processing.
		162
		163	While request submission and execution is fully asynchronous, result
		164	processing is not and relies on the perl interpreter calling C<poll_cb>
		165	(or another function with the same effect).
		166
		167	=item result
		168
		169	The request results are processed synchronously by C<poll_cb>.
		170
		171	The C<poll_cb> function will process all outstanding aio requests by
		172	calling their callbacks, freeing memory associated with them and managing
		173	any groups they are contained in.
		174
		175	=item done
		176
		177	Request has reached the end of its lifetime and holds no resources anymore
		178	(except possibly for the Perl object, but its connection to the actual
		179	aio request is severed and calling its methods will either do nothing or
		180	result in a runtime error).
		181
		182	=back
54		183
55	=cut	184	=cut
56		185
57	package IO::AIO;	186	package IO::AIO;
58		187
		188	no warnings;
		189	use strict 'vars';
		190
59	use base 'Exporter';	191	use base 'Exporter';
60		192
61	use Fcntl ();
62
63	BEGIN {	193	BEGIN {
64	$VERSION = 0.3;	194	our $VERSION = '2.4';
65		195
66	@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink	196	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat
67	aio_fsync aio_fdatasync aio_readahead);	197	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);	198	aio_readlink aio_fsync aio_fdatasync aio_readahead aio_rename aio_link
		199	aio_move aio_copy aio_group aio_nop aio_mknod aio_load aio_rmtree aio_mkdir
		200	aio_chown aio_chmod aio_utime);
		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", $VERSION);
72	}	211	}
73		212
74	=head1 FUNCTIONS	213	=head1 FUNCTIONS
75		214
76	=head2 AIO FUNCTIONS	215	=head2 AIO REQUEST FUNCTIONS
77		216
78	All the C<aio_*> calls are more or less thin wrappers around the syscall	217	All the C<aio_*> calls are more or less thin wrappers around the syscall
79	with the same name (sans C<aio_>). The arguments are similar or identical,	218	with the same name (sans C<aio_>). The arguments are similar or identical,
80	and they all accept an additional C<$callback> argument which must be	219	and they all accept an additional (and optional) C<$callback> argument
81	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
82	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
83	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
84	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	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
90	directory could have changed. Alternatively, you can make sure that you	233	request is being executed, the current working directory could have
91	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
		277	=item aioreq_nice $pri_adjust
		278
		279	Similar to C<aioreq_pri>, but subtracts the given value from the current
		280	priority, so the effect is cumulative.
		281
		282
95	=item aio_open $pathname, $flags, $mode, $callback	283	=item aio_open $pathname, $flags, $mode, $callback->($fh)
96		284
97	Asynchronously open or create a file and call the callback with a newly	285	Asynchronously open or create a file and call the callback with a newly
98	created filehandle for the file.	286	created filehandle for the file.
99		287
100	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,	288	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
101	for an explanation.	289	for an explanation.
102		290
103	The C<$mode> argument is a bitmask. See the C<Fcntl> module for a	291	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
104	list. They are the same as used in C<sysopen>.	292	list. They are the same as used by C<sysopen>.
		293
		294	Likewise, C<$mode> specifies the mode of the newly created file, if it
		295	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
		296	except that it is mandatory (i.e. use C<0> if you don't create new files,
		297	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		298	by the umask in effect then the request is being executed, so better never
		299	change the umask.
105		300
106	Example:	301	Example:
107		302
108	aio_open "/etc/passwd", O_RDONLY, 0, sub {	303	aio_open "/etc/passwd", O_RDONLY, 0, sub {
109	if ($_[0]) {	304	if ($_[0]) {
…		…
112	} else {	307	} else {
113	die "open failed: $!\n";	308	die "open failed: $!\n";
114	}	309	}
115	};	310	};
116		311
		312
117	=item aio_close $fh, $callback	313	=item aio_close $fh, $callback->($status)
118		314
119	Asynchronously close a file and call the callback with the result	315	Asynchronously close a file and call the callback with the result
120	code. I<WARNING:> although accepted, you should not pass in a perl	316	code. I<WARNING:> although accepted, you should not pass in a perl
121	filehandle here, as perl will likely close the file descriptor itself when	317	filehandle here, as perl will likely close the file descriptor another
122	the filehandle is destroyed. Normally, you can safely call perls C<close>	318	time when the filehandle is destroyed. Normally, you can safely call perls
123	or just let filehandles go out of scope.	319	C<close> or just let filehandles go out of scope.
124		320
		321	This is supposed to be a bug in the API, so that might change. It's
		322	therefore best to avoid this function.
		323
		324
125	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	325	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
126		326
127	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	327	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
128		328
129	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	329	Reads or writes C<length> bytes from the specified C<fh> and C<offset>
130	into the scalar given by C<data> and offset C<dataoffset> and calls the	330	into the scalar given by C<data> and offset C<dataoffset> and calls the
131	callback without the actual number of bytes read (or -1 on error, just	331	callback without the actual number of bytes read (or -1 on error, just
132	like the syscall).	332	like the syscall). If C<offset> is undefined, then the current file offset
		333	will be used (and updated), otherwise the file offset will not be changed
		334	by these calls.
133		335
		336	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		337	is outstanding. Modifying it can result in segfaults or World War III (if
		338	the necessary/optional hardware is installed).
		339
134	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at	340	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
135	offset C<0> within the scalar:	341	offset C<0> within the scalar:
136		342
137	aio_read $fh, 7, 15, $buffer, 0, sub {	343	aio_read $fh, 7, 15, $buffer, 0, sub {
138	$_[0] > 0 or die "read error: $!";	344	$_[0] > 0 or die "read error: $!";
139	print "read $_[0] bytes: <$buffer>\n";	345	print "read $_[0] bytes: <$buffer>\n";
140	};	346	};
141		347
		348
		349	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		350
		351	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		352	reading at byte offset C<$in_offset>, and starts writing at the current
		353	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		354	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		355	other.
		356
		357	This call tries to make use of a native C<sendfile> syscall to provide
		358	zero-copy operation. For this to work, C<$out_fh> should refer to a
		359	socket, and C<$in_fh> should refer to mmap'able file.
		360
		361	If the native sendfile call fails or is not implemented, it will be
		362	emulated, so you can call C<aio_sendfile> on any type of filehandle
		363	regardless of the limitations of the operating system.
		364
		365	Please note, however, that C<aio_sendfile> can read more bytes from
		366	C<$in_fh> than are written, and there is no way to find out how many
		367	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only
		368	provides the number of bytes written to C<$out_fh>. Only if the result
		369	value equals C<$length> one can assume that C<$length> bytes have been
		370	read.
		371
		372
142	=item aio_readahead $fh,$offset,$length, $callback	373	=item aio_readahead $fh,$offset,$length, $callback->($retval)
143		374
144	Asynchronously reads the specified byte range into the page cache, using
145	the C<readahead> syscall. If that syscall doesn't exist the status will be
146	C<-1> and C<$!> is set to ENOSYS.
147
148	readahead() populates the page cache with data from a file so that	375	C<aio_readahead> populates the page cache with data from a file so that
149	subsequent reads from that file will not block on disk I/O. The C<$offset>	376	subsequent reads from that file will not block on disk I/O. The C<$offset>
150	argument specifies the starting point from which data is to be read and	377	argument specifies the starting point from which data is to be read and
151	C<$length> specifies the number of bytes to be read. I/O is performed in	378	C<$length> specifies the number of bytes to be read. I/O is performed in
152	whole pages, so that offset is effectively rounded down to a page boundary	379	whole pages, so that offset is effectively rounded down to a page boundary
153	and bytes are read up to the next page boundary greater than or equal to	380	and bytes are read up to the next page boundary greater than or equal to
154	(off-set+length). aio_readahead() does not read beyond the end of the	381	(off-set+length). C<aio_readahead> does not read beyond the end of the
155	file. The current file offset of the file is left unchanged.	382	file. The current file offset of the file is left unchanged.
156		383
		384	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
		385	emulated by simply reading the data, which would have a similar effect.
		386
		387
157	=item aio_stat $fh_or_path, $callback	388	=item aio_stat $fh_or_path, $callback->($status)
158		389
159	=item aio_lstat $fh, $callback	390	=item aio_lstat $fh, $callback->($status)
160		391
161	Works like perl's C<stat> or C<lstat> in void context. The callback will	392	Works like perl's C<stat> or C<lstat> in void context. The callback will
162	be called after the stat and the results will be available using C<stat _>	393	be called after the stat and the results will be available using C<stat _>
163	or C<-s _> etc...	394	or C<-s _> etc...
164		395
…		…
174	aio_stat "/etc/passwd", sub {	405	aio_stat "/etc/passwd", sub {
175	$_[0] and die "stat failed: $!";	406	$_[0] and die "stat failed: $!";
176	print "size is ", -s _, "\n";	407	print "size is ", -s _, "\n";
177	};	408	};
178		409
		410
		411	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		412
		413	Works like perl's C<utime> function (including the special case of $atime
		414	and $mtime being undef). Fractional times are supported if the underlying
		415	syscalls support them.
		416
		417	When called with a pathname, uses utimes(2) if available, otherwise
		418	utime(2). If called on a file descriptor, uses futimes(2) if available,
		419	otherwise returns ENOSYS, so this is not portable.
		420
		421	Examples:
		422
		423	# set atime and mtime to current time (basically touch(1)):
		424	aio_utime "path", undef, undef;
		425	# set atime to current time and mtime to beginning of the epoch:
		426	aio_utime "path", time, undef; # undef==0
		427
		428
		429	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		430
		431	Works like perl's C<chown> function, except that C<undef> for either $uid
		432	or $gid is being interpreted as "do not change" (but -1 can also be used).
		433
		434	Examples:
		435
		436	# same as "chown root path" in the shell:
		437	aio_chown "path", 0, -1;
		438	# same as above:
		439	aio_chown "path", 0, undef;
		440
		441
		442	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		443
		444	Works like perl's C<chmod> function.
		445
		446
179	=item aio_unlink $pathname, $callback	447	=item aio_unlink $pathname, $callback->($status)
180		448
181	Asynchronously unlink (delete) a file and call the callback with the	449	Asynchronously unlink (delete) a file and call the callback with the
182	result code.	450	result code.
183		451
		452
		453	=item aio_mknod $path, $mode, $dev, $callback->($status)
		454
		455	[EXPERIMENTAL]
		456
		457	Asynchronously create a device node (or fifo). See mknod(2).
		458
		459	The only (POSIX-) portable way of calling this function is:
		460
		461	aio_mknod $path, IO::AIO::S_IFIFO \| $mode, 0, sub { ...
		462
		463
		464	=item aio_link $srcpath, $dstpath, $callback->($status)
		465
		466	Asynchronously create a new link to the existing object at C<$srcpath> at
		467	the path C<$dstpath> and call the callback with the result code.
		468
		469
		470	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		471
		472	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		473	the path C<$dstpath> and call the callback with the result code.
		474
		475
		476	=item aio_readlink $path, $callback->($link)
		477
		478	Asynchronously read the symlink specified by C<$path> and pass it to
		479	the callback. If an error occurs, nothing or undef gets passed to the
		480	callback.
		481
		482
		483	=item aio_rename $srcpath, $dstpath, $callback->($status)
		484
		485	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		486	rename(2) and call the callback with the result code.
		487
		488
		489	=item aio_mkdir $pathname, $mode, $callback->($status)
		490
		491	Asynchronously mkdir (create) a directory and call the callback with
		492	the result code. C<$mode> will be modified by the umask at the time the
		493	request is executed, so do not change your umask.
		494
		495
		496	=item aio_rmdir $pathname, $callback->($status)
		497
		498	Asynchronously rmdir (delete) a directory and call the callback with the
		499	result code.
		500
		501
		502	=item aio_readdir $pathname, $callback->($entries)
		503
		504	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		505	directory (i.e. opendir + readdir + closedir). The entries will not be
		506	sorted, and will B<NOT> include the C<.> and C<..> entries.
		507
		508	The callback a single argument which is either C<undef> or an array-ref
		509	with the filenames.
		510
		511
		512	=item aio_load $path, $data, $callback->($status)
		513
		514	This is a composite request that tries to fully load the given file into
		515	memory. Status is the same as with aio_read.
		516
		517	=cut
		518
		519	sub aio_load($$;$) {
		520	aio_block {
		521	my ($path, undef, $cb) = @_;
		522	my $data = \$_[1];
		523
		524	my $pri = aioreq_pri;
		525	my $grp = aio_group $cb;
		526
		527	aioreq_pri $pri;
		528	add $grp aio_open $path, O_RDONLY, 0, sub {
		529	my $fh = shift
		530	or return $grp->result (-1);
		531
		532	aioreq_pri $pri;
		533	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		534	$grp->result ($_[0]);
		535	};
		536	};
		537
		538	$grp
		539	}
		540	}
		541
		542	=item aio_copy $srcpath, $dstpath, $callback->($status)
		543
		544	Try to copy the I<file> (directories not supported as either source or
		545	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		546	the C<0> (error) or C<-1> ok.
		547
		548	This is a composite request that it creates the destination file with
		549	mode 0200 and copies the contents of the source file into it using
		550	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		551	uid/gid, in that order.
		552
		553	If an error occurs, the partial destination file will be unlinked, if
		554	possible, except when setting atime, mtime, access mode and uid/gid, where
		555	errors are being ignored.
		556
		557	=cut
		558
		559	sub aio_copy($$;$) {
		560	aio_block {
		561	my ($src, $dst, $cb) = @_;
		562
		563	my $pri = aioreq_pri;
		564	my $grp = aio_group $cb;
		565
		566	aioreq_pri $pri;
		567	add $grp aio_open $src, O_RDONLY, 0, sub {
		568	if (my $src_fh = $_[0]) {
		569	my @stat = stat $src_fh;
		570
		571	aioreq_pri $pri;
		572	add $grp aio_open $dst, O_CREAT \| O_WRONLY \| O_TRUNC, 0200, sub {
		573	if (my $dst_fh = $_[0]) {
		574	aioreq_pri $pri;
		575	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		576	if ($_[0] == $stat[7]) {
		577	$grp->result (0);
		578	close $src_fh;
		579
		580	# those should not normally block. should. should.
		581	utime $stat[8], $stat[9], $dst;
		582	chmod $stat[2] & 07777, $dst_fh;
		583	chown $stat[4], $stat[5], $dst_fh;
		584	close $dst_fh;
		585	} else {
		586	$grp->result (-1);
		587	close $src_fh;
		588	close $dst_fh;
		589
		590	aioreq $pri;
		591	add $grp aio_unlink $dst;
		592	}
		593	};
		594	} else {
		595	$grp->result (-1);
		596	}
		597	},
		598
		599	} else {
		600	$grp->result (-1);
		601	}
		602	};
		603
		604	$grp
		605	}
		606	}
		607
		608	=item aio_move $srcpath, $dstpath, $callback->($status)
		609
		610	Try to move the I<file> (directories not supported as either source or
		611	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		612	the C<0> (error) or C<-1> ok.
		613
		614	This is a composite request that tries to rename(2) the file first. If
		615	rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if
		616	that is successful, unlinking the C<$srcpath>.
		617
		618	=cut
		619
		620	sub aio_move($$;$) {
		621	aio_block {
		622	my ($src, $dst, $cb) = @_;
		623
		624	my $pri = aioreq_pri;
		625	my $grp = aio_group $cb;
		626
		627	aioreq_pri $pri;
		628	add $grp aio_rename $src, $dst, sub {
		629	if ($_[0] && $! == EXDEV) {
		630	aioreq_pri $pri;
		631	add $grp aio_copy $src, $dst, sub {
		632	$grp->result ($_[0]);
		633
		634	if (!$_[0]) {
		635	aioreq_pri $pri;
		636	add $grp aio_unlink $src;
		637	}
		638	};
		639	} else {
		640	$grp->result ($_[0]);
		641	}
		642	};
		643
		644	$grp
		645	}
		646	}
		647
		648	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		649
		650	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		651	efficiently separate the entries of directory C<$path> into two sets of
		652	names, directories you can recurse into (directories), and ones you cannot
		653	recurse into (everything else, including symlinks to directories).
		654
		655	C<aio_scandir> is a composite request that creates of many sub requests_
		656	C<$maxreq> specifies the maximum number of outstanding aio requests that
		657	this function generates. If it is C<< <= 0 >>, then a suitable default
		658	will be chosen (currently 4).
		659
		660	On error, the callback is called without arguments, otherwise it receives
		661	two array-refs with path-relative entry names.
		662
		663	Example:
		664
		665	aio_scandir $dir, 0, sub {
		666	my ($dirs, $nondirs) = @_;
		667	print "real directories: @$dirs\n";
		668	print "everything else: @$nondirs\n";
		669	};
		670
		671	Implementation notes.
		672
		673	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		674
		675	After reading the directory, the modification time, size etc. of the
		676	directory before and after the readdir is checked, and if they match (and
		677	isn't the current time), the link count will be used to decide how many
		678	entries are directories (if >= 2). Otherwise, no knowledge of the number
		679	of subdirectories will be assumed.
		680
		681	Then entries will be sorted into likely directories (everything without
		682	a non-initial dot currently) and likely non-directories (everything
		683	else). Then every entry plus an appended C</.> will be C<stat>'ed,
		684	likely directories first. If that succeeds, it assumes that the entry
		685	is a directory or a symlink to directory (which will be checked
		686	seperately). This is often faster than stat'ing the entry itself because
		687	filesystems might detect the type of the entry without reading the inode
		688	data (e.g. ext2fs filetype feature).
		689
		690	If the known number of directories (link count - 2) has been reached, the
		691	rest of the entries is assumed to be non-directories.
		692
		693	This only works with certainty on POSIX (= UNIX) filesystems, which
		694	fortunately are the vast majority of filesystems around.
		695
		696	It will also likely work on non-POSIX filesystems with reduced efficiency
		697	as those tend to return 0 or 1 as link counts, which disables the
		698	directory counting heuristic.
		699
		700	=cut
		701
		702	sub aio_scandir($$;$) {
		703	aio_block {
		704	my ($path, $maxreq, $cb) = @_;
		705
		706	my $pri = aioreq_pri;
		707
		708	my $grp = aio_group $cb;
		709
		710	$maxreq = 4 if $maxreq <= 0;
		711
		712	# stat once
		713	aioreq_pri $pri;
		714	add $grp aio_stat $path, sub {
		715	return $grp->result () if $_[0];
		716	my $now = time;
		717	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		718
		719	# read the directory entries
		720	aioreq_pri $pri;
		721	add $grp aio_readdir $path, sub {
		722	my $entries = shift
		723	or return $grp->result ();
		724
		725	# stat the dir another time
		726	aioreq_pri $pri;
		727	add $grp aio_stat $path, sub {
		728	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		729
		730	my $ndirs;
		731
		732	# take the slow route if anything looks fishy
		733	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		734	$ndirs = -1;
		735	} else {
		736	# if nlink == 2, we are finished
		737	# on non-posix-fs's, we rely on nlink < 2
		738	$ndirs = (stat _)[3] - 2
		739	or return $grp->result ([], $entries);
		740	}
		741
		742	# sort into likely dirs and likely nondirs
		743	# dirs == files without ".", short entries first
		744	$entries = [map $_->[0],
		745	sort { $b->[1] cmp $a->[1] }
		746	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
		747	@$entries];
		748
		749	my (@dirs, @nondirs);
		750
		751	my $statgrp = add $grp aio_group sub {
		752	$grp->result (\@dirs, \@nondirs);
		753	};
		754
		755	limit $statgrp $maxreq;
		756	feed $statgrp sub {
		757	return unless @$entries;
		758	my $entry = pop @$entries;
		759
		760	aioreq_pri $pri;
		761	add $statgrp aio_stat "$path/$entry/.", sub {
		762	if ($_[0] < 0) {
		763	push @nondirs, $entry;
		764	} else {
		765	# need to check for real directory
		766	aioreq_pri $pri;
		767	add $statgrp aio_lstat "$path/$entry", sub {
		768	if (-d _) {
		769	push @dirs, $entry;
		770
		771	unless (--$ndirs) {
		772	push @nondirs, @$entries;
		773	feed $statgrp;
		774	}
		775	} else {
		776	push @nondirs, $entry;
		777	}
		778	}
		779	}
		780	};
		781	};
		782	};
		783	};
		784	};
		785
		786	$grp
		787	}
		788	}
		789
		790	=item aio_rmtree $path, $callback->($status)
		791
		792	Delete a directory tree starting (and including) C<$path>, return the
		793	status of the final C<rmdir> only. This is a composite request that
		794	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		795	everything else.
		796
		797	=cut
		798
		799	sub aio_rmtree;
		800	sub aio_rmtree($;$) {
		801	aio_block {
		802	my ($path, $cb) = @_;
		803
		804	my $pri = aioreq_pri;
		805	my $grp = aio_group $cb;
		806
		807	aioreq_pri $pri;
		808	add $grp aio_scandir $path, 0, sub {
		809	my ($dirs, $nondirs) = @_;
		810
		811	my $dirgrp = aio_group sub {
		812	add $grp aio_rmdir $path, sub {
		813	$grp->result ($_[0]);
		814	};
		815	};
		816
		817	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		818	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		819
		820	add $grp $dirgrp;
		821	};
		822
		823	$grp
		824	}
		825	}
		826
184	=item aio_fsync $fh, $callback	827	=item aio_fsync $fh, $callback->($status)
185		828
186	Asynchronously call fsync on the given filehandle and call the callback	829	Asynchronously call fsync on the given filehandle and call the callback
187	with the fsync result code.	830	with the fsync result code.
188		831
189	=item aio_fdatasync $fh, $callback	832	=item aio_fdatasync $fh, $callback->($status)
190		833
191	Asynchronously call fdatasync on the given filehandle and call the	834	Asynchronously call fdatasync on the given filehandle and call the
192	callback with the fdatasync result code.	835	callback with the fdatasync result code.
193		836
		837	If this call isn't available because your OS lacks it or it couldn't be
		838	detected, it will be emulated by calling C<fsync> instead.
		839
		840	=item aio_group $callback->(...)
		841
		842	This is a very special aio request: Instead of doing something, it is a
		843	container for other aio requests, which is useful if you want to bundle
		844	many requests into a single, composite, request with a definite callback
		845	and the ability to cancel the whole request with its subrequests.
		846
		847	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		848	for more info.
		849
		850	Example:
		851
		852	my $grp = aio_group sub {
		853	print "all stats done\n";
		854	};
		855
		856	add $grp
		857	(aio_stat ...),
		858	(aio_stat ...),
		859	...;
		860
		861	=item aio_nop $callback->()
		862
		863	This is a special request - it does nothing in itself and is only used for
		864	side effects, such as when you want to add a dummy request to a group so
		865	that finishing the requests in the group depends on executing the given
		866	code.
		867
		868	While this request does nothing, it still goes through the execution
		869	phase and still requires a worker thread. Thus, the callback will not
		870	be executed immediately but only after other requests in the queue have
		871	entered their execution phase. This can be used to measure request
		872	latency.
		873
		874	=item IO::AIO::aio_busy $fractional_seconds, $callback->() NOT EXPORTED
		875
		876	Mainly used for debugging and benchmarking, this aio request puts one of
		877	the request workers to sleep for the given time.
		878
		879	While it is theoretically handy to have simple I/O scheduling requests
		880	like sleep and file handle readable/writable, the overhead this creates is
		881	immense (it blocks a thread for a long time) so do not use this function
		882	except to put your application under artificial I/O pressure.
		883
194	=back	884	=back
195		885
		886	=head2 IO::AIO::REQ CLASS
		887
		888	All non-aggregate C<aio_*> functions return an object of this class when
		889	called in non-void context.
		890
		891	=over 4
		892
		893	=item cancel $req
		894
		895	Cancels the request, if possible. Has the effect of skipping execution
		896	when entering the B<execute> state and skipping calling the callback when
		897	entering the the B<result> state, but will leave the request otherwise
		898	untouched. That means that requests that currently execute will not be
		899	stopped and resources held by the request will not be freed prematurely.
		900
		901	=item cb $req $callback->(...)
		902
		903	Replace (or simply set) the callback registered to the request.
		904
		905	=back
		906
		907	=head2 IO::AIO::GRP CLASS
		908
		909	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		910	objects of this class, too.
		911
		912	A IO::AIO::GRP object is a special request that can contain multiple other
		913	aio requests.
		914
		915	You create one by calling the C<aio_group> constructing function with a
		916	callback that will be called when all contained requests have entered the
		917	C<done> state:
		918
		919	my $grp = aio_group sub {
		920	print "all requests are done\n";
		921	};
		922
		923	You add requests by calling the C<add> method with one or more
		924	C<IO::AIO::REQ> objects:
		925
		926	$grp->add (aio_unlink "...");
		927
		928	add $grp aio_stat "...", sub {
		929	$_[0] or return $grp->result ("error");
		930
		931	# add another request dynamically, if first succeeded
		932	add $grp aio_open "...", sub {
		933	$grp->result ("ok");
		934	};
		935	};
		936
		937	This makes it very easy to create composite requests (see the source of
		938	C<aio_move> for an application) that work and feel like simple requests.
		939
		940	=over 4
		941
		942	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		943	C<IO::AIO::poll_cb>, just like any other request.
		944
		945	=item * They can be canceled like any other request. Canceling will cancel not
		946	only the request itself, but also all requests it contains.
		947
		948	=item * They can also can also be added to other IO::AIO::GRP objects.
		949
		950	=item * You must not add requests to a group from within the group callback (or
		951	any later time).
		952
		953	=back
		954
		955	Their lifetime, simplified, looks like this: when they are empty, they
		956	will finish very quickly. If they contain only requests that are in the
		957	C<done> state, they will also finish. Otherwise they will continue to
		958	exist.
		959
		960	That means after creating a group you have some time to add requests. And
		961	in the callbacks of those requests, you can add further requests to the
		962	group. And only when all those requests have finished will the the group
		963	itself finish.
		964
		965	=over 4
		966
		967	=item add $grp ...
		968
		969	=item $grp->add (...)
		970
		971	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		972	be added, including other groups, as long as you do not create circular
		973	dependencies.
		974
		975	Returns all its arguments.
		976
		977	=item $grp->cancel_subs
		978
		979	Cancel all subrequests and clears any feeder, but not the group request
		980	itself. Useful when you queued a lot of events but got a result early.
		981
		982	=item $grp->result (...)
		983
		984	Set the result value(s) that will be passed to the group callback when all
		985	subrequests have finished and set thre groups errno to the current value
		986	of errno (just like calling C<errno> without an error number). By default,
		987	no argument will be passed and errno is zero.
		988
		989	=item $grp->errno ([$errno])
		990
		991	Sets the group errno value to C<$errno>, or the current value of errno
		992	when the argument is missing.
		993
		994	Every aio request has an associated errno value that is restored when
		995	the callback is invoked. This method lets you change this value from its
		996	default (0).
		997
		998	Calling C<result> will also set errno, so make sure you either set C<$!>
		999	before the call to C<result>, or call c<errno> after it.
		1000
		1001	=item feed $grp $callback->($grp)
		1002
		1003	Sets a feeder/generator on this group: every group can have an attached
		1004	generator that generates requests if idle. The idea behind this is that,
		1005	although you could just queue as many requests as you want in a group,
		1006	this might starve other requests for a potentially long time. For
		1007	example, C<aio_scandir> might generate hundreds of thousands C<aio_stat>
		1008	requests, delaying any later requests for a long time.
		1009
		1010	To avoid this, and allow incremental generation of requests, you can
		1011	instead a group and set a feeder on it that generates those requests. The
		1012	feed callback will be called whenever there are few enough (see C<limit>,
		1013	below) requests active in the group itself and is expected to queue more
		1014	requests.
		1015
		1016	The feed callback can queue as many requests as it likes (i.e. C<add> does
		1017	not impose any limits).
		1018
		1019	If the feed does not queue more requests when called, it will be
		1020	automatically removed from the group.
		1021
		1022	If the feed limit is C<0>, it will be set to C<2> automatically.
		1023
		1024	Example:
		1025
		1026	# stat all files in @files, but only ever use four aio requests concurrently:
		1027
		1028	my $grp = aio_group sub { print "finished\n" };
		1029	limit $grp 4;
		1030	feed $grp sub {
		1031	my $file = pop @files
		1032	or return;
		1033
		1034	add $grp aio_stat $file, sub { ... };
		1035	};
		1036
		1037	=item limit $grp $num
		1038
		1039	Sets the feeder limit for the group: The feeder will be called whenever
		1040	the group contains less than this many requests.
		1041
		1042	Setting the limit to C<0> will pause the feeding process.
		1043
		1044	=back
		1045
196	=head2 SUPPORT FUNCTIONS	1046	=head2 SUPPORT FUNCTIONS
197		1047
		1048	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1049
198	=over 4	1050	=over 4
199		1051
200	=item $fileno = IO::AIO::poll_fileno	1052	=item $fileno = IO::AIO::poll_fileno
201		1053
202	Return the I<request result pipe filehandle>. This filehandle must be	1054	Return the I<request result pipe file descriptor>. This filehandle must be
203	polled for reading by some mechanism outside this module (e.g. Event	1055	polled for reading by some mechanism outside this module (e.g. Event or
204	or select, see below). If the pipe becomes readable you have to call	1056	select, see below or the SYNOPSIS). If the pipe becomes readable you have
205	C<poll_cb> to check the results.	1057	to call C<poll_cb> to check the results.
206		1058
207	See C<poll_cb> for an example.	1059	See C<poll_cb> for an example.
208		1060
209	=item IO::AIO::poll_cb	1061	=item IO::AIO::poll_cb
210		1062
211	Process all outstanding events on the result pipe. You have to call this	1063	Process some outstanding events on the result pipe. You have to call this
212	regularly. Returns the number of events processed. Returns immediately	1064	regularly. Returns the number of events processed. Returns immediately
213	when no events are outstanding.	1065	when no events are outstanding. The amount of events processed depends on
		1066	the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
214		1067
215	You can use Event to multiplex, e.g.:	1068	If not all requests were processed for whatever reason, the filehandle
		1069	will still be ready when C<poll_cb> returns.
		1070
		1071	Example: Install an Event watcher that automatically calls
		1072	IO::AIO::poll_cb with high priority:
216		1073
217	Event->io (fd => IO::AIO::poll_fileno,	1074	Event->io (fd => IO::AIO::poll_fileno,
218	poll => 'r', async => 1,	1075	poll => 'r', async => 1,
219	cb => \&IO::AIO::poll_cb);	1076	cb => \&IO::AIO::poll_cb);
220		1077
		1078	=item IO::AIO::max_poll_reqs $nreqs
		1079
		1080	=item IO::AIO::max_poll_time $seconds
		1081
		1082	These set the maximum number of requests (default C<0>, meaning infinity)
		1083	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1084	the maximum amount of time (default C<0>, meaning infinity) spent in
		1085	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1086	of time C<poll_cb> is allowed to use).
		1087
		1088	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1089	syscall per request processed, which is not normally a problem unless your
		1090	callbacks are really really fast or your OS is really really slow (I am
		1091	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1092
		1093	Setting these is useful if you want to ensure some level of
		1094	interactiveness when perl is not fast enough to process all requests in
		1095	time.
		1096
		1097	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
		1098
		1099	Example: Install an Event watcher that automatically calls
		1100	IO::AIO::poll_cb with low priority, to ensure that other parts of the
		1101	program get the CPU sometimes even under high AIO load.
		1102
		1103	# try not to spend much more than 0.1s in poll_cb
		1104	IO::AIO::max_poll_time 0.1;
		1105
		1106	# use a low priority so other tasks have priority
		1107	Event->io (fd => IO::AIO::poll_fileno,
		1108	poll => 'r', nice => 1,
		1109	cb => &IO::AIO::poll_cb);
		1110
221	=item IO::AIO::poll_wait	1111	=item IO::AIO::poll_wait
222		1112
		1113	If there are any outstanding requests and none of them in the result
223	Wait till the result filehandle becomes ready for reading (simply does a	1114	phase, wait till the result filehandle becomes ready for reading (simply
224	select on the filehandle. This is useful if you want to synchronously wait	1115	does a C<select> on the filehandle. This is useful if you want to
225	for some requests to finish).	1116	synchronously wait for some requests to finish).
226		1117
227	See C<nreqs> for an example.	1118	See C<nreqs> for an example.
228		1119
		1120	=item IO::AIO::poll
		1121
		1122	Waits until some requests have been handled.
		1123
		1124	Returns the number of requests processed, but is otherwise strictly
		1125	equivalent to:
		1126
		1127	IO::AIO::poll_wait, IO::AIO::poll_cb
		1128
229	=item IO::AIO::nreqs	1129	=item IO::AIO::flush
230		1130
231	Returns the number of requests currently outstanding.	1131	Wait till all outstanding AIO requests have been handled.
232		1132
233	Example: wait till there are no outstanding requests anymore:	1133	Strictly equivalent to:
234		1134
235	IO::AIO::poll_wait, IO::AIO::poll_cb	1135	IO::AIO::poll_wait, IO::AIO::poll_cb
236	while IO::AIO::nreqs;	1136	while IO::AIO::nreqs;
237		1137
238	=item IO::AIO::flush	1138	=back
239		1139
240	Wait till all outstanding AIO requests have been handled.	1140	=head3 CONTROLLING THE NUMBER OF THREADS
		1141
		1142	=over
241		1143
242	=item IO::AIO::min_parallel $nthreads	1144	=item IO::AIO::min_parallel $nthreads
243		1145
244	Set the minimum number of AIO threads to C<$nthreads>. The default is	1146	Set the minimum number of AIO threads to C<$nthreads>. The current
245	C<1>, which means a single asynchronous operation can be done at one time	1147	default is C<8>, which means eight asynchronous operations can execute
246	(the number of outstanding operations, however, is unlimited).	1148	concurrently at any one time (the number of outstanding requests,
		1149	however, is unlimited).
247		1150
		1151	IO::AIO starts threads only on demand, when an AIO request is queued and
		1152	no free thread exists. Please note that queueing up a hundred requests can
		1153	create demand for a hundred threads, even if it turns out that everything
		1154	is in the cache and could have been processed faster by a single thread.
		1155
248	It is recommended to keep the number of threads low, as some Linux	1156	It is recommended to keep the number of threads relatively low, as some
249	kernel versions will scale negatively with the number of threads (higher	1157	Linux kernel versions will scale negatively with the number of threads
250	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32	1158	(higher parallelity => MUCH higher latency). With current Linux 2.6
251	threads should be fine.	1159	versions, 4-32 threads should be fine.
252		1160
253	Under normal circumstances you don't need to call this function, as this	1161	Under most circumstances you don't need to call this function, as the
254	module automatically starts some threads (the exact number might change,	1162	module selects a default that is suitable for low to moderate load.
255	and is currently 4).
256		1163
257	=item IO::AIO::max_parallel $nthreads	1164	=item IO::AIO::max_parallel $nthreads
258		1165
259	Sets the maximum number of AIO threads to C<$nthreads>. If more than	1166	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
260	the specified number of threads are currently running, kill them. This	1167	specified number of threads are currently running, this function kills
261	function blocks until the limit is reached.	1168	them. This function blocks until the limit is reached.
		1169
		1170	While C<$nthreads> are zero, aio requests get queued but not executed
		1171	until the number of threads has been increased again.
262		1172
263	This module automatically runs C<max_parallel 0> at program end, to ensure	1173	This module automatically runs C<max_parallel 0> at program end, to ensure
264	that all threads are killed and that there are no outstanding requests.	1174	that all threads are killed and that there are no outstanding requests.
265		1175
266	Under normal circumstances you don't need to call this function.	1176	Under normal circumstances you don't need to call this function.
267		1177
		1178	=item IO::AIO::max_idle $nthreads
		1179
		1180	Limit the number of threads (default: 4) that are allowed to idle (i.e.,
		1181	threads that did not get a request to process within 10 seconds). That
		1182	means if a thread becomes idle while C<$nthreads> other threads are also
		1183	idle, it will free its resources and exit.
		1184
		1185	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1186	to allow for extremely high load situations, but want to free resources
		1187	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1188
		1189	The default is probably ok in most situations, especially if thread
		1190	creation is fast. If thread creation is very slow on your system you might
		1191	want to use larger values.
		1192
268	=item $oldnreqs = IO::AIO::max_outstanding $nreqs	1193	=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs
		1194
		1195	This is a very bad function to use in interactive programs because it
		1196	blocks, and a bad way to reduce concurrency because it is inexact: Better
		1197	use an C<aio_group> together with a feed callback.
269		1198
270	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1199	Sets the maximum number of outstanding requests to C<$nreqs>. If you
271	try to queue up more than this number of requests, the caller will block until	1200	to queue up more than this number of requests, the next call to the
272	some requests have been handled.	1201	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
		1202	function will block until the limit is no longer exceeded.
273		1203
274	The default is very large, so normally there is no practical limit. If you	1204	The default value is very large, so there is no practical limit on the
275	queue up many requests in a loop it it often improves speed if you set	1205	number of outstanding requests.
276	this to a relatively low number, such as C<100>.
277		1206
278	Under normal circumstances you don't need to call this function.	1207	You can still queue as many requests as you want. Therefore,
		1208	C<max_oustsanding> is mainly useful in simple scripts (with low values) or
		1209	as a stop gap to shield against fatal memory overflow (with large values).
		1210
		1211	=back
		1212
		1213	=head3 STATISTICAL INFORMATION
		1214
		1215	=over
		1216
		1217	=item IO::AIO::nreqs
		1218
		1219	Returns the number of requests currently in the ready, execute or pending
		1220	states (i.e. for which their callback has not been invoked yet).
		1221
		1222	Example: wait till there are no outstanding requests anymore:
		1223
		1224	IO::AIO::poll_wait, IO::AIO::poll_cb
		1225	while IO::AIO::nreqs;
		1226
		1227	=item IO::AIO::nready
		1228
		1229	Returns the number of requests currently in the ready state (not yet
		1230	executed).
		1231
		1232	=item IO::AIO::npending
		1233
		1234	Returns the number of requests currently in the pending state (executed,
		1235	but not yet processed by poll_cb).
279		1236
280	=back	1237	=back
281		1238
282	=cut	1239	=cut
283		1240
284	# support function to convert a fd into a perl filehandle	1241	# support function to convert a fd into a perl filehandle
285	sub _fd2fh {	1242	sub _fd2fh {
286	return undef if $_[0] < 0;	1243	return undef if $_[0] < 0;
287		1244
288	# try to be perl5.6-compatible	1245	# try to generate nice filehandles
289	local *AIO_FH;	1246	my $sym = "IO::AIO::fd#$_[0]";
290	open AIO_FH, "+<&=$_[0]"	1247	local *$sym;
		1248
		1249	open *$sym, "+<&=$_[0]" # usually works under any unix
		1250	or open *$sym, "<&=$_[0]" # cygwin needs this
		1251	or open *$sym, ">&=$_[0]" # or this
291	or return undef;	1252	or return undef;
292		1253
293	*AIO_FH	1254	*$sym
294	}	1255	}
295		1256
296	min_parallel 4;	1257	min_parallel 8;
297		1258
298	END {	1259	END { flush }
299	max_parallel 0;
300	}
301		1260
302	1;	1261	1;
303		1262
		1263	=head2 FORK BEHAVIOUR
		1264
		1265	This module should do "the right thing" when the process using it forks:
		1266
		1267	Before the fork, IO::AIO enters a quiescent state where no requests
		1268	can be added in other threads and no results will be processed. After
		1269	the fork the parent simply leaves the quiescent state and continues
		1270	request/result processing, while the child frees the request/result queue
		1271	(so that the requests started before the fork will only be handled in the
		1272	parent). Threads will be started on demand until the limit set in the
		1273	parent process has been reached again.
		1274
		1275	In short: the parent will, after a short pause, continue as if fork had
		1276	not been called, while the child will act as if IO::AIO has not been used
		1277	yet.
		1278
		1279	=head2 MEMORY USAGE
		1280
		1281	Per-request usage:
		1282
		1283	Each aio request uses - depending on your architecture - around 100-200
		1284	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1285	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1286	scalars and other data passed into aio requests will also be locked and
		1287	will consume memory till the request has entered the done state.
		1288
		1289	This is now awfully much, so queuing lots of requests is not usually a
		1290	problem.
		1291
		1292	Per-thread usage:
		1293
		1294	In the execution phase, some aio requests require more memory for
		1295	temporary buffers, and each thread requires a stack and other data
		1296	structures (usually around 16k-128k, depending on the OS).
		1297
		1298	=head1 KNOWN BUGS
		1299
		1300	Known bugs will be fixed in the next release.
		1301
304	=head1 SEE ALSO	1302	=head1 SEE ALSO
305		1303
306	L<Coro>, L<Linux::AIO>.	1304	L<Coro::AIO>.
307		1305
308	=head1 AUTHOR	1306	=head1 AUTHOR
309		1307
310	Marc Lehmann <schmorp@schmorp.de>	1308	Marc Lehmann <schmorp@schmorp.de>
311	http://home.schmorp.de/	1309	http://home.schmorp.de/

Diff Legend

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

Comparing IO-AIO/AIO.pm (file contents): Revision 1.12 by root, Mon Jul 11 01:03:17 2005 UTC vs. Revision 1.108 by root, Fri Jun 1 13:25:50 2007 UTC

Diff Legend

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.12 by root, Mon Jul 11 01:03:17 2005 UTC vs.
Revision 1.108 by root, Fri Jun 1 13:25:50 2007 UTC