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

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.1 by root, Sun Jul 10 17:07:44 2005 UTC vs.
Revision 1.82 by root, Fri Oct 27 20:10:06 2006 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use IO::AIO;	7	use IO::AIO;
8		8
		9	aio_open "/etc/passwd", O_RDONLY, 0, sub {
		10	my ($fh) = @_;
		11	...
		12	};
		13
		14	aio_unlink "/tmp/file", sub { };
		15
		16	aio_read $fh, 30000, 1024, $buffer, 0, sub {
		17	$_[0] > 0 or die "read error: $!";
		18	};
		19
		20	# version 2+ has request and group objects
		21	use IO::AIO 2;
		22
		23	aioreq_pri 4; # give next request a very high priority
		24	my $req = aio_unlink "/tmp/file", sub { };
		25	$req->cancel; # cancel request if still in queue
		26
		27	my $grp = aio_group sub { print "all stats done\n" };
		28	add $grp aio_stat "..." for ...;
		29
		30	# AnyEvent integration
		31	open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
		32	my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
		33
		34	# Event integration
		35	Event->io (fd => IO::AIO::poll_fileno,
		36	poll => 'r',
		37	cb => \&IO::AIO::poll_cb);
		38
		39	# Glib/Gtk2 integration
		40	add_watch Glib::IO IO::AIO::poll_fileno,
		41	in => sub { IO::AIO::poll_cb; 1 };
		42
		43	# Tk integration
		44	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		45	readable => \&IO::AIO::poll_cb);
		46
		47	# Danga::Socket integration
		48	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		49	\&IO::AIO::poll_cb);
		50
9	=head1 DESCRIPTION	51	=head1 DESCRIPTION
10		52
11	This module implements asynchronous I/O using whatever means your	53	This module implements asynchronous I/O using whatever means your
12	operating system supports. Currently, it falls back to Linux::AIO if that	54	operating system supports.
13	module is available, or uses pthreads to emulato aio functionality.
14		55
15	Currently, in this module a number of threads are started that execute	56	In this version, a number of threads are started that execute your
16	your read/writes and signal their completion. You don't need thread	57	requests and signal their completion. You don't need thread support
17	support in your libc or perl, and the threads created by this module will	58	in perl, and the threads created by this module will not be visible
18	not be visible to the pthreads library.	59	to perl. In the future, this module might make use of the native aio
		60	functions available on many operating systems. However, they are often
		61	not well-supported or restricted (Linux doesn't allow them on normal
		62	files currently, for example), and they would only support aio_read and
		63	aio_write, so the remaining functionality would have to be implemented
		64	using threads anyway.
19		65
20	Although the module will work with in the presence of other threads, it is	66	Although the module will work with in the presence of other (Perl-)
21	not reentrant, so use appropriate locking yourself.	67	threads, it is currently not reentrant in any way, so use appropriate
		68	locking yourself, always call C<poll_cb> from within the same thread, or
		69	never call C<poll_cb> (or other C<aio_> functions) recursively.
22		70
23	=head2 API NOTES	71	=head1 REQUEST ANATOMY AND LIFETIME
		72
		73	Every C<aio_*> function creates a request. which is a C data structure not
		74	directly visible to Perl.
		75
		76	If called in non-void context, every request function returns a Perl
		77	object representing the request. In void context, nothing is returned,
		78	which saves a bit of memory.
		79
		80	The perl object is a fairly standard ref-to-hash object. The hash contents
		81	are not used by IO::AIO so you are free to store anything you like in it.
		82
		83	During their existance, aio requests travel through the following states,
		84	in order:
		85
		86	=over 4
		87
		88	=item ready
		89
		90	Immediately after a request is created it is put into the ready state,
		91	waiting for a thread to execute it.
		92
		93	=item execute
		94
		95	A thread has accepted the request for processing and is currently
		96	executing it (e.g. blocking in read).
		97
		98	=item pending
		99
		100	The request has been executed and is waiting for result processing.
		101
		102	While request submission and execution is fully asynchronous, result
		103	processing is not and relies on the perl interpreter calling C<poll_cb>
		104	(or another function with the same effect).
		105
		106	=item result
		107
		108	The request results are processed synchronously by C<poll_cb>.
		109
		110	The C<poll_cb> function will process all outstanding aio requests by
		111	calling their callbacks, freeing memory associated with them and managing
		112	any groups they are contained in.
		113
		114	=item done
		115
		116	Request has reached the end of its lifetime and holds no resources anymore
		117	(except possibly for the Perl object, but its connection to the actual
		118	aio request is severed and calling its methods will either do nothing or
		119	result in a runtime error).
		120
		121	=cut
		122
		123	package IO::AIO;
		124
		125	no warnings;
		126	use strict 'vars';
		127
		128	use base 'Exporter';
		129
		130	BEGIN {
		131	our $VERSION = '2.0';
		132
		133	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat
		134	aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink
		135	aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move
		136	aio_copy aio_group aio_nop aio_mknod);
		137	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
		138	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
		139	min_parallel max_parallel nreqs nready npending);
		140
		141	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
		142
		143	require XSLoader;
		144	XSLoader::load ("IO::AIO", $VERSION);
		145	}
		146
		147	=head1 FUNCTIONS
		148
		149	=head2 AIO FUNCTIONS
24		150
25	All the C<aio_*> calls are more or less thin wrappers around the syscall	151	All the C<aio_*> calls are more or less thin wrappers around the syscall
26	with the same name (sans C<aio_>). The arguments are similar or identical,	152	with the same name (sans C<aio_>). The arguments are similar or identical,
27	and they all accept an additional C<$callback> argument which must be	153	and they all accept an additional (and optional) C<$callback> argument
28	a code reference. This code reference will get called with the syscall	154	which must be a code reference. This code reference will get called with
29	return code (e.g. most syscalls return C<-1> on error, unlike perl, which	155	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
30	usually delivers "false") as it's sole argument when the given syscall has	156	perl, which usually delivers "false") as it's sole argument when the given
31	been executed asynchronously.	157	syscall has been executed asynchronously.
32		158
33	All functions that expect a filehandle will also accept a file descriptor.	159	All functions expecting a filehandle keep a copy of the filehandle
		160	internally until the request has finished.
34		161
		162	All requests return objects of type L<IO::AIO::REQ> that allow further
		163	manipulation of those requests while they are in-flight.
		164
35	The filenames you pass to these routines I<must> be absolute. The reason	165	The pathnames you pass to these routines I<must> be absolute and
36	is that at the time the request is being executed, the current working	166	encoded in byte form. The reason for the former is that at the time the
37	directory could have changed. Alternatively, you can make sure that you	167	request is being executed, the current working directory could have
		168	changed. Alternatively, you can make sure that you never change the
38	never change the current working directory.	169	current working directory.
		170
		171	To encode pathnames to byte form, either make sure you either: a)
		172	always pass in filenames you got from outside (command line, readdir
		173	etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode
		174	your pathnames to the locale (or other) encoding in effect in the user
		175	environment, d) use Glib::filename_from_unicode on unicode filenames or e)
		176	use something else.
39		177
40	=over 4	178	=over 4
41		179
42	=cut	180	=item $prev_pri = aioreq_pri [$pri]
43		181
44	package IO::AIO;	182	Returns the priority value that would be used for the next request and, if
		183	C<$pri> is given, sets the priority for the next aio request.
45		184
46	use base 'Exporter';	185	The default priority is C<0>, the minimum and maximum priorities are C<-4>
		186	and C<4>, respectively. Requests with higher priority will be serviced
		187	first.
47		188
48	BEGIN {	189	The priority will be reset to C<0> after each call to one of the C<aio_*>
49	$VERSION = 0.1;	190	functions.
50		191
51	@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink	192	Example: open a file with low priority, then read something from it with
52	aio_fsync aio_fdatasync aio_readahead);	193	higher priority so the read request is serviced before other low priority
53	@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs);	194	open requests (potentially spamming the cache):
54		195
55	require XSLoader;	196	aioreq_pri -3;
56	XSLoader::load IO::AIO, $VERSION;	197	aio_open ..., sub {
57	}	198	return unless $_[0];
58		199
59	=item IO::AIO::min_parallel $nthreads	200	aioreq_pri -2;
		201	aio_read $_[0], ..., sub {
		202	...
		203	};
		204	};
60		205
61	Set the minimum number of AIO threads to C<$nthreads>. The default is	206	=item aioreq_nice $pri_adjust
62	C<1>, which means a single asynchronous operation can be done at one time
63	(the number of outstanding operations, however, is unlimited).
64		207
65	It is recommended to keep the number of threads low, as some linux	208	Similar to C<aioreq_pri>, but subtracts the given value from the current
66	kernel versions will scale negatively with the number of threads (higher	209	priority, so effects are cumulative.
67	parallelity => MUCH higher latency).
68		210
69	Under normal circumstances you don't need to call this function, as this
70	module automatically starts a single async thread.
71
72	=item IO::AIO::max_parallel $nthreads
73
74	Sets the maximum number of AIO threads to C<$nthreads>. If more than
75	the specified number of threads are currently running, kill them. This
76	function blocks until the limit is reached.
77
78	This module automatically runs C<max_parallel 0> at program end, to ensure
79	that all threads are killed and that there are no outstanding requests.
80
81	Under normal circumstances you don't need to call this function.
82
83	=item $fileno = IO::AIO::poll_fileno
84
85	Return the I<request result pipe filehandle>. This filehandle must be
86	polled for reading by some mechanism outside this module (e.g. Event
87	or select, see below). If the pipe becomes readable you have to call
88	C<poll_cb> to check the results.
89
90	See C<poll_cb> for an example.
91
92	=item IO::AIO::poll_cb
93
94	Process all outstanding events on the result pipe. You have to call this
95	regularly. Returns the number of events processed. Returns immediately
96	when no events are outstanding.
97
98	You can use Event to multiplex, e.g.:
99
100	Event->io (fd => IO::AIO::poll_fileno,
101	poll => 'r', async => 1,
102	cb => \&IO::AIO::poll_cb);
103
104	=item IO::AIO::poll_wait
105
106	Wait till the result filehandle becomes ready for reading (simply does a
107	select on the filehandle. This is useful if you want to synchronously wait
108	for some requests to finish).
109
110	See C<nreqs> for an example.
111
112	=item IO::AIO::nreqs
113
114	Returns the number of requests currently outstanding.
115
116	Example: wait till there are no outstanding requests anymore:
117
118	IO::AIO::poll_wait, IO::AIO::poll_cb
119	while IO::AIO::nreqs;
120
121	=item aio_open $pathname, $flags, $mode, $callback	211	=item aio_open $pathname, $flags, $mode, $callback->($fh)
122		212
123	Asynchronously open or create a file and call the callback with the	213	Asynchronously open or create a file and call the callback with a newly
124	filedescriptor (NOT a perl filehandle, sorry for that, but watch out, this	214	created filehandle for the file.
125	might change in the future).
126		215
127	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,	216	The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
128	for an explanation.	217	for an explanation.
129		218
130	The C<$mode> argument is a bitmask. See the C<Fcntl> module for a	219	The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
131	list. They are the same as used in C<sysopen>.	220	list. They are the same as used by C<sysopen>.
		221
		222	Likewise, C<$mode> specifies the mode of the newly created file, if it
		223	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
		224	except that it is mandatory (i.e. use C<0> if you don't create new files,
		225	and C<0666> or C<0777> if you do).
132		226
133	Example:	227	Example:
134		228
135	aio_open "/etc/passwd", O_RDONLY, 0, sub {	229	aio_open "/etc/passwd", O_RDONLY, 0, sub {
136	if ($_[0] >= 0) {	230	if ($_[0]) {
137	open my $fh, "<&$_[0]"; # create a copy for perl
138	aio_close $_[0], sub { }; # close the aio handle
139	print "open successful, fh is $fh\n";	231	print "open successful, fh is $_[0]\n";
140	...	232	...
141	} else {	233	} else {
142	die "open failed: $!\n";	234	die "open failed: $!\n";
143	}	235	}
144	};	236	};
145		237
146	=item aio_close $fh, $callback	238	=item aio_close $fh, $callback->($status)
147		239
148	Asynchronously close a file and call the callback with the result code.	240	Asynchronously close a file and call the callback with the result
		241	code. I<WARNING:> although accepted, you should not pass in a perl
		242	filehandle here, as perl will likely close the file descriptor another
		243	time when the filehandle is destroyed. Normally, you can safely call perls
		244	C<close> or just let filehandles go out of scope.
149		245
		246	This is supposed to be a bug in the API, so that might change. It's
		247	therefore best to avoid this function.
		248
150	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback	249	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
151		250
152	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback	251	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
153		252
154	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	253	Reads or writes C<length> bytes from the specified C<fh> and C<offset>
155	into the scalar given by C<data> and offset C<dataoffset> and calls the	254	into the scalar given by C<data> and offset C<dataoffset> and calls the
156	callback without the actual number of bytes read (or -1 on error, just	255	callback without the actual number of bytes read (or -1 on error, just
157	like the syscall).	256	like the syscall).
158		257
		258	The C<$data> scalar I<MUST NOT> be modified in any way while the request
		259	is outstanding. Modifying it can result in segfaults or WW3 (if the
		260	necessary/optional hardware is installed).
		261
159	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at	262	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
160	offset C<0> within the scalar:	263	offset C<0> within the scalar:
161		264
162	aio_read $fh, 7, 15, $buffer, 0, sub {	265	aio_read $fh, 7, 15, $buffer, 0, sub {
163	$_[0] >= 0 or die "read error: $!";	266	$_[0] > 0 or die "read error: $!";
164	print "read <$buffer>\n";	267	print "read $_[0] bytes: <$buffer>\n";
165	};	268	};
166		269
		270	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		271
		272	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
		273	reading at byte offset C<$in_offset>, and starts writing at the current
		274	file offset of C<$out_fh>. Because of that, it is not safe to issue more
		275	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
		276	other.
		277
		278	This call tries to make use of a native C<sendfile> syscall to provide
		279	zero-copy operation. For this to work, C<$out_fh> should refer to a
		280	socket, and C<$in_fh> should refer to mmap'able file.
		281
		282	If the native sendfile call fails or is not implemented, it will be
		283	emulated, so you can call C<aio_sendfile> on any type of filehandle
		284	regardless of the limitations of the operating system.
		285
		286	Please note, however, that C<aio_sendfile> can read more bytes from
		287	C<$in_fh> than are written, and there is no way to find out how many
		288	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only
		289	provides the number of bytes written to C<$out_fh>. Only if the result
		290	value equals C<$length> one can assume that C<$length> bytes have been
		291	read.
		292
167	=item aio_readahead $fh,$offset,$length, $callback	293	=item aio_readahead $fh,$offset,$length, $callback->($retval)
168		294
169	Asynchronously reads the specified byte range into the page cache, using
170	the C<readahead> syscall. If that syscall doesn't exist the status will be
171	C<-1> and C<$!> is set to ENOSYS.
172
173	readahead() populates the page cache with data from a file so that	295	C<aio_readahead> populates the page cache with data from a file so that
174	subsequent reads from that file will not block on disk I/O. The C<$offset>	296	subsequent reads from that file will not block on disk I/O. The C<$offset>
175	argument specifies the starting point from which data is to be read and	297	argument specifies the starting point from which data is to be read and
176	C<$length> specifies the number of bytes to be read. I/O is performed in	298	C<$length> specifies the number of bytes to be read. I/O is performed in
177	whole pages, so that offset is effectively rounded down to a page boundary	299	whole pages, so that offset is effectively rounded down to a page boundary
178	and bytes are read up to the next page boundary greater than or equal to	300	and bytes are read up to the next page boundary greater than or equal to
179	(off-set+length). aio_readahead() does not read beyond the end of the	301	(off-set+length). C<aio_readahead> does not read beyond the end of the
180	file. The current file offset of the file is left unchanged.	302	file. The current file offset of the file is left unchanged.
181		303
		304	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
		305	emulated by simply reading the data, which would have a similar effect.
		306
182	=item aio_stat $fh_or_path, $callback	307	=item aio_stat $fh_or_path, $callback->($status)
183		308
184	=item aio_lstat $fh, $callback	309	=item aio_lstat $fh, $callback->($status)
185		310
186	Works like perl's C<stat> or C<lstat> in void context. The callback will	311	Works like perl's C<stat> or C<lstat> in void context. The callback will
187	be called after the stat and the results will be available using C<stat _>	312	be called after the stat and the results will be available using C<stat _>
188	or C<-s _> etc...	313	or C<-s _> etc...
189		314
…		…
199	aio_stat "/etc/passwd", sub {	324	aio_stat "/etc/passwd", sub {
200	$_[0] and die "stat failed: $!";	325	$_[0] and die "stat failed: $!";
201	print "size is ", -s _, "\n";	326	print "size is ", -s _, "\n";
202	};	327	};
203		328
204	=item aio_unlink $pathname, $callback	329	=item aio_unlink $pathname, $callback->($status)
205		330
206	Asynchronously unlink (delete) a file and call the callback with the	331	Asynchronously unlink (delete) a file and call the callback with the
207	result code.	332	result code.
208		333
		334	=item aio_mknod $path, $mode, $dev, $callback->($status)
		335
		336	Asynchronously create a device node (or fifo). See mknod(2): the only
		337	portable value for C<$mode> is C<S_IFIFO> ored with permissions, and C<0>
		338	for C<$dev>.
		339
		340	=item aio_link $srcpath, $dstpath, $callback->($status)
		341
		342	Asynchronously create a new link to the existing object at C<$srcpath> at
		343	the path C<$dstpath> and call the callback with the result code.
		344
		345	=item aio_symlink $srcpath, $dstpath, $callback->($status)
		346
		347	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
		348	the path C<$dstpath> and call the callback with the result code.
		349
		350	=item aio_rename $srcpath, $dstpath, $callback->($status)
		351
		352	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
		353	rename(2) and call the callback with the result code.
		354
		355	=item aio_rmdir $pathname, $callback->($status)
		356
		357	Asynchronously rmdir (delete) a directory and call the callback with the
		358	result code.
		359
		360	=item aio_readdir $pathname, $callback->($entries)
		361
		362	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
		363	directory (i.e. opendir + readdir + closedir). The entries will not be
		364	sorted, and will B<NOT> include the C<.> and C<..> entries.
		365
		366	The callback a single argument which is either C<undef> or an array-ref
		367	with the filenames.
		368
		369	=item aio_copy $srcpath, $dstpath, $callback->($status)
		370
		371	Try to copy the I<file> (directories not supported as either source or
		372	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		373	the C<0> (error) or C<-1> ok.
		374
		375	This is a composite request that it creates the destination file with
		376	mode 0200 and copies the contents of the source file into it using
		377	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		378	uid/gid, in that order.
		379
		380	If an error occurs, the partial destination file will be unlinked, if
		381	possible, except when setting atime, mtime, access mode and uid/gid, where
		382	errors are being ignored.
		383
		384	=cut
		385
		386	sub aio_copy($$;$) {
		387	my ($src, $dst, $cb) = @_;
		388
		389	my $pri = aioreq_pri;
		390	my $grp = aio_group $cb;
		391
		392	aioreq_pri $pri;
		393	add $grp aio_open $src, O_RDONLY, 0, sub {
		394	if (my $src_fh = $_[0]) {
		395	my @stat = stat $src_fh;
		396
		397	aioreq_pri $pri;
		398	add $grp aio_open $dst, O_CREAT \| O_WRONLY \| O_TRUNC, 0200, sub {
		399	if (my $dst_fh = $_[0]) {
		400	aioreq_pri $pri;
		401	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		402	if ($_[0] == $stat[7]) {
		403	$grp->result (0);
		404	close $src_fh;
		405
		406	# those should not normally block. should. should.
		407	utime $stat[8], $stat[9], $dst;
		408	chmod $stat[2] & 07777, $dst_fh;
		409	chown $stat[4], $stat[5], $dst_fh;
		410	close $dst_fh;
		411	} else {
		412	$grp->result (-1);
		413	close $src_fh;
		414	close $dst_fh;
		415
		416	aioreq $pri;
		417	add $grp aio_unlink $dst;
		418	}
		419	};
		420	} else {
		421	$grp->result (-1);
		422	}
		423	},
		424
		425	} else {
		426	$grp->result (-1);
		427	}
		428	};
		429
		430	$grp
		431	}
		432
		433	=item aio_move $srcpath, $dstpath, $callback->($status)
		434
		435	Try to move the I<file> (directories not supported as either source or
		436	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		437	the C<0> (error) or C<-1> ok.
		438
		439	This is a composite request that tries to rename(2) the file first. If
		440	rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if
		441	that is successful, unlinking the C<$srcpath>.
		442
		443	=cut
		444
		445	sub aio_move($$;$) {
		446	my ($src, $dst, $cb) = @_;
		447
		448	my $pri = aioreq_pri;
		449	my $grp = aio_group $cb;
		450
		451	aioreq_pri $pri;
		452	add $grp aio_rename $src, $dst, sub {
		453	if ($_[0] && $! == EXDEV) {
		454	aioreq_pri $pri;
		455	add $grp aio_copy $src, $dst, sub {
		456	$grp->result ($_[0]);
		457
		458	if (!$_[0]) {
		459	aioreq_pri $pri;
		460	add $grp aio_unlink $src;
		461	}
		462	};
		463	} else {
		464	$grp->result ($_[0]);
		465	}
		466	};
		467
		468	$grp
		469	}
		470
		471	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
		472
		473	Scans a directory (similar to C<aio_readdir>) but additionally tries to
		474	efficiently separate the entries of directory C<$path> into two sets of
		475	names, directories you can recurse into (directories), and ones you cannot
		476	recurse into (everything else, including symlinks to directories).
		477
		478	C<aio_scandir> is a composite request that creates of many sub requests_
		479	C<$maxreq> specifies the maximum number of outstanding aio requests that
		480	this function generates. If it is C<< <= 0 >>, then a suitable default
		481	will be chosen (currently 4).
		482
		483	On error, the callback is called without arguments, otherwise it receives
		484	two array-refs with path-relative entry names.
		485
		486	Example:
		487
		488	aio_scandir $dir, 0, sub {
		489	my ($dirs, $nondirs) = @_;
		490	print "real directories: @$dirs\n";
		491	print "everything else: @$nondirs\n";
		492	};
		493
		494	Implementation notes.
		495
		496	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
		497
		498	After reading the directory, the modification time, size etc. of the
		499	directory before and after the readdir is checked, and if they match (and
		500	isn't the current time), the link count will be used to decide how many
		501	entries are directories (if >= 2). Otherwise, no knowledge of the number
		502	of subdirectories will be assumed.
		503
		504	Then entries will be sorted into likely directories (everything without
		505	a non-initial dot currently) and likely non-directories (everything
		506	else). Then every entry plus an appended C</.> will be C<stat>'ed,
		507	likely directories first. If that succeeds, it assumes that the entry
		508	is a directory or a symlink to directory (which will be checked
		509	seperately). This is often faster than stat'ing the entry itself because
		510	filesystems might detect the type of the entry without reading the inode
		511	data (e.g. ext2fs filetype feature).
		512
		513	If the known number of directories (link count - 2) has been reached, the
		514	rest of the entries is assumed to be non-directories.
		515
		516	This only works with certainty on POSIX (= UNIX) filesystems, which
		517	fortunately are the vast majority of filesystems around.
		518
		519	It will also likely work on non-POSIX filesystems with reduced efficiency
		520	as those tend to return 0 or 1 as link counts, which disables the
		521	directory counting heuristic.
		522
		523	=cut
		524
		525	sub aio_scandir($$$) {
		526	my ($path, $maxreq, $cb) = @_;
		527
		528	my $pri = aioreq_pri;
		529
		530	my $grp = aio_group $cb;
		531
		532	$maxreq = 4 if $maxreq <= 0;
		533
		534	# stat once
		535	aioreq_pri $pri;
		536	add $grp aio_stat $path, sub {
		537	return $grp->result () if $_[0];
		538	my $now = time;
		539	my $hash1 = join ":", (stat _)[0,1,3,7,9];
		540
		541	# read the directory entries
		542	aioreq_pri $pri;
		543	add $grp aio_readdir $path, sub {
		544	my $entries = shift
		545	or return $grp->result ();
		546
		547	# stat the dir another time
		548	aioreq_pri $pri;
		549	add $grp aio_stat $path, sub {
		550	my $hash2 = join ":", (stat _)[0,1,3,7,9];
		551
		552	my $ndirs;
		553
		554	# take the slow route if anything looks fishy
		555	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
		556	$ndirs = -1;
		557	} else {
		558	# if nlink == 2, we are finished
		559	# on non-posix-fs's, we rely on nlink < 2
		560	$ndirs = (stat _)[3] - 2
		561	or return $grp->result ([], $entries);
		562	}
		563
		564	# sort into likely dirs and likely nondirs
		565	# dirs == files without ".", short entries first
		566	$entries = [map $_->[0],
		567	sort { $b->[1] cmp $a->[1] }
		568	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
		569	@$entries];
		570
		571	my (@dirs, @nondirs);
		572
		573	my $statgrp = add $grp aio_group sub {
		574	$grp->result (\@dirs, \@nondirs);
		575	};
		576
		577	limit $statgrp $maxreq;
		578	feed $statgrp sub {
		579	return unless @$entries;
		580	my $entry = pop @$entries;
		581
		582	aioreq_pri $pri;
		583	add $statgrp aio_stat "$path/$entry/.", sub {
		584	if ($_[0] < 0) {
		585	push @nondirs, $entry;
		586	} else {
		587	# need to check for real directory
		588	aioreq_pri $pri;
		589	add $statgrp aio_lstat "$path/$entry", sub {
		590	if (-d _) {
		591	push @dirs, $entry;
		592
		593	unless (--$ndirs) {
		594	push @nondirs, @$entries;
		595	feed $statgrp;
		596	}
		597	} else {
		598	push @nondirs, $entry;
		599	}
		600	}
		601	}
		602	};
		603	};
		604	};
		605	};
		606	};
		607
		608	$grp
		609	}
		610
209	=item aio_fsync $fh, $callback	611	=item aio_fsync $fh, $callback->($status)
210		612
211	Asynchronously call fsync on the given filehandle and call the callback	613	Asynchronously call fsync on the given filehandle and call the callback
212	with the fsync result code.	614	with the fsync result code.
213		615
214	=item aio_fdatasync $fh, $callback	616	=item aio_fdatasync $fh, $callback->($status)
215		617
216	Asynchronously call fdatasync on the given filehandle and call the	618	Asynchronously call fdatasync on the given filehandle and call the
217	callback with the fdatasync result code.	619	callback with the fdatasync result code.
218		620
		621	If this call isn't available because your OS lacks it or it couldn't be
		622	detected, it will be emulated by calling C<fsync> instead.
		623
		624	=item aio_group $callback->(...)
		625
		626	This is a very special aio request: Instead of doing something, it is a
		627	container for other aio requests, which is useful if you want to bundle
		628	many requests into a single, composite, request with a definite callback
		629	and the ability to cancel the whole request with its subrequests.
		630
		631	Returns an object of class L<IO::AIO::GRP>. See its documentation below
		632	for more info.
		633
		634	Example:
		635
		636	my $grp = aio_group sub {
		637	print "all stats done\n";
		638	};
		639
		640	add $grp
		641	(aio_stat ...),
		642	(aio_stat ...),
		643	...;
		644
		645	=item aio_nop $callback->()
		646
		647	This is a special request - it does nothing in itself and is only used for
		648	side effects, such as when you want to add a dummy request to a group so
		649	that finishing the requests in the group depends on executing the given
		650	code.
		651
		652	While this request does nothing, it still goes through the execution
		653	phase and still requires a worker thread. Thus, the callback will not
		654	be executed immediately but only after other requests in the queue have
		655	entered their execution phase. This can be used to measure request
		656	latency.
		657
		658	=item IO::AIO::aio_busy $fractional_seconds, $callback->() NOT EXPORTED
		659
		660	Mainly used for debugging and benchmarking, this aio request puts one of
		661	the request workers to sleep for the given time.
		662
		663	While it is theoretically handy to have simple I/O scheduling requests
		664	like sleep and file handle readable/writable, the overhead this creates is
		665	immense (it blocks a thread for a long time) so do not use this function
		666	except to put your application under artificial I/O pressure.
		667
		668	=back
		669
		670	=head2 IO::AIO::REQ CLASS
		671
		672	All non-aggregate C<aio_*> functions return an object of this class when
		673	called in non-void context.
		674
		675	=over 4
		676
		677	=item cancel $req
		678
		679	Cancels the request, if possible. Has the effect of skipping execution
		680	when entering the B<execute> state and skipping calling the callback when
		681	entering the the B<result> state, but will leave the request otherwise
		682	untouched. That means that requests that currently execute will not be
		683	stopped and resources held by the request will not be freed prematurely.
		684
		685	=item cb $req $callback->(...)
		686
		687	Replace (or simply set) the callback registered to the request.
		688
		689	=back
		690
		691	=head2 IO::AIO::GRP CLASS
		692
		693	This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
		694	objects of this class, too.
		695
		696	A IO::AIO::GRP object is a special request that can contain multiple other
		697	aio requests.
		698
		699	You create one by calling the C<aio_group> constructing function with a
		700	callback that will be called when all contained requests have entered the
		701	C<done> state:
		702
		703	my $grp = aio_group sub {
		704	print "all requests are done\n";
		705	};
		706
		707	You add requests by calling the C<add> method with one or more
		708	C<IO::AIO::REQ> objects:
		709
		710	$grp->add (aio_unlink "...");
		711
		712	add $grp aio_stat "...", sub {
		713	$_[0] or return $grp->result ("error");
		714
		715	# add another request dynamically, if first succeeded
		716	add $grp aio_open "...", sub {
		717	$grp->result ("ok");
		718	};
		719	};
		720
		721	This makes it very easy to create composite requests (see the source of
		722	C<aio_move> for an application) that work and feel like simple requests.
		723
		724	=over 4
		725
		726	=item * The IO::AIO::GRP objects will be cleaned up during calls to
		727	C<IO::AIO::poll_cb>, just like any other request.
		728
		729	=item * They can be canceled like any other request. Canceling will cancel not
		730	only the request itself, but also all requests it contains.
		731
		732	=item * They can also can also be added to other IO::AIO::GRP objects.
		733
		734	=item * You must not add requests to a group from within the group callback (or
		735	any later time).
		736
		737	=back
		738
		739	Their lifetime, simplified, looks like this: when they are empty, they
		740	will finish very quickly. If they contain only requests that are in the
		741	C<done> state, they will also finish. Otherwise they will continue to
		742	exist.
		743
		744	That means after creating a group you have some time to add requests. And
		745	in the callbacks of those requests, you can add further requests to the
		746	group. And only when all those requests have finished will the the group
		747	itself finish.
		748
		749	=over 4
		750
		751	=item add $grp ...
		752
		753	=item $grp->add (...)
		754
		755	Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
		756	be added, including other groups, as long as you do not create circular
		757	dependencies.
		758
		759	Returns all its arguments.
		760
		761	=item $grp->cancel_subs
		762
		763	Cancel all subrequests and clears any feeder, but not the group request
		764	itself. Useful when you queued a lot of events but got a result early.
		765
		766	=item $grp->result (...)
		767
		768	Set the result value(s) that will be passed to the group callback when all
		769	subrequests have finished and set thre groups errno to the current value
		770	of errno (just like calling C<errno> without an error number). By default,
		771	no argument will be passed and errno is zero.
		772
		773	=item $grp->errno ([$errno])
		774
		775	Sets the group errno value to C<$errno>, or the current value of errno
		776	when the argument is missing.
		777
		778	Every aio request has an associated errno value that is restored when
		779	the callback is invoked. This method lets you change this value from its
		780	default (0).
		781
		782	Calling C<result> will also set errno, so make sure you either set C<$!>
		783	before the call to C<result>, or call c<errno> after it.
		784
		785	=item feed $grp $callback->($grp)
		786
		787	Sets a feeder/generator on this group: every group can have an attached
		788	generator that generates requests if idle. The idea behind this is that,
		789	although you could just queue as many requests as you want in a group,
		790	this might starve other requests for a potentially long time. For
		791	example, C<aio_scandir> might generate hundreds of thousands C<aio_stat>
		792	requests, delaying any later requests for a long time.
		793
		794	To avoid this, and allow incremental generation of requests, you can
		795	instead a group and set a feeder on it that generates those requests. The
		796	feed callback will be called whenever there are few enough (see C<limit>,
		797	below) requests active in the group itself and is expected to queue more
		798	requests.
		799
		800	The feed callback can queue as many requests as it likes (i.e. C<add> does
		801	not impose any limits).
		802
		803	If the feed does not queue more requests when called, it will be
		804	automatically removed from the group.
		805
		806	If the feed limit is C<0>, it will be set to C<2> automatically.
		807
		808	Example:
		809
		810	# stat all files in @files, but only ever use four aio requests concurrently:
		811
		812	my $grp = aio_group sub { print "finished\n" };
		813	limit $grp 4;
		814	feed $grp sub {
		815	my $file = pop @files
		816	or return;
		817
		818	add $grp aio_stat $file, sub { ... };
		819	};
		820
		821	=item limit $grp $num
		822
		823	Sets the feeder limit for the group: The feeder will be called whenever
		824	the group contains less than this many requests.
		825
		826	Setting the limit to C<0> will pause the feeding process.
		827
		828	=back
		829
		830	=head2 SUPPORT FUNCTIONS
		831
		832	=over 4
		833
		834	=item $fileno = IO::AIO::poll_fileno
		835
		836	Return the I<request result pipe file descriptor>. This filehandle must be
		837	polled for reading by some mechanism outside this module (e.g. Event or
		838	select, see below or the SYNOPSIS). If the pipe becomes readable you have
		839	to call C<poll_cb> to check the results.
		840
		841	See C<poll_cb> for an example.
		842
		843	=item IO::AIO::poll_cb
		844
		845	Process all outstanding events on the result pipe. You have to call this
		846	regularly. Returns the number of events processed. Returns immediately
		847	when no events are outstanding.
		848
		849	If not all requests were processed for whatever reason, the filehandle
		850	will still be ready when C<poll_cb> returns.
		851
		852	Example: Install an Event watcher that automatically calls
		853	IO::AIO::poll_cb with high priority:
		854
		855	Event->io (fd => IO::AIO::poll_fileno,
		856	poll => 'r', async => 1,
		857	cb => \&IO::AIO::poll_cb);
		858
		859	=item IO::AIO::poll_some $max_requests
		860
		861	Similar to C<poll_cb>, but only processes up to C<$max_requests> requests
		862	at a time.
		863
		864	Useful if you want to ensure some level of interactiveness when perl is
		865	not fast enough to process all requests in time.
		866
		867	Example: Install an Event watcher that automatically calls
		868	IO::AIO::poll_some with low priority, to ensure that other parts of the
		869	program get the CPU sometimes even under high AIO load.
		870
		871	Event->io (fd => IO::AIO::poll_fileno,
		872	poll => 'r', nice => 1,
		873	cb => sub { IO::AIO::poll_some 256 });
		874
		875	=item IO::AIO::poll_wait
		876
		877	Wait till the result filehandle becomes ready for reading (simply does a
		878	C<select> on the filehandle. This is useful if you want to synchronously wait
		879	for some requests to finish).
		880
		881	See C<nreqs> for an example.
		882
		883	=item IO::AIO::nreqs
		884
		885	Returns the number of requests currently in the ready, execute or pending
		886	states (i.e. for which their callback has not been invoked yet).
		887
		888	Example: wait till there are no outstanding requests anymore:
		889
		890	IO::AIO::poll_wait, IO::AIO::poll_cb
		891	while IO::AIO::nreqs;
		892
		893	=item IO::AIO::nready
		894
		895	Returns the number of requests currently in the ready state (not yet
		896	executed).
		897
		898	=item IO::AIO::npending
		899
		900	Returns the number of requests currently in the pending state (executed,
		901	but not yet processed by poll_cb).
		902
		903	=item IO::AIO::flush
		904
		905	Wait till all outstanding AIO requests have been handled.
		906
		907	Strictly equivalent to:
		908
		909	IO::AIO::poll_wait, IO::AIO::poll_cb
		910	while IO::AIO::nreqs;
		911
		912	=item IO::AIO::poll
		913
		914	Waits until some requests have been handled.
		915
		916	Strictly equivalent to:
		917
		918	IO::AIO::poll_wait, IO::AIO::poll_cb
		919	if IO::AIO::nreqs;
		920
		921	=item IO::AIO::min_parallel $nthreads
		922
		923	Set the minimum number of AIO threads to C<$nthreads>. The current
		924	default is C<8>, which means eight asynchronous operations can execute
		925	concurrently at any one time (the number of outstanding requests,
		926	however, is unlimited).
		927
		928	IO::AIO starts threads only on demand, when an AIO request is queued and
		929	no free thread exists.
		930
		931	It is recommended to keep the number of threads relatively low, as some
		932	Linux kernel versions will scale negatively with the number of threads
		933	(higher parallelity => MUCH higher latency). With current Linux 2.6
		934	versions, 4-32 threads should be fine.
		935
		936	Under most circumstances you don't need to call this function, as the
		937	module selects a default that is suitable for low to moderate load.
		938
		939	=item IO::AIO::max_parallel $nthreads
		940
		941	Sets the maximum number of AIO threads to C<$nthreads>. If more than the
		942	specified number of threads are currently running, this function kills
		943	them. This function blocks until the limit is reached.
		944
		945	While C<$nthreads> are zero, aio requests get queued but not executed
		946	until the number of threads has been increased again.
		947
		948	This module automatically runs C<max_parallel 0> at program end, to ensure
		949	that all threads are killed and that there are no outstanding requests.
		950
		951	Under normal circumstances you don't need to call this function.
		952
		953	=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs
		954
		955	This is a very bad function to use in interactive programs because it
		956	blocks, and a bad way to reduce concurrency because it is inexact: Better
		957	use an C<aio_group> together with a feed callback.
		958
		959	Sets the maximum number of outstanding requests to C<$nreqs>. If you
		960	to queue up more than this number of requests, the next call to the
		961	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
		962	function will block until the limit is no longer exceeded.
		963
		964	The default value is very large, so there is no practical limit on the
		965	number of outstanding requests.
		966
		967	You can still queue as many requests as you want. Therefore,
		968	C<max_oustsanding> is mainly useful in simple scripts (with low values) or
		969	as a stop gap to shield against fatal memory overflow (with large values).
		970
		971	=back
		972
219	=cut	973	=cut
220		974
		975	# support function to convert a fd into a perl filehandle
		976	sub _fd2fh {
		977	return undef if $_[0] < 0;
		978
		979	# try to generate nice filehandles
		980	my $sym = "IO::AIO::fd#$_[0]";
		981	local *$sym;
		982
		983	open *$sym, "+<&=$_[0]" # usually works under any unix
		984	or open *$sym, "<&=$_[0]" # cygwin needs this
		985	or open *$sym, ">&=$_[0]" # or this
		986	or return undef;
		987
		988	*$sym
		989	}
		990
221	min_parallel 4;	991	min_parallel 8;
222		992
223	END {	993	END {
224	max_parallel 0;	994	flush;
225	}	995	};
226		996
227	1;	997	1;
228		998
229	=back	999	=head2 FORK BEHAVIOUR
230		1000
		1001	This module should do "the right thing" when the process using it forks:
		1002
		1003	Before the fork, IO::AIO enters a quiescent state where no requests
		1004	can be added in other threads and no results will be processed. After
		1005	the fork the parent simply leaves the quiescent state and continues
		1006	request/result processing, while the child frees the request/result queue
		1007	(so that the requests started before the fork will only be handled in the
		1008	parent). Threads will be started on demand until the limit set in the
		1009	parent process has been reached again.
		1010
		1011	In short: the parent will, after a short pause, continue as if fork had
		1012	not been called, while the child will act as if IO::AIO has not been used
		1013	yet.
		1014
		1015	=head2 MEMORY USAGE
		1016
		1017	Per-request usage:
		1018
		1019	Each aio request uses - depending on your architecture - around 100-200
		1020	bytes of memory. In addition, stat requests need a stat buffer (possibly
		1021	a few hundred bytes), readdir requires a result buffer and so on. Perl
		1022	scalars and other data passed into aio requests will also be locked and
		1023	will consume memory till the request has entered the done state.
		1024
		1025	This is now awfully much, so queuing lots of requests is not usually a
		1026	problem.
		1027
		1028	Per-thread usage:
		1029
		1030	In the execution phase, some aio requests require more memory for
		1031	temporary buffers, and each thread requires a stack and other data
		1032	structures (usually around 16k-128k, depending on the OS).
		1033
231	=head1 BUGS	1034	=head1 KNOWN BUGS
232		1035
233	- aio_open gives a fd, but all other functions expect a perl filehandle.	1036	Known bugs will be fixed in the next release.
234		1037
235	=head1 SEE ALSO	1038	=head1 SEE ALSO
236		1039
237	L<Coro>, L<Linux::AIO>.	1040	L<Coro::AIO>.
238		1041
239	=head1 AUTHOR	1042	=head1 AUTHOR
240		1043
241	Marc Lehmann <schmorp@schmorp.de>	1044	Marc Lehmann <schmorp@schmorp.de>
242	http://home.schmorp.de/	1045	http://home.schmorp.de/

Diff Legend

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

Comparing IO-AIO/AIO.pm (file contents): Revision 1.1 by root, Sun Jul 10 17:07:44 2005 UTC vs. Revision 1.82 by root, Fri Oct 27 20:10:06 2006 UTC

Diff Legend

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.1 by root, Sun Jul 10 17:07:44 2005 UTC vs.
Revision 1.82 by root, Fri Oct 27 20:10:06 2006 UTC