ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.81 by root, Fri Oct 27 19:17:23 2006 UTC vs.
Revision 1.219 by root, Tue Mar 27 18:54:45 2012 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use IO::AIO;	7	use IO::AIO;
8		8
9	aio_open "/etc/passwd", O_RDONLY, 0, sub {	9	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
10	my ($fh) = @_;	10	my $fh = shift
		11	or die "/etc/passwd: $!";
11	...	12	...
12	};	13	};
13		14
14	aio_unlink "/tmp/file", sub { };	15	aio_unlink "/tmp/file", sub { };
15		16
…		…
25	$req->cancel; # cancel request if still in queue	26	$req->cancel; # cancel request if still in queue
26		27
27	my $grp = aio_group sub { print "all stats done\n" };	28	my $grp = aio_group sub { print "all stats done\n" };
28	add $grp aio_stat "..." for ...;	29	add $grp aio_stat "..." for ...;
29		30
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
51	=head1 DESCRIPTION	31	=head1 DESCRIPTION
52		32
53	This module implements asynchronous I/O using whatever means your	33	This module implements asynchronous I/O using whatever means your
54	operating system supports.	34	operating system supports. It is implemented as an interface to C<libeio>
		35	(L<http://software.schmorp.de/pkg/libeio.html>).
		36
		37	Asynchronous means that operations that can normally block your program
		38	(e.g. reading from disk) will be done asynchronously: the operation
		39	will still block, but you can do something else in the meantime. This
		40	is extremely useful for programs that need to stay interactive even
		41	when doing heavy I/O (GUI programs, high performance network servers
		42	etc.), but can also be used to easily do operations in parallel that are
		43	normally done sequentially, e.g. stat'ing many files, which is much faster
		44	on a RAID volume or over NFS when you do a number of stat operations
		45	concurrently.
		46
		47	While most of this works on all types of file descriptors (for
		48	example sockets), using these functions on file descriptors that
		49	support nonblocking operation (again, sockets, pipes etc.) is
		50	very inefficient. Use an event loop for that (such as the L<EV>
		51	module): IO::AIO will naturally fit into such an event loop itself.
55		52
56	In this version, a number of threads are started that execute your	53	In this version, a number of threads are started that execute your
57	requests and signal their completion. You don't need thread support	54	requests and signal their completion. You don't need thread support
58	in perl, and the threads created by this module will not be visible	55	in perl, and the threads created by this module will not be visible
59	to perl. In the future, this module might make use of the native aio	56	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	57	functions available on many operating systems. However, they are often
61	not well-supported or restricted (Linux doesn't allow them on normal	58	not well-supported or restricted (GNU/Linux doesn't allow them on normal
62	files currently, for example), and they would only support aio_read and	59	files currently, for example), and they would only support aio_read and
63	aio_write, so the remaining functionality would have to be implemented	60	aio_write, so the remaining functionality would have to be implemented
64	using threads anyway.	61	using threads anyway.
65		62
66	Although the module will work with in the presence of other (Perl-)	63	Although the module will work in the presence of other (Perl-) threads,
67	threads, it is currently not reentrant in any way, so use appropriate	64	it is currently not reentrant in any way, so use appropriate locking
68	locking yourself, always call C<poll_cb> from within the same thread, or	65	yourself, always call C<poll_cb> from within the same thread, or never
69	never call C<poll_cb> (or other C<aio_> functions) recursively.	66	call C<poll_cb> (or other C<aio_> functions) recursively.
		67
		68	=head2 EXAMPLE
		69
		70	This is a simple example that uses the EV module and loads
		71	F</etc/passwd> asynchronously:
		72
		73	use Fcntl;
		74	use EV;
		75	use IO::AIO;
		76
		77	# register the IO::AIO callback with EV
		78	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		79
		80	# queue the request to open /etc/passwd
		81	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
		82	my $fh = shift
		83	or die "error while opening: $!";
		84
		85	# stat'ing filehandles is generally non-blocking
		86	my $size = -s $fh;
		87
		88	# queue a request to read the file
		89	my $contents;
		90	aio_read $fh, 0, $size, $contents, 0, sub {
		91	$_[0] == $size
		92	or die "short read: $!";
		93
		94	close $fh;
		95
		96	# file contents now in $contents
		97	print $contents;
		98
		99	# exit event loop and program
		100	EV::unloop;
		101	};
		102	};
		103
		104	# possibly queue up other requests, or open GUI windows,
		105	# check for sockets etc. etc.
		106
		107	# process events as long as there are some:
		108	EV::loop;
70		109
71	=head1 REQUEST ANATOMY AND LIFETIME	110	=head1 REQUEST ANATOMY AND LIFETIME
72		111
73	Every C<aio_*> function creates a request. which is a C data structure not	112	Every C<aio_*> function creates a request. which is a C data structure not
74	directly visible to Perl.	113	directly visible to Perl.
…		…
116	Request has reached the end of its lifetime and holds no resources anymore	155	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	156	(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	157	aio request is severed and calling its methods will either do nothing or
119	result in a runtime error).	158	result in a runtime error).
120		159
		160	=back
		161
121	=cut	162	=cut
122		163
123	package IO::AIO;	164	package IO::AIO;
124		165
125	no warnings;	166	use Carp ();
126	use strict 'vars';	167
		168	use common::sense;
127		169
128	use base 'Exporter';	170	use base 'Exporter';
129		171
130	BEGIN {	172	BEGIN {
131	our $VERSION = '2.0';	173	our $VERSION = '4.12';
132		174
133	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat	175	our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close
134	aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink	176	aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
135	aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move	177	aio_scandir aio_symlink aio_readlink aio_realpath aio_sync
		178	aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range aio_fallocate
		179	aio_pathsync aio_readahead
		180	aio_rename aio_link aio_move aio_copy aio_group
		181	aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
		182	aio_chmod aio_utime aio_truncate
		183	aio_msync aio_mtouch aio_mlock aio_mlockall
		184	aio_statvfs
136	aio_group aio_nop);	185	aio_wd);
		186
137	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));	187	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
138	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush	188	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
139	min_parallel max_parallel nreqs nready npending);	189	min_parallel max_parallel max_idle idle_timeout
		190	nreqs nready npending nthreads
		191	max_poll_time max_poll_reqs
		192	sendfile fadvise madvise
		193	mmap munmap munlock munlockall);
		194
		195	push @AIO_REQ, qw(aio_busy); # not exported
140		196
141	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';	197	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
142		198
143	require XSLoader;	199	require XSLoader;
144	XSLoader::load ("IO::AIO", $VERSION);	200	XSLoader::load ("IO::AIO", $VERSION);
145	}	201	}
146		202
147	=head1 FUNCTIONS	203	=head1 FUNCTIONS
148		204
149	=head2 AIO FUNCTIONS	205	=head2 QUICK OVERVIEW
		206
		207	This section simply lists the prototypes of the most important functions
		208	for quick reference. See the following sections for function-by-function
		209	documentation.
		210
		211	aio_wd $pathname, $callback->($wd)
		212	aio_open $pathname, $flags, $mode, $callback->($fh)
		213	aio_close $fh, $callback->($status)
		214	aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		215	aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
		216	aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
		217	aio_readahead $fh,$offset,$length, $callback->($retval)
		218	aio_stat $fh_or_path, $callback->($status)
		219	aio_lstat $fh, $callback->($status)
		220	aio_statvfs $fh_or_path, $callback->($statvfs)
		221	aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		222	aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		223	aio_truncate $fh_or_path, $offset, $callback->($status)
		224	aio_chmod $fh_or_path, $mode, $callback->($status)
		225	aio_unlink $pathname, $callback->($status)
		226	aio_mknod $pathname, $mode, $dev, $callback->($status)
		227	aio_link $srcpath, $dstpath, $callback->($status)
		228	aio_symlink $srcpath, $dstpath, $callback->($status)
		229	aio_readlink $pathname, $callback->($link)
		230	aio_realpath $pathname, $callback->($link)
		231	aio_rename $srcpath, $dstpath, $callback->($status)
		232	aio_mkdir $pathname, $mode, $callback->($status)
		233	aio_rmdir $pathname, $callback->($status)
		234	aio_readdir $pathname, $callback->($entries)
		235	aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		236	IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
		237	IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
		238	aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		239	aio_load $pathname, $data, $callback->($status)
		240	aio_copy $srcpath, $dstpath, $callback->($status)
		241	aio_move $srcpath, $dstpath, $callback->($status)
		242	aio_rmtree $pathname, $callback->($status)
		243	aio_sync $callback->($status)
		244	aio_syncfs $fh, $callback->($status)
		245	aio_fsync $fh, $callback->($status)
		246	aio_fdatasync $fh, $callback->($status)
		247	aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		248	aio_pathsync $pathname, $callback->($status)
		249	aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		250	aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		251	aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		252	aio_mlockall $flags, $callback->($status)
		253	aio_group $callback->(...)
		254	aio_nop $callback->()
		255
		256	$prev_pri = aioreq_pri [$pri]
		257	aioreq_nice $pri_adjust
		258
		259	IO::AIO::poll_wait
		260	IO::AIO::poll_cb
		261	IO::AIO::poll
		262	IO::AIO::flush
		263	IO::AIO::max_poll_reqs $nreqs
		264	IO::AIO::max_poll_time $seconds
		265	IO::AIO::min_parallel $nthreads
		266	IO::AIO::max_parallel $nthreads
		267	IO::AIO::max_idle $nthreads
		268	IO::AIO::idle_timeout $seconds
		269	IO::AIO::max_outstanding $maxreqs
		270	IO::AIO::nreqs
		271	IO::AIO::nready
		272	IO::AIO::npending
		273
		274	IO::AIO::sendfile $ofh, $ifh, $offset, $count
		275	IO::AIO::fadvise $fh, $offset, $len, $advice
		276	IO::AIO::madvise $scalar, $offset, $length, $advice
		277	IO::AIO::mprotect $scalar, $offset, $length, $protect
		278	IO::AIO::munlock $scalar, $offset = 0, $length = undef
		279	IO::AIO::munlockall
		280
		281	=head2 API NOTES
150		282
151	All the C<aio_*> calls are more or less thin wrappers around the syscall	283	All the C<aio_*> calls are more or less thin wrappers around the syscall
152	with the same name (sans C<aio_>). The arguments are similar or identical,	284	with the same name (sans C<aio_>). The arguments are similar or identical,
153	and they all accept an additional (and optional) C<$callback> argument	285	and they all accept an additional (and optional) C<$callback> argument
154	which must be a code reference. This code reference will get called with	286	which must be a code reference. This code reference will be called after
155	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	287	the syscall has been executed in an asynchronous fashion. The results
156	perl, which usually delivers "false") as it's sole argument when the given	288	of the request will be passed as arguments to the callback (and, if an
157	syscall has been executed asynchronously.	289	error occured, in C<$!>) - for most requests the syscall return code (e.g.
		290	most syscalls return C<-1> on error, unlike perl, which usually delivers
		291	"false").
		292
		293	Some requests (such as C<aio_readdir>) pass the actual results and
		294	communicate failures by passing C<undef>.
158		295
159	All functions expecting a filehandle keep a copy of the filehandle	296	All functions expecting a filehandle keep a copy of the filehandle
160	internally until the request has finished.	297	internally until the request has finished.
161		298
162	All requests return objects of type L<IO::AIO::REQ> that allow further	299	All functions return request objects of type L<IO::AIO::REQ> that allow
163	manipulation of those requests while they are in-flight.	300	further manipulation of those requests while they are in-flight.
164		301
165	The pathnames you pass to these routines I<must> be absolute and	302	The pathnames you pass to these routines I<should> be absolute. The
166	encoded in byte form. The reason for the former is that at the time the	303	reason for this is that at the time the request is being executed, the
167	request is being executed, the current working directory could have	304	current working directory could have changed. Alternatively, you can
168	changed. Alternatively, you can make sure that you never change the	305	make sure that you never change the current working directory anywhere
169	current working directory.	306	in the program and then use relative paths. You can also take advantage
		307	of IO::AIOs working directory abstraction, that lets you specify paths
		308	relative to some previously-opened "working directory object" - see the
		309	description of the C<IO::AIO::WD> class later in this document.
170		310
171	To encode pathnames to byte form, either make sure you either: a)	311	To encode pathnames as octets, either make sure you either: a) always pass
172	always pass in filenames you got from outside (command line, readdir	312	in filenames you got from outside (command line, readdir etc.) without
173	etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode	313	tinkering, b) are in your native filesystem encoding, c) use the Encode
174	your pathnames to the locale (or other) encoding in effect in the user	314	module and encode your pathnames to the locale (or other) encoding in
175	environment, d) use Glib::filename_from_unicode on unicode filenames or e)	315	effect in the user environment, d) use Glib::filename_from_unicode on
176	use something else.	316	unicode filenames or e) use something else to ensure your scalar has the
		317	correct contents.
		318
		319	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		320	handles correctly whether it is set or not.
		321
		322	=head2 AIO REQUEST FUNCTIONS
177		323
178	=over 4	324	=over 4
179		325
180	=item $prev_pri = aioreq_pri [$pri]	326	=item $prev_pri = aioreq_pri [$pri]
181		327
…		…
201	aio_read $_[0], ..., sub {	347	aio_read $_[0], ..., sub {
202	...	348	...
203	};	349	};
204	};	350	};
205		351
		352
206	=item aioreq_nice $pri_adjust	353	=item aioreq_nice $pri_adjust
207		354
208	Similar to C<aioreq_pri>, but subtracts the given value from the current	355	Similar to C<aioreq_pri>, but subtracts the given value from the current
209	priority, so effects are cumulative.	356	priority, so the effect is cumulative.
		357
210		358
211	=item aio_open $pathname, $flags, $mode, $callback->($fh)	359	=item aio_open $pathname, $flags, $mode, $callback->($fh)
212		360
213	Asynchronously open or create a file and call the callback with a newly	361	Asynchronously open or create a file and call the callback with a newly
214	created filehandle for the file.	362	created filehandle for the file.
…		…
220	list. They are the same as used by C<sysopen>.	368	list. They are the same as used by C<sysopen>.
221		369
222	Likewise, C<$mode> specifies the mode of the newly created file, if it	370	Likewise, C<$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>,	371	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,	372	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).	373	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		374	by the umask in effect then the request is being executed, so better never
		375	change the umask.
226		376
227	Example:	377	Example:
228		378
229	aio_open "/etc/passwd", O_RDONLY, 0, sub {	379	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
230	if ($_[0]) {	380	if ($_[0]) {
231	print "open successful, fh is $_[0]\n";	381	print "open successful, fh is $_[0]\n";
232	...	382	...
233	} else {	383	} else {
234	die "open failed: $!\n";	384	die "open failed: $!\n";
235	}	385	}
236	};	386	};
237		387
		388	In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
		389	C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
		390	following POSIX and non-POSIX constants are available (missing ones on
		391	your system are, as usual, C<0>):
		392
		393	C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
		394	C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
		395	C<O_RSYNC>, C<O_SYNC> and C<O_TTY_INIT>.
		396
		397
238	=item aio_close $fh, $callback->($status)	398	=item aio_close $fh, $callback->($status)
239		399
240	Asynchronously close a file and call the callback with the result	400	Asynchronously close a file and call the callback with the result
241	code. I<WARNING:> although accepted, you should not pass in a perl	401	code.
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.
245		402
246	This is supposed to be a bug in the API, so that might change. It's	403	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
247	therefore best to avoid this function.	404	closing the file descriptor associated with the filehandle itself.
		405
		406	Therefore, C<aio_close> will not close the filehandle - instead it will
		407	use dup2 to overwrite the file descriptor with the write-end of a pipe
		408	(the pipe fd will be created on demand and will be cached).
		409
		410	Or in other words: the file descriptor will be closed, but it will not be
		411	free for reuse until the perl filehandle is closed.
		412
		413	=cut
248		414
249	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	415	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
250		416
251	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	417	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
252		418
253	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	419	Reads or writes C<$length> bytes from or to the specified C<$fh> and
254	into the scalar given by C<data> and offset C<dataoffset> and calls the	420	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset>
255	callback without the actual number of bytes read (or -1 on error, just	421	and calls the callback without the actual number of bytes read (or -1 on
256	like the syscall).	422	error, just like the syscall).
		423
		424	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		425	offset plus the actual number of bytes read.
		426
		427	If C<$offset> is undefined, then the current file descriptor offset will
		428	be used (and updated), otherwise the file descriptor offset will not be
		429	changed by these calls.
		430
		431	If C<$length> is undefined in C<aio_write>, use the remaining length of
		432	C<$data>.
		433
		434	If C<$dataoffset> is less than zero, it will be counted from the end of
		435	C<$data>.
257		436
258	The C<$data> scalar I<MUST NOT> be modified in any way while the request	437	The C<$data> scalar I<MUST NOT> be modified in any way while the request
259	is outstanding. Modifying it can result in segfaults or WW3 (if the	438	is outstanding. Modifying it can result in segfaults or World War III (if
260	necessary/optional hardware is installed).	439	the necessary/optional hardware is installed).
261		440
262	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	441	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
263	offset C<0> within the scalar:	442	offset C<0> within the scalar:
264		443
265	aio_read $fh, 7, 15, $buffer, 0, sub {	444	aio_read $fh, 7, 15, $buffer, 0, sub {
266	$_[0] > 0 or die "read error: $!";	445	$_[0] > 0 or die "read error: $!";
267	print "read $_[0] bytes: <$buffer>\n";	446	print "read $_[0] bytes: <$buffer>\n";
268	};	447	};
269		448
270	=item aio_move $srcpath, $dstpath, $callback->($status)
271
272	Try to move the I<file> (directories not supported as either source or
273	destination) from C<$srcpath> to C<$dstpath> and call the callback with
274	the C<0> (error) or C<-1> ok.
275
276	This is a composite request that tries to rename(2) the file first. If
277	rename files with C<EXDEV>, it creates the destination file with mode 0200
278	and copies the contents of the source file into it using C<aio_sendfile>,
279	followed by restoring atime, mtime, access mode and uid/gid, in that
280	order, and unlinking the C<$srcpath>.
281
282	If an error occurs, the partial destination file will be unlinked, if
283	possible, except when setting atime, mtime, access mode and uid/gid, where
284	errors are being ignored.
285
286	=cut
287
288	sub aio_move($$$) {
289	my ($src, $dst, $cb) = @_;
290
291	my $pri = aioreq_pri;
292	my $grp = aio_group $cb;
293
294	aioreq_pri $pri;
295	add $grp aio_rename $src, $dst, sub {
296	if ($_[0] && $! == EXDEV) {
297	aioreq_pri $pri;
298	add $grp aio_open $src, O_RDONLY, 0, sub {
299	if (my $src_fh = $_[0]) {
300	my @stat = stat $src_fh;
301
302	aioreq_pri $pri;
303	add $grp aio_open $dst, O_WRONLY, 0200, sub {
304	if (my $dst_fh = $_[0]) {
305	aioreq_pri $pri;
306	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
307	close $src_fh;
308
309	if ($_[0] == $stat[7]) {
310	utime $stat[8], $stat[9], $dst;
311	chmod $stat[2] & 07777, $dst_fh;
312	chown $stat[4], $stat[5], $dst_fh;
313	close $dst_fh;
314
315	aioreq_pri $pri;
316	add $grp aio_unlink $src, sub {
317	$grp->result ($_[0]);
318	};
319	} else {
320	my $errno = $!;
321	aioreq_pri $pri;
322	add $grp aio_unlink $dst, sub {
323	$! = $errno;
324	$grp->result (-1);
325	};
326	}
327	};
328	} else {
329	$grp->result (-1);
330	}
331	},
332
333	} else {
334	$grp->result (-1);
335	}
336	};
337	} else {
338	$grp->result ($_[0]);
339	}
340	};
341
342	$grp
343	}
344		449
345	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)	450	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
346		451
347	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts	452	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
348	reading at byte offset C<$in_offset>, and starts writing at the current	453	reading at byte offset C<$in_offset>, and starts writing at the current
349	file offset of C<$out_fh>. Because of that, it is not safe to issue more	454	file offset of C<$out_fh>. Because of that, it is not safe to issue more
350	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each	455	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
351	other.	456	other. The same C<$in_fh> works fine though, as this function does not
		457	move or use the file offset of C<$in_fh>.
352		458
		459	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		460	are written, and there is no way to find out how many more bytes have been
		461	read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
		462	number of bytes written to C<$out_fh>. Only if the result value equals
		463	C<$length> one can assume that C<$length> bytes have been read.
		464
		465	Unlike with other C<aio_> functions, it makes a lot of sense to use
		466	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		467	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		468	the socket I/O will be non-blocking. Note, however, that you can run
		469	into a trap where C<aio_sendfile> reads some data with readahead, then
		470	fails to write all data, and when the socket is ready the next time, the
		471	data in the cache is already lost, forcing C<aio_sendfile> to again hit
		472	the disk. Explicit C<aio_read> + C<aio_write> let's you better control
		473	resource usage.
		474
353	This call tries to make use of a native C<sendfile> syscall to provide	475	This call tries to make use of a native C<sendfile>-like syscall to
354	zero-copy operation. For this to work, C<$out_fh> should refer to a	476	provide zero-copy operation. For this to work, C<$out_fh> should refer to
355	socket, and C<$in_fh> should refer to mmap'able file.	477	a socket, and C<$in_fh> should refer to an mmap'able file.
356		478
357	If the native sendfile call fails or is not implemented, it will be	479	If a native sendfile cannot be found or it fails with C<ENOSYS>,
358	emulated, so you can call C<aio_sendfile> on any type of filehandle	480	C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
		481	C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
359	regardless of the limitations of the operating system.	482	type of filehandle regardless of the limitations of the operating system.
360		483
361	Please note, however, that C<aio_sendfile> can read more bytes from	484	As native sendfile syscalls (as practically any non-POSIX interface hacked
362	C<$in_fh> than are written, and there is no way to find out how many	485	together in a hurry to improve benchmark numbers) tend to be rather buggy
363	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only	486	on many systems, this implementation tries to work around some known bugs
364	provides the number of bytes written to C<$out_fh>. Only if the result	487	in Linux and FreeBSD kernels (probably others, too), but that might fail,
365	value equals C<$length> one can assume that C<$length> bytes have been	488	so you really really should check the return value of C<aio_sendfile> -
366	read.	489	fewre bytes than expected might have been transferred.
		490
367		491
368	=item aio_readahead $fh,$offset,$length, $callback->($retval)	492	=item aio_readahead $fh,$offset,$length, $callback->($retval)
369		493
370	C<aio_readahead> populates the page cache with data from a file so that	494	C<aio_readahead> populates the page cache with data from a file so that
371	subsequent reads from that file will not block on disk I/O. The C<$offset>	495	subsequent reads from that file will not block on disk I/O. The C<$offset>
…		…
377	file. The current file offset of the file is left unchanged.	501	file. The current file offset of the file is left unchanged.
378		502
379	If that syscall doesn't exist (likely if your OS isn't Linux) it will be	503	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
380	emulated by simply reading the data, which would have a similar effect.	504	emulated by simply reading the data, which would have a similar effect.
381		505
		506
382	=item aio_stat $fh_or_path, $callback->($status)	507	=item aio_stat $fh_or_path, $callback->($status)
383		508
384	=item aio_lstat $fh, $callback->($status)	509	=item aio_lstat $fh, $callback->($status)
385		510
386	Works like perl's C<stat> or C<lstat> in void context. The callback will	511	Works like perl's C<stat> or C<lstat> in void context. The callback will
…		…
391	for an explanation.	516	for an explanation.
392		517
393	Currently, the stats are always 64-bit-stats, i.e. instead of returning an	518	Currently, the stats are always 64-bit-stats, i.e. instead of returning an
394	error when stat'ing a large file, the results will be silently truncated	519	error when stat'ing a large file, the results will be silently truncated
395	unless perl itself is compiled with large file support.	520	unless perl itself is compiled with large file support.
		521
		522	To help interpret the mode and dev/rdev stat values, IO::AIO offers the
		523	following constants and functions (if not implemented, the constants will
		524	be C<0> and the functions will either C<croak> or fall back on traditional
		525	behaviour).
		526
		527	C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
		528	C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
		529	C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
396		530
397	Example: Print the length of F</etc/passwd>:	531	Example: Print the length of F</etc/passwd>:
398		532
399	aio_stat "/etc/passwd", sub {	533	aio_stat "/etc/passwd", sub {
400	$_[0] and die "stat failed: $!";	534	$_[0] and die "stat failed: $!";
401	print "size is ", -s _, "\n";	535	print "size is ", -s _, "\n";
402	};	536	};
403		537
		538
		539	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		540
		541	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		542	whether a file handle or path was passed.
		543
		544	On success, the callback is passed a hash reference with the following
		545	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		546	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		547	is passed.
		548
		549	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		550	C<ST_NOSUID>.
		551
		552	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		553	their correct value when available, or to C<0> on systems that do
		554	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		555	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		556	C<ST_NODIRATIME> and C<ST_RELATIME>.
		557
		558	Example: stat C</wd> and dump out the data if successful.
		559
		560	aio_statvfs "/wd", sub {
		561	my $f = $_[0]
		562	or die "statvfs: $!";
		563
		564	use Data::Dumper;
		565	say Dumper $f;
		566	};
		567
		568	# result:
		569	{
		570	bsize => 1024,
		571	bfree => 4333064312,
		572	blocks => 10253828096,
		573	files => 2050765568,
		574	flag => 4096,
		575	favail => 2042092649,
		576	bavail => 4333064312,
		577	ffree => 2042092649,
		578	namemax => 255,
		579	frsize => 1024,
		580	fsid => 1810
		581	}
		582
		583
		584	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		585
		586	Works like perl's C<utime> function (including the special case of $atime
		587	and $mtime being undef). Fractional times are supported if the underlying
		588	syscalls support them.
		589
		590	When called with a pathname, uses utimes(2) if available, otherwise
		591	utime(2). If called on a file descriptor, uses futimes(2) if available,
		592	otherwise returns ENOSYS, so this is not portable.
		593
		594	Examples:
		595
		596	# set atime and mtime to current time (basically touch(1)):
		597	aio_utime "path", undef, undef;
		598	# set atime to current time and mtime to beginning of the epoch:
		599	aio_utime "path", time, undef; # undef==0
		600
		601
		602	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		603
		604	Works like perl's C<chown> function, except that C<undef> for either $uid
		605	or $gid is being interpreted as "do not change" (but -1 can also be used).
		606
		607	Examples:
		608
		609	# same as "chown root path" in the shell:
		610	aio_chown "path", 0, -1;
		611	# same as above:
		612	aio_chown "path", 0, undef;
		613
		614
		615	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		616
		617	Works like truncate(2) or ftruncate(2).
		618
		619
		620	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		621
		622	Works like perl's C<chmod> function.
		623
		624
404	=item aio_unlink $pathname, $callback->($status)	625	=item aio_unlink $pathname, $callback->($status)
405		626
406	Asynchronously unlink (delete) a file and call the callback with the	627	Asynchronously unlink (delete) a file and call the callback with the
407	result code.	628	result code.
408		629
		630
		631	=item aio_mknod $pathname, $mode, $dev, $callback->($status)
		632
		633	[EXPERIMENTAL]
		634
		635	Asynchronously create a device node (or fifo). See mknod(2).
		636
		637	The only (POSIX-) portable way of calling this function is:
		638
		639	aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		640
		641	See C<aio_stat> for info about some potentially helpful extra constants
		642	and functions.
		643
409	=item aio_link $srcpath, $dstpath, $callback->($status)	644	=item aio_link $srcpath, $dstpath, $callback->($status)
410		645
411	Asynchronously create a new link to the existing object at C<$srcpath> at	646	Asynchronously create a new link to the existing object at C<$srcpath> at
412	the path C<$dstpath> and call the callback with the result code.	647	the path C<$dstpath> and call the callback with the result code.
413		648
		649
414	=item aio_symlink $srcpath, $dstpath, $callback->($status)	650	=item aio_symlink $srcpath, $dstpath, $callback->($status)
415		651
416	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at	652	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
417	the path C<$dstpath> and call the callback with the result code.	653	the path C<$dstpath> and call the callback with the result code.
418		654
		655
		656	=item aio_readlink $pathname, $callback->($link)
		657
		658	Asynchronously read the symlink specified by C<$path> and pass it to
		659	the callback. If an error occurs, nothing or undef gets passed to the
		660	callback.
		661
		662
		663	=item aio_realpath $pathname, $callback->($path)
		664
		665	Asynchronously make the path absolute and resolve any symlinks in
		666	C<$path>. The resulting path only consists of directories (Same as
		667	L<Cwd::realpath>).
		668
		669	This request can be used to get the absolute path of the current working
		670	directory by passing it a path of F<.> (a single dot).
		671
		672
419	=item aio_rename $srcpath, $dstpath, $callback->($status)	673	=item aio_rename $srcpath, $dstpath, $callback->($status)
420		674
421	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as	675	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
422	rename(2) and call the callback with the result code.	676	rename(2) and call the callback with the result code.
423		677
		678
		679	=item aio_mkdir $pathname, $mode, $callback->($status)
		680
		681	Asynchronously mkdir (create) a directory and call the callback with
		682	the result code. C<$mode> will be modified by the umask at the time the
		683	request is executed, so do not change your umask.
		684
		685
424	=item aio_rmdir $pathname, $callback->($status)	686	=item aio_rmdir $pathname, $callback->($status)
425		687
426	Asynchronously rmdir (delete) a directory and call the callback with the	688	Asynchronously rmdir (delete) a directory and call the callback with the
427	result code.	689	result code.
		690
428		691
429	=item aio_readdir $pathname, $callback->($entries)	692	=item aio_readdir $pathname, $callback->($entries)
430		693
431	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire	694	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
432	directory (i.e. opendir + readdir + closedir). The entries will not be	695	directory (i.e. opendir + readdir + closedir). The entries will not be
433	sorted, and will B<NOT> include the C<.> and C<..> entries.	696	sorted, and will B<NOT> include the C<.> and C<..> entries.
434		697
435	The callback a single argument which is either C<undef> or an array-ref	698	The callback is passed a single argument which is either C<undef> or an
436	with the filenames.	699	array-ref with the filenames.
437		700
		701
		702	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		703
		704	Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
		705	tune behaviour and output format. In case of an error, C<$entries> will be
		706	C<undef>.
		707
		708	The flags are a combination of the following constants, ORed together (the
		709	flags will also be passed to the callback, possibly modified):
		710
		711	=over 4
		712
		713	=item IO::AIO::READDIR_DENTS
		714
		715	When this flag is off, then the callback gets an arrayref consisting of
		716	names only (as with C<aio_readdir>), otherwise it gets an arrayref with
		717	C<[$name, $type, $inode]> arrayrefs, each describing a single directory
		718	entry in more detail.
		719
		720	C<$name> is the name of the entry.
		721
		722	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		723
		724	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		725	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		726	C<IO::AIO::DT_WHT>.
		727
		728	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
		729	know, you have to run stat yourself. Also, for speed reasons, the C<$type>
		730	scalars are read-only: you can not modify them.
		731
		732	C<$inode> is the inode number (which might not be exact on systems with 64
		733	bit inode numbers and 32 bit perls). This field has unspecified content on
		734	systems that do not deliver the inode information.
		735
		736	=item IO::AIO::READDIR_DIRS_FIRST
		737
		738	When this flag is set, then the names will be returned in an order where
		739	likely directories come first, in optimal stat order. This is useful when
		740	you need to quickly find directories, or you want to find all directories
		741	while avoiding to stat() each entry.
		742
		743	If the system returns type information in readdir, then this is used
		744	to find directories directly. Otherwise, likely directories are names
		745	beginning with ".", or otherwise names with no dots, of which names with
		746	short names are tried first.
		747
		748	=item IO::AIO::READDIR_STAT_ORDER
		749
		750	When this flag is set, then the names will be returned in an order
		751	suitable for stat()'ing each one. That is, when you plan to stat()
		752	all files in the given directory, then the returned order will likely
		753	be fastest.
		754
		755	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
		756	the likely dirs come first, resulting in a less optimal stat order.
		757
		758	=item IO::AIO::READDIR_FOUND_UNKNOWN
		759
		760	This flag should not be set when calling C<aio_readdirx>. Instead, it
		761	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		762	C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
		763	C<$type>'s are known, which can be used to speed up some algorithms.
		764
		765	=back
		766
		767
		768	=item aio_load $pathname, $data, $callback->($status)
		769
		770	This is a composite request that tries to fully load the given file into
		771	memory. Status is the same as with aio_read.
		772
		773	=cut
		774
		775	sub aio_load($$;$) {
		776	my ($path, undef, $cb) = @_;
		777	my $data = \$_[1];
		778
		779	my $pri = aioreq_pri;
		780	my $grp = aio_group $cb;
		781
		782	aioreq_pri $pri;
		783	add $grp aio_open $path, O_RDONLY, 0, sub {
		784	my $fh = shift
		785	or return $grp->result (-1);
		786
		787	aioreq_pri $pri;
		788	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		789	$grp->result ($_[0]);
		790	};
		791	};
		792
		793	$grp
		794	}
		795
		796	=item aio_copy $srcpath, $dstpath, $callback->($status)
		797
		798	Try to copy the I<file> (directories not supported as either source or
		799	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		800	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		801
		802	This is a composite request that creates the destination file with
		803	mode 0200 and copies the contents of the source file into it using
		804	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
		805	uid/gid, in that order.
		806
		807	If an error occurs, the partial destination file will be unlinked, if
		808	possible, except when setting atime, mtime, access mode and uid/gid, where
		809	errors are being ignored.
		810
		811	=cut
		812
		813	sub aio_copy($$;$) {
		814	my ($src, $dst, $cb) = @_;
		815
		816	my $pri = aioreq_pri;
		817	my $grp = aio_group $cb;
		818
		819	aioreq_pri $pri;
		820	add $grp aio_open $src, O_RDONLY, 0, sub {
		821	if (my $src_fh = $_[0]) {
		822	my @stat = stat $src_fh; # hmm, might block over nfs?
		823
		824	aioreq_pri $pri;
		825	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
		826	if (my $dst_fh = $_[0]) {
		827	aioreq_pri $pri;
		828	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
		829	if ($_[0] == $stat[7]) {
		830	$grp->result (0);
		831	close $src_fh;
		832
		833	my $ch = sub {
		834	aioreq_pri $pri;
		835	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		836	aioreq_pri $pri;
		837	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		838	aioreq_pri $pri;
		839	add $grp aio_close $dst_fh;
		840	}
		841	};
		842	};
		843
		844	aioreq_pri $pri;
		845	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		846	if ($_[0] < 0 && $! == ENOSYS) {
		847	aioreq_pri $pri;
		848	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
		849	} else {
		850	$ch->();
		851	}
		852	};
		853	} else {
		854	$grp->result (-1);
		855	close $src_fh;
		856	close $dst_fh;
		857
		858	aioreq $pri;
		859	add $grp aio_unlink $dst;
		860	}
		861	};
		862	} else {
		863	$grp->result (-1);
		864	}
		865	},
		866
		867	} else {
		868	$grp->result (-1);
		869	}
		870	};
		871
		872	$grp
		873	}
		874
		875	=item aio_move $srcpath, $dstpath, $callback->($status)
		876
		877	Try to move the I<file> (directories not supported as either source or
		878	destination) from C<$srcpath> to C<$dstpath> and call the callback with
		879	a status of C<0> (ok) or C<-1> (error, see C<$!>).
		880
		881	This is a composite request that tries to rename(2) the file first; if
		882	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
		883	that is successful, unlinks the C<$srcpath>.
		884
		885	=cut
		886
		887	sub aio_move($$;$) {
		888	my ($src, $dst, $cb) = @_;
		889
		890	my $pri = aioreq_pri;
		891	my $grp = aio_group $cb;
		892
		893	aioreq_pri $pri;
		894	add $grp aio_rename $src, $dst, sub {
		895	if ($_[0] && $! == EXDEV) {
		896	aioreq_pri $pri;
		897	add $grp aio_copy $src, $dst, sub {
		898	$grp->result ($_[0]);
		899
		900	unless ($_[0]) {
		901	aioreq_pri $pri;
		902	add $grp aio_unlink $src;
		903	}
		904	};
		905	} else {
		906	$grp->result ($_[0]);
		907	}
		908	};
		909
		910	$grp
		911	}
		912
438	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)	913	=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
439		914
440	Scans a directory (similar to C<aio_readdir>) but additionally tries to	915	Scans a directory (similar to C<aio_readdir>) but additionally tries to
441	efficiently separate the entries of directory C<$path> into two sets of	916	efficiently separate the entries of directory C<$path> into two sets of
442	names, directories you can recurse into (directories), and ones you cannot	917	names, directories you can recurse into (directories), and ones you cannot
443	recurse into (everything else, including symlinks to directories).	918	recurse into (everything else, including symlinks to directories).
…		…
460		935
461	Implementation notes.	936	Implementation notes.
462		937
463	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.	938	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
464		939
		940	If readdir returns file type information, then this is used directly to
		941	find directories.
		942
465	After reading the directory, the modification time, size etc. of the	943	Otherwise, after reading the directory, the modification time, size etc.
466	directory before and after the readdir is checked, and if they match (and	944	of the directory before and after the readdir is checked, and if they
467	isn't the current time), the link count will be used to decide how many	945	match (and isn't the current time), the link count will be used to decide
468	entries are directories (if >= 2). Otherwise, no knowledge of the number	946	how many entries are directories (if >= 2). Otherwise, no knowledge of the
469	of subdirectories will be assumed.	947	number of subdirectories will be assumed.
470		948
471	Then entries will be sorted into likely directories (everything without	949	Then entries will be sorted into likely directories a non-initial dot
472	a non-initial dot currently) and likely non-directories (everything	950	currently) and likely non-directories (see C<aio_readdirx>). Then every
473	else). Then every entry plus an appended C</.> will be C<stat>'ed,	951	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
474	likely directories first. If that succeeds, it assumes that the entry	952	in order of their inode numbers. If that succeeds, it assumes that the
475	is a directory or a symlink to directory (which will be checked	953	entry is a directory or a symlink to directory (which will be checked
476	seperately). This is often faster than stat'ing the entry itself because	954	separately). This is often faster than stat'ing the entry itself because
477	filesystems might detect the type of the entry without reading the inode	955	filesystems might detect the type of the entry without reading the inode
478	data (e.g. ext2fs filetype feature).	956	data (e.g. ext2fs filetype feature), even on systems that cannot return
		957	the filetype information on readdir.
479		958
480	If the known number of directories (link count - 2) has been reached, the	959	If the known number of directories (link count - 2) has been reached, the
481	rest of the entries is assumed to be non-directories.	960	rest of the entries is assumed to be non-directories.
482		961
483	This only works with certainty on POSIX (= UNIX) filesystems, which	962	This only works with certainty on POSIX (= UNIX) filesystems, which
…		…
487	as those tend to return 0 or 1 as link counts, which disables the	966	as those tend to return 0 or 1 as link counts, which disables the
488	directory counting heuristic.	967	directory counting heuristic.
489		968
490	=cut	969	=cut
491		970
492	sub aio_scandir($$$) {	971	sub aio_scandir($$;$) {
493	my ($path, $maxreq, $cb) = @_;	972	my ($path, $maxreq, $cb) = @_;
494		973
495	my $pri = aioreq_pri;	974	my $pri = aioreq_pri;
496		975
497	my $grp = aio_group $cb;	976	my $grp = aio_group $cb;
498		977
499	$maxreq = 4 if $maxreq <= 0;	978	$maxreq = 4 if $maxreq <= 0;
500		979
501	# stat once	980	# get a wd object
502	aioreq_pri $pri;	981	aioreq_pri $pri;
503	add $grp aio_stat $path, sub {	982	add $grp aio_wd $path, sub {
		983	$_[0]
504	return $grp->result () if $_[0];	984	or return $grp->result ();
505	my $now = time;
506	my $hash1 = join ":", (stat _)[0,1,3,7,9];
507		985
508	# read the directory entries	986	my $wd = [shift, "."];
		987
		988	# stat once
509	aioreq_pri $pri;	989	aioreq_pri $pri;
510	add $grp aio_readdir $path, sub {	990	add $grp aio_stat $wd, sub {
511	my $entries = shift
512	or return $grp->result ();	991	return $grp->result () if $_[0];
		992	my $now = time;
		993	my $hash1 = join ":", (stat _)[0,1,3,7,9];
513		994
514	# stat the dir another time	995	# read the directory entries
515	aioreq_pri $pri;	996	aioreq_pri $pri;
		997	add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
		998	my $entries = shift
		999	or return $grp->result ();
		1000
		1001	# stat the dir another time
		1002	aioreq_pri $pri;
516	add $grp aio_stat $path, sub {	1003	add $grp aio_stat $wd, sub {
517	my $hash2 = join ":", (stat _)[0,1,3,7,9];	1004	my $hash2 = join ":", (stat _)[0,1,3,7,9];
518		1005
519	my $ndirs;	1006	my $ndirs;
520		1007
521	# take the slow route if anything looks fishy	1008	# take the slow route if anything looks fishy
522	if ($hash1 ne $hash2 or (stat _)[9] == $now) {	1009	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
523	$ndirs = -1;	1010	$ndirs = -1;
524	} else {	1011	} else {
525	# if nlink == 2, we are finished	1012	# if nlink == 2, we are finished
526	# on non-posix-fs's, we rely on nlink < 2	1013	# for non-posix-fs's, we rely on nlink < 2
527	$ndirs = (stat _)[3] - 2	1014	$ndirs = (stat _)[3] - 2
528	or return $grp->result ([], $entries);	1015	or return $grp->result ([], $entries);
529	}	1016	}
530		1017
531	# sort into likely dirs and likely nondirs
532	# dirs == files without ".", short entries first
533	$entries = [map $_->[0],
534	sort { $b->[1] cmp $a->[1] }
535	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
536	@$entries];
537
538	my (@dirs, @nondirs);	1018	my (@dirs, @nondirs);
539		1019
540	my $statgrp = add $grp aio_group sub {	1020	my $statgrp = add $grp aio_group sub {
541	$grp->result (\@dirs, \@nondirs);	1021	$grp->result (\@dirs, \@nondirs);
542	};	1022	};
543		1023
544	limit $statgrp $maxreq;	1024	limit $statgrp $maxreq;
545	feed $statgrp sub {	1025	feed $statgrp sub {
546	return unless @$entries;	1026	return unless @$entries;
547	my $entry = pop @$entries;	1027	my $entry = shift @$entries;
548		1028
549	aioreq_pri $pri;	1029	aioreq_pri $pri;
		1030	$wd->[1] = "$entry/.";
550	add $statgrp aio_stat "$path/$entry/.", sub {	1031	add $statgrp aio_stat $wd, sub {
551	if ($_[0] < 0) {	1032	if ($_[0] < 0) {
552	push @nondirs, $entry;	1033	push @nondirs, $entry;
553	} else {	1034	} else {
554	# need to check for real directory	1035	# need to check for real directory
555	aioreq_pri $pri;	1036	aioreq_pri $pri;
		1037	$wd->[1] = $entry;
556	add $statgrp aio_lstat "$path/$entry", sub {	1038	add $statgrp aio_lstat $wd, sub {
557	if (-d _) {	1039	if (-d _) {
558	push @dirs, $entry;	1040	push @dirs, $entry;
559		1041
560	unless (--$ndirs) {	1042	unless (--$ndirs) {
561	push @nondirs, @$entries;	1043	push @nondirs, @$entries;
562	feed $statgrp;	1044	feed $statgrp;
		1045	}
		1046	} else {
		1047	push @nondirs, $entry;
563	}	1048	}
564	} else {
565	push @nondirs, $entry;
566	}	1049	}
567	}	1050	}
568	}	1051	};
569	};	1052	};
570	};	1053	};
571	};	1054	};
572	};	1055	};
573	};	1056	};
574		1057
575	$grp	1058	$grp
576	}	1059	}
577		1060
		1061	=item aio_rmtree $pathname, $callback->($status)
		1062
		1063	Delete a directory tree starting (and including) C<$path>, return the
		1064	status of the final C<rmdir> only. This is a composite request that
		1065	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1066	everything else.
		1067
		1068	=cut
		1069
		1070	sub aio_rmtree;
		1071	sub aio_rmtree($;$) {
		1072	my ($path, $cb) = @_;
		1073
		1074	my $pri = aioreq_pri;
		1075	my $grp = aio_group $cb;
		1076
		1077	aioreq_pri $pri;
		1078	add $grp aio_scandir $path, 0, sub {
		1079	my ($dirs, $nondirs) = @_;
		1080
		1081	my $dirgrp = aio_group sub {
		1082	add $grp aio_rmdir $path, sub {
		1083	$grp->result ($_[0]);
		1084	};
		1085	};
		1086
		1087	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1088	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1089
		1090	add $grp $dirgrp;
		1091	};
		1092
		1093	$grp
		1094	}
		1095
		1096	=item aio_sync $callback->($status)
		1097
		1098	Asynchronously call sync and call the callback when finished.
		1099
578	=item aio_fsync $fh, $callback->($status)	1100	=item aio_fsync $fh, $callback->($status)
579		1101
580	Asynchronously call fsync on the given filehandle and call the callback	1102	Asynchronously call fsync on the given filehandle and call the callback
581	with the fsync result code.	1103	with the fsync result code.
582		1104
…		…
585	Asynchronously call fdatasync on the given filehandle and call the	1107	Asynchronously call fdatasync on the given filehandle and call the
586	callback with the fdatasync result code.	1108	callback with the fdatasync result code.
587		1109
588	If this call isn't available because your OS lacks it or it couldn't be	1110	If this call isn't available because your OS lacks it or it couldn't be
589	detected, it will be emulated by calling C<fsync> instead.	1111	detected, it will be emulated by calling C<fsync> instead.
		1112
		1113	=item aio_syncfs $fh, $callback->($status)
		1114
		1115	Asynchronously call the syncfs syscall to sync the filesystem associated
		1116	to the given filehandle and call the callback with the syncfs result
		1117	code. If syncfs is not available, calls sync(), but returns C<-1> and sets
		1118	errno to C<ENOSYS> nevertheless.
		1119
		1120	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1121
		1122	Sync the data portion of the file specified by C<$offset> and C<$length>
		1123	to disk (but NOT the metadata), by calling the Linux-specific
		1124	sync_file_range call. If sync_file_range is not available or it returns
		1125	ENOSYS, then fdatasync or fsync is being substituted.
		1126
		1127	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1128	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1129	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1130	manpage for details.
		1131
		1132	=item aio_pathsync $pathname, $callback->($status)
		1133
		1134	This request tries to open, fsync and close the given path. This is a
		1135	composite request intended to sync directories after directory operations
		1136	(E.g. rename). This might not work on all operating systems or have any
		1137	specific effect, but usually it makes sure that directory changes get
		1138	written to disc. It works for anything that can be opened for read-only,
		1139	not just directories.
		1140
		1141	Future versions of this function might fall back to other methods when
		1142	C<fsync> on the directory fails (such as calling C<sync>).
		1143
		1144	Passes C<0> when everything went ok, and C<-1> on error.
		1145
		1146	=cut
		1147
		1148	sub aio_pathsync($;$) {
		1149	my ($path, $cb) = @_;
		1150
		1151	my $pri = aioreq_pri;
		1152	my $grp = aio_group $cb;
		1153
		1154	aioreq_pri $pri;
		1155	add $grp aio_open $path, O_RDONLY, 0, sub {
		1156	my ($fh) = @_;
		1157	if ($fh) {
		1158	aioreq_pri $pri;
		1159	add $grp aio_fsync $fh, sub {
		1160	$grp->result ($_[0]);
		1161
		1162	aioreq_pri $pri;
		1163	add $grp aio_close $fh;
		1164	};
		1165	} else {
		1166	$grp->result (-1);
		1167	}
		1168	};
		1169
		1170	$grp
		1171	}
		1172
		1173	=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1174
		1175	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1176	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1177	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1178	scalar must only be modified in-place while an aio operation is pending on
		1179	it).
		1180
		1181	It calls the C<msync> function of your OS, if available, with the memory
		1182	area starting at C<$offset> in the string and ending C<$length> bytes
		1183	later. If C<$length> is negative, counts from the end, and if C<$length>
		1184	is C<undef>, then it goes till the end of the string. The flags can be
		1185	a combination of C<IO::AIO::MS_ASYNC>, C<IO::AIO::MS_INVALIDATE> and
		1186	C<IO::AIO::MS_SYNC>.
		1187
		1188	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1189
		1190	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1191	scalars.
		1192
		1193	It touches (reads or writes) all memory pages in the specified
		1194	range inside the scalar. All caveats and parameters are the same
		1195	as for C<aio_msync>, above, except for flags, which must be either
		1196	C<0> (which reads all pages and ensures they are instantiated) or
		1197	C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
		1198	writing an octet from it, which dirties the page).
		1199
		1200	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1201
		1202	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1203	scalars.
		1204
		1205	It reads in all the pages of the underlying storage into memory (if any)
		1206	and locks them, so they are not getting swapped/paged out or removed.
		1207
		1208	If C<$length> is undefined, then the scalar will be locked till the end.
		1209
		1210	On systems that do not implement C<mlock>, this function returns C<-1>
		1211	and sets errno to C<ENOSYS>.
		1212
		1213	Note that the corresponding C<munlock> is synchronous and is
		1214	documented under L<MISCELLANEOUS FUNCTIONS>.
		1215
		1216	Example: open a file, mmap and mlock it - both will be undone when
		1217	C<$data> gets destroyed.
		1218
		1219	open my $fh, "<", $path or die "$path: $!";
		1220	my $data;
		1221	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1222	aio_mlock $data; # mlock in background
		1223
		1224	=item aio_mlockall $flags, $callback->($status)
		1225
		1226	Calls the C<mlockall> function with the given C<$flags> (a combination of
		1227	C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
		1228
		1229	On systems that do not implement C<mlockall>, this function returns C<-1>
		1230	and sets errno to C<ENOSYS>.
		1231
		1232	Note that the corresponding C<munlockall> is synchronous and is
		1233	documented under L<MISCELLANEOUS FUNCTIONS>.
		1234
		1235	Example: asynchronously lock all current and future pages into memory.
		1236
		1237	aio_mlockall IO::AIO::MCL_FUTURE;
590		1238
591	=item aio_group $callback->(...)	1239	=item aio_group $callback->(...)
592		1240
593	This is a very special aio request: Instead of doing something, it is a	1241	This is a very special aio request: Instead of doing something, it is a
594	container for other aio requests, which is useful if you want to bundle	1242	container for other aio requests, which is useful if you want to bundle
…		…
632	immense (it blocks a thread for a long time) so do not use this function	1280	immense (it blocks a thread for a long time) so do not use this function
633	except to put your application under artificial I/O pressure.	1281	except to put your application under artificial I/O pressure.
634		1282
635	=back	1283	=back
636		1284
		1285
		1286	=head2 IO::AIO::WD - multiple working directories
		1287
		1288	Your process only has one current working directory, which is used by all
		1289	threads. This makes it hard to use relative paths (some other component
		1290	could call C<chdir> at any time, and it is hard to control when the path
		1291	will be used by IO::AIO).
		1292
		1293	One solution for this is to always use absolute paths. This usually works,
		1294	but can be quite slow (the kernel has to walk the whole path on every
		1295	access), and can also be a hassle to implement.
		1296
		1297	Newer POSIX systems have a number of functions (openat, fdopendir,
		1298	futimensat and so on) that make it possible to specify working directories
		1299	per operation.
		1300
		1301	For portability, and because the clowns who "designed", or shall I write,
		1302	perpetrated this new interface were obviously half-drunk, this abstraction
		1303	cannot be perfect, though.
		1304
		1305	IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
		1306	object. This object stores the canonicalised, absolute version of the
		1307	path, and on systems that allow it, also a directory file descriptor.
		1308
		1309	Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
		1310	or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
		1311	object and a pathname instead (or the IO::AIO::WD object alone, which
		1312	gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
		1313	IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
		1314	to that IO::AIO::WD object.
		1315
		1316	For example, to get a wd object for F</etc> and then stat F<passwd>
		1317	inside, you would write:
		1318
		1319	aio_wd "/etc", sub {
		1320	my $etcdir = shift;
		1321
		1322	# although $etcdir can be undef on error, there is generally no reason
		1323	# to check for errors here, as aio_stat will fail with ENOENT
		1324	# when $etcdir is undef.
		1325
		1326	aio_stat [$etcdir, "passwd"], sub {
		1327	# yay
		1328	};
		1329	};
		1330
		1331	That C<aio_wd> is a request and not a normal function shows that creating
		1332	an IO::AIO::WD object is itself a potentially blocking operation, which is
		1333	why it is done asynchronously.
		1334
		1335	To stat the directory obtained with C<aio_wd> above, one could write
		1336	either of the following three request calls:
		1337
		1338	aio_lstat "/etc" , sub { ... # pathname as normal string
		1339	aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
		1340	aio_lstat $wd , sub { ... # shorthand for the previous
		1341
		1342	As with normal pathnames, IO::AIO keeps a copy of the working directory
		1343	object and the pathname string, so you could write the following without
		1344	causing any issues due to C<$path> getting reused:
		1345
		1346	my $path = [$wd, undef];
		1347
		1348	for my $name (qw(abc def ghi)) {
		1349	$path->[1] = $name;
		1350	aio_stat $path, sub {
		1351	# ...
		1352	};
		1353	}
		1354
		1355	There are some caveats: when directories get renamed (or deleted), the
		1356	pathname string doesn't change, so will point to the new directory (or
		1357	nowhere at all), while the directory fd, if available on the system,
		1358	will still point to the original directory. Most functions accepting a
		1359	pathname will use the directory fd on newer systems, and the string on
		1360	older systems. Some functions (such as realpath) will always rely on the
		1361	string form of the pathname.
		1362
		1363	So this fucntionality is mainly useful to get some protection against
		1364	C<chdir>, to easily get an absolute path out of a relative path for future
		1365	reference, and to speed up doing many operations in the same directory
		1366	(e.g. when stat'ing all files in a directory).
		1367
		1368	The following functions implement this working directory abstraction:
		1369
		1370	=over 4
		1371
		1372	=item aio_wd $pathname, $callback->($wd)
		1373
		1374	Asynchonously canonicalise the given pathname and convert it to an
		1375	IO::AIO::WD object representing it. If possible and supported on the
		1376	system, also open a directory fd to speed up pathname resolution relative
		1377	to this working directory.
		1378
		1379	If something goes wrong, then C<undef> is passwd to the callback instead
		1380	of a working directory object and C<$!> is set appropriately. Since
		1381	passing C<undef> as working directory component of a pathname fails the
		1382	request with C<ENOENT>, there is often no need for error checking in the
		1383	C<aio_wd> callback, as future requests using the value will fail in the
		1384	expected way.
		1385
		1386	If this call isn't available because your OS lacks it or it couldn't be
		1387	detected, it will be emulated by calling C<fsync> instead.
		1388
		1389	=item IO::AIO::CWD
		1390
		1391	This is a compiletime constant (object) that represents the process
		1392	current working directory.
		1393
		1394	Specifying this object as working directory object for a pathname is as
		1395	if the pathname would be specified directly, without a directory object,
		1396	e.g., these calls are functionally identical:
		1397
		1398	aio_stat "somefile", sub { ... };
		1399	aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
		1400
		1401	=back
		1402
		1403
637	=head2 IO::AIO::REQ CLASS	1404	=head2 IO::AIO::REQ CLASS
638		1405
639	All non-aggregate C<aio_*> functions return an object of this class when	1406	All non-aggregate C<aio_*> functions return an object of this class when
640	called in non-void context.	1407	called in non-void context.
641		1408
…		…
644	=item cancel $req	1411	=item cancel $req
645		1412
646	Cancels the request, if possible. Has the effect of skipping execution	1413	Cancels the request, if possible. Has the effect of skipping execution
647	when entering the B<execute> state and skipping calling the callback when	1414	when entering the B<execute> state and skipping calling the callback when
648	entering the the B<result> state, but will leave the request otherwise	1415	entering the the B<result> state, but will leave the request otherwise
649	untouched. That means that requests that currently execute will not be	1416	untouched (with the exception of readdir). That means that requests that
650	stopped and resources held by the request will not be freed prematurely.	1417	currently execute will not be stopped and resources held by the request
		1418	will not be freed prematurely.
651		1419
652	=item cb $req $callback->(...)	1420	=item cb $req $callback->(...)
653		1421
654	Replace (or simply set) the callback registered to the request.	1422	Replace (or simply set) the callback registered to the request.
655		1423
…		…
706	Their lifetime, simplified, looks like this: when they are empty, they	1474	Their lifetime, simplified, looks like this: when they are empty, they
707	will finish very quickly. If they contain only requests that are in the	1475	will finish very quickly. If they contain only requests that are in the
708	C<done> state, they will also finish. Otherwise they will continue to	1476	C<done> state, they will also finish. Otherwise they will continue to
709	exist.	1477	exist.
710		1478
711	That means after creating a group you have some time to add requests. And	1479	That means after creating a group you have some time to add requests
712	in the callbacks of those requests, you can add further requests to the	1480	(precisely before the callback has been invoked, which is only done within
713	group. And only when all those requests have finished will the the group	1481	the C<poll_cb>). And in the callbacks of those requests, you can add
714	itself finish.	1482	further requests to the group. And only when all those requests have
		1483	finished will the the group itself finish.
715		1484
716	=over 4	1485	=over 4
717		1486
718	=item add $grp ...	1487	=item add $grp ...
719		1488
…		…
728	=item $grp->cancel_subs	1497	=item $grp->cancel_subs
729		1498
730	Cancel all subrequests and clears any feeder, but not the group request	1499	Cancel all subrequests and clears any feeder, but not the group request
731	itself. Useful when you queued a lot of events but got a result early.	1500	itself. Useful when you queued a lot of events but got a result early.
732		1501
		1502	The group request will finish normally (you cannot add requests to the
		1503	group).
		1504
733	=item $grp->result (...)	1505	=item $grp->result (...)
734		1506
735	Set the result value(s) that will be passed to the group callback when all	1507	Set the result value(s) that will be passed to the group callback when all
736	subrequests have finished and set thre groups errno to the current value	1508	subrequests have finished and set the groups errno to the current value
737	of errno (just like calling C<errno> without an error number). By default,	1509	of errno (just like calling C<errno> without an error number). By default,
738	no argument will be passed and errno is zero.	1510	no argument will be passed and errno is zero.
739		1511
740	=item $grp->errno ([$errno])	1512	=item $grp->errno ([$errno])
741		1513
…		…
752	=item feed $grp $callback->($grp)	1524	=item feed $grp $callback->($grp)
753		1525
754	Sets a feeder/generator on this group: every group can have an attached	1526	Sets a feeder/generator on this group: every group can have an attached
755	generator that generates requests if idle. The idea behind this is that,	1527	generator that generates requests if idle. The idea behind this is that,
756	although you could just queue as many requests as you want in a group,	1528	although you could just queue as many requests as you want in a group,
757	this might starve other requests for a potentially long time. For	1529	this might starve other requests for a potentially long time. For example,
758	example, C<aio_scandir> might generate hundreds of thousands C<aio_stat>	1530	C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
759	requests, delaying any later requests for a long time.	1531	requests, delaying any later requests for a long time.
760		1532
761	To avoid this, and allow incremental generation of requests, you can	1533	To avoid this, and allow incremental generation of requests, you can
762	instead a group and set a feeder on it that generates those requests. The	1534	instead a group and set a feeder on it that generates those requests. The
763	feed callback will be called whenever there are few enough (see C<limit>,	1535	feed callback will be called whenever there are few enough (see C<limit>,
…		…
768	not impose any limits).	1540	not impose any limits).
769		1541
770	If the feed does not queue more requests when called, it will be	1542	If the feed does not queue more requests when called, it will be
771	automatically removed from the group.	1543	automatically removed from the group.
772		1544
773	If the feed limit is C<0>, it will be set to C<2> automatically.	1545	If the feed limit is C<0> when this method is called, it will be set to
		1546	C<2> automatically.
774		1547
775	Example:	1548	Example:
776		1549
777	# stat all files in @files, but only ever use four aio requests concurrently:	1550	# stat all files in @files, but only ever use four aio requests concurrently:
778		1551
…		…
790	Sets the feeder limit for the group: The feeder will be called whenever	1563	Sets the feeder limit for the group: The feeder will be called whenever
791	the group contains less than this many requests.	1564	the group contains less than this many requests.
792		1565
793	Setting the limit to C<0> will pause the feeding process.	1566	Setting the limit to C<0> will pause the feeding process.
794		1567
		1568	The default value for the limit is C<0>, but note that setting a feeder
		1569	automatically bumps it up to C<2>.
		1570
795	=back	1571	=back
796		1572
797	=head2 SUPPORT FUNCTIONS	1573	=head2 SUPPORT FUNCTIONS
798		1574
		1575	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
		1576
799	=over 4	1577	=over 4
800		1578
801	=item $fileno = IO::AIO::poll_fileno	1579	=item $fileno = IO::AIO::poll_fileno
802		1580
803	Return the I<request result pipe file descriptor>. This filehandle must be	1581	Return the I<request result pipe file descriptor>. This filehandle must be
804	polled for reading by some mechanism outside this module (e.g. Event or	1582	polled for reading by some mechanism outside this module (e.g. EV, Glib,
805	select, see below or the SYNOPSIS). If the pipe becomes readable you have	1583	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
806	to call C<poll_cb> to check the results.	1584	you have to call C<poll_cb> to check the results.
807		1585
808	See C<poll_cb> for an example.	1586	See C<poll_cb> for an example.
809		1587
810	=item IO::AIO::poll_cb	1588	=item IO::AIO::poll_cb
811		1589
812	Process all outstanding events on the result pipe. You have to call this	1590	Process some outstanding events on the result pipe. You have to call
813	regularly. Returns the number of events processed. Returns immediately	1591	this regularly. Returns C<0> if all events could be processed (or there
814	when no events are outstanding.	1592	were no events to process), or C<-1> if it returned earlier for whatever
		1593	reason. Returns immediately when no events are outstanding. The amount of
		1594	events processed depends on the settings of C<IO::AIO::max_poll_req> and
		1595	C<IO::AIO::max_poll_time>.
815		1596
816	If not all requests were processed for whatever reason, the filehandle	1597	If not all requests were processed for whatever reason, the filehandle
817	will still be ready when C<poll_cb> returns.	1598	will still be ready when C<poll_cb> returns, so normally you don't have to
		1599	do anything special to have it called later.
		1600
		1601	Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
		1602	ready, it can be beneficial to call this function from loops which submit
		1603	a lot of requests, to make sure the results get processed when they become
		1604	available and not just when the loop is finished and the event loop takes
		1605	over again. This function returns very fast when there are no outstanding
		1606	requests.
818		1607
819	Example: Install an Event watcher that automatically calls	1608	Example: Install an Event watcher that automatically calls
820	IO::AIO::poll_cb with high priority:	1609	IO::AIO::poll_cb with high priority (more examples can be found in the
		1610	SYNOPSIS section, at the top of this document):
821		1611
822	Event->io (fd => IO::AIO::poll_fileno,	1612	Event->io (fd => IO::AIO::poll_fileno,
823	poll => 'r', async => 1,	1613	poll => 'r', async => 1,
824	cb => \&IO::AIO::poll_cb);	1614	cb => \&IO::AIO::poll_cb);
825		1615
826	=item IO::AIO::poll_some $max_requests	1616	=item IO::AIO::poll_wait
827		1617
828	Similar to C<poll_cb>, but only processes up to C<$max_requests> requests	1618	If there are any outstanding requests and none of them in the result
829	at a time.	1619	phase, wait till the result filehandle becomes ready for reading (simply
		1620	does a C<select> on the filehandle. This is useful if you want to
		1621	synchronously wait for some requests to finish).
830		1622
831	Useful if you want to ensure some level of interactiveness when perl is	1623	See C<nreqs> for an example.
832	not fast enough to process all requests in time.	1624
		1625	=item IO::AIO::poll
		1626
		1627	Waits until some requests have been handled.
		1628
		1629	Returns the number of requests processed, but is otherwise strictly
		1630	equivalent to:
		1631
		1632	IO::AIO::poll_wait, IO::AIO::poll_cb
		1633
		1634	=item IO::AIO::flush
		1635
		1636	Wait till all outstanding AIO requests have been handled.
		1637
		1638	Strictly equivalent to:
		1639
		1640	IO::AIO::poll_wait, IO::AIO::poll_cb
		1641	while IO::AIO::nreqs;
		1642
		1643	=item IO::AIO::max_poll_reqs $nreqs
		1644
		1645	=item IO::AIO::max_poll_time $seconds
		1646
		1647	These set the maximum number of requests (default C<0>, meaning infinity)
		1648	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
		1649	the maximum amount of time (default C<0>, meaning infinity) spent in
		1650	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
		1651	of time C<poll_cb> is allowed to use).
		1652
		1653	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1654	syscall per request processed, which is not normally a problem unless your
		1655	callbacks are really really fast or your OS is really really slow (I am
		1656	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1657
		1658	Setting these is useful if you want to ensure some level of
		1659	interactiveness when perl is not fast enough to process all requests in
		1660	time.
		1661
		1662	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
833		1663
834	Example: Install an Event watcher that automatically calls	1664	Example: Install an Event watcher that automatically calls
835	IO::AIO::poll_some with low priority, to ensure that other parts of the	1665	IO::AIO::poll_cb with low priority, to ensure that other parts of the
836	program get the CPU sometimes even under high AIO load.	1666	program get the CPU sometimes even under high AIO load.
837		1667
		1668	# try not to spend much more than 0.1s in poll_cb
		1669	IO::AIO::max_poll_time 0.1;
		1670
		1671	# use a low priority so other tasks have priority
838	Event->io (fd => IO::AIO::poll_fileno,	1672	Event->io (fd => IO::AIO::poll_fileno,
839	poll => 'r', nice => 1,	1673	poll => 'r', nice => 1,
840	cb => sub { IO::AIO::poll_some 256 });	1674	cb => &IO::AIO::poll_cb);
841		1675
842	=item IO::AIO::poll_wait	1676	=back
843		1677
844	Wait till the result filehandle becomes ready for reading (simply does a	1678	=head3 CONTROLLING THE NUMBER OF THREADS
845	C<select> on the filehandle. This is useful if you want to synchronously wait
846	for some requests to finish).
847		1679
848	See C<nreqs> for an example.	1680	=over
849
850	=item IO::AIO::nreqs
851
852	Returns the number of requests currently in the ready, execute or pending
853	states (i.e. for which their callback has not been invoked yet).
854
855	Example: wait till there are no outstanding requests anymore:
856
857	IO::AIO::poll_wait, IO::AIO::poll_cb
858	while IO::AIO::nreqs;
859
860	=item IO::AIO::nready
861
862	Returns the number of requests currently in the ready state (not yet
863	executed).
864
865	=item IO::AIO::npending
866
867	Returns the number of requests currently in the pending state (executed,
868	but not yet processed by poll_cb).
869
870	=item IO::AIO::flush
871
872	Wait till all outstanding AIO requests have been handled.
873
874	Strictly equivalent to:
875
876	IO::AIO::poll_wait, IO::AIO::poll_cb
877	while IO::AIO::nreqs;
878
879	=item IO::AIO::poll
880
881	Waits until some requests have been handled.
882
883	Strictly equivalent to:
884
885	IO::AIO::poll_wait, IO::AIO::poll_cb
886	if IO::AIO::nreqs;
887		1681
888	=item IO::AIO::min_parallel $nthreads	1682	=item IO::AIO::min_parallel $nthreads
889		1683
890	Set the minimum number of AIO threads to C<$nthreads>. The current	1684	Set the minimum number of AIO threads to C<$nthreads>. The current
891	default is C<8>, which means eight asynchronous operations can execute	1685	default is C<8>, which means eight asynchronous operations can execute
892	concurrently at any one time (the number of outstanding requests,	1686	concurrently at any one time (the number of outstanding requests,
893	however, is unlimited).	1687	however, is unlimited).
894		1688
895	IO::AIO starts threads only on demand, when an AIO request is queued and	1689	IO::AIO starts threads only on demand, when an AIO request is queued and
896	no free thread exists.	1690	no free thread exists. Please note that queueing up a hundred requests can
		1691	create demand for a hundred threads, even if it turns out that everything
		1692	is in the cache and could have been processed faster by a single thread.
897		1693
898	It is recommended to keep the number of threads relatively low, as some	1694	It is recommended to keep the number of threads relatively low, as some
899	Linux kernel versions will scale negatively with the number of threads	1695	Linux kernel versions will scale negatively with the number of threads
900	(higher parallelity => MUCH higher latency). With current Linux 2.6	1696	(higher parallelity => MUCH higher latency). With current Linux 2.6
901	versions, 4-32 threads should be fine.	1697	versions, 4-32 threads should be fine.
…		…
915	This module automatically runs C<max_parallel 0> at program end, to ensure	1711	This module automatically runs C<max_parallel 0> at program end, to ensure
916	that all threads are killed and that there are no outstanding requests.	1712	that all threads are killed and that there are no outstanding requests.
917		1713
918	Under normal circumstances you don't need to call this function.	1714	Under normal circumstances you don't need to call this function.
919		1715
		1716	=item IO::AIO::max_idle $nthreads
		1717
		1718	Limit the number of threads (default: 4) that are allowed to idle
		1719	(i.e., threads that did not get a request to process within the idle
		1720	timeout (default: 10 seconds). That means if a thread becomes idle while
		1721	C<$nthreads> other threads are also idle, it will free its resources and
		1722	exit.
		1723
		1724	This is useful when you allow a large number of threads (e.g. 100 or 1000)
		1725	to allow for extremely high load situations, but want to free resources
		1726	under normal circumstances (1000 threads can easily consume 30MB of RAM).
		1727
		1728	The default is probably ok in most situations, especially if thread
		1729	creation is fast. If thread creation is very slow on your system you might
		1730	want to use larger values.
		1731
		1732	=item IO::AIO::idle_timeout $seconds
		1733
		1734	Sets the minimum idle timeout (default 10) after which worker threads are
		1735	allowed to exit. SEe C<IO::AIO::max_idle>.
		1736
920	=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs	1737	=item IO::AIO::max_outstanding $maxreqs
		1738
		1739	Sets the maximum number of outstanding requests to C<$nreqs>. If
		1740	you do queue up more than this number of requests, the next call to
		1741	C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
		1742	C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
		1743	longer exceeded.
		1744
		1745	In other words, this setting does not enforce a queue limit, but can be
		1746	used to make poll functions block if the limit is exceeded.
921		1747
922	This is a very bad function to use in interactive programs because it	1748	This is a very bad function to use in interactive programs because it
923	blocks, and a bad way to reduce concurrency because it is inexact: Better	1749	blocks, and a bad way to reduce concurrency because it is inexact: Better
924	use an C<aio_group> together with a feed callback.	1750	use an C<aio_group> together with a feed callback.
925		1751
926	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1752	It's main use is in scripts without an event loop - when you want to stat
927	to queue up more than this number of requests, the next call to the	1753	a lot of files, you can write somehting like this:
928	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
929	function will block until the limit is no longer exceeded.
930		1754
931	The default value is very large, so there is no practical limit on the	1755	IO::AIO::max_outstanding 32;
932	number of outstanding requests.
933		1756
934	You can still queue as many requests as you want. Therefore,	1757	for my $path (...) {
935	C<max_oustsanding> is mainly useful in simple scripts (with low values) or	1758	aio_stat $path , ...;
936	as a stop gap to shield against fatal memory overflow (with large values).	1759	IO::AIO::poll_cb;
		1760	}
		1761
		1762	IO::AIO::flush;
		1763
		1764	The call to C<poll_cb> inside the loop will normally return instantly, but
		1765	as soon as more thna C<32> reqeusts are in-flight, it will block until
		1766	some requests have been handled. This keeps the loop from pushing a large
		1767	number of C<aio_stat> requests onto the queue.
		1768
		1769	The default value for C<max_outstanding> is very large, so there is no
		1770	practical limit on the number of outstanding requests.
937		1771
938	=back	1772	=back
939		1773
		1774	=head3 STATISTICAL INFORMATION
		1775
		1776	=over
		1777
		1778	=item IO::AIO::nreqs
		1779
		1780	Returns the number of requests currently in the ready, execute or pending
		1781	states (i.e. for which their callback has not been invoked yet).
		1782
		1783	Example: wait till there are no outstanding requests anymore:
		1784
		1785	IO::AIO::poll_wait, IO::AIO::poll_cb
		1786	while IO::AIO::nreqs;
		1787
		1788	=item IO::AIO::nready
		1789
		1790	Returns the number of requests currently in the ready state (not yet
		1791	executed).
		1792
		1793	=item IO::AIO::npending
		1794
		1795	Returns the number of requests currently in the pending state (executed,
		1796	but not yet processed by poll_cb).
		1797
		1798	=back
		1799
		1800	=head3 MISCELLANEOUS FUNCTIONS
		1801
		1802	IO::AIO implements some functions that might be useful, but are not
		1803	asynchronous.
		1804
		1805	=over 4
		1806
		1807	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		1808
		1809	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		1810	but is blocking (this makes most sense if you know the input data is
		1811	likely cached already and the output filehandle is set to non-blocking
		1812	operations).
		1813
		1814	Returns the number of bytes copied, or C<-1> on error.
		1815
		1816	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		1817
		1818	Simply calls the C<posix_fadvise> function (see its
		1819	manpage for details). The following advice constants are
		1820	available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		1821	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		1822	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		1823
		1824	On systems that do not implement C<posix_fadvise>, this function returns
		1825	ENOSYS, otherwise the return value of C<posix_fadvise>.
		1826
		1827	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		1828
		1829	Simply calls the C<posix_madvise> function (see its
		1830	manpage for details). The following advice constants are
		1831	available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		1832	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
		1833
		1834	On systems that do not implement C<posix_madvise>, this function returns
		1835	ENOSYS, otherwise the return value of C<posix_madvise>.
		1836
		1837	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		1838
		1839	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		1840	$scalar (see its manpage for details). The following protect
		1841	constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		1842	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		1843
		1844	On systems that do not implement C<mprotect>, this function returns
		1845	ENOSYS, otherwise the return value of C<mprotect>.
		1846
		1847	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		1848
		1849	Memory-maps a file (or anonymous memory range) and attaches it to the
		1850	given C<$scalar>, which will act like a string scalar.
		1851
		1852	The only operations allowed on the scalar are C<substr>/C<vec> that don't
		1853	change the string length, and most read-only operations such as copying it
		1854	or searching it with regexes and so on.
		1855
		1856	Anything else is unsafe and will, at best, result in memory leaks.
		1857
		1858	The memory map associated with the C<$scalar> is automatically removed
		1859	when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
		1860	C<IO::AIO::munmap> functions are called.
		1861
		1862	This calls the C<mmap>(2) function internally. See your system's manual
		1863	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		1864
		1865	The C<$length> must be larger than zero and smaller than the actual
		1866	filesize.
		1867
		1868	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		1869	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		1870
		1871	C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
		1872	C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
		1873	not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
		1874	(which is set to C<MAP_ANON> if your system only provides this
		1875	constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
		1876	C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
		1877	C<IO::AIO::MAP_NONBLOCK>
		1878
		1879	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		1880
		1881	C<$offset> is the offset from the start of the file - it generally must be
		1882	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		1883
		1884	Example:
		1885
		1886	use Digest::MD5;
		1887	use IO::AIO;
		1888
		1889	open my $fh, "<verybigfile"
		1890	or die "$!";
		1891
		1892	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		1893	or die "verybigfile: $!";
		1894
		1895	my $fast_md5 = md5 $data;
		1896
		1897	=item IO::AIO::munmap $scalar
		1898
		1899	Removes a previous mmap and undefines the C<$scalar>.
		1900
		1901	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		1902
		1903	Calls the C<munlock> function, undoing the effects of a previous
		1904	C<aio_mlock> call (see its description for details).
		1905
		1906	=item IO::AIO::munlockall
		1907
		1908	Calls the C<munlockall> function.
		1909
		1910	On systems that do not implement C<munlockall>, this function returns
		1911	ENOSYS, otherwise the return value of C<munlockall>.
		1912
		1913	=back
		1914
940	=cut	1915	=cut
941		1916
942	# support function to convert a fd into a perl filehandle
943	sub _fd2fh {
944	return undef if $_[0] < 0;
945
946	# try to generate nice filehandles
947	my $sym = "IO::AIO::fd#$_[0]";
948	local *$sym;
949
950	open *$sym, "+<&=$_[0]" # usually works under any unix
951	or open *$sym, "<&=$_[0]" # cygwin needs this
952	or open *$sym, ">&=$_[0]" # or this
953	or return undef;
954
955	*$sym
956	}
957
958	min_parallel 8;	1917	min_parallel 8;
959		1918
		1919	END { flush }
		1920
960	1;	1921	1;
961		1922
		1923	=head1 EVENT LOOP INTEGRATION
		1924
		1925	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		1926	automatically into many event loops:
		1927
		1928	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		1929	use AnyEvent::AIO;
		1930
		1931	You can also integrate IO::AIO manually into many event loops, here are
		1932	some examples of how to do this:
		1933
		1934	# EV integration
		1935	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		1936
		1937	# Event integration
		1938	Event->io (fd => IO::AIO::poll_fileno,
		1939	poll => 'r',
		1940	cb => \&IO::AIO::poll_cb);
		1941
		1942	# Glib/Gtk2 integration
		1943	add_watch Glib::IO IO::AIO::poll_fileno,
		1944	in => sub { IO::AIO::poll_cb; 1 };
		1945
		1946	# Tk integration
		1947	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		1948	readable => \&IO::AIO::poll_cb);
		1949
		1950	# Danga::Socket integration
		1951	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		1952	\&IO::AIO::poll_cb);
		1953
962	=head2 FORK BEHAVIOUR	1954	=head2 FORK BEHAVIOUR
963		1955
964	This module should do "the right thing" when the process using it forks:	1956	Usage of pthreads in a program changes the semantics of fork
		1957	considerably. Specifically, only async-safe functions can be called after
		1958	fork. Perl doesn't know about this, so in general, you cannot call fork
		1959	with defined behaviour in perl if pthreads are involved. IO::AIO uses
		1960	pthreads, so this applies, but many other extensions and (for inexplicable
		1961	reasons) perl itself often is linked against pthreads, so this limitation
		1962	applies to quite a lot of perls.
965		1963
966	Before the fork, IO::AIO enters a quiescent state where no requests	1964	This module no longer tries to fight your OS, or POSIX. That means IO::AIO
967	can be added in other threads and no results will be processed. After	1965	only works in the process that loaded it. Forking is fully supported, but
968	the fork the parent simply leaves the quiescent state and continues	1966	using IO::AIO in the child is not.
969	request/result processing, while the child frees the request/result queue
970	(so that the requests started before the fork will only be handled in the
971	parent). Threads will be started on demand until the limit set in the
972	parent process has been reached again.
973		1967
974	In short: the parent will, after a short pause, continue as if fork had	1968	You might get around by not I<using> IO::AIO before (or after)
975	not been called, while the child will act as if IO::AIO has not been used	1969	forking. You could also try to call the L<IO::AIO::reinit> function in the
976	yet.	1970	child:
		1971
		1972	=over 4
		1973
		1974	=item IO::AIO::reinit
		1975
		1976	Abandons all current requests and I/O threads and simply reinitialises all
		1977	data structures. This is not an operation supported by any standards, but
		1978	happens to work on GNU/Linux and some newer BSD systems.
		1979
		1980	The only reasonable use for this function is to call it after forking, if
		1981	C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
		1982	the process will result in undefined behaviour. Calling it at any time
		1983	will also result in any undefined (by POSIX) behaviour.
		1984
		1985	=back
977		1986
978	=head2 MEMORY USAGE	1987	=head2 MEMORY USAGE
979		1988
980	Per-request usage:	1989	Per-request usage:
981		1990
…		…
983	bytes of memory. In addition, stat requests need a stat buffer (possibly	1992	bytes of memory. In addition, stat requests need a stat buffer (possibly
984	a few hundred bytes), readdir requires a result buffer and so on. Perl	1993	a few hundred bytes), readdir requires a result buffer and so on. Perl
985	scalars and other data passed into aio requests will also be locked and	1994	scalars and other data passed into aio requests will also be locked and
986	will consume memory till the request has entered the done state.	1995	will consume memory till the request has entered the done state.
987		1996
988	This is now awfully much, so queuing lots of requests is not usually a	1997	This is not awfully much, so queuing lots of requests is not usually a
989	problem.	1998	problem.
990		1999
991	Per-thread usage:	2000	Per-thread usage:
992		2001
993	In the execution phase, some aio requests require more memory for	2002	In the execution phase, some aio requests require more memory for
…		…
998		2007
999	Known bugs will be fixed in the next release.	2008	Known bugs will be fixed in the next release.
1000		2009
1001	=head1 SEE ALSO	2010	=head1 SEE ALSO
1002		2011
1003	L<Coro::AIO>.	2012	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		2013	more natural syntax.
1004		2014
1005	=head1 AUTHOR	2015	=head1 AUTHOR
1006		2016
1007	Marc Lehmann <schmorp@schmorp.de>	2017	Marc Lehmann <schmorp@schmorp.de>
1008	http://home.schmorp.de/	2018	http://home.schmorp.de/

Diff Legend

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