ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/IO-AIO/AIO.pm

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.86 by root, Sat Oct 28 23:32:29 2006 UTC vs.
Revision 1.211 by root, Tue Sep 27 12:10:29 2011 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>).
55		36
56	Asynchronous means that operations that can normally block your program	37	Asynchronous means that operations that can normally block your program
57	(e.g. reading from disk) will be done asynchronously: the operation	38	(e.g. reading from disk) will be done asynchronously: the operation
58	will still block, but you can do something else in the meantime. This	39	will still block, but you can do something else in the meantime. This
59	is extremely useful for programs that need to stay interactive even	40	is extremely useful for programs that need to stay interactive even
…		…
61	etc.), but can also be used to easily do operations in parallel that are	42	etc.), but can also be used to easily do operations in parallel that are
62	normally done sequentially, e.g. stat'ing many files, which is much faster	43	normally done sequentially, e.g. stat'ing many files, which is much faster
63	on a RAID volume or over NFS when you do a number of stat operations	44	on a RAID volume or over NFS when you do a number of stat operations
64	concurrently.	45	concurrently.
65		46
66	While this works on all types of file descriptors (for example sockets),	47	While most of this works on all types of file descriptors (for
67	using these functions on file descriptors that support nonblocking	48	example sockets), using these functions on file descriptors that
68	operation (again, sockets, pipes etc.) is very inefficient. Use an event	49	support nonblocking operation (again, sockets, pipes etc.) is
69	loop for that (such as the L<Event|Event> module): IO::AIO will naturally	50	very inefficient. Use an event loop for that (such as the L<EV>
70	fit into such an event loop itself.	51	module): IO::AIO will naturally fit into such an event loop itself.
71		52
72	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
73	requests and signal their completion. You don't need thread support	54	requests and signal their completion. You don't need thread support
74	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
75	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
…		…
77	not well-supported or restricted (GNU/Linux doesn't allow them on normal	58	not well-supported or restricted (GNU/Linux doesn't allow them on normal
78	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
79	aio_write, so the remaining functionality would have to be implemented	60	aio_write, so the remaining functionality would have to be implemented
80	using threads anyway.	61	using threads anyway.
81		62
82	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,
83	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
84	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
85	never call C<poll_cb> (or other C<aio_> functions) recursively.	66	call C<poll_cb> (or other C<aio_> functions) recursively.
86		67
87	=head2 EXAMPLE	68	=head2 EXAMPLE
88		69
89	This is a simple example that uses the Event module and loads	70	This is a simple example that uses the EV module and loads
90	F</etc/passwd> asynchronously:	71	F</etc/passwd> asynchronously:
91		72
92	use Fcntl;	73	use Fcntl;
93	use Event;	74	use EV;
94	use IO::AIO;	75	use IO::AIO;
95		76
96	# register the IO::AIO callback with Event	77	# register the IO::AIO callback with EV
97	Event->io (fd => IO::AIO::poll_fileno,	78	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
98	poll => 'r',
99	cb => \&IO::AIO::poll_cb);
100		79
101	# queue the request to open /etc/passwd	80	# queue the request to open /etc/passwd
102	aio_open "/etc/passwd", O_RDONLY, 0, sub {	81	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
103	my $fh = $_[0]	82	my $fh = shift
104	or die "error while opening: $!";	83	or die "error while opening: $!";
105		84
106	# stat'ing filehandles is generally non-blocking	85	# stat'ing filehandles is generally non-blocking
107	my $size = -s $fh;	86	my $size = -s $fh;
108		87
…		…
116		95
117	# file contents now in $contents	96	# file contents now in $contents
118	print $contents;	97	print $contents;
119		98
120	# exit event loop and program	99	# exit event loop and program
121	Event::unloop;	100	EV::unloop;
122	};	101	};
123	};	102	};
124		103
125	# possibly queue up other requests, or open GUI windows,	104	# possibly queue up other requests, or open GUI windows,
126	# check for sockets etc. etc.	105	# check for sockets etc. etc.
127		106
128	# process events as long as there are some:	107	# process events as long as there are some:
129	Event::loop;	108	EV::loop;
130		109
131	=head1 REQUEST ANATOMY AND LIFETIME	110	=head1 REQUEST ANATOMY AND LIFETIME
132		111
133	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
134	directly visible to Perl.	113	directly visible to Perl.
…		…
176	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
177	(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
178	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
179	result in a runtime error).	158	result in a runtime error).
180		159
		160	=back
		161
181	=cut	162	=cut
182		163
183	package IO::AIO;	164	package IO::AIO;
184		165
185	no warnings;	166	use Carp ();
186	use strict 'vars';	167
		168	use common::sense;
187		169
188	use base 'Exporter';	170	use base 'Exporter';
189		171
190	BEGIN {	172	BEGIN {
191	our $VERSION = '2.1';	173	our $VERSION = '4.0';
192		174
193	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
194	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
195	aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move	177	aio_scandir aio_symlink aio_readlink aio_realpath aio_sync
196	aio_copy aio_group aio_nop aio_mknod);	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
		185	aio_wd);
		186
197	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));	187	our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
198	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush	188	our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
199	min_parallel max_parallel max_idle	189	min_parallel max_parallel max_idle idle_timeout
200	nreqs nready npending nthreads	190	nreqs nready npending nthreads
201	max_poll_time max_poll_reqs);	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
202		196
203	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';	197	@IO::AIO::GRP::ISA = 'IO::AIO::REQ';
204		198
205	require XSLoader;	199	require XSLoader;
206	XSLoader::load ("IO::AIO", $VERSION);	200	XSLoader::load ("IO::AIO", $VERSION);
207	}	201	}
208		202
209	=head1 FUNCTIONS	203	=head1 FUNCTIONS
210		204
		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_load $pathname, $data, $callback->($status)
		239	aio_copy $srcpath, $dstpath, $callback->($status)
		240	aio_move $srcpath, $dstpath, $callback->($status)
		241	aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
		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, $length = undef
		279	IO::AIO::munlockall
		280
211	=head2 AIO FUNCTIONS	281	=head2 AIO REQUEST FUNCTIONS
212		282
213	All the C<aio_*> calls are more or less thin wrappers around the syscall	283	All the C<aio_*> calls are more or less thin wrappers around the syscall
214	with the same name (sans C<aio_>). The arguments are similar or identical,	284	with the same name (sans C<aio_>). The arguments are similar or identical,
215	and they all accept an additional (and optional) C<$callback> argument	285	and they all accept an additional (and optional) C<$callback> argument
216	which must be a code reference. This code reference will get called with	286	which must be a code reference. This code reference will get called with
217	the syscall return code (e.g. most syscalls return C<-1> on error, unlike	287	the syscall return code (e.g. most syscalls return C<-1> on error, unlike
218	perl, which usually delivers "false") as it's sole argument when the given	288	perl, which usually delivers "false") as its sole argument after the given
219	syscall has been executed asynchronously.	289	syscall has been executed asynchronously.
220		290
221	All functions expecting a filehandle keep a copy of the filehandle	291	All functions expecting a filehandle keep a copy of the filehandle
222	internally until the request has finished.	292	internally until the request has finished.
223		293
224	All requests return objects of type L<IO::AIO::REQ> that allow further	294	All functions return request objects of type L<IO::AIO::REQ> that allow
225	manipulation of those requests while they are in-flight.	295	further manipulation of those requests while they are in-flight.
226		296
227	The pathnames you pass to these routines I<must> be absolute and	297	The pathnames you pass to these routines I<should> be absolute. The
228	encoded in byte form. The reason for the former is that at the time the	298	reason for this is that at the time the request is being executed, the
229	request is being executed, the current working directory could have	299	current working directory could have changed. Alternatively, you can make
230	changed. Alternatively, you can make sure that you never change the	300	sure that you never change the current working directory anywhere in
231	current working directory.	301	the program and then use relative paths. Lastly, you can take advantage
		302	of IO::AIOs working directory abstraction - see the description of the
		303	C<IO::AIO::WD> class later in this document.
232		304
233	To encode pathnames to byte form, either make sure you either: a)	305	To encode pathnames as octets, either make sure you either: a) always pass
234	always pass in filenames you got from outside (command line, readdir	306	in filenames you got from outside (command line, readdir etc.) without
235	etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode	307	tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
236	your pathnames to the locale (or other) encoding in effect in the user	308	your pathnames to the locale (or other) encoding in effect in the user
237	environment, d) use Glib::filename_from_unicode on unicode filenames or e)	309	environment, d) use Glib::filename_from_unicode on unicode filenames or e)
238	use something else.	310	use something else to ensure your scalar has the correct contents.
		311
		312	This works, btw. independent of the internal UTF-8 bit, which IO::AIO
		313	handles correctly whether it is set or not.
239		314
240	=over 4	315	=over 4
241		316
242	=item $prev_pri = aioreq_pri [$pri]	317	=item $prev_pri = aioreq_pri [$pri]
243		318
…		…
263	aio_read $_[0], ..., sub {	338	aio_read $_[0], ..., sub {
264	...	339	...
265	};	340	};
266	};	341	};
267		342
		343
268	=item aioreq_nice $pri_adjust	344	=item aioreq_nice $pri_adjust
269		345
270	Similar to C<aioreq_pri>, but subtracts the given value from the current	346	Similar to C<aioreq_pri>, but subtracts the given value from the current
271	priority, so effects are cumulative.	347	priority, so the effect is cumulative.
		348
272		349
273	=item aio_open $pathname, $flags, $mode, $callback->($fh)	350	=item aio_open $pathname, $flags, $mode, $callback->($fh)
274		351
275	Asynchronously open or create a file and call the callback with a newly	352	Asynchronously open or create a file and call the callback with a newly
276	created filehandle for the file.	353	created filehandle for the file.
…		…
282	list. They are the same as used by C<sysopen>.	359	list. They are the same as used by C<sysopen>.
283		360
284	Likewise, C<$mode> specifies the mode of the newly created file, if it	361	Likewise, C<$mode> specifies the mode of the newly created file, if it
285	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,	362	didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
286	except that it is mandatory (i.e. use C<0> if you don't create new files,	363	except that it is mandatory (i.e. use C<0> if you don't create new files,
287	and C<0666> or C<0777> if you do).	364	and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
		365	by the umask in effect then the request is being executed, so better never
		366	change the umask.
288		367
289	Example:	368	Example:
290		369
291	aio_open "/etc/passwd", O_RDONLY, 0, sub {	370	aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
292	if ($_[0]) {	371	if ($_[0]) {
293	print "open successful, fh is $_[0]\n";	372	print "open successful, fh is $_[0]\n";
294	...	373	...
295	} else {	374	} else {
296	die "open failed: $!\n";	375	die "open failed: $!\n";
297	}	376	}
298	};	377	};
299		378
		379	In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
		380	C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
		381	following POSIX and non-POSIX constants are available (missing ones on
		382	your system are, as usual, C<0>):
		383
		384	C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
		385	C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
		386	C<O_RSYNC>, C<O_SYNC> and C<O_TTY_INIT>.
		387
		388
300	=item aio_close $fh, $callback->($status)	389	=item aio_close $fh, $callback->($status)
301		390
302	Asynchronously close a file and call the callback with the result	391	Asynchronously close a file and call the callback with the result
303	code. I<WARNING:> although accepted, you should not pass in a perl	392	code.
304	filehandle here, as perl will likely close the file descriptor another
305	time when the filehandle is destroyed. Normally, you can safely call perls
306	C<close> or just let filehandles go out of scope.
307		393
308	This is supposed to be a bug in the API, so that might change. It's	394	Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
309	therefore best to avoid this function.	395	closing the file descriptor associated with the filehandle itself.
		396
		397	Therefore, C<aio_close> will not close the filehandle - instead it will
		398	use dup2 to overwrite the file descriptor with the write-end of a pipe
		399	(the pipe fd will be created on demand and will be cached).
		400
		401	Or in other words: the file descriptor will be closed, but it will not be
		402	free for reuse until the perl filehandle is closed.
		403
		404	=cut
310		405
311	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	406	=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
312		407
313	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)	408	=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
314		409
315	Reads or writes C<length> bytes from the specified C<fh> and C<offset>	410	Reads or writes C<$length> bytes from or to the specified C<$fh> and
316	into the scalar given by C<data> and offset C<dataoffset> and calls the	411	C<$offset> into the scalar given by C<$data> and offset C<$dataoffset>
317	callback without the actual number of bytes read (or -1 on error, just	412	and calls the callback without the actual number of bytes read (or -1 on
318	like the syscall).	413	error, just like the syscall).
		414
		415	C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
		416	offset plus the actual number of bytes read.
		417
		418	If C<$offset> is undefined, then the current file descriptor offset will
		419	be used (and updated), otherwise the file descriptor offset will not be
		420	changed by these calls.
		421
		422	If C<$length> is undefined in C<aio_write>, use the remaining length of
		423	C<$data>.
		424
		425	If C<$dataoffset> is less than zero, it will be counted from the end of
		426	C<$data>.
319		427
320	The C<$data> scalar I<MUST NOT> be modified in any way while the request	428	The C<$data> scalar I<MUST NOT> be modified in any way while the request
321	is outstanding. Modifying it can result in segfaults or WW3 (if the	429	is outstanding. Modifying it can result in segfaults or World War III (if
322	necessary/optional hardware is installed).	430	the necessary/optional hardware is installed).
323		431
324	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at	432	Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at
325	offset C<0> within the scalar:	433	offset C<0> within the scalar:
326		434
327	aio_read $fh, 7, 15, $buffer, 0, sub {	435	aio_read $fh, 7, 15, $buffer, 0, sub {
328	$_[0] > 0 or die "read error: $!";	436	$_[0] > 0 or die "read error: $!";
329	print "read $_[0] bytes: <$buffer>\n";	437	print "read $_[0] bytes: <$buffer>\n";
330	};	438	};
331		439
		440
332	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)	441	=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
333		442
334	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts	443	Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
335	reading at byte offset C<$in_offset>, and starts writing at the current	444	reading at byte offset C<$in_offset>, and starts writing at the current
336	file offset of C<$out_fh>. Because of that, it is not safe to issue more	445	file offset of C<$out_fh>. Because of that, it is not safe to issue more
337	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each	446	than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
338	other.	447	other. The same C<$in_fh> works fine though, as this function does not
		448	move or use the file offset of C<$in_fh>.
339		449
		450	Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
		451	are written, and there is no way to find out how many more bytes have been
		452	read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
		453	number of bytes written to C<$out_fh>. Only if the result value equals
		454	C<$length> one can assume that C<$length> bytes have been read.
		455
		456	Unlike with other C<aio_> functions, it makes a lot of sense to use
		457	C<aio_sendfile> on non-blocking sockets, as long as one end (typically
		458	the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
		459	the socket I/O will be non-blocking. Note, however, that you can run
		460	into a trap where C<aio_sendfile> reads some data with readahead, then
		461	fails to write all data, and when the socket is ready the next time, the
		462	data in the cache is already lost, forcing C<aio_sendfile> to again hit
		463	the disk. Explicit C<aio_read> + C<aio_write> let's you better control
		464	resource usage.
		465
340	This call tries to make use of a native C<sendfile> syscall to provide	466	This call tries to make use of a native C<sendfile>-like syscall to
341	zero-copy operation. For this to work, C<$out_fh> should refer to a	467	provide zero-copy operation. For this to work, C<$out_fh> should refer to
342	socket, and C<$in_fh> should refer to mmap'able file.	468	a socket, and C<$in_fh> should refer to an mmap'able file.
343		469
344	If the native sendfile call fails or is not implemented, it will be	470	If a native sendfile cannot be found or it fails with C<ENOSYS>,
345	emulated, so you can call C<aio_sendfile> on any type of filehandle	471	C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
		472	C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
346	regardless of the limitations of the operating system.	473	type of filehandle regardless of the limitations of the operating system.
347		474
348	Please note, however, that C<aio_sendfile> can read more bytes from	475	As native sendfile syscalls (as practically any non-POSIX interface hacked
349	C<$in_fh> than are written, and there is no way to find out how many	476	together in a hurry to improve benchmark numbers) tend to be rather buggy
350	bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only	477	on many systems, this implementation tries to work around some known bugs
351	provides the number of bytes written to C<$out_fh>. Only if the result	478	in Linux and FreeBSD kernels (probably others, too), but that might fail,
352	value equals C<$length> one can assume that C<$length> bytes have been	479	so you really really should check the return value of C<aio_sendfile> -
353	read.	480	fewre bytes than expected might have been transferred.
		481
354		482
355	=item aio_readahead $fh,$offset,$length, $callback->($retval)	483	=item aio_readahead $fh,$offset,$length, $callback->($retval)
356		484
357	C<aio_readahead> populates the page cache with data from a file so that	485	C<aio_readahead> populates the page cache with data from a file so that
358	subsequent reads from that file will not block on disk I/O. The C<$offset>	486	subsequent reads from that file will not block on disk I/O. The C<$offset>
…		…
364	file. The current file offset of the file is left unchanged.	492	file. The current file offset of the file is left unchanged.
365		493
366	If that syscall doesn't exist (likely if your OS isn't Linux) it will be	494	If that syscall doesn't exist (likely if your OS isn't Linux) it will be
367	emulated by simply reading the data, which would have a similar effect.	495	emulated by simply reading the data, which would have a similar effect.
368		496
		497
369	=item aio_stat $fh_or_path, $callback->($status)	498	=item aio_stat $fh_or_path, $callback->($status)
370		499
371	=item aio_lstat $fh, $callback->($status)	500	=item aio_lstat $fh, $callback->($status)
372		501
373	Works like perl's C<stat> or C<lstat> in void context. The callback will	502	Works like perl's C<stat> or C<lstat> in void context. The callback will
…		…
378	for an explanation.	507	for an explanation.
379		508
380	Currently, the stats are always 64-bit-stats, i.e. instead of returning an	509	Currently, the stats are always 64-bit-stats, i.e. instead of returning an
381	error when stat'ing a large file, the results will be silently truncated	510	error when stat'ing a large file, the results will be silently truncated
382	unless perl itself is compiled with large file support.	511	unless perl itself is compiled with large file support.
		512
		513	To help interpret the mode and dev/rdev stat values, IO::AIO offers the
		514	following constants and functions (if not implemented, the constants will
		515	be C<0> and the functions will either C<croak> or fall back on traditional
		516	behaviour).
		517
		518	C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
		519	C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
		520	C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
383		521
384	Example: Print the length of F</etc/passwd>:	522	Example: Print the length of F</etc/passwd>:
385		523
386	aio_stat "/etc/passwd", sub {	524	aio_stat "/etc/passwd", sub {
387	$_[0] and die "stat failed: $!";	525	$_[0] and die "stat failed: $!";
388	print "size is ", -s _, "\n";	526	print "size is ", -s _, "\n";
389	};	527	};
390		528
		529
		530	=item aio_statvfs $fh_or_path, $callback->($statvfs)
		531
		532	Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
		533	whether a file handle or path was passed.
		534
		535	On success, the callback is passed a hash reference with the following
		536	members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
		537	C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
		538	is passed.
		539
		540	The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
		541	C<ST_NOSUID>.
		542
		543	The following non-POSIX IO::AIO::ST_* flag masks are defined to
		544	their correct value when available, or to C<0> on systems that do
		545	not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
		546	C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
		547	C<ST_NODIRATIME> and C<ST_RELATIME>.
		548
		549	Example: stat C</wd> and dump out the data if successful.
		550
		551	aio_statvfs "/wd", sub {
		552	my $f = $_[0]
		553	or die "statvfs: $!";
		554
		555	use Data::Dumper;
		556	say Dumper $f;
		557	};
		558
		559	# result:
		560	{
		561	bsize => 1024,
		562	bfree => 4333064312,
		563	blocks => 10253828096,
		564	files => 2050765568,
		565	flag => 4096,
		566	favail => 2042092649,
		567	bavail => 4333064312,
		568	ffree => 2042092649,
		569	namemax => 255,
		570	frsize => 1024,
		571	fsid => 1810
		572	}
		573
		574
		575	=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
		576
		577	Works like perl's C<utime> function (including the special case of $atime
		578	and $mtime being undef). Fractional times are supported if the underlying
		579	syscalls support them.
		580
		581	When called with a pathname, uses utimes(2) if available, otherwise
		582	utime(2). If called on a file descriptor, uses futimes(2) if available,
		583	otherwise returns ENOSYS, so this is not portable.
		584
		585	Examples:
		586
		587	# set atime and mtime to current time (basically touch(1)):
		588	aio_utime "path", undef, undef;
		589	# set atime to current time and mtime to beginning of the epoch:
		590	aio_utime "path", time, undef; # undef==0
		591
		592
		593	=item aio_chown $fh_or_path, $uid, $gid, $callback->($status)
		594
		595	Works like perl's C<chown> function, except that C<undef> for either $uid
		596	or $gid is being interpreted as "do not change" (but -1 can also be used).
		597
		598	Examples:
		599
		600	# same as "chown root path" in the shell:
		601	aio_chown "path", 0, -1;
		602	# same as above:
		603	aio_chown "path", 0, undef;
		604
		605
		606	=item aio_truncate $fh_or_path, $offset, $callback->($status)
		607
		608	Works like truncate(2) or ftruncate(2).
		609
		610
		611	=item aio_chmod $fh_or_path, $mode, $callback->($status)
		612
		613	Works like perl's C<chmod> function.
		614
		615
391	=item aio_unlink $pathname, $callback->($status)	616	=item aio_unlink $pathname, $callback->($status)
392		617
393	Asynchronously unlink (delete) a file and call the callback with the	618	Asynchronously unlink (delete) a file and call the callback with the
394	result code.	619	result code.
395		620
		621
396	=item aio_mknod $path, $mode, $dev, $callback->($status)	622	=item aio_mknod $pathname, $mode, $dev, $callback->($status)
397		623
398	[EXPERIMENTAL]	624	[EXPERIMENTAL]
399		625
400	Asynchronously create a device node (or fifo). See mknod(2).	626	Asynchronously create a device node (or fifo). See mknod(2).
401		627
402	The only (POSIX-) portable way of calling this function is:	628	The only (POSIX-) portable way of calling this function is:
403		629
404	aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...	630	aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
		631
		632	See C<aio_stat> for info about some potentially helpful extra constants
		633	and functions.
405		634
406	=item aio_link $srcpath, $dstpath, $callback->($status)	635	=item aio_link $srcpath, $dstpath, $callback->($status)
407		636
408	Asynchronously create a new link to the existing object at C<$srcpath> at	637	Asynchronously create a new link to the existing object at C<$srcpath> at
409	the path C<$dstpath> and call the callback with the result code.	638	the path C<$dstpath> and call the callback with the result code.
410		639
		640
411	=item aio_symlink $srcpath, $dstpath, $callback->($status)	641	=item aio_symlink $srcpath, $dstpath, $callback->($status)
412		642
413	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at	643	Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
414	the path C<$dstpath> and call the callback with the result code.	644	the path C<$dstpath> and call the callback with the result code.
415		645
		646
		647	=item aio_readlink $pathname, $callback->($link)
		648
		649	Asynchronously read the symlink specified by C<$path> and pass it to
		650	the callback. If an error occurs, nothing or undef gets passed to the
		651	callback.
		652
		653
		654	=item aio_realpath $pathname, $callback->($path)
		655
		656	Asynchronously make the path absolute and resolve any symlinks in
		657	C<$path>. The resulting path only consists of directories (Same as
		658	L<Cwd::realpath>).
		659
		660	This request can be used to get the absolute path of the current working
		661	directory by passing it a path of F<.> (a single dot).
		662
		663
416	=item aio_rename $srcpath, $dstpath, $callback->($status)	664	=item aio_rename $srcpath, $dstpath, $callback->($status)
417		665
418	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as	666	Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
419	rename(2) and call the callback with the result code.	667	rename(2) and call the callback with the result code.
420		668
		669
		670	=item aio_mkdir $pathname, $mode, $callback->($status)
		671
		672	Asynchronously mkdir (create) a directory and call the callback with
		673	the result code. C<$mode> will be modified by the umask at the time the
		674	request is executed, so do not change your umask.
		675
		676
421	=item aio_rmdir $pathname, $callback->($status)	677	=item aio_rmdir $pathname, $callback->($status)
422		678
423	Asynchronously rmdir (delete) a directory and call the callback with the	679	Asynchronously rmdir (delete) a directory and call the callback with the
424	result code.	680	result code.
		681
425		682
426	=item aio_readdir $pathname, $callback->($entries)	683	=item aio_readdir $pathname, $callback->($entries)
427		684
428	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire	685	Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
429	directory (i.e. opendir + readdir + closedir). The entries will not be	686	directory (i.e. opendir + readdir + closedir). The entries will not be
430	sorted, and will B<NOT> include the C<.> and C<..> entries.	687	sorted, and will B<NOT> include the C<.> and C<..> entries.
431		688
432	The callback a single argument which is either C<undef> or an array-ref	689	The callback is passed a single argument which is either C<undef> or an
433	with the filenames.	690	array-ref with the filenames.
		691
		692
		693	=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
		694
		695	Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
		696	tune behaviour and output format. In case of an error, C<$entries> will be
		697	C<undef>.
		698
		699	The flags are a combination of the following constants, ORed together (the
		700	flags will also be passed to the callback, possibly modified):
		701
		702	=over 4
		703
		704	=item IO::AIO::READDIR_DENTS
		705
		706	When this flag is off, then the callback gets an arrayref consisting of
		707	names only (as with C<aio_readdir>), otherwise it gets an arrayref with
		708	C<[$name, $type, $inode]> arrayrefs, each describing a single directory
		709	entry in more detail.
		710
		711	C<$name> is the name of the entry.
		712
		713	C<$type> is one of the C<IO::AIO::DT_xxx> constants:
		714
		715	C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
		716	C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
		717	C<IO::AIO::DT_WHT>.
		718
		719	C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
		720	know, you have to run stat yourself. Also, for speed reasons, the C<$type>
		721	scalars are read-only: you can not modify them.
		722
		723	C<$inode> is the inode number (which might not be exact on systems with 64
		724	bit inode numbers and 32 bit perls). This field has unspecified content on
		725	systems that do not deliver the inode information.
		726
		727	=item IO::AIO::READDIR_DIRS_FIRST
		728
		729	When this flag is set, then the names will be returned in an order where
		730	likely directories come first, in optimal stat order. This is useful when
		731	you need to quickly find directories, or you want to find all directories
		732	while avoiding to stat() each entry.
		733
		734	If the system returns type information in readdir, then this is used
		735	to find directories directly. Otherwise, likely directories are names
		736	beginning with ".", or otherwise names with no dots, of which names with
		737	short names are tried first.
		738
		739	=item IO::AIO::READDIR_STAT_ORDER
		740
		741	When this flag is set, then the names will be returned in an order
		742	suitable for stat()'ing each one. That is, when you plan to stat()
		743	all files in the given directory, then the returned order will likely
		744	be fastest.
		745
		746	If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
		747	the likely dirs come first, resulting in a less optimal stat order.
		748
		749	=item IO::AIO::READDIR_FOUND_UNKNOWN
		750
		751	This flag should not be set when calling C<aio_readdirx>. Instead, it
		752	is being set by C<aio_readdirx>, when any of the C<$type>'s found were
		753	C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
		754	C<$type>'s are known, which can be used to speed up some algorithms.
		755
		756	=back
		757
		758
		759	=item aio_load $pathname, $data, $callback->($status)
		760
		761	This is a composite request that tries to fully load the given file into
		762	memory. Status is the same as with aio_read.
		763
		764	=cut
		765
		766	sub aio_load($$;$) {
		767	my ($path, undef, $cb) = @_;
		768	my $data = \$_[1];
		769
		770	my $pri = aioreq_pri;
		771	my $grp = aio_group $cb;
		772
		773	aioreq_pri $pri;
		774	add $grp aio_open $path, O_RDONLY, 0, sub {
		775	my $fh = shift
		776	or return $grp->result (-1);
		777
		778	aioreq_pri $pri;
		779	add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
		780	$grp->result ($_[0]);
		781	};
		782	};
		783
		784	$grp
		785	}
434		786
435	=item aio_copy $srcpath, $dstpath, $callback->($status)	787	=item aio_copy $srcpath, $dstpath, $callback->($status)
436		788
437	Try to copy the I<file> (directories not supported as either source or	789	Try to copy the I<file> (directories not supported as either source or
438	destination) from C<$srcpath> to C<$dstpath> and call the callback with	790	destination) from C<$srcpath> to C<$dstpath> and call the callback with
439	the C<0> (error) or C<-1> ok.	791	a status of C<0> (ok) or C<-1> (error, see C<$!>).
440		792
441	This is a composite request that it creates the destination file with	793	This is a composite request that creates the destination file with
442	mode 0200 and copies the contents of the source file into it using	794	mode 0200 and copies the contents of the source file into it using
443	C<aio_sendfile>, followed by restoring atime, mtime, access mode and	795	C<aio_sendfile>, followed by restoring atime, mtime, access mode and
444	uid/gid, in that order.	796	uid/gid, in that order.
445		797
446	If an error occurs, the partial destination file will be unlinked, if	798	If an error occurs, the partial destination file will be unlinked, if
…		…
456	my $grp = aio_group $cb;	808	my $grp = aio_group $cb;
457		809
458	aioreq_pri $pri;	810	aioreq_pri $pri;
459	add $grp aio_open $src, O_RDONLY, 0, sub {	811	add $grp aio_open $src, O_RDONLY, 0, sub {
460	if (my $src_fh = $_[0]) {	812	if (my $src_fh = $_[0]) {
461	my @stat = stat $src_fh;	813	my @stat = stat $src_fh; # hmm, might block over nfs?
462		814
463	aioreq_pri $pri;	815	aioreq_pri $pri;
464	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {	816	add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
465	if (my $dst_fh = $_[0]) {	817	if (my $dst_fh = $_[0]) {
466	aioreq_pri $pri;	818	aioreq_pri $pri;
467	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {	819	add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
468	if ($_[0] == $stat[7]) {	820	if ($_[0] == $stat[7]) {
469	$grp->result (0);	821	$grp->result (0);
470	close $src_fh;	822	close $src_fh;
471		823
472	# those should not normally block. should. should.	824	my $ch = sub {
		825	aioreq_pri $pri;
		826	add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
		827	aioreq_pri $pri;
		828	add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
		829	aioreq_pri $pri;
		830	add $grp aio_close $dst_fh;
		831	}
		832	};
		833	};
		834
		835	aioreq_pri $pri;
		836	add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
		837	if ($_[0] < 0 && $! == ENOSYS) {
		838	aioreq_pri $pri;
473	utime $stat[8], $stat[9], $dst;	839	add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
474	chmod $stat[2] & 07777, $dst_fh;	840	} else {
475	chown $stat[4], $stat[5], $dst_fh;	841	$ch->();
		842	}
476	close $dst_fh;	843	};
477	} else {	844	} else {
478	$grp->result (-1);	845	$grp->result (-1);
479	close $src_fh;	846	close $src_fh;
480	close $dst_fh;	847	close $dst_fh;
481		848
…		…
498		865
499	=item aio_move $srcpath, $dstpath, $callback->($status)	866	=item aio_move $srcpath, $dstpath, $callback->($status)
500		867
501	Try to move the I<file> (directories not supported as either source or	868	Try to move the I<file> (directories not supported as either source or
502	destination) from C<$srcpath> to C<$dstpath> and call the callback with	869	destination) from C<$srcpath> to C<$dstpath> and call the callback with
503	the C<0> (error) or C<-1> ok.	870	a status of C<0> (ok) or C<-1> (error, see C<$!>).
504		871
505	This is a composite request that tries to rename(2) the file first. If	872	This is a composite request that tries to rename(2) the file first; if
506	rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if	873	rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
507	that is successful, unlinking the C<$srcpath>.	874	that is successful, unlinks the C<$srcpath>.
508		875
509	=cut	876	=cut
510		877
511	sub aio_move($$;$) {	878	sub aio_move($$;$) {
512	my ($src, $dst, $cb) = @_;	879	my ($src, $dst, $cb) = @_;
…		…
519	if ($_[0] && $! == EXDEV) {	886	if ($_[0] && $! == EXDEV) {
520	aioreq_pri $pri;	887	aioreq_pri $pri;
521	add $grp aio_copy $src, $dst, sub {	888	add $grp aio_copy $src, $dst, sub {
522	$grp->result ($_[0]);	889	$grp->result ($_[0]);
523		890
524	if (!$_[0]) {	891	unless ($_[0]) {
525	aioreq_pri $pri;	892	aioreq_pri $pri;
526	add $grp aio_unlink $src;	893	add $grp aio_unlink $src;
527	}	894	}
528	};	895	};
529	} else {	896	} else {
…		…
532	};	899	};
533		900
534	$grp	901	$grp
535	}	902	}
536		903
537	=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)	904	=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
538		905
539	Scans a directory (similar to C<aio_readdir>) but additionally tries to	906	Scans a directory (similar to C<aio_readdir>) but additionally tries to
540	efficiently separate the entries of directory C<$path> into two sets of	907	efficiently separate the entries of directory C<$path> into two sets of
541	names, directories you can recurse into (directories), and ones you cannot	908	names, directories you can recurse into (directories), and ones you cannot
542	recurse into (everything else, including symlinks to directories).	909	recurse into (everything else, including symlinks to directories).
…		…
559		926
560	Implementation notes.	927	Implementation notes.
561		928
562	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.	929	The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
563		930
		931	If readdir returns file type information, then this is used directly to
		932	find directories.
		933
564	After reading the directory, the modification time, size etc. of the	934	Otherwise, after reading the directory, the modification time, size etc.
565	directory before and after the readdir is checked, and if they match (and	935	of the directory before and after the readdir is checked, and if they
566	isn't the current time), the link count will be used to decide how many	936	match (and isn't the current time), the link count will be used to decide
567	entries are directories (if >= 2). Otherwise, no knowledge of the number	937	how many entries are directories (if >= 2). Otherwise, no knowledge of the
568	of subdirectories will be assumed.	938	number of subdirectories will be assumed.
569		939
570	Then entries will be sorted into likely directories (everything without	940	Then entries will be sorted into likely directories a non-initial dot
571	a non-initial dot currently) and likely non-directories (everything	941	currently) and likely non-directories (see C<aio_readdirx>). Then every
572	else). Then every entry plus an appended C</.> will be C<stat>'ed,	942	entry plus an appended C</.> will be C<stat>'ed, likely directories first,
573	likely directories first. If that succeeds, it assumes that the entry	943	in order of their inode numbers. If that succeeds, it assumes that the
574	is a directory or a symlink to directory (which will be checked	944	entry is a directory or a symlink to directory (which will be checked
575	seperately). This is often faster than stat'ing the entry itself because	945	separately). This is often faster than stat'ing the entry itself because
576	filesystems might detect the type of the entry without reading the inode	946	filesystems might detect the type of the entry without reading the inode
577	data (e.g. ext2fs filetype feature).	947	data (e.g. ext2fs filetype feature), even on systems that cannot return
		948	the filetype information on readdir.
578		949
579	If the known number of directories (link count - 2) has been reached, the	950	If the known number of directories (link count - 2) has been reached, the
580	rest of the entries is assumed to be non-directories.	951	rest of the entries is assumed to be non-directories.
581		952
582	This only works with certainty on POSIX (= UNIX) filesystems, which	953	This only works with certainty on POSIX (= UNIX) filesystems, which
…		…
586	as those tend to return 0 or 1 as link counts, which disables the	957	as those tend to return 0 or 1 as link counts, which disables the
587	directory counting heuristic.	958	directory counting heuristic.
588		959
589	=cut	960	=cut
590		961
591	sub aio_scandir($$$) {	962	sub aio_scandir($$;$) {
592	my ($path, $maxreq, $cb) = @_;	963	my ($path, $maxreq, $cb) = @_;
593		964
594	my $pri = aioreq_pri;	965	my $pri = aioreq_pri;
595		966
596	my $grp = aio_group $cb;	967	my $grp = aio_group $cb;
597		968
598	$maxreq = 4 if $maxreq <= 0;	969	$maxreq = 4 if $maxreq <= 0;
599		970
600	# stat once	971	# get a wd object
		972
601	aioreq_pri $pri;	973	aioreq_pri $pri;
602	add $grp aio_stat $path, sub {	974	add $grp aio_wd $path, sub {
603	return $grp->result () if $_[0];	975	my $wd = [shift, "."];
604	my $now = time;
605	my $hash1 = join ":", (stat _)[0,1,3,7,9];
606		976
607	# read the directory entries	977	# stat once
608	aioreq_pri $pri;	978	aioreq_pri $pri;
609	add $grp aio_readdir $path, sub {	979	add $grp aio_stat $wd, sub {
610	my $entries = shift
611	or return $grp->result ();	980	return $grp->result () if $_[0];
		981	my $now = time;
		982	my $hash1 = join ":", (stat _)[0,1,3,7,9];
612		983
613	# stat the dir another time	984	# read the directory entries
614	aioreq_pri $pri;	985	aioreq_pri $pri;
		986	add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
		987	my $entries = shift
		988	or return $grp->result ();
		989
		990	# stat the dir another time
		991	aioreq_pri $pri;
615	add $grp aio_stat $path, sub {	992	add $grp aio_stat $wd, sub {
616	my $hash2 = join ":", (stat _)[0,1,3,7,9];	993	my $hash2 = join ":", (stat _)[0,1,3,7,9];
617		994
618	my $ndirs;	995	my $ndirs;
619		996
620	# take the slow route if anything looks fishy	997	# take the slow route if anything looks fishy
621	if ($hash1 ne $hash2 or (stat _)[9] == $now) {	998	if ($hash1 ne $hash2 or (stat _)[9] == $now) {
622	$ndirs = -1;	999	$ndirs = -1;
623	} else {	1000	} else {
624	# if nlink == 2, we are finished	1001	# if nlink == 2, we are finished
625	# on non-posix-fs's, we rely on nlink < 2	1002	# for non-posix-fs's, we rely on nlink < 2
626	$ndirs = (stat _)[3] - 2	1003	$ndirs = (stat _)[3] - 2
627	or return $grp->result ([], $entries);	1004	or return $grp->result ([], $entries);
628	}	1005	}
629		1006
630	# sort into likely dirs and likely nondirs
631	# dirs == files without ".", short entries first
632	$entries = [map $_->[0],
633	sort { $b->[1] cmp $a->[1] }
634	map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
635	@$entries];
636
637	my (@dirs, @nondirs);	1007	my (@dirs, @nondirs);
638		1008
639	my $statgrp = add $grp aio_group sub {	1009	my $statgrp = add $grp aio_group sub {
640	$grp->result (\@dirs, \@nondirs);	1010	$grp->result (\@dirs, \@nondirs);
641	};	1011	};
642		1012
643	limit $statgrp $maxreq;	1013	limit $statgrp $maxreq;
644	feed $statgrp sub {	1014	feed $statgrp sub {
645	return unless @$entries;	1015	return unless @$entries;
646	my $entry = pop @$entries;	1016	my $entry = shift @$entries;
647		1017
648	aioreq_pri $pri;	1018	aioreq_pri $pri;
		1019	$wd->[1] = "$entry/.";
649	add $statgrp aio_stat "$path/$entry/.", sub {	1020	add $statgrp aio_stat $wd, sub {
650	if ($_[0] < 0) {	1021	if ($_[0] < 0) {
651	push @nondirs, $entry;	1022	push @nondirs, $entry;
652	} else {	1023	} else {
653	# need to check for real directory	1024	# need to check for real directory
654	aioreq_pri $pri;	1025	aioreq_pri $pri;
		1026	$wd->[1] = $entry;
655	add $statgrp aio_lstat "$path/$entry", sub {	1027	add $statgrp aio_lstat $wd, sub {
656	if (-d _) {	1028	if (-d _) {
657	push @dirs, $entry;	1029	push @dirs, $entry;
658		1030
659	unless (--$ndirs) {	1031	unless (--$ndirs) {
660	push @nondirs, @$entries;	1032	push @nondirs, @$entries;
661	feed $statgrp;	1033	feed $statgrp;
		1034	}
		1035	} else {
		1036	push @nondirs, $entry;
662	}	1037	}
663	} else {
664	push @nondirs, $entry;
665	}	1038	}
666	}	1039	}
667	}	1040	};
668	};	1041	};
669	};	1042	};
670	};	1043	};
671	};	1044	};
672	};	1045	};
673		1046
674	$grp	1047	$grp
675	}	1048	}
676		1049
		1050	=item aio_rmtree $pathname, $callback->($status)
		1051
		1052	Delete a directory tree starting (and including) C<$path>, return the
		1053	status of the final C<rmdir> only. This is a composite request that
		1054	uses C<aio_scandir> to recurse into and rmdir directories, and unlink
		1055	everything else.
		1056
		1057	=cut
		1058
		1059	sub aio_rmtree;
		1060	sub aio_rmtree($;$) {
		1061	my ($path, $cb) = @_;
		1062
		1063	my $pri = aioreq_pri;
		1064	my $grp = aio_group $cb;
		1065
		1066	aioreq_pri $pri;
		1067	add $grp aio_scandir $path, 0, sub {
		1068	my ($dirs, $nondirs) = @_;
		1069
		1070	my $dirgrp = aio_group sub {
		1071	add $grp aio_rmdir $path, sub {
		1072	$grp->result ($_[0]);
		1073	};
		1074	};
		1075
		1076	(aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
		1077	(aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
		1078
		1079	add $grp $dirgrp;
		1080	};
		1081
		1082	$grp
		1083	}
		1084
		1085	=item aio_sync $callback->($status)
		1086
		1087	Asynchronously call sync and call the callback when finished.
		1088
677	=item aio_fsync $fh, $callback->($status)	1089	=item aio_fsync $fh, $callback->($status)
678		1090
679	Asynchronously call fsync on the given filehandle and call the callback	1091	Asynchronously call fsync on the given filehandle and call the callback
680	with the fsync result code.	1092	with the fsync result code.
681		1093
…		…
684	Asynchronously call fdatasync on the given filehandle and call the	1096	Asynchronously call fdatasync on the given filehandle and call the
685	callback with the fdatasync result code.	1097	callback with the fdatasync result code.
686		1098
687	If this call isn't available because your OS lacks it or it couldn't be	1099	If this call isn't available because your OS lacks it or it couldn't be
688	detected, it will be emulated by calling C<fsync> instead.	1100	detected, it will be emulated by calling C<fsync> instead.
		1101
		1102	=item aio_syncfs $fh, $callback->($status)
		1103
		1104	Asynchronously call the syncfs syscall to sync the filesystem associated
		1105	to the given filehandle and call the callback with the syncfs result
		1106	code. If syncfs is not available, calls sync(), but returns C<-1> and sets
		1107	errno to C<ENOSYS> nevertheless.
		1108
		1109	=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
		1110
		1111	Sync the data portion of the file specified by C<$offset> and C<$length>
		1112	to disk (but NOT the metadata), by calling the Linux-specific
		1113	sync_file_range call. If sync_file_range is not available or it returns
		1114	ENOSYS, then fdatasync or fsync is being substituted.
		1115
		1116	C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
		1117	C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
		1118	C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
		1119	manpage for details.
		1120
		1121	=item aio_pathsync $pathname, $callback->($status)
		1122
		1123	This request tries to open, fsync and close the given path. This is a
		1124	composite request intended to sync directories after directory operations
		1125	(E.g. rename). This might not work on all operating systems or have any
		1126	specific effect, but usually it makes sure that directory changes get
		1127	written to disc. It works for anything that can be opened for read-only,
		1128	not just directories.
		1129
		1130	Future versions of this function might fall back to other methods when
		1131	C<fsync> on the directory fails (such as calling C<sync>).
		1132
		1133	Passes C<0> when everything went ok, and C<-1> on error.
		1134
		1135	=cut
		1136
		1137	sub aio_pathsync($;$) {
		1138	my ($path, $cb) = @_;
		1139
		1140	my $pri = aioreq_pri;
		1141	my $grp = aio_group $cb;
		1142
		1143	aioreq_pri $pri;
		1144	add $grp aio_open $path, O_RDONLY, 0, sub {
		1145	my ($fh) = @_;
		1146	if ($fh) {
		1147	aioreq_pri $pri;
		1148	add $grp aio_fsync $fh, sub {
		1149	$grp->result ($_[0]);
		1150
		1151	aioreq_pri $pri;
		1152	add $grp aio_close $fh;
		1153	};
		1154	} else {
		1155	$grp->result (-1);
		1156	}
		1157	};
		1158
		1159	$grp
		1160	}
		1161
		1162	=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1163
		1164	This is a rather advanced IO::AIO call, which only works on mmap(2)ed
		1165	scalars (see the C<IO::AIO::mmap> function, although it also works on data
		1166	scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
		1167	scalar must only be modified in-place while an aio operation is pending on
		1168	it).
		1169
		1170	It calls the C<msync> function of your OS, if available, with the memory
		1171	area starting at C<$offset> in the string and ending C<$length> bytes
		1172	later. If C<$length> is negative, counts from the end, and if C<$length>
		1173	is C<undef>, then it goes till the end of the string. The flags can be
		1174	a combination of C<IO::AIO::MS_ASYNC>, C<IO::AIO::MS_INVALIDATE> and
		1175	C<IO::AIO::MS_SYNC>.
		1176
		1177	=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
		1178
		1179	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1180	scalars.
		1181
		1182	It touches (reads or writes) all memory pages in the specified
		1183	range inside the scalar. All caveats and parameters are the same
		1184	as for C<aio_msync>, above, except for flags, which must be either
		1185	C<0> (which reads all pages and ensures they are instantiated) or
		1186	C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
		1187	writing an octet from it, which dirties the page).
		1188
		1189	=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
		1190
		1191	This is a rather advanced IO::AIO call, which works best on mmap(2)ed
		1192	scalars.
		1193
		1194	It reads in all the pages of the underlying storage into memory (if any)
		1195	and locks them, so they are not getting swapped/paged out or removed.
		1196
		1197	If C<$length> is undefined, then the scalar will be locked till the end.
		1198
		1199	On systems that do not implement C<mlock>, this function returns C<-1>
		1200	and sets errno to C<ENOSYS>.
		1201
		1202	Note that the corresponding C<munlock> is synchronous and is
		1203	documented under L<MISCELLANEOUS FUNCTIONS>.
		1204
		1205	Example: open a file, mmap and mlock it - both will be undone when
		1206	C<$data> gets destroyed.
		1207
		1208	open my $fh, "<", $path or die "$path: $!";
		1209	my $data;
		1210	IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
		1211	aio_mlock $data; # mlock in background
		1212
		1213	=item aio_mlockall $flags, $callback->($status)
		1214
		1215	Calls the C<mlockall> function with the given C<$flags> (a combination of
		1216	C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
		1217
		1218	On systems that do not implement C<mlockall>, this function returns C<-1>
		1219	and sets errno to C<ENOSYS>.
		1220
		1221	Note that the corresponding C<munlockall> is synchronous and is
		1222	documented under L<MISCELLANEOUS FUNCTIONS>.
		1223
		1224	Example: asynchronously lock all current and future pages into memory.
		1225
		1226	aio_mlockall IO::AIO::MCL_FUTURE;
689		1227
690	=item aio_group $callback->(...)	1228	=item aio_group $callback->(...)
691		1229
692	This is a very special aio request: Instead of doing something, it is a	1230	This is a very special aio request: Instead of doing something, it is a
693	container for other aio requests, which is useful if you want to bundle	1231	container for other aio requests, which is useful if you want to bundle
…		…
731	immense (it blocks a thread for a long time) so do not use this function	1269	immense (it blocks a thread for a long time) so do not use this function
732	except to put your application under artificial I/O pressure.	1270	except to put your application under artificial I/O pressure.
733		1271
734	=back	1272	=back
735		1273
		1274
		1275	=head2 IO::AIO::WD - multiple working directories
		1276
		1277	Your process only has one current working directory, which is used by all
		1278	threads. This makes it hard to use relative paths (some other component
		1279	could call C<chdir> at any time, and it is hard to control when the path
		1280	will be used by IO::AIO).
		1281
		1282	One solution for this is to always use absolute paths. This usually works,
		1283	but can be quite slow (the kernel has to walk the whole path on every
		1284	access), and can also be a hassle to implement.
		1285
		1286	Newer POSIX systems have a number of functions (openat, fdopendir,
		1287	futimensat and so on) that make it possible to specify working directories
		1288	per operation.
		1289
		1290	For portability, and because the clowns who "designed", or shall I write,
		1291	perpetrated this new interface were obviously half-drunk, this abstraction
		1292	cannot be perfect, though.
		1293
		1294	IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
		1295	object. This object stores the canonicalised, absolute version of the
		1296	path, and on systems that allow it, also a directory file descriptor.
		1297
		1298	Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
		1299	or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
		1300	object and a pathname instead. If the pathname is absolute, the
		1301	IO::AIO::WD objetc is ignored, otherwise the pathname is resolved relative
		1302	to that IO::AIO::WD object.
		1303
		1304	For example, to get a wd object for F</etc> and then stat F<passwd>
		1305	inside, you would write:
		1306
		1307	aio_wd "/etc", sub {
		1308	my $etcdir = shift;
		1309
		1310	# although $etcdir can be undef on error, there is generally no reason
		1311	# to check for errors here, as aio_stat will fail with ENOENT
		1312	# when $etcdir is undef.
		1313
		1314	aio_stat [$etcdir, "passwd"], sub {
		1315	# yay
		1316	};
		1317	};
		1318
		1319	This shows that creating an IO::AIO::WD object is itself a potentially
		1320	blocking operation, which is why it is done asynchronously.
		1321
		1322	As with normal pathnames, IO::AIO keeps a copy of the working directory
		1323	object and the pathname string, so you could write the following without
		1324	causing any issues due to C<$path> getting reused:
		1325
		1326	my $path = [$wd, undef];
		1327
		1328	for my $name (qw(abc def ghi)) {
		1329	$path->[1] = $name;
		1330	aio_stat $path, sub {
		1331	# ...
		1332	};
		1333	}
		1334
		1335	There are some caveats: when directories get renamed (or deleted), the
		1336	pathname string doesn't change, so will point to the new directory (or
		1337	nowhere at all), while the directory fd, if available on the system,
		1338	will still point to the original directory. Most functions accepting a
		1339	pathname will use the directory fd on newer systems, and the string on
		1340	older systems. Some functions (such as realpath) will always rely on the
		1341	string form of the pathname.
		1342
		1343	So this fucntionality is mainly useful to get some protection against
		1344	C<chdir>, to easily get an absolute path out of a relative path for future
		1345	reference, and to speed up doing many operations in the same directory
		1346	(e.g. when stat'ing all files in a directory).
		1347
		1348	The following functions implement this working directory abstraction:
		1349
		1350	=over 4
		1351
		1352	=item aio_wd $pathname, $callback->($wd)
		1353
		1354	Asynchonously canonicalise the given pathname and convert it to an
		1355	IO::AIO::WD object representing it. If possible and supported on the
		1356	system, also open a directory fd to speed up pathname resolution relative
		1357	to this working directory.
		1358
		1359	If something goes wrong, then C<undef> is passwd to the callback instead
		1360	of a working directory object and C<$!> is set appropriately. Since
		1361	passing C<undef> as working directory component of a pathname fails the
		1362	request with C<ENOENT>, there is often no need for error checking in the
		1363	C<aio_wd> callback, as future requests using the value will fail in the
		1364	expected way.
		1365
		1366	If this call isn't available because your OS lacks it or it couldn't be
		1367	detected, it will be emulated by calling C<fsync> instead.
		1368
		1369	=item IO::AIO::CWD
		1370
		1371	This is a compiletime constant (object) that represents the process
		1372	current working directory.
		1373
		1374	Specifying this object as working directory object for a pathname is as
		1375	if the pathname would be specified directly, without a directory object,
		1376	e.g., these calls are functionally identical:
		1377
		1378	aio_stat "somefile", sub { ... };
		1379	aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
		1380
		1381	=back
		1382
		1383
736	=head2 IO::AIO::REQ CLASS	1384	=head2 IO::AIO::REQ CLASS
737		1385
738	All non-aggregate C<aio_*> functions return an object of this class when	1386	All non-aggregate C<aio_*> functions return an object of this class when
739	called in non-void context.	1387	called in non-void context.
740		1388
…		…
743	=item cancel $req	1391	=item cancel $req
744		1392
745	Cancels the request, if possible. Has the effect of skipping execution	1393	Cancels the request, if possible. Has the effect of skipping execution
746	when entering the B<execute> state and skipping calling the callback when	1394	when entering the B<execute> state and skipping calling the callback when
747	entering the the B<result> state, but will leave the request otherwise	1395	entering the the B<result> state, but will leave the request otherwise
748	untouched. That means that requests that currently execute will not be	1396	untouched (with the exception of readdir). That means that requests that
749	stopped and resources held by the request will not be freed prematurely.	1397	currently execute will not be stopped and resources held by the request
		1398	will not be freed prematurely.
750		1399
751	=item cb $req $callback->(...)	1400	=item cb $req $callback->(...)
752		1401
753	Replace (or simply set) the callback registered to the request.	1402	Replace (or simply set) the callback registered to the request.
754		1403
…		…
805	Their lifetime, simplified, looks like this: when they are empty, they	1454	Their lifetime, simplified, looks like this: when they are empty, they
806	will finish very quickly. If they contain only requests that are in the	1455	will finish very quickly. If they contain only requests that are in the
807	C<done> state, they will also finish. Otherwise they will continue to	1456	C<done> state, they will also finish. Otherwise they will continue to
808	exist.	1457	exist.
809		1458
810	That means after creating a group you have some time to add requests. And	1459	That means after creating a group you have some time to add requests
811	in the callbacks of those requests, you can add further requests to the	1460	(precisely before the callback has been invoked, which is only done within
812	group. And only when all those requests have finished will the the group	1461	the C<poll_cb>). And in the callbacks of those requests, you can add
813	itself finish.	1462	further requests to the group. And only when all those requests have
		1463	finished will the the group itself finish.
814		1464
815	=over 4	1465	=over 4
816		1466
817	=item add $grp ...	1467	=item add $grp ...
818		1468
…		…
827	=item $grp->cancel_subs	1477	=item $grp->cancel_subs
828		1478
829	Cancel all subrequests and clears any feeder, but not the group request	1479	Cancel all subrequests and clears any feeder, but not the group request
830	itself. Useful when you queued a lot of events but got a result early.	1480	itself. Useful when you queued a lot of events but got a result early.
831		1481
		1482	The group request will finish normally (you cannot add requests to the
		1483	group).
		1484
832	=item $grp->result (...)	1485	=item $grp->result (...)
833		1486
834	Set the result value(s) that will be passed to the group callback when all	1487	Set the result value(s) that will be passed to the group callback when all
835	subrequests have finished and set thre groups errno to the current value	1488	subrequests have finished and set the groups errno to the current value
836	of errno (just like calling C<errno> without an error number). By default,	1489	of errno (just like calling C<errno> without an error number). By default,
837	no argument will be passed and errno is zero.	1490	no argument will be passed and errno is zero.
838		1491
839	=item $grp->errno ([$errno])	1492	=item $grp->errno ([$errno])
840		1493
…		…
851	=item feed $grp $callback->($grp)	1504	=item feed $grp $callback->($grp)
852		1505
853	Sets a feeder/generator on this group: every group can have an attached	1506	Sets a feeder/generator on this group: every group can have an attached
854	generator that generates requests if idle. The idea behind this is that,	1507	generator that generates requests if idle. The idea behind this is that,
855	although you could just queue as many requests as you want in a group,	1508	although you could just queue as many requests as you want in a group,
856	this might starve other requests for a potentially long time. For	1509	this might starve other requests for a potentially long time. For example,
857	example, C<aio_scandir> might generate hundreds of thousands C<aio_stat>	1510	C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
858	requests, delaying any later requests for a long time.	1511	requests, delaying any later requests for a long time.
859		1512
860	To avoid this, and allow incremental generation of requests, you can	1513	To avoid this, and allow incremental generation of requests, you can
861	instead a group and set a feeder on it that generates those requests. The	1514	instead a group and set a feeder on it that generates those requests. The
862	feed callback will be called whenever there are few enough (see C<limit>,	1515	feed callback will be called whenever there are few enough (see C<limit>,
…		…
867	not impose any limits).	1520	not impose any limits).
868		1521
869	If the feed does not queue more requests when called, it will be	1522	If the feed does not queue more requests when called, it will be
870	automatically removed from the group.	1523	automatically removed from the group.
871		1524
872	If the feed limit is C<0>, it will be set to C<2> automatically.	1525	If the feed limit is C<0> when this method is called, it will be set to
		1526	C<2> automatically.
873		1527
874	Example:	1528	Example:
875		1529
876	# stat all files in @files, but only ever use four aio requests concurrently:	1530	# stat all files in @files, but only ever use four aio requests concurrently:
877		1531
…		…
889	Sets the feeder limit for the group: The feeder will be called whenever	1543	Sets the feeder limit for the group: The feeder will be called whenever
890	the group contains less than this many requests.	1544	the group contains less than this many requests.
891		1545
892	Setting the limit to C<0> will pause the feeding process.	1546	Setting the limit to C<0> will pause the feeding process.
893		1547
		1548	The default value for the limit is C<0>, but note that setting a feeder
		1549	automatically bumps it up to C<2>.
		1550
894	=back	1551	=back
895		1552
896	=head2 SUPPORT FUNCTIONS	1553	=head2 SUPPORT FUNCTIONS
897		1554
898	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION	1555	=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
…		…
900	=over 4	1557	=over 4
901		1558
902	=item $fileno = IO::AIO::poll_fileno	1559	=item $fileno = IO::AIO::poll_fileno
903		1560
904	Return the I<request result pipe file descriptor>. This filehandle must be	1561	Return the I<request result pipe file descriptor>. This filehandle must be
905	polled for reading by some mechanism outside this module (e.g. Event or	1562	polled for reading by some mechanism outside this module (e.g. EV, Glib,
906	select, see below or the SYNOPSIS). If the pipe becomes readable you have	1563	select and so on, see below or the SYNOPSIS). If the pipe becomes readable
907	to call C<poll_cb> to check the results.	1564	you have to call C<poll_cb> to check the results.
908		1565
909	See C<poll_cb> for an example.	1566	See C<poll_cb> for an example.
910		1567
911	=item IO::AIO::poll_cb	1568	=item IO::AIO::poll_cb
912		1569
913	Process some outstanding events on the result pipe. You have to call this	1570	Process some outstanding events on the result pipe. You have to call
914	regularly. Returns the number of events processed. Returns immediately	1571	this regularly. Returns C<0> if all events could be processed (or there
915	when no events are outstanding. The amount of events processed depends on	1572	were no events to process), or C<-1> if it returned earlier for whatever
916	the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.	1573	reason. Returns immediately when no events are outstanding. The amount of
		1574	events processed depends on the settings of C<IO::AIO::max_poll_req> and
		1575	C<IO::AIO::max_poll_time>.
917		1576
918	If not all requests were processed for whatever reason, the filehandle	1577	If not all requests were processed for whatever reason, the filehandle
919	will still be ready when C<poll_cb> returns.	1578	will still be ready when C<poll_cb> returns, so normally you don't have to
		1579	do anything special to have it called later.
		1580
		1581	Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
		1582	ready, it can be beneficial to call this function from loops which submit
		1583	a lot of requests, to make sure the results get processed when they become
		1584	available and not just when the loop is finished and the event loop takes
		1585	over again. This function returns very fast when there are no outstanding
		1586	requests.
920		1587
921	Example: Install an Event watcher that automatically calls	1588	Example: Install an Event watcher that automatically calls
922	IO::AIO::poll_cb with high priority:	1589	IO::AIO::poll_cb with high priority (more examples can be found in the
		1590	SYNOPSIS section, at the top of this document):
923		1591
924	Event->io (fd => IO::AIO::poll_fileno,	1592	Event->io (fd => IO::AIO::poll_fileno,
925	poll => 'r', async => 1,	1593	poll => 'r', async => 1,
926	cb => \&IO::AIO::poll_cb);	1594	cb => \&IO::AIO::poll_cb);
		1595
		1596	=item IO::AIO::poll_wait
		1597
		1598	If there are any outstanding requests and none of them in the result
		1599	phase, wait till the result filehandle becomes ready for reading (simply
		1600	does a C<select> on the filehandle. This is useful if you want to
		1601	synchronously wait for some requests to finish).
		1602
		1603	See C<nreqs> for an example.
		1604
		1605	=item IO::AIO::poll
		1606
		1607	Waits until some requests have been handled.
		1608
		1609	Returns the number of requests processed, but is otherwise strictly
		1610	equivalent to:
		1611
		1612	IO::AIO::poll_wait, IO::AIO::poll_cb
		1613
		1614	=item IO::AIO::flush
		1615
		1616	Wait till all outstanding AIO requests have been handled.
		1617
		1618	Strictly equivalent to:
		1619
		1620	IO::AIO::poll_wait, IO::AIO::poll_cb
		1621	while IO::AIO::nreqs;
927		1622
928	=item IO::AIO::max_poll_reqs $nreqs	1623	=item IO::AIO::max_poll_reqs $nreqs
929		1624
930	=item IO::AIO::max_poll_time $seconds	1625	=item IO::AIO::max_poll_time $seconds
931		1626
…		…
933	that are being processed by C<IO::AIO::poll_cb> in one call, respectively	1628	that are being processed by C<IO::AIO::poll_cb> in one call, respectively
934	the maximum amount of time (default C<0>, meaning infinity) spent in	1629	the maximum amount of time (default C<0>, meaning infinity) spent in
935	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount	1630	C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
936	of time C<poll_cb> is allowed to use).	1631	of time C<poll_cb> is allowed to use).
937		1632
		1633	Setting C<max_poll_time> to a non-zero value creates an overhead of one
		1634	syscall per request processed, which is not normally a problem unless your
		1635	callbacks are really really fast or your OS is really really slow (I am
		1636	not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
		1637
938	Setting these is useful if you want to ensure some level of	1638	Setting these is useful if you want to ensure some level of
939	interactiveness when perl is not fast enough to process all requests in	1639	interactiveness when perl is not fast enough to process all requests in
940	time.	1640	time.
941		1641
942	For interactive programs, values such as C<0.01> to C<0.1> should be fine.	1642	For interactive programs, values such as C<0.01> to C<0.1> should be fine.
943		1643
944	Example: Install an Event watcher that automatically calls	1644	Example: Install an Event watcher that automatically calls
945	IO::AIO::poll_some with low priority, to ensure that other parts of the	1645	IO::AIO::poll_cb with low priority, to ensure that other parts of the
946	program get the CPU sometimes even under high AIO load.	1646	program get the CPU sometimes even under high AIO load.
947		1647
948	# try not to spend much more than 0.1s in poll_cb	1648	# try not to spend much more than 0.1s in poll_cb
949	IO::AIO::max_poll_time 0.1;	1649	IO::AIO::max_poll_time 0.1;
950		1650
951	# use a low priority so other tasks have priority	1651	# use a low priority so other tasks have priority
952	Event->io (fd => IO::AIO::poll_fileno,	1652	Event->io (fd => IO::AIO::poll_fileno,
953	poll => 'r', nice => 1,	1653	poll => 'r', nice => 1,
954	cb => &IO::AIO::poll_cb);	1654	cb => &IO::AIO::poll_cb);
955		1655
956	=item IO::AIO::poll_wait	1656	=back
957
958	Wait till the result filehandle becomes ready for reading (simply does a
959	C<select> on the filehandle. This is useful if you want to synchronously
960	wait for some requests to finish).
961
962	See C<nreqs> for an example.
963
964	=item IO::AIO::poll
965
966	Waits until some requests have been handled.
967
968	Strictly equivalent to:
969
970	IO::AIO::poll_wait, IO::AIO::poll_cb
971	if IO::AIO::nreqs;
972
973	=item IO::AIO::flush
974
975	Wait till all outstanding AIO requests have been handled.
976
977	Strictly equivalent to:
978
979	IO::AIO::poll_wait, IO::AIO::poll_cb
980	while IO::AIO::nreqs;
981		1657
982	=head3 CONTROLLING THE NUMBER OF THREADS	1658	=head3 CONTROLLING THE NUMBER OF THREADS
		1659
		1660	=over
983		1661
984	=item IO::AIO::min_parallel $nthreads	1662	=item IO::AIO::min_parallel $nthreads
985		1663
986	Set the minimum number of AIO threads to C<$nthreads>. The current	1664	Set the minimum number of AIO threads to C<$nthreads>. The current
987	default is C<8>, which means eight asynchronous operations can execute	1665	default is C<8>, which means eight asynchronous operations can execute
…		…
1015		1693
1016	Under normal circumstances you don't need to call this function.	1694	Under normal circumstances you don't need to call this function.
1017		1695
1018	=item IO::AIO::max_idle $nthreads	1696	=item IO::AIO::max_idle $nthreads
1019		1697
1020	Limit the number of threads (default: 4) that are allowed to idle (i.e.,	1698	Limit the number of threads (default: 4) that are allowed to idle
1021	threads that did not get a request to process within 10 seconds). That	1699	(i.e., threads that did not get a request to process within the idle
1022	means if a thread becomes idle while C<$nthreads> other threads are also	1700	timeout (default: 10 seconds). That means if a thread becomes idle while
1023	idle, it will free its resources and exit.	1701	C<$nthreads> other threads are also idle, it will free its resources and
		1702	exit.
1024		1703
1025	This is useful when you allow a large number of threads (e.g. 100 or 1000)	1704	This is useful when you allow a large number of threads (e.g. 100 or 1000)
1026	to allow for extremely high load situations, but want to free resources	1705	to allow for extremely high load situations, but want to free resources
1027	under normal circumstances (1000 threads can easily consume 30MB of RAM).	1706	under normal circumstances (1000 threads can easily consume 30MB of RAM).
1028		1707
1029	The default is probably ok in most situations, especially if thread	1708	The default is probably ok in most situations, especially if thread
1030	creation is fast. If thread creation is very slow on your system you might	1709	creation is fast. If thread creation is very slow on your system you might
1031	want to use larger values.	1710	want to use larger values.
1032		1711
		1712	=item IO::AIO::idle_timeout $seconds
		1713
		1714	Sets the minimum idle timeout (default 10) after which worker threads are
		1715	allowed to exit. SEe C<IO::AIO::max_idle>.
		1716
1033	=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs	1717	=item IO::AIO::max_outstanding $maxreqs
		1718
		1719	Sets the maximum number of outstanding requests to C<$nreqs>. If
		1720	you do queue up more than this number of requests, the next call to
		1721	C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
		1722	C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
		1723	longer exceeded.
		1724
		1725	In other words, this setting does not enforce a queue limit, but can be
		1726	used to make poll functions block if the limit is exceeded.
1034		1727
1035	This is a very bad function to use in interactive programs because it	1728	This is a very bad function to use in interactive programs because it
1036	blocks, and a bad way to reduce concurrency because it is inexact: Better	1729	blocks, and a bad way to reduce concurrency because it is inexact: Better
1037	use an C<aio_group> together with a feed callback.	1730	use an C<aio_group> together with a feed callback.
1038		1731
1039	Sets the maximum number of outstanding requests to C<$nreqs>. If you	1732	It's main use is in scripts without an event loop - when you want to stat
1040	to queue up more than this number of requests, the next call to the	1733	a lot of files, you can write somehting like this:
1041	C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
1042	function will block until the limit is no longer exceeded.
1043		1734
1044	The default value is very large, so there is no practical limit on the	1735	IO::AIO::max_outstanding 32;
1045	number of outstanding requests.
1046		1736
1047	You can still queue as many requests as you want. Therefore,	1737	for my $path (...) {
1048	C<max_oustsanding> is mainly useful in simple scripts (with low values) or	1738	aio_stat $path , ...;
1049	as a stop gap to shield against fatal memory overflow (with large values).	1739	IO::AIO::poll_cb;
		1740	}
		1741
		1742	IO::AIO::flush;
		1743
		1744	The call to C<poll_cb> inside the loop will normally return instantly, but
		1745	as soon as more thna C<32> reqeusts are in-flight, it will block until
		1746	some requests have been handled. This keeps the loop from pushing a large
		1747	number of C<aio_stat> requests onto the queue.
		1748
		1749	The default value for C<max_outstanding> is very large, so there is no
		1750	practical limit on the number of outstanding requests.
		1751
		1752	=back
1050		1753
1051	=head3 STATISTICAL INFORMATION	1754	=head3 STATISTICAL INFORMATION
		1755
		1756	=over
1052		1757
1053	=item IO::AIO::nreqs	1758	=item IO::AIO::nreqs
1054		1759
1055	Returns the number of requests currently in the ready, execute or pending	1760	Returns the number of requests currently in the ready, execute or pending
1056	states (i.e. for which their callback has not been invoked yet).	1761	states (i.e. for which their callback has not been invoked yet).
…		…
1070	Returns the number of requests currently in the pending state (executed,	1775	Returns the number of requests currently in the pending state (executed,
1071	but not yet processed by poll_cb).	1776	but not yet processed by poll_cb).
1072		1777
1073	=back	1778	=back
1074		1779
		1780	=head3 MISCELLANEOUS FUNCTIONS
		1781
		1782	IO::AIO implements some functions that might be useful, but are not
		1783	asynchronous.
		1784
		1785	=over 4
		1786
		1787	=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
		1788
		1789	Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
		1790	but is blocking (this makes most sense if you know the input data is
		1791	likely cached already and the output filehandle is set to non-blocking
		1792	operations).
		1793
		1794	Returns the number of bytes copied, or C<-1> on error.
		1795
		1796	=item IO::AIO::fadvise $fh, $offset, $len, $advice
		1797
		1798	Simply calls the C<posix_fadvise> function (see its
		1799	manpage for details). The following advice constants are
		1800	available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
		1801	C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
		1802	C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
		1803
		1804	On systems that do not implement C<posix_fadvise>, this function returns
		1805	ENOSYS, otherwise the return value of C<posix_fadvise>.
		1806
		1807	=item IO::AIO::madvise $scalar, $offset, $len, $advice
		1808
		1809	Simply calls the C<posix_madvise> function (see its
		1810	manpage for details). The following advice constants are
		1811	available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
		1812	C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
		1813
		1814	On systems that do not implement C<posix_madvise>, this function returns
		1815	ENOSYS, otherwise the return value of C<posix_madvise>.
		1816
		1817	=item IO::AIO::mprotect $scalar, $offset, $len, $protect
		1818
		1819	Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
		1820	$scalar (see its manpage for details). The following protect
		1821	constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
		1822	C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
		1823
		1824	On systems that do not implement C<mprotect>, this function returns
		1825	ENOSYS, otherwise the return value of C<mprotect>.
		1826
		1827	=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
		1828
		1829	Memory-maps a file (or anonymous memory range) and attaches it to the
		1830	given C<$scalar>, which will act like a string scalar.
		1831
		1832	The only operations allowed on the scalar are C<substr>/C<vec> that don't
		1833	change the string length, and most read-only operations such as copying it
		1834	or searching it with regexes and so on.
		1835
		1836	Anything else is unsafe and will, at best, result in memory leaks.
		1837
		1838	The memory map associated with the C<$scalar> is automatically removed
		1839	when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
		1840	C<IO::AIO::munmap> functions are called.
		1841
		1842	This calls the C<mmap>(2) function internally. See your system's manual
		1843	page for details on the C<$length>, C<$prot> and C<$flags> parameters.
		1844
		1845	The C<$length> must be larger than zero and smaller than the actual
		1846	filesize.
		1847
		1848	C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
		1849	C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
		1850
		1851	C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
		1852	C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
		1853	not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
		1854	(which is set to C<MAP_ANON> if your system only provides this
		1855	constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
		1856	C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
		1857	C<IO::AIO::MAP_NONBLOCK>
		1858
		1859	If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
		1860
		1861	C<$offset> is the offset from the start of the file - it generally must be
		1862	a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
		1863
		1864	Example:
		1865
		1866	use Digest::MD5;
		1867	use IO::AIO;
		1868
		1869	open my $fh, "<verybigfile"
		1870	or die "$!";
		1871
		1872	IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
		1873	or die "verybigfile: $!";
		1874
		1875	my $fast_md5 = md5 $data;
		1876
		1877	=item IO::AIO::munmap $scalar
		1878
		1879	Removes a previous mmap and undefines the C<$scalar>.
		1880
		1881	=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
		1882
		1883	Calls the C<munlock> function, undoing the effects of a previous
		1884	C<aio_mlock> call (see its description for details).
		1885
		1886	=item IO::AIO::munlockall
		1887
		1888	Calls the C<munlockall> function.
		1889
		1890	On systems that do not implement C<munlockall>, this function returns
		1891	ENOSYS, otherwise the return value of C<munlockall>.
		1892
		1893	=back
		1894
1075	=cut	1895	=cut
1076		1896
1077	# support function to convert a fd into a perl filehandle
1078	sub _fd2fh {
1079	return undef if $_[0] < 0;
1080
1081	# try to generate nice filehandles
1082	my $sym = "IO::AIO::fd#$_[0]";
1083	local *$sym;
1084
1085	open *$sym, "+<&=$_[0]" # usually works under any unix
1086	or open *$sym, "<&=$_[0]" # cygwin needs this
1087	or open *$sym, ">&=$_[0]" # or this
1088	or return undef;
1089
1090	*$sym
1091	}
1092
1093	min_parallel 8;	1897	min_parallel 8;
1094		1898
1095	END {	1899	END { flush }
1096	min_parallel 1;
1097	flush;
1098	};
1099		1900
1100	1;	1901	1;
1101		1902
		1903	=head1 EVENT LOOP INTEGRATION
		1904
		1905	It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
		1906	automatically into many event loops:
		1907
		1908	# AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
		1909	use AnyEvent::AIO;
		1910
		1911	You can also integrate IO::AIO manually into many event loops, here are
		1912	some examples of how to do this:
		1913
		1914	# EV integration
		1915	my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
		1916
		1917	# Event integration
		1918	Event->io (fd => IO::AIO::poll_fileno,
		1919	poll => 'r',
		1920	cb => \&IO::AIO::poll_cb);
		1921
		1922	# Glib/Gtk2 integration
		1923	add_watch Glib::IO IO::AIO::poll_fileno,
		1924	in => sub { IO::AIO::poll_cb; 1 };
		1925
		1926	# Tk integration
		1927	Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
		1928	readable => \&IO::AIO::poll_cb);
		1929
		1930	# Danga::Socket integration
		1931	Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
		1932	\&IO::AIO::poll_cb);
		1933
1102	=head2 FORK BEHAVIOUR	1934	=head2 FORK BEHAVIOUR
1103		1935
1104	This module should do "the right thing" when the process using it forks:	1936	Usage of pthreads in a program changes the semantics of fork
		1937	considerably. Specifically, only async-safe functions can be called after
		1938	fork. Perl doesn't know about this, so in general, you cannot call fork
		1939	with defined behaviour in perl if pthreads are involved. IO::AIO uses
		1940	pthreads, so this applies, but many other extensions and (for inexplicable
		1941	reasons) perl itself often is linked against pthreads, so this limitation
		1942	applies to quite a lot of perls.
1105		1943
1106	Before the fork, IO::AIO enters a quiescent state where no requests	1944	This module no longer tries to fight your OS, or POSIX. That means IO::AIO
1107	can be added in other threads and no results will be processed. After	1945	only works in the process that loaded it. Forking is fully supported, but
1108	the fork the parent simply leaves the quiescent state and continues	1946	using IO::AIO in the child is not.
1109	request/result processing, while the child frees the request/result queue
1110	(so that the requests started before the fork will only be handled in the
1111	parent). Threads will be started on demand until the limit set in the
1112	parent process has been reached again.
1113		1947
1114	In short: the parent will, after a short pause, continue as if fork had	1948	You might get around by not I<using> IO::AIO before (or after)
1115	not been called, while the child will act as if IO::AIO has not been used	1949	forking. You could also try to call the L<IO::AIO::reinit> function in the
1116	yet.	1950	child:
		1951
		1952	=over 4
		1953
		1954	=item IO::AIO::reinit
		1955
		1956	Abandons all current requests and I/O threads and simply reinitialises all
		1957	data structures. This is not an operation supported by any standards, but
		1958	happens to work on GNU/Linux and some newer BSD systems.
		1959
		1960	The only reasonable use for this function is to call it after forking, if
		1961	C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
		1962	the process will result in undefined behaviour. Calling it at any time
		1963	will also result in any undefined (by POSIX) behaviour.
		1964
		1965	=back
1117		1966
1118	=head2 MEMORY USAGE	1967	=head2 MEMORY USAGE
1119		1968
1120	Per-request usage:	1969	Per-request usage:
1121		1970
…		…
1123	bytes of memory. In addition, stat requests need a stat buffer (possibly	1972	bytes of memory. In addition, stat requests need a stat buffer (possibly
1124	a few hundred bytes), readdir requires a result buffer and so on. Perl	1973	a few hundred bytes), readdir requires a result buffer and so on. Perl
1125	scalars and other data passed into aio requests will also be locked and	1974	scalars and other data passed into aio requests will also be locked and
1126	will consume memory till the request has entered the done state.	1975	will consume memory till the request has entered the done state.
1127		1976
1128	This is now awfully much, so queuing lots of requests is not usually a	1977	This is not awfully much, so queuing lots of requests is not usually a
1129	problem.	1978	problem.
1130		1979
1131	Per-thread usage:	1980	Per-thread usage:
1132		1981
1133	In the execution phase, some aio requests require more memory for	1982	In the execution phase, some aio requests require more memory for
…		…
1138		1987
1139	Known bugs will be fixed in the next release.	1988	Known bugs will be fixed in the next release.
1140		1989
1141	=head1 SEE ALSO	1990	=head1 SEE ALSO
1142		1991
1143	L<Coro::AIO>.	1992	L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
		1993	more natural syntax.
1144		1994
1145	=head1 AUTHOR	1995	=head1 AUTHOR
1146		1996
1147	Marc Lehmann <schmorp@schmorp.de>	1997	Marc Lehmann <schmorp@schmorp.de>
1148	http://home.schmorp.de/	1998	http://home.schmorp.de/

Diff Legend

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