ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/IO-AIO/AIO.pm
(Generate patch)

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.161 by root, Tue Jul 14 00:32:27 2009 UTC vs.
Revision 1.281 by root, Tue Feb 20 04:32:59 2018 UTC

1=head1 NAME 1=head1 NAME
2 2
3IO::AIO - Asynchronous Input/Output 3IO::AIO - Asynchronous/Advanced Input/Output
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 = shift 10 my $fh = shift
11 or die "/etc/passwd: $!"; 11 or die "/etc/passwd: $!";
12 ... 12 ...
13 }; 13 };
14 14
25 my $req = aio_unlink "/tmp/file", sub { }; 25 my $req = aio_unlink "/tmp/file", sub { };
26 $req->cancel; # cancel request if still in queue 26 $req->cancel; # cancel request if still in queue
27 27
28 my $grp = aio_group sub { print "all stats done\n" }; 28 my $grp = aio_group sub { print "all stats done\n" };
29 add $grp aio_stat "..." for ...; 29 add $grp aio_stat "..." for ...;
30
31 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
32 use AnyEvent::AIO;
33
34 # EV integration
35 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
36
37 # Event integration
38 Event->io (fd => IO::AIO::poll_fileno,
39 poll => 'r',
40 cb => \&IO::AIO::poll_cb);
41
42 # Glib/Gtk2 integration
43 add_watch Glib::IO IO::AIO::poll_fileno,
44 in => sub { IO::AIO::poll_cb; 1 };
45
46 # Tk integration
47 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
48 readable => \&IO::AIO::poll_cb);
49
50 # Danga::Socket integration
51 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
52 \&IO::AIO::poll_cb);
53 30
54=head1 DESCRIPTION 31=head1 DESCRIPTION
55 32
56This module implements asynchronous I/O using whatever means your 33This module implements asynchronous I/O using whatever means your
57operating system supports. It is implemented as an interface to C<libeio> 34operating system supports. It is implemented as an interface to C<libeio>
81not well-supported or restricted (GNU/Linux doesn't allow them on normal 58not well-supported or restricted (GNU/Linux doesn't allow them on normal
82files currently, for example), and they would only support aio_read and 59files currently, for example), and they would only support aio_read and
83aio_write, so the remaining functionality would have to be implemented 60aio_write, so the remaining functionality would have to be implemented
84using threads anyway. 61using threads anyway.
85 62
63In addition to asynchronous I/O, this module also exports some rather
64arcane interfaces, such as C<madvise> or linux's C<splice> system call,
65which is why the C<A> in C<AIO> can also mean I<advanced>.
66
86Although the module will work in the presence of other (Perl-) threads, 67Although the module will work in the presence of other (Perl-) threads,
87it is currently not reentrant in any way, so use appropriate locking 68it is currently not reentrant in any way, so use appropriate locking
88yourself, always call C<poll_cb> from within the same thread, or never 69yourself, always call C<poll_cb> from within the same thread, or never
89call C<poll_cb> (or other C<aio_> functions) recursively. 70call C<poll_cb> (or other C<aio_> functions) recursively.
90 71
91=head2 EXAMPLE 72=head2 EXAMPLE
92 73
93This is a simple example that uses the EV module and loads 74This is a simple example that uses the EV module and loads
94F</etc/passwd> asynchronously: 75F</etc/passwd> asynchronously:
95 76
96 use Fcntl;
97 use EV; 77 use EV;
98 use IO::AIO; 78 use IO::AIO;
99 79
100 # register the IO::AIO callback with EV 80 # register the IO::AIO callback with EV
101 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb; 81 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
102 82
103 # queue the request to open /etc/passwd 83 # queue the request to open /etc/passwd
104 aio_open "/etc/passwd", O_RDONLY, 0, sub { 84 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
105 my $fh = shift 85 my $fh = shift
106 or die "error while opening: $!"; 86 or die "error while opening: $!";
107 87
108 # stat'ing filehandles is generally non-blocking 88 # stat'ing filehandles is generally non-blocking
109 my $size = -s $fh; 89 my $size = -s $fh;
118 98
119 # file contents now in $contents 99 # file contents now in $contents
120 print $contents; 100 print $contents;
121 101
122 # exit event loop and program 102 # exit event loop and program
123 EV::unloop; 103 EV::break;
124 }; 104 };
125 }; 105 };
126 106
127 # possibly queue up other requests, or open GUI windows, 107 # possibly queue up other requests, or open GUI windows,
128 # check for sockets etc. etc. 108 # check for sockets etc. etc.
129 109
130 # process events as long as there are some: 110 # process events as long as there are some:
131 EV::loop; 111 EV::run;
132 112
133=head1 REQUEST ANATOMY AND LIFETIME 113=head1 REQUEST ANATOMY AND LIFETIME
134 114
135Every C<aio_*> function creates a request. which is a C data structure not 115Every C<aio_*> function creates a request. which is a C data structure not
136directly visible to Perl. 116directly visible to Perl.
191use common::sense; 171use common::sense;
192 172
193use base 'Exporter'; 173use base 'Exporter';
194 174
195BEGIN { 175BEGIN {
196 our $VERSION = '3.261'; 176 our $VERSION = 4.4;
197 177
198 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close 178 our @AIO_REQ = qw(aio_sendfile aio_seek aio_read aio_write aio_open aio_close
199 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx 179 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
200 aio_scandir aio_symlink aio_readlink aio_sync aio_fsync 180 aio_scandir aio_symlink aio_readlink aio_realpath aio_fcntl aio_ioctl
201 aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead 181 aio_sync aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range
182 aio_pathsync aio_readahead aio_fiemap aio_allocate
202 aio_rename aio_link aio_move aio_copy aio_group 183 aio_rename aio_rename2 aio_link aio_move aio_copy aio_group
203 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown 184 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
204 aio_chmod aio_utime aio_truncate); 185 aio_chmod aio_utime aio_truncate
186 aio_msync aio_mtouch aio_mlock aio_mlockall
187 aio_statvfs
188 aio_slurp
189 aio_wd);
205 190
206 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); 191 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
207 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 192 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
208 min_parallel max_parallel max_idle 193 min_parallel max_parallel max_idle idle_timeout
209 nreqs nready npending nthreads 194 nreqs nready npending nthreads
210 max_poll_time max_poll_reqs 195 max_poll_time max_poll_reqs
211 sendfile fadvise); 196 sendfile fadvise madvise
197 mmap munmap munlock munlockall);
212 198
213 push @AIO_REQ, qw(aio_busy); # not exported 199 push @AIO_REQ, qw(aio_busy); # not exported
214 200
215 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 201 @IO::AIO::GRP::ISA = 'IO::AIO::REQ';
216 202
218 XSLoader::load ("IO::AIO", $VERSION); 204 XSLoader::load ("IO::AIO", $VERSION);
219} 205}
220 206
221=head1 FUNCTIONS 207=head1 FUNCTIONS
222 208
223=head2 AIO REQUEST FUNCTIONS 209=head2 QUICK OVERVIEW
210
211This section simply lists the prototypes most of the functions for
212quick reference. See the following sections for function-by-function
213documentation.
214
215 aio_wd $pathname, $callback->($wd)
216 aio_open $pathname, $flags, $mode, $callback->($fh)
217 aio_close $fh, $callback->($status)
218 aio_seek $fh,$offset,$whence, $callback->($offs)
219 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
220 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
221 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
222 aio_readahead $fh,$offset,$length, $callback->($retval)
223 aio_stat $fh_or_path, $callback->($status)
224 aio_lstat $fh, $callback->($status)
225 aio_statvfs $fh_or_path, $callback->($statvfs)
226 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
227 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
228 aio_chmod $fh_or_path, $mode, $callback->($status)
229 aio_truncate $fh_or_path, $offset, $callback->($status)
230 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
231 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
232 aio_unlink $pathname, $callback->($status)
233 aio_mknod $pathname, $mode, $dev, $callback->($status)
234 aio_link $srcpath, $dstpath, $callback->($status)
235 aio_symlink $srcpath, $dstpath, $callback->($status)
236 aio_readlink $pathname, $callback->($link)
237 aio_realpath $pathname, $callback->($path)
238 aio_rename $srcpath, $dstpath, $callback->($status)
239 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
240 aio_mkdir $pathname, $mode, $callback->($status)
241 aio_rmdir $pathname, $callback->($status)
242 aio_readdir $pathname, $callback->($entries)
243 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
244 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
245 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
246 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
247 aio_load $pathname, $data, $callback->($status)
248 aio_copy $srcpath, $dstpath, $callback->($status)
249 aio_move $srcpath, $dstpath, $callback->($status)
250 aio_rmtree $pathname, $callback->($status)
251 aio_fcntl $fh, $cmd, $arg, $callback->($status)
252 aio_ioctl $fh, $request, $buf, $callback->($status)
253 aio_sync $callback->($status)
254 aio_syncfs $fh, $callback->($status)
255 aio_fsync $fh, $callback->($status)
256 aio_fdatasync $fh, $callback->($status)
257 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
258 aio_pathsync $pathname, $callback->($status)
259 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
260 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
261 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
262 aio_mlockall $flags, $callback->($status)
263 aio_group $callback->(...)
264 aio_nop $callback->()
265
266 $prev_pri = aioreq_pri [$pri]
267 aioreq_nice $pri_adjust
268
269 IO::AIO::poll_wait
270 IO::AIO::poll_cb
271 IO::AIO::poll
272 IO::AIO::flush
273 IO::AIO::max_poll_reqs $nreqs
274 IO::AIO::max_poll_time $seconds
275 IO::AIO::min_parallel $nthreads
276 IO::AIO::max_parallel $nthreads
277 IO::AIO::max_idle $nthreads
278 IO::AIO::idle_timeout $seconds
279 IO::AIO::max_outstanding $maxreqs
280 IO::AIO::nreqs
281 IO::AIO::nready
282 IO::AIO::npending
283 $nfd = IO::AIO::get_fdlimit [EXPERIMENTAL]
284 IO::AIO::min_fdlimit $nfd [EXPERIMENTAL]
285
286 IO::AIO::sendfile $ofh, $ifh, $offset, $count
287 IO::AIO::fadvise $fh, $offset, $len, $advice
288 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
289 IO::AIO::munmap $scalar
290 IO::AIO::madvise $scalar, $offset, $length, $advice
291 IO::AIO::mprotect $scalar, $offset, $length, $protect
292 IO::AIO::munlock $scalar, $offset = 0, $length = undef
293 IO::AIO::munlockall
294
295=head2 API NOTES
224 296
225All the C<aio_*> calls are more or less thin wrappers around the syscall 297All the C<aio_*> calls are more or less thin wrappers around the syscall
226with the same name (sans C<aio_>). The arguments are similar or identical, 298with the same name (sans C<aio_>). The arguments are similar or identical,
227and they all accept an additional (and optional) C<$callback> argument 299and they all accept an additional (and optional) C<$callback> argument
228which must be a code reference. This code reference will get called with 300which must be a code reference. This code reference will be called after
229the syscall return code (e.g. most syscalls return C<-1> on error, unlike 301the syscall has been executed in an asynchronous fashion. The results
230perl, which usually delivers "false") as its sole argument after the given 302of the request will be passed as arguments to the callback (and, if an
231syscall has been executed asynchronously. 303error occured, in C<$!>) - for most requests the syscall return code (e.g.
304most syscalls return C<-1> on error, unlike perl, which usually delivers
305"false").
306
307Some requests (such as C<aio_readdir>) pass the actual results and
308communicate failures by passing C<undef>.
232 309
233All functions expecting a filehandle keep a copy of the filehandle 310All functions expecting a filehandle keep a copy of the filehandle
234internally until the request has finished. 311internally until the request has finished.
235 312
236All functions return request objects of type L<IO::AIO::REQ> that allow 313All functions return request objects of type L<IO::AIO::REQ> that allow
237further manipulation of those requests while they are in-flight. 314further manipulation of those requests while they are in-flight.
238 315
239The pathnames you pass to these routines I<must> be absolute and 316The pathnames you pass to these routines I<should> be absolute. The
240encoded as octets. The reason for the former is that at the time the 317reason for this is that at the time the request is being executed, the
241request is being executed, the current working directory could have 318current working directory could have changed. Alternatively, you can
242changed. Alternatively, you can make sure that you never change the 319make sure that you never change the current working directory anywhere
243current working directory anywhere in the program and then use relative 320in the program and then use relative paths. You can also take advantage
244paths. 321of IO::AIOs working directory abstraction, that lets you specify paths
322relative to some previously-opened "working directory object" - see the
323description of the C<IO::AIO::WD> class later in this document.
245 324
246To encode pathnames as octets, either make sure you either: a) always pass 325To encode pathnames as octets, either make sure you either: a) always pass
247in filenames you got from outside (command line, readdir etc.) without 326in filenames you got from outside (command line, readdir etc.) without
248tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode 327tinkering, b) are in your native filesystem encoding, c) use the Encode
249your pathnames to the locale (or other) encoding in effect in the user 328module and encode your pathnames to the locale (or other) encoding in
250environment, d) use Glib::filename_from_unicode on unicode filenames or e) 329effect in the user environment, d) use Glib::filename_from_unicode on
251use something else to ensure your scalar has the correct contents. 330unicode filenames or e) use something else to ensure your scalar has the
331correct contents.
252 332
253This works, btw. independent of the internal UTF-8 bit, which IO::AIO 333This works, btw. independent of the internal UTF-8 bit, which IO::AIO
254handles correctly whether it is set or not. 334handles correctly whether it is set or not.
335
336=head2 AIO REQUEST FUNCTIONS
255 337
256=over 4 338=over 4
257 339
258=item $prev_pri = aioreq_pri [$pri] 340=item $prev_pri = aioreq_pri [$pri]
259 341
289 371
290 372
291=item aio_open $pathname, $flags, $mode, $callback->($fh) 373=item aio_open $pathname, $flags, $mode, $callback->($fh)
292 374
293Asynchronously open or create a file and call the callback with a newly 375Asynchronously open or create a file and call the callback with a newly
294created filehandle for the file. 376created filehandle for the file (or C<undef> in case of an error).
295 377
296The pathname passed to C<aio_open> must be absolute. See API NOTES, above, 378The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
297for an explanation. 379for an explanation.
298 380
299The C<$flags> argument is a bitmask. See the C<Fcntl> module for a 381The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
306by the umask in effect then the request is being executed, so better never 388by the umask in effect then the request is being executed, so better never
307change the umask. 389change the umask.
308 390
309Example: 391Example:
310 392
311 aio_open "/etc/passwd", O_RDONLY, 0, sub { 393 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
312 if ($_[0]) { 394 if ($_[0]) {
313 print "open successful, fh is $_[0]\n"; 395 print "open successful, fh is $_[0]\n";
314 ... 396 ...
315 } else { 397 } else {
316 die "open failed: $!\n"; 398 die "open failed: $!\n";
317 } 399 }
318 }; 400 };
319 401
402In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
403C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
404following POSIX and non-POSIX constants are available (missing ones on
405your system are, as usual, C<0>):
406
407C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
408C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
409C<O_RSYNC>, C<O_SYNC>, C<O_PATH>, C<O_TMPFILE>, and C<O_TTY_INIT>.
410
320 411
321=item aio_close $fh, $callback->($status) 412=item aio_close $fh, $callback->($status)
322 413
323Asynchronously close a file and call the callback with the result 414Asynchronously close a file and call the callback with the result
324code. 415code.
333Or in other words: the file descriptor will be closed, but it will not be 424Or in other words: the file descriptor will be closed, but it will not be
334free for reuse until the perl filehandle is closed. 425free for reuse until the perl filehandle is closed.
335 426
336=cut 427=cut
337 428
429=item aio_seek $fh, $offset, $whence, $callback->($offs)
430
431Seeks the filehandle to the new C<$offset>, similarly to perl's
432C<sysseek>. The C<$whence> can use the traditional values (C<0> for
433C<IO::AIO::SEEK_SET>, C<1> for C<IO::AIO::SEEK_CUR> or C<2> for
434C<IO::AIO::SEEK_END>).
435
436The resulting absolute offset will be passed to the callback, or C<-1> in
437case of an error.
438
439In theory, the C<$whence> constants could be different than the
440corresponding values from L<Fcntl>, but perl guarantees they are the same,
441so don't panic.
442
443As a GNU/Linux (and maybe Solaris) extension, also the constants
444C<IO::AIO::SEEK_DATA> and C<IO::AIO::SEEK_HOLE> are available, if they
445could be found. No guarantees about suitability for use in C<aio_seek> or
446Perl's C<sysseek> can be made though, although I would naively assume they
447"just work".
448
338=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 449=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
339 450
340=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 451=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
341 452
342Reads or writes C<$length> bytes from or to the specified C<$fh> and 453Reads or writes C<$length> bytes from or to the specified C<$fh> and
343C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> 454C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> and
344and calls the callback without the actual number of bytes read (or -1 on 455calls the callback with the actual number of bytes transferred (or -1 on
345error, just like the syscall). 456error, just like the syscall).
346 457
347C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to 458C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
348offset plus the actual number of bytes read. 459offset plus the actual number of bytes read.
349 460
374 485
375Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 486Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
376reading at byte offset C<$in_offset>, and starts writing at the current 487reading at byte offset C<$in_offset>, and starts writing at the current
377file offset of C<$out_fh>. Because of that, it is not safe to issue more 488file offset of C<$out_fh>. Because of that, it is not safe to issue more
378than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each 489than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
379other. 490other. The same C<$in_fh> works fine though, as this function does not
491move or use the file offset of C<$in_fh>.
380 492
493Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
494are written, and there is no way to find out how many more bytes have been
495read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
496number of bytes written to C<$out_fh>. Only if the result value equals
497C<$length> one can assume that C<$length> bytes have been read.
498
499Unlike with other C<aio_> functions, it makes a lot of sense to use
500C<aio_sendfile> on non-blocking sockets, as long as one end (typically
501the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
502the socket I/O will be non-blocking. Note, however, that you can run
503into a trap where C<aio_sendfile> reads some data with readahead, then
504fails to write all data, and when the socket is ready the next time, the
505data in the cache is already lost, forcing C<aio_sendfile> to again hit
506the disk. Explicit C<aio_read> + C<aio_write> let's you better control
507resource usage.
508
381This call tries to make use of a native C<sendfile> syscall to provide 509This call tries to make use of a native C<sendfile>-like syscall to
382zero-copy operation. For this to work, C<$out_fh> should refer to a 510provide zero-copy operation. For this to work, C<$out_fh> should refer to
383socket, and C<$in_fh> should refer to mmap'able file. 511a socket, and C<$in_fh> should refer to an mmap'able file.
384 512
385If the native sendfile call fails or is not implemented, it will be 513If a native sendfile cannot be found or it fails with C<ENOSYS>,
386emulated, so you can call C<aio_sendfile> on any type of filehandle 514C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
515C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
387regardless of the limitations of the operating system. 516type of filehandle regardless of the limitations of the operating system.
388 517
389Please note, however, that C<aio_sendfile> can read more bytes from 518As native sendfile syscalls (as practically any non-POSIX interface hacked
390C<$in_fh> than are written, and there is no way to find out how many 519together in a hurry to improve benchmark numbers) tend to be rather buggy
391bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only 520on many systems, this implementation tries to work around some known bugs
392provides the number of bytes written to C<$out_fh>. Only if the result 521in Linux and FreeBSD kernels (probably others, too), but that might fail,
393value equals C<$length> one can assume that C<$length> bytes have been 522so you really really should check the return value of C<aio_sendfile> -
394read. 523fewer bytes than expected might have been transferred.
395 524
396 525
397=item aio_readahead $fh,$offset,$length, $callback->($retval) 526=item aio_readahead $fh,$offset,$length, $callback->($retval)
398 527
399C<aio_readahead> populates the page cache with data from a file so that 528C<aio_readahead> populates the page cache with data from a file so that
403whole pages, so that offset is effectively rounded down to a page boundary 532whole pages, so that offset is effectively rounded down to a page boundary
404and bytes are read up to the next page boundary greater than or equal to 533and bytes are read up to the next page boundary greater than or equal to
405(off-set+length). C<aio_readahead> does not read beyond the end of the 534(off-set+length). C<aio_readahead> does not read beyond the end of the
406file. The current file offset of the file is left unchanged. 535file. The current file offset of the file is left unchanged.
407 536
408If that syscall doesn't exist (likely if your OS isn't Linux) it will be 537If that syscall doesn't exist (likely if your kernel isn't Linux) it will
409emulated by simply reading the data, which would have a similar effect. 538be emulated by simply reading the data, which would have a similar effect.
410 539
411 540
412=item aio_stat $fh_or_path, $callback->($status) 541=item aio_stat $fh_or_path, $callback->($status)
413 542
414=item aio_lstat $fh, $callback->($status) 543=item aio_lstat $fh, $callback->($status)
421for an explanation. 550for an explanation.
422 551
423Currently, the stats are always 64-bit-stats, i.e. instead of returning an 552Currently, the stats are always 64-bit-stats, i.e. instead of returning an
424error when stat'ing a large file, the results will be silently truncated 553error when stat'ing a large file, the results will be silently truncated
425unless perl itself is compiled with large file support. 554unless perl itself is compiled with large file support.
555
556To help interpret the mode and dev/rdev stat values, IO::AIO offers the
557following constants and functions (if not implemented, the constants will
558be C<0> and the functions will either C<croak> or fall back on traditional
559behaviour).
560
561C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
562C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
563C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
426 564
427Example: Print the length of F</etc/passwd>: 565Example: Print the length of F</etc/passwd>:
428 566
429 aio_stat "/etc/passwd", sub { 567 aio_stat "/etc/passwd", sub {
430 $_[0] and die "stat failed: $!"; 568 $_[0] and die "stat failed: $!";
431 print "size is ", -s _, "\n"; 569 print "size is ", -s _, "\n";
432 }; 570 };
433 571
434 572
573=item aio_statvfs $fh_or_path, $callback->($statvfs)
574
575Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
576whether a file handle or path was passed.
577
578On success, the callback is passed a hash reference with the following
579members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
580C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
581is passed.
582
583The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
584C<ST_NOSUID>.
585
586The following non-POSIX IO::AIO::ST_* flag masks are defined to
587their correct value when available, or to C<0> on systems that do
588not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
589C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
590C<ST_NODIRATIME> and C<ST_RELATIME>.
591
592Example: stat C</wd> and dump out the data if successful.
593
594 aio_statvfs "/wd", sub {
595 my $f = $_[0]
596 or die "statvfs: $!";
597
598 use Data::Dumper;
599 say Dumper $f;
600 };
601
602 # result:
603 {
604 bsize => 1024,
605 bfree => 4333064312,
606 blocks => 10253828096,
607 files => 2050765568,
608 flag => 4096,
609 favail => 2042092649,
610 bavail => 4333064312,
611 ffree => 2042092649,
612 namemax => 255,
613 frsize => 1024,
614 fsid => 1810
615 }
616
435=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 617=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
436 618
437Works like perl's C<utime> function (including the special case of $atime 619Works like perl's C<utime> function (including the special case of $atime
438and $mtime being undef). Fractional times are supported if the underlying 620and $mtime being undef). Fractional times are supported if the underlying
439syscalls support them. 621syscalls support them.
466=item aio_truncate $fh_or_path, $offset, $callback->($status) 648=item aio_truncate $fh_or_path, $offset, $callback->($status)
467 649
468Works like truncate(2) or ftruncate(2). 650Works like truncate(2) or ftruncate(2).
469 651
470 652
653=item aio_allocate $fh, $mode, $offset, $len, $callback->($status)
654
655Allocates or frees disk space according to the C<$mode> argument. See the
656linux C<fallocate> documentation for details.
657
658C<$mode> is usually C<0> or C<IO::AIO::FALLOC_FL_KEEP_SIZE> to allocate
659space, or C<IO::AIO::FALLOC_FL_PUNCH_HOLE | IO::AIO::FALLOC_FL_KEEP_SIZE>,
660to deallocate a file range.
661
662IO::AIO also supports C<FALLOC_FL_COLLAPSE_RANGE>, to remove a range
663(without leaving a hole), C<FALLOC_FL_ZERO_RANGE>, to zero a range,
664C<FALLOC_FL_INSERT_RANGE> to insert a range and C<FALLOC_FL_UNSHARE_RANGE>
665to unshare shared blocks (see your L<fallocate(2)> manpage).
666
667The file system block size used by C<fallocate> is presumably the
668C<f_bsize> returned by C<statvfs>, but different filesystems and filetypes
669can dictate other limitations.
670
671If C<fallocate> isn't available or cannot be emulated (currently no
672emulation will be attempted), passes C<-1> and sets C<$!> to C<ENOSYS>.
673
674
471=item aio_chmod $fh_or_path, $mode, $callback->($status) 675=item aio_chmod $fh_or_path, $mode, $callback->($status)
472 676
473Works like perl's C<chmod> function. 677Works like perl's C<chmod> function.
474 678
475 679
477 681
478Asynchronously unlink (delete) a file and call the callback with the 682Asynchronously unlink (delete) a file and call the callback with the
479result code. 683result code.
480 684
481 685
482=item aio_mknod $path, $mode, $dev, $callback->($status) 686=item aio_mknod $pathname, $mode, $dev, $callback->($status)
483 687
484[EXPERIMENTAL] 688[EXPERIMENTAL]
485 689
486Asynchronously create a device node (or fifo). See mknod(2). 690Asynchronously create a device node (or fifo). See mknod(2).
487 691
488The only (POSIX-) portable way of calling this function is: 692The only (POSIX-) portable way of calling this function is:
489 693
490 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 694 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
491 695
696See C<aio_stat> for info about some potentially helpful extra constants
697and functions.
492 698
493=item aio_link $srcpath, $dstpath, $callback->($status) 699=item aio_link $srcpath, $dstpath, $callback->($status)
494 700
495Asynchronously create a new link to the existing object at C<$srcpath> at 701Asynchronously create a new link to the existing object at C<$srcpath> at
496the path C<$dstpath> and call the callback with the result code. 702the path C<$dstpath> and call the callback with the result code.
500 706
501Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 707Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
502the path C<$dstpath> and call the callback with the result code. 708the path C<$dstpath> and call the callback with the result code.
503 709
504 710
505=item aio_readlink $path, $callback->($link) 711=item aio_readlink $pathname, $callback->($link)
506 712
507Asynchronously read the symlink specified by C<$path> and pass it to 713Asynchronously read the symlink specified by C<$path> and pass it to
508the callback. If an error occurs, nothing or undef gets passed to the 714the callback. If an error occurs, nothing or undef gets passed to the
509callback. 715callback.
510 716
511 717
718=item aio_realpath $pathname, $callback->($path)
719
720Asynchronously make the path absolute and resolve any symlinks in
721C<$path>. The resulting path only consists of directories (same as
722L<Cwd::realpath>).
723
724This request can be used to get the absolute path of the current working
725directory by passing it a path of F<.> (a single dot).
726
727
512=item aio_rename $srcpath, $dstpath, $callback->($status) 728=item aio_rename $srcpath, $dstpath, $callback->($status)
513 729
514Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 730Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
515rename(2) and call the callback with the result code. 731rename(2) and call the callback with the result code.
732
733On systems that support the AIO::WD working directory abstraction
734natively, the case C<[$wd, "."]> as C<$srcpath> is specialcased - instead
735of failing, C<rename> is called on the absolute path of C<$wd>.
736
737
738=item aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
739
740Basically a version of C<aio_rename> with an additional C<$flags>
741argument. Calling this with C<$flags=0> is the same as calling
742C<aio_rename>.
743
744Non-zero flags are currently only supported on GNU/Linux systems that
745support renameat2. Other systems fail with C<ENOSYS> in this case.
746
747The following constants are available (missing ones are, as usual C<0>),
748see renameat2(2) for details:
749
750C<IO::AIO::RENAME_NOREPLACE>, C<IO::AIO::RENAME_EXCHANGE>
751and C<IO::AIO::RENAME_WHITEOUT>.
516 752
517 753
518=item aio_mkdir $pathname, $mode, $callback->($status) 754=item aio_mkdir $pathname, $mode, $callback->($status)
519 755
520Asynchronously mkdir (create) a directory and call the callback with 756Asynchronously mkdir (create) a directory and call the callback with
525=item aio_rmdir $pathname, $callback->($status) 761=item aio_rmdir $pathname, $callback->($status)
526 762
527Asynchronously rmdir (delete) a directory and call the callback with the 763Asynchronously rmdir (delete) a directory and call the callback with the
528result code. 764result code.
529 765
766On systems that support the AIO::WD working directory abstraction
767natively, the case C<[$wd, "."]> is specialcased - instead of failing,
768C<rmdir> is called on the absolute path of C<$wd>.
769
530 770
531=item aio_readdir $pathname, $callback->($entries) 771=item aio_readdir $pathname, $callback->($entries)
532 772
533Unlike the POSIX call of the same name, C<aio_readdir> reads an entire 773Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
534directory (i.e. opendir + readdir + closedir). The entries will not be 774directory (i.e. opendir + readdir + closedir). The entries will not be
538array-ref with the filenames. 778array-ref with the filenames.
539 779
540 780
541=item aio_readdirx $pathname, $flags, $callback->($entries, $flags) 781=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
542 782
543Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune 783Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
544behaviour and output format. In case of an error, C<$entries> will be 784tune behaviour and output format. In case of an error, C<$entries> will be
545C<undef>. 785C<undef>.
546 786
547The flags are a combination of the following constants, ORed together (the 787The flags are a combination of the following constants, ORed together (the
548flags will also be passed to the callback, possibly modified): 788flags will also be passed to the callback, possibly modified):
549 789
550=over 4 790=over 4
551 791
552=item IO::AIO::READDIR_DENTS 792=item IO::AIO::READDIR_DENTS
553 793
554When this flag is off, then the callback gets an arrayref with of names 794When this flag is off, then the callback gets an arrayref consisting of
555only (as with C<aio_readdir>), otherwise it gets an arrayref with 795names only (as with C<aio_readdir>), otherwise it gets an arrayref with
556C<[$name, $type, $inode]> arrayrefs, each describing a single directory 796C<[$name, $type, $inode]> arrayrefs, each describing a single directory
557entry in more detail. 797entry in more detail.
558 798
559C<$name> is the name of the entry. 799C<$name> is the name of the entry.
560 800
573systems that do not deliver the inode information. 813systems that do not deliver the inode information.
574 814
575=item IO::AIO::READDIR_DIRS_FIRST 815=item IO::AIO::READDIR_DIRS_FIRST
576 816
577When this flag is set, then the names will be returned in an order where 817When this flag is set, then the names will be returned in an order where
578likely directories come first. This is useful when you need to quickly 818likely directories come first, in optimal stat order. This is useful when
579find directories, or you want to find all directories while avoiding to 819you need to quickly find directories, or you want to find all directories
580stat() each entry. 820while avoiding to stat() each entry.
581 821
582If the system returns type information in readdir, then this is used 822If the system returns type information in readdir, then this is used
583to find directories directly. Otherwise, likely directories are files 823to find directories directly. Otherwise, likely directories are names
584beginning with ".", or otherwise files with no dots, of which files with 824beginning with ".", or otherwise names with no dots, of which names with
585short names are tried first. 825short names are tried first.
586 826
587=item IO::AIO::READDIR_STAT_ORDER 827=item IO::AIO::READDIR_STAT_ORDER
588 828
589When this flag is set, then the names will be returned in an order 829When this flag is set, then the names will be returned in an order
596 836
597=item IO::AIO::READDIR_FOUND_UNKNOWN 837=item IO::AIO::READDIR_FOUND_UNKNOWN
598 838
599This flag should not be set when calling C<aio_readdirx>. Instead, it 839This flag should not be set when calling C<aio_readdirx>. Instead, it
600is being set by C<aio_readdirx>, when any of the C<$type>'s found were 840is being set by C<aio_readdirx>, when any of the C<$type>'s found were
601C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all 841C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
602C<$type>'s are known, which can be used to speed up some algorithms. 842C<$type>'s are known, which can be used to speed up some algorithms.
603 843
604=back 844=back
605 845
606 846
847=item aio_slurp $pathname, $offset, $length, $data, $callback->($status)
848
849Opens, reads and closes the given file. The data is put into C<$data>,
850which is resized as required.
851
852If C<$offset> is negative, then it is counted from the end of the file.
853
854If C<$length> is zero, then the remaining length of the file is
855used. Also, in this case, the same limitations to modifying C<$data> apply
856as when IO::AIO::mmap is used, i.e. it must only be modified in-place
857with C<substr>. If the size of the file is known, specifying a non-zero
858C<$length> results in a performance advantage.
859
860This request is similar to the older C<aio_load> request, but since it is
861a single request, it might be more efficient to use.
862
863Example: load F</etc/passwd> into C<$passwd>.
864
865 my $passwd;
866 aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
867 $_[0] >= 0
868 or die "/etc/passwd: $!\n";
869
870 printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
871 print $passwd;
872 };
873 IO::AIO::flush;
874
875
607=item aio_load $path, $data, $callback->($status) 876=item aio_load $pathname, $data, $callback->($status)
608 877
609This is a composite request that tries to fully load the given file into 878This is a composite request that tries to fully load the given file into
610memory. Status is the same as with aio_read. 879memory. Status is the same as with aio_read.
880
881Using C<aio_slurp> might be more efficient, as it is a single request.
611 882
612=cut 883=cut
613 884
614sub aio_load($$;$) { 885sub aio_load($$;$) {
615 my ($path, undef, $cb) = @_; 886 my ($path, undef, $cb) = @_;
634 905
635=item aio_copy $srcpath, $dstpath, $callback->($status) 906=item aio_copy $srcpath, $dstpath, $callback->($status)
636 907
637Try to copy the I<file> (directories not supported as either source or 908Try to copy the I<file> (directories not supported as either source or
638destination) from C<$srcpath> to C<$dstpath> and call the callback with 909destination) from C<$srcpath> to C<$dstpath> and call the callback with
639the C<0> (error) or C<-1> ok. 910a status of C<0> (ok) or C<-1> (error, see C<$!>).
911
912Existing destination files will be truncated.
640 913
641This is a composite request that creates the destination file with 914This is a composite request that creates the destination file with
642mode 0200 and copies the contents of the source file into it using 915mode 0200 and copies the contents of the source file into it using
643C<aio_sendfile>, followed by restoring atime, mtime, access mode and 916C<aio_sendfile>, followed by restoring atime, mtime, access mode and
644uid/gid, in that order. 917uid/gid, in that order.
656 my $grp = aio_group $cb; 929 my $grp = aio_group $cb;
657 930
658 aioreq_pri $pri; 931 aioreq_pri $pri;
659 add $grp aio_open $src, O_RDONLY, 0, sub { 932 add $grp aio_open $src, O_RDONLY, 0, sub {
660 if (my $src_fh = $_[0]) { 933 if (my $src_fh = $_[0]) {
661 my @stat = stat $src_fh; # hmm, might bock over nfs? 934 my @stat = stat $src_fh; # hmm, might block over nfs?
662 935
663 aioreq_pri $pri; 936 aioreq_pri $pri;
664 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub { 937 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
665 if (my $dst_fh = $_[0]) { 938 if (my $dst_fh = $_[0]) {
666 aioreq_pri $pri; 939 aioreq_pri $pri;
713 986
714=item aio_move $srcpath, $dstpath, $callback->($status) 987=item aio_move $srcpath, $dstpath, $callback->($status)
715 988
716Try to move the I<file> (directories not supported as either source or 989Try to move the I<file> (directories not supported as either source or
717destination) from C<$srcpath> to C<$dstpath> and call the callback with 990destination) from C<$srcpath> to C<$dstpath> and call the callback with
718the C<0> (error) or C<-1> ok. 991a status of C<0> (ok) or C<-1> (error, see C<$!>).
719 992
720This is a composite request that tries to rename(2) the file first; if 993This is a composite request that tries to rename(2) the file first; if
721rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if 994rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
722that is successful, unlinks the C<$srcpath>. 995that is successful, unlinks the C<$srcpath>.
723 996
734 if ($_[0] && $! == EXDEV) { 1007 if ($_[0] && $! == EXDEV) {
735 aioreq_pri $pri; 1008 aioreq_pri $pri;
736 add $grp aio_copy $src, $dst, sub { 1009 add $grp aio_copy $src, $dst, sub {
737 $grp->result ($_[0]); 1010 $grp->result ($_[0]);
738 1011
739 if (!$_[0]) { 1012 unless ($_[0]) {
740 aioreq_pri $pri; 1013 aioreq_pri $pri;
741 add $grp aio_unlink $src; 1014 add $grp aio_unlink $src;
742 } 1015 }
743 }; 1016 };
744 } else { 1017 } else {
747 }; 1020 };
748 1021
749 $grp 1022 $grp
750} 1023}
751 1024
752=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 1025=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
753 1026
754Scans a directory (similar to C<aio_readdir>) but additionally tries to 1027Scans a directory (similar to C<aio_readdir>) but additionally tries to
755efficiently separate the entries of directory C<$path> into two sets of 1028efficiently separate the entries of directory C<$path> into two sets of
756names, directories you can recurse into (directories), and ones you cannot 1029names, directories you can recurse into (directories), and ones you cannot
757recurse into (everything else, including symlinks to directories). 1030recurse into (everything else, including symlinks to directories).
758 1031
759C<aio_scandir> is a composite request that creates of many sub requests_ 1032C<aio_scandir> is a composite request that generates many sub requests.
760C<$maxreq> specifies the maximum number of outstanding aio requests that 1033C<$maxreq> specifies the maximum number of outstanding aio requests that
761this function generates. If it is C<< <= 0 >>, then a suitable default 1034this function generates. If it is C<< <= 0 >>, then a suitable default
762will be chosen (currently 4). 1035will be chosen (currently 4).
763 1036
764On error, the callback is called without arguments, otherwise it receives 1037On error, the callback is called without arguments, otherwise it receives
788Then entries will be sorted into likely directories a non-initial dot 1061Then entries will be sorted into likely directories a non-initial dot
789currently) and likely non-directories (see C<aio_readdirx>). Then every 1062currently) and likely non-directories (see C<aio_readdirx>). Then every
790entry plus an appended C</.> will be C<stat>'ed, likely directories first, 1063entry plus an appended C</.> will be C<stat>'ed, likely directories first,
791in order of their inode numbers. If that succeeds, it assumes that the 1064in order of their inode numbers. If that succeeds, it assumes that the
792entry is a directory or a symlink to directory (which will be checked 1065entry is a directory or a symlink to directory (which will be checked
793seperately). This is often faster than stat'ing the entry itself because 1066separately). This is often faster than stat'ing the entry itself because
794filesystems might detect the type of the entry without reading the inode 1067filesystems might detect the type of the entry without reading the inode
795data (e.g. ext2fs filetype feature), even on systems that cannot return 1068data (e.g. ext2fs filetype feature), even on systems that cannot return
796the filetype information on readdir. 1069the filetype information on readdir.
797 1070
798If the known number of directories (link count - 2) has been reached, the 1071If the known number of directories (link count - 2) has been reached, the
814 1087
815 my $grp = aio_group $cb; 1088 my $grp = aio_group $cb;
816 1089
817 $maxreq = 4 if $maxreq <= 0; 1090 $maxreq = 4 if $maxreq <= 0;
818 1091
819 # stat once 1092 # get a wd object
820 aioreq_pri $pri; 1093 aioreq_pri $pri;
821 add $grp aio_stat $path, sub { 1094 add $grp aio_wd $path, sub {
1095 $_[0]
822 return $grp->result () if $_[0]; 1096 or return $grp->result ();
823 my $now = time;
824 my $hash1 = join ":", (stat _)[0,1,3,7,9];
825 1097
826 # read the directory entries 1098 my $wd = [shift, "."];
1099
1100 # stat once
827 aioreq_pri $pri; 1101 aioreq_pri $pri;
828 add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub { 1102 add $grp aio_stat $wd, sub {
829 my $entries = shift
830 or return $grp->result (); 1103 return $grp->result () if $_[0];
1104 my $now = time;
1105 my $hash1 = join ":", (stat _)[0,1,3,7,9];
831 1106
832 # stat the dir another time 1107 # read the directory entries
833 aioreq_pri $pri; 1108 aioreq_pri $pri;
1109 add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
1110 my $entries = shift
1111 or return $grp->result ();
1112
1113 # stat the dir another time
1114 aioreq_pri $pri;
834 add $grp aio_stat $path, sub { 1115 add $grp aio_stat $wd, sub {
835 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 1116 my $hash2 = join ":", (stat _)[0,1,3,7,9];
836 1117
837 my $ndirs; 1118 my $ndirs;
838 1119
839 # take the slow route if anything looks fishy 1120 # take the slow route if anything looks fishy
840 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 1121 if ($hash1 ne $hash2 or (stat _)[9] == $now) {
841 $ndirs = -1; 1122 $ndirs = -1;
842 } else { 1123 } else {
843 # if nlink == 2, we are finished 1124 # if nlink == 2, we are finished
844 # for non-posix-fs's, we rely on nlink < 2 1125 # for non-posix-fs's, we rely on nlink < 2
845 $ndirs = (stat _)[3] - 2 1126 $ndirs = (stat _)[3] - 2
846 or return $grp->result ([], $entries); 1127 or return $grp->result ([], $entries);
847 } 1128 }
848 1129
849 my (@dirs, @nondirs); 1130 my (@dirs, @nondirs);
850 1131
851 my $statgrp = add $grp aio_group sub { 1132 my $statgrp = add $grp aio_group sub {
852 $grp->result (\@dirs, \@nondirs); 1133 $grp->result (\@dirs, \@nondirs);
853 }; 1134 };
854 1135
855 limit $statgrp $maxreq; 1136 limit $statgrp $maxreq;
856 feed $statgrp sub { 1137 feed $statgrp sub {
857 return unless @$entries; 1138 return unless @$entries;
858 my $entry = shift @$entries; 1139 my $entry = shift @$entries;
859 1140
860 aioreq_pri $pri; 1141 aioreq_pri $pri;
1142 $wd->[1] = "$entry/.";
861 add $statgrp aio_stat "$path/$entry/.", sub { 1143 add $statgrp aio_stat $wd, sub {
862 if ($_[0] < 0) { 1144 if ($_[0] < 0) {
863 push @nondirs, $entry; 1145 push @nondirs, $entry;
864 } else { 1146 } else {
865 # need to check for real directory 1147 # need to check for real directory
866 aioreq_pri $pri; 1148 aioreq_pri $pri;
1149 $wd->[1] = $entry;
867 add $statgrp aio_lstat "$path/$entry", sub { 1150 add $statgrp aio_lstat $wd, sub {
868 if (-d _) { 1151 if (-d _) {
869 push @dirs, $entry; 1152 push @dirs, $entry;
870 1153
871 unless (--$ndirs) { 1154 unless (--$ndirs) {
872 push @nondirs, @$entries; 1155 push @nondirs, @$entries;
873 feed $statgrp; 1156 feed $statgrp;
1157 }
1158 } else {
1159 push @nondirs, $entry;
874 } 1160 }
875 } else {
876 push @nondirs, $entry;
877 } 1161 }
878 } 1162 }
879 } 1163 };
880 }; 1164 };
881 }; 1165 };
882 }; 1166 };
883 }; 1167 };
884 }; 1168 };
885 1169
886 $grp 1170 $grp
887} 1171}
888 1172
889=item aio_rmtree $path, $callback->($status) 1173=item aio_rmtree $pathname, $callback->($status)
890 1174
891Delete a directory tree starting (and including) C<$path>, return the 1175Delete a directory tree starting (and including) C<$path>, return the
892status of the final C<rmdir> only. This is a composite request that 1176status of the final C<rmdir> only. This is a composite request that
893uses C<aio_scandir> to recurse into and rmdir directories, and unlink 1177uses C<aio_scandir> to recurse into and rmdir directories, and unlink
894everything else. 1178everything else.
895 1179
896=cut 1180=cut
897 1181
919 }; 1203 };
920 1204
921 $grp 1205 $grp
922} 1206}
923 1207
1208=item aio_fcntl $fh, $cmd, $arg, $callback->($status)
1209
1210=item aio_ioctl $fh, $request, $buf, $callback->($status)
1211
1212These work just like the C<fcntl> and C<ioctl> built-in functions, except
1213they execute asynchronously and pass the return value to the callback.
1214
1215Both calls can be used for a lot of things, some of which make more sense
1216to run asynchronously in their own thread, while some others make less
1217sense. For example, calls that block waiting for external events, such
1218as locking, will also lock down an I/O thread while it is waiting, which
1219can deadlock the whole I/O system. At the same time, there might be no
1220alternative to using a thread to wait.
1221
1222So in general, you should only use these calls for things that do
1223(filesystem) I/O, not for things that wait for other events (network,
1224other processes), although if you are careful and know what you are doing,
1225you still can.
1226
1227The following constants are available (missing ones are, as usual C<0>):
1228
1229C<F_DUPFD_CLOEXEC>,
1230
1231C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>,
1232
1233C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>.
1234
1235C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>,
1236C<FS_IOC_FIEMAP>.
1237
1238C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>,
1239C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>.
1240
1241C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>,
1242C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>,
1243C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>,
1244C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>,
1245C<FS_FL_USER_MODIFIABLE>.
1246
1247C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>,
1248C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>,
1249C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>,
1250C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>,
1251
924=item aio_sync $callback->($status) 1252=item aio_sync $callback->($status)
925 1253
926Asynchronously call sync and call the callback when finished. 1254Asynchronously call sync and call the callback when finished.
927 1255
928=item aio_fsync $fh, $callback->($status) 1256=item aio_fsync $fh, $callback->($status)
935Asynchronously call fdatasync on the given filehandle and call the 1263Asynchronously call fdatasync on the given filehandle and call the
936callback with the fdatasync result code. 1264callback with the fdatasync result code.
937 1265
938If this call isn't available because your OS lacks it or it couldn't be 1266If this call isn't available because your OS lacks it or it couldn't be
939detected, it will be emulated by calling C<fsync> instead. 1267detected, it will be emulated by calling C<fsync> instead.
1268
1269=item aio_syncfs $fh, $callback->($status)
1270
1271Asynchronously call the syncfs syscall to sync the filesystem associated
1272to the given filehandle and call the callback with the syncfs result
1273code. If syncfs is not available, calls sync(), but returns C<-1> and sets
1274errno to C<ENOSYS> nevertheless.
940 1275
941=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 1276=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
942 1277
943Sync the data portion of the file specified by C<$offset> and C<$length> 1278Sync the data portion of the file specified by C<$offset> and C<$length>
944to disk (but NOT the metadata), by calling the Linux-specific 1279to disk (but NOT the metadata), by calling the Linux-specific
948C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>, 1283C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
949C<IO::AIO::SYNC_FILE_RANGE_WRITE> and 1284C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
950C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range 1285C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
951manpage for details. 1286manpage for details.
952 1287
953=item aio_pathsync $path, $callback->($status) 1288=item aio_pathsync $pathname, $callback->($status)
954 1289
955This request tries to open, fsync and close the given path. This is a 1290This request tries to open, fsync and close the given path. This is a
956composite request intended to sync directories after directory operations 1291composite request intended to sync directories after directory operations
957(E.g. rename). This might not work on all operating systems or have any 1292(E.g. rename). This might not work on all operating systems or have any
958specific effect, but usually it makes sure that directory changes get 1293specific effect, but usually it makes sure that directory changes get
959written to disc. It works for anything that can be opened for read-only, 1294written to disc. It works for anything that can be opened for read-only,
960not just directories. 1295not just directories.
1296
1297Future versions of this function might fall back to other methods when
1298C<fsync> on the directory fails (such as calling C<sync>).
961 1299
962Passes C<0> when everything went ok, and C<-1> on error. 1300Passes C<0> when everything went ok, and C<-1> on error.
963 1301
964=cut 1302=cut
965 1303
986 }; 1324 };
987 1325
988 $grp 1326 $grp
989} 1327}
990 1328
1329=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
1330
1331This is a rather advanced IO::AIO call, which only works on mmap(2)ed
1332scalars (see the C<IO::AIO::mmap> function, although it also works on data
1333scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
1334scalar must only be modified in-place while an aio operation is pending on
1335it).
1336
1337It calls the C<msync> function of your OS, if available, with the memory
1338area starting at C<$offset> in the string and ending C<$length> bytes
1339later. If C<$length> is negative, counts from the end, and if C<$length>
1340is C<undef>, then it goes till the end of the string. The flags can be
1341either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional
1342C<IO::AIO::MS_INVALIDATE>.
1343
1344=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
1345
1346This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1347scalars.
1348
1349It touches (reads or writes) all memory pages in the specified
1350range inside the scalar. All caveats and parameters are the same
1351as for C<aio_msync>, above, except for flags, which must be either
1352C<0> (which reads all pages and ensures they are instantiated) or
1353C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and
1354writing an octet from it, which dirties the page).
1355
1356=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1357
1358This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1359scalars.
1360
1361It reads in all the pages of the underlying storage into memory (if any)
1362and locks them, so they are not getting swapped/paged out or removed.
1363
1364If C<$length> is undefined, then the scalar will be locked till the end.
1365
1366On systems that do not implement C<mlock>, this function returns C<-1>
1367and sets errno to C<ENOSYS>.
1368
1369Note that the corresponding C<munlock> is synchronous and is
1370documented under L<MISCELLANEOUS FUNCTIONS>.
1371
1372Example: open a file, mmap and mlock it - both will be undone when
1373C<$data> gets destroyed.
1374
1375 open my $fh, "<", $path or die "$path: $!";
1376 my $data;
1377 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1378 aio_mlock $data; # mlock in background
1379
1380=item aio_mlockall $flags, $callback->($status)
1381
1382Calls the C<mlockall> function with the given C<$flags> (a combination of
1383C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
1384
1385On systems that do not implement C<mlockall>, this function returns C<-1>
1386and sets errno to C<ENOSYS>.
1387
1388Note that the corresponding C<munlockall> is synchronous and is
1389documented under L<MISCELLANEOUS FUNCTIONS>.
1390
1391Example: asynchronously lock all current and future pages into memory.
1392
1393 aio_mlockall IO::AIO::MCL_FUTURE;
1394
1395=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
1396
1397Queries the extents of the given file (by calling the Linux C<FIEMAP>
1398ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If
1399the ioctl is not available on your OS, then this request will fail with
1400C<ENOSYS>.
1401
1402C<$start> is the starting offset to query extents for, C<$length> is the
1403size of the range to query - if it is C<undef>, then the whole file will
1404be queried.
1405
1406C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or
1407C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also
1408exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query
1409the data portion.
1410
1411C<$count> is the maximum number of extent records to return. If it is
1412C<undef>, then IO::AIO queries all extents of the range. As a very special
1413case, if it is C<0>, then the callback receives the number of extents
1414instead of the extents themselves (which is unreliable, see below).
1415
1416If an error occurs, the callback receives no arguments. The special
1417C<errno> value C<IO::AIO::EBADR> is available to test for flag errors.
1418
1419Otherwise, the callback receives an array reference with extent
1420structures. Each extent structure is an array reference itself, with the
1421following members:
1422
1423 [$logical, $physical, $length, $flags]
1424
1425Flags is any combination of the following flag values (typically either C<0>
1426or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)):
1427
1428C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>,
1429C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>,
1430C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>,
1431C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>,
1432C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or
1433C<IO::AIO::FIEMAP_EXTENT_SHARED>.
1434
1435At the time of this writing (Linux 3.2), this request is unreliable unless
1436C<$count> is C<undef>, as the kernel has all sorts of bugs preventing
1437it to return all extents of a range for files with a large number of
1438extents. The code (only) works around all these issues if C<$count> is
1439C<undef>.
1440
991=item aio_group $callback->(...) 1441=item aio_group $callback->(...)
992 1442
993This is a very special aio request: Instead of doing something, it is a 1443This is a very special aio request: Instead of doing something, it is a
994container for other aio requests, which is useful if you want to bundle 1444container for other aio requests, which is useful if you want to bundle
995many requests into a single, composite, request with a definite callback 1445many requests into a single, composite, request with a definite callback
1031like sleep and file handle readable/writable, the overhead this creates is 1481like sleep and file handle readable/writable, the overhead this creates is
1032immense (it blocks a thread for a long time) so do not use this function 1482immense (it blocks a thread for a long time) so do not use this function
1033except to put your application under artificial I/O pressure. 1483except to put your application under artificial I/O pressure.
1034 1484
1035=back 1485=back
1486
1487
1488=head2 IO::AIO::WD - multiple working directories
1489
1490Your process only has one current working directory, which is used by all
1491threads. This makes it hard to use relative paths (some other component
1492could call C<chdir> at any time, and it is hard to control when the path
1493will be used by IO::AIO).
1494
1495One solution for this is to always use absolute paths. This usually works,
1496but can be quite slow (the kernel has to walk the whole path on every
1497access), and can also be a hassle to implement.
1498
1499Newer POSIX systems have a number of functions (openat, fdopendir,
1500futimensat and so on) that make it possible to specify working directories
1501per operation.
1502
1503For portability, and because the clowns who "designed", or shall I write,
1504perpetrated this new interface were obviously half-drunk, this abstraction
1505cannot be perfect, though.
1506
1507IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
1508object. This object stores the canonicalised, absolute version of the
1509path, and on systems that allow it, also a directory file descriptor.
1510
1511Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
1512or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
1513object and a pathname instead (or the IO::AIO::WD object alone, which
1514gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
1515IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
1516to that IO::AIO::WD object.
1517
1518For example, to get a wd object for F</etc> and then stat F<passwd>
1519inside, you would write:
1520
1521 aio_wd "/etc", sub {
1522 my $etcdir = shift;
1523
1524 # although $etcdir can be undef on error, there is generally no reason
1525 # to check for errors here, as aio_stat will fail with ENOENT
1526 # when $etcdir is undef.
1527
1528 aio_stat [$etcdir, "passwd"], sub {
1529 # yay
1530 };
1531 };
1532
1533The fact that C<aio_wd> is a request and not a normal function shows that
1534creating an IO::AIO::WD object is itself a potentially blocking operation,
1535which is why it is done asynchronously.
1536
1537To stat the directory obtained with C<aio_wd> above, one could write
1538either of the following three request calls:
1539
1540 aio_lstat "/etc" , sub { ... # pathname as normal string
1541 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1542 aio_lstat $wd , sub { ... # shorthand for the previous
1543
1544As with normal pathnames, IO::AIO keeps a copy of the working directory
1545object and the pathname string, so you could write the following without
1546causing any issues due to C<$path> getting reused:
1547
1548 my $path = [$wd, undef];
1549
1550 for my $name (qw(abc def ghi)) {
1551 $path->[1] = $name;
1552 aio_stat $path, sub {
1553 # ...
1554 };
1555 }
1556
1557There are some caveats: when directories get renamed (or deleted), the
1558pathname string doesn't change, so will point to the new directory (or
1559nowhere at all), while the directory fd, if available on the system,
1560will still point to the original directory. Most functions accepting a
1561pathname will use the directory fd on newer systems, and the string on
1562older systems. Some functions (such as C<aio_realpath>) will always rely on
1563the string form of the pathname.
1564
1565So this functionality is mainly useful to get some protection against
1566C<chdir>, to easily get an absolute path out of a relative path for future
1567reference, and to speed up doing many operations in the same directory
1568(e.g. when stat'ing all files in a directory).
1569
1570The following functions implement this working directory abstraction:
1571
1572=over 4
1573
1574=item aio_wd $pathname, $callback->($wd)
1575
1576Asynchonously canonicalise the given pathname and convert it to an
1577IO::AIO::WD object representing it. If possible and supported on the
1578system, also open a directory fd to speed up pathname resolution relative
1579to this working directory.
1580
1581If something goes wrong, then C<undef> is passwd to the callback instead
1582of a working directory object and C<$!> is set appropriately. Since
1583passing C<undef> as working directory component of a pathname fails the
1584request with C<ENOENT>, there is often no need for error checking in the
1585C<aio_wd> callback, as future requests using the value will fail in the
1586expected way.
1587
1588=item IO::AIO::CWD
1589
1590This is a compiletime constant (object) that represents the process
1591current working directory.
1592
1593Specifying this object as working directory object for a pathname is as if
1594the pathname would be specified directly, without a directory object. For
1595example, these calls are functionally identical:
1596
1597 aio_stat "somefile", sub { ... };
1598 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1599
1600=back
1601
1602To recover the path associated with an IO::AIO::WD object, you can use
1603C<aio_realpath>:
1604
1605 aio_realpath $wd, sub {
1606 warn "path is $_[0]\n";
1607 };
1608
1609Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir>
1610sometimes, fall back to using an absolue path.
1036 1611
1037=head2 IO::AIO::REQ CLASS 1612=head2 IO::AIO::REQ CLASS
1038 1613
1039All non-aggregate C<aio_*> functions return an object of this class when 1614All non-aggregate C<aio_*> functions return an object of this class when
1040called in non-void context. 1615called in non-void context.
1130=item $grp->cancel_subs 1705=item $grp->cancel_subs
1131 1706
1132Cancel all subrequests and clears any feeder, but not the group request 1707Cancel all subrequests and clears any feeder, but not the group request
1133itself. Useful when you queued a lot of events but got a result early. 1708itself. Useful when you queued a lot of events but got a result early.
1134 1709
1710The group request will finish normally (you cannot add requests to the
1711group).
1712
1135=item $grp->result (...) 1713=item $grp->result (...)
1136 1714
1137Set the result value(s) that will be passed to the group callback when all 1715Set the result value(s) that will be passed to the group callback when all
1138subrequests have finished and set the groups errno to the current value 1716subrequests have finished and set the groups errno to the current value
1139of errno (just like calling C<errno> without an error number). By default, 1717of errno (just like calling C<errno> without an error number). By default,
1155 1733
1156Sets a feeder/generator on this group: every group can have an attached 1734Sets a feeder/generator on this group: every group can have an attached
1157generator that generates requests if idle. The idea behind this is that, 1735generator that generates requests if idle. The idea behind this is that,
1158although you could just queue as many requests as you want in a group, 1736although you could just queue as many requests as you want in a group,
1159this might starve other requests for a potentially long time. For example, 1737this might starve other requests for a potentially long time. For example,
1160C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests, 1738C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
1161delaying any later requests for a long time. 1739requests, delaying any later requests for a long time.
1162 1740
1163To avoid this, and allow incremental generation of requests, you can 1741To avoid this, and allow incremental generation of requests, you can
1164instead a group and set a feeder on it that generates those requests. The 1742instead a group and set a feeder on it that generates those requests. The
1165feed callback will be called whenever there are few enough (see C<limit>, 1743feed callback will be called whenever there are few enough (see C<limit>,
1166below) requests active in the group itself and is expected to queue more 1744below) requests active in the group itself and is expected to queue more
1215 1793
1216See C<poll_cb> for an example. 1794See C<poll_cb> for an example.
1217 1795
1218=item IO::AIO::poll_cb 1796=item IO::AIO::poll_cb
1219 1797
1220Process some outstanding events on the result pipe. You have to call this 1798Process some requests that have reached the result phase (i.e. they have
1221regularly. Returns C<0> if all events could be processed, or C<-1> if it 1799been executed but the results are not yet reported). You have to call
1222returned earlier for whatever reason. Returns immediately when no events 1800this "regularly" to finish outstanding requests.
1223are outstanding. The amount of events processed depends on the settings of
1224C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
1225 1801
1802Returns C<0> if all events could be processed (or there were no
1803events to process), or C<-1> if it returned earlier for whatever
1804reason. Returns immediately when no events are outstanding. The amount
1805of events processed depends on the settings of C<IO::AIO::max_poll_req>,
1806C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>.
1807
1226If not all requests were processed for whatever reason, the filehandle 1808If not all requests were processed for whatever reason, the poll file
1227will still be ready when C<poll_cb> returns, so normally you don't have to 1809descriptor will still be ready when C<poll_cb> returns, so normally you
1228do anything special to have it called later. 1810don't have to do anything special to have it called later.
1811
1812Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
1813ready, it can be beneficial to call this function from loops which submit
1814a lot of requests, to make sure the results get processed when they become
1815available and not just when the loop is finished and the event loop takes
1816over again. This function returns very fast when there are no outstanding
1817requests.
1229 1818
1230Example: Install an Event watcher that automatically calls 1819Example: Install an Event watcher that automatically calls
1231IO::AIO::poll_cb with high priority (more examples can be found in the 1820IO::AIO::poll_cb with high priority (more examples can be found in the
1232SYNOPSIS section, at the top of this document): 1821SYNOPSIS section, at the top of this document):
1233 1822
1234 Event->io (fd => IO::AIO::poll_fileno, 1823 Event->io (fd => IO::AIO::poll_fileno,
1235 poll => 'r', async => 1, 1824 poll => 'r', async => 1,
1236 cb => \&IO::AIO::poll_cb); 1825 cb => \&IO::AIO::poll_cb);
1826
1827=item IO::AIO::poll_wait
1828
1829Wait until either at least one request is in the result phase or no
1830requests are outstanding anymore.
1831
1832This is useful if you want to synchronously wait for some requests to
1833become ready, without actually handling them.
1834
1835See C<nreqs> for an example.
1836
1837=item IO::AIO::poll
1838
1839Waits until some requests have been handled.
1840
1841Returns the number of requests processed, but is otherwise strictly
1842equivalent to:
1843
1844 IO::AIO::poll_wait, IO::AIO::poll_cb
1845
1846=item IO::AIO::flush
1847
1848Wait till all outstanding AIO requests have been handled.
1849
1850Strictly equivalent to:
1851
1852 IO::AIO::poll_wait, IO::AIO::poll_cb
1853 while IO::AIO::nreqs;
1237 1854
1238=item IO::AIO::max_poll_reqs $nreqs 1855=item IO::AIO::max_poll_reqs $nreqs
1239 1856
1240=item IO::AIO::max_poll_time $seconds 1857=item IO::AIO::max_poll_time $seconds
1241 1858
1266 # use a low priority so other tasks have priority 1883 # use a low priority so other tasks have priority
1267 Event->io (fd => IO::AIO::poll_fileno, 1884 Event->io (fd => IO::AIO::poll_fileno,
1268 poll => 'r', nice => 1, 1885 poll => 'r', nice => 1,
1269 cb => &IO::AIO::poll_cb); 1886 cb => &IO::AIO::poll_cb);
1270 1887
1271=item IO::AIO::poll_wait
1272
1273If there are any outstanding requests and none of them in the result
1274phase, wait till the result filehandle becomes ready for reading (simply
1275does a C<select> on the filehandle. This is useful if you want to
1276synchronously wait for some requests to finish).
1277
1278See C<nreqs> for an example.
1279
1280=item IO::AIO::poll
1281
1282Waits until some requests have been handled.
1283
1284Returns the number of requests processed, but is otherwise strictly
1285equivalent to:
1286
1287 IO::AIO::poll_wait, IO::AIO::poll_cb
1288
1289=item IO::AIO::flush
1290
1291Wait till all outstanding AIO requests have been handled.
1292
1293Strictly equivalent to:
1294
1295 IO::AIO::poll_wait, IO::AIO::poll_cb
1296 while IO::AIO::nreqs;
1297
1298=back 1888=back
1299 1889
1300=head3 CONTROLLING THE NUMBER OF THREADS 1890=head3 CONTROLLING THE NUMBER OF THREADS
1301 1891
1302=over 1892=over
1335 1925
1336Under normal circumstances you don't need to call this function. 1926Under normal circumstances you don't need to call this function.
1337 1927
1338=item IO::AIO::max_idle $nthreads 1928=item IO::AIO::max_idle $nthreads
1339 1929
1340Limit the number of threads (default: 4) that are allowed to idle (i.e., 1930Limit the number of threads (default: 4) that are allowed to idle
1341threads that did not get a request to process within 10 seconds). That 1931(i.e., threads that did not get a request to process within the idle
1342means if a thread becomes idle while C<$nthreads> other threads are also 1932timeout (default: 10 seconds). That means if a thread becomes idle while
1343idle, it will free its resources and exit. 1933C<$nthreads> other threads are also idle, it will free its resources and
1934exit.
1344 1935
1345This is useful when you allow a large number of threads (e.g. 100 or 1000) 1936This is useful when you allow a large number of threads (e.g. 100 or 1000)
1346to allow for extremely high load situations, but want to free resources 1937to allow for extremely high load situations, but want to free resources
1347under normal circumstances (1000 threads can easily consume 30MB of RAM). 1938under normal circumstances (1000 threads can easily consume 30MB of RAM).
1348 1939
1349The default is probably ok in most situations, especially if thread 1940The default is probably ok in most situations, especially if thread
1350creation is fast. If thread creation is very slow on your system you might 1941creation is fast. If thread creation is very slow on your system you might
1351want to use larger values. 1942want to use larger values.
1352 1943
1944=item IO::AIO::idle_timeout $seconds
1945
1946Sets the minimum idle timeout (default 10) after which worker threads are
1947allowed to exit. SEe C<IO::AIO::max_idle>.
1948
1353=item IO::AIO::max_outstanding $maxreqs 1949=item IO::AIO::max_outstanding $maxreqs
1950
1951Sets the maximum number of outstanding requests to C<$nreqs>. If
1952you do queue up more than this number of requests, the next call to
1953C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
1954C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
1955longer exceeded.
1956
1957In other words, this setting does not enforce a queue limit, but can be
1958used to make poll functions block if the limit is exceeded.
1354 1959
1355This is a very bad function to use in interactive programs because it 1960This is a very bad function to use in interactive programs because it
1356blocks, and a bad way to reduce concurrency because it is inexact: Better 1961blocks, and a bad way to reduce concurrency because it is inexact: Better
1357use an C<aio_group> together with a feed callback. 1962use an C<aio_group> together with a feed callback.
1358 1963
1359Sets the maximum number of outstanding requests to C<$nreqs>. If you 1964Its main use is in scripts without an event loop - when you want to stat
1360do queue up more than this number of requests, the next call to the 1965a lot of files, you can write something like this:
1361C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
1362function will block until the limit is no longer exceeded.
1363 1966
1364The default value is very large, so there is no practical limit on the 1967 IO::AIO::max_outstanding 32;
1365number of outstanding requests.
1366 1968
1367You can still queue as many requests as you want. Therefore, 1969 for my $path (...) {
1368C<max_outstanding> is mainly useful in simple scripts (with low values) or 1970 aio_stat $path , ...;
1369as a stop gap to shield against fatal memory overflow (with large values). 1971 IO::AIO::poll_cb;
1972 }
1973
1974 IO::AIO::flush;
1975
1976The call to C<poll_cb> inside the loop will normally return instantly, but
1977as soon as more thna C<32> reqeusts are in-flight, it will block until
1978some requests have been handled. This keeps the loop from pushing a large
1979number of C<aio_stat> requests onto the queue.
1980
1981The default value for C<max_outstanding> is very large, so there is no
1982practical limit on the number of outstanding requests.
1370 1983
1371=back 1984=back
1372 1985
1373=head3 STATISTICAL INFORMATION 1986=head3 STATISTICAL INFORMATION
1374 1987
1396 2009
1397=back 2010=back
1398 2011
1399=head3 MISCELLANEOUS FUNCTIONS 2012=head3 MISCELLANEOUS FUNCTIONS
1400 2013
1401IO::AIO implements some functions that might be useful, but are not 2014IO::AIO implements some functions that are useful when you want to use
1402asynchronous. 2015some "Advanced I/O" function not available to in Perl, without going the
2016"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*>
2017counterpart.
1403 2018
1404=over 4 2019=over 4
2020
2021=item $numfd = IO::AIO::get_fdlimit
2022
2023This function is I<EXPERIMENTAL> and subject to change.
2024
2025Tries to find the current file descriptor limit and returns it, or
2026C<undef> and sets C<$!> in case of an error. The limit is one larger than
2027the highest valid file descriptor number.
2028
2029=item IO::AIO::min_fdlimit [$numfd]
2030
2031This function is I<EXPERIMENTAL> and subject to change.
2032
2033Try to increase the current file descriptor limit(s) to at least C<$numfd>
2034by changing the soft or hard file descriptor resource limit. If C<$numfd>
2035is missing, it will try to set a very high limit, although this is not
2036recommended when you know the actual minimum that you require.
2037
2038If the limit cannot be raised enough, the function makes a best-effort
2039attempt to increase the limit as much as possible, using various
2040tricks, while still failing. You can query the resulting limit using
2041C<IO::AIO::get_fdlimit>.
2042
2043If an error occurs, returns C<undef> and sets C<$!>, otherwise returns
2044true.
1405 2045
1406=item IO::AIO::sendfile $ofh, $ifh, $offset, $count 2046=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
1407 2047
1408Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>, 2048Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
1409but is blocking (this makes most sense if you know the input data is 2049but is blocking (this makes most sense if you know the input data is
1412 2052
1413Returns the number of bytes copied, or C<-1> on error. 2053Returns the number of bytes copied, or C<-1> on error.
1414 2054
1415=item IO::AIO::fadvise $fh, $offset, $len, $advice 2055=item IO::AIO::fadvise $fh, $offset, $len, $advice
1416 2056
1417Simply calls the C<posix_fadvise> function (see it's 2057Simply calls the C<posix_fadvise> function (see its
1418manpage for details). The following advice constants are 2058manpage for details). The following advice constants are
1419avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>, 2059available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
1420C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>, 2060C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
1421C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>. 2061C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
1422 2062
1423On systems that do not implement C<posix_fadvise>, this function returns 2063On systems that do not implement C<posix_fadvise>, this function returns
1424ENOSYS, otherwise the return value of C<posix_fadvise>. 2064ENOSYS, otherwise the return value of C<posix_fadvise>.
1425 2065
2066=item IO::AIO::madvise $scalar, $offset, $len, $advice
2067
2068Simply calls the C<posix_madvise> function (see its
2069manpage for details). The following advice constants are
2070available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
2071C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>,
2072C<IO::AIO::MADV_DONTNEED>.
2073
2074If C<$offset> is negative, counts from the end. If C<$length> is negative,
2075the remaining length of the C<$scalar> is used. If possible, C<$length>
2076will be reduced to fit into the C<$scalar>.
2077
2078On systems that do not implement C<posix_madvise>, this function returns
2079ENOSYS, otherwise the return value of C<posix_madvise>.
2080
2081=item IO::AIO::mprotect $scalar, $offset, $len, $protect
2082
2083Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
2084$scalar (see its manpage for details). The following protect
2085constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
2086C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
2087
2088If C<$offset> is negative, counts from the end. If C<$length> is negative,
2089the remaining length of the C<$scalar> is used. If possible, C<$length>
2090will be reduced to fit into the C<$scalar>.
2091
2092On systems that do not implement C<mprotect>, this function returns
2093ENOSYS, otherwise the return value of C<mprotect>.
2094
2095=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
2096
2097Memory-maps a file (or anonymous memory range) and attaches it to the
2098given C<$scalar>, which will act like a string scalar. Returns true on
2099success, and false otherwise.
2100
2101The scalar must exist, but its contents do not matter - this means you
2102cannot use a nonexistant array or hash element. When in doubt, C<undef>
2103the scalar first.
2104
2105The only operations allowed on the mmapped scalar are C<substr>/C<vec>,
2106which don't change the string length, and most read-only operations such
2107as copying it or searching it with regexes and so on.
2108
2109Anything else is unsafe and will, at best, result in memory leaks.
2110
2111The memory map associated with the C<$scalar> is automatically removed
2112when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap>
2113or C<IO::AIO::munmap> functions are called on it.
2114
2115This calls the C<mmap>(2) function internally. See your system's manual
2116page for details on the C<$length>, C<$prot> and C<$flags> parameters.
2117
2118The C<$length> must be larger than zero and smaller than the actual
2119filesize.
2120
2121C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
2122C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
2123
2124C<$flags> can be a combination of
2125C<IO::AIO::MAP_SHARED> or
2126C<IO::AIO::MAP_PRIVATE>,
2127or a number of system-specific flags (when not available, the are C<0>):
2128C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant),
2129C<IO::AIO::MAP_LOCKED>,
2130C<IO::AIO::MAP_NORESERVE>,
2131C<IO::AIO::MAP_POPULATE>,
2132C<IO::AIO::MAP_NONBLOCK>,
2133C<IO::AIO::MAP_FIXED>,
2134C<IO::AIO::MAP_GROWSDOWN>,
2135C<IO::AIO::MAP_32BIT>,
2136C<IO::AIO::MAP_HUGETLB> or
2137C<IO::AIO::MAP_STACK>.
2138
2139If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
2140
2141C<$offset> is the offset from the start of the file - it generally must be
2142a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
2143
2144Example:
2145
2146 use Digest::MD5;
2147 use IO::AIO;
2148
2149 open my $fh, "<verybigfile"
2150 or die "$!";
2151
2152 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
2153 or die "verybigfile: $!";
2154
2155 my $fast_md5 = md5 $data;
2156
2157=item IO::AIO::munmap $scalar
2158
2159Removes a previous mmap and undefines the C<$scalar>.
2160
2161=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
2162
2163Calls the C<munlock> function, undoing the effects of a previous
2164C<aio_mlock> call (see its description for details).
2165
2166=item IO::AIO::munlockall
2167
2168Calls the C<munlockall> function.
2169
2170On systems that do not implement C<munlockall>, this function returns
2171ENOSYS, otherwise the return value of C<munlockall>.
2172
2173=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
2174
2175Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or
2176C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they
2177should be the file offset.
2178
2179C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might
2180silently corrupt the data in this case.
2181
2182The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>,
2183C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and
2184C<IO::AIO::SPLICE_F_GIFT>.
2185
2186See the C<splice(2)> manpage for details.
2187
2188=item IO::AIO::tee $r_fh, $w_fh, $length, $flags
2189
2190Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the
2191description for C<IO::AIO::splice> above for details.
2192
2193=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
2194
2195Attempts to query or change the pipe buffer size. Obviously works only
2196on pipes, and currently works only on GNU/Linux systems, and fails with
2197C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer
2198size on other systems, drop me a note.
2199
2200=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
2201
2202This is a direct interface to the Linux L<pipe2(2)> system call. If
2203C<$flags> is missing or C<0>, then this should be the same as a call to
2204perl's built-in C<pipe> function and create a new pipe, and works on
2205systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe
2206(..., 4096, O_BINARY)>.
2207
2208If C<$flags> is non-zero, it tries to invoke the pipe2 system call with
2209the given flags (Linux 2.6.27, glibc 2.9).
2210
2211On success, the read and write file handles are returned.
2212
2213On error, nothing will be returned. If the pipe2 syscall is missing and
2214C<$flags> is non-zero, fails with C<ENOSYS>.
2215
2216Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the
2217time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and
2218C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported.
2219
2220Example: create a pipe race-free w.r.t. threads and fork:
2221
2222 my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
2223 or die "pipe2: $!\n";
2224
2225=item $fh = IO::AIO::eventfd ([$initval, [$flags]])
2226
2227This is a direct interface to the Linux L<eventfd(2)> system call. The
2228(unhelpful) defaults for C<$initval> and C<$flags> are C<0> for both.
2229
2230On success, the new eventfd filehandle is returned, otherwise returns
2231C<undef>. If the eventfd syscall is missing, fails with C<ENOSYS>.
2232
2233Please refer to L<eventfd(2)> for more info on this call.
2234
2235The following symbol flag values are available: C<IO::AIO::EFD_CLOEXEC>,
2236C<IO::AIO::EFD_NONBLOCK> and C<IO::AIO::EFD_SEMAPHORE> (Linux 2.6.30).
2237
1426=back 2238=back
1427 2239
1428=cut 2240=cut
1429 2241
1430min_parallel 8; 2242min_parallel 8;
1431 2243
1432END { flush } 2244END { flush }
1433 2245
14341; 22461;
1435 2247
2248=head1 EVENT LOOP INTEGRATION
2249
2250It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
2251automatically into many event loops:
2252
2253 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
2254 use AnyEvent::AIO;
2255
2256You can also integrate IO::AIO manually into many event loops, here are
2257some examples of how to do this:
2258
2259 # EV integration
2260 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
2261
2262 # Event integration
2263 Event->io (fd => IO::AIO::poll_fileno,
2264 poll => 'r',
2265 cb => \&IO::AIO::poll_cb);
2266
2267 # Glib/Gtk2 integration
2268 add_watch Glib::IO IO::AIO::poll_fileno,
2269 in => sub { IO::AIO::poll_cb; 1 };
2270
2271 # Tk integration
2272 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
2273 readable => \&IO::AIO::poll_cb);
2274
2275 # Danga::Socket integration
2276 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
2277 \&IO::AIO::poll_cb);
2278
1436=head2 FORK BEHAVIOUR 2279=head2 FORK BEHAVIOUR
1437 2280
1438This module should do "the right thing" when the process using it forks: 2281Usage of pthreads in a program changes the semantics of fork
2282considerably. Specifically, only async-safe functions can be called after
2283fork. Perl doesn't know about this, so in general, you cannot call fork
2284with defined behaviour in perl if pthreads are involved. IO::AIO uses
2285pthreads, so this applies, but many other extensions and (for inexplicable
2286reasons) perl itself often is linked against pthreads, so this limitation
2287applies to quite a lot of perls.
1439 2288
1440Before the fork, IO::AIO enters a quiescent state where no requests 2289This module no longer tries to fight your OS, or POSIX. That means IO::AIO
1441can be added in other threads and no results will be processed. After 2290only works in the process that loaded it. Forking is fully supported, but
1442the fork the parent simply leaves the quiescent state and continues 2291using IO::AIO in the child is not.
1443request/result processing, while the child frees the request/result queue
1444(so that the requests started before the fork will only be handled in the
1445parent). Threads will be started on demand until the limit set in the
1446parent process has been reached again.
1447 2292
1448In short: the parent will, after a short pause, continue as if fork had 2293You might get around by not I<using> IO::AIO before (or after)
1449not been called, while the child will act as if IO::AIO has not been used 2294forking. You could also try to call the L<IO::AIO::reinit> function in the
1450yet. 2295child:
2296
2297=over 4
2298
2299=item IO::AIO::reinit
2300
2301Abandons all current requests and I/O threads and simply reinitialises all
2302data structures. This is not an operation supported by any standards, but
2303happens to work on GNU/Linux and some newer BSD systems.
2304
2305The only reasonable use for this function is to call it after forking, if
2306C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
2307the process will result in undefined behaviour. Calling it at any time
2308will also result in any undefined (by POSIX) behaviour.
2309
2310=back
1451 2311
1452=head2 MEMORY USAGE 2312=head2 MEMORY USAGE
1453 2313
1454Per-request usage: 2314Per-request usage:
1455 2315

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines