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 1.170 by root, Sat Jan 2 14:11:32 2010 UTC vs.
Revision 1.212 by root, Thu Sep 29 22:42:15 2011 UTC

… …
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use IO::AIO; 7 use IO::AIO;
8 8
9 aio_open "/etc/passwd", O_RDONLY, 0, sub { 9 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
10 my $fh = 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>
… …
99 76
100 # register the IO::AIO callback with EV 77 # register the IO::AIO callback with EV
101 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb; 78 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
102 79
103 # queue the request to open /etc/passwd 80 # queue the request to open /etc/passwd
104 aio_open "/etc/passwd", O_RDONLY, 0, sub { 81 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
105 my $fh = shift 82 my $fh = shift
106 or die "error while opening: $!"; 83 or die "error while opening: $!";
107 84
108 # stat'ing filehandles is generally non-blocking 85 # stat'ing filehandles is generally non-blocking
109 my $size = -s $fh; 86 my $size = -s $fh;
… …
191use common::sense; 168use common::sense;
192 169
193use base 'Exporter'; 170use base 'Exporter';
194 171
195BEGIN { 172BEGIN {
196 our $VERSION = '3.31'; 173 our $VERSION = '4.0';
197 174
198 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close 175 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close
199 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx 176 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
200 aio_scandir aio_symlink aio_readlink aio_sync aio_fsync 177 aio_scandir aio_symlink aio_readlink aio_realpath aio_sync
201 aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead 178 aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range aio_fallocate
179 aio_pathsync aio_readahead
202 aio_rename aio_link aio_move aio_copy aio_group 180 aio_rename aio_link aio_move aio_copy aio_group
203 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown 181 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
204 aio_chmod aio_utime aio_truncate 182 aio_chmod aio_utime aio_truncate
205 aio_msync aio_mtouch); 183 aio_msync aio_mtouch aio_mlock aio_mlockall
184 aio_statvfs
185 aio_wd);
206 186
207 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); 187 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
208 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 188 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
209 min_parallel max_parallel max_idle 189 min_parallel max_parallel max_idle idle_timeout
210 nreqs nready npending nthreads 190 nreqs nready npending nthreads
211 max_poll_time max_poll_reqs 191 max_poll_time max_poll_reqs
212 sendfile fadvise); 192 sendfile fadvise madvise
193 mmap munmap munlock munlockall);
213 194
214 push @AIO_REQ, qw(aio_busy); # not exported 195 push @AIO_REQ, qw(aio_busy); # not exported
215 196
216 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 197 @IO::AIO::GRP::ISA = 'IO::AIO::REQ';
217 198
… …
219 XSLoader::load ("IO::AIO", $VERSION); 200 XSLoader::load ("IO::AIO", $VERSION);
220} 201}
221 202
222=head1 FUNCTIONS 203=head1 FUNCTIONS
223 204
205=head2 QUICK OVERVIEW
206
207This section simply lists the prototypes of the most important functions
208for quick reference. See the following sections for function-by-function
209documentation.
210
211 aio_wd $pathname, $callback->($wd)
212 aio_open $pathname, $flags, $mode, $callback->($fh)
213 aio_close $fh, $callback->($status)
214 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
215 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
216 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
217 aio_readahead $fh,$offset,$length, $callback->($retval)
218 aio_stat $fh_or_path, $callback->($status)
219 aio_lstat $fh, $callback->($status)
220 aio_statvfs $fh_or_path, $callback->($statvfs)
221 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
222 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
223 aio_truncate $fh_or_path, $offset, $callback->($status)
224 aio_chmod $fh_or_path, $mode, $callback->($status)
225 aio_unlink $pathname, $callback->($status)
226 aio_mknod $pathname, $mode, $dev, $callback->($status)
227 aio_link $srcpath, $dstpath, $callback->($status)
228 aio_symlink $srcpath, $dstpath, $callback->($status)
229 aio_readlink $pathname, $callback->($link)
230 aio_realpath $pathname, $callback->($link)
231 aio_rename $srcpath, $dstpath, $callback->($status)
232 aio_mkdir $pathname, $mode, $callback->($status)
233 aio_rmdir $pathname, $callback->($status)
234 aio_readdir $pathname, $callback->($entries)
235 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
236 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
237 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
238 aio_load $pathname, $data, $callback->($status)
239 aio_copy $srcpath, $dstpath, $callback->($status)
240 aio_move $srcpath, $dstpath, $callback->($status)
241 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
242 aio_rmtree $pathname, $callback->($status)
243 aio_sync $callback->($status)
244 aio_syncfs $fh, $callback->($status)
245 aio_fsync $fh, $callback->($status)
246 aio_fdatasync $fh, $callback->($status)
247 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
248 aio_pathsync $pathname, $callback->($status)
249 aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
250 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
251 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
252 aio_mlockall $flags, $callback->($status)
253 aio_group $callback->(...)
254 aio_nop $callback->()
255
256 $prev_pri = aioreq_pri [$pri]
257 aioreq_nice $pri_adjust
258
259 IO::AIO::poll_wait
260 IO::AIO::poll_cb
261 IO::AIO::poll
262 IO::AIO::flush
263 IO::AIO::max_poll_reqs $nreqs
264 IO::AIO::max_poll_time $seconds
265 IO::AIO::min_parallel $nthreads
266 IO::AIO::max_parallel $nthreads
267 IO::AIO::max_idle $nthreads
268 IO::AIO::idle_timeout $seconds
269 IO::AIO::max_outstanding $maxreqs
270 IO::AIO::nreqs
271 IO::AIO::nready
272 IO::AIO::npending
273
274 IO::AIO::sendfile $ofh, $ifh, $offset, $count
275 IO::AIO::fadvise $fh, $offset, $len, $advice
276 IO::AIO::madvise $scalar, $offset, $length, $advice
277 IO::AIO::mprotect $scalar, $offset, $length, $protect
278 IO::AIO::munlock $scalar, $offset = 0, $length = undef
279 IO::AIO::munlockall
280
224=head2 AIO REQUEST FUNCTIONS 281=head2 AIO REQUEST FUNCTIONS
225 282
226All the C<aio_*> calls are more or less thin wrappers around the syscall 283All the C<aio_*> calls are more or less thin wrappers around the syscall
227with the same name (sans C<aio_>). The arguments are similar or identical, 284with the same name (sans C<aio_>). The arguments are similar or identical,
228and they all accept an additional (and optional) C<$callback> argument 285and they all accept an additional (and optional) C<$callback> argument
229which must be a code reference. This code reference will get called with 286which must be a code reference. This code reference will be called after
230the syscall return code (e.g. most syscalls return C<-1> on error, unlike 287the syscall has been executed in an asynchronous fashion. The results
231perl, which usually delivers "false") as its sole argument after the given 288of the request will be passed as arguments to the callback (and, if an
232syscall has been executed asynchronously. 289error occured, in C<$!>) - for most requests the syscall return code (e.g.
290most syscalls return C<-1> on error, unlike perl, which usually delivers
291"false").
292
293Some requests (such as C<aio_readdir>) pass the actual results and
294communicate failures by passing C<undef>.
233 295
234All functions expecting a filehandle keep a copy of the filehandle 296All functions expecting a filehandle keep a copy of the filehandle
235internally until the request has finished. 297internally until the request has finished.
236 298
237All functions return request objects of type L<IO::AIO::REQ> that allow 299All functions return request objects of type L<IO::AIO::REQ> that allow
238further manipulation of those requests while they are in-flight. 300further manipulation of those requests while they are in-flight.
239 301
240The pathnames you pass to these routines I<must> be absolute and 302The pathnames you pass to these routines I<should> be absolute. The
241encoded as octets. The reason for the former is that at the time the 303reason for this is that at the time the request is being executed, the
242request is being executed, the current working directory could have 304current working directory could have changed. Alternatively, you can
243changed. Alternatively, you can make sure that you never change the 305make sure that you never change the current working directory anywhere
244current working directory anywhere in the program and then use relative 306in the program and then use relative paths. You can also take advantage
245paths. 307of IO::AIOs working directory abstraction, that lets you specify paths
308relative to some previously-opened "working directory object" - see the
309description of the C<IO::AIO::WD> class later in this document.
246 310
247To encode pathnames as octets, either make sure you either: a) always pass 311To encode pathnames as octets, either make sure you either: a) always pass
248in filenames you got from outside (command line, readdir etc.) without 312in filenames you got from outside (command line, readdir etc.) without
249tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode 313tinkering, b) are in your native filesystem encoding, c) use the Encode
250your pathnames to the locale (or other) encoding in effect in the user 314module and encode your pathnames to the locale (or other) encoding in
251environment, d) use Glib::filename_from_unicode on unicode filenames or e) 315effect in the user environment, d) use Glib::filename_from_unicode on
252use something else to ensure your scalar has the correct contents. 316unicode filenames or e) use something else to ensure your scalar has the
317correct contents.
253 318
254This works, btw. independent of the internal UTF-8 bit, which IO::AIO 319This works, btw. independent of the internal UTF-8 bit, which IO::AIO
255handles correctly whether it is set or not. 320handles correctly whether it is set or not.
256 321
257=over 4 322=over 4
… …
307by the umask in effect then the request is being executed, so better never 372by the umask in effect then the request is being executed, so better never
308change the umask. 373change the umask.
309 374
310Example: 375Example:
311 376
312 aio_open "/etc/passwd", O_RDONLY, 0, sub { 377 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
313 if ($_[0]) { 378 if ($_[0]) {
314 print "open successful, fh is $_[0]\n"; 379 print "open successful, fh is $_[0]\n";
315 ... 380 ...
316 } else { 381 } else {
317 die "open failed: $!\n"; 382 die "open failed: $!\n";
318 } 383 }
319 }; 384 };
320 385
386In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
387C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
388following POSIX and non-POSIX constants are available (missing ones on
389your system are, as usual, C<0>):
390
391C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
392C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
393C<O_RSYNC>, C<O_SYNC> and C<O_TTY_INIT>.
394
321 395
322=item aio_close $fh, $callback->($status) 396=item aio_close $fh, $callback->($status)
323 397
324Asynchronously close a file and call the callback with the result 398Asynchronously close a file and call the callback with the result
325code. 399code.
… …
375 449
376Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 450Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
377reading at byte offset C<$in_offset>, and starts writing at the current 451reading at byte offset C<$in_offset>, and starts writing at the current
378file offset of C<$out_fh>. Because of that, it is not safe to issue more 452file offset of C<$out_fh>. Because of that, it is not safe to issue more
379than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each 453than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
380other. 454other. The same C<$in_fh> works fine though, as this function does not
455move or use the file offset of C<$in_fh>.
381 456
457Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
458are written, and there is no way to find out how many more bytes have been
459read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
460number of bytes written to C<$out_fh>. Only if the result value equals
461C<$length> one can assume that C<$length> bytes have been read.
462
463Unlike with other C<aio_> functions, it makes a lot of sense to use
464C<aio_sendfile> on non-blocking sockets, as long as one end (typically
465the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
466the socket I/O will be non-blocking. Note, however, that you can run
467into a trap where C<aio_sendfile> reads some data with readahead, then
468fails to write all data, and when the socket is ready the next time, the
469data in the cache is already lost, forcing C<aio_sendfile> to again hit
470the disk. Explicit C<aio_read> + C<aio_write> let's you better control
471resource usage.
472
382This call tries to make use of a native C<sendfile> syscall to provide 473This call tries to make use of a native C<sendfile>-like syscall to
383zero-copy operation. For this to work, C<$out_fh> should refer to a 474provide zero-copy operation. For this to work, C<$out_fh> should refer to
384socket, and C<$in_fh> should refer to mmap'able file. 475a socket, and C<$in_fh> should refer to an mmap'able file.
385 476
386If a native sendfile cannot be found or it fails with C<ENOSYS>, 477If a native sendfile cannot be found or it fails with C<ENOSYS>,
387C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or C<ENOTSOCK>, 478C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
388it will be emulated, so you can call C<aio_sendfile> on any type of 479C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
389filehandle regardless of the limitations of the operating system. 480type of filehandle regardless of the limitations of the operating system.
390 481
391Please note, however, that C<aio_sendfile> can read more bytes from 482As native sendfile syscalls (as practically any non-POSIX interface hacked
392C<$in_fh> than are written, and there is no way to find out how many 483together in a hurry to improve benchmark numbers) tend to be rather buggy
393bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only 484on many systems, this implementation tries to work around some known bugs
394provides the number of bytes written to C<$out_fh>. Only if the result 485in Linux and FreeBSD kernels (probably others, too), but that might fail,
395value equals C<$length> one can assume that C<$length> bytes have been 486so you really really should check the return value of C<aio_sendfile> -
396read. 487fewre bytes than expected might have been transferred.
397 488
398 489
399=item aio_readahead $fh,$offset,$length, $callback->($retval) 490=item aio_readahead $fh,$offset,$length, $callback->($retval)
400 491
401C<aio_readahead> populates the page cache with data from a file so that 492C<aio_readahead> populates the page cache with data from a file so that
… …
424 515
425Currently, the stats are always 64-bit-stats, i.e. instead of returning an 516Currently, the stats are always 64-bit-stats, i.e. instead of returning an
426error when stat'ing a large file, the results will be silently truncated 517error when stat'ing a large file, the results will be silently truncated
427unless perl itself is compiled with large file support. 518unless perl itself is compiled with large file support.
428 519
520To help interpret the mode and dev/rdev stat values, IO::AIO offers the
521following constants and functions (if not implemented, the constants will
522be C<0> and the functions will either C<croak> or fall back on traditional
523behaviour).
524
525C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
526C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
527C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
528
429Example: Print the length of F</etc/passwd>: 529Example: Print the length of F</etc/passwd>:
430 530
431 aio_stat "/etc/passwd", sub { 531 aio_stat "/etc/passwd", sub {
432 $_[0] and die "stat failed: $!"; 532 $_[0] and die "stat failed: $!";
433 print "size is ", -s _, "\n"; 533 print "size is ", -s _, "\n";
434 }; 534 };
435 535
436 536
537=item aio_statvfs $fh_or_path, $callback->($statvfs)
538
539Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
540whether a file handle or path was passed.
541
542On success, the callback is passed a hash reference with the following
543members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
544C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
545is passed.
546
547The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
548C<ST_NOSUID>.
549
550The following non-POSIX IO::AIO::ST_* flag masks are defined to
551their correct value when available, or to C<0> on systems that do
552not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
553C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
554C<ST_NODIRATIME> and C<ST_RELATIME>.
555
556Example: stat C</wd> and dump out the data if successful.
557
558 aio_statvfs "/wd", sub {
559 my $f = $_[0]
560 or die "statvfs: $!";
561
562 use Data::Dumper;
563 say Dumper $f;
564 };
565
566 # result:
567 {
568 bsize => 1024,
569 bfree => 4333064312,
570 blocks => 10253828096,
571 files => 2050765568,
572 flag => 4096,
573 favail => 2042092649,
574 bavail => 4333064312,
575 ffree => 2042092649,
576 namemax => 255,
577 frsize => 1024,
578 fsid => 1810
579 }
580
581
437=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 582=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
438 583
439Works like perl's C<utime> function (including the special case of $atime 584Works like perl's C<utime> function (including the special case of $atime
440and $mtime being undef). Fractional times are supported if the underlying 585and $mtime being undef). Fractional times are supported if the underlying
441syscalls support them. 586syscalls support them.
… …
479 624
480Asynchronously unlink (delete) a file and call the callback with the 625Asynchronously unlink (delete) a file and call the callback with the
481result code. 626result code.
482 627
483 628
484=item aio_mknod $path, $mode, $dev, $callback->($status) 629=item aio_mknod $pathname, $mode, $dev, $callback->($status)
485 630
486[EXPERIMENTAL] 631[EXPERIMENTAL]
487 632
488Asynchronously create a device node (or fifo). See mknod(2). 633Asynchronously create a device node (or fifo). See mknod(2).
489 634
490The only (POSIX-) portable way of calling this function is: 635The only (POSIX-) portable way of calling this function is:
491 636
492 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 637 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
493 638
639See C<aio_stat> for info about some potentially helpful extra constants
640and functions.
494 641
495=item aio_link $srcpath, $dstpath, $callback->($status) 642=item aio_link $srcpath, $dstpath, $callback->($status)
496 643
497Asynchronously create a new link to the existing object at C<$srcpath> at 644Asynchronously create a new link to the existing object at C<$srcpath> at
498the path C<$dstpath> and call the callback with the result code. 645the path C<$dstpath> and call the callback with the result code.
… …
502 649
503Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 650Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
504the path C<$dstpath> and call the callback with the result code. 651the path C<$dstpath> and call the callback with the result code.
505 652
506 653
507=item aio_readlink $path, $callback->($link) 654=item aio_readlink $pathname, $callback->($link)
508 655
509Asynchronously read the symlink specified by C<$path> and pass it to 656Asynchronously read the symlink specified by C<$path> and pass it to
510the callback. If an error occurs, nothing or undef gets passed to the 657the callback. If an error occurs, nothing or undef gets passed to the
511callback. 658callback.
512 659
513 660
661=item aio_realpath $pathname, $callback->($path)
662
663Asynchronously make the path absolute and resolve any symlinks in
664C<$path>. The resulting path only consists of directories (Same as
665L<Cwd::realpath>).
666
667This request can be used to get the absolute path of the current working
668directory by passing it a path of F<.> (a single dot).
669
670
514=item aio_rename $srcpath, $dstpath, $callback->($status) 671=item aio_rename $srcpath, $dstpath, $callback->($status)
515 672
516Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 673Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
517rename(2) and call the callback with the result code. 674rename(2) and call the callback with the result code.
518 675
… …
540array-ref with the filenames. 697array-ref with the filenames.
541 698
542 699
543=item aio_readdirx $pathname, $flags, $callback->($entries, $flags) 700=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
544 701
545Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune 702Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
546behaviour and output format. In case of an error, C<$entries> will be 703tune behaviour and output format. In case of an error, C<$entries> will be
547C<undef>. 704C<undef>.
548 705
549The flags are a combination of the following constants, ORed together (the 706The flags are a combination of the following constants, ORed together (the
550flags will also be passed to the callback, possibly modified): 707flags will also be passed to the callback, possibly modified):
551 708
552=over 4 709=over 4
553 710
554=item IO::AIO::READDIR_DENTS 711=item IO::AIO::READDIR_DENTS
555 712
556When this flag is off, then the callback gets an arrayref with of names 713When this flag is off, then the callback gets an arrayref consisting of
557only (as with C<aio_readdir>), otherwise it gets an arrayref with 714names only (as with C<aio_readdir>), otherwise it gets an arrayref with
558C<[$name, $type, $inode]> arrayrefs, each describing a single directory 715C<[$name, $type, $inode]> arrayrefs, each describing a single directory
559entry in more detail. 716entry in more detail.
560 717
561C<$name> is the name of the entry. 718C<$name> is the name of the entry.
562 719
… …
575systems that do not deliver the inode information. 732systems that do not deliver the inode information.
576 733
577=item IO::AIO::READDIR_DIRS_FIRST 734=item IO::AIO::READDIR_DIRS_FIRST
578 735
579When this flag is set, then the names will be returned in an order where 736When this flag is set, then the names will be returned in an order where
580likely directories come first. This is useful when you need to quickly 737likely directories come first, in optimal stat order. This is useful when
581find directories, or you want to find all directories while avoiding to 738you need to quickly find directories, or you want to find all directories
582stat() each entry. 739while avoiding to stat() each entry.
583 740
584If the system returns type information in readdir, then this is used 741If the system returns type information in readdir, then this is used
585to find directories directly. Otherwise, likely directories are files 742to find directories directly. Otherwise, likely directories are names
586beginning with ".", or otherwise files with no dots, of which files with 743beginning with ".", or otherwise names with no dots, of which names with
587short names are tried first. 744short names are tried first.
588 745
589=item IO::AIO::READDIR_STAT_ORDER 746=item IO::AIO::READDIR_STAT_ORDER
590 747
591When this flag is set, then the names will be returned in an order 748When this flag is set, then the names will be returned in an order
… …
598 755
599=item IO::AIO::READDIR_FOUND_UNKNOWN 756=item IO::AIO::READDIR_FOUND_UNKNOWN
600 757
601This flag should not be set when calling C<aio_readdirx>. Instead, it 758This flag should not be set when calling C<aio_readdirx>. Instead, it
602is being set by C<aio_readdirx>, when any of the C<$type>'s found were 759is being set by C<aio_readdirx>, when any of the C<$type>'s found were
603C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all 760C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
604C<$type>'s are known, which can be used to speed up some algorithms. 761C<$type>'s are known, which can be used to speed up some algorithms.
605 762
606=back 763=back
607 764
608 765
609=item aio_load $path, $data, $callback->($status) 766=item aio_load $pathname, $data, $callback->($status)
610 767
611This is a composite request that tries to fully load the given file into 768This is a composite request that tries to fully load the given file into
612memory. Status is the same as with aio_read. 769memory. Status is the same as with aio_read.
613 770
614=cut 771=cut
… …
736 if ($_[0] && $! == EXDEV) { 893 if ($_[0] && $! == EXDEV) {
737 aioreq_pri $pri; 894 aioreq_pri $pri;
738 add $grp aio_copy $src, $dst, sub { 895 add $grp aio_copy $src, $dst, sub {
739 $grp->result ($_[0]); 896 $grp->result ($_[0]);
740 897
741 if (!$_[0]) { 898 unless ($_[0]) {
742 aioreq_pri $pri; 899 aioreq_pri $pri;
743 add $grp aio_unlink $src; 900 add $grp aio_unlink $src;
744 } 901 }
745 }; 902 };
746 } else { 903 } else {
… …
749 }; 906 };
750 907
751 $grp 908 $grp
752} 909}
753 910
754=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 911=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
755 912
756Scans a directory (similar to C<aio_readdir>) but additionally tries to 913Scans a directory (similar to C<aio_readdir>) but additionally tries to
757efficiently separate the entries of directory C<$path> into two sets of 914efficiently separate the entries of directory C<$path> into two sets of
758names, directories you can recurse into (directories), and ones you cannot 915names, directories you can recurse into (directories), and ones you cannot
759recurse into (everything else, including symlinks to directories). 916recurse into (everything else, including symlinks to directories).
… …
790Then entries will be sorted into likely directories a non-initial dot 947Then entries will be sorted into likely directories a non-initial dot
791currently) and likely non-directories (see C<aio_readdirx>). Then every 948currently) and likely non-directories (see C<aio_readdirx>). Then every
792entry plus an appended C</.> will be C<stat>'ed, likely directories first, 949entry plus an appended C</.> will be C<stat>'ed, likely directories first,
793in order of their inode numbers. If that succeeds, it assumes that the 950in order of their inode numbers. If that succeeds, it assumes that the
794entry is a directory or a symlink to directory (which will be checked 951entry is a directory or a symlink to directory (which will be checked
795seperately). This is often faster than stat'ing the entry itself because 952separately). This is often faster than stat'ing the entry itself because
796filesystems might detect the type of the entry without reading the inode 953filesystems might detect the type of the entry without reading the inode
797data (e.g. ext2fs filetype feature), even on systems that cannot return 954data (e.g. ext2fs filetype feature), even on systems that cannot return
798the filetype information on readdir. 955the filetype information on readdir.
799 956
800If the known number of directories (link count - 2) has been reached, the 957If the known number of directories (link count - 2) has been reached, the
… …
816 973
817 my $grp = aio_group $cb; 974 my $grp = aio_group $cb;
818 975
819 $maxreq = 4 if $maxreq <= 0; 976 $maxreq = 4 if $maxreq <= 0;
820 977
821 # stat once 978 # get a wd object
822 aioreq_pri $pri; 979 aioreq_pri $pri;
823 add $grp aio_stat $path, sub { 980 add $grp aio_wd $path, sub {
981 $_[0]
824 return $grp->result () if $_[0]; 982 or return $grp->result ();
825 my $now = time;
826 my $hash1 = join ":", (stat _)[0,1,3,7,9];
827 983
828 # read the directory entries 984 my $wd = [shift, "."];
985
986 # stat once
829 aioreq_pri $pri; 987 aioreq_pri $pri;
830 add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub { 988 add $grp aio_stat $wd, sub {
831 my $entries = shift
832 or return $grp->result (); 989 return $grp->result () if $_[0];
990 my $now = time;
991 my $hash1 = join ":", (stat _)[0,1,3,7,9];
833 992
834 # stat the dir another time 993 # read the directory entries
835 aioreq_pri $pri; 994 aioreq_pri $pri;
995 add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
996 my $entries = shift
997 or return $grp->result ();
998
999 # stat the dir another time
1000 aioreq_pri $pri;
836 add $grp aio_stat $path, sub { 1001 add $grp aio_stat $wd, sub {
837 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 1002 my $hash2 = join ":", (stat _)[0,1,3,7,9];
838 1003
839 my $ndirs; 1004 my $ndirs;
840 1005
841 # take the slow route if anything looks fishy 1006 # take the slow route if anything looks fishy
842 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 1007 if ($hash1 ne $hash2 or (stat _)[9] == $now) {
843 $ndirs = -1; 1008 $ndirs = -1;
844 } else { 1009 } else {
845 # if nlink == 2, we are finished 1010 # if nlink == 2, we are finished
846 # for non-posix-fs's, we rely on nlink < 2 1011 # for non-posix-fs's, we rely on nlink < 2
847 $ndirs = (stat _)[3] - 2 1012 $ndirs = (stat _)[3] - 2
848 or return $grp->result ([], $entries); 1013 or return $grp->result ([], $entries);
849 } 1014 }
850 1015
851 my (@dirs, @nondirs); 1016 my (@dirs, @nondirs);
852 1017
853 my $statgrp = add $grp aio_group sub { 1018 my $statgrp = add $grp aio_group sub {
854 $grp->result (\@dirs, \@nondirs); 1019 $grp->result (\@dirs, \@nondirs);
855 }; 1020 };
856 1021
857 limit $statgrp $maxreq; 1022 limit $statgrp $maxreq;
858 feed $statgrp sub { 1023 feed $statgrp sub {
859 return unless @$entries; 1024 return unless @$entries;
860 my $entry = shift @$entries; 1025 my $entry = shift @$entries;
861 1026
862 aioreq_pri $pri; 1027 aioreq_pri $pri;
1028 $wd->[1] = "$entry/.";
863 add $statgrp aio_stat "$path/$entry/.", sub { 1029 add $statgrp aio_stat $wd, sub {
864 if ($_[0] < 0) { 1030 if ($_[0] < 0) {
865 push @nondirs, $entry; 1031 push @nondirs, $entry;
866 } else { 1032 } else {
867 # need to check for real directory 1033 # need to check for real directory
868 aioreq_pri $pri; 1034 aioreq_pri $pri;
1035 $wd->[1] = $entry;
869 add $statgrp aio_lstat "$path/$entry", sub { 1036 add $statgrp aio_lstat $wd, sub {
870 if (-d _) { 1037 if (-d _) {
871 push @dirs, $entry; 1038 push @dirs, $entry;
872 1039
873 unless (--$ndirs) { 1040 unless (--$ndirs) {
874 push @nondirs, @$entries; 1041 push @nondirs, @$entries;
875 feed $statgrp; 1042 feed $statgrp;
1043 }
1044 } else {
1045 push @nondirs, $entry;
876 } 1046 }
877 } else {
878 push @nondirs, $entry;
879 } 1047 }
880 } 1048 }
881 } 1049 };
882 }; 1050 };
883 }; 1051 };
884 }; 1052 };
885 }; 1053 };
886 }; 1054 };
887 1055
888 $grp 1056 $grp
889} 1057}
890 1058
891=item aio_rmtree $path, $callback->($status) 1059=item aio_rmtree $pathname, $callback->($status)
892 1060
893Delete a directory tree starting (and including) C<$path>, return the 1061Delete a directory tree starting (and including) C<$path>, return the
894status of the final C<rmdir> only. This is a composite request that 1062status of the final C<rmdir> only. This is a composite request that
895uses C<aio_scandir> to recurse into and rmdir directories, and unlink 1063uses C<aio_scandir> to recurse into and rmdir directories, and unlink
896everything else. 1064everything else.
… …
938callback with the fdatasync result code. 1106callback with the fdatasync result code.
939 1107
940If this call isn't available because your OS lacks it or it couldn't be 1108If this call isn't available because your OS lacks it or it couldn't be
941detected, it will be emulated by calling C<fsync> instead. 1109detected, it will be emulated by calling C<fsync> instead.
942 1110
1111=item aio_syncfs $fh, $callback->($status)
1112
1113Asynchronously call the syncfs syscall to sync the filesystem associated
1114to the given filehandle and call the callback with the syncfs result
1115code. If syncfs is not available, calls sync(), but returns C<-1> and sets
1116errno to C<ENOSYS> nevertheless.
1117
943=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 1118=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
944 1119
945Sync the data portion of the file specified by C<$offset> and C<$length> 1120Sync the data portion of the file specified by C<$offset> and C<$length>
946to disk (but NOT the metadata), by calling the Linux-specific 1121to disk (but NOT the metadata), by calling the Linux-specific
947sync_file_range call. If sync_file_range is not available or it returns 1122sync_file_range call. If sync_file_range is not available or it returns
… …
950C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>, 1125C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
951C<IO::AIO::SYNC_FILE_RANGE_WRITE> and 1126C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
952C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range 1127C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
953manpage for details. 1128manpage for details.
954 1129
955=item aio_pathsync $path, $callback->($status) 1130=item aio_pathsync $pathname, $callback->($status)
956 1131
957This request tries to open, fsync and close the given path. This is a 1132This request tries to open, fsync and close the given path. This is a
958composite request intended to sync directories after directory operations 1133composite request intended to sync directories after directory operations
959(E.g. rename). This might not work on all operating systems or have any 1134(E.g. rename). This might not work on all operating systems or have any
960specific effect, but usually it makes sure that directory changes get 1135specific effect, but usually it makes sure that directory changes get
… …
994} 1169}
995 1170
996=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 1171=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
997 1172
998This is a rather advanced IO::AIO call, which only works on mmap(2)ed 1173This is a rather advanced IO::AIO call, which only works on mmap(2)ed
999scalars (see the L<Sys::Mmap> or L<Mmap> modules for details on this, note 1174scalars (see the C<IO::AIO::mmap> function, although it also works on data
1175scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
1000that the scalar must only be modified in-place while an aio operation is 1176scalar must only be modified in-place while an aio operation is pending on
1001pending on it). 1177it).
1002 1178
1003It calls the C<msync> function of your OS, if available, with the memory 1179It calls the C<msync> function of your OS, if available, with the memory
1004area starting at C<$offset> in the string and ending C<$length> bytes 1180area starting at C<$offset> in the string and ending C<$length> bytes
1005later. If C<$length> is negative, counts from the end, and if C<$length> 1181later. If C<$length> is negative, counts from the end, and if C<$length>
1006is C<undef>, then it goes till the end of the string. The flags can be 1182is C<undef>, then it goes till the end of the string. The flags can be
… …
1017as for C<aio_msync>, above, except for flags, which must be either 1193as for C<aio_msync>, above, except for flags, which must be either
1018C<0> (which reads all pages and ensures they are instantiated) or 1194C<0> (which reads all pages and ensures they are instantiated) or
1019C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and 1195C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
1020writing an octet from it, which dirties the page). 1196writing an octet from it, which dirties the page).
1021 1197
1198=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1199
1200This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1201scalars.
1202
1203It reads in all the pages of the underlying storage into memory (if any)
1204and locks them, so they are not getting swapped/paged out or removed.
1205
1206If C<$length> is undefined, then the scalar will be locked till the end.
1207
1208On systems that do not implement C<mlock>, this function returns C<-1>
1209and sets errno to C<ENOSYS>.
1210
1211Note that the corresponding C<munlock> is synchronous and is
1212documented under L<MISCELLANEOUS FUNCTIONS>.
1213
1214Example: open a file, mmap and mlock it - both will be undone when
1215C<$data> gets destroyed.
1216
1217 open my $fh, "<", $path or die "$path: $!";
1218 my $data;
1219 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1220 aio_mlock $data; # mlock in background
1221
1222=item aio_mlockall $flags, $callback->($status)
1223
1224Calls the C<mlockall> function with the given C<$flags> (a combination of
1225C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
1226
1227On systems that do not implement C<mlockall>, this function returns C<-1>
1228and sets errno to C<ENOSYS>.
1229
1230Note that the corresponding C<munlockall> is synchronous and is
1231documented under L<MISCELLANEOUS FUNCTIONS>.
1232
1233Example: asynchronously lock all current and future pages into memory.
1234
1235 aio_mlockall IO::AIO::MCL_FUTURE;
1236
1022=item aio_group $callback->(...) 1237=item aio_group $callback->(...)
1023 1238
1024This is a very special aio request: Instead of doing something, it is a 1239This is a very special aio request: Instead of doing something, it is a
1025container for other aio requests, which is useful if you want to bundle 1240container for other aio requests, which is useful if you want to bundle
1026many requests into a single, composite, request with a definite callback 1241many requests into a single, composite, request with a definite callback
… …
1062like sleep and file handle readable/writable, the overhead this creates is 1277like sleep and file handle readable/writable, the overhead this creates is
1063immense (it blocks a thread for a long time) so do not use this function 1278immense (it blocks a thread for a long time) so do not use this function
1064except to put your application under artificial I/O pressure. 1279except to put your application under artificial I/O pressure.
1065 1280
1066=back 1281=back
1282
1283
1284=head2 IO::AIO::WD - multiple working directories
1285
1286Your process only has one current working directory, which is used by all
1287threads. This makes it hard to use relative paths (some other component
1288could call C<chdir> at any time, and it is hard to control when the path
1289will be used by IO::AIO).
1290
1291One solution for this is to always use absolute paths. This usually works,
1292but can be quite slow (the kernel has to walk the whole path on every
1293access), and can also be a hassle to implement.
1294
1295Newer POSIX systems have a number of functions (openat, fdopendir,
1296futimensat and so on) that make it possible to specify working directories
1297per operation.
1298
1299For portability, and because the clowns who "designed", or shall I write,
1300perpetrated this new interface were obviously half-drunk, this abstraction
1301cannot be perfect, though.
1302
1303IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
1304object. This object stores the canonicalised, absolute version of the
1305path, and on systems that allow it, also a directory file descriptor.
1306
1307Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
1308or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
1309object and a pathname instead. If the pathname is absolute, the
1310IO::AIO::WD objetc is ignored, otherwise the pathname is resolved relative
1311to that IO::AIO::WD object.
1312
1313For example, to get a wd object for F</etc> and then stat F<passwd>
1314inside, you would write:
1315
1316 aio_wd "/etc", sub {
1317 my $etcdir = shift;
1318
1319 # although $etcdir can be undef on error, there is generally no reason
1320 # to check for errors here, as aio_stat will fail with ENOENT
1321 # when $etcdir is undef.
1322
1323 aio_stat [$etcdir, "passwd"], sub {
1324 # yay
1325 };
1326 };
1327
1328This shows that creating an IO::AIO::WD object is itself a potentially
1329blocking operation, which is why it is done asynchronously.
1330
1331As with normal pathnames, IO::AIO keeps a copy of the working directory
1332object and the pathname string, so you could write the following without
1333causing any issues due to C<$path> getting reused:
1334
1335 my $path = [$wd, undef];
1336
1337 for my $name (qw(abc def ghi)) {
1338 $path->[1] = $name;
1339 aio_stat $path, sub {
1340 # ...
1341 };
1342 }
1343
1344There are some caveats: when directories get renamed (or deleted), the
1345pathname string doesn't change, so will point to the new directory (or
1346nowhere at all), while the directory fd, if available on the system,
1347will still point to the original directory. Most functions accepting a
1348pathname will use the directory fd on newer systems, and the string on
1349older systems. Some functions (such as realpath) will always rely on the
1350string form of the pathname.
1351
1352So this fucntionality is mainly useful to get some protection against
1353C<chdir>, to easily get an absolute path out of a relative path for future
1354reference, and to speed up doing many operations in the same directory
1355(e.g. when stat'ing all files in a directory).
1356
1357The following functions implement this working directory abstraction:
1358
1359=over 4
1360
1361=item aio_wd $pathname, $callback->($wd)
1362
1363Asynchonously canonicalise the given pathname and convert it to an
1364IO::AIO::WD object representing it. If possible and supported on the
1365system, also open a directory fd to speed up pathname resolution relative
1366to this working directory.
1367
1368If something goes wrong, then C<undef> is passwd to the callback instead
1369of a working directory object and C<$!> is set appropriately. Since
1370passing C<undef> as working directory component of a pathname fails the
1371request with C<ENOENT>, there is often no need for error checking in the
1372C<aio_wd> callback, as future requests using the value will fail in the
1373expected way.
1374
1375If this call isn't available because your OS lacks it or it couldn't be
1376detected, it will be emulated by calling C<fsync> instead.
1377
1378=item IO::AIO::CWD
1379
1380This is a compiletime constant (object) that represents the process
1381current working directory.
1382
1383Specifying this object as working directory object for a pathname is as
1384if the pathname would be specified directly, without a directory object,
1385e.g., these calls are functionally identical:
1386
1387 aio_stat "somefile", sub { ... };
1388 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1389
1390=back
1391
1067 1392
1068=head2 IO::AIO::REQ CLASS 1393=head2 IO::AIO::REQ CLASS
1069 1394
1070All non-aggregate C<aio_*> functions return an object of this class when 1395All non-aggregate C<aio_*> functions return an object of this class when
1071called in non-void context. 1396called in non-void context.
… …
1189 1514
1190Sets a feeder/generator on this group: every group can have an attached 1515Sets a feeder/generator on this group: every group can have an attached
1191generator that generates requests if idle. The idea behind this is that, 1516generator that generates requests if idle. The idea behind this is that,
1192although you could just queue as many requests as you want in a group, 1517although you could just queue as many requests as you want in a group,
1193this might starve other requests for a potentially long time. For example, 1518this might starve other requests for a potentially long time. For example,
1194C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests, 1519C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
1195delaying any later requests for a long time. 1520requests, delaying any later requests for a long time.
1196 1521
1197To avoid this, and allow incremental generation of requests, you can 1522To avoid this, and allow incremental generation of requests, you can
1198instead a group and set a feeder on it that generates those requests. The 1523instead a group and set a feeder on it that generates those requests. The
1199feed callback will be called whenever there are few enough (see C<limit>, 1524feed callback will be called whenever there are few enough (see C<limit>,
1200below) requests active in the group itself and is expected to queue more 1525below) requests active in the group itself and is expected to queue more
… …
1249 1574
1250See C<poll_cb> for an example. 1575See C<poll_cb> for an example.
1251 1576
1252=item IO::AIO::poll_cb 1577=item IO::AIO::poll_cb
1253 1578
1254Process some outstanding events on the result pipe. You have to call this 1579Process some outstanding events on the result pipe. You have to call
1255regularly. Returns C<0> if all events could be processed, or C<-1> if it 1580this regularly. Returns C<0> if all events could be processed (or there
1256returned earlier for whatever reason. Returns immediately when no events 1581were no events to process), or C<-1> if it returned earlier for whatever
1257are outstanding. The amount of events processed depends on the settings of 1582reason. Returns immediately when no events are outstanding. The amount of
1258C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>. 1583events processed depends on the settings of C<IO::AIO::max_poll_req> and
1584C<IO::AIO::max_poll_time>.
1259 1585
1260If not all requests were processed for whatever reason, the filehandle 1586If not all requests were processed for whatever reason, the filehandle
1261will still be ready when C<poll_cb> returns, so normally you don't have to 1587will still be ready when C<poll_cb> returns, so normally you don't have to
1262do anything special to have it called later. 1588do anything special to have it called later.
1263 1589
1590Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
1591ready, it can be beneficial to call this function from loops which submit
1592a lot of requests, to make sure the results get processed when they become
1593available and not just when the loop is finished and the event loop takes
1594over again. This function returns very fast when there are no outstanding
1595requests.
1596
1264Example: Install an Event watcher that automatically calls 1597Example: Install an Event watcher that automatically calls
1265IO::AIO::poll_cb with high priority (more examples can be found in the 1598IO::AIO::poll_cb with high priority (more examples can be found in the
1266SYNOPSIS section, at the top of this document): 1599SYNOPSIS section, at the top of this document):
1267 1600
1268 Event->io (fd => IO::AIO::poll_fileno, 1601 Event->io (fd => IO::AIO::poll_fileno,
1269 poll => 'r', async => 1, 1602 poll => 'r', async => 1,
1270 cb => \&IO::AIO::poll_cb); 1603 cb => \&IO::AIO::poll_cb);
1604
1605=item IO::AIO::poll_wait
1606
1607If there are any outstanding requests and none of them in the result
1608phase, wait till the result filehandle becomes ready for reading (simply
1609does a C<select> on the filehandle. This is useful if you want to
1610synchronously wait for some requests to finish).
1611
1612See C<nreqs> for an example.
1613
1614=item IO::AIO::poll
1615
1616Waits until some requests have been handled.
1617
1618Returns the number of requests processed, but is otherwise strictly
1619equivalent to:
1620
1621 IO::AIO::poll_wait, IO::AIO::poll_cb
1622
1623=item IO::AIO::flush
1624
1625Wait till all outstanding AIO requests have been handled.
1626
1627Strictly equivalent to:
1628
1629 IO::AIO::poll_wait, IO::AIO::poll_cb
1630 while IO::AIO::nreqs;
1271 1631
1272=item IO::AIO::max_poll_reqs $nreqs 1632=item IO::AIO::max_poll_reqs $nreqs
1273 1633
1274=item IO::AIO::max_poll_time $seconds 1634=item IO::AIO::max_poll_time $seconds
1275 1635
… …
1300 # use a low priority so other tasks have priority 1660 # use a low priority so other tasks have priority
1301 Event->io (fd => IO::AIO::poll_fileno, 1661 Event->io (fd => IO::AIO::poll_fileno,
1302 poll => 'r', nice => 1, 1662 poll => 'r', nice => 1,
1303 cb => &IO::AIO::poll_cb); 1663 cb => &IO::AIO::poll_cb);
1304 1664
1305=item IO::AIO::poll_wait
1306
1307If there are any outstanding requests and none of them in the result
1308phase, wait till the result filehandle becomes ready for reading (simply
1309does a C<select> on the filehandle. This is useful if you want to
1310synchronously wait for some requests to finish).
1311
1312See C<nreqs> for an example.
1313
1314=item IO::AIO::poll
1315
1316Waits until some requests have been handled.
1317
1318Returns the number of requests processed, but is otherwise strictly
1319equivalent to:
1320
1321 IO::AIO::poll_wait, IO::AIO::poll_cb
1322
1323=item IO::AIO::flush
1324
1325Wait till all outstanding AIO requests have been handled.
1326
1327Strictly equivalent to:
1328
1329 IO::AIO::poll_wait, IO::AIO::poll_cb
1330 while IO::AIO::nreqs;
1331
1332=back 1665=back
1333 1666
1334=head3 CONTROLLING THE NUMBER OF THREADS 1667=head3 CONTROLLING THE NUMBER OF THREADS
1335 1668
1336=over 1669=over
… …
1369 1702
1370Under normal circumstances you don't need to call this function. 1703Under normal circumstances you don't need to call this function.
1371 1704
1372=item IO::AIO::max_idle $nthreads 1705=item IO::AIO::max_idle $nthreads
1373 1706
1374Limit the number of threads (default: 4) that are allowed to idle (i.e., 1707Limit the number of threads (default: 4) that are allowed to idle
1375threads that did not get a request to process within 10 seconds). That 1708(i.e., threads that did not get a request to process within the idle
1376means if a thread becomes idle while C<$nthreads> other threads are also 1709timeout (default: 10 seconds). That means if a thread becomes idle while
1377idle, it will free its resources and exit. 1710C<$nthreads> other threads are also idle, it will free its resources and
1711exit.
1378 1712
1379This is useful when you allow a large number of threads (e.g. 100 or 1000) 1713This is useful when you allow a large number of threads (e.g. 100 or 1000)
1380to allow for extremely high load situations, but want to free resources 1714to allow for extremely high load situations, but want to free resources
1381under normal circumstances (1000 threads can easily consume 30MB of RAM). 1715under normal circumstances (1000 threads can easily consume 30MB of RAM).
1382 1716
1383The default is probably ok in most situations, especially if thread 1717The default is probably ok in most situations, especially if thread
1384creation is fast. If thread creation is very slow on your system you might 1718creation is fast. If thread creation is very slow on your system you might
1385want to use larger values. 1719want to use larger values.
1386 1720
1721=item IO::AIO::idle_timeout $seconds
1722
1723Sets the minimum idle timeout (default 10) after which worker threads are
1724allowed to exit. SEe C<IO::AIO::max_idle>.
1725
1387=item IO::AIO::max_outstanding $maxreqs 1726=item IO::AIO::max_outstanding $maxreqs
1727
1728Sets the maximum number of outstanding requests to C<$nreqs>. If
1729you do queue up more than this number of requests, the next call to
1730C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
1731C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
1732longer exceeded.
1733
1734In other words, this setting does not enforce a queue limit, but can be
1735used to make poll functions block if the limit is exceeded.
1388 1736
1389This is a very bad function to use in interactive programs because it 1737This is a very bad function to use in interactive programs because it
1390blocks, and a bad way to reduce concurrency because it is inexact: Better 1738blocks, and a bad way to reduce concurrency because it is inexact: Better
1391use an C<aio_group> together with a feed callback. 1739use an C<aio_group> together with a feed callback.
1392 1740
1393Sets the maximum number of outstanding requests to C<$nreqs>. If you 1741It's main use is in scripts without an event loop - when you want to stat
1394do queue up more than this number of requests, the next call to the 1742a lot of files, you can write somehting like this:
1395C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
1396function will block until the limit is no longer exceeded.
1397 1743
1398The default value is very large, so there is no practical limit on the 1744 IO::AIO::max_outstanding 32;
1399number of outstanding requests.
1400 1745
1401You can still queue as many requests as you want. Therefore, 1746 for my $path (...) {
1402C<max_outstanding> is mainly useful in simple scripts (with low values) or 1747 aio_stat $path , ...;
1403as a stop gap to shield against fatal memory overflow (with large values). 1748 IO::AIO::poll_cb;
1749 }
1750
1751 IO::AIO::flush;
1752
1753The call to C<poll_cb> inside the loop will normally return instantly, but
1754as soon as more thna C<32> reqeusts are in-flight, it will block until
1755some requests have been handled. This keeps the loop from pushing a large
1756number of C<aio_stat> requests onto the queue.
1757
1758The default value for C<max_outstanding> is very large, so there is no
1759practical limit on the number of outstanding requests.
1404 1760
1405=back 1761=back
1406 1762
1407=head3 STATISTICAL INFORMATION 1763=head3 STATISTICAL INFORMATION
1408 1764
… …
1446 1802
1447Returns the number of bytes copied, or C<-1> on error. 1803Returns the number of bytes copied, or C<-1> on error.
1448 1804
1449=item IO::AIO::fadvise $fh, $offset, $len, $advice 1805=item IO::AIO::fadvise $fh, $offset, $len, $advice
1450 1806
1451Simply calls the C<posix_fadvise> function (see it's 1807Simply calls the C<posix_fadvise> function (see its
1452manpage for details). The following advice constants are 1808manpage for details). The following advice constants are
1453avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>, 1809available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
1454C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>, 1810C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
1455C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>. 1811C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
1456 1812
1457On systems that do not implement C<posix_fadvise>, this function returns 1813On systems that do not implement C<posix_fadvise>, this function returns
1458ENOSYS, otherwise the return value of C<posix_fadvise>. 1814ENOSYS, otherwise the return value of C<posix_fadvise>.
1459 1815
1816=item IO::AIO::madvise $scalar, $offset, $len, $advice
1817
1818Simply calls the C<posix_madvise> function (see its
1819manpage for details). The following advice constants are
1820available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
1821C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
1822
1823On systems that do not implement C<posix_madvise>, this function returns
1824ENOSYS, otherwise the return value of C<posix_madvise>.
1825
1826=item IO::AIO::mprotect $scalar, $offset, $len, $protect
1827
1828Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
1829$scalar (see its manpage for details). The following protect
1830constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
1831C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
1832
1833On systems that do not implement C<mprotect>, this function returns
1834ENOSYS, otherwise the return value of C<mprotect>.
1835
1836=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1837
1838Memory-maps a file (or anonymous memory range) and attaches it to the
1839given C<$scalar>, which will act like a string scalar.
1840
1841The only operations allowed on the scalar are C<substr>/C<vec> that don't
1842change the string length, and most read-only operations such as copying it
1843or searching it with regexes and so on.
1844
1845Anything else is unsafe and will, at best, result in memory leaks.
1846
1847The memory map associated with the C<$scalar> is automatically removed
1848when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
1849C<IO::AIO::munmap> functions are called.
1850
1851This calls the C<mmap>(2) function internally. See your system's manual
1852page for details on the C<$length>, C<$prot> and C<$flags> parameters.
1853
1854The C<$length> must be larger than zero and smaller than the actual
1855filesize.
1856
1857C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
1858C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
1859
1860C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
1861C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
1862not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
1863(which is set to C<MAP_ANON> if your system only provides this
1864constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
1865C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
1866C<IO::AIO::MAP_NONBLOCK>
1867
1868If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
1869
1870C<$offset> is the offset from the start of the file - it generally must be
1871a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
1872
1873Example:
1874
1875 use Digest::MD5;
1876 use IO::AIO;
1877
1878 open my $fh, "<verybigfile"
1879 or die "$!";
1880
1881 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1882 or die "verybigfile: $!";
1883
1884 my $fast_md5 = md5 $data;
1885
1886=item IO::AIO::munmap $scalar
1887
1888Removes a previous mmap and undefines the C<$scalar>.
1889
1890=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
1891
1892Calls the C<munlock> function, undoing the effects of a previous
1893C<aio_mlock> call (see its description for details).
1894
1895=item IO::AIO::munlockall
1896
1897Calls the C<munlockall> function.
1898
1899On systems that do not implement C<munlockall>, this function returns
1900ENOSYS, otherwise the return value of C<munlockall>.
1901
1460=back 1902=back
1461 1903
1462=cut 1904=cut
1463 1905
1464min_parallel 8; 1906min_parallel 8;
1465 1907
1466END { flush } 1908END { flush }
1467 1909
14681; 19101;
1469 1911
1912=head1 EVENT LOOP INTEGRATION
1913
1914It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
1915automatically into many event loops:
1916
1917 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
1918 use AnyEvent::AIO;
1919
1920You can also integrate IO::AIO manually into many event loops, here are
1921some examples of how to do this:
1922
1923 # EV integration
1924 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
1925
1926 # Event integration
1927 Event->io (fd => IO::AIO::poll_fileno,
1928 poll => 'r',
1929 cb => \&IO::AIO::poll_cb);
1930
1931 # Glib/Gtk2 integration
1932 add_watch Glib::IO IO::AIO::poll_fileno,
1933 in => sub { IO::AIO::poll_cb; 1 };
1934
1935 # Tk integration
1936 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
1937 readable => \&IO::AIO::poll_cb);
1938
1939 # Danga::Socket integration
1940 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
1941 \&IO::AIO::poll_cb);
1942
1470=head2 FORK BEHAVIOUR 1943=head2 FORK BEHAVIOUR
1471 1944
1472This module should do "the right thing" when the process using it forks: 1945Usage of pthreads in a program changes the semantics of fork
1946considerably. Specifically, only async-safe functions can be called after
1947fork. Perl doesn't know about this, so in general, you cannot call fork
1948with defined behaviour in perl if pthreads are involved. IO::AIO uses
1949pthreads, so this applies, but many other extensions and (for inexplicable
1950reasons) perl itself often is linked against pthreads, so this limitation
1951applies to quite a lot of perls.
1473 1952
1474Before the fork, IO::AIO enters a quiescent state where no requests 1953This module no longer tries to fight your OS, or POSIX. That means IO::AIO
1475can be added in other threads and no results will be processed. After 1954only works in the process that loaded it. Forking is fully supported, but
1476the fork the parent simply leaves the quiescent state and continues 1955using IO::AIO in the child is not.
1477request/result processing, while the child frees the request/result queue
1478(so that the requests started before the fork will only be handled in the
1479parent). Threads will be started on demand until the limit set in the
1480parent process has been reached again.
1481 1956
1482In short: the parent will, after a short pause, continue as if fork had 1957You might get around by not I<using> IO::AIO before (or after)
1483not been called, while the child will act as if IO::AIO has not been used 1958forking. You could also try to call the L<IO::AIO::reinit> function in the
1484yet. 1959child:
1960
1961=over 4
1962
1963=item IO::AIO::reinit
1964
1965Abandons all current requests and I/O threads and simply reinitialises all
1966data structures. This is not an operation supported by any standards, but
1967happens to work on GNU/Linux and some newer BSD systems.
1968
1969The only reasonable use for this function is to call it after forking, if
1970C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
1971the process will result in undefined behaviour. Calling it at any time
1972will also result in any undefined (by POSIX) behaviour.
1973
1974=back
1485 1975
1486=head2 MEMORY USAGE 1976=head2 MEMORY USAGE
1487 1977
1488Per-request usage: 1978Per-request usage:
1489 1979

Diff Legend

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