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 1.151 by root, Fri Jun 12 00:43:16 2009 UTC vs.
Revision 1.209 by root, Tue Sep 27 00:41:51 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
… …
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 30
31 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
32 use AnyEvent::AIO;
33
34 # EV integration
35 my $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
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. 34operating system supports. It is implemented as an interface to C<libeio>
35(L<http://software.schmorp.de/pkg/libeio.html>).
58 36
59Asynchronous means that operations that can normally block your program 37Asynchronous means that operations that can normally block your program
60(e.g. reading from disk) will be done asynchronously: the operation 38(e.g. reading from disk) will be done asynchronously: the operation
61will still block, but you can do something else in the meantime. This 39will still block, but you can do something else in the meantime. This
62is extremely useful for programs that need to stay interactive even 40is extremely useful for programs that need to stay interactive even
… …
66on a RAID volume or over NFS when you do a number of stat operations 44on a RAID volume or over NFS when you do a number of stat operations
67concurrently. 45concurrently.
68 46
69While most of this works on all types of file descriptors (for 47While most of this works on all types of file descriptors (for
70example sockets), using these functions on file descriptors that 48example sockets), using these functions on file descriptors that
71support nonblocking operation (again, sockets, pipes etc.) is very 49support nonblocking operation (again, sockets, pipes etc.) is
72inefficient. Use an event loop for that (such as the L<Event|Event> 50very inefficient. Use an event loop for that (such as the L<EV>
73module): IO::AIO will naturally fit into such an event loop itself. 51module): IO::AIO will naturally fit into such an event loop itself.
74 52
75In this version, a number of threads are started that execute your 53In this version, a number of threads are started that execute your
76requests and signal their completion. You don't need thread support 54requests and signal their completion. You don't need thread support
77in perl, and the threads created by this module will not be visible 55in perl, and the threads created by this module will not be visible
… …
87yourself, always call C<poll_cb> from within the same thread, or never 65yourself, always call C<poll_cb> from within the same thread, or never
88call C<poll_cb> (or other C<aio_> functions) recursively. 66call C<poll_cb> (or other C<aio_> functions) recursively.
89 67
90=head2 EXAMPLE 68=head2 EXAMPLE
91 69
92This is a simple example that uses the Event module and loads 70This is a simple example that uses the EV module and loads
93F</etc/passwd> asynchronously: 71F</etc/passwd> asynchronously:
94 72
95 use Fcntl; 73 use Fcntl;
96 use Event; 74 use EV;
97 use IO::AIO; 75 use IO::AIO;
98 76
99 # register the IO::AIO callback with Event 77 # register the IO::AIO callback with EV
100 Event->io (fd => IO::AIO::poll_fileno, 78 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
101 poll => 'r',
102 cb => \&IO::AIO::poll_cb);
103 79
104 # queue the request to open /etc/passwd 80 # queue the request to open /etc/passwd
105 aio_open "/etc/passwd", O_RDONLY, 0, sub { 81 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
106 my $fh = shift 82 my $fh = shift
107 or die "error while opening: $!"; 83 or die "error while opening: $!";
108 84
109 # stat'ing filehandles is generally non-blocking 85 # stat'ing filehandles is generally non-blocking
110 my $size = -s $fh; 86 my $size = -s $fh;
… …
119 95
120 # file contents now in $contents 96 # file contents now in $contents
121 print $contents; 97 print $contents;
122 98
123 # exit event loop and program 99 # exit event loop and program
124 Event::unloop; 100 EV::unloop;
125 }; 101 };
126 }; 102 };
127 103
128 # possibly queue up other requests, or open GUI windows, 104 # possibly queue up other requests, or open GUI windows,
129 # check for sockets etc. etc. 105 # check for sockets etc. etc.
130 106
131 # process events as long as there are some: 107 # process events as long as there are some:
132 Event::loop; 108 EV::loop;
133 109
134=head1 REQUEST ANATOMY AND LIFETIME 110=head1 REQUEST ANATOMY AND LIFETIME
135 111
136Every C<aio_*> function creates a request. which is a C data structure not 112Every C<aio_*> function creates a request. which is a C data structure not
137directly visible to Perl. 113directly visible to Perl.
… …
187 163
188package IO::AIO; 164package IO::AIO;
189 165
190use Carp (); 166use Carp ();
191 167
192no warnings; 168use common::sense;
193use strict 'vars';
194 169
195use base 'Exporter'; 170use base 'Exporter';
196 171
197BEGIN { 172BEGIN {
198 our $VERSION = '3.2'; 173 our $VERSION = '4.0';
199 174
200 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
201 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
202 aio_scandir aio_symlink aio_readlink aio_sync aio_fsync 177 aio_scandir aio_symlink aio_readlink aio_realpath aio_sync
203 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
204 aio_rename aio_link aio_move aio_copy aio_group 180 aio_rename aio_link aio_move aio_copy aio_group
205 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
206 aio_chmod aio_utime aio_truncate); 182 aio_chmod aio_utime aio_truncate
183 aio_msync aio_mtouch aio_mlock aio_mlockall
184 aio_statvfs
185 aio_wd);
207 186
208 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); 187 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
209 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 188 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
210 min_parallel max_parallel max_idle 189 min_parallel max_parallel max_idle idle_timeout
211 nreqs nready npending nthreads 190 nreqs nready npending nthreads
212 max_poll_time max_poll_reqs); 191 max_poll_time max_poll_reqs
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
218 require XSLoader; 199 require XSLoader;
219 XSLoader::load ("IO::AIO", $VERSION); 200 XSLoader::load ("IO::AIO", $VERSION);
220} 201}
221 202
222=head1 FUNCTIONS 203=head1 FUNCTIONS
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
223 280
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,
… …
235internally until the request has finished. 292internally until the request has finished.
236 293
237All functions return request objects of type L<IO::AIO::REQ> that allow 294All functions return request objects of type L<IO::AIO::REQ> that allow
238further manipulation of those requests while they are in-flight. 295further manipulation of those requests while they are in-flight.
239 296
240The pathnames you pass to these routines I<must> be absolute and 297The 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 298reason for this is that at the time the request is being executed, the
242request is being executed, the current working directory could have 299current working directory could have changed. Alternatively, you can make
243changed. Alternatively, you can make sure that you never change the 300sure that you never change the current working directory anywhere in
244current working directory anywhere in the program and then use relative 301the program and then use relative paths. Lastly, you can take advantage
245paths. 302of IO::AIOs working directory abstraction - see the description of the
303C<IO::AIO::WD> class later in this document.
246 304
247To encode pathnames as octets, either make sure you either: a) always pass 305To encode pathnames as octets, either make sure you either: a) always pass
248in filenames you got from outside (command line, readdir etc.) without 306in 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 307tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
250your pathnames to the locale (or other) encoding in effect in the user 308your pathnames to the locale (or other) encoding in effect in the user
… …
307by the umask in effect then the request is being executed, so better never 365by the umask in effect then the request is being executed, so better never
308change the umask. 366change the umask.
309 367
310Example: 368Example:
311 369
312 aio_open "/etc/passwd", O_RDONLY, 0, sub { 370 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
313 if ($_[0]) { 371 if ($_[0]) {
314 print "open successful, fh is $_[0]\n"; 372 print "open successful, fh is $_[0]\n";
315 ... 373 ...
316 } else { 374 } else {
317 die "open failed: $!\n"; 375 die "open failed: $!\n";
318 } 376 }
319 }; 377 };
320 378
379In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
380C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
381following POSIX and non-POSIX constants are available (missing ones on
382your system are, as usual, C<0>):
383
384C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
385C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
386C<O_RSYNC>, C<O_SYNC> and C<O_TTY_INIT>.
387
321 388
322=item aio_close $fh, $callback->($status) 389=item aio_close $fh, $callback->($status)
323 390
324Asynchronously close a file and call the callback with the result 391Asynchronously close a file and call the callback with the result
325code. 392code.
… …
375 442
376Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 443Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
377reading at byte offset C<$in_offset>, and starts writing at the current 444reading at byte offset C<$in_offset>, and starts writing at the current
378file offset of C<$out_fh>. Because of that, it is not safe to issue more 445file offset of C<$out_fh>. Because of that, it is not safe to issue more
379than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each 446than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
380other. 447other. The same C<$in_fh> works fine though, as this function does not
448move or use the file offset of C<$in_fh>.
381 449
450Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
451are written, and there is no way to find out how many more bytes have been
452read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
453number of bytes written to C<$out_fh>. Only if the result value equals
454C<$length> one can assume that C<$length> bytes have been read.
455
456Unlike with other C<aio_> functions, it makes a lot of sense to use
457C<aio_sendfile> on non-blocking sockets, as long as one end (typically
458the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
459the socket I/O will be non-blocking. Note, however, that you can run
460into a trap where C<aio_sendfile> reads some data with readahead, then
461fails to write all data, and when the socket is ready the next time, the
462data in the cache is already lost, forcing C<aio_sendfile> to again hit
463the disk. Explicit C<aio_read> + C<aio_write> let's you better control
464resource usage.
465
382This call tries to make use of a native C<sendfile> syscall to provide 466This call tries to make use of a native C<sendfile>-like syscall to
383zero-copy operation. For this to work, C<$out_fh> should refer to a 467provide zero-copy operation. For this to work, C<$out_fh> should refer to
384socket, and C<$in_fh> should refer to mmap'able file. 468a socket, and C<$in_fh> should refer to an mmap'able file.
385 469
386If the native sendfile call fails or is not implemented, it will be 470If a native sendfile cannot be found or it fails with C<ENOSYS>,
387emulated, so you can call C<aio_sendfile> on any type of filehandle 471C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
472C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
388regardless of the limitations of the operating system. 473type of filehandle regardless of the limitations of the operating system.
389 474
390Please note, however, that C<aio_sendfile> can read more bytes from 475As native sendfile syscalls (as practically any non-POSIX interface hacked
391C<$in_fh> than are written, and there is no way to find out how many 476together in a hurry to improve benchmark numbers) tend to be rather buggy
392bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only 477on many systems, this implementation tries to work around some known bugs
393provides the number of bytes written to C<$out_fh>. Only if the result 478in Linux and FreeBSD kernels (probably others, too), but that might fail,
394value equals C<$length> one can assume that C<$length> bytes have been 479so you really really should check the return value of C<aio_sendfile> -
395read. 480fewre bytes than expected might have been transferred.
396 481
397 482
398=item aio_readahead $fh,$offset,$length, $callback->($retval) 483=item aio_readahead $fh,$offset,$length, $callback->($retval)
399 484
400C<aio_readahead> populates the page cache with data from a file so that 485C<aio_readahead> populates the page cache with data from a file so that
… …
423 508
424Currently, the stats are always 64-bit-stats, i.e. instead of returning an 509Currently, the stats are always 64-bit-stats, i.e. instead of returning an
425error when stat'ing a large file, the results will be silently truncated 510error when stat'ing a large file, the results will be silently truncated
426unless perl itself is compiled with large file support. 511unless perl itself is compiled with large file support.
427 512
513To help interpret the mode and dev/rdev stat values, IO::AIO offers the
514following constants and functions (if not implemented, the constants will
515be C<0> and the functions will either C<croak> or fall back on traditional
516behaviour).
517
518C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
519C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
520C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
521
428Example: Print the length of F</etc/passwd>: 522Example: Print the length of F</etc/passwd>:
429 523
430 aio_stat "/etc/passwd", sub { 524 aio_stat "/etc/passwd", sub {
431 $_[0] and die "stat failed: $!"; 525 $_[0] and die "stat failed: $!";
432 print "size is ", -s _, "\n"; 526 print "size is ", -s _, "\n";
433 }; 527 };
434 528
435 529
530=item aio_statvfs $fh_or_path, $callback->($statvfs)
531
532Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
533whether a file handle or path was passed.
534
535On success, the callback is passed a hash reference with the following
536members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
537C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
538is passed.
539
540The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
541C<ST_NOSUID>.
542
543The following non-POSIX IO::AIO::ST_* flag masks are defined to
544their correct value when available, or to C<0> on systems that do
545not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
546C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
547C<ST_NODIRATIME> and C<ST_RELATIME>.
548
549Example: stat C</wd> and dump out the data if successful.
550
551 aio_statvfs "/wd", sub {
552 my $f = $_[0]
553 or die "statvfs: $!";
554
555 use Data::Dumper;
556 say Dumper $f;
557 };
558
559 # result:
560 {
561 bsize => 1024,
562 bfree => 4333064312,
563 blocks => 10253828096,
564 files => 2050765568,
565 flag => 4096,
566 favail => 2042092649,
567 bavail => 4333064312,
568 ffree => 2042092649,
569 namemax => 255,
570 frsize => 1024,
571 fsid => 1810
572 }
573
574
436=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 575=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
437 576
438Works like perl's C<utime> function (including the special case of $atime 577Works like perl's C<utime> function (including the special case of $atime
439and $mtime being undef). Fractional times are supported if the underlying 578and $mtime being undef). Fractional times are supported if the underlying
440syscalls support them. 579syscalls support them.
… …
478 617
479Asynchronously unlink (delete) a file and call the callback with the 618Asynchronously unlink (delete) a file and call the callback with the
480result code. 619result code.
481 620
482 621
483=item aio_mknod $path, $mode, $dev, $callback->($status) 622=item aio_mknod $pathname, $mode, $dev, $callback->($status)
484 623
485[EXPERIMENTAL] 624[EXPERIMENTAL]
486 625
487Asynchronously create a device node (or fifo). See mknod(2). 626Asynchronously create a device node (or fifo). See mknod(2).
488 627
489The only (POSIX-) portable way of calling this function is: 628The only (POSIX-) portable way of calling this function is:
490 629
491 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 630 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
492 631
632See C<aio_stat> for info about some potentially helpful extra constants
633and functions.
493 634
494=item aio_link $srcpath, $dstpath, $callback->($status) 635=item aio_link $srcpath, $dstpath, $callback->($status)
495 636
496Asynchronously create a new link to the existing object at C<$srcpath> at 637Asynchronously create a new link to the existing object at C<$srcpath> at
497the path C<$dstpath> and call the callback with the result code. 638the path C<$dstpath> and call the callback with the result code.
… …
501 642
502Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 643Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
503the path C<$dstpath> and call the callback with the result code. 644the path C<$dstpath> and call the callback with the result code.
504 645
505 646
506=item aio_readlink $path, $callback->($link) 647=item aio_readlink $pathname, $callback->($link)
507 648
508Asynchronously read the symlink specified by C<$path> and pass it to 649Asynchronously read the symlink specified by C<$path> and pass it to
509the callback. If an error occurs, nothing or undef gets passed to the 650the callback. If an error occurs, nothing or undef gets passed to the
510callback. 651callback.
511 652
512 653
654=item aio_realpath $pathname, $callback->($path)
655
656Asynchronously make the path absolute and resolve any symlinks in
657C<$path>. The resulting path only consists of directories (Same as
658L<Cwd::realpath>).
659
660This request can be used to get the absolute path of the current working
661directory by passing it a path of F<.> (a single dot).
662
663
513=item aio_rename $srcpath, $dstpath, $callback->($status) 664=item aio_rename $srcpath, $dstpath, $callback->($status)
514 665
515Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 666Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
516rename(2) and call the callback with the result code. 667rename(2) and call the callback with the result code.
517 668
… …
539array-ref with the filenames. 690array-ref with the filenames.
540 691
541 692
542=item aio_readdirx $pathname, $flags, $callback->($entries, $flags) 693=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
543 694
544Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune 695Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
545behaviour and output format. In case of an error, C<$entries> will be 696tune behaviour and output format. In case of an error, C<$entries> will be
546C<undef>. 697C<undef>.
547 698
548The flags are a combination of the following constants, ORed together (the 699The flags are a combination of the following constants, ORed together (the
549flags will also be passed to the callback, possibly modified): 700flags will also be passed to the callback, possibly modified):
550 701
551=over 4 702=over 4
552 703
553=item IO::AIO::READDIR_DENTS 704=item IO::AIO::READDIR_DENTS
554 705
555When this flag is off, then the callback gets an arrayref with of names 706When this flag is off, then the callback gets an arrayref consisting of
556only (as with C<aio_readdir>), otherwise it gets an arrayref with 707names only (as with C<aio_readdir>), otherwise it gets an arrayref with
557C<[$name, $type, $inode]> arrayrefs, each describing a single directory 708C<[$name, $type, $inode]> arrayrefs, each describing a single directory
558entry in more detail. 709entry in more detail.
559 710
560C<$name> is the name of the entry. 711C<$name> is the name of the entry.
561 712
… …
568C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to 719C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
569know, you have to run stat yourself. Also, for speed reasons, the C<$type> 720know, you have to run stat yourself. Also, for speed reasons, the C<$type>
570scalars are read-only: you can not modify them. 721scalars are read-only: you can not modify them.
571 722
572C<$inode> is the inode number (which might not be exact on systems with 64 723C<$inode> is the inode number (which might not be exact on systems with 64
573bit inode numbers and 32 bit perls). On systems that do not deliver the 724bit inode numbers and 32 bit perls). This field has unspecified content on
574inode information, this will always be zero. 725systems that do not deliver the inode information.
575 726
576=item IO::AIO::READDIR_DIRS_FIRST 727=item IO::AIO::READDIR_DIRS_FIRST
577 728
578When this flag is set, then the names will be returned in an order where 729When this flag is set, then the names will be returned in an order where
579likely directories come first. This is useful when you need to quickly 730likely directories come first, in optimal stat order. This is useful when
580find directories, or you want to find all directories while avoiding to 731you need to quickly find directories, or you want to find all directories
581stat() each entry. 732while avoiding to stat() each entry.
582 733
583If the system returns type information in readdir, then this is used 734If the system returns type information in readdir, then this is used
584to find directories directly. Otherwise, likely directories are files 735to find directories directly. Otherwise, likely directories are names
585beginning with ".", or otherwise files with no dots, of which files with 736beginning with ".", or otherwise names with no dots, of which names with
586short names are tried first. 737short names are tried first.
587 738
588=item IO::AIO::READDIR_STAT_ORDER 739=item IO::AIO::READDIR_STAT_ORDER
589 740
590When this flag is set, then the names will be returned in an order 741When this flag is set, then the names will be returned in an order
… …
597 748
598=item IO::AIO::READDIR_FOUND_UNKNOWN 749=item IO::AIO::READDIR_FOUND_UNKNOWN
599 750
600This flag should not be set when calling C<aio_readdirx>. Instead, it 751This flag should not be set when calling C<aio_readdirx>. Instead, it
601is being set by C<aio_readdirx>, when any of the C<$type>'s found were 752is being set by C<aio_readdirx>, when any of the C<$type>'s found were
602C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all 753C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
603C<$type>'s are known, which can be used to speed up some algorithms. 754C<$type>'s are known, which can be used to speed up some algorithms.
604 755
605=back 756=back
606 757
607 758
608=item aio_load $path, $data, $callback->($status) 759=item aio_load $pathname, $data, $callback->($status)
609 760
610This is a composite request that tries to fully load the given file into 761This is a composite request that tries to fully load the given file into
611memory. Status is the same as with aio_read. 762memory. Status is the same as with aio_read.
612 763
613=cut 764=cut
… …
635 786
636=item aio_copy $srcpath, $dstpath, $callback->($status) 787=item aio_copy $srcpath, $dstpath, $callback->($status)
637 788
638Try to copy the I<file> (directories not supported as either source or 789Try to copy the I<file> (directories not supported as either source or
639destination) from C<$srcpath> to C<$dstpath> and call the callback with 790destination) from C<$srcpath> to C<$dstpath> and call the callback with
640the C<0> (error) or C<-1> ok. 791a status of C<0> (ok) or C<-1> (error, see C<$!>).
641 792
642This is a composite request that creates the destination file with 793This is a composite request that creates the destination file with
643mode 0200 and copies the contents of the source file into it using 794mode 0200 and copies the contents of the source file into it using
644C<aio_sendfile>, followed by restoring atime, mtime, access mode and 795C<aio_sendfile>, followed by restoring atime, mtime, access mode and
645uid/gid, in that order. 796uid/gid, in that order.
… …
657 my $grp = aio_group $cb; 808 my $grp = aio_group $cb;
658 809
659 aioreq_pri $pri; 810 aioreq_pri $pri;
660 add $grp aio_open $src, O_RDONLY, 0, sub { 811 add $grp aio_open $src, O_RDONLY, 0, sub {
661 if (my $src_fh = $_[0]) { 812 if (my $src_fh = $_[0]) {
662 my @stat = stat $src_fh; # hmm, might bock over nfs? 813 my @stat = stat $src_fh; # hmm, might block over nfs?
663 814
664 aioreq_pri $pri; 815 aioreq_pri $pri;
665 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub { 816 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
666 if (my $dst_fh = $_[0]) { 817 if (my $dst_fh = $_[0]) {
667 aioreq_pri $pri; 818 aioreq_pri $pri;
… …
714 865
715=item aio_move $srcpath, $dstpath, $callback->($status) 866=item aio_move $srcpath, $dstpath, $callback->($status)
716 867
717Try to move the I<file> (directories not supported as either source or 868Try to move the I<file> (directories not supported as either source or
718destination) from C<$srcpath> to C<$dstpath> and call the callback with 869destination) from C<$srcpath> to C<$dstpath> and call the callback with
719the C<0> (error) or C<-1> ok. 870a status of C<0> (ok) or C<-1> (error, see C<$!>).
720 871
721This is a composite request that tries to rename(2) the file first; if 872This is a composite request that tries to rename(2) the file first; if
722rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if 873rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
723that is successful, unlinks the C<$srcpath>. 874that is successful, unlinks the C<$srcpath>.
724 875
… …
735 if ($_[0] && $! == EXDEV) { 886 if ($_[0] && $! == EXDEV) {
736 aioreq_pri $pri; 887 aioreq_pri $pri;
737 add $grp aio_copy $src, $dst, sub { 888 add $grp aio_copy $src, $dst, sub {
738 $grp->result ($_[0]); 889 $grp->result ($_[0]);
739 890
740 if (!$_[0]) { 891 unless ($_[0]) {
741 aioreq_pri $pri; 892 aioreq_pri $pri;
742 add $grp aio_unlink $src; 893 add $grp aio_unlink $src;
743 } 894 }
744 }; 895 };
745 } else { 896 } else {
… …
748 }; 899 };
749 900
750 $grp 901 $grp
751} 902}
752 903
753=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 904=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
754 905
755Scans a directory (similar to C<aio_readdir>) but additionally tries to 906Scans a directory (similar to C<aio_readdir>) but additionally tries to
756efficiently separate the entries of directory C<$path> into two sets of 907efficiently separate the entries of directory C<$path> into two sets of
757names, directories you can recurse into (directories), and ones you cannot 908names, directories you can recurse into (directories), and ones you cannot
758recurse into (everything else, including symlinks to directories). 909recurse into (everything else, including symlinks to directories).
… …
789Then entries will be sorted into likely directories a non-initial dot 940Then entries will be sorted into likely directories a non-initial dot
790currently) and likely non-directories (see C<aio_readdirx>). Then every 941currently) and likely non-directories (see C<aio_readdirx>). Then every
791entry plus an appended C</.> will be C<stat>'ed, likely directories first, 942entry plus an appended C</.> will be C<stat>'ed, likely directories first,
792in order of their inode numbers. If that succeeds, it assumes that the 943in order of their inode numbers. If that succeeds, it assumes that the
793entry is a directory or a symlink to directory (which will be checked 944entry is a directory or a symlink to directory (which will be checked
794seperately). This is often faster than stat'ing the entry itself because 945separately). This is often faster than stat'ing the entry itself because
795filesystems might detect the type of the entry without reading the inode 946filesystems might detect the type of the entry without reading the inode
796data (e.g. ext2fs filetype feature), even on systems that cannot return 947data (e.g. ext2fs filetype feature), even on systems that cannot return
797the filetype information on readdir. 948the filetype information on readdir.
798 949
799If the known number of directories (link count - 2) has been reached, the 950If the known number of directories (link count - 2) has been reached, the
… …
885 }; 1036 };
886 1037
887 $grp 1038 $grp
888} 1039}
889 1040
890=item aio_rmtree $path, $callback->($status) 1041=item aio_rmtree $pathname, $callback->($status)
891 1042
892Delete a directory tree starting (and including) C<$path>, return the 1043Delete a directory tree starting (and including) C<$path>, return the
893status of the final C<rmdir> only. This is a composite request that 1044status of the final C<rmdir> only. This is a composite request that
894uses C<aio_scandir> to recurse into and rmdir directories, and unlink 1045uses C<aio_scandir> to recurse into and rmdir directories, and unlink
895everything else. 1046everything else.
… …
937callback with the fdatasync result code. 1088callback with the fdatasync result code.
938 1089
939If this call isn't available because your OS lacks it or it couldn't be 1090If this call isn't available because your OS lacks it or it couldn't be
940detected, it will be emulated by calling C<fsync> instead. 1091detected, it will be emulated by calling C<fsync> instead.
941 1092
1093=item aio_syncfs $fh, $callback->($status)
1094
1095Asynchronously call the syncfs syscall to sync the filesystem associated
1096to the given filehandle and call the callback with the syncfs result
1097code. If syncfs is not available, calls sync(), but returns C<-1> and sets
1098errno to C<ENOSYS> nevertheless.
1099
942=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 1100=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
943 1101
944Sync the data portion of the file specified by C<$offset> and C<$length> 1102Sync the data portion of the file specified by C<$offset> and C<$length>
945to disk (but NOT the metadata), by calling the Linux-specific 1103to disk (but NOT the metadata), by calling the Linux-specific
946sync_file_range call. If sync_file_range is not available or it returns 1104sync_file_range call. If sync_file_range is not available or it returns
… …
949C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>, 1107C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
950C<IO::AIO::SYNC_FILE_RANGE_WRITE> and 1108C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
951C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range 1109C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
952manpage for details. 1110manpage for details.
953 1111
954=item aio_pathsync $path, $callback->($status) 1112=item aio_pathsync $pathname, $callback->($status)
955 1113
956This request tries to open, fsync and close the given path. This is a 1114This request tries to open, fsync and close the given path. This is a
957composite request intended to sync directories after directory operations 1115composite request intended to sync directories after directory operations
958(E.g. rename). This might not work on all operating systems or have any 1116(E.g. rename). This might not work on all operating systems or have any
959specific effect, but usually it makes sure that directory changes get 1117specific effect, but usually it makes sure that directory changes get
960written to disc. It works for anything that can be opened for read-only, 1118written to disc. It works for anything that can be opened for read-only,
961not just directories. 1119not just directories.
1120
1121Future versions of this function might fall back to other methods when
1122C<fsync> on the directory fails (such as calling C<sync>).
962 1123
963Passes C<0> when everything went ok, and C<-1> on error. 1124Passes C<0> when everything went ok, and C<-1> on error.
964 1125
965=cut 1126=cut
966 1127
… …
987 }; 1148 };
988 1149
989 $grp 1150 $grp
990} 1151}
991 1152
1153=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
1154
1155This is a rather advanced IO::AIO call, which only works on mmap(2)ed
1156scalars (see the C<IO::AIO::mmap> function, although it also works on data
1157scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
1158scalar must only be modified in-place while an aio operation is pending on
1159it).
1160
1161It calls the C<msync> function of your OS, if available, with the memory
1162area starting at C<$offset> in the string and ending C<$length> bytes
1163later. If C<$length> is negative, counts from the end, and if C<$length>
1164is C<undef>, then it goes till the end of the string. The flags can be
1165a combination of C<IO::AIO::MS_ASYNC>, C<IO::AIO::MS_INVALIDATE> and
1166C<IO::AIO::MS_SYNC>.
1167
1168=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
1169
1170This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1171scalars.
1172
1173It touches (reads or writes) all memory pages in the specified
1174range inside the scalar. All caveats and parameters are the same
1175as for C<aio_msync>, above, except for flags, which must be either
1176C<0> (which reads all pages and ensures they are instantiated) or
1177C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
1178writing an octet from it, which dirties the page).
1179
1180=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1181
1182This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1183scalars.
1184
1185It reads in all the pages of the underlying storage into memory (if any)
1186and locks them, so they are not getting swapped/paged out or removed.
1187
1188If C<$length> is undefined, then the scalar will be locked till the end.
1189
1190On systems that do not implement C<mlock>, this function returns C<-1>
1191and sets errno to C<ENOSYS>.
1192
1193Note that the corresponding C<munlock> is synchronous and is
1194documented under L<MISCELLANEOUS FUNCTIONS>.
1195
1196Example: open a file, mmap and mlock it - both will be undone when
1197C<$data> gets destroyed.
1198
1199 open my $fh, "<", $path or die "$path: $!";
1200 my $data;
1201 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1202 aio_mlock $data; # mlock in background
1203
1204=item aio_mlockall $flags, $callback->($status)
1205
1206Calls the C<mlockall> function with the given C<$flags> (a combination of
1207C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
1208
1209On systems that do not implement C<mlockall>, this function returns C<-1>
1210and sets errno to C<ENOSYS>.
1211
1212Note that the corresponding C<munlockall> is synchronous and is
1213documented under L<MISCELLANEOUS FUNCTIONS>.
1214
1215Example: asynchronously lock all current and future pages into memory.
1216
1217 aio_mlockall IO::AIO::MCL_FUTURE;
1218
992=item aio_group $callback->(...) 1219=item aio_group $callback->(...)
993 1220
994This is a very special aio request: Instead of doing something, it is a 1221This is a very special aio request: Instead of doing something, it is a
995container for other aio requests, which is useful if you want to bundle 1222container for other aio requests, which is useful if you want to bundle
996many requests into a single, composite, request with a definite callback 1223many requests into a single, composite, request with a definite callback
… …
1032like sleep and file handle readable/writable, the overhead this creates is 1259like sleep and file handle readable/writable, the overhead this creates is
1033immense (it blocks a thread for a long time) so do not use this function 1260immense (it blocks a thread for a long time) so do not use this function
1034except to put your application under artificial I/O pressure. 1261except to put your application under artificial I/O pressure.
1035 1262
1036=back 1263=back
1264
1265
1266=head2 IO::AIO::WD - multiple working directories
1267
1268Your process only has one current working directory, which is used by all
1269threads. This makes it hard to use relative paths (some other component
1270could call C<chdir> at any time, and it is hard to control when the path
1271will be used by IO::AIO).
1272
1273One solution for this is to always use absolute paths. This usually works,
1274but can be quite slow (the kernel has to walk the whole path on every
1275access), and can also be a hassle to implement.
1276
1277Newer POSIX systems have a number of functions (openat, fdopendir,
1278futimensat and so on) that make it possible to specify working directories
1279per operation.
1280
1281For portability, and because the clowns who "designed", or shall I write,
1282perpetrated this new interface were obviously half-drunk, this abstraction
1283cannot be perfect, though.
1284
1285IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
1286object. This object stores the canonicalised, absolute version of the
1287path, and on systems that allow it, also a directory file descriptor.
1288
1289Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
1290or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
1291object and a pathname instead. If the pathname is absolute, the
1292IO::AIO::WD objetc is ignored, otherwise the pathname is resolved relative
1293to that IO::AIO::WD object.
1294
1295For example, to get a wd object for F</etc> and then stat F<passwd>
1296inside, you would write:
1297
1298 aio_wd "/etc", sub {
1299 my $etcdir = shift;
1300
1301 # although $etcdir can be undef on error, there is generally no reason
1302 # to check for errors here, as aio_stat will fail with ENOENT
1303 # when $etcdir is undef.
1304
1305 aio_stat [$etcdir, "passwd"], sub {
1306 # yay
1307 };
1308 };
1309
1310This shows that creating an IO::AIO::WD object is itself a potentially
1311blocking operation, which is why it is done asynchronously.
1312
1313As with normal pathnames, IO::AIO keeps a copy of the working directory
1314object and the pathname string, so you could write the following without
1315causing any issues due to C<$path> getting reused:
1316
1317 my $path = [$wd, undef];
1318
1319 for my $name (qw(abc def ghi)) {
1320 $path->[1] = $name;
1321 aio_stat $path, sub {
1322 # ...
1323 };
1324 }
1325
1326There are some caveats: when directories get renamed (or deleted), the
1327pathname string doesn't change, so will point to the new directory (or
1328nowhere at all), while the directory fd, if available on the system,
1329will still point to the original directory. Most functions accepting a
1330pathname will use the directory fd on newer systems, and the string on
1331older systems. Some functions (such as realpath) will always rely on the
1332string form of the pathname.
1333
1334So this fucntionality is mainly useful to get some protection against
1335C<chdir>, to easily get an absolute path out of a relative path for future
1336reference, and to speed up doing many operations in the same directory
1337(e.g. when stat'ing all files in a directory).
1338
1339The following functions implement this working directory abstraction:
1340
1341=over 4
1342
1343=item aio_wd $pathname, $callback->($wd)
1344
1345Asynchonously canonicalise the given pathname and convert it to an
1346IO::AIO::WD object representing it. If possible and supported on the
1347system, also open a directory fd to speed up pathname resolution relative
1348to this working directory.
1349
1350If something goes wrong, then C<undef> is passwd to the callback instead
1351of a working directory object and C<$!> is set appropriately. Since
1352passing C<undef> as working directory component of a pathname fails the
1353request with C<ENOENT>, there is often no need for error checking in the
1354C<aio_wd> callback, as future requests using the value will fail in the
1355expected way.
1356
1357If this call isn't available because your OS lacks it or it couldn't be
1358detected, it will be emulated by calling C<fsync> instead.
1359
1360=item IO::AIO::CWD
1361
1362This is a compiletime constant (object) that represents the process
1363current working directory.
1364
1365Specifying this object as working directory object for a pathname is as
1366if the pathname would be specified directly, without a directory object,
1367e.g., these calls are functionally identical:
1368
1369 aio_stat "somefile", sub { ... };
1370 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1371
1372=back
1373
1037 1374
1038=head2 IO::AIO::REQ CLASS 1375=head2 IO::AIO::REQ CLASS
1039 1376
1040All non-aggregate C<aio_*> functions return an object of this class when 1377All non-aggregate C<aio_*> functions return an object of this class when
1041called in non-void context. 1378called in non-void context.
… …
1131=item $grp->cancel_subs 1468=item $grp->cancel_subs
1132 1469
1133Cancel all subrequests and clears any feeder, but not the group request 1470Cancel all subrequests and clears any feeder, but not the group request
1134itself. Useful when you queued a lot of events but got a result early. 1471itself. Useful when you queued a lot of events but got a result early.
1135 1472
1473The group request will finish normally (you cannot add requests to the
1474group).
1475
1136=item $grp->result (...) 1476=item $grp->result (...)
1137 1477
1138Set the result value(s) that will be passed to the group callback when all 1478Set the result value(s) that will be passed to the group callback when all
1139subrequests have finished and set the groups errno to the current value 1479subrequests have finished and set the groups errno to the current value
1140of errno (just like calling C<errno> without an error number). By default, 1480of errno (just like calling C<errno> without an error number). By default,
… …
1208=over 4 1548=over 4
1209 1549
1210=item $fileno = IO::AIO::poll_fileno 1550=item $fileno = IO::AIO::poll_fileno
1211 1551
1212Return the I<request result pipe file descriptor>. This filehandle must be 1552Return the I<request result pipe file descriptor>. This filehandle must be
1213polled for reading by some mechanism outside this module (e.g. Event or 1553polled for reading by some mechanism outside this module (e.g. EV, Glib,
1214select, see below or the SYNOPSIS). If the pipe becomes readable you have 1554select and so on, see below or the SYNOPSIS). If the pipe becomes readable
1215to call C<poll_cb> to check the results. 1555you have to call C<poll_cb> to check the results.
1216 1556
1217See C<poll_cb> for an example. 1557See C<poll_cb> for an example.
1218 1558
1219=item IO::AIO::poll_cb 1559=item IO::AIO::poll_cb
1220 1560
1221Process some outstanding events on the result pipe. You have to call this 1561Process some outstanding events on the result pipe. You have to call
1222regularly. Returns C<0> if all events could be processed, or C<-1> if it 1562this regularly. Returns C<0> if all events could be processed (or there
1223returned earlier for whatever reason. Returns immediately when no events 1563were no events to process), or C<-1> if it returned earlier for whatever
1224are outstanding. The amount of events processed depends on the settings of 1564reason. Returns immediately when no events are outstanding. The amount of
1225C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>. 1565events processed depends on the settings of C<IO::AIO::max_poll_req> and
1566C<IO::AIO::max_poll_time>.
1226 1567
1227If not all requests were processed for whatever reason, the filehandle 1568If not all requests were processed for whatever reason, the filehandle
1228will still be ready when C<poll_cb> returns, so normally you don't have to 1569will still be ready when C<poll_cb> returns, so normally you don't have to
1229do anything special to have it called later. 1570do anything special to have it called later.
1230 1571
1572Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
1573ready, it can be beneficial to call this function from loops which submit
1574a lot of requests, to make sure the results get processed when they become
1575available and not just when the loop is finished and the event loop takes
1576over again. This function returns very fast when there are no outstanding
1577requests.
1578
1231Example: Install an Event watcher that automatically calls 1579Example: Install an Event watcher that automatically calls
1232IO::AIO::poll_cb with high priority: 1580IO::AIO::poll_cb with high priority (more examples can be found in the
1581SYNOPSIS section, at the top of this document):
1233 1582
1234 Event->io (fd => IO::AIO::poll_fileno, 1583 Event->io (fd => IO::AIO::poll_fileno,
1235 poll => 'r', async => 1, 1584 poll => 'r', async => 1,
1236 cb => \&IO::AIO::poll_cb); 1585 cb => \&IO::AIO::poll_cb);
1586
1587=item IO::AIO::poll_wait
1588
1589If there are any outstanding requests and none of them in the result
1590phase, wait till the result filehandle becomes ready for reading (simply
1591does a C<select> on the filehandle. This is useful if you want to
1592synchronously wait for some requests to finish).
1593
1594See C<nreqs> for an example.
1595
1596=item IO::AIO::poll
1597
1598Waits until some requests have been handled.
1599
1600Returns the number of requests processed, but is otherwise strictly
1601equivalent to:
1602
1603 IO::AIO::poll_wait, IO::AIO::poll_cb
1604
1605=item IO::AIO::flush
1606
1607Wait till all outstanding AIO requests have been handled.
1608
1609Strictly equivalent to:
1610
1611 IO::AIO::poll_wait, IO::AIO::poll_cb
1612 while IO::AIO::nreqs;
1237 1613
1238=item IO::AIO::max_poll_reqs $nreqs 1614=item IO::AIO::max_poll_reqs $nreqs
1239 1615
1240=item IO::AIO::max_poll_time $seconds 1616=item IO::AIO::max_poll_time $seconds
1241 1617
… …
1266 # use a low priority so other tasks have priority 1642 # use a low priority so other tasks have priority
1267 Event->io (fd => IO::AIO::poll_fileno, 1643 Event->io (fd => IO::AIO::poll_fileno,
1268 poll => 'r', nice => 1, 1644 poll => 'r', nice => 1,
1269 cb => &IO::AIO::poll_cb); 1645 cb => &IO::AIO::poll_cb);
1270 1646
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 1647=back
1299 1648
1300=head3 CONTROLLING THE NUMBER OF THREADS 1649=head3 CONTROLLING THE NUMBER OF THREADS
1301 1650
1302=over 1651=over
… …
1335 1684
1336Under normal circumstances you don't need to call this function. 1685Under normal circumstances you don't need to call this function.
1337 1686
1338=item IO::AIO::max_idle $nthreads 1687=item IO::AIO::max_idle $nthreads
1339 1688
1340Limit the number of threads (default: 4) that are allowed to idle (i.e., 1689Limit the number of threads (default: 4) that are allowed to idle
1341threads that did not get a request to process within 10 seconds). That 1690(i.e., threads that did not get a request to process within the idle
1342means if a thread becomes idle while C<$nthreads> other threads are also 1691timeout (default: 10 seconds). That means if a thread becomes idle while
1343idle, it will free its resources and exit. 1692C<$nthreads> other threads are also idle, it will free its resources and
1693exit.
1344 1694
1345This is useful when you allow a large number of threads (e.g. 100 or 1000) 1695This 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 1696to allow for extremely high load situations, but want to free resources
1347under normal circumstances (1000 threads can easily consume 30MB of RAM). 1697under normal circumstances (1000 threads can easily consume 30MB of RAM).
1348 1698
1349The default is probably ok in most situations, especially if thread 1699The default is probably ok in most situations, especially if thread
1350creation is fast. If thread creation is very slow on your system you might 1700creation is fast. If thread creation is very slow on your system you might
1351want to use larger values. 1701want to use larger values.
1352 1702
1703=item IO::AIO::idle_timeout $seconds
1704
1705Sets the minimum idle timeout (default 10) after which worker threads are
1706allowed to exit. SEe C<IO::AIO::max_idle>.
1707
1353=item IO::AIO::max_outstanding $maxreqs 1708=item IO::AIO::max_outstanding $maxreqs
1709
1710Sets the maximum number of outstanding requests to C<$nreqs>. If
1711you do queue up more than this number of requests, the next call to
1712C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
1713C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
1714longer exceeded.
1715
1716In other words, this setting does not enforce a queue limit, but can be
1717used to make poll functions block if the limit is exceeded.
1354 1718
1355This is a very bad function to use in interactive programs because it 1719This 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 1720blocks, and a bad way to reduce concurrency because it is inexact: Better
1357use an C<aio_group> together with a feed callback. 1721use an C<aio_group> together with a feed callback.
1358 1722
1359Sets the maximum number of outstanding requests to C<$nreqs>. If you 1723It's 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 1724a lot of files, you can write somehting 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 1725
1364The default value is very large, so there is no practical limit on the 1726 IO::AIO::max_outstanding 32;
1365number of outstanding requests.
1366 1727
1367You can still queue as many requests as you want. Therefore, 1728 for my $path (...) {
1368C<max_outstanding> is mainly useful in simple scripts (with low values) or 1729 aio_stat $path , ...;
1369as a stop gap to shield against fatal memory overflow (with large values). 1730 IO::AIO::poll_cb;
1731 }
1732
1733 IO::AIO::flush;
1734
1735The call to C<poll_cb> inside the loop will normally return instantly, but
1736as soon as more thna C<32> reqeusts are in-flight, it will block until
1737some requests have been handled. This keeps the loop from pushing a large
1738number of C<aio_stat> requests onto the queue.
1739
1740The default value for C<max_outstanding> is very large, so there is no
1741practical limit on the number of outstanding requests.
1370 1742
1371=back 1743=back
1372 1744
1373=head3 STATISTICAL INFORMATION 1745=head3 STATISTICAL INFORMATION
1374 1746
… …
1394Returns the number of requests currently in the pending state (executed, 1766Returns the number of requests currently in the pending state (executed,
1395but not yet processed by poll_cb). 1767but not yet processed by poll_cb).
1396 1768
1397=back 1769=back
1398 1770
1771=head3 MISCELLANEOUS FUNCTIONS
1772
1773IO::AIO implements some functions that might be useful, but are not
1774asynchronous.
1775
1776=over 4
1777
1778=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
1779
1780Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
1781but is blocking (this makes most sense if you know the input data is
1782likely cached already and the output filehandle is set to non-blocking
1783operations).
1784
1785Returns the number of bytes copied, or C<-1> on error.
1786
1787=item IO::AIO::fadvise $fh, $offset, $len, $advice
1788
1789Simply calls the C<posix_fadvise> function (see its
1790manpage for details). The following advice constants are
1791available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
1792C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
1793C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
1794
1795On systems that do not implement C<posix_fadvise>, this function returns
1796ENOSYS, otherwise the return value of C<posix_fadvise>.
1797
1798=item IO::AIO::madvise $scalar, $offset, $len, $advice
1799
1800Simply calls the C<posix_madvise> function (see its
1801manpage for details). The following advice constants are
1802available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
1803C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
1804
1805On systems that do not implement C<posix_madvise>, this function returns
1806ENOSYS, otherwise the return value of C<posix_madvise>.
1807
1808=item IO::AIO::mprotect $scalar, $offset, $len, $protect
1809
1810Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
1811$scalar (see its manpage for details). The following protect
1812constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
1813C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
1814
1815On systems that do not implement C<mprotect>, this function returns
1816ENOSYS, otherwise the return value of C<mprotect>.
1817
1818=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1819
1820Memory-maps a file (or anonymous memory range) and attaches it to the
1821given C<$scalar>, which will act like a string scalar.
1822
1823The only operations allowed on the scalar are C<substr>/C<vec> that don't
1824change the string length, and most read-only operations such as copying it
1825or searching it with regexes and so on.
1826
1827Anything else is unsafe and will, at best, result in memory leaks.
1828
1829The memory map associated with the C<$scalar> is automatically removed
1830when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
1831C<IO::AIO::munmap> functions are called.
1832
1833This calls the C<mmap>(2) function internally. See your system's manual
1834page for details on the C<$length>, C<$prot> and C<$flags> parameters.
1835
1836The C<$length> must be larger than zero and smaller than the actual
1837filesize.
1838
1839C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
1840C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
1841
1842C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
1843C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
1844not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
1845(which is set to C<MAP_ANON> if your system only provides this
1846constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
1847C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
1848C<IO::AIO::MAP_NONBLOCK>
1849
1850If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
1851
1852C<$offset> is the offset from the start of the file - it generally must be
1853a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
1854
1855Example:
1856
1857 use Digest::MD5;
1858 use IO::AIO;
1859
1860 open my $fh, "<verybigfile"
1861 or die "$!";
1862
1863 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1864 or die "verybigfile: $!";
1865
1866 my $fast_md5 = md5 $data;
1867
1868=item IO::AIO::munmap $scalar
1869
1870Removes a previous mmap and undefines the C<$scalar>.
1871
1872=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
1873
1874Calls the C<munlock> function, undoing the effects of a previous
1875C<aio_mlock> call (see its description for details).
1876
1877=item IO::AIO::munlockall
1878
1879Calls the C<munlockall> function.
1880
1881On systems that do not implement C<munlockall>, this function returns
1882ENOSYS, otherwise the return value of C<munlockall>.
1883
1884=back
1885
1399=cut 1886=cut
1400 1887
1401min_parallel 8; 1888min_parallel 8;
1402 1889
1403END { flush } 1890END { flush }
1404 1891
14051; 18921;
1406 1893
1894=head1 EVENT LOOP INTEGRATION
1895
1896It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
1897automatically into many event loops:
1898
1899 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
1900 use AnyEvent::AIO;
1901
1902You can also integrate IO::AIO manually into many event loops, here are
1903some examples of how to do this:
1904
1905 # EV integration
1906 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
1907
1908 # Event integration
1909 Event->io (fd => IO::AIO::poll_fileno,
1910 poll => 'r',
1911 cb => \&IO::AIO::poll_cb);
1912
1913 # Glib/Gtk2 integration
1914 add_watch Glib::IO IO::AIO::poll_fileno,
1915 in => sub { IO::AIO::poll_cb; 1 };
1916
1917 # Tk integration
1918 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
1919 readable => \&IO::AIO::poll_cb);
1920
1921 # Danga::Socket integration
1922 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
1923 \&IO::AIO::poll_cb);
1924
1407=head2 FORK BEHAVIOUR 1925=head2 FORK BEHAVIOUR
1408 1926
1409This module should do "the right thing" when the process using it forks: 1927Usage of pthreads in a program changes the semantics of fork
1928considerably. Specifically, only async-safe functions can be called after
1929fork. Perl doesn't know about this, so in general, you cannot call fork
1930with defined behaviour in perl if pthreads are involved. IO::AIO uses
1931pthreads, so this applies, but many other extensions and (for inexplicable
1932reasons) perl itself often is linked against pthreads, so this limitation
1933applies to quite a lot of perls.
1410 1934
1411Before the fork, IO::AIO enters a quiescent state where no requests 1935This module no longer tries to fight your OS, or POSIX. That means IO::AIO
1412can be added in other threads and no results will be processed. After 1936only works in the process that loaded it. Forking is fully supported, but
1413the fork the parent simply leaves the quiescent state and continues 1937using IO::AIO in the child is not.
1414request/result processing, while the child frees the request/result queue
1415(so that the requests started before the fork will only be handled in the
1416parent). Threads will be started on demand until the limit set in the
1417parent process has been reached again.
1418 1938
1419In short: the parent will, after a short pause, continue as if fork had 1939You might get around by not I<using> IO::AIO before (or after)
1420not been called, while the child will act as if IO::AIO has not been used 1940forking. You could also try to call the L<IO::AIO::reinit> function in the
1421yet. 1941child:
1942
1943=over 4
1944
1945=item IO::AIO::reinit
1946
1947Abandons all current requests and I/O threads and simply reinitialises all
1948data structures. This is not an operation supported by any standards, but
1949happens to work on GNU/Linux and some newer BSD systems.
1950
1951The only reasonable use for this function is to call it after forking, if
1952C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
1953the process will result in undefined behaviour. Calling it at any time
1954will also result in any undefined (by POSIX) behaviour.
1955
1956=back
1422 1957
1423=head2 MEMORY USAGE 1958=head2 MEMORY USAGE
1424 1959
1425Per-request usage: 1960Per-request usage:
1426 1961

Diff Legend

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