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 1.117 by root, Sat Oct 6 14:05:19 2007 UTC vs.
Revision 1.158 by root, Wed Jun 17 06:33:07 2009 UTC

… …
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
31 # AnyEvent integration 34 # EV integration
32 open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!"; 35 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
33 my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
34 36
35 # Event integration 37 # Event integration
36 Event->io (fd => IO::AIO::poll_fileno, 38 Event->io (fd => IO::AIO::poll_fileno,
37 poll => 'r', 39 poll => 'r',
38 cb => \&IO::AIO::poll_cb); 40 cb => \&IO::AIO::poll_cb);
… …
50 \&IO::AIO::poll_cb); 52 \&IO::AIO::poll_cb);
51 53
52=head1 DESCRIPTION 54=head1 DESCRIPTION
53 55
54This module implements asynchronous I/O using whatever means your 56This module implements asynchronous I/O using whatever means your
55operating system supports. 57operating system supports. It is implemented as an interface to C<libeio>
58(L<http://software.schmorp.de/pkg/libeio.html>).
56 59
57Asynchronous means that operations that can normally block your program 60Asynchronous means that operations that can normally block your program
58(e.g. reading from disk) will be done asynchronously: the operation 61(e.g. reading from disk) will be done asynchronously: the operation
59will still block, but you can do something else in the meantime. This 62will still block, but you can do something else in the meantime. This
60is extremely useful for programs that need to stay interactive even 63is extremely useful for programs that need to stay interactive even
… …
64on a RAID volume or over NFS when you do a number of stat operations 67on a RAID volume or over NFS when you do a number of stat operations
65concurrently. 68concurrently.
66 69
67While most of this works on all types of file descriptors (for 70While most of this works on all types of file descriptors (for
68example sockets), using these functions on file descriptors that 71example sockets), using these functions on file descriptors that
69support nonblocking operation (again, sockets, pipes etc.) is very 72support nonblocking operation (again, sockets, pipes etc.) is
70inefficient. Use an event loop for that (such as the L<Event|Event> 73very inefficient. Use an event loop for that (such as the L<EV>
71module): IO::AIO will naturally fit into such an event loop itself. 74module): IO::AIO will naturally fit into such an event loop itself.
72 75
73In this version, a number of threads are started that execute your 76In this version, a number of threads are started that execute your
74requests and signal their completion. You don't need thread support 77requests and signal their completion. You don't need thread support
75in perl, and the threads created by this module will not be visible 78in perl, and the threads created by this module will not be visible
… …
85yourself, always call C<poll_cb> from within the same thread, or never 88yourself, always call C<poll_cb> from within the same thread, or never
86call C<poll_cb> (or other C<aio_> functions) recursively. 89call C<poll_cb> (or other C<aio_> functions) recursively.
87 90
88=head2 EXAMPLE 91=head2 EXAMPLE
89 92
90This is a simple example that uses the Event module and loads 93This is a simple example that uses the EV module and loads
91F</etc/passwd> asynchronously: 94F</etc/passwd> asynchronously:
92 95
93 use Fcntl; 96 use Fcntl;
94 use Event; 97 use EV;
95 use IO::AIO; 98 use IO::AIO;
96 99
97 # register the IO::AIO callback with Event 100 # register the IO::AIO callback with EV
98 Event->io (fd => IO::AIO::poll_fileno, 101 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
99 poll => 'r',
100 cb => \&IO::AIO::poll_cb);
101 102
102 # queue the request to open /etc/passwd 103 # queue the request to open /etc/passwd
103 aio_open "/etc/passwd", O_RDONLY, 0, sub { 104 aio_open "/etc/passwd", O_RDONLY, 0, sub {
104 my $fh = shift 105 my $fh = shift
105 or die "error while opening: $!"; 106 or die "error while opening: $!";
… …
117 118
118 # file contents now in $contents 119 # file contents now in $contents
119 print $contents; 120 print $contents;
120 121
121 # exit event loop and program 122 # exit event loop and program
122 Event::unloop; 123 EV::unloop;
123 }; 124 };
124 }; 125 };
125 126
126 # possibly queue up other requests, or open GUI windows, 127 # possibly queue up other requests, or open GUI windows,
127 # check for sockets etc. etc. 128 # check for sockets etc. etc.
128 129
129 # process events as long as there are some: 130 # process events as long as there are some:
130 Event::loop; 131 EV::loop;
131 132
132=head1 REQUEST ANATOMY AND LIFETIME 133=head1 REQUEST ANATOMY AND LIFETIME
133 134
134Every C<aio_*> function creates a request. which is a C data structure not 135Every C<aio_*> function creates a request. which is a C data structure not
135directly visible to Perl. 136directly visible to Perl.
… …
191use strict 'vars'; 192use strict 'vars';
192 193
193use base 'Exporter'; 194use base 'Exporter';
194 195
195BEGIN { 196BEGIN {
196 our $VERSION = '2.51'; 197 our $VERSION = '3.25';
197 198
198 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat 199 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close
199 aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink 200 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
200 aio_readlink aio_fsync aio_fdatasync aio_readahead aio_rename aio_link 201 aio_scandir aio_symlink aio_readlink aio_sync aio_fsync
202 aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead
203 aio_rename aio_link aio_move aio_copy aio_group
201 aio_move aio_copy aio_group aio_nop aio_mknod aio_load aio_rmtree aio_mkdir 204 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
202 aio_chown aio_chmod aio_utime aio_truncate); 205 aio_chmod aio_utime aio_truncate);
206
203 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice aio_block)); 207 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
204 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 208 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
205 min_parallel max_parallel max_idle 209 min_parallel max_parallel max_idle
206 nreqs nready npending nthreads 210 nreqs nready npending nthreads
207 max_poll_time max_poll_reqs); 211 max_poll_time max_poll_reqs
212 sendfile fadvise);
213
214 push @AIO_REQ, qw(aio_busy); # not exported
208 215
209 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 216 @IO::AIO::GRP::ISA = 'IO::AIO::REQ';
210 217
211 require XSLoader; 218 require XSLoader;
212 XSLoader::load ("IO::AIO", $VERSION); 219 XSLoader::load ("IO::AIO", $VERSION);
… …
219All the C<aio_*> calls are more or less thin wrappers around the syscall 226All the C<aio_*> calls are more or less thin wrappers around the syscall
220with the same name (sans C<aio_>). The arguments are similar or identical, 227with the same name (sans C<aio_>). The arguments are similar or identical,
221and they all accept an additional (and optional) C<$callback> argument 228and they all accept an additional (and optional) C<$callback> argument
222which must be a code reference. This code reference will get called with 229which must be a code reference. This code reference will get called with
223the syscall return code (e.g. most syscalls return C<-1> on error, unlike 230the syscall return code (e.g. most syscalls return C<-1> on error, unlike
224perl, which usually delivers "false") as it's sole argument when the given 231perl, which usually delivers "false") as its sole argument after the given
225syscall has been executed asynchronously. 232syscall has been executed asynchronously.
226 233
227All functions expecting a filehandle keep a copy of the filehandle 234All functions expecting a filehandle keep a copy of the filehandle
228internally until the request has finished. 235internally until the request has finished.
229 236
… …
243your pathnames to the locale (or other) encoding in effect in the user 250your pathnames to the locale (or other) encoding in effect in the user
244environment, d) use Glib::filename_from_unicode on unicode filenames or e) 251environment, d) use Glib::filename_from_unicode on unicode filenames or e)
245use something else to ensure your scalar has the correct contents. 252use something else to ensure your scalar has the correct contents.
246 253
247This works, btw. independent of the internal UTF-8 bit, which IO::AIO 254This works, btw. independent of the internal UTF-8 bit, which IO::AIO
248handles correctly wether it is set or not. 255handles correctly whether it is set or not.
249 256
250=over 4 257=over 4
251 258
252=item $prev_pri = aioreq_pri [$pri] 259=item $prev_pri = aioreq_pri [$pri]
253 260
… …
316 323
317Asynchronously close a file and call the callback with the result 324Asynchronously close a file and call the callback with the result
318code. 325code.
319 326
320Unfortunately, you can't do this to perl. Perl I<insists> very strongly on 327Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
321closing the file descriptor associated with the filehandle itself. Here is 328closing the file descriptor associated with the filehandle itself.
322what aio_close will try:
323 329
324 1. dup()licate the fd 330Therefore, C<aio_close> will not close the filehandle - instead it will
325 2. asynchronously close() the duplicated fd 331use dup2 to overwrite the file descriptor with the write-end of a pipe
326 3. dup()licate the fd once more 332(the pipe fd will be created on demand and will be cached).
327 4. let perl close() the filehandle
328 5. asynchronously close the duplicated fd
329 333
330The idea is that the first close() flushes stuff to disk that closing an 334Or in other words: the file descriptor will be closed, but it will not be
331fd will flush, so when perl closes the fd, nothing much will need to be 335free for reuse until the perl filehandle is closed.
332flushed. The second async. close() will then flush stuff to disk that
333closing the last fd to the file will flush.
334
335Just FYI, SuSv3 has this to say on close:
336
337 All outstanding record locks owned by the process on the file
338 associated with the file descriptor shall be removed.
339
340 If fildes refers to a socket, close() shall cause the socket to be
341 destroyed. ... close() shall block for up to the current linger
342 interval until all data is transmitted.
343 [this actually sounds like a specification bug, but who knows]
344
345And at least Linux additionally actually flushes stuff on every close,
346even when the file itself is still open.
347
348Sounds enourmously inefficient and complicated? Yes... please show me how
349to nuke perl's fd out of existence...
350 336
351=cut 337=cut
352 338
353sub aio_close($;$) {
354 aio_block {
355 my ($fh, $cb) = @_;
356
357 my $pri = aioreq_pri;
358 my $grp = aio_group $cb;
359
360 my $fd = fileno $fh;
361
362 defined $fd or Carp::croak "aio_close called with fd-less filehandle";
363
364 # if the dups fail we will simply get EBADF
365 my $fd2 = _dup $fd;
366 aioreq_pri $pri;
367 add $grp _aio_close $fd2, sub {
368 my $fd2 = _dup $fd;
369 close $fh;
370 aioreq_pri $pri;
371 add $grp _aio_close $fd2, sub {
372 $grp->result ($_[0]);
373 };
374 };
375
376 $grp
377 }
378}
379
380
381=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 339=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
382 340
383=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 341=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
384 342
385Reads or writes C<$length> bytes from the specified C<$fh> and C<$offset> 343Reads or writes C<$length> bytes from or to the specified C<$fh> and
386into the scalar given by C<$data> and offset C<$dataoffset> and calls the 344C<$offset> into the scalar given by C<$data> and offset C<$dataoffset>
387callback without the actual number of bytes read (or -1 on error, just 345and calls the callback without the actual number of bytes read (or -1 on
388like the syscall). 346error, just like the syscall).
347
348C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
349offset plus the actual number of bytes read.
389 350
390If C<$offset> is undefined, then the current file descriptor offset will 351If C<$offset> is undefined, then the current file descriptor offset will
391be used (and updated), otherwise the file descriptor offset will not be 352be used (and updated), otherwise the file descriptor offset will not be
392changed by these calls. 353changed by these calls.
393 354
394If C<$length> is undefined in C<aio_write>, use the remaining length of C<$data>. 355If C<$length> is undefined in C<aio_write>, use the remaining length of
356C<$data>.
395 357
396If C<$dataoffset> is less than zero, it will be counted from the end of 358If C<$dataoffset> is less than zero, it will be counted from the end of
397C<$data>. 359C<$data>.
398 360
399The C<$data> scalar I<MUST NOT> be modified in any way while the request 361The C<$data> scalar I<MUST NOT> be modified in any way while the request
… …
571 533
572Unlike the POSIX call of the same name, C<aio_readdir> reads an entire 534Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
573directory (i.e. opendir + readdir + closedir). The entries will not be 535directory (i.e. opendir + readdir + closedir). The entries will not be
574sorted, and will B<NOT> include the C<.> and C<..> entries. 536sorted, and will B<NOT> include the C<.> and C<..> entries.
575 537
576The callback a single argument which is either C<undef> or an array-ref 538The callback is passed a single argument which is either C<undef> or an
577with the filenames. 539array-ref with the filenames.
540
541
542=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
543
544Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune
545behaviour and output format. In case of an error, C<$entries> will be
546C<undef>.
547
548The flags are a combination of the following constants, ORed together (the
549flags will also be passed to the callback, possibly modified):
550
551=over 4
552
553=item IO::AIO::READDIR_DENTS
554
555When this flag is off, then the callback gets an arrayref with of names
556only (as with C<aio_readdir>), otherwise it gets an arrayref with
557C<[$name, $type, $inode]> arrayrefs, each describing a single directory
558entry in more detail.
559
560C<$name> is the name of the entry.
561
562C<$type> is one of the C<IO::AIO::DT_xxx> constants:
563
564C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
565C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
566C<IO::AIO::DT_WHT>.
567
568C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need to
569know, you have to run stat yourself. Also, for speed reasons, the C<$type>
570scalars are read-only: you can not modify them.
571
572C<$inode> is the inode number (which might not be exact on systems with 64
573bit inode numbers and 32 bit perls). This field has unspecified content on
574systems that do not deliver the inode information.
575
576=item IO::AIO::READDIR_DIRS_FIRST
577
578When this flag is set, then the names will be returned in an order where
579likely directories come first. This is useful when you need to quickly
580find directories, or you want to find all directories while avoiding to
581stat() each entry.
582
583If the system returns type information in readdir, then this is used
584to find directories directly. Otherwise, likely directories are files
585beginning with ".", or otherwise files with no dots, of which files with
586short names are tried first.
587
588=item IO::AIO::READDIR_STAT_ORDER
589
590When this flag is set, then the names will be returned in an order
591suitable for stat()'ing each one. That is, when you plan to stat()
592all files in the given directory, then the returned order will likely
593be fastest.
594
595If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, then
596the likely dirs come first, resulting in a less optimal stat order.
597
598=item IO::AIO::READDIR_FOUND_UNKNOWN
599
600This flag should not be set when calling C<aio_readdirx>. Instead, it
601is being set by C<aio_readdirx>, when any of the C<$type>'s found were
602C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all
603C<$type>'s are known, which can be used to speed up some algorithms.
604
605=back
578 606
579 607
580=item aio_load $path, $data, $callback->($status) 608=item aio_load $path, $data, $callback->($status)
581 609
582This is a composite request that tries to fully load the given file into 610This is a composite request that tries to fully load the given file into
583memory. Status is the same as with aio_read. 611memory. Status is the same as with aio_read.
584 612
585=cut 613=cut
586 614
587sub aio_load($$;$) { 615sub aio_load($$;$) {
588 aio_block {
589 my ($path, undef, $cb) = @_; 616 my ($path, undef, $cb) = @_;
590 my $data = \$_[1]; 617 my $data = \$_[1];
591 618
592 my $pri = aioreq_pri; 619 my $pri = aioreq_pri;
593 my $grp = aio_group $cb; 620 my $grp = aio_group $cb;
621
622 aioreq_pri $pri;
623 add $grp aio_open $path, O_RDONLY, 0, sub {
624 my $fh = shift
625 or return $grp->result (-1);
594 626
595 aioreq_pri $pri; 627 aioreq_pri $pri;
596 add $grp aio_open $path, O_RDONLY, 0, sub {
597 my $fh = shift
598 or return $grp->result (-1);
599
600 aioreq_pri $pri;
601 add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub { 628 add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
602 $grp->result ($_[0]); 629 $grp->result ($_[0]);
603 };
604 }; 630 };
605
606 $grp
607 } 631 };
632
633 $grp
608} 634}
609 635
610=item aio_copy $srcpath, $dstpath, $callback->($status) 636=item aio_copy $srcpath, $dstpath, $callback->($status)
611 637
612Try to copy the I<file> (directories not supported as either source or 638Try to copy the I<file> (directories not supported as either source or
613destination) from C<$srcpath> to C<$dstpath> and call the callback with 639destination) from C<$srcpath> to C<$dstpath> and call the callback with
614the C<0> (error) or C<-1> ok. 640the C<0> (error) or C<-1> ok.
615 641
616This is a composite request that it creates the destination file with 642This is a composite request that creates the destination file with
617mode 0200 and copies the contents of the source file into it using 643mode 0200 and copies the contents of the source file into it using
618C<aio_sendfile>, followed by restoring atime, mtime, access mode and 644C<aio_sendfile>, followed by restoring atime, mtime, access mode and
619uid/gid, in that order. 645uid/gid, in that order.
620 646
621If an error occurs, the partial destination file will be unlinked, if 647If an error occurs, the partial destination file will be unlinked, if
… …
623errors are being ignored. 649errors are being ignored.
624 650
625=cut 651=cut
626 652
627sub aio_copy($$;$) { 653sub aio_copy($$;$) {
628 aio_block {
629 my ($src, $dst, $cb) = @_; 654 my ($src, $dst, $cb) = @_;
630 655
631 my $pri = aioreq_pri; 656 my $pri = aioreq_pri;
632 my $grp = aio_group $cb; 657 my $grp = aio_group $cb;
633 658
634 aioreq_pri $pri; 659 aioreq_pri $pri;
635 add $grp aio_open $src, O_RDONLY, 0, sub { 660 add $grp aio_open $src, O_RDONLY, 0, sub {
636 if (my $src_fh = $_[0]) { 661 if (my $src_fh = $_[0]) {
637 my @stat = stat $src_fh; 662 my @stat = stat $src_fh; # hmm, might bock over nfs?
638 663
639 aioreq_pri $pri; 664 aioreq_pri $pri;
640 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub { 665 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
641 if (my $dst_fh = $_[0]) { 666 if (my $dst_fh = $_[0]) {
642 aioreq_pri $pri; 667 aioreq_pri $pri;
643 add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub { 668 add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
644 if ($_[0] == $stat[7]) { 669 if ($_[0] == $stat[7]) {
645 $grp->result (0); 670 $grp->result (0);
646 close $src_fh; 671 close $src_fh;
647 672
648 # those should not normally block. should. should.
649 utime $stat[8], $stat[9], $dst;
650 chmod $stat[2] & 07777, $dst_fh;
651 chown $stat[4], $stat[5], $dst_fh;
652 close $dst_fh;
653 } else { 673 my $ch = sub {
654 $grp->result (-1);
655 close $src_fh;
656 close $dst_fh;
657
658 aioreq $pri; 674 aioreq_pri $pri;
675 add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
676 aioreq_pri $pri;
677 add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
678 aioreq_pri $pri;
659 add $grp aio_unlink $dst; 679 add $grp aio_close $dst_fh;
680 }
681 };
660 } 682 };
683
684 aioreq_pri $pri;
685 add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
686 if ($_[0] < 0 && $! == ENOSYS) {
687 aioreq_pri $pri;
688 add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
689 } else {
690 $ch->();
691 }
692 };
693 } else {
694 $grp->result (-1);
695 close $src_fh;
696 close $dst_fh;
697
698 aioreq $pri;
699 add $grp aio_unlink $dst;
661 }; 700 }
662 } else {
663 $grp->result (-1);
664 } 701 };
702 } else {
703 $grp->result (-1);
665 }, 704 }
666
667 } else {
668 $grp->result (-1);
669 } 705 },
706
707 } else {
708 $grp->result (-1);
670 }; 709 }
671
672 $grp
673 } 710 };
711
712 $grp
674} 713}
675 714
676=item aio_move $srcpath, $dstpath, $callback->($status) 715=item aio_move $srcpath, $dstpath, $callback->($status)
677 716
678Try to move the I<file> (directories not supported as either source or 717Try to move the I<file> (directories not supported as either source or
679destination) from C<$srcpath> to C<$dstpath> and call the callback with 718destination) from C<$srcpath> to C<$dstpath> and call the callback with
680the C<0> (error) or C<-1> ok. 719the C<0> (error) or C<-1> ok.
681 720
682This is a composite request that tries to rename(2) the file first. If 721This is a composite request that tries to rename(2) the file first; if
683rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if 722rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
684that is successful, unlinking the C<$srcpath>. 723that is successful, unlinks the C<$srcpath>.
685 724
686=cut 725=cut
687 726
688sub aio_move($$;$) { 727sub aio_move($$;$) {
689 aio_block {
690 my ($src, $dst, $cb) = @_; 728 my ($src, $dst, $cb) = @_;
691 729
692 my $pri = aioreq_pri; 730 my $pri = aioreq_pri;
693 my $grp = aio_group $cb; 731 my $grp = aio_group $cb;
694 732
695 aioreq_pri $pri; 733 aioreq_pri $pri;
696 add $grp aio_rename $src, $dst, sub { 734 add $grp aio_rename $src, $dst, sub {
697 if ($_[0] && $! == EXDEV) { 735 if ($_[0] && $! == EXDEV) {
698 aioreq_pri $pri; 736 aioreq_pri $pri;
699 add $grp aio_copy $src, $dst, sub { 737 add $grp aio_copy $src, $dst, sub {
700 $grp->result ($_[0]);
701
702 if (!$_[0]) {
703 aioreq_pri $pri;
704 add $grp aio_unlink $src;
705 }
706 };
707 } else {
708 $grp->result ($_[0]); 738 $grp->result ($_[0]);
739
740 if (!$_[0]) {
741 aioreq_pri $pri;
742 add $grp aio_unlink $src;
743 }
709 } 744 };
745 } else {
746 $grp->result ($_[0]);
710 }; 747 }
711
712 $grp
713 } 748 };
749
750 $grp
714} 751}
715 752
716=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 753=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
717 754
718Scans a directory (similar to C<aio_readdir>) but additionally tries to 755Scans a directory (similar to C<aio_readdir>) but additionally tries to
… …
738 775
739Implementation notes. 776Implementation notes.
740 777
741The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can. 778The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
742 779
780If readdir returns file type information, then this is used directly to
781find directories.
782
743After reading the directory, the modification time, size etc. of the 783Otherwise, after reading the directory, the modification time, size etc.
744directory before and after the readdir is checked, and if they match (and 784of the directory before and after the readdir is checked, and if they
745isn't the current time), the link count will be used to decide how many 785match (and isn't the current time), the link count will be used to decide
746entries are directories (if >= 2). Otherwise, no knowledge of the number 786how many entries are directories (if >= 2). Otherwise, no knowledge of the
747of subdirectories will be assumed. 787number of subdirectories will be assumed.
748 788
749Then entries will be sorted into likely directories (everything without 789Then entries will be sorted into likely directories a non-initial dot
750a non-initial dot currently) and likely non-directories (everything 790currently) and likely non-directories (see C<aio_readdirx>). Then every
751else). Then every entry plus an appended C</.> will be C<stat>'ed, 791entry plus an appended C</.> will be C<stat>'ed, likely directories first,
752likely directories first. If that succeeds, it assumes that the entry 792in order of their inode numbers. If that succeeds, it assumes that the
753is a directory or a symlink to directory (which will be checked 793entry is a directory or a symlink to directory (which will be checked
754seperately). This is often faster than stat'ing the entry itself because 794seperately). This is often faster than stat'ing the entry itself because
755filesystems might detect the type of the entry without reading the inode 795filesystems might detect the type of the entry without reading the inode
756data (e.g. ext2fs filetype feature). 796data (e.g. ext2fs filetype feature), even on systems that cannot return
797the filetype information on readdir.
757 798
758If the known number of directories (link count - 2) has been reached, the 799If the known number of directories (link count - 2) has been reached, the
759rest of the entries is assumed to be non-directories. 800rest of the entries is assumed to be non-directories.
760 801
761This only works with certainty on POSIX (= UNIX) filesystems, which 802This only works with certainty on POSIX (= UNIX) filesystems, which
… …
766directory counting heuristic. 807directory counting heuristic.
767 808
768=cut 809=cut
769 810
770sub aio_scandir($$;$) { 811sub aio_scandir($$;$) {
771 aio_block {
772 my ($path, $maxreq, $cb) = @_; 812 my ($path, $maxreq, $cb) = @_;
773 813
774 my $pri = aioreq_pri; 814 my $pri = aioreq_pri;
775 815
776 my $grp = aio_group $cb; 816 my $grp = aio_group $cb;
777 817
778 $maxreq = 4 if $maxreq <= 0; 818 $maxreq = 4 if $maxreq <= 0;
779 819
780 # stat once 820 # stat once
821 aioreq_pri $pri;
822 add $grp aio_stat $path, sub {
823 return $grp->result () if $_[0];
824 my $now = time;
825 my $hash1 = join ":", (stat _)[0,1,3,7,9];
826
827 # read the directory entries
781 aioreq_pri $pri; 828 aioreq_pri $pri;
782 add $grp aio_stat $path, sub { 829 add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub {
830 my $entries = shift
783 return $grp->result () if $_[0]; 831 or return $grp->result ();
784 my $now = time;
785 my $hash1 = join ":", (stat _)[0,1,3,7,9];
786 832
787 # read the directory entries 833 # stat the dir another time
788 aioreq_pri $pri; 834 aioreq_pri $pri;
789 add $grp aio_readdir $path, sub {
790 my $entries = shift
791 or return $grp->result ();
792
793 # stat the dir another time
794 aioreq_pri $pri;
795 add $grp aio_stat $path, sub { 835 add $grp aio_stat $path, sub {
796 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 836 my $hash2 = join ":", (stat _)[0,1,3,7,9];
797 837
798 my $ndirs; 838 my $ndirs;
799 839
800 # take the slow route if anything looks fishy 840 # take the slow route if anything looks fishy
801 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 841 if ($hash1 ne $hash2 or (stat _)[9] == $now) {
802 $ndirs = -1; 842 $ndirs = -1;
803 } else { 843 } else {
804 # if nlink == 2, we are finished 844 # if nlink == 2, we are finished
805 # on non-posix-fs's, we rely on nlink < 2 845 # for non-posix-fs's, we rely on nlink < 2
806 $ndirs = (stat _)[3] - 2 846 $ndirs = (stat _)[3] - 2
807 or return $grp->result ([], $entries); 847 or return $grp->result ([], $entries);
808 } 848 }
809 849
810 # sort into likely dirs and likely nondirs
811 # dirs == files without ".", short entries first
812 $entries = [map $_->[0],
813 sort { $b->[1] cmp $a->[1] }
814 map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
815 @$entries];
816
817 my (@dirs, @nondirs); 850 my (@dirs, @nondirs);
818 851
819 my $statgrp = add $grp aio_group sub { 852 my $statgrp = add $grp aio_group sub {
820 $grp->result (\@dirs, \@nondirs); 853 $grp->result (\@dirs, \@nondirs);
821 }; 854 };
822 855
823 limit $statgrp $maxreq; 856 limit $statgrp $maxreq;
824 feed $statgrp sub { 857 feed $statgrp sub {
825 return unless @$entries; 858 return unless @$entries;
826 my $entry = pop @$entries; 859 my $entry = shift @$entries;
827 860
828 aioreq_pri $pri; 861 aioreq_pri $pri;
829 add $statgrp aio_stat "$path/$entry/.", sub { 862 add $statgrp aio_stat "$path/$entry/.", sub {
830 if ($_[0] < 0) { 863 if ($_[0] < 0) {
831 push @nondirs, $entry; 864 push @nondirs, $entry;
832 } else { 865 } else {
833 # need to check for real directory 866 # need to check for real directory
834 aioreq_pri $pri; 867 aioreq_pri $pri;
835 add $statgrp aio_lstat "$path/$entry", sub { 868 add $statgrp aio_lstat "$path/$entry", sub {
836 if (-d _) { 869 if (-d _) {
837 push @dirs, $entry; 870 push @dirs, $entry;
838 871
839 unless (--$ndirs) { 872 unless (--$ndirs) {
840 push @nondirs, @$entries; 873 push @nondirs, @$entries;
841 feed $statgrp; 874 feed $statgrp;
842 }
843 } else {
844 push @nondirs, $entry;
845 } 875 }
876 } else {
877 push @nondirs, $entry;
846 } 878 }
847 } 879 }
848 }; 880 }
849 }; 881 };
850 }; 882 };
851 }; 883 };
852 }; 884 };
853
854 $grp
855 } 885 };
886
887 $grp
856} 888}
857 889
858=item aio_rmtree $path, $callback->($status) 890=item aio_rmtree $path, $callback->($status)
859 891
860Delete a directory tree starting (and including) C<$path>, return the 892Delete a directory tree starting (and including) C<$path>, return the
… …
864 896
865=cut 897=cut
866 898
867sub aio_rmtree; 899sub aio_rmtree;
868sub aio_rmtree($;$) { 900sub aio_rmtree($;$) {
869 aio_block {
870 my ($path, $cb) = @_; 901 my ($path, $cb) = @_;
871 902
872 my $pri = aioreq_pri; 903 my $pri = aioreq_pri;
873 my $grp = aio_group $cb; 904 my $grp = aio_group $cb;
874 905
875 aioreq_pri $pri; 906 aioreq_pri $pri;
876 add $grp aio_scandir $path, 0, sub { 907 add $grp aio_scandir $path, 0, sub {
877 my ($dirs, $nondirs) = @_; 908 my ($dirs, $nondirs) = @_;
878 909
879 my $dirgrp = aio_group sub { 910 my $dirgrp = aio_group sub {
880 add $grp aio_rmdir $path, sub { 911 add $grp aio_rmdir $path, sub {
881 $grp->result ($_[0]); 912 $grp->result ($_[0]);
882 };
883 }; 913 };
884
885 (aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
886 (aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
887
888 add $grp $dirgrp;
889 }; 914 };
890 915
891 $grp 916 (aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
917 (aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
918
919 add $grp $dirgrp;
892 } 920 };
921
922 $grp
893} 923}
924
925=item aio_sync $callback->($status)
926
927Asynchronously call sync and call the callback when finished.
894 928
895=item aio_fsync $fh, $callback->($status) 929=item aio_fsync $fh, $callback->($status)
896 930
897Asynchronously call fsync on the given filehandle and call the callback 931Asynchronously call fsync on the given filehandle and call the callback
898with the fsync result code. 932with the fsync result code.
… …
902Asynchronously call fdatasync on the given filehandle and call the 936Asynchronously call fdatasync on the given filehandle and call the
903callback with the fdatasync result code. 937callback with the fdatasync result code.
904 938
905If this call isn't available because your OS lacks it or it couldn't be 939If this call isn't available because your OS lacks it or it couldn't be
906detected, it will be emulated by calling C<fsync> instead. 940detected, it will be emulated by calling C<fsync> instead.
941
942=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
943
944Sync the data portion of the file specified by C<$offset> and C<$length>
945to disk (but NOT the metadata), by calling the Linux-specific
946sync_file_range call. If sync_file_range is not available or it returns
947ENOSYS, then fdatasync or fsync is being substituted.
948
949C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
950C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
951C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
952manpage for details.
953
954=item aio_pathsync $path, $callback->($status)
955
956This request tries to open, fsync and close the given path. This is a
957composite request intended to sync directories after directory operations
958(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
960written to disc. It works for anything that can be opened for read-only,
961not just directories.
962
963Passes C<0> when everything went ok, and C<-1> on error.
964
965=cut
966
967sub aio_pathsync($;$) {
968 my ($path, $cb) = @_;
969
970 my $pri = aioreq_pri;
971 my $grp = aio_group $cb;
972
973 aioreq_pri $pri;
974 add $grp aio_open $path, O_RDONLY, 0, sub {
975 my ($fh) = @_;
976 if ($fh) {
977 aioreq_pri $pri;
978 add $grp aio_fsync $fh, sub {
979 $grp->result ($_[0]);
980
981 aioreq_pri $pri;
982 add $grp aio_close $fh;
983 };
984 } else {
985 $grp->result (-1);
986 }
987 };
988
989 $grp
990}
907 991
908=item aio_group $callback->(...) 992=item aio_group $callback->(...)
909 993
910This is a very special aio request: Instead of doing something, it is a 994This is a very special aio request: Instead of doing something, it is a
911container for other aio requests, which is useful if you want to bundle 995container for other aio requests, which is useful if you want to bundle
… …
961=item cancel $req 1045=item cancel $req
962 1046
963Cancels the request, if possible. Has the effect of skipping execution 1047Cancels the request, if possible. Has the effect of skipping execution
964when entering the B<execute> state and skipping calling the callback when 1048when entering the B<execute> state and skipping calling the callback when
965entering the the B<result> state, but will leave the request otherwise 1049entering the the B<result> state, but will leave the request otherwise
966untouched. That means that requests that currently execute will not be 1050untouched (with the exception of readdir). That means that requests that
967stopped and resources held by the request will not be freed prematurely. 1051currently execute will not be stopped and resources held by the request
1052will not be freed prematurely.
968 1053
969=item cb $req $callback->(...) 1054=item cb $req $callback->(...)
970 1055
971Replace (or simply set) the callback registered to the request. 1056Replace (or simply set) the callback registered to the request.
972 1057
… …
1023Their lifetime, simplified, looks like this: when they are empty, they 1108Their lifetime, simplified, looks like this: when they are empty, they
1024will finish very quickly. If they contain only requests that are in the 1109will finish very quickly. If they contain only requests that are in the
1025C<done> state, they will also finish. Otherwise they will continue to 1110C<done> state, they will also finish. Otherwise they will continue to
1026exist. 1111exist.
1027 1112
1028That means after creating a group you have some time to add requests. And 1113That means after creating a group you have some time to add requests
1029in the callbacks of those requests, you can add further requests to the 1114(precisely before the callback has been invoked, which is only done within
1030group. And only when all those requests have finished will the the group 1115the C<poll_cb>). And in the callbacks of those requests, you can add
1031itself finish. 1116further requests to the group. And only when all those requests have
1117finished will the the group itself finish.
1032 1118
1033=over 4 1119=over 4
1034 1120
1035=item add $grp ... 1121=item add $grp ...
1036 1122
… …
1048itself. Useful when you queued a lot of events but got a result early. 1134itself. Useful when you queued a lot of events but got a result early.
1049 1135
1050=item $grp->result (...) 1136=item $grp->result (...)
1051 1137
1052Set the result value(s) that will be passed to the group callback when all 1138Set the result value(s) that will be passed to the group callback when all
1053subrequests have finished and set thre groups errno to the current value 1139subrequests have finished and set the groups errno to the current value
1054of errno (just like calling C<errno> without an error number). By default, 1140of errno (just like calling C<errno> without an error number). By default,
1055no argument will be passed and errno is zero. 1141no argument will be passed and errno is zero.
1056 1142
1057=item $grp->errno ([$errno]) 1143=item $grp->errno ([$errno])
1058 1144
… …
1069=item feed $grp $callback->($grp) 1155=item feed $grp $callback->($grp)
1070 1156
1071Sets a feeder/generator on this group: every group can have an attached 1157Sets a feeder/generator on this group: every group can have an attached
1072generator that generates requests if idle. The idea behind this is that, 1158generator that generates requests if idle. The idea behind this is that,
1073although you could just queue as many requests as you want in a group, 1159although you could just queue as many requests as you want in a group,
1074this might starve other requests for a potentially long time. For 1160this might starve other requests for a potentially long time. For example,
1075example, C<aio_scandir> might generate hundreds of thousands C<aio_stat> 1161C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests,
1076requests, delaying any later requests for a long time. 1162delaying any later requests for a long time.
1077 1163
1078To avoid this, and allow incremental generation of requests, you can 1164To avoid this, and allow incremental generation of requests, you can
1079instead a group and set a feeder on it that generates those requests. The 1165instead a group and set a feeder on it that generates those requests. The
1080feed callback will be called whenever there are few enough (see C<limit>, 1166feed callback will be called whenever there are few enough (see C<limit>,
1081below) requests active in the group itself and is expected to queue more 1167below) requests active in the group itself and is expected to queue more
… …
1085not impose any limits). 1171not impose any limits).
1086 1172
1087If the feed does not queue more requests when called, it will be 1173If the feed does not queue more requests when called, it will be
1088automatically removed from the group. 1174automatically removed from the group.
1089 1175
1090If the feed limit is C<0>, it will be set to C<2> automatically. 1176If the feed limit is C<0> when this method is called, it will be set to
1177C<2> automatically.
1091 1178
1092Example: 1179Example:
1093 1180
1094 # stat all files in @files, but only ever use four aio requests concurrently: 1181 # stat all files in @files, but only ever use four aio requests concurrently:
1095 1182
… …
1107Sets the feeder limit for the group: The feeder will be called whenever 1194Sets the feeder limit for the group: The feeder will be called whenever
1108the group contains less than this many requests. 1195the group contains less than this many requests.
1109 1196
1110Setting the limit to C<0> will pause the feeding process. 1197Setting the limit to C<0> will pause the feeding process.
1111 1198
1199The default value for the limit is C<0>, but note that setting a feeder
1200automatically bumps it up to C<2>.
1201
1112=back 1202=back
1113 1203
1114=head2 SUPPORT FUNCTIONS 1204=head2 SUPPORT FUNCTIONS
1115 1205
1116=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION 1206=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
… …
1118=over 4 1208=over 4
1119 1209
1120=item $fileno = IO::AIO::poll_fileno 1210=item $fileno = IO::AIO::poll_fileno
1121 1211
1122Return the I<request result pipe file descriptor>. This filehandle must be 1212Return the I<request result pipe file descriptor>. This filehandle must be
1123polled for reading by some mechanism outside this module (e.g. Event or 1213polled for reading by some mechanism outside this module (e.g. EV, Glib,
1124select, see below or the SYNOPSIS). If the pipe becomes readable you have 1214select and so on, see below or the SYNOPSIS). If the pipe becomes readable
1125to call C<poll_cb> to check the results. 1215you have to call C<poll_cb> to check the results.
1126 1216
1127See C<poll_cb> for an example. 1217See C<poll_cb> for an example.
1128 1218
1129=item IO::AIO::poll_cb 1219=item IO::AIO::poll_cb
1130 1220
1131Process some outstanding events on the result pipe. You have to call this 1221Process some outstanding events on the result pipe. You have to call this
1132regularly. Returns the number of events processed. Returns immediately 1222regularly. Returns C<0> if all events could be processed, or C<-1> if it
1223returned earlier for whatever reason. Returns immediately when no events
1133when no events are outstanding. The amount of events processed depends on 1224are outstanding. The amount of events processed depends on the settings of
1134the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>. 1225C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
1135 1226
1136If not all requests were processed for whatever reason, the filehandle 1227If not all requests were processed for whatever reason, the filehandle
1137will still be ready when C<poll_cb> returns. 1228will still be ready when C<poll_cb> returns, so normally you don't have to
1229do anything special to have it called later.
1138 1230
1139Example: Install an Event watcher that automatically calls 1231Example: Install an Event watcher that automatically calls
1140IO::AIO::poll_cb with high priority: 1232IO::AIO::poll_cb with high priority (more examples can be found in the
1233SYNOPSIS section, at the top of this document):
1141 1234
1142 Event->io (fd => IO::AIO::poll_fileno, 1235 Event->io (fd => IO::AIO::poll_fileno,
1143 poll => 'r', async => 1, 1236 poll => 'r', async => 1,
1144 cb => \&IO::AIO::poll_cb); 1237 cb => \&IO::AIO::poll_cb);
1145 1238
… …
1256 1349
1257The default is probably ok in most situations, especially if thread 1350The default is probably ok in most situations, especially if thread
1258creation is fast. If thread creation is very slow on your system you might 1351creation is fast. If thread creation is very slow on your system you might
1259want to use larger values. 1352want to use larger values.
1260 1353
1261=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 1354=item IO::AIO::max_outstanding $maxreqs
1262 1355
1263This is a very bad function to use in interactive programs because it 1356This is a very bad function to use in interactive programs because it
1264blocks, and a bad way to reduce concurrency because it is inexact: Better 1357blocks, and a bad way to reduce concurrency because it is inexact: Better
1265use an C<aio_group> together with a feed callback. 1358use an C<aio_group> together with a feed callback.
1266 1359
… …
1271 1364
1272The default value is very large, so there is no practical limit on the 1365The default value is very large, so there is no practical limit on the
1273number of outstanding requests. 1366number of outstanding requests.
1274 1367
1275You can still queue as many requests as you want. Therefore, 1368You can still queue as many requests as you want. Therefore,
1276C<max_oustsanding> is mainly useful in simple scripts (with low values) or 1369C<max_outstanding> is mainly useful in simple scripts (with low values) or
1277as a stop gap to shield against fatal memory overflow (with large values). 1370as a stop gap to shield against fatal memory overflow (with large values).
1278 1371
1279=back 1372=back
1280 1373
1281=head3 STATISTICAL INFORMATION 1374=head3 STATISTICAL INFORMATION
… …
1299 1392
1300=item IO::AIO::npending 1393=item IO::AIO::npending
1301 1394
1302Returns the number of requests currently in the pending state (executed, 1395Returns the number of requests currently in the pending state (executed,
1303but not yet processed by poll_cb). 1396but not yet processed by poll_cb).
1397
1398=back
1399
1400=head3 MISCELLANEOUS FUNCTIONS
1401
1402IO::AIO implements some functions that might be useful, but are not
1403asynchronous.
1404
1405=over 4
1406
1407=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
1408
1409Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
1410but is blocking (this makes most sense if you know the input data is
1411likely cached already and the output filehandle is set to non-blocking
1412operations).
1413
1414Returns the number of bytes copied, or C<-1> on error.
1415
1416=item IO::AIO::fadvise $fh, $offset, $len, $advice
1417
1418Simply calls the C<posix_fadvise> function (see it's
1419manpage for details). The following advice constants are
1420avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
1421C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
1422C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
1423
1424On systems that do not implement C<posix_fadvise>, this function returns
1425ENOSYS, otherwise the return value of C<posix_fadvise>.
1304 1426
1305=back 1427=back
1306 1428
1307=cut 1429=cut
1308 1430
… …
1351 1473
1352Known bugs will be fixed in the next release. 1474Known bugs will be fixed in the next release.
1353 1475
1354=head1 SEE ALSO 1476=head1 SEE ALSO
1355 1477
1356L<Coro::AIO>. 1478L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
1479more natural syntax.
1357 1480
1358=head1 AUTHOR 1481=head1 AUTHOR
1359 1482
1360 Marc Lehmann <schmorp@schmorp.de> 1483 Marc Lehmann <schmorp@schmorp.de>
1361 http://home.schmorp.de/ 1484 http://home.schmorp.de/

Diff Legend

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