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 1.62 by root, Sun Oct 22 21:13:47 2006 UTC vs.
Revision 1.101 by root, Sun Jan 7 22:59:57 2007 UTC

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", O_RDONLY, 0, sub {
10 my ($fh) = @_; 10 my $fh = shift
11 or die "/etc/passwd: $!";
11 ... 12 ...
12 }; 13 };
13 14
14 aio_unlink "/tmp/file", sub { }; 15 aio_unlink "/tmp/file", sub { };
15 16
18 }; 19 };
19 20
20 # version 2+ has request and group objects 21 # version 2+ has request and group objects
21 use IO::AIO 2; 22 use IO::AIO 2;
22 23
24 aioreq_pri 4; # give next request a very high priority
23 my $req = aio_unlink "/tmp/file", sub { }; 25 my $req = aio_unlink "/tmp/file", sub { };
24 $req->cancel; # cancel request if still in queue 26 $req->cancel; # cancel request if still in queue
25 27
26 my $grp = aio_group sub { print "all stats done\n" }; 28 my $grp = aio_group sub { print "all stats done\n" };
27 add $grp aio_stat "..." for ...; 29 add $grp aio_stat "..." for ...;
50=head1 DESCRIPTION 52=head1 DESCRIPTION
51 53
52This module implements asynchronous I/O using whatever means your 54This module implements asynchronous I/O using whatever means your
53operating system supports. 55operating system supports.
54 56
57Asynchronous means that operations that can normally block your program
58(e.g. reading from disk) will be done asynchronously: the operation
59will still block, but you can do something else in the meantime. This
60is extremely useful for programs that need to stay interactive even
61when doing heavy I/O (GUI programs, high performance network servers
62etc.), but can also be used to easily do operations in parallel that are
63normally done sequentially, e.g. stat'ing many files, which is much faster
64on a RAID volume or over NFS when you do a number of stat operations
65concurrently.
66
67While most of this works on all types of file descriptors (for example
68sockets), using these functions on file descriptors that support
69nonblocking operation (again, sockets, pipes etc.) is very inefficient or
70might not work (aio_read fails on sockets/pipes/fifos). Use an event loop
71for that (such as the L<Event|Event> module): IO::AIO will naturally fit
72into such an event loop itself.
73
55Currently, a number of threads are started that execute your read/writes 74In this version, a number of threads are started that execute your
56and signal their completion. You don't need thread support in your libc or 75requests and signal their completion. You don't need thread support
57perl, and the threads created by this module will not be visible to the 76in perl, and the threads created by this module will not be visible
58pthreads library. In the future, this module might make use of the native 77to perl. In the future, this module might make use of the native aio
59aio functions available on many operating systems. However, they are often 78functions available on many operating systems. However, they are often
60not well-supported (Linux doesn't allow them on normal files currently, 79not well-supported or restricted (GNU/Linux doesn't allow them on normal
61for example), and they would only support aio_read and aio_write, so the 80files currently, for example), and they would only support aio_read and
62remaining functionality would have to be implemented using threads anyway. 81aio_write, so the remaining functionality would have to be implemented
82using threads anyway.
63 83
64Although the module will work with in the presence of other threads, it is 84Although the module will work with in the presence of other (Perl-)
65currently not reentrant, so use appropriate locking yourself, always call 85threads, it is currently not reentrant in any way, so use appropriate
66C<poll_cb> from within the same thread, or never call C<poll_cb> (or other 86locking yourself, always call C<poll_cb> from within the same thread, or
67C<aio_> functions) recursively. 87never call C<poll_cb> (or other C<aio_> functions) recursively.
88
89=head2 EXAMPLE
90
91This is a simple example that uses the Event module and loads
92F</etc/passwd> asynchronously:
93
94 use Fcntl;
95 use Event;
96 use IO::AIO;
97
98 # register the IO::AIO callback with Event
99 Event->io (fd => IO::AIO::poll_fileno,
100 poll => 'r',
101 cb => \&IO::AIO::poll_cb);
102
103 # queue the request to open /etc/passwd
104 aio_open "/etc/passwd", O_RDONLY, 0, sub {
105 my $fh = shift
106 or die "error while opening: $!";
107
108 # stat'ing filehandles is generally non-blocking
109 my $size = -s $fh;
110
111 # queue a request to read the file
112 my $contents;
113 aio_read $fh, 0, $size, $contents, 0, sub {
114 $_[0] == $size
115 or die "short read: $!";
116
117 close $fh;
118
119 # file contents now in $contents
120 print $contents;
121
122 # exit event loop and program
123 Event::unloop;
124 };
125 };
126
127 # possibly queue up other requests, or open GUI windows,
128 # check for sockets etc. etc.
129
130 # process events as long as there are some:
131 Event::loop;
132
133=head1 REQUEST ANATOMY AND LIFETIME
134
135Every C<aio_*> function creates a request. which is a C data structure not
136directly visible to Perl.
137
138If called in non-void context, every request function returns a Perl
139object representing the request. In void context, nothing is returned,
140which saves a bit of memory.
141
142The perl object is a fairly standard ref-to-hash object. The hash contents
143are not used by IO::AIO so you are free to store anything you like in it.
144
145During their existance, aio requests travel through the following states,
146in order:
147
148=over 4
149
150=item ready
151
152Immediately after a request is created it is put into the ready state,
153waiting for a thread to execute it.
154
155=item execute
156
157A thread has accepted the request for processing and is currently
158executing it (e.g. blocking in read).
159
160=item pending
161
162The request has been executed and is waiting for result processing.
163
164While request submission and execution is fully asynchronous, result
165processing is not and relies on the perl interpreter calling C<poll_cb>
166(or another function with the same effect).
167
168=item result
169
170The request results are processed synchronously by C<poll_cb>.
171
172The C<poll_cb> function will process all outstanding aio requests by
173calling their callbacks, freeing memory associated with them and managing
174any groups they are contained in.
175
176=item done
177
178Request has reached the end of its lifetime and holds no resources anymore
179(except possibly for the Perl object, but its connection to the actual
180aio request is severed and calling its methods will either do nothing or
181result in a runtime error).
182
183=back
68 184
69=cut 185=cut
70 186
71package IO::AIO; 187package IO::AIO;
72 188
74use strict 'vars'; 190use strict 'vars';
75 191
76use base 'Exporter'; 192use base 'Exporter';
77 193
78BEGIN { 194BEGIN {
79 our $VERSION = '2.0'; 195 our $VERSION = '2.32';
80 196
81 our @EXPORT = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat 197 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat
82 aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink 198 aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink
83 aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move 199 aio_readlink aio_fsync aio_fdatasync aio_readahead aio_rename aio_link
84 aio_group); 200 aio_move aio_copy aio_group aio_nop aio_mknod aio_load aio_rmtree aio_mkdir);
85 our @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs); 201 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice aio_block));
202 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
203 min_parallel max_parallel max_idle
204 nreqs nready npending nthreads
205 max_poll_time max_poll_reqs);
86 206
87 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 207 @IO::AIO::GRP::ISA = 'IO::AIO::REQ';
88 208
89 require XSLoader; 209 require XSLoader;
90 XSLoader::load ("IO::AIO", $VERSION); 210 XSLoader::load ("IO::AIO", $VERSION);
91} 211}
92 212
93=head1 FUNCTIONS 213=head1 FUNCTIONS
94 214
95=head2 AIO FUNCTIONS 215=head2 AIO REQUEST FUNCTIONS
96 216
97All the C<aio_*> calls are more or less thin wrappers around the syscall 217All the C<aio_*> calls are more or less thin wrappers around the syscall
98with the same name (sans C<aio_>). The arguments are similar or identical, 218with the same name (sans C<aio_>). The arguments are similar or identical,
99and they all accept an additional (and optional) C<$callback> argument 219and they all accept an additional (and optional) C<$callback> argument
100which must be a code reference. This code reference will get called with 220which must be a code reference. This code reference will get called with
103syscall has been executed asynchronously. 223syscall has been executed asynchronously.
104 224
105All functions expecting a filehandle keep a copy of the filehandle 225All functions expecting a filehandle keep a copy of the filehandle
106internally until the request has finished. 226internally until the request has finished.
107 227
108All requests return objects of type L<IO::AIO::REQ> that allow further 228All functions return request objects of type L<IO::AIO::REQ> that allow
109manipulation of those requests while they are in-flight. 229further manipulation of those requests while they are in-flight.
110 230
111The pathnames you pass to these routines I<must> be absolute and 231The pathnames you pass to these routines I<must> be absolute and
112encoded in byte form. The reason for the former is that at the time the 232encoded as octets. The reason for the former is that at the time the
113request is being executed, the current working directory could have 233request is being executed, the current working directory could have
114changed. Alternatively, you can make sure that you never change the 234changed. Alternatively, you can make sure that you never change the
115current working directory. 235current working directory anywhere in the program and then use relative
236paths.
116 237
117To encode pathnames to byte form, either make sure you either: a) 238To encode pathnames as octets, either make sure you either: a) always pass
118always pass in filenames you got from outside (command line, readdir 239in filenames you got from outside (command line, readdir etc.) without
119etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode 240tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
120your pathnames to the locale (or other) encoding in effect in the user 241your pathnames to the locale (or other) encoding in effect in the user
121environment, d) use Glib::filename_from_unicode on unicode filenames or e) 242environment, d) use Glib::filename_from_unicode on unicode filenames or e)
122use something else. 243use something else to ensure your scalar has the correct contents.
244
245This works, btw. independent of the internal UTF-8 bit, which IO::AIO
246handles correctly wether it is set or not.
123 247
124=over 4 248=over 4
249
250=item $prev_pri = aioreq_pri [$pri]
251
252Returns the priority value that would be used for the next request and, if
253C<$pri> is given, sets the priority for the next aio request.
254
255The default priority is C<0>, the minimum and maximum priorities are C<-4>
256and C<4>, respectively. Requests with higher priority will be serviced
257first.
258
259The priority will be reset to C<0> after each call to one of the C<aio_*>
260functions.
261
262Example: open a file with low priority, then read something from it with
263higher priority so the read request is serviced before other low priority
264open requests (potentially spamming the cache):
265
266 aioreq_pri -3;
267 aio_open ..., sub {
268 return unless $_[0];
269
270 aioreq_pri -2;
271 aio_read $_[0], ..., sub {
272 ...
273 };
274 };
275
276=item aioreq_nice $pri_adjust
277
278Similar to C<aioreq_pri>, but subtracts the given value from the current
279priority, so the effect is cumulative.
125 280
126=item aio_open $pathname, $flags, $mode, $callback->($fh) 281=item aio_open $pathname, $flags, $mode, $callback->($fh)
127 282
128Asynchronously open or create a file and call the callback with a newly 283Asynchronously open or create a file and call the callback with a newly
129created filehandle for the file. 284created filehandle for the file.
135list. They are the same as used by C<sysopen>. 290list. They are the same as used by C<sysopen>.
136 291
137Likewise, C<$mode> specifies the mode of the newly created file, if it 292Likewise, C<$mode> specifies the mode of the newly created file, if it
138didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>, 293didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>,
139except that it is mandatory (i.e. use C<0> if you don't create new files, 294except that it is mandatory (i.e. use C<0> if you don't create new files,
140and C<0666> or C<0777> if you do). 295and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified
296by the umask in effect then the request is being executed, so better never
297change the umask.
141 298
142Example: 299Example:
143 300
144 aio_open "/etc/passwd", O_RDONLY, 0, sub { 301 aio_open "/etc/passwd", O_RDONLY, 0, sub {
145 if ($_[0]) { 302 if ($_[0]) {
180 aio_read $fh, 7, 15, $buffer, 0, sub { 337 aio_read $fh, 7, 15, $buffer, 0, sub {
181 $_[0] > 0 or die "read error: $!"; 338 $_[0] > 0 or die "read error: $!";
182 print "read $_[0] bytes: <$buffer>\n"; 339 print "read $_[0] bytes: <$buffer>\n";
183 }; 340 };
184 341
185=item aio_move $srcpath, $dstpath, $callback->($status)
186
187[EXPERIMENTAL due to internal aio_group use]
188
189Try to move the I<file> (directories not supported as either source or
190destination) from C<$srcpath> to C<$dstpath> and call the callback with
191the C<0> (error) or C<-1> ok.
192
193This is a composite request that tries to rename(2) the file first. If
194rename files with C<EXDEV>, it creates the destination file with mode 0200
195and copies the contents of the source file into it using C<aio_sendfile>,
196followed by restoring atime, mtime, access mode and uid/gid, in that
197order, and unlinking the C<$srcpath>.
198
199If an error occurs, the partial destination file will be unlinked, if
200possible, except when setting atime, mtime, access mode and uid/gid, where
201errors are being ignored.
202
203=cut
204
205sub aio_move($$$) {
206 my ($src, $dst, $cb) = @_;
207
208 my $grp = aio_group $cb;
209
210 add $grp aio_rename $src, $dst, sub {
211 if ($_[0] && $! == EXDEV) {
212 add $grp aio_open $src, O_RDONLY, 0, sub {
213 if (my $src_fh = $_[0]) {
214 my @stat = stat $src_fh;
215
216 add $grp aio_open $dst, O_WRONLY, 0200, sub {
217 if (my $dst_fh = $_[0]) {
218 add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
219 close $src_fh;
220
221 if ($_[0] == $stat[7]) {
222 utime $stat[8], $stat[9], $dst;
223 chmod $stat[2] & 07777, $dst_fh;
224 chown $stat[4], $stat[5], $dst_fh;
225 close $dst_fh;
226
227 add $grp aio_unlink $src, sub {
228 $grp->result ($_[0]);
229 };
230 } else {
231 my $errno = $!;
232 add $grp aio_unlink $dst, sub {
233 $! = $errno;
234 $grp->result (-1);
235 };
236 }
237 };
238 } else {
239 $grp->result (-1);
240 }
241 },
242
243 } else {
244 $grp->result (-1);
245 }
246 };
247 } else {
248 $grp->result ($_[0]);
249 }
250 };
251
252 $grp
253}
254
255=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 342=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
256 343
257Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 344Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
258reading at byte offset C<$in_offset>, and starts writing at the current 345reading at byte offset C<$in_offset>, and starts writing at the current
259file offset of C<$out_fh>. Because of that, it is not safe to issue more 346file offset of C<$out_fh>. Because of that, it is not safe to issue more
314=item aio_unlink $pathname, $callback->($status) 401=item aio_unlink $pathname, $callback->($status)
315 402
316Asynchronously unlink (delete) a file and call the callback with the 403Asynchronously unlink (delete) a file and call the callback with the
317result code. 404result code.
318 405
406=item aio_mknod $path, $mode, $dev, $callback->($status)
407
408[EXPERIMENTAL]
409
410Asynchronously create a device node (or fifo). See mknod(2).
411
412The only (POSIX-) portable way of calling this function is:
413
414 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...
415
319=item aio_link $srcpath, $dstpath, $callback->($status) 416=item aio_link $srcpath, $dstpath, $callback->($status)
320 417
321Asynchronously create a new link to the existing object at C<$srcpath> at 418Asynchronously create a new link to the existing object at C<$srcpath> at
322the path C<$dstpath> and call the callback with the result code. 419the path C<$dstpath> and call the callback with the result code.
323 420
324=item aio_symlink $srcpath, $dstpath, $callback->($status) 421=item aio_symlink $srcpath, $dstpath, $callback->($status)
325 422
326Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 423Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
327the path C<$dstpath> and call the callback with the result code. 424the path C<$dstpath> and call the callback with the result code.
328 425
426=item aio_readlink $path, $callback->($link)
427
428Asynchronously read the symlink specified by C<$path> and pass it to
429the callback. If an error occurs, nothing or undef gets passed to the
430callback.
431
329=item aio_rename $srcpath, $dstpath, $callback->($status) 432=item aio_rename $srcpath, $dstpath, $callback->($status)
330 433
331Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 434Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
332rename(2) and call the callback with the result code. 435rename(2) and call the callback with the result code.
436
437=item aio_mkdir $pathname, $mode, $callback->($status)
438
439Asynchronously mkdir (create) a directory and call the callback with
440the result code. C<$mode> will be modified by the umask at the time the
441request is executed, so do not change your umask.
333 442
334=item aio_rmdir $pathname, $callback->($status) 443=item aio_rmdir $pathname, $callback->($status)
335 444
336Asynchronously rmdir (delete) a directory and call the callback with the 445Asynchronously rmdir (delete) a directory and call the callback with the
337result code. 446result code.
343sorted, and will B<NOT> include the C<.> and C<..> entries. 452sorted, and will B<NOT> include the C<.> and C<..> entries.
344 453
345The callback a single argument which is either C<undef> or an array-ref 454The callback a single argument which is either C<undef> or an array-ref
346with the filenames. 455with the filenames.
347 456
457=item aio_load $path, $data, $callback->($status)
458
459This is a composite request that tries to fully load the given file into
460memory. Status is the same as with aio_read.
461
462=cut
463
464sub aio_load($$;$) {
465 aio_block {
466 my ($path, undef, $cb) = @_;
467 my $data = \$_[1];
468
469 my $pri = aioreq_pri;
470 my $grp = aio_group $cb;
471
472 aioreq_pri $pri;
473 add $grp aio_open $path, O_RDONLY, 0, sub {
474 my ($fh) = @_
475 or return $grp->result (-1);
476
477 aioreq_pri $pri;
478 add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
479 $grp->result ($_[0]);
480 };
481 };
482
483 $grp
484 }
485}
486
487=item aio_copy $srcpath, $dstpath, $callback->($status)
488
489Try to copy the I<file> (directories not supported as either source or
490destination) from C<$srcpath> to C<$dstpath> and call the callback with
491the C<0> (error) or C<-1> ok.
492
493This is a composite request that it creates the destination file with
494mode 0200 and copies the contents of the source file into it using
495C<aio_sendfile>, followed by restoring atime, mtime, access mode and
496uid/gid, in that order.
497
498If an error occurs, the partial destination file will be unlinked, if
499possible, except when setting atime, mtime, access mode and uid/gid, where
500errors are being ignored.
501
502=cut
503
504sub aio_copy($$;$) {
505 aio_block {
506 my ($src, $dst, $cb) = @_;
507
508 my $pri = aioreq_pri;
509 my $grp = aio_group $cb;
510
511 aioreq_pri $pri;
512 add $grp aio_open $src, O_RDONLY, 0, sub {
513 if (my $src_fh = $_[0]) {
514 my @stat = stat $src_fh;
515
516 aioreq_pri $pri;
517 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
518 if (my $dst_fh = $_[0]) {
519 aioreq_pri $pri;
520 add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
521 if ($_[0] == $stat[7]) {
522 $grp->result (0);
523 close $src_fh;
524
525 # those should not normally block. should. should.
526 utime $stat[8], $stat[9], $dst;
527 chmod $stat[2] & 07777, $dst_fh;
528 chown $stat[4], $stat[5], $dst_fh;
529 close $dst_fh;
530 } else {
531 $grp->result (-1);
532 close $src_fh;
533 close $dst_fh;
534
535 aioreq $pri;
536 add $grp aio_unlink $dst;
537 }
538 };
539 } else {
540 $grp->result (-1);
541 }
542 },
543
544 } else {
545 $grp->result (-1);
546 }
547 };
548
549 $grp
550 }
551}
552
553=item aio_move $srcpath, $dstpath, $callback->($status)
554
555Try to move the I<file> (directories not supported as either source or
556destination) from C<$srcpath> to C<$dstpath> and call the callback with
557the C<0> (error) or C<-1> ok.
558
559This is a composite request that tries to rename(2) the file first. If
560rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if
561that is successful, unlinking the C<$srcpath>.
562
563=cut
564
565sub aio_move($$;$) {
566 aio_block {
567 my ($src, $dst, $cb) = @_;
568
569 my $pri = aioreq_pri;
570 my $grp = aio_group $cb;
571
572 aioreq_pri $pri;
573 add $grp aio_rename $src, $dst, sub {
574 if ($_[0] && $! == EXDEV) {
575 aioreq_pri $pri;
576 add $grp aio_copy $src, $dst, sub {
577 $grp->result ($_[0]);
578
579 if (!$_[0]) {
580 aioreq_pri $pri;
581 add $grp aio_unlink $src;
582 }
583 };
584 } else {
585 $grp->result ($_[0]);
586 }
587 };
588
589 $grp
590 }
591}
592
348=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 593=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
349 594
350[EXPERIMENTAL due to internal aio_group use]
351
352Scans a directory (similar to C<aio_readdir>) but additionally tries to 595Scans a directory (similar to C<aio_readdir>) but additionally tries to
353separate the entries of directory C<$path> into two sets of names, ones 596efficiently separate the entries of directory C<$path> into two sets of
354you can recurse into (directories or links to them), and ones you cannot 597names, directories you can recurse into (directories), and ones you cannot
355recurse into (everything else). 598recurse into (everything else, including symlinks to directories).
356 599
357C<aio_scandir> is a composite request that creates of many sub requests_ 600C<aio_scandir> is a composite request that creates of many sub requests_
358C<$maxreq> specifies the maximum number of outstanding aio requests that 601C<$maxreq> specifies the maximum number of outstanding aio requests that
359this function generates. If it is C<< <= 0 >>, then a suitable default 602this function generates. If it is C<< <= 0 >>, then a suitable default
360will be chosen (currently 6). 603will be chosen (currently 4).
361 604
362On error, the callback is called without arguments, otherwise it receives 605On error, the callback is called without arguments, otherwise it receives
363two array-refs with path-relative entry names. 606two array-refs with path-relative entry names.
364 607
365Example: 608Example:
399as those tend to return 0 or 1 as link counts, which disables the 642as those tend to return 0 or 1 as link counts, which disables the
400directory counting heuristic. 643directory counting heuristic.
401 644
402=cut 645=cut
403 646
404sub aio_scandir($$$) { 647sub aio_scandir($$;$) {
648 aio_block {
405 my ($path, $maxreq, $cb) = @_; 649 my ($path, $maxreq, $cb) = @_;
406 650
651 my $pri = aioreq_pri;
652
407 my $grp = aio_group $cb; 653 my $grp = aio_group $cb;
408 654
409 $maxreq = 6 if $maxreq <= 0; 655 $maxreq = 4 if $maxreq <= 0;
410 656
411 # stat once 657 # stat once
658 aioreq_pri $pri;
412 add $grp aio_stat $path, sub { 659 add $grp aio_stat $path, sub {
413 return $grp->result () if $_[0]; 660 return $grp->result () if $_[0];
414 my $now = time; 661 my $now = time;
415 my $hash1 = join ":", (stat _)[0,1,3,7,9]; 662 my $hash1 = join ":", (stat _)[0,1,3,7,9];
416 663
417 # read the directory entries 664 # read the directory entries
665 aioreq_pri $pri;
418 add $grp aio_readdir $path, sub { 666 add $grp aio_readdir $path, sub {
419 my $entries = shift 667 my $entries = shift
420 or return $grp->result (); 668 or return $grp->result ();
421 669
422 # stat the dir another time 670 # stat the dir another time
671 aioreq_pri $pri;
423 add $grp aio_stat $path, sub { 672 add $grp aio_stat $path, sub {
424 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 673 my $hash2 = join ":", (stat _)[0,1,3,7,9];
425 674
426 my $ndirs; 675 my $ndirs;
427 676
428 # take the slow route if anything looks fishy 677 # take the slow route if anything looks fishy
429 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 678 if ($hash1 ne $hash2 or (stat _)[9] == $now) {
430 $ndirs = -1; 679 $ndirs = -1;
431 } else { 680 } else {
432 # if nlink == 2, we are finished 681 # if nlink == 2, we are finished
433 # on non-posix-fs's, we rely on nlink < 2 682 # on non-posix-fs's, we rely on nlink < 2
434 $ndirs = (stat _)[3] - 2 683 $ndirs = (stat _)[3] - 2
435 or return $grp->result ([], $entries); 684 or return $grp->result ([], $entries);
436 } 685 }
437 686
438 # sort into likely dirs and likely nondirs 687 # sort into likely dirs and likely nondirs
439 # dirs == files without ".", short entries first 688 # dirs == files without ".", short entries first
440 $entries = [map $_->[0], 689 $entries = [map $_->[0],
441 sort { $b->[1] cmp $a->[1] } 690 sort { $b->[1] cmp $a->[1] }
442 map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length], 691 map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
443 @$entries]; 692 @$entries];
444 693
445 my (@dirs, @nondirs); 694 my (@dirs, @nondirs);
446 695
447 my ($statcb, $schedcb);
448 my $nreq = 0;
449
450 my $statgrp = add $grp aio_group; 696 my $statgrp = add $grp aio_group sub {
451
452 $schedcb = sub {
453 if (@$entries) {
454 if ($nreq < $maxreq) {
455 my $ent = pop @$entries;
456 $nreq++;
457 add $statgrp aio_stat "$path/$ent/.", sub { $statcb->($_[0], $ent) };
458 }
459 } elsif (!$nreq) {
460 # finished
461 $statgrp->cancel;
462 undef $statcb;
463 undef $schedcb;
464 $grp->result (\@dirs, \@nondirs); 697 $grp->result (\@dirs, \@nondirs);
465 } 698 };
699
700 limit $statgrp $maxreq;
701 feed $statgrp sub {
702 return unless @$entries;
703 my $entry = pop @$entries;
704
705 aioreq_pri $pri;
706 add $statgrp aio_stat "$path/$entry/.", sub {
707 if ($_[0] < 0) {
708 push @nondirs, $entry;
709 } else {
710 # need to check for real directory
711 aioreq_pri $pri;
712 add $statgrp aio_lstat "$path/$entry", sub {
713 if (-d _) {
714 push @dirs, $entry;
715
716 unless (--$ndirs) {
717 push @nondirs, @$entries;
718 feed $statgrp;
719 }
720 } else {
721 push @nondirs, $entry;
722 }
723 }
724 }
725 };
726 };
466 }; 727 };
467 $statcb = sub {
468 my ($status, $entry) = @_;
469
470 if ($status < 0) {
471 $nreq--;
472 push @nondirs, $entry;
473 &$schedcb;
474 } else {
475 # need to check for real directory
476 add $grp aio_lstat "$path/$entry", sub {
477 $nreq--;
478
479 if (-d _) {
480 push @dirs, $entry;
481
482 if (!--$ndirs) {
483 push @nondirs, @$entries;
484 $entries = [];
485 }
486 } else {
487 push @nondirs, $entry;
488 }
489
490 &$schedcb;
491 }
492 }
493 };
494
495 &$schedcb while @$entries && $nreq < $maxreq;
496 }; 728 };
497 }; 729 };
730
731 $grp
498 }; 732 }
733}
499 734
735=item aio_rmtree $path, $callback->($status)
736
737Delete a directory tree starting (and including) C<$path>, return the
738status of the final C<rmdir> only. This is a composite request that
739uses C<aio_scandir> to recurse into and rmdir directories, and unlink
740everything else.
741
742=cut
743
744sub aio_rmtree;
745sub aio_rmtree($;$) {
746 aio_block {
747 my ($path, $cb) = @_;
748
749 my $pri = aioreq_pri;
750 my $grp = aio_group $cb;
751
752 aioreq_pri $pri;
753 add $grp aio_scandir $path, 0, sub {
754 my ($dirs, $nondirs) = @_;
755
756 my $dirgrp = aio_group sub {
757 add $grp aio_rmdir $path, sub {
758 $grp->result ($_[0]);
759 };
760 };
761
762 (aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
763 (aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
764
765 add $grp $dirgrp;
766 };
767
500 $grp 768 $grp
769 }
501} 770}
502 771
503=item aio_fsync $fh, $callback->($status) 772=item aio_fsync $fh, $callback->($status)
504 773
505Asynchronously call fsync on the given filehandle and call the callback 774Asynchronously call fsync on the given filehandle and call the callback
513If this call isn't available because your OS lacks it or it couldn't be 782If this call isn't available because your OS lacks it or it couldn't be
514detected, it will be emulated by calling C<fsync> instead. 783detected, it will be emulated by calling C<fsync> instead.
515 784
516=item aio_group $callback->(...) 785=item aio_group $callback->(...)
517 786
518[EXPERIMENTAL]
519
520This is a very special aio request: Instead of doing something, it is a 787This is a very special aio request: Instead of doing something, it is a
521container for other aio requests, which is useful if you want to bundle 788container for other aio requests, which is useful if you want to bundle
522many requests into a single, composite, request. 789many requests into a single, composite, request with a definite callback
790and the ability to cancel the whole request with its subrequests.
523 791
524Returns an object of class L<IO::AIO::GRP>. See its documentation below 792Returns an object of class L<IO::AIO::GRP>. See its documentation below
525for more info. 793for more info.
526 794
527Example: 795Example:
533 add $grp 801 add $grp
534 (aio_stat ...), 802 (aio_stat ...),
535 (aio_stat ...), 803 (aio_stat ...),
536 ...; 804 ...;
537 805
806=item aio_nop $callback->()
807
808This is a special request - it does nothing in itself and is only used for
809side effects, such as when you want to add a dummy request to a group so
810that finishing the requests in the group depends on executing the given
811code.
812
813While this request does nothing, it still goes through the execution
814phase and still requires a worker thread. Thus, the callback will not
815be executed immediately but only after other requests in the queue have
816entered their execution phase. This can be used to measure request
817latency.
818
538=item IO::AIO::aio_sleep $fractional_seconds, $callback->() *NOT EXPORTED* 819=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
539 820
540Mainly used for debugging and benchmarking, this aio request puts one of 821Mainly used for debugging and benchmarking, this aio request puts one of
541the request workers to sleep for the given time. 822the request workers to sleep for the given time.
542 823
543While it is theoretically handy to have simple I/O scheduling requests 824While it is theoretically handy to have simple I/O scheduling requests
544like sleep and file handle readable/writable, the overhead this creates 825like sleep and file handle readable/writable, the overhead this creates is
545is immense, so do not use this function except to put your application 826immense (it blocks a thread for a long time) so do not use this function
546under artificial I/O pressure. 827except to put your application under artificial I/O pressure.
547 828
548=back 829=back
549 830
550=head2 IO::AIO::REQ CLASS 831=head2 IO::AIO::REQ CLASS
551 832
552All non-aggregate C<aio_*> functions return an object of this class when 833All non-aggregate C<aio_*> functions return an object of this class when
553called in non-void context. 834called in non-void context.
554 835
555A request always moves through the following five states in its lifetime,
556in order: B<ready> (request has been created, but has not been executed
557yet), B<execute> (request is currently being executed), B<pending>
558(request has been executed but callback has not been called yet),
559B<result> (results are being processed synchronously, includes calling the
560callback) and B<done> (request has reached the end of its lifetime and
561holds no resources anymore).
562
563=over 4 836=over 4
564 837
565=item $req->cancel 838=item cancel $req
566 839
567Cancels the request, if possible. Has the effect of skipping execution 840Cancels the request, if possible. Has the effect of skipping execution
568when entering the B<execute> state and skipping calling the callback when 841when entering the B<execute> state and skipping calling the callback when
569entering the the B<result> state, but will leave the request otherwise 842entering the the B<result> state, but will leave the request otherwise
570untouched. That means that requests that currently execute will not be 843untouched. That means that requests that currently execute will not be
571stopped and resources held by the request will not be freed prematurely. 844stopped and resources held by the request will not be freed prematurely.
572 845
846=item cb $req $callback->(...)
847
848Replace (or simply set) the callback registered to the request.
849
573=back 850=back
574 851
575=head2 IO::AIO::GRP CLASS 852=head2 IO::AIO::GRP CLASS
576 853
577This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to 854This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to
616=item * They can also can also be added to other IO::AIO::GRP objects. 893=item * They can also can also be added to other IO::AIO::GRP objects.
617 894
618=item * You must not add requests to a group from within the group callback (or 895=item * You must not add requests to a group from within the group callback (or
619any later time). 896any later time).
620 897
621=item * This does not harmonise well with C<max_outstanding>, so best do
622not combine C<aio_group> with it. Groups and feeders are recommended for
623this kind of concurrency-limiting.
624
625=back 898=back
626 899
627Their lifetime, simplified, looks like this: when they are empty, they 900Their lifetime, simplified, looks like this: when they are empty, they
628will finish very quickly. If they contain only requests that are in the 901will finish very quickly. If they contain only requests that are in the
629C<done> state, they will also finish. Otherwise they will continue to 902C<done> state, they will also finish. Otherwise they will continue to
634group. And only when all those requests have finished will the the group 907group. And only when all those requests have finished will the the group
635itself finish. 908itself finish.
636 909
637=over 4 910=over 4
638 911
912=item add $grp ...
913
639=item $grp->add (...) 914=item $grp->add (...)
640
641=item add $grp ...
642 915
643Add one or more requests to the group. Any type of L<IO::AIO::REQ> can 916Add one or more requests to the group. Any type of L<IO::AIO::REQ> can
644be added, including other groups, as long as you do not create circular 917be added, including other groups, as long as you do not create circular
645dependencies. 918dependencies.
646 919
647Returns all its arguments. 920Returns all its arguments.
648 921
922=item $grp->cancel_subs
923
924Cancel all subrequests and clears any feeder, but not the group request
925itself. Useful when you queued a lot of events but got a result early.
926
649=item $grp->result (...) 927=item $grp->result (...)
650 928
651Set the result value(s) that will be passed to the group callback when all 929Set the result value(s) that will be passed to the group callback when all
652subrequests have finished. By default, no argument will be passed. 930subrequests have finished and set thre groups errno to the current value
931of errno (just like calling C<errno> without an error number). By default,
932no argument will be passed and errno is zero.
653 933
654=item $grp->set_feeder ($callback->($grp)) 934=item $grp->errno ([$errno])
655 935
656[VERY EXPERIMENTAL] 936Sets the group errno value to C<$errno>, or the current value of errno
937when the argument is missing.
938
939Every aio request has an associated errno value that is restored when
940the callback is invoked. This method lets you change this value from its
941default (0).
942
943Calling C<result> will also set errno, so make sure you either set C<$!>
944before the call to C<result>, or call c<errno> after it.
945
946=item feed $grp $callback->($grp)
657 947
658Sets a feeder/generator on this group: every group can have an attached 948Sets a feeder/generator on this group: every group can have an attached
659generator that generates requests if idle. The idea behind this is that, 949generator that generates requests if idle. The idea behind this is that,
660although you could just queue as many requests as you want in a group, 950although you could just queue as many requests as you want in a group,
661this might starve other requests for a potentially long time. For 951this might starve other requests for a potentially long time. For
662example, C<aio_scandir> might generate hundreds of thousands C<aio_stat> 952example, C<aio_scandir> might generate hundreds of thousands C<aio_stat>
663requests, delaying any later requests for a long time. 953requests, delaying any later requests for a long time.
664 954
665To avoid this, and allow incremental generation of requests, you can 955To avoid this, and allow incremental generation of requests, you can
666instead a group and set a feeder on it that generates those requests. The 956instead a group and set a feeder on it that generates those requests. The
667feeder will be called whenever there are few enough (see C<feeder_limit>, 957feed callback will be called whenever there are few enough (see C<limit>,
668below) requests active in the group itself and is expected to queue more 958below) requests active in the group itself and is expected to queue more
669requests. 959requests.
670 960
671The feeder can queue as many requests as it likes (i.e. C<add> does not 961The feed callback can queue as many requests as it likes (i.e. C<add> does
672impose any limits). 962not impose any limits).
673 963
674If the feeder does not queue more requests when called, it will be 964If the feed does not queue more requests when called, it will be
675automatically removed from the group. 965automatically removed from the group.
676 966
677If the feeder limit is C<0>, it will be set to C<2> automatically. 967If the feed limit is C<0>, it will be set to C<2> automatically.
678 968
679Example: 969Example:
680 970
681 # stat all files in @files, but only ever use four aio requests concurrently: 971 # stat all files in @files, but only ever use four aio requests concurrently:
682 972
683 my $grp = aio_group sub { print "finished\n" }; 973 my $grp = aio_group sub { print "finished\n" };
684 $grp->feeder_limit (4); 974 limit $grp 4;
685 $grp->set_feeder (sub { 975 feed $grp sub {
686 my $file = pop @files 976 my $file = pop @files
687 or return; 977 or return;
688 978
689 add $grp aio_stat $file, sub { ... }; 979 add $grp aio_stat $file, sub { ... };
690 }); 980 };
691 981
692=item $grp->feeder_limit ($num) 982=item limit $grp $num
693 983
694Sets the feeder limit for the group: The feeder will be called whenever 984Sets the feeder limit for the group: The feeder will be called whenever
695the group contains less than this many requests. 985the group contains less than this many requests.
696 986
697Setting the limit to C<0> will pause the feeding process. 987Setting the limit to C<0> will pause the feeding process.
698 988
699=back 989=back
700 990
701=head2 SUPPORT FUNCTIONS 991=head2 SUPPORT FUNCTIONS
992
993=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
702 994
703=over 4 995=over 4
704 996
705=item $fileno = IO::AIO::poll_fileno 997=item $fileno = IO::AIO::poll_fileno
706 998
711 1003
712See C<poll_cb> for an example. 1004See C<poll_cb> for an example.
713 1005
714=item IO::AIO::poll_cb 1006=item IO::AIO::poll_cb
715 1007
716Process all outstanding events on the result pipe. You have to call this 1008Process some outstanding events on the result pipe. You have to call this
717regularly. Returns the number of events processed. Returns immediately 1009regularly. Returns the number of events processed. Returns immediately
718when no events are outstanding. 1010when no events are outstanding. The amount of events processed depends on
1011the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
1012
1013If not all requests were processed for whatever reason, the filehandle
1014will still be ready when C<poll_cb> returns.
719 1015
720Example: Install an Event watcher that automatically calls 1016Example: Install an Event watcher that automatically calls
721IO::AIO::poll_cb with high priority: 1017IO::AIO::poll_cb with high priority:
722 1018
723 Event->io (fd => IO::AIO::poll_fileno, 1019 Event->io (fd => IO::AIO::poll_fileno,
724 poll => 'r', async => 1, 1020 poll => 'r', async => 1,
725 cb => \&IO::AIO::poll_cb); 1021 cb => \&IO::AIO::poll_cb);
726 1022
1023=item IO::AIO::max_poll_reqs $nreqs
1024
1025=item IO::AIO::max_poll_time $seconds
1026
1027These set the maximum number of requests (default C<0>, meaning infinity)
1028that are being processed by C<IO::AIO::poll_cb> in one call, respectively
1029the maximum amount of time (default C<0>, meaning infinity) spent in
1030C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
1031of time C<poll_cb> is allowed to use).
1032
1033Setting C<max_poll_time> to a non-zero value creates an overhead of one
1034syscall per request processed, which is not normally a problem unless your
1035callbacks are really really fast or your OS is really really slow (I am
1036not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
1037
1038Setting these is useful if you want to ensure some level of
1039interactiveness when perl is not fast enough to process all requests in
1040time.
1041
1042For interactive programs, values such as C<0.01> to C<0.1> should be fine.
1043
1044Example: Install an Event watcher that automatically calls
1045IO::AIO::poll_cb with low priority, to ensure that other parts of the
1046program get the CPU sometimes even under high AIO load.
1047
1048 # try not to spend much more than 0.1s in poll_cb
1049 IO::AIO::max_poll_time 0.1;
1050
1051 # use a low priority so other tasks have priority
1052 Event->io (fd => IO::AIO::poll_fileno,
1053 poll => 'r', nice => 1,
1054 cb => &IO::AIO::poll_cb);
1055
727=item IO::AIO::poll_wait 1056=item IO::AIO::poll_wait
728 1057
1058If there are any outstanding requests and none of them in the result
729Wait till the result filehandle becomes ready for reading (simply does a 1059phase, wait till the result filehandle becomes ready for reading (simply
730C<select> on the filehandle. This is useful if you want to synchronously wait 1060does a C<select> on the filehandle. This is useful if you want to
731for some requests to finish). 1061synchronously wait for some requests to finish).
732 1062
733See C<nreqs> for an example. 1063See C<nreqs> for an example.
734 1064
1065=item IO::AIO::poll
1066
1067Waits until some requests have been handled.
1068
1069Returns the number of requests processed, but is otherwise strictly
1070equivalent to:
1071
1072 IO::AIO::poll_wait, IO::AIO::poll_cb
1073
735=item IO::AIO::nreqs 1074=item IO::AIO::flush
736 1075
737Returns the number of requests currently outstanding (i.e. for which their 1076Wait till all outstanding AIO requests have been handled.
738callback has not been invoked yet).
739 1077
740Example: wait till there are no outstanding requests anymore: 1078Strictly equivalent to:
741 1079
742 IO::AIO::poll_wait, IO::AIO::poll_cb 1080 IO::AIO::poll_wait, IO::AIO::poll_cb
743 while IO::AIO::nreqs; 1081 while IO::AIO::nreqs;
744 1082
745=item IO::AIO::flush 1083=head3 CONTROLLING THE NUMBER OF THREADS
746
747Wait till all outstanding AIO requests have been handled.
748
749Strictly equivalent to:
750
751 IO::AIO::poll_wait, IO::AIO::poll_cb
752 while IO::AIO::nreqs;
753
754=item IO::AIO::poll
755
756Waits until some requests have been handled.
757
758Strictly equivalent to:
759
760 IO::AIO::poll_wait, IO::AIO::poll_cb
761 if IO::AIO::nreqs;
762 1084
763=item IO::AIO::min_parallel $nthreads 1085=item IO::AIO::min_parallel $nthreads
764 1086
765Set the minimum number of AIO threads to C<$nthreads>. The current 1087Set the minimum number of AIO threads to C<$nthreads>. The current
766default is C<8>, which means eight asynchronous operations can execute 1088default is C<8>, which means eight asynchronous operations can execute
767concurrently at any one time (the number of outstanding requests, 1089concurrently at any one time (the number of outstanding requests,
768however, is unlimited). 1090however, is unlimited).
769 1091
770IO::AIO starts threads only on demand, when an AIO request is queued and 1092IO::AIO starts threads only on demand, when an AIO request is queued and
771no free thread exists. 1093no free thread exists. Please note that queueing up a hundred requests can
1094create demand for a hundred threads, even if it turns out that everything
1095is in the cache and could have been processed faster by a single thread.
772 1096
773It is recommended to keep the number of threads relatively low, as some 1097It is recommended to keep the number of threads relatively low, as some
774Linux kernel versions will scale negatively with the number of threads 1098Linux kernel versions will scale negatively with the number of threads
775(higher parallelity => MUCH higher latency). With current Linux 2.6 1099(higher parallelity => MUCH higher latency). With current Linux 2.6
776versions, 4-32 threads should be fine. 1100versions, 4-32 threads should be fine.
790This module automatically runs C<max_parallel 0> at program end, to ensure 1114This module automatically runs C<max_parallel 0> at program end, to ensure
791that all threads are killed and that there are no outstanding requests. 1115that all threads are killed and that there are no outstanding requests.
792 1116
793Under normal circumstances you don't need to call this function. 1117Under normal circumstances you don't need to call this function.
794 1118
1119=item IO::AIO::max_idle $nthreads
1120
1121Limit the number of threads (default: 4) that are allowed to idle (i.e.,
1122threads that did not get a request to process within 10 seconds). That
1123means if a thread becomes idle while C<$nthreads> other threads are also
1124idle, it will free its resources and exit.
1125
1126This is useful when you allow a large number of threads (e.g. 100 or 1000)
1127to allow for extremely high load situations, but want to free resources
1128under normal circumstances (1000 threads can easily consume 30MB of RAM).
1129
1130The default is probably ok in most situations, especially if thread
1131creation is fast. If thread creation is very slow on your system you might
1132want to use larger values.
1133
795=item $oldnreqs = IO::AIO::max_outstanding $nreqs 1134=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs
796 1135
797[DEPRECATED] 1136This is a very bad function to use in interactive programs because it
1137blocks, and a bad way to reduce concurrency because it is inexact: Better
1138use an C<aio_group> together with a feed callback.
798 1139
799Sets the maximum number of outstanding requests to C<$nreqs>. If you 1140Sets the maximum number of outstanding requests to C<$nreqs>. If you
800try to queue up more than this number of requests, the caller will block until 1141to queue up more than this number of requests, the next call to the
801some requests have been handled. 1142C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
1143function will block until the limit is no longer exceeded.
802 1144
803The default is very large, so normally there is no practical limit. If you 1145The default value is very large, so there is no practical limit on the
804queue up many requests in a loop it often improves speed if you set 1146number of outstanding requests.
805this to a relatively low number, such as C<100>.
806 1147
807This function does not work well together with C<aio_group>'s, and their 1148You can still queue as many requests as you want. Therefore,
808feeder interface is better suited to limiting concurrency, so do not use 1149C<max_oustsanding> is mainly useful in simple scripts (with low values) or
809this function. 1150as a stop gap to shield against fatal memory overflow (with large values).
810 1151
811Under normal circumstances you don't need to call this function. 1152=head3 STATISTICAL INFORMATION
1153
1154=item IO::AIO::nreqs
1155
1156Returns the number of requests currently in the ready, execute or pending
1157states (i.e. for which their callback has not been invoked yet).
1158
1159Example: wait till there are no outstanding requests anymore:
1160
1161 IO::AIO::poll_wait, IO::AIO::poll_cb
1162 while IO::AIO::nreqs;
1163
1164=item IO::AIO::nready
1165
1166Returns the number of requests currently in the ready state (not yet
1167executed).
1168
1169=item IO::AIO::npending
1170
1171Returns the number of requests currently in the pending state (executed,
1172but not yet processed by poll_cb).
812 1173
813=back 1174=back
814 1175
815=cut 1176=cut
816 1177
830 *$sym 1191 *$sym
831} 1192}
832 1193
833min_parallel 8; 1194min_parallel 8;
834 1195
835END { 1196END { flush }
836 max_parallel 0;
837}
838 1197
8391; 11981;
840 1199
841=head2 FORK BEHAVIOUR 1200=head2 FORK BEHAVIOUR
842 1201
843This module should do "the right thing" when the process using it forks: 1202This module should do "the right thing" when the process using it forks:
844 1203
845Before the fork, IO::AIO enters a quiescent state where no requests 1204Before the fork, IO::AIO enters a quiescent state where no requests
846can be added in other threads and no results will be processed. After 1205can be added in other threads and no results will be processed. After
847the fork the parent simply leaves the quiescent state and continues 1206the fork the parent simply leaves the quiescent state and continues
848request/result processing, while the child clears the request/result 1207request/result processing, while the child frees the request/result queue
849queue (so the requests started before the fork will only be handled in 1208(so that the requests started before the fork will only be handled in the
850the parent). Threads will be started on demand until the limit ste in the 1209parent). Threads will be started on demand until the limit set in the
851parent process has been reached again. 1210parent process has been reached again.
852 1211
853In short: the parent will, after a short pause, continue as if fork had 1212In short: the parent will, after a short pause, continue as if fork had
854not been called, while the child will act as if IO::AIO has not been used 1213not been called, while the child will act as if IO::AIO has not been used
855yet. 1214yet.
856 1215
857=head2 MEMORY USAGE 1216=head2 MEMORY USAGE
858 1217
1218Per-request usage:
1219
859Each aio request uses - depending on your architecture - around 128 bytes 1220Each aio request uses - depending on your architecture - around 100-200
860of memory. In addition, stat requests need a stat buffer (possibly a few 1221bytes of memory. In addition, stat requests need a stat buffer (possibly
861hundred bytes). Perl scalars and other data passed into aio requests will 1222a few hundred bytes), readdir requires a result buffer and so on. Perl
862also be locked. 1223scalars and other data passed into aio requests will also be locked and
1224will consume memory till the request has entered the done state.
863 1225
864This is now awfully much, so queuing lots of requests is not usually a 1226This is now awfully much, so queuing lots of requests is not usually a
865problem. 1227problem.
866 1228
867Each thread needs a stack area which is usually around 16k, sometimes much 1229Per-thread usage:
868larger, depending on the OS. 1230
1231In the execution phase, some aio requests require more memory for
1232temporary buffers, and each thread requires a stack and other data
1233structures (usually around 16k-128k, depending on the OS).
1234
1235=head1 KNOWN BUGS
1236
1237Known bugs will be fixed in the next release.
869 1238
870=head1 SEE ALSO 1239=head1 SEE ALSO
871 1240
872L<Coro>, L<Linux::AIO> (obsolete). 1241L<Coro::AIO>.
873 1242
874=head1 AUTHOR 1243=head1 AUTHOR
875 1244
876 Marc Lehmann <schmorp@schmorp.de> 1245 Marc Lehmann <schmorp@schmorp.de>
877 http://home.schmorp.de/ 1246 http://home.schmorp.de/

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines