… | |
… | |
15 | |
15 | |
16 | aio_read $fh, 30000, 1024, $buffer, 0, sub { |
16 | aio_read $fh, 30000, 1024, $buffer, 0, sub { |
17 | $_[0] > 0 or die "read error: $!"; |
17 | $_[0] > 0 or die "read error: $!"; |
18 | }; |
18 | }; |
19 | |
19 | |
20 | # AnyEvent |
20 | # version 2+ has request and group objects |
|
|
21 | use IO::AIO 2; |
|
|
22 | |
|
|
23 | aioreq_pri 4; # give next request a very high priority |
|
|
24 | my $req = aio_unlink "/tmp/file", sub { }; |
|
|
25 | $req->cancel; # cancel request if still in queue |
|
|
26 | |
|
|
27 | my $grp = aio_group sub { print "all stats done\n" }; |
|
|
28 | add $grp aio_stat "..." for ...; |
|
|
29 | |
|
|
30 | # AnyEvent integration |
21 | open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!"; |
31 | open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!"; |
22 | my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb }); |
32 | my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb }); |
23 | |
33 | |
24 | # Event |
34 | # Event integration |
25 | Event->io (fd => IO::AIO::poll_fileno, |
35 | Event->io (fd => IO::AIO::poll_fileno, |
26 | poll => 'r', |
36 | poll => 'r', |
27 | cb => \&IO::AIO::poll_cb); |
37 | cb => \&IO::AIO::poll_cb); |
28 | |
38 | |
29 | # Glib/Gtk2 |
39 | # Glib/Gtk2 integration |
30 | add_watch Glib::IO IO::AIO::poll_fileno, |
40 | add_watch Glib::IO IO::AIO::poll_fileno, |
31 | in => sub { IO::AIO::poll_cb; 1 }; |
41 | in => sub { IO::AIO::poll_cb; 1 }; |
32 | |
42 | |
33 | # Tk |
43 | # Tk integration |
34 | Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "", |
44 | Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "", |
35 | readable => \&IO::AIO::poll_cb); |
45 | readable => \&IO::AIO::poll_cb); |
36 | |
46 | |
37 | # Danga::Socket |
47 | # Danga::Socket integration |
38 | Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => |
48 | Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => |
39 | \&IO::AIO::poll_cb); |
49 | \&IO::AIO::poll_cb); |
40 | |
50 | |
41 | |
|
|
42 | =head1 DESCRIPTION |
51 | =head1 DESCRIPTION |
43 | |
52 | |
44 | This module implements asynchronous I/O using whatever means your |
53 | This module implements asynchronous I/O using whatever means your |
45 | operating system supports. |
54 | operating system supports. |
46 | |
55 | |
47 | Currently, a number of threads are started that execute your read/writes |
56 | Currently, a number of threads are started that execute your read/writes |
48 | and signal their completion. You don't need thread support in your libc or |
57 | and signal their completion. You don't need thread support in perl, and |
49 | perl, and the threads created by this module will not be visible to the |
58 | the threads created by this module will not be visible to perl. In the |
50 | pthreads library. In the future, this module might make use of the native |
59 | future, this module might make use of the native aio functions available |
51 | aio functions available on many operating systems. However, they are often |
60 | on many operating systems. However, they are often not well-supported |
52 | not well-supported (Linux doesn't allow them on normal files currently, |
61 | (Linux doesn't allow them on normal files currently, for example), |
53 | for example), and they would only support aio_read and aio_write, so the |
62 | and they would only support aio_read and aio_write, so the remaining |
54 | remaining functionality would have to be implemented using threads anyway. |
63 | functionality would have to be implemented using threads anyway. |
55 | |
64 | |
56 | Although the module will work with in the presence of other threads, it is |
65 | Although the module will work with in the presence of other threads, |
57 | currently not reentrant, so use appropriate locking yourself, always call |
66 | it is currently not reentrant in any way, so use appropriate locking |
58 | C<poll_cb> from within the same thread, or never call C<poll_cb> (or other |
67 | yourself, always call C<poll_cb> from within the same thread, or never |
59 | C<aio_> functions) recursively. |
68 | call C<poll_cb> (or other C<aio_> functions) recursively. |
60 | |
69 | |
61 | =cut |
70 | =cut |
62 | |
71 | |
63 | package IO::AIO; |
72 | package IO::AIO; |
64 | |
73 | |
65 | no warnings; |
74 | no warnings; |
|
|
75 | use strict 'vars'; |
66 | |
76 | |
67 | use base 'Exporter'; |
77 | use base 'Exporter'; |
68 | |
78 | |
69 | use Fcntl (); |
|
|
70 | |
|
|
71 | BEGIN { |
79 | BEGIN { |
72 | $VERSION = '1.8'; |
80 | our $VERSION = '2.0'; |
73 | |
81 | |
74 | @EXPORT = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat |
82 | our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat |
75 | aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink |
83 | aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink |
76 | aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move); |
84 | aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move |
77 | @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel |
85 | aio_group aio_nop); |
78 | max_outstanding nreqs); |
86 | our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); |
|
|
87 | our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush |
|
|
88 | min_parallel max_parallel max_outstanding nreqs); |
|
|
89 | |
|
|
90 | @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; |
79 | |
91 | |
80 | require XSLoader; |
92 | require XSLoader; |
81 | XSLoader::load IO::AIO, $VERSION; |
93 | XSLoader::load ("IO::AIO", $VERSION); |
82 | } |
94 | } |
83 | |
95 | |
84 | =head1 FUNCTIONS |
96 | =head1 FUNCTIONS |
85 | |
97 | |
86 | =head2 AIO FUNCTIONS |
98 | =head2 AIO FUNCTIONS |
… | |
… | |
94 | syscall has been executed asynchronously. |
106 | syscall has been executed asynchronously. |
95 | |
107 | |
96 | All functions expecting a filehandle keep a copy of the filehandle |
108 | All functions expecting a filehandle keep a copy of the filehandle |
97 | internally until the request has finished. |
109 | internally until the request has finished. |
98 | |
110 | |
|
|
111 | All requests return objects of type L<IO::AIO::REQ> that allow further |
|
|
112 | manipulation of those requests while they are in-flight. |
|
|
113 | |
99 | The pathnames you pass to these routines I<must> be absolute and |
114 | The pathnames you pass to these routines I<must> be absolute and |
100 | encoded in byte form. The reason for the former is that at the time the |
115 | encoded in byte form. The reason for the former is that at the time the |
101 | request is being executed, the current working directory could have |
116 | request is being executed, the current working directory could have |
102 | changed. Alternatively, you can make sure that you never change the |
117 | changed. Alternatively, you can make sure that you never change the |
103 | current working directory. |
118 | current working directory. |
… | |
… | |
108 | your pathnames to the locale (or other) encoding in effect in the user |
123 | your pathnames to the locale (or other) encoding in effect in the user |
109 | environment, d) use Glib::filename_from_unicode on unicode filenames or e) |
124 | environment, d) use Glib::filename_from_unicode on unicode filenames or e) |
110 | use something else. |
125 | use something else. |
111 | |
126 | |
112 | =over 4 |
127 | =over 4 |
|
|
128 | |
|
|
129 | =item aioreq_pri $pri |
|
|
130 | |
|
|
131 | Sets the priority for the next aio request. The default priority |
|
|
132 | is C<0>, the minimum and maximum priorities are C<-4> and C<4>, |
|
|
133 | respectively. Requests with higher priority will be serviced first. |
|
|
134 | |
|
|
135 | The priority will be reset to C<0> after each call to one of the C<aio_> |
|
|
136 | functions. |
|
|
137 | |
|
|
138 | Example: open a file with low priority, then read something from it with |
|
|
139 | higher priority so the read request is serviced before other low priority |
|
|
140 | open requests (potentially spamming the cache): |
|
|
141 | |
|
|
142 | aioreq_pri -3; |
|
|
143 | aio_open ..., sub { |
|
|
144 | return unless $_[0]; |
|
|
145 | |
|
|
146 | aioreq_pri -2; |
|
|
147 | aio_read $_[0], ..., sub { |
|
|
148 | ... |
|
|
149 | }; |
|
|
150 | }; |
|
|
151 | |
|
|
152 | =item aioreq_nice $pri_adjust |
|
|
153 | |
|
|
154 | Similar to C<aioreq_pri>, but subtracts the given value from the current |
|
|
155 | priority, so effects are cumulative. |
113 | |
156 | |
114 | =item aio_open $pathname, $flags, $mode, $callback->($fh) |
157 | =item aio_open $pathname, $flags, $mode, $callback->($fh) |
115 | |
158 | |
116 | Asynchronously open or create a file and call the callback with a newly |
159 | Asynchronously open or create a file and call the callback with a newly |
117 | created filehandle for the file. |
160 | created filehandle for the file. |
… | |
… | |
170 | print "read $_[0] bytes: <$buffer>\n"; |
213 | print "read $_[0] bytes: <$buffer>\n"; |
171 | }; |
214 | }; |
172 | |
215 | |
173 | =item aio_move $srcpath, $dstpath, $callback->($status) |
216 | =item aio_move $srcpath, $dstpath, $callback->($status) |
174 | |
217 | |
175 | [EXPERIMENTAL] |
|
|
176 | |
|
|
177 | Try to move the I<file> (directories not supported as either source or destination) |
218 | Try to move the I<file> (directories not supported as either source or |
178 | from C<$srcpath> to C<$dstpath> and call the callback with the C<0> (error) or C<-1> ok. |
219 | destination) from C<$srcpath> to C<$dstpath> and call the callback with |
|
|
220 | the C<0> (error) or C<-1> ok. |
179 | |
221 | |
180 | This is a composite request that tries to rename(2) the file first. If |
222 | This is a composite request that tries to rename(2) the file first. If |
181 | rename files with C<EXDEV>, it creates the destination file with mode 0200 |
223 | rename files with C<EXDEV>, it creates the destination file with mode 0200 |
182 | and copies the contents of the source file into it using C<aio_sendfile>, |
224 | and copies the contents of the source file into it using C<aio_sendfile>, |
183 | followed by restoring atime, mtime, access mode and uid/gid, in that |
225 | followed by restoring atime, mtime, access mode and uid/gid, in that |
… | |
… | |
190 | =cut |
232 | =cut |
191 | |
233 | |
192 | sub aio_move($$$) { |
234 | sub aio_move($$$) { |
193 | my ($src, $dst, $cb) = @_; |
235 | my ($src, $dst, $cb) = @_; |
194 | |
236 | |
|
|
237 | my $grp = aio_group $cb; |
|
|
238 | |
195 | aio_rename $src, $dst, sub { |
239 | add $grp aio_rename $src, $dst, sub { |
196 | if ($_[0] && $! == Errno::EXDEV) { |
240 | if ($_[0] && $! == EXDEV) { |
197 | aio_open $src, O_RDONLY, 0, sub { |
241 | add $grp aio_open $src, O_RDONLY, 0, sub { |
198 | if (my $src_fh = $_[0]) { |
242 | if (my $src_fh = $_[0]) { |
199 | my @stat = stat $src_fh; |
243 | my @stat = stat $src_fh; |
200 | |
244 | |
201 | aio_open $dst, O_WRONLY, 0200, sub { |
245 | add $grp aio_open $dst, O_WRONLY, 0200, sub { |
202 | if (my $dst_fh = $_[0]) { |
246 | if (my $dst_fh = $_[0]) { |
203 | aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub { |
247 | add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub { |
204 | close $src_fh; |
248 | close $src_fh; |
205 | |
249 | |
206 | if ($_[0] == $stat[7]) { |
250 | if ($_[0] == $stat[7]) { |
207 | utime $stat[8], $stat[9], $dst; |
251 | utime $stat[8], $stat[9], $dst; |
208 | chmod $stat[2] & 07777, $dst_fh; |
252 | chmod $stat[2] & 07777, $dst_fh; |
209 | chown $stat[4], $stat[5], $dst_fh; |
253 | chown $stat[4], $stat[5], $dst_fh; |
210 | close $dst_fh; |
254 | close $dst_fh; |
211 | |
255 | |
212 | aio_unlink $src, sub { |
256 | add $grp aio_unlink $src, sub { |
213 | $cb->($_[0]); |
257 | $grp->result ($_[0]); |
214 | }; |
258 | }; |
215 | } else { |
259 | } else { |
216 | my $errno = $!; |
260 | my $errno = $!; |
217 | aio_unlink $dst, sub { |
261 | add $grp aio_unlink $dst, sub { |
218 | $! = $errno; |
262 | $! = $errno; |
219 | $cb->(-1); |
263 | $grp->result (-1); |
220 | }; |
264 | }; |
221 | } |
265 | } |
222 | }; |
266 | }; |
223 | } else { |
267 | } else { |
224 | $cb->(-1); |
268 | $grp->result (-1); |
225 | } |
269 | } |
226 | }, |
270 | }, |
227 | |
271 | |
228 | } else { |
272 | } else { |
229 | $cb->(-1); |
273 | $grp->result (-1); |
230 | } |
274 | } |
231 | }; |
275 | }; |
232 | } else { |
276 | } else { |
233 | $cb->($_[0]); |
277 | $grp->result ($_[0]); |
234 | } |
278 | } |
235 | }; |
279 | }; |
|
|
280 | |
|
|
281 | $grp |
236 | } |
282 | } |
237 | |
283 | |
238 | =item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) |
284 | =item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) |
239 | |
285 | |
240 | Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts |
286 | Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts |
… | |
… | |
328 | The callback a single argument which is either C<undef> or an array-ref |
374 | The callback a single argument which is either C<undef> or an array-ref |
329 | with the filenames. |
375 | with the filenames. |
330 | |
376 | |
331 | =item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) |
377 | =item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) |
332 | |
378 | |
333 | Scans a directory (similar to C<aio_readdir>) and tries to separate the |
379 | Scans a directory (similar to C<aio_readdir>) but additionally tries to |
334 | entries of directory C<$path> into two sets of names, ones you can recurse |
380 | separate the entries of directory C<$path> into two sets of names, ones |
335 | into (directories), and ones you cannot recurse into (everything else). |
381 | you can recurse into (directories or links to them), and ones you cannot |
|
|
382 | recurse into (everything else). |
336 | |
383 | |
337 | C<aio_scandir> is a composite request that consists of many |
384 | C<aio_scandir> is a composite request that creates of many sub requests_ |
338 | aio-primitives. C<$maxreq> specifies the maximum number of outstanding |
385 | C<$maxreq> specifies the maximum number of outstanding aio requests that |
339 | aio requests that this function generates. If it is C<< <= 0 >>, then a |
386 | this function generates. If it is C<< <= 0 >>, then a suitable default |
340 | suitable default will be chosen (currently 8). |
387 | will be chosen (currently 6). |
341 | |
388 | |
342 | On error, the callback is called without arguments, otherwise it receives |
389 | On error, the callback is called without arguments, otherwise it receives |
343 | two array-refs with path-relative entry names. |
390 | two array-refs with path-relative entry names. |
344 | |
391 | |
345 | Example: |
392 | Example: |
… | |
… | |
353 | Implementation notes. |
400 | Implementation notes. |
354 | |
401 | |
355 | The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can. |
402 | The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can. |
356 | |
403 | |
357 | After reading the directory, the modification time, size etc. of the |
404 | After reading the directory, the modification time, size etc. of the |
358 | directory before and after the readdir is checked, and if they match, the |
405 | directory before and after the readdir is checked, and if they match (and |
359 | link count will be used to decide how many entries are directories (if |
406 | isn't the current time), the link count will be used to decide how many |
360 | >= 2). Otherwise, no knowledge of the number of subdirectories will be |
407 | entries are directories (if >= 2). Otherwise, no knowledge of the number |
361 | assumed. |
408 | of subdirectories will be assumed. |
362 | |
409 | |
363 | Then entires will be sorted into likely directories (everything without a |
410 | Then entries will be sorted into likely directories (everything without |
364 | non-initial dot) and likely non-directories (everything else). Then every |
411 | a non-initial dot currently) and likely non-directories (everything |
365 | entry + C</.> will be C<stat>'ed, likely directories first. This is often |
412 | else). Then every entry plus an appended C</.> will be C<stat>'ed, |
|
|
413 | likely directories first. If that succeeds, it assumes that the entry |
|
|
414 | is a directory or a symlink to directory (which will be checked |
|
|
415 | seperately). This is often faster than stat'ing the entry itself because |
366 | faster because filesystems might detect the type of the entry without |
416 | filesystems might detect the type of the entry without reading the inode |
367 | reading the inode data (e.g. ext2fs filetype feature). If that succeeds, |
417 | data (e.g. ext2fs filetype feature). |
368 | it assumes that the entry is a directory or a symlink to directory (which |
|
|
369 | will be checked seperately). |
|
|
370 | |
418 | |
371 | If the known number of directories has been reached, the rest of the |
419 | If the known number of directories (link count - 2) has been reached, the |
372 | entries is assumed to be non-directories. |
420 | rest of the entries is assumed to be non-directories. |
|
|
421 | |
|
|
422 | This only works with certainty on POSIX (= UNIX) filesystems, which |
|
|
423 | fortunately are the vast majority of filesystems around. |
|
|
424 | |
|
|
425 | It will also likely work on non-POSIX filesystems with reduced efficiency |
|
|
426 | as those tend to return 0 or 1 as link counts, which disables the |
|
|
427 | directory counting heuristic. |
373 | |
428 | |
374 | =cut |
429 | =cut |
375 | |
430 | |
376 | sub aio_scandir($$$) { |
431 | sub aio_scandir($$$) { |
377 | my ($path, $maxreq, $cb) = @_; |
432 | my ($path, $maxreq, $cb) = @_; |
378 | |
433 | |
|
|
434 | my $grp = aio_group $cb; |
|
|
435 | |
379 | $maxreq = 8 if $maxreq <= 0; |
436 | $maxreq = 6 if $maxreq <= 0; |
380 | |
437 | |
381 | # stat once |
438 | # stat once |
382 | aio_stat $path, sub { |
439 | add $grp aio_stat $path, sub { |
383 | return $cb->() if $_[0]; |
440 | return $grp->result () if $_[0]; |
|
|
441 | my $now = time; |
384 | my $hash1 = join ":", (stat _)[0,1,3,7,9]; |
442 | my $hash1 = join ":", (stat _)[0,1,3,7,9]; |
385 | |
443 | |
386 | # read the directory entries |
444 | # read the directory entries |
387 | aio_readdir $path, sub { |
445 | add $grp aio_readdir $path, sub { |
388 | my $entries = shift |
446 | my $entries = shift |
389 | or return $cb->(); |
447 | or return $grp->result (); |
390 | |
448 | |
391 | # stat the dir another time |
449 | # stat the dir another time |
392 | aio_stat $path, sub { |
450 | add $grp aio_stat $path, sub { |
393 | my $hash2 = join ":", (stat _)[0,1,3,7,9]; |
451 | my $hash2 = join ":", (stat _)[0,1,3,7,9]; |
394 | |
452 | |
395 | my $ndirs; |
453 | my $ndirs; |
396 | |
454 | |
397 | # take the slow route if anything looks fishy |
455 | # take the slow route if anything looks fishy |
398 | if ($hash1 ne $hash2) { |
456 | if ($hash1 ne $hash2 or (stat _)[9] == $now) { |
399 | $ndirs = -1; |
457 | $ndirs = -1; |
400 | } else { |
458 | } else { |
401 | # if nlink == 2, we are finished |
459 | # if nlink == 2, we are finished |
402 | # on non-posix-fs's, we rely on nlink < 2 |
460 | # on non-posix-fs's, we rely on nlink < 2 |
403 | $ndirs = (stat _)[3] - 2 |
461 | $ndirs = (stat _)[3] - 2 |
404 | or return $cb->([], $entries); |
462 | or return $grp->result ([], $entries); |
405 | } |
463 | } |
406 | |
464 | |
407 | # sort into likely dirs and likely nondirs |
465 | # sort into likely dirs and likely nondirs |
408 | # dirs == files without ".", short entries first |
466 | # dirs == files without ".", short entries first |
409 | $entries = [map $_->[0], |
467 | $entries = [map $_->[0], |
… | |
… | |
414 | my (@dirs, @nondirs); |
472 | my (@dirs, @nondirs); |
415 | |
473 | |
416 | my ($statcb, $schedcb); |
474 | my ($statcb, $schedcb); |
417 | my $nreq = 0; |
475 | my $nreq = 0; |
418 | |
476 | |
|
|
477 | my $statgrp = add $grp aio_group; |
|
|
478 | |
419 | $schedcb = sub { |
479 | $schedcb = sub { |
420 | if (@$entries) { |
480 | if (@$entries) { |
421 | if ($nreq < $maxreq) { |
481 | if ($nreq < $maxreq) { |
422 | my $ent = pop @$entries; |
482 | my $ent = pop @$entries; |
423 | $nreq++; |
483 | $nreq++; |
424 | aio_stat "$path/$ent/.", sub { $statcb->($_[0], $ent) }; |
484 | add $statgrp aio_stat "$path/$ent/.", sub { $statcb->($_[0], $ent) }; |
425 | } |
485 | } |
426 | } elsif (!$nreq) { |
486 | } elsif (!$nreq) { |
427 | # finished |
487 | # finished |
|
|
488 | $statgrp->cancel; |
428 | undef $statcb; |
489 | undef $statcb; |
429 | undef $schedcb; |
490 | undef $schedcb; |
430 | $cb->(\@dirs, \@nondirs) if $cb; |
491 | $grp->result (\@dirs, \@nondirs); |
431 | undef $cb; |
|
|
432 | } |
492 | } |
433 | }; |
493 | }; |
434 | $statcb = sub { |
494 | $statcb = sub { |
435 | my ($status, $entry) = @_; |
495 | my ($status, $entry) = @_; |
436 | |
496 | |
… | |
… | |
438 | $nreq--; |
498 | $nreq--; |
439 | push @nondirs, $entry; |
499 | push @nondirs, $entry; |
440 | &$schedcb; |
500 | &$schedcb; |
441 | } else { |
501 | } else { |
442 | # need to check for real directory |
502 | # need to check for real directory |
443 | aio_lstat "$path/$entry", sub { |
503 | add $grp aio_lstat "$path/$entry", sub { |
444 | $nreq--; |
504 | $nreq--; |
445 | |
505 | |
446 | if (-d _) { |
506 | if (-d _) { |
447 | push @dirs, $entry; |
507 | push @dirs, $entry; |
448 | |
508 | |
… | |
… | |
461 | |
521 | |
462 | &$schedcb while @$entries && $nreq < $maxreq; |
522 | &$schedcb while @$entries && $nreq < $maxreq; |
463 | }; |
523 | }; |
464 | }; |
524 | }; |
465 | }; |
525 | }; |
|
|
526 | |
|
|
527 | $grp |
466 | } |
528 | } |
467 | |
529 | |
468 | =item aio_fsync $fh, $callback->($status) |
530 | =item aio_fsync $fh, $callback->($status) |
469 | |
531 | |
470 | Asynchronously call fsync on the given filehandle and call the callback |
532 | Asynchronously call fsync on the given filehandle and call the callback |
… | |
… | |
475 | Asynchronously call fdatasync on the given filehandle and call the |
537 | Asynchronously call fdatasync on the given filehandle and call the |
476 | callback with the fdatasync result code. |
538 | callback with the fdatasync result code. |
477 | |
539 | |
478 | If this call isn't available because your OS lacks it or it couldn't be |
540 | If this call isn't available because your OS lacks it or it couldn't be |
479 | detected, it will be emulated by calling C<fsync> instead. |
541 | detected, it will be emulated by calling C<fsync> instead. |
|
|
542 | |
|
|
543 | =item aio_group $callback->(...) |
|
|
544 | |
|
|
545 | [EXPERIMENTAL] |
|
|
546 | |
|
|
547 | This is a very special aio request: Instead of doing something, it is a |
|
|
548 | container for other aio requests, which is useful if you want to bundle |
|
|
549 | many requests into a single, composite, request with a definite callback |
|
|
550 | and the ability to cancel the whole request with its subrequests. |
|
|
551 | |
|
|
552 | Returns an object of class L<IO::AIO::GRP>. See its documentation below |
|
|
553 | for more info. |
|
|
554 | |
|
|
555 | Example: |
|
|
556 | |
|
|
557 | my $grp = aio_group sub { |
|
|
558 | print "all stats done\n"; |
|
|
559 | }; |
|
|
560 | |
|
|
561 | add $grp |
|
|
562 | (aio_stat ...), |
|
|
563 | (aio_stat ...), |
|
|
564 | ...; |
|
|
565 | |
|
|
566 | =item aio_nop $callback->() |
|
|
567 | |
|
|
568 | This is a special request - it does nothing in itself and is only used for |
|
|
569 | side effects, such as when you want to add a dummy request to a group so |
|
|
570 | that finishing the requests in the group depends on executing the given |
|
|
571 | code. |
|
|
572 | |
|
|
573 | While this request does nothing, it still goes through the execution |
|
|
574 | phase and still requires a worker thread. Thus, the callback will not |
|
|
575 | be executed immediately but only after other requests in the queue have |
|
|
576 | entered their execution phase. This can be used to measure request |
|
|
577 | latency. |
|
|
578 | |
|
|
579 | =item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED* |
|
|
580 | |
|
|
581 | Mainly used for debugging and benchmarking, this aio request puts one of |
|
|
582 | the request workers to sleep for the given time. |
|
|
583 | |
|
|
584 | While it is theoretically handy to have simple I/O scheduling requests |
|
|
585 | like sleep and file handle readable/writable, the overhead this creates is |
|
|
586 | immense (it blocks a thread for a long time) so do not use this function |
|
|
587 | except to put your application under artificial I/O pressure. |
|
|
588 | |
|
|
589 | =back |
|
|
590 | |
|
|
591 | =head2 IO::AIO::REQ CLASS |
|
|
592 | |
|
|
593 | All non-aggregate C<aio_*> functions return an object of this class when |
|
|
594 | called in non-void context. |
|
|
595 | |
|
|
596 | A request always moves through the following five states in its lifetime, |
|
|
597 | in order: B<ready> (request has been created, but has not been executed |
|
|
598 | yet), B<execute> (request is currently being executed), B<pending> |
|
|
599 | (request has been executed but callback has not been called yet), |
|
|
600 | B<result> (results are being processed synchronously, includes calling the |
|
|
601 | callback) and B<done> (request has reached the end of its lifetime and |
|
|
602 | holds no resources anymore). |
|
|
603 | |
|
|
604 | =over 4 |
|
|
605 | |
|
|
606 | =item cancel $req |
|
|
607 | |
|
|
608 | Cancels the request, if possible. Has the effect of skipping execution |
|
|
609 | when entering the B<execute> state and skipping calling the callback when |
|
|
610 | entering the the B<result> state, but will leave the request otherwise |
|
|
611 | untouched. That means that requests that currently execute will not be |
|
|
612 | stopped and resources held by the request will not be freed prematurely. |
|
|
613 | |
|
|
614 | =item cb $req $callback->(...) |
|
|
615 | |
|
|
616 | Replace (or simply set) the callback registered to the request. |
|
|
617 | |
|
|
618 | =back |
|
|
619 | |
|
|
620 | =head2 IO::AIO::GRP CLASS |
|
|
621 | |
|
|
622 | This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to |
|
|
623 | objects of this class, too. |
|
|
624 | |
|
|
625 | A IO::AIO::GRP object is a special request that can contain multiple other |
|
|
626 | aio requests. |
|
|
627 | |
|
|
628 | You create one by calling the C<aio_group> constructing function with a |
|
|
629 | callback that will be called when all contained requests have entered the |
|
|
630 | C<done> state: |
|
|
631 | |
|
|
632 | my $grp = aio_group sub { |
|
|
633 | print "all requests are done\n"; |
|
|
634 | }; |
|
|
635 | |
|
|
636 | You add requests by calling the C<add> method with one or more |
|
|
637 | C<IO::AIO::REQ> objects: |
|
|
638 | |
|
|
639 | $grp->add (aio_unlink "..."); |
|
|
640 | |
|
|
641 | add $grp aio_stat "...", sub { |
|
|
642 | $_[0] or return $grp->result ("error"); |
|
|
643 | |
|
|
644 | # add another request dynamically, if first succeeded |
|
|
645 | add $grp aio_open "...", sub { |
|
|
646 | $grp->result ("ok"); |
|
|
647 | }; |
|
|
648 | }; |
|
|
649 | |
|
|
650 | This makes it very easy to create composite requests (see the source of |
|
|
651 | C<aio_move> for an application) that work and feel like simple requests. |
|
|
652 | |
|
|
653 | =over 4 |
|
|
654 | |
|
|
655 | =item * The IO::AIO::GRP objects will be cleaned up during calls to |
|
|
656 | C<IO::AIO::poll_cb>, just like any other request. |
|
|
657 | |
|
|
658 | =item * They can be canceled like any other request. Canceling will cancel not |
|
|
659 | only the request itself, but also all requests it contains. |
|
|
660 | |
|
|
661 | =item * They can also can also be added to other IO::AIO::GRP objects. |
|
|
662 | |
|
|
663 | =item * You must not add requests to a group from within the group callback (or |
|
|
664 | any later time). |
|
|
665 | |
|
|
666 | =item * This does not harmonise well with C<max_outstanding>, so best do |
|
|
667 | not combine C<aio_group> with it. Groups and feeders are recommended for |
|
|
668 | this kind of concurrency-limiting. |
|
|
669 | |
|
|
670 | =back |
|
|
671 | |
|
|
672 | Their lifetime, simplified, looks like this: when they are empty, they |
|
|
673 | will finish very quickly. If they contain only requests that are in the |
|
|
674 | C<done> state, they will also finish. Otherwise they will continue to |
|
|
675 | exist. |
|
|
676 | |
|
|
677 | That means after creating a group you have some time to add requests. And |
|
|
678 | in the callbacks of those requests, you can add further requests to the |
|
|
679 | group. And only when all those requests have finished will the the group |
|
|
680 | itself finish. |
|
|
681 | |
|
|
682 | =over 4 |
|
|
683 | |
|
|
684 | =item add $grp ... |
|
|
685 | |
|
|
686 | =item $grp->add (...) |
|
|
687 | |
|
|
688 | Add one or more requests to the group. Any type of L<IO::AIO::REQ> can |
|
|
689 | be added, including other groups, as long as you do not create circular |
|
|
690 | dependencies. |
|
|
691 | |
|
|
692 | Returns all its arguments. |
|
|
693 | |
|
|
694 | =item $grp->result (...) |
|
|
695 | |
|
|
696 | Set the result value(s) that will be passed to the group callback when all |
|
|
697 | subrequests have finished. By default, no argument will be passed. |
|
|
698 | |
|
|
699 | =item feed $grp $callback->($grp) |
|
|
700 | |
|
|
701 | [VERY EXPERIMENTAL] |
|
|
702 | |
|
|
703 | Sets a feeder/generator on this group: every group can have an attached |
|
|
704 | generator that generates requests if idle. The idea behind this is that, |
|
|
705 | although you could just queue as many requests as you want in a group, |
|
|
706 | this might starve other requests for a potentially long time. For |
|
|
707 | example, C<aio_scandir> might generate hundreds of thousands C<aio_stat> |
|
|
708 | requests, delaying any later requests for a long time. |
|
|
709 | |
|
|
710 | To avoid this, and allow incremental generation of requests, you can |
|
|
711 | instead a group and set a feeder on it that generates those requests. The |
|
|
712 | feed callback will be called whenever there are few enough (see C<limit>, |
|
|
713 | below) requests active in the group itself and is expected to queue more |
|
|
714 | requests. |
|
|
715 | |
|
|
716 | The feed callback can queue as many requests as it likes (i.e. C<add> does |
|
|
717 | not impose any limits). |
|
|
718 | |
|
|
719 | If the feed does not queue more requests when called, it will be |
|
|
720 | automatically removed from the group. |
|
|
721 | |
|
|
722 | If the feed limit is C<0>, it will be set to C<2> automatically. |
|
|
723 | |
|
|
724 | Example: |
|
|
725 | |
|
|
726 | # stat all files in @files, but only ever use four aio requests concurrently: |
|
|
727 | |
|
|
728 | my $grp = aio_group sub { print "finished\n" }; |
|
|
729 | limit $grp 4; |
|
|
730 | feed $grp sub { |
|
|
731 | my $file = pop @files |
|
|
732 | or return; |
|
|
733 | |
|
|
734 | add $grp aio_stat $file, sub { ... }; |
|
|
735 | }; |
|
|
736 | |
|
|
737 | =item limit $grp $num |
|
|
738 | |
|
|
739 | Sets the feeder limit for the group: The feeder will be called whenever |
|
|
740 | the group contains less than this many requests. |
|
|
741 | |
|
|
742 | Setting the limit to C<0> will pause the feeding process. |
480 | |
743 | |
481 | =back |
744 | =back |
482 | |
745 | |
483 | =head2 SUPPORT FUNCTIONS |
746 | =head2 SUPPORT FUNCTIONS |
484 | |
747 | |
… | |
… | |
542 | IO::AIO::poll_wait, IO::AIO::poll_cb |
805 | IO::AIO::poll_wait, IO::AIO::poll_cb |
543 | if IO::AIO::nreqs; |
806 | if IO::AIO::nreqs; |
544 | |
807 | |
545 | =item IO::AIO::min_parallel $nthreads |
808 | =item IO::AIO::min_parallel $nthreads |
546 | |
809 | |
547 | Set the minimum number of AIO threads to C<$nthreads>. The current default |
810 | Set the minimum number of AIO threads to C<$nthreads>. The current |
548 | is C<4>, which means four asynchronous operations can be done at one time |
811 | default is C<8>, which means eight asynchronous operations can execute |
549 | (the number of outstanding operations, however, is unlimited). |
812 | concurrently at any one time (the number of outstanding requests, |
|
|
813 | however, is unlimited). |
550 | |
814 | |
551 | IO::AIO starts threads only on demand, when an AIO request is queued and |
815 | IO::AIO starts threads only on demand, when an AIO request is queued and |
552 | no free thread exists. |
816 | no free thread exists. |
553 | |
817 | |
554 | It is recommended to keep the number of threads low, as some Linux |
818 | It is recommended to keep the number of threads relatively low, as some |
555 | kernel versions will scale negatively with the number of threads (higher |
819 | Linux kernel versions will scale negatively with the number of threads |
556 | parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32 |
820 | (higher parallelity => MUCH higher latency). With current Linux 2.6 |
557 | threads should be fine. |
821 | versions, 4-32 threads should be fine. |
558 | |
822 | |
559 | Under most circumstances you don't need to call this function, as the |
823 | Under most circumstances you don't need to call this function, as the |
560 | module selects a default that is suitable for low to moderate load. |
824 | module selects a default that is suitable for low to moderate load. |
561 | |
825 | |
562 | =item IO::AIO::max_parallel $nthreads |
826 | =item IO::AIO::max_parallel $nthreads |
… | |
… | |
573 | |
837 | |
574 | Under normal circumstances you don't need to call this function. |
838 | Under normal circumstances you don't need to call this function. |
575 | |
839 | |
576 | =item $oldnreqs = IO::AIO::max_outstanding $nreqs |
840 | =item $oldnreqs = IO::AIO::max_outstanding $nreqs |
577 | |
841 | |
|
|
842 | [DEPRECATED] |
|
|
843 | |
578 | Sets the maximum number of outstanding requests to C<$nreqs>. If you |
844 | Sets the maximum number of outstanding requests to C<$nreqs>. If you |
579 | try to queue up more than this number of requests, the caller will block until |
845 | try to queue up more than this number of requests, the caller will block until |
580 | some requests have been handled. |
846 | some requests have been handled. |
581 | |
847 | |
582 | The default is very large, so normally there is no practical limit. If you |
848 | The default is very large, so normally there is no practical limit. If you |
583 | queue up many requests in a loop it often improves speed if you set |
849 | queue up many requests in a loop it often improves speed if you set |
584 | this to a relatively low number, such as C<100>. |
850 | this to a relatively low number, such as C<100>. |
|
|
851 | |
|
|
852 | This function does not work well together with C<aio_group>'s, and their |
|
|
853 | feeder interface is better suited to limiting concurrency, so do not use |
|
|
854 | this function. |
585 | |
855 | |
586 | Under normal circumstances you don't need to call this function. |
856 | Under normal circumstances you don't need to call this function. |
587 | |
857 | |
588 | =back |
858 | =back |
589 | |
859 | |
… | |
… | |
603 | or return undef; |
873 | or return undef; |
604 | |
874 | |
605 | *$sym |
875 | *$sym |
606 | } |
876 | } |
607 | |
877 | |
608 | min_parallel 4; |
878 | min_parallel 8; |
609 | |
879 | |
610 | END { |
880 | END { |
611 | max_parallel 0; |
881 | max_parallel 0; |
612 | } |
882 | } |
613 | |
883 | |
614 | 1; |
884 | 1; |
615 | |
885 | |
616 | =head2 FORK BEHAVIOUR |
886 | =head2 FORK BEHAVIOUR |
|
|
887 | |
|
|
888 | This module should do "the right thing" when the process using it forks: |
617 | |
889 | |
618 | Before the fork, IO::AIO enters a quiescent state where no requests |
890 | Before the fork, IO::AIO enters a quiescent state where no requests |
619 | can be added in other threads and no results will be processed. After |
891 | can be added in other threads and no results will be processed. After |
620 | the fork the parent simply leaves the quiescent state and continues |
892 | the fork the parent simply leaves the quiescent state and continues |
621 | request/result processing, while the child clears the request/result |
893 | request/result processing, while the child clears the request/result |
622 | queue (so the requests started before the fork will only be handled in |
894 | queue (so the requests started before the fork will only be handled in |
623 | the parent). Threats will be started on demand until the limit ste in the |
895 | the parent). Threads will be started on demand until the limit ste in the |
624 | parent process has been reached again. |
896 | parent process has been reached again. |
625 | |
897 | |
|
|
898 | In short: the parent will, after a short pause, continue as if fork had |
|
|
899 | not been called, while the child will act as if IO::AIO has not been used |
|
|
900 | yet. |
|
|
901 | |
|
|
902 | =head2 MEMORY USAGE |
|
|
903 | |
|
|
904 | Each aio request uses - depending on your architecture - around 128 bytes |
|
|
905 | of memory. In addition, stat requests need a stat buffer (possibly a few |
|
|
906 | hundred bytes). Perl scalars and other data passed into aio requests will |
|
|
907 | also be locked. |
|
|
908 | |
|
|
909 | This is now awfully much, so queuing lots of requests is not usually a |
|
|
910 | problem. |
|
|
911 | |
|
|
912 | Each thread needs a stack area which is usually around 16k, sometimes much |
|
|
913 | larger, depending on the OS. |
|
|
914 | |
626 | =head1 SEE ALSO |
915 | =head1 SEE ALSO |
627 | |
916 | |
628 | L<Coro>, L<Linux::AIO>. |
917 | L<Coro::AIO>. |
629 | |
918 | |
630 | =head1 AUTHOR |
919 | =head1 AUTHOR |
631 | |
920 | |
632 | Marc Lehmann <schmorp@schmorp.de> |
921 | Marc Lehmann <schmorp@schmorp.de> |
633 | http://home.schmorp.de/ |
922 | http://home.schmorp.de/ |