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 1.175 by root, Sun Jan 10 20:37:33 2010 UTC vs.
Revision 1.219 by root, Tue Mar 27 18:54:45 2012 UTC

… …
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use IO::AIO; 7 use IO::AIO;
8 8
9 aio_open "/etc/passwd", O_RDONLY, 0, sub { 9 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
10 my $fh = shift 10 my $fh = shift
11 or die "/etc/passwd: $!"; 11 or die "/etc/passwd: $!";
12 ... 12 ...
13 }; 13 };
14 14
… …
76 76
77 # register the IO::AIO callback with EV 77 # register the IO::AIO callback with EV
78 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb; 78 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
79 79
80 # queue the request to open /etc/passwd 80 # queue the request to open /etc/passwd
81 aio_open "/etc/passwd", O_RDONLY, 0, sub { 81 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
82 my $fh = shift 82 my $fh = shift
83 or die "error while opening: $!"; 83 or die "error while opening: $!";
84 84
85 # stat'ing filehandles is generally non-blocking 85 # stat'ing filehandles is generally non-blocking
86 my $size = -s $fh; 86 my $size = -s $fh;
… …
168use common::sense; 168use common::sense;
169 169
170use base 'Exporter'; 170use base 'Exporter';
171 171
172BEGIN { 172BEGIN {
173 our $VERSION = '3.5'; 173 our $VERSION = '4.12';
174 174
175 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close 175 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close
176 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx 176 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
177 aio_scandir aio_symlink aio_readlink aio_sync aio_fsync 177 aio_scandir aio_symlink aio_readlink aio_realpath aio_sync
178 aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead 178 aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range aio_fallocate
179 aio_pathsync aio_readahead
179 aio_rename aio_link aio_move aio_copy aio_group 180 aio_rename aio_link aio_move aio_copy aio_group
180 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown 181 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
181 aio_chmod aio_utime aio_truncate 182 aio_chmod aio_utime aio_truncate
182 aio_msync aio_mtouch aio_statvfs); 183 aio_msync aio_mtouch aio_mlock aio_mlockall
184 aio_statvfs
185 aio_wd);
183 186
184 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); 187 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
185 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 188 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
186 min_parallel max_parallel max_idle 189 min_parallel max_parallel max_idle idle_timeout
187 nreqs nready npending nthreads 190 nreqs nready npending nthreads
188 max_poll_time max_poll_reqs 191 max_poll_time max_poll_reqs
189 sendfile fadvise); 192 sendfile fadvise madvise
193 mmap munmap munlock munlockall);
190 194
191 push @AIO_REQ, qw(aio_busy); # not exported 195 push @AIO_REQ, qw(aio_busy); # not exported
192 196
193 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 197 @IO::AIO::GRP::ISA = 'IO::AIO::REQ';
194 198
… …
202 206
203This section simply lists the prototypes of the most important functions 207This section simply lists the prototypes of the most important functions
204for quick reference. See the following sections for function-by-function 208for quick reference. See the following sections for function-by-function
205documentation. 209documentation.
206 210
211 aio_wd $pathname, $callback->($wd)
207 aio_open $pathname, $flags, $mode, $callback->($fh) 212 aio_open $pathname, $flags, $mode, $callback->($fh)
208 aio_close $fh, $callback->($status) 213 aio_close $fh, $callback->($status)
209 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 214 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
210 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 215 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
211 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 216 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
… …
216 aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 221 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
217 aio_chown $fh_or_path, $uid, $gid, $callback->($status) 222 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
218 aio_truncate $fh_or_path, $offset, $callback->($status) 223 aio_truncate $fh_or_path, $offset, $callback->($status)
219 aio_chmod $fh_or_path, $mode, $callback->($status) 224 aio_chmod $fh_or_path, $mode, $callback->($status)
220 aio_unlink $pathname, $callback->($status) 225 aio_unlink $pathname, $callback->($status)
221 aio_mknod $path, $mode, $dev, $callback->($status) 226 aio_mknod $pathname, $mode, $dev, $callback->($status)
222 aio_link $srcpath, $dstpath, $callback->($status) 227 aio_link $srcpath, $dstpath, $callback->($status)
223 aio_symlink $srcpath, $dstpath, $callback->($status) 228 aio_symlink $srcpath, $dstpath, $callback->($status)
224 aio_readlink $path, $callback->($link) 229 aio_readlink $pathname, $callback->($link)
230 aio_realpath $pathname, $callback->($link)
225 aio_rename $srcpath, $dstpath, $callback->($status) 231 aio_rename $srcpath, $dstpath, $callback->($status)
226 aio_mkdir $pathname, $mode, $callback->($status) 232 aio_mkdir $pathname, $mode, $callback->($status)
227 aio_rmdir $pathname, $callback->($status) 233 aio_rmdir $pathname, $callback->($status)
228 aio_readdir $pathname, $callback->($entries) 234 aio_readdir $pathname, $callback->($entries)
229 aio_readdirx $pathname, $flags, $callback->($entries, $flags) 235 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
230 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST 236 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
231 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN 237 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
238 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
232 aio_load $path, $data, $callback->($status) 239 aio_load $pathname, $data, $callback->($status)
233 aio_copy $srcpath, $dstpath, $callback->($status) 240 aio_copy $srcpath, $dstpath, $callback->($status)
234 aio_move $srcpath, $dstpath, $callback->($status) 241 aio_move $srcpath, $dstpath, $callback->($status)
235 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
236 aio_rmtree $path, $callback->($status) 242 aio_rmtree $pathname, $callback->($status)
237 aio_sync $callback->($status) 243 aio_sync $callback->($status)
244 aio_syncfs $fh, $callback->($status)
238 aio_fsync $fh, $callback->($status) 245 aio_fsync $fh, $callback->($status)
239 aio_fdatasync $fh, $callback->($status) 246 aio_fdatasync $fh, $callback->($status)
240 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 247 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
241 aio_pathsync $path, $callback->($status) 248 aio_pathsync $pathname, $callback->($status)
242 aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 249 aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
243 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 250 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
251 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
252 aio_mlockall $flags, $callback->($status)
244 aio_group $callback->(...) 253 aio_group $callback->(...)
245 aio_nop $callback->() 254 aio_nop $callback->()
246 255
247 $prev_pri = aioreq_pri [$pri] 256 $prev_pri = aioreq_pri [$pri]
248 aioreq_nice $pri_adjust 257 aioreq_nice $pri_adjust
… …
254 IO::AIO::max_poll_reqs $nreqs 263 IO::AIO::max_poll_reqs $nreqs
255 IO::AIO::max_poll_time $seconds 264 IO::AIO::max_poll_time $seconds
256 IO::AIO::min_parallel $nthreads 265 IO::AIO::min_parallel $nthreads
257 IO::AIO::max_parallel $nthreads 266 IO::AIO::max_parallel $nthreads
258 IO::AIO::max_idle $nthreads 267 IO::AIO::max_idle $nthreads
268 IO::AIO::idle_timeout $seconds
259 IO::AIO::max_outstanding $maxreqs 269 IO::AIO::max_outstanding $maxreqs
260 IO::AIO::nreqs 270 IO::AIO::nreqs
261 IO::AIO::nready 271 IO::AIO::nready
262 IO::AIO::npending 272 IO::AIO::npending
263 273
264 IO::AIO::sendfile $ofh, $ifh, $offset, $count 274 IO::AIO::sendfile $ofh, $ifh, $offset, $count
265 IO::AIO::fadvise $fh, $offset, $len, $advice 275 IO::AIO::fadvise $fh, $offset, $len, $advice
266 IO::AIO::mlockall $flags 276 IO::AIO::madvise $scalar, $offset, $length, $advice
277 IO::AIO::mprotect $scalar, $offset, $length, $protect
278 IO::AIO::munlock $scalar, $offset = 0, $length = undef
267 IO::AIO::munlockall 279 IO::AIO::munlockall
268 280
269=head2 AIO REQUEST FUNCTIONS 281=head2 API NOTES
270 282
271All the C<aio_*> calls are more or less thin wrappers around the syscall 283All the C<aio_*> calls are more or less thin wrappers around the syscall
272with the same name (sans C<aio_>). The arguments are similar or identical, 284with the same name (sans C<aio_>). The arguments are similar or identical,
273and they all accept an additional (and optional) C<$callback> argument 285and they all accept an additional (and optional) C<$callback> argument
274which must be a code reference. This code reference will get called with 286which must be a code reference. This code reference will be called after
275the syscall return code (e.g. most syscalls return C<-1> on error, unlike 287the syscall has been executed in an asynchronous fashion. The results
276perl, which usually delivers "false") as its sole argument after the given 288of the request will be passed as arguments to the callback (and, if an
277syscall has been executed asynchronously. 289error occured, in C<$!>) - for most requests the syscall return code (e.g.
290most syscalls return C<-1> on error, unlike perl, which usually delivers
291"false").
292
293Some requests (such as C<aio_readdir>) pass the actual results and
294communicate failures by passing C<undef>.
278 295
279All functions expecting a filehandle keep a copy of the filehandle 296All functions expecting a filehandle keep a copy of the filehandle
280internally until the request has finished. 297internally until the request has finished.
281 298
282All functions return request objects of type L<IO::AIO::REQ> that allow 299All functions return request objects of type L<IO::AIO::REQ> that allow
283further manipulation of those requests while they are in-flight. 300further manipulation of those requests while they are in-flight.
284 301
285The pathnames you pass to these routines I<must> be absolute and 302The pathnames you pass to these routines I<should> be absolute. The
286encoded as octets. The reason for the former is that at the time the 303reason for this is that at the time the request is being executed, the
287request is being executed, the current working directory could have 304current working directory could have changed. Alternatively, you can
288changed. Alternatively, you can make sure that you never change the 305make sure that you never change the current working directory anywhere
289current working directory anywhere in the program and then use relative 306in the program and then use relative paths. You can also take advantage
290paths. 307of IO::AIOs working directory abstraction, that lets you specify paths
308relative to some previously-opened "working directory object" - see the
309description of the C<IO::AIO::WD> class later in this document.
291 310
292To encode pathnames as octets, either make sure you either: a) always pass 311To encode pathnames as octets, either make sure you either: a) always pass
293in filenames you got from outside (command line, readdir etc.) without 312in filenames you got from outside (command line, readdir etc.) without
294tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode 313tinkering, b) are in your native filesystem encoding, c) use the Encode
295your pathnames to the locale (or other) encoding in effect in the user 314module and encode your pathnames to the locale (or other) encoding in
296environment, d) use Glib::filename_from_unicode on unicode filenames or e) 315effect in the user environment, d) use Glib::filename_from_unicode on
297use something else to ensure your scalar has the correct contents. 316unicode filenames or e) use something else to ensure your scalar has the
317correct contents.
298 318
299This works, btw. independent of the internal UTF-8 bit, which IO::AIO 319This works, btw. independent of the internal UTF-8 bit, which IO::AIO
300handles correctly whether it is set or not. 320handles correctly whether it is set or not.
321
322=head2 AIO REQUEST FUNCTIONS
301 323
302=over 4 324=over 4
303 325
304=item $prev_pri = aioreq_pri [$pri] 326=item $prev_pri = aioreq_pri [$pri]
305 327
… …
352by the umask in effect then the request is being executed, so better never 374by the umask in effect then the request is being executed, so better never
353change the umask. 375change the umask.
354 376
355Example: 377Example:
356 378
357 aio_open "/etc/passwd", O_RDONLY, 0, sub { 379 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
358 if ($_[0]) { 380 if ($_[0]) {
359 print "open successful, fh is $_[0]\n"; 381 print "open successful, fh is $_[0]\n";
360 ... 382 ...
361 } else { 383 } else {
362 die "open failed: $!\n"; 384 die "open failed: $!\n";
363 } 385 }
364 }; 386 };
365 387
388In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
389C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
390following POSIX and non-POSIX constants are available (missing ones on
391your system are, as usual, C<0>):
392
393C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
394C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
395C<O_RSYNC>, C<O_SYNC> and C<O_TTY_INIT>.
396
366 397
367=item aio_close $fh, $callback->($status) 398=item aio_close $fh, $callback->($status)
368 399
369Asynchronously close a file and call the callback with the result 400Asynchronously close a file and call the callback with the result
370code. 401code.
… …
420 451
421Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 452Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
422reading at byte offset C<$in_offset>, and starts writing at the current 453reading at byte offset C<$in_offset>, and starts writing at the current
423file offset of C<$out_fh>. Because of that, it is not safe to issue more 454file offset of C<$out_fh>. Because of that, it is not safe to issue more
424than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each 455than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
425other. 456other. The same C<$in_fh> works fine though, as this function does not
457move or use the file offset of C<$in_fh>.
426 458
459Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
460are written, and there is no way to find out how many more bytes have been
461read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
462number of bytes written to C<$out_fh>. Only if the result value equals
463C<$length> one can assume that C<$length> bytes have been read.
464
465Unlike with other C<aio_> functions, it makes a lot of sense to use
466C<aio_sendfile> on non-blocking sockets, as long as one end (typically
467the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
468the socket I/O will be non-blocking. Note, however, that you can run
469into a trap where C<aio_sendfile> reads some data with readahead, then
470fails to write all data, and when the socket is ready the next time, the
471data in the cache is already lost, forcing C<aio_sendfile> to again hit
472the disk. Explicit C<aio_read> + C<aio_write> let's you better control
473resource usage.
474
427This call tries to make use of a native C<sendfile> syscall to provide 475This call tries to make use of a native C<sendfile>-like syscall to
428zero-copy operation. For this to work, C<$out_fh> should refer to a 476provide zero-copy operation. For this to work, C<$out_fh> should refer to
429socket, and C<$in_fh> should refer to mmap'able file. 477a socket, and C<$in_fh> should refer to an mmap'able file.
430 478
431If a native sendfile cannot be found or it fails with C<ENOSYS>, 479If a native sendfile cannot be found or it fails with C<ENOSYS>,
432C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or C<ENOTSOCK>, 480C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
433it will be emulated, so you can call C<aio_sendfile> on any type of 481C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
434filehandle regardless of the limitations of the operating system. 482type of filehandle regardless of the limitations of the operating system.
435 483
436Please note, however, that C<aio_sendfile> can read more bytes from 484As native sendfile syscalls (as practically any non-POSIX interface hacked
437C<$in_fh> than are written, and there is no way to find out how many 485together in a hurry to improve benchmark numbers) tend to be rather buggy
438bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only 486on many systems, this implementation tries to work around some known bugs
439provides the number of bytes written to C<$out_fh>. Only if the result 487in Linux and FreeBSD kernels (probably others, too), but that might fail,
440value equals C<$length> one can assume that C<$length> bytes have been 488so you really really should check the return value of C<aio_sendfile> -
441read. 489fewre bytes than expected might have been transferred.
442 490
443 491
444=item aio_readahead $fh,$offset,$length, $callback->($retval) 492=item aio_readahead $fh,$offset,$length, $callback->($retval)
445 493
446C<aio_readahead> populates the page cache with data from a file so that 494C<aio_readahead> populates the page cache with data from a file so that
… …
468for an explanation. 516for an explanation.
469 517
470Currently, the stats are always 64-bit-stats, i.e. instead of returning an 518Currently, the stats are always 64-bit-stats, i.e. instead of returning an
471error when stat'ing a large file, the results will be silently truncated 519error when stat'ing a large file, the results will be silently truncated
472unless perl itself is compiled with large file support. 520unless perl itself is compiled with large file support.
521
522To help interpret the mode and dev/rdev stat values, IO::AIO offers the
523following constants and functions (if not implemented, the constants will
524be C<0> and the functions will either C<croak> or fall back on traditional
525behaviour).
526
527C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
528C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
529C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
473 530
474Example: Print the length of F</etc/passwd>: 531Example: Print the length of F</etc/passwd>:
475 532
476 aio_stat "/etc/passwd", sub { 533 aio_stat "/etc/passwd", sub {
477 $_[0] and die "stat failed: $!"; 534 $_[0] and die "stat failed: $!";
… …
569 626
570Asynchronously unlink (delete) a file and call the callback with the 627Asynchronously unlink (delete) a file and call the callback with the
571result code. 628result code.
572 629
573 630
574=item aio_mknod $path, $mode, $dev, $callback->($status) 631=item aio_mknod $pathname, $mode, $dev, $callback->($status)
575 632
576[EXPERIMENTAL] 633[EXPERIMENTAL]
577 634
578Asynchronously create a device node (or fifo). See mknod(2). 635Asynchronously create a device node (or fifo). See mknod(2).
579 636
580The only (POSIX-) portable way of calling this function is: 637The only (POSIX-) portable way of calling this function is:
581 638
582 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 639 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
583 640
641See C<aio_stat> for info about some potentially helpful extra constants
642and functions.
584 643
585=item aio_link $srcpath, $dstpath, $callback->($status) 644=item aio_link $srcpath, $dstpath, $callback->($status)
586 645
587Asynchronously create a new link to the existing object at C<$srcpath> at 646Asynchronously create a new link to the existing object at C<$srcpath> at
588the path C<$dstpath> and call the callback with the result code. 647the path C<$dstpath> and call the callback with the result code.
… …
592 651
593Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 652Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
594the path C<$dstpath> and call the callback with the result code. 653the path C<$dstpath> and call the callback with the result code.
595 654
596 655
597=item aio_readlink $path, $callback->($link) 656=item aio_readlink $pathname, $callback->($link)
598 657
599Asynchronously read the symlink specified by C<$path> and pass it to 658Asynchronously read the symlink specified by C<$path> and pass it to
600the callback. If an error occurs, nothing or undef gets passed to the 659the callback. If an error occurs, nothing or undef gets passed to the
601callback. 660callback.
602 661
603 662
663=item aio_realpath $pathname, $callback->($path)
664
665Asynchronously make the path absolute and resolve any symlinks in
666C<$path>. The resulting path only consists of directories (Same as
667L<Cwd::realpath>).
668
669This request can be used to get the absolute path of the current working
670directory by passing it a path of F<.> (a single dot).
671
672
604=item aio_rename $srcpath, $dstpath, $callback->($status) 673=item aio_rename $srcpath, $dstpath, $callback->($status)
605 674
606Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 675Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
607rename(2) and call the callback with the result code. 676rename(2) and call the callback with the result code.
608 677
… …
630array-ref with the filenames. 699array-ref with the filenames.
631 700
632 701
633=item aio_readdirx $pathname, $flags, $callback->($entries, $flags) 702=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
634 703
635Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune 704Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
636behaviour and output format. In case of an error, C<$entries> will be 705tune behaviour and output format. In case of an error, C<$entries> will be
637C<undef>. 706C<undef>.
638 707
639The flags are a combination of the following constants, ORed together (the 708The flags are a combination of the following constants, ORed together (the
640flags will also be passed to the callback, possibly modified): 709flags will also be passed to the callback, possibly modified):
641 710
642=over 4 711=over 4
643 712
644=item IO::AIO::READDIR_DENTS 713=item IO::AIO::READDIR_DENTS
645 714
646When this flag is off, then the callback gets an arrayref with of names 715When this flag is off, then the callback gets an arrayref consisting of
647only (as with C<aio_readdir>), otherwise it gets an arrayref with 716names only (as with C<aio_readdir>), otherwise it gets an arrayref with
648C<[$name, $type, $inode]> arrayrefs, each describing a single directory 717C<[$name, $type, $inode]> arrayrefs, each describing a single directory
649entry in more detail. 718entry in more detail.
650 719
651C<$name> is the name of the entry. 720C<$name> is the name of the entry.
652 721
… …
665systems that do not deliver the inode information. 734systems that do not deliver the inode information.
666 735
667=item IO::AIO::READDIR_DIRS_FIRST 736=item IO::AIO::READDIR_DIRS_FIRST
668 737
669When this flag is set, then the names will be returned in an order where 738When this flag is set, then the names will be returned in an order where
670likely directories come first. This is useful when you need to quickly 739likely directories come first, in optimal stat order. This is useful when
671find directories, or you want to find all directories while avoiding to 740you need to quickly find directories, or you want to find all directories
672stat() each entry. 741while avoiding to stat() each entry.
673 742
674If the system returns type information in readdir, then this is used 743If the system returns type information in readdir, then this is used
675to find directories directly. Otherwise, likely directories are files 744to find directories directly. Otherwise, likely directories are names
676beginning with ".", or otherwise files with no dots, of which files with 745beginning with ".", or otherwise names with no dots, of which names with
677short names are tried first. 746short names are tried first.
678 747
679=item IO::AIO::READDIR_STAT_ORDER 748=item IO::AIO::READDIR_STAT_ORDER
680 749
681When this flag is set, then the names will be returned in an order 750When this flag is set, then the names will be returned in an order
… …
688 757
689=item IO::AIO::READDIR_FOUND_UNKNOWN 758=item IO::AIO::READDIR_FOUND_UNKNOWN
690 759
691This flag should not be set when calling C<aio_readdirx>. Instead, it 760This flag should not be set when calling C<aio_readdirx>. Instead, it
692is being set by C<aio_readdirx>, when any of the C<$type>'s found were 761is being set by C<aio_readdirx>, when any of the C<$type>'s found were
693C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all 762C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
694C<$type>'s are known, which can be used to speed up some algorithms. 763C<$type>'s are known, which can be used to speed up some algorithms.
695 764
696=back 765=back
697 766
698 767
699=item aio_load $path, $data, $callback->($status) 768=item aio_load $pathname, $data, $callback->($status)
700 769
701This is a composite request that tries to fully load the given file into 770This is a composite request that tries to fully load the given file into
702memory. Status is the same as with aio_read. 771memory. Status is the same as with aio_read.
703 772
704=cut 773=cut
… …
826 if ($_[0] && $! == EXDEV) { 895 if ($_[0] && $! == EXDEV) {
827 aioreq_pri $pri; 896 aioreq_pri $pri;
828 add $grp aio_copy $src, $dst, sub { 897 add $grp aio_copy $src, $dst, sub {
829 $grp->result ($_[0]); 898 $grp->result ($_[0]);
830 899
831 if (!$_[0]) { 900 unless ($_[0]) {
832 aioreq_pri $pri; 901 aioreq_pri $pri;
833 add $grp aio_unlink $src; 902 add $grp aio_unlink $src;
834 } 903 }
835 }; 904 };
836 } else { 905 } else {
… …
839 }; 908 };
840 909
841 $grp 910 $grp
842} 911}
843 912
844=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 913=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
845 914
846Scans a directory (similar to C<aio_readdir>) but additionally tries to 915Scans a directory (similar to C<aio_readdir>) but additionally tries to
847efficiently separate the entries of directory C<$path> into two sets of 916efficiently separate the entries of directory C<$path> into two sets of
848names, directories you can recurse into (directories), and ones you cannot 917names, directories you can recurse into (directories), and ones you cannot
849recurse into (everything else, including symlinks to directories). 918recurse into (everything else, including symlinks to directories).
… …
880Then entries will be sorted into likely directories a non-initial dot 949Then entries will be sorted into likely directories a non-initial dot
881currently) and likely non-directories (see C<aio_readdirx>). Then every 950currently) and likely non-directories (see C<aio_readdirx>). Then every
882entry plus an appended C</.> will be C<stat>'ed, likely directories first, 951entry plus an appended C</.> will be C<stat>'ed, likely directories first,
883in order of their inode numbers. If that succeeds, it assumes that the 952in order of their inode numbers. If that succeeds, it assumes that the
884entry is a directory or a symlink to directory (which will be checked 953entry is a directory or a symlink to directory (which will be checked
885seperately). This is often faster than stat'ing the entry itself because 954separately). This is often faster than stat'ing the entry itself because
886filesystems might detect the type of the entry without reading the inode 955filesystems might detect the type of the entry without reading the inode
887data (e.g. ext2fs filetype feature), even on systems that cannot return 956data (e.g. ext2fs filetype feature), even on systems that cannot return
888the filetype information on readdir. 957the filetype information on readdir.
889 958
890If the known number of directories (link count - 2) has been reached, the 959If the known number of directories (link count - 2) has been reached, the
… …
906 975
907 my $grp = aio_group $cb; 976 my $grp = aio_group $cb;
908 977
909 $maxreq = 4 if $maxreq <= 0; 978 $maxreq = 4 if $maxreq <= 0;
910 979
911 # stat once 980 # get a wd object
912 aioreq_pri $pri; 981 aioreq_pri $pri;
913 add $grp aio_stat $path, sub { 982 add $grp aio_wd $path, sub {
983 $_[0]
914 return $grp->result () if $_[0]; 984 or return $grp->result ();
915 my $now = time;
916 my $hash1 = join ":", (stat _)[0,1,3,7,9];
917 985
918 # read the directory entries 986 my $wd = [shift, "."];
987
988 # stat once
919 aioreq_pri $pri; 989 aioreq_pri $pri;
920 add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub { 990 add $grp aio_stat $wd, sub {
921 my $entries = shift
922 or return $grp->result (); 991 return $grp->result () if $_[0];
992 my $now = time;
993 my $hash1 = join ":", (stat _)[0,1,3,7,9];
923 994
924 # stat the dir another time 995 # read the directory entries
925 aioreq_pri $pri; 996 aioreq_pri $pri;
997 add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
998 my $entries = shift
999 or return $grp->result ();
1000
1001 # stat the dir another time
1002 aioreq_pri $pri;
926 add $grp aio_stat $path, sub { 1003 add $grp aio_stat $wd, sub {
927 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 1004 my $hash2 = join ":", (stat _)[0,1,3,7,9];
928 1005
929 my $ndirs; 1006 my $ndirs;
930 1007
931 # take the slow route if anything looks fishy 1008 # take the slow route if anything looks fishy
932 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 1009 if ($hash1 ne $hash2 or (stat _)[9] == $now) {
933 $ndirs = -1; 1010 $ndirs = -1;
934 } else { 1011 } else {
935 # if nlink == 2, we are finished 1012 # if nlink == 2, we are finished
936 # for non-posix-fs's, we rely on nlink < 2 1013 # for non-posix-fs's, we rely on nlink < 2
937 $ndirs = (stat _)[3] - 2 1014 $ndirs = (stat _)[3] - 2
938 or return $grp->result ([], $entries); 1015 or return $grp->result ([], $entries);
939 } 1016 }
940 1017
941 my (@dirs, @nondirs); 1018 my (@dirs, @nondirs);
942 1019
943 my $statgrp = add $grp aio_group sub { 1020 my $statgrp = add $grp aio_group sub {
944 $grp->result (\@dirs, \@nondirs); 1021 $grp->result (\@dirs, \@nondirs);
945 }; 1022 };
946 1023
947 limit $statgrp $maxreq; 1024 limit $statgrp $maxreq;
948 feed $statgrp sub { 1025 feed $statgrp sub {
949 return unless @$entries; 1026 return unless @$entries;
950 my $entry = shift @$entries; 1027 my $entry = shift @$entries;
951 1028
952 aioreq_pri $pri; 1029 aioreq_pri $pri;
1030 $wd->[1] = "$entry/.";
953 add $statgrp aio_stat "$path/$entry/.", sub { 1031 add $statgrp aio_stat $wd, sub {
954 if ($_[0] < 0) { 1032 if ($_[0] < 0) {
955 push @nondirs, $entry; 1033 push @nondirs, $entry;
956 } else { 1034 } else {
957 # need to check for real directory 1035 # need to check for real directory
958 aioreq_pri $pri; 1036 aioreq_pri $pri;
1037 $wd->[1] = $entry;
959 add $statgrp aio_lstat "$path/$entry", sub { 1038 add $statgrp aio_lstat $wd, sub {
960 if (-d _) { 1039 if (-d _) {
961 push @dirs, $entry; 1040 push @dirs, $entry;
962 1041
963 unless (--$ndirs) { 1042 unless (--$ndirs) {
964 push @nondirs, @$entries; 1043 push @nondirs, @$entries;
965 feed $statgrp; 1044 feed $statgrp;
1045 }
1046 } else {
1047 push @nondirs, $entry;
966 } 1048 }
967 } else {
968 push @nondirs, $entry;
969 } 1049 }
970 } 1050 }
971 } 1051 };
972 }; 1052 };
973 }; 1053 };
974 }; 1054 };
975 }; 1055 };
976 }; 1056 };
977 1057
978 $grp 1058 $grp
979} 1059}
980 1060
981=item aio_rmtree $path, $callback->($status) 1061=item aio_rmtree $pathname, $callback->($status)
982 1062
983Delete a directory tree starting (and including) C<$path>, return the 1063Delete a directory tree starting (and including) C<$path>, return the
984status of the final C<rmdir> only. This is a composite request that 1064status of the final C<rmdir> only. This is a composite request that
985uses C<aio_scandir> to recurse into and rmdir directories, and unlink 1065uses C<aio_scandir> to recurse into and rmdir directories, and unlink
986everything else. 1066everything else.
… …
1028callback with the fdatasync result code. 1108callback with the fdatasync result code.
1029 1109
1030If this call isn't available because your OS lacks it or it couldn't be 1110If this call isn't available because your OS lacks it or it couldn't be
1031detected, it will be emulated by calling C<fsync> instead. 1111detected, it will be emulated by calling C<fsync> instead.
1032 1112
1113=item aio_syncfs $fh, $callback->($status)
1114
1115Asynchronously call the syncfs syscall to sync the filesystem associated
1116to the given filehandle and call the callback with the syncfs result
1117code. If syncfs is not available, calls sync(), but returns C<-1> and sets
1118errno to C<ENOSYS> nevertheless.
1119
1033=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 1120=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
1034 1121
1035Sync the data portion of the file specified by C<$offset> and C<$length> 1122Sync the data portion of the file specified by C<$offset> and C<$length>
1036to disk (but NOT the metadata), by calling the Linux-specific 1123to disk (but NOT the metadata), by calling the Linux-specific
1037sync_file_range call. If sync_file_range is not available or it returns 1124sync_file_range call. If sync_file_range is not available or it returns
… …
1040C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>, 1127C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
1041C<IO::AIO::SYNC_FILE_RANGE_WRITE> and 1128C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
1042C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range 1129C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
1043manpage for details. 1130manpage for details.
1044 1131
1045=item aio_pathsync $path, $callback->($status) 1132=item aio_pathsync $pathname, $callback->($status)
1046 1133
1047This request tries to open, fsync and close the given path. This is a 1134This request tries to open, fsync and close the given path. This is a
1048composite request intended to sync directories after directory operations 1135composite request intended to sync directories after directory operations
1049(E.g. rename). This might not work on all operating systems or have any 1136(E.g. rename). This might not work on all operating systems or have any
1050specific effect, but usually it makes sure that directory changes get 1137specific effect, but usually it makes sure that directory changes get
… …
1084} 1171}
1085 1172
1086=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 1173=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
1087 1174
1088This is a rather advanced IO::AIO call, which only works on mmap(2)ed 1175This is a rather advanced IO::AIO call, which only works on mmap(2)ed
1089scalars (see the L<Sys::Mmap> or L<Mmap> modules for details on this, note 1176scalars (see the C<IO::AIO::mmap> function, although it also works on data
1177scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
1090that the scalar must only be modified in-place while an aio operation is 1178scalar must only be modified in-place while an aio operation is pending on
1091pending on it). 1179it).
1092 1180
1093It calls the C<msync> function of your OS, if available, with the memory 1181It calls the C<msync> function of your OS, if available, with the memory
1094area starting at C<$offset> in the string and ending C<$length> bytes 1182area starting at C<$offset> in the string and ending C<$length> bytes
1095later. If C<$length> is negative, counts from the end, and if C<$length> 1183later. If C<$length> is negative, counts from the end, and if C<$length>
1096is C<undef>, then it goes till the end of the string. The flags can be 1184is C<undef>, then it goes till the end of the string. The flags can be
… …
1107as for C<aio_msync>, above, except for flags, which must be either 1195as for C<aio_msync>, above, except for flags, which must be either
1108C<0> (which reads all pages and ensures they are instantiated) or 1196C<0> (which reads all pages and ensures they are instantiated) or
1109C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and 1197C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and
1110writing an octet from it, which dirties the page). 1198writing an octet from it, which dirties the page).
1111 1199
1200=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1201
1202This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1203scalars.
1204
1205It reads in all the pages of the underlying storage into memory (if any)
1206and locks them, so they are not getting swapped/paged out or removed.
1207
1208If C<$length> is undefined, then the scalar will be locked till the end.
1209
1210On systems that do not implement C<mlock>, this function returns C<-1>
1211and sets errno to C<ENOSYS>.
1212
1213Note that the corresponding C<munlock> is synchronous and is
1214documented under L<MISCELLANEOUS FUNCTIONS>.
1215
1216Example: open a file, mmap and mlock it - both will be undone when
1217C<$data> gets destroyed.
1218
1219 open my $fh, "<", $path or die "$path: $!";
1220 my $data;
1221 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1222 aio_mlock $data; # mlock in background
1223
1224=item aio_mlockall $flags, $callback->($status)
1225
1226Calls the C<mlockall> function with the given C<$flags> (a combination of
1227C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL_FUTURE>).
1228
1229On systems that do not implement C<mlockall>, this function returns C<-1>
1230and sets errno to C<ENOSYS>.
1231
1232Note that the corresponding C<munlockall> is synchronous and is
1233documented under L<MISCELLANEOUS FUNCTIONS>.
1234
1235Example: asynchronously lock all current and future pages into memory.
1236
1237 aio_mlockall IO::AIO::MCL_FUTURE;
1238
1112=item aio_group $callback->(...) 1239=item aio_group $callback->(...)
1113 1240
1114This is a very special aio request: Instead of doing something, it is a 1241This is a very special aio request: Instead of doing something, it is a
1115container for other aio requests, which is useful if you want to bundle 1242container for other aio requests, which is useful if you want to bundle
1116many requests into a single, composite, request with a definite callback 1243many requests into a single, composite, request with a definite callback
… …
1152like sleep and file handle readable/writable, the overhead this creates is 1279like sleep and file handle readable/writable, the overhead this creates is
1153immense (it blocks a thread for a long time) so do not use this function 1280immense (it blocks a thread for a long time) so do not use this function
1154except to put your application under artificial I/O pressure. 1281except to put your application under artificial I/O pressure.
1155 1282
1156=back 1283=back
1284
1285
1286=head2 IO::AIO::WD - multiple working directories
1287
1288Your process only has one current working directory, which is used by all
1289threads. This makes it hard to use relative paths (some other component
1290could call C<chdir> at any time, and it is hard to control when the path
1291will be used by IO::AIO).
1292
1293One solution for this is to always use absolute paths. This usually works,
1294but can be quite slow (the kernel has to walk the whole path on every
1295access), and can also be a hassle to implement.
1296
1297Newer POSIX systems have a number of functions (openat, fdopendir,
1298futimensat and so on) that make it possible to specify working directories
1299per operation.
1300
1301For portability, and because the clowns who "designed", or shall I write,
1302perpetrated this new interface were obviously half-drunk, this abstraction
1303cannot be perfect, though.
1304
1305IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
1306object. This object stores the canonicalised, absolute version of the
1307path, and on systems that allow it, also a directory file descriptor.
1308
1309Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
1310or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
1311object and a pathname instead (or the IO::AIO::WD object alone, which
1312gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
1313IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
1314to that IO::AIO::WD object.
1315
1316For example, to get a wd object for F</etc> and then stat F<passwd>
1317inside, you would write:
1318
1319 aio_wd "/etc", sub {
1320 my $etcdir = shift;
1321
1322 # although $etcdir can be undef on error, there is generally no reason
1323 # to check for errors here, as aio_stat will fail with ENOENT
1324 # when $etcdir is undef.
1325
1326 aio_stat [$etcdir, "passwd"], sub {
1327 # yay
1328 };
1329 };
1330
1331That C<aio_wd> is a request and not a normal function shows that creating
1332an IO::AIO::WD object is itself a potentially blocking operation, which is
1333why it is done asynchronously.
1334
1335To stat the directory obtained with C<aio_wd> above, one could write
1336either of the following three request calls:
1337
1338 aio_lstat "/etc" , sub { ... # pathname as normal string
1339 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1340 aio_lstat $wd , sub { ... # shorthand for the previous
1341
1342As with normal pathnames, IO::AIO keeps a copy of the working directory
1343object and the pathname string, so you could write the following without
1344causing any issues due to C<$path> getting reused:
1345
1346 my $path = [$wd, undef];
1347
1348 for my $name (qw(abc def ghi)) {
1349 $path->[1] = $name;
1350 aio_stat $path, sub {
1351 # ...
1352 };
1353 }
1354
1355There are some caveats: when directories get renamed (or deleted), the
1356pathname string doesn't change, so will point to the new directory (or
1357nowhere at all), while the directory fd, if available on the system,
1358will still point to the original directory. Most functions accepting a
1359pathname will use the directory fd on newer systems, and the string on
1360older systems. Some functions (such as realpath) will always rely on the
1361string form of the pathname.
1362
1363So this fucntionality is mainly useful to get some protection against
1364C<chdir>, to easily get an absolute path out of a relative path for future
1365reference, and to speed up doing many operations in the same directory
1366(e.g. when stat'ing all files in a directory).
1367
1368The following functions implement this working directory abstraction:
1369
1370=over 4
1371
1372=item aio_wd $pathname, $callback->($wd)
1373
1374Asynchonously canonicalise the given pathname and convert it to an
1375IO::AIO::WD object representing it. If possible and supported on the
1376system, also open a directory fd to speed up pathname resolution relative
1377to this working directory.
1378
1379If something goes wrong, then C<undef> is passwd to the callback instead
1380of a working directory object and C<$!> is set appropriately. Since
1381passing C<undef> as working directory component of a pathname fails the
1382request with C<ENOENT>, there is often no need for error checking in the
1383C<aio_wd> callback, as future requests using the value will fail in the
1384expected way.
1385
1386If this call isn't available because your OS lacks it or it couldn't be
1387detected, it will be emulated by calling C<fsync> instead.
1388
1389=item IO::AIO::CWD
1390
1391This is a compiletime constant (object) that represents the process
1392current working directory.
1393
1394Specifying this object as working directory object for a pathname is as
1395if the pathname would be specified directly, without a directory object,
1396e.g., these calls are functionally identical:
1397
1398 aio_stat "somefile", sub { ... };
1399 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1400
1401=back
1402
1157 1403
1158=head2 IO::AIO::REQ CLASS 1404=head2 IO::AIO::REQ CLASS
1159 1405
1160All non-aggregate C<aio_*> functions return an object of this class when 1406All non-aggregate C<aio_*> functions return an object of this class when
1161called in non-void context. 1407called in non-void context.
… …
1279 1525
1280Sets a feeder/generator on this group: every group can have an attached 1526Sets a feeder/generator on this group: every group can have an attached
1281generator that generates requests if idle. The idea behind this is that, 1527generator that generates requests if idle. The idea behind this is that,
1282although you could just queue as many requests as you want in a group, 1528although you could just queue as many requests as you want in a group,
1283this might starve other requests for a potentially long time. For example, 1529this might starve other requests for a potentially long time. For example,
1284C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests, 1530C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
1285delaying any later requests for a long time. 1531requests, delaying any later requests for a long time.
1286 1532
1287To avoid this, and allow incremental generation of requests, you can 1533To avoid this, and allow incremental generation of requests, you can
1288instead a group and set a feeder on it that generates those requests. The 1534instead a group and set a feeder on it that generates those requests. The
1289feed callback will be called whenever there are few enough (see C<limit>, 1535feed callback will be called whenever there are few enough (see C<limit>,
1290below) requests active in the group itself and is expected to queue more 1536below) requests active in the group itself and is expected to queue more
… …
1339 1585
1340See C<poll_cb> for an example. 1586See C<poll_cb> for an example.
1341 1587
1342=item IO::AIO::poll_cb 1588=item IO::AIO::poll_cb
1343 1589
1344Process some outstanding events on the result pipe. You have to call this 1590Process some outstanding events on the result pipe. You have to call
1345regularly. Returns C<0> if all events could be processed, or C<-1> if it 1591this regularly. Returns C<0> if all events could be processed (or there
1346returned earlier for whatever reason. Returns immediately when no events 1592were no events to process), or C<-1> if it returned earlier for whatever
1347are outstanding. The amount of events processed depends on the settings of 1593reason. Returns immediately when no events are outstanding. The amount of
1348C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>. 1594events processed depends on the settings of C<IO::AIO::max_poll_req> and
1595C<IO::AIO::max_poll_time>.
1349 1596
1350If not all requests were processed for whatever reason, the filehandle 1597If not all requests were processed for whatever reason, the filehandle
1351will still be ready when C<poll_cb> returns, so normally you don't have to 1598will still be ready when C<poll_cb> returns, so normally you don't have to
1352do anything special to have it called later. 1599do anything special to have it called later.
1600
1601Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
1602ready, it can be beneficial to call this function from loops which submit
1603a lot of requests, to make sure the results get processed when they become
1604available and not just when the loop is finished and the event loop takes
1605over again. This function returns very fast when there are no outstanding
1606requests.
1353 1607
1354Example: Install an Event watcher that automatically calls 1608Example: Install an Event watcher that automatically calls
1355IO::AIO::poll_cb with high priority (more examples can be found in the 1609IO::AIO::poll_cb with high priority (more examples can be found in the
1356SYNOPSIS section, at the top of this document): 1610SYNOPSIS section, at the top of this document):
1357 1611
… …
1459 1713
1460Under normal circumstances you don't need to call this function. 1714Under normal circumstances you don't need to call this function.
1461 1715
1462=item IO::AIO::max_idle $nthreads 1716=item IO::AIO::max_idle $nthreads
1463 1717
1464Limit the number of threads (default: 4) that are allowed to idle (i.e., 1718Limit the number of threads (default: 4) that are allowed to idle
1465threads that did not get a request to process within 10 seconds). That 1719(i.e., threads that did not get a request to process within the idle
1466means if a thread becomes idle while C<$nthreads> other threads are also 1720timeout (default: 10 seconds). That means if a thread becomes idle while
1467idle, it will free its resources and exit. 1721C<$nthreads> other threads are also idle, it will free its resources and
1722exit.
1468 1723
1469This is useful when you allow a large number of threads (e.g. 100 or 1000) 1724This is useful when you allow a large number of threads (e.g. 100 or 1000)
1470to allow for extremely high load situations, but want to free resources 1725to allow for extremely high load situations, but want to free resources
1471under normal circumstances (1000 threads can easily consume 30MB of RAM). 1726under normal circumstances (1000 threads can easily consume 30MB of RAM).
1472 1727
1473The default is probably ok in most situations, especially if thread 1728The default is probably ok in most situations, especially if thread
1474creation is fast. If thread creation is very slow on your system you might 1729creation is fast. If thread creation is very slow on your system you might
1475want to use larger values. 1730want to use larger values.
1476 1731
1732=item IO::AIO::idle_timeout $seconds
1733
1734Sets the minimum idle timeout (default 10) after which worker threads are
1735allowed to exit. SEe C<IO::AIO::max_idle>.
1736
1477=item IO::AIO::max_outstanding $maxreqs 1737=item IO::AIO::max_outstanding $maxreqs
1738
1739Sets the maximum number of outstanding requests to C<$nreqs>. If
1740you do queue up more than this number of requests, the next call to
1741C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
1742C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
1743longer exceeded.
1744
1745In other words, this setting does not enforce a queue limit, but can be
1746used to make poll functions block if the limit is exceeded.
1478 1747
1479This is a very bad function to use in interactive programs because it 1748This is a very bad function to use in interactive programs because it
1480blocks, and a bad way to reduce concurrency because it is inexact: Better 1749blocks, and a bad way to reduce concurrency because it is inexact: Better
1481use an C<aio_group> together with a feed callback. 1750use an C<aio_group> together with a feed callback.
1482 1751
1483Sets the maximum number of outstanding requests to C<$nreqs>. If you 1752It's main use is in scripts without an event loop - when you want to stat
1484do queue up more than this number of requests, the next call to the 1753a lot of files, you can write somehting like this:
1485C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
1486function will block until the limit is no longer exceeded.
1487 1754
1488The default value is very large, so there is no practical limit on the 1755 IO::AIO::max_outstanding 32;
1489number of outstanding requests.
1490 1756
1491You can still queue as many requests as you want. Therefore, 1757 for my $path (...) {
1492C<max_outstanding> is mainly useful in simple scripts (with low values) or 1758 aio_stat $path , ...;
1493as a stop gap to shield against fatal memory overflow (with large values). 1759 IO::AIO::poll_cb;
1760 }
1761
1762 IO::AIO::flush;
1763
1764The call to C<poll_cb> inside the loop will normally return instantly, but
1765as soon as more thna C<32> reqeusts are in-flight, it will block until
1766some requests have been handled. This keeps the loop from pushing a large
1767number of C<aio_stat> requests onto the queue.
1768
1769The default value for C<max_outstanding> is very large, so there is no
1770practical limit on the number of outstanding requests.
1494 1771
1495=back 1772=back
1496 1773
1497=head3 STATISTICAL INFORMATION 1774=head3 STATISTICAL INFORMATION
1498 1775
… …
1536 1813
1537Returns the number of bytes copied, or C<-1> on error. 1814Returns the number of bytes copied, or C<-1> on error.
1538 1815
1539=item IO::AIO::fadvise $fh, $offset, $len, $advice 1816=item IO::AIO::fadvise $fh, $offset, $len, $advice
1540 1817
1541Simply calls the C<posix_fadvise> function (see it's 1818Simply calls the C<posix_fadvise> function (see its
1542manpage for details). The following advice constants are 1819manpage for details). The following advice constants are
1543avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>, 1820available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
1544C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>, 1821C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
1545C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>. 1822C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
1546 1823
1547On systems that do not implement C<posix_fadvise>, this function returns 1824On systems that do not implement C<posix_fadvise>, this function returns
1548ENOSYS, otherwise the return value of C<posix_fadvise>. 1825ENOSYS, otherwise the return value of C<posix_fadvise>.
1549 1826
1550=item IO::AIO::mlockall $flags 1827=item IO::AIO::madvise $scalar, $offset, $len, $advice
1551 1828
1552Calls the C<mlockall> function with the given C<$flags> (a combination of 1829Simply calls the C<posix_madvise> function (see its
1553C<IO::AIO::MCL_CURRENT> and C<IO::AIO::MCL__FUTURE>). 1830manpage for details). The following advice constants are
1831available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
1832C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>.
1554 1833
1834On systems that do not implement C<posix_madvise>, this function returns
1835ENOSYS, otherwise the return value of C<posix_madvise>.
1836
1837=item IO::AIO::mprotect $scalar, $offset, $len, $protect
1838
1839Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
1840$scalar (see its manpage for details). The following protect
1841constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
1842C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
1843
1555On systems that do not implement C<mlockall>, this function returns 1844On systems that do not implement C<mprotect>, this function returns
1556ENOSYS, otherwise the return value of C<mlockall>. 1845ENOSYS, otherwise the return value of C<mprotect>.
1846
1847=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1848
1849Memory-maps a file (or anonymous memory range) and attaches it to the
1850given C<$scalar>, which will act like a string scalar.
1851
1852The only operations allowed on the scalar are C<substr>/C<vec> that don't
1853change the string length, and most read-only operations such as copying it
1854or searching it with regexes and so on.
1855
1856Anything else is unsafe and will, at best, result in memory leaks.
1857
1858The memory map associated with the C<$scalar> is automatically removed
1859when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or
1860C<IO::AIO::munmap> functions are called.
1861
1862This calls the C<mmap>(2) function internally. See your system's manual
1863page for details on the C<$length>, C<$prot> and C<$flags> parameters.
1864
1865The C<$length> must be larger than zero and smaller than the actual
1866filesize.
1867
1868C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
1869C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
1870
1871C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or
1872C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when
1873not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS>
1874(which is set to C<MAP_ANON> if your system only provides this
1875constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>,
1876C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or
1877C<IO::AIO::MAP_NONBLOCK>
1878
1879If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
1880
1881C<$offset> is the offset from the start of the file - it generally must be
1882a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
1883
1884Example:
1885
1886 use Digest::MD5;
1887 use IO::AIO;
1888
1889 open my $fh, "<verybigfile"
1890 or die "$!";
1891
1892 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1893 or die "verybigfile: $!";
1894
1895 my $fast_md5 = md5 $data;
1896
1897=item IO::AIO::munmap $scalar
1898
1899Removes a previous mmap and undefines the C<$scalar>.
1900
1901=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
1902
1903Calls the C<munlock> function, undoing the effects of a previous
1904C<aio_mlock> call (see its description for details).
1557 1905
1558=item IO::AIO::munlockall 1906=item IO::AIO::munlockall
1559 1907
1560Calls the C<munlockall> function. 1908Calls the C<munlockall> function.
1561 1909
… …
1603 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => 1951 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
1604 \&IO::AIO::poll_cb); 1952 \&IO::AIO::poll_cb);
1605 1953
1606=head2 FORK BEHAVIOUR 1954=head2 FORK BEHAVIOUR
1607 1955
1608This module should do "the right thing" when the process using it forks: 1956Usage of pthreads in a program changes the semantics of fork
1957considerably. Specifically, only async-safe functions can be called after
1958fork. Perl doesn't know about this, so in general, you cannot call fork
1959with defined behaviour in perl if pthreads are involved. IO::AIO uses
1960pthreads, so this applies, but many other extensions and (for inexplicable
1961reasons) perl itself often is linked against pthreads, so this limitation
1962applies to quite a lot of perls.
1609 1963
1610Before the fork, IO::AIO enters a quiescent state where no requests 1964This module no longer tries to fight your OS, or POSIX. That means IO::AIO
1611can be added in other threads and no results will be processed. After 1965only works in the process that loaded it. Forking is fully supported, but
1612the fork the parent simply leaves the quiescent state and continues 1966using IO::AIO in the child is not.
1613request/result processing, while the child frees the request/result queue
1614(so that the requests started before the fork will only be handled in the
1615parent). Threads will be started on demand until the limit set in the
1616parent process has been reached again.
1617 1967
1618In short: the parent will, after a short pause, continue as if fork had 1968You might get around by not I<using> IO::AIO before (or after)
1619not been called, while the child will act as if IO::AIO has not been used 1969forking. You could also try to call the L<IO::AIO::reinit> function in the
1620yet. 1970child:
1971
1972=over 4
1973
1974=item IO::AIO::reinit
1975
1976Abandons all current requests and I/O threads and simply reinitialises all
1977data structures. This is not an operation supported by any standards, but
1978happens to work on GNU/Linux and some newer BSD systems.
1979
1980The only reasonable use for this function is to call it after forking, if
1981C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
1982the process will result in undefined behaviour. Calling it at any time
1983will also result in any undefined (by POSIX) behaviour.
1984
1985=back
1621 1986
1622=head2 MEMORY USAGE 1987=head2 MEMORY USAGE
1623 1988
1624Per-request usage: 1989Per-request usage:
1625 1990

Diff Legend

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