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 1.187 by root, Fri Feb 11 00:05:17 2011 UTC vs.
Revision 1.279 by root, Sat Jan 6 01:04:42 2018 UTC

1=head1 NAME 1=head1 NAME
2 2
3IO::AIO - Asynchronous Input/Output 3IO::AIO - Asynchronous/Advanced Input/Output
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use IO::AIO; 7 use IO::AIO;
8 8
58not well-supported or restricted (GNU/Linux doesn't allow them on normal 58not well-supported or restricted (GNU/Linux doesn't allow them on normal
59files currently, for example), and they would only support aio_read and 59files currently, for example), and they would only support aio_read and
60aio_write, so the remaining functionality would have to be implemented 60aio_write, so the remaining functionality would have to be implemented
61using threads anyway. 61using threads anyway.
62 62
63In addition to asynchronous I/O, this module also exports some rather
64arcane interfaces, such as C<madvise> or linux's C<splice> system call,
65which is why the C<A> in C<AIO> can also mean I<advanced>.
66
63Although the module will work in the presence of other (Perl-) threads, 67Although the module will work in the presence of other (Perl-) threads,
64it is currently not reentrant in any way, so use appropriate locking 68it is currently not reentrant in any way, so use appropriate locking
65yourself, always call C<poll_cb> from within the same thread, or never 69yourself, always call C<poll_cb> from within the same thread, or never
66call C<poll_cb> (or other C<aio_> functions) recursively. 70call C<poll_cb> (or other C<aio_> functions) recursively.
67 71
68=head2 EXAMPLE 72=head2 EXAMPLE
69 73
70This is a simple example that uses the EV module and loads 74This is a simple example that uses the EV module and loads
71F</etc/passwd> asynchronously: 75F</etc/passwd> asynchronously:
72 76
73 use Fcntl;
74 use EV; 77 use EV;
75 use IO::AIO; 78 use IO::AIO;
76 79
77 # register the IO::AIO callback with EV 80 # register the IO::AIO callback with EV
78 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb; 81 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
95 98
96 # file contents now in $contents 99 # file contents now in $contents
97 print $contents; 100 print $contents;
98 101
99 # exit event loop and program 102 # exit event loop and program
100 EV::unloop; 103 EV::break;
101 }; 104 };
102 }; 105 };
103 106
104 # possibly queue up other requests, or open GUI windows, 107 # possibly queue up other requests, or open GUI windows,
105 # check for sockets etc. etc. 108 # check for sockets etc. etc.
106 109
107 # process events as long as there are some: 110 # process events as long as there are some:
108 EV::loop; 111 EV::run;
109 112
110=head1 REQUEST ANATOMY AND LIFETIME 113=head1 REQUEST ANATOMY AND LIFETIME
111 114
112Every C<aio_*> function creates a request. which is a C data structure not 115Every C<aio_*> function creates a request. which is a C data structure not
113directly visible to Perl. 116directly visible to Perl.
168use common::sense; 171use common::sense;
169 172
170use base 'Exporter'; 173use base 'Exporter';
171 174
172BEGIN { 175BEGIN {
173 our $VERSION = '3.72'; 176 our $VERSION = 4.4;
174 177
175 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close 178 our @AIO_REQ = qw(aio_sendfile aio_seek aio_read aio_write aio_open aio_close
176 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx 179 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
177 aio_scandir aio_symlink aio_readlink aio_sync aio_fsync 180 aio_scandir aio_symlink aio_readlink aio_realpath aio_fcntl aio_ioctl
178 aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead 181 aio_sync aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range
182 aio_pathsync aio_readahead aio_fiemap aio_allocate
179 aio_rename aio_link aio_move aio_copy aio_group 183 aio_rename aio_rename2 aio_link aio_move aio_copy aio_group
180 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown 184 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
181 aio_chmod aio_utime aio_truncate 185 aio_chmod aio_utime aio_truncate
182 aio_msync aio_mtouch aio_mlock aio_mlockall 186 aio_msync aio_mtouch aio_mlock aio_mlockall
183 aio_statvfs); 187 aio_statvfs
188 aio_slurp
189 aio_wd);
184 190
185 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); 191 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
186 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 192 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
187 min_parallel max_parallel max_idle 193 min_parallel max_parallel max_idle idle_timeout
188 nreqs nready npending nthreads 194 nreqs nready npending nthreads
189 max_poll_time max_poll_reqs 195 max_poll_time max_poll_reqs
190 sendfile fadvise madvise 196 sendfile fadvise madvise
191 mmap munmap munlock munlockall); 197 mmap munmap munlock munlockall);
192 198
200 206
201=head1 FUNCTIONS 207=head1 FUNCTIONS
202 208
203=head2 QUICK OVERVIEW 209=head2 QUICK OVERVIEW
204 210
205This section simply lists the prototypes of the most important functions 211This section simply lists the prototypes most of the functions for
206for quick reference. See the following sections for function-by-function 212quick reference. See the following sections for function-by-function
207documentation. 213documentation.
208 214
215 aio_wd $pathname, $callback->($wd)
209 aio_open $pathname, $flags, $mode, $callback->($fh) 216 aio_open $pathname, $flags, $mode, $callback->($fh)
210 aio_close $fh, $callback->($status) 217 aio_close $fh, $callback->($status)
218 aio_seek $fh,$offset,$whence, $callback->($offs)
211 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 219 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
212 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 220 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
213 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 221 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
214 aio_readahead $fh,$offset,$length, $callback->($retval) 222 aio_readahead $fh,$offset,$length, $callback->($retval)
215 aio_stat $fh_or_path, $callback->($status) 223 aio_stat $fh_or_path, $callback->($status)
216 aio_lstat $fh, $callback->($status) 224 aio_lstat $fh, $callback->($status)
217 aio_statvfs $fh_or_path, $callback->($statvfs) 225 aio_statvfs $fh_or_path, $callback->($statvfs)
218 aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 226 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
219 aio_chown $fh_or_path, $uid, $gid, $callback->($status) 227 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
228 aio_chmod $fh_or_path, $mode, $callback->($status)
220 aio_truncate $fh_or_path, $offset, $callback->($status) 229 aio_truncate $fh_or_path, $offset, $callback->($status)
221 aio_chmod $fh_or_path, $mode, $callback->($status) 230 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
231 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
222 aio_unlink $pathname, $callback->($status) 232 aio_unlink $pathname, $callback->($status)
223 aio_mknod $path, $mode, $dev, $callback->($status) 233 aio_mknod $pathname, $mode, $dev, $callback->($status)
224 aio_link $srcpath, $dstpath, $callback->($status) 234 aio_link $srcpath, $dstpath, $callback->($status)
225 aio_symlink $srcpath, $dstpath, $callback->($status) 235 aio_symlink $srcpath, $dstpath, $callback->($status)
226 aio_readlink $path, $callback->($link) 236 aio_readlink $pathname, $callback->($link)
237 aio_realpath $pathname, $callback->($path)
227 aio_rename $srcpath, $dstpath, $callback->($status) 238 aio_rename $srcpath, $dstpath, $callback->($status)
239 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
228 aio_mkdir $pathname, $mode, $callback->($status) 240 aio_mkdir $pathname, $mode, $callback->($status)
229 aio_rmdir $pathname, $callback->($status) 241 aio_rmdir $pathname, $callback->($status)
230 aio_readdir $pathname, $callback->($entries) 242 aio_readdir $pathname, $callback->($entries)
231 aio_readdirx $pathname, $flags, $callback->($entries, $flags) 243 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
232 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST 244 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
233 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN 245 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
246 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
234 aio_load $path, $data, $callback->($status) 247 aio_load $pathname, $data, $callback->($status)
235 aio_copy $srcpath, $dstpath, $callback->($status) 248 aio_copy $srcpath, $dstpath, $callback->($status)
236 aio_move $srcpath, $dstpath, $callback->($status) 249 aio_move $srcpath, $dstpath, $callback->($status)
237 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
238 aio_rmtree $path, $callback->($status) 250 aio_rmtree $pathname, $callback->($status)
251 aio_fcntl $fh, $cmd, $arg, $callback->($status)
252 aio_ioctl $fh, $request, $buf, $callback->($status)
239 aio_sync $callback->($status) 253 aio_sync $callback->($status)
254 aio_syncfs $fh, $callback->($status)
240 aio_fsync $fh, $callback->($status) 255 aio_fsync $fh, $callback->($status)
241 aio_fdatasync $fh, $callback->($status) 256 aio_fdatasync $fh, $callback->($status)
242 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 257 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
243 aio_pathsync $path, $callback->($status) 258 aio_pathsync $pathname, $callback->($status)
244 aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 259 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
245 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 260 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
246 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status) 261 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
247 aio_mlockall $flags, $callback->($status) 262 aio_mlockall $flags, $callback->($status)
248 aio_group $callback->(...) 263 aio_group $callback->(...)
249 aio_nop $callback->() 264 aio_nop $callback->()
258 IO::AIO::max_poll_reqs $nreqs 273 IO::AIO::max_poll_reqs $nreqs
259 IO::AIO::max_poll_time $seconds 274 IO::AIO::max_poll_time $seconds
260 IO::AIO::min_parallel $nthreads 275 IO::AIO::min_parallel $nthreads
261 IO::AIO::max_parallel $nthreads 276 IO::AIO::max_parallel $nthreads
262 IO::AIO::max_idle $nthreads 277 IO::AIO::max_idle $nthreads
278 IO::AIO::idle_timeout $seconds
263 IO::AIO::max_outstanding $maxreqs 279 IO::AIO::max_outstanding $maxreqs
264 IO::AIO::nreqs 280 IO::AIO::nreqs
265 IO::AIO::nready 281 IO::AIO::nready
266 IO::AIO::npending 282 IO::AIO::npending
283 $nfd = IO::AIO::get_fdlimit [EXPERIMENTAL]
284 IO::AIO::min_fdlimit $nfd [EXPERIMENTAL]
267 285
268 IO::AIO::sendfile $ofh, $ifh, $offset, $count 286 IO::AIO::sendfile $ofh, $ifh, $offset, $count
269 IO::AIO::fadvise $fh, $offset, $len, $advice 287 IO::AIO::fadvise $fh, $offset, $len, $advice
288 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
289 IO::AIO::munmap $scalar
270 IO::AIO::madvise $scalar, $offset, $length, $advice 290 IO::AIO::madvise $scalar, $offset, $length, $advice
271 IO::AIO::mprotect $scalar, $offset, $length, $protect 291 IO::AIO::mprotect $scalar, $offset, $length, $protect
272 IO::AIO::munlock $scalar, $offset = 0, $length = undef 292 IO::AIO::munlock $scalar, $offset = 0, $length = undef
273 IO::AIO::munlockall 293 IO::AIO::munlockall
274 294
275=head2 AIO REQUEST FUNCTIONS 295=head2 API NOTES
276 296
277All the C<aio_*> calls are more or less thin wrappers around the syscall 297All the C<aio_*> calls are more or less thin wrappers around the syscall
278with the same name (sans C<aio_>). The arguments are similar or identical, 298with the same name (sans C<aio_>). The arguments are similar or identical,
279and they all accept an additional (and optional) C<$callback> argument 299and they all accept an additional (and optional) C<$callback> argument
280which must be a code reference. This code reference will get called with 300which must be a code reference. This code reference will be called after
281the syscall return code (e.g. most syscalls return C<-1> on error, unlike 301the syscall has been executed in an asynchronous fashion. The results
282perl, which usually delivers "false") as its sole argument after the given 302of the request will be passed as arguments to the callback (and, if an
283syscall has been executed asynchronously. 303error occured, in C<$!>) - for most requests the syscall return code (e.g.
304most syscalls return C<-1> on error, unlike perl, which usually delivers
305"false").
306
307Some requests (such as C<aio_readdir>) pass the actual results and
308communicate failures by passing C<undef>.
284 309
285All functions expecting a filehandle keep a copy of the filehandle 310All functions expecting a filehandle keep a copy of the filehandle
286internally until the request has finished. 311internally until the request has finished.
287 312
288All functions return request objects of type L<IO::AIO::REQ> that allow 313All functions return request objects of type L<IO::AIO::REQ> that allow
289further manipulation of those requests while they are in-flight. 314further manipulation of those requests while they are in-flight.
290 315
291The pathnames you pass to these routines I<must> be absolute and 316The pathnames you pass to these routines I<should> be absolute. The
292encoded as octets. The reason for the former is that at the time the 317reason for this is that at the time the request is being executed, the
293request is being executed, the current working directory could have 318current working directory could have changed. Alternatively, you can
294changed. Alternatively, you can make sure that you never change the 319make sure that you never change the current working directory anywhere
295current working directory anywhere in the program and then use relative 320in the program and then use relative paths. You can also take advantage
296paths. 321of IO::AIOs working directory abstraction, that lets you specify paths
322relative to some previously-opened "working directory object" - see the
323description of the C<IO::AIO::WD> class later in this document.
297 324
298To encode pathnames as octets, either make sure you either: a) always pass 325To encode pathnames as octets, either make sure you either: a) always pass
299in filenames you got from outside (command line, readdir etc.) without 326in filenames you got from outside (command line, readdir etc.) without
300tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode 327tinkering, b) are in your native filesystem encoding, c) use the Encode
301your pathnames to the locale (or other) encoding in effect in the user 328module and encode your pathnames to the locale (or other) encoding in
302environment, d) use Glib::filename_from_unicode on unicode filenames or e) 329effect in the user environment, d) use Glib::filename_from_unicode on
303use something else to ensure your scalar has the correct contents. 330unicode filenames or e) use something else to ensure your scalar has the
331correct contents.
304 332
305This works, btw. independent of the internal UTF-8 bit, which IO::AIO 333This works, btw. independent of the internal UTF-8 bit, which IO::AIO
306handles correctly whether it is set or not. 334handles correctly whether it is set or not.
335
336=head2 AIO REQUEST FUNCTIONS
307 337
308=over 4 338=over 4
309 339
310=item $prev_pri = aioreq_pri [$pri] 340=item $prev_pri = aioreq_pri [$pri]
311 341
341 371
342 372
343=item aio_open $pathname, $flags, $mode, $callback->($fh) 373=item aio_open $pathname, $flags, $mode, $callback->($fh)
344 374
345Asynchronously open or create a file and call the callback with a newly 375Asynchronously open or create a file and call the callback with a newly
346created filehandle for the file. 376created filehandle for the file (or C<undef> in case of an error).
347 377
348The pathname passed to C<aio_open> must be absolute. See API NOTES, above, 378The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
349for an explanation. 379for an explanation.
350 380
351The C<$flags> argument is a bitmask. See the C<Fcntl> module for a 381The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
367 } else { 397 } else {
368 die "open failed: $!\n"; 398 die "open failed: $!\n";
369 } 399 }
370 }; 400 };
371 401
402In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
403C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
404following POSIX and non-POSIX constants are available (missing ones on
405your system are, as usual, C<0>):
406
407C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
408C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
409C<O_RSYNC>, C<O_SYNC>, C<O_PATH>, C<O_TMPFILE>, and C<O_TTY_INIT>.
410
372 411
373=item aio_close $fh, $callback->($status) 412=item aio_close $fh, $callback->($status)
374 413
375Asynchronously close a file and call the callback with the result 414Asynchronously close a file and call the callback with the result
376code. 415code.
385Or in other words: the file descriptor will be closed, but it will not be 424Or in other words: the file descriptor will be closed, but it will not be
386free for reuse until the perl filehandle is closed. 425free for reuse until the perl filehandle is closed.
387 426
388=cut 427=cut
389 428
429=item aio_seek $fh, $offset, $whence, $callback->($offs)
430
431Seeks the filehandle to the new C<$offset>, similarly to perl's
432C<sysseek>. The C<$whence> can use the traditional values (C<0> for
433C<IO::AIO::SEEK_SET>, C<1> for C<IO::AIO::SEEK_CUR> or C<2> for
434C<IO::AIO::SEEK_END>).
435
436The resulting absolute offset will be passed to the callback, or C<-1> in
437case of an error.
438
439In theory, the C<$whence> constants could be different than the
440corresponding values from L<Fcntl>, but perl guarantees they are the same,
441so don't panic.
442
443As a GNU/Linux (and maybe Solaris) extension, also the constants
444C<IO::AIO::SEEK_DATA> and C<IO::AIO::SEEK_HOLE> are available, if they
445could be found. No guarantees about suitability for use in C<aio_seek> or
446Perl's C<sysseek> can be made though, although I would naively assume they
447"just work".
448
390=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 449=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
391 450
392=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 451=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
393 452
394Reads or writes C<$length> bytes from or to the specified C<$fh> and 453Reads or writes C<$length> bytes from or to the specified C<$fh> and
395C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> 454C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> and
396and calls the callback without the actual number of bytes read (or -1 on 455calls the callback with the actual number of bytes transferred (or -1 on
397error, just like the syscall). 456error, just like the syscall).
398 457
399C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to 458C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
400offset plus the actual number of bytes read. 459offset plus the actual number of bytes read.
401 460
426 485
427Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 486Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
428reading at byte offset C<$in_offset>, and starts writing at the current 487reading at byte offset C<$in_offset>, and starts writing at the current
429file offset of C<$out_fh>. Because of that, it is not safe to issue more 488file offset of C<$out_fh>. Because of that, it is not safe to issue more
430than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each 489than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
431other. 490other. The same C<$in_fh> works fine though, as this function does not
491move or use the file offset of C<$in_fh>.
432 492
433Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than 493Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
434are written, and there is no way to find out how many bytes have been read 494are written, and there is no way to find out how many more bytes have been
435from C<aio_sendfile> alone, as C<aio_sendfile> only provides the number of 495read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
436bytes written to C<$out_fh>. Only if the result value equals C<$length> 496number of bytes written to C<$out_fh>. Only if the result value equals
437one can assume that C<$length> bytes have been read. 497C<$length> one can assume that C<$length> bytes have been read.
438 498
439Unlike with other C<aio_> functions, it makes a lot of sense to use 499Unlike with other C<aio_> functions, it makes a lot of sense to use
440C<aio_sendfile> on non-blocking sockets, as long as one end (typically 500C<aio_sendfile> on non-blocking sockets, as long as one end (typically
441the C<$in_fh>) is a file - the file I/O will then be asynchronous, while 501the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
442the socket I/O will be non-blocking. Note, however, that you can run into 502the socket I/O will be non-blocking. Note, however, that you can run
443a trap where C<aio_sendfile> reads some data with readahead, then fails 503into a trap where C<aio_sendfile> reads some data with readahead, then
444to write all data, and when the socket is ready the next time, the data 504fails to write all data, and when the socket is ready the next time, the
445in the cache is already lost, forcing C<aio_sendfile> to again hit the 505data in the cache is already lost, forcing C<aio_sendfile> to again hit
446disk. Explicit C<aio_read> + C<aio_write> let's you control resource usage 506the disk. Explicit C<aio_read> + C<aio_write> let's you better control
447much better. 507resource usage.
448 508
449This call tries to make use of a native C<sendfile> syscall to provide 509This call tries to make use of a native C<sendfile>-like syscall to
450zero-copy operation. For this to work, C<$out_fh> should refer to a 510provide zero-copy operation. For this to work, C<$out_fh> should refer to
451socket, and C<$in_fh> should refer to an mmap'able file. 511a socket, and C<$in_fh> should refer to an mmap'able file.
452 512
453If a native sendfile cannot be found or it fails with C<ENOSYS>, 513If a native sendfile cannot be found or it fails with C<ENOSYS>,
454C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or C<ENOTSOCK>, 514C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
455it will be emulated, so you can call C<aio_sendfile> on any type of 515C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
456filehandle regardless of the limitations of the operating system. 516type of filehandle regardless of the limitations of the operating system.
517
518As native sendfile syscalls (as practically any non-POSIX interface hacked
519together in a hurry to improve benchmark numbers) tend to be rather buggy
520on many systems, this implementation tries to work around some known bugs
521in Linux and FreeBSD kernels (probably others, too), but that might fail,
522so you really really should check the return value of C<aio_sendfile> -
523fewer bytes than expected might have been transferred.
457 524
458 525
459=item aio_readahead $fh,$offset,$length, $callback->($retval) 526=item aio_readahead $fh,$offset,$length, $callback->($retval)
460 527
461C<aio_readahead> populates the page cache with data from a file so that 528C<aio_readahead> populates the page cache with data from a file so that
465whole pages, so that offset is effectively rounded down to a page boundary 532whole pages, so that offset is effectively rounded down to a page boundary
466and bytes are read up to the next page boundary greater than or equal to 533and bytes are read up to the next page boundary greater than or equal to
467(off-set+length). C<aio_readahead> does not read beyond the end of the 534(off-set+length). C<aio_readahead> does not read beyond the end of the
468file. The current file offset of the file is left unchanged. 535file. The current file offset of the file is left unchanged.
469 536
470If that syscall doesn't exist (likely if your OS isn't Linux) it will be 537If that syscall doesn't exist (likely if your kernel isn't Linux) it will
471emulated by simply reading the data, which would have a similar effect. 538be emulated by simply reading the data, which would have a similar effect.
472 539
473 540
474=item aio_stat $fh_or_path, $callback->($status) 541=item aio_stat $fh_or_path, $callback->($status)
475 542
476=item aio_lstat $fh, $callback->($status) 543=item aio_lstat $fh, $callback->($status)
545 namemax => 255, 612 namemax => 255,
546 frsize => 1024, 613 frsize => 1024,
547 fsid => 1810 614 fsid => 1810
548 } 615 }
549 616
617Here is a (likely partial - send me updates!) list of fsid values used by
618Linux - it is safe to hardcode these when C<$^O> is C<linux>:
619
620 0x0000adf5 adfs
621 0x0000adff affs
622 0x5346414f afs
623 0x09041934 anon-inode filesystem
624 0x00000187 autofs
625 0x42465331 befs
626 0x1badface bfs
627 0x42494e4d binfmt_misc
628 0x9123683e btrfs
629 0x0027e0eb cgroupfs
630 0xff534d42 cifs
631 0x73757245 coda
632 0x012ff7b7 coh
633 0x28cd3d45 cramfs
634 0x453dcd28 cramfs-wend (wrong endianness)
635 0x64626720 debugfs
636 0x00001373 devfs
637 0x00001cd1 devpts
638 0x0000f15f ecryptfs
639 0x00414a53 efs
640 0x0000137d ext
641 0x0000ef53 ext2/ext3/ext4
642 0x0000ef51 ext2
643 0xf2f52010 f2fs
644 0x00004006 fat
645 0x65735546 fuseblk
646 0x65735543 fusectl
647 0x0bad1dea futexfs
648 0x01161970 gfs2
649 0x47504653 gpfs
650 0x00004244 hfs
651 0xf995e849 hpfs
652 0x00c0ffee hostfs
653 0x958458f6 hugetlbfs
654 0x2bad1dea inotifyfs
655 0x00009660 isofs
656 0x000072b6 jffs2
657 0x3153464a jfs
658 0x6b414653 k-afs
659 0x0bd00bd0 lustre
660 0x0000137f minix
661 0x0000138f minix 30 char names
662 0x00002468 minix v2
663 0x00002478 minix v2 30 char names
664 0x00004d5a minix v3
665 0x19800202 mqueue
666 0x00004d44 msdos
667 0x0000564c novell
668 0x00006969 nfs
669 0x6e667364 nfsd
670 0x00003434 nilfs
671 0x5346544e ntfs
672 0x00009fa1 openprom
673 0x7461636F ocfs2
674 0x00009fa0 proc
675 0x6165676c pstorefs
676 0x0000002f qnx4
677 0x68191122 qnx6
678 0x858458f6 ramfs
679 0x52654973 reiserfs
680 0x00007275 romfs
681 0x67596969 rpc_pipefs
682 0x73636673 securityfs
683 0xf97cff8c selinux
684 0x0000517b smb
685 0x534f434b sockfs
686 0x73717368 squashfs
687 0x62656572 sysfs
688 0x012ff7b6 sysv2
689 0x012ff7b5 sysv4
690 0x01021994 tmpfs
691 0x15013346 udf
692 0x00011954 ufs
693 0x54190100 ufs byteswapped
694 0x00009fa2 usbdevfs
695 0x01021997 v9fs
696 0xa501fcf5 vxfs
697 0xabba1974 xenfs
698 0x012ff7b4 xenix
699 0x58465342 xfs
700 0x012fd16d xia
550 701
551=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 702=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
552 703
553Works like perl's C<utime> function (including the special case of $atime 704Works like perl's C<utime> function (including the special case of $atime
554and $mtime being undef). Fractional times are supported if the underlying 705and $mtime being undef). Fractional times are supported if the underlying
582=item aio_truncate $fh_or_path, $offset, $callback->($status) 733=item aio_truncate $fh_or_path, $offset, $callback->($status)
583 734
584Works like truncate(2) or ftruncate(2). 735Works like truncate(2) or ftruncate(2).
585 736
586 737
738=item aio_allocate $fh, $mode, $offset, $len, $callback->($status)
739
740Allocates or frees disk space according to the C<$mode> argument. See the
741linux C<fallocate> documentation for details.
742
743C<$mode> is usually C<0> or C<IO::AIO::FALLOC_FL_KEEP_SIZE> to allocate
744space, or C<IO::AIO::FALLOC_FL_PUNCH_HOLE | IO::AIO::FALLOC_FL_KEEP_SIZE>,
745to deallocate a file range.
746
747IO::AIO also supports C<FALLOC_FL_COLLAPSE_RANGE>, to remove a range
748(without leaving a hole), C<FALLOC_FL_ZERO_RANGE>, to zero a range,
749C<FALLOC_FL_INSERT_RANGE> to insert a range and C<FALLOC_FL_UNSHARE_RANGE>
750to unshare shared blocks (see your L<fallocate(2)> manpage).
751
752The file system block size used by C<fallocate> is presumably the
753C<f_bsize> returned by C<statvfs>, but different filesystems and filetypes
754can dictate other limitations.
755
756If C<fallocate> isn't available or cannot be emulated (currently no
757emulation will be attempted), passes C<-1> and sets C<$!> to C<ENOSYS>.
758
759
587=item aio_chmod $fh_or_path, $mode, $callback->($status) 760=item aio_chmod $fh_or_path, $mode, $callback->($status)
588 761
589Works like perl's C<chmod> function. 762Works like perl's C<chmod> function.
590 763
591 764
593 766
594Asynchronously unlink (delete) a file and call the callback with the 767Asynchronously unlink (delete) a file and call the callback with the
595result code. 768result code.
596 769
597 770
598=item aio_mknod $path, $mode, $dev, $callback->($status) 771=item aio_mknod $pathname, $mode, $dev, $callback->($status)
599 772
600[EXPERIMENTAL] 773[EXPERIMENTAL]
601 774
602Asynchronously create a device node (or fifo). See mknod(2). 775Asynchronously create a device node (or fifo). See mknod(2).
603 776
604The only (POSIX-) portable way of calling this function is: 777The only (POSIX-) portable way of calling this function is:
605 778
606 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 779 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
607 780
608See C<aio_stat> for info about some potentially helpful extra constants 781See C<aio_stat> for info about some potentially helpful extra constants
609and functions. 782and functions.
610 783
611=item aio_link $srcpath, $dstpath, $callback->($status) 784=item aio_link $srcpath, $dstpath, $callback->($status)
618 791
619Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 792Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
620the path C<$dstpath> and call the callback with the result code. 793the path C<$dstpath> and call the callback with the result code.
621 794
622 795
623=item aio_readlink $path, $callback->($link) 796=item aio_readlink $pathname, $callback->($link)
624 797
625Asynchronously read the symlink specified by C<$path> and pass it to 798Asynchronously read the symlink specified by C<$path> and pass it to
626the callback. If an error occurs, nothing or undef gets passed to the 799the callback. If an error occurs, nothing or undef gets passed to the
627callback. 800callback.
628 801
629 802
803=item aio_realpath $pathname, $callback->($path)
804
805Asynchronously make the path absolute and resolve any symlinks in
806C<$path>. The resulting path only consists of directories (same as
807L<Cwd::realpath>).
808
809This request can be used to get the absolute path of the current working
810directory by passing it a path of F<.> (a single dot).
811
812
630=item aio_rename $srcpath, $dstpath, $callback->($status) 813=item aio_rename $srcpath, $dstpath, $callback->($status)
631 814
632Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 815Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
633rename(2) and call the callback with the result code. 816rename(2) and call the callback with the result code.
817
818On systems that support the AIO::WD working directory abstraction
819natively, the case C<[$wd, "."]> as C<$srcpath> is specialcased - instead
820of failing, C<rename> is called on the absolute path of C<$wd>.
821
822
823=item aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
824
825Basically a version of C<aio_rename> with an additional C<$flags>
826argument. Calling this with C<$flags=0> is the same as calling
827C<aio_rename>.
828
829Non-zero flags are currently only supported on GNU/Linux systems that
830support renameat2. Other systems fail with C<ENOSYS> in this case.
831
832The following constants are available (missing ones are, as usual C<0>),
833see renameat2(2) for details:
834
835C<IO::AIO::RENAME_NOREPLACE>, C<IO::AIO::RENAME_EXCHANGE>
836and C<IO::AIO::RENAME_WHITEOUT>.
634 837
635 838
636=item aio_mkdir $pathname, $mode, $callback->($status) 839=item aio_mkdir $pathname, $mode, $callback->($status)
637 840
638Asynchronously mkdir (create) a directory and call the callback with 841Asynchronously mkdir (create) a directory and call the callback with
643=item aio_rmdir $pathname, $callback->($status) 846=item aio_rmdir $pathname, $callback->($status)
644 847
645Asynchronously rmdir (delete) a directory and call the callback with the 848Asynchronously rmdir (delete) a directory and call the callback with the
646result code. 849result code.
647 850
851On systems that support the AIO::WD working directory abstraction
852natively, the case C<[$wd, "."]> is specialcased - instead of failing,
853C<rmdir> is called on the absolute path of C<$wd>.
854
648 855
649=item aio_readdir $pathname, $callback->($entries) 856=item aio_readdir $pathname, $callback->($entries)
650 857
651Unlike the POSIX call of the same name, C<aio_readdir> reads an entire 858Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
652directory (i.e. opendir + readdir + closedir). The entries will not be 859directory (i.e. opendir + readdir + closedir). The entries will not be
656array-ref with the filenames. 863array-ref with the filenames.
657 864
658 865
659=item aio_readdirx $pathname, $flags, $callback->($entries, $flags) 866=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
660 867
661Quite similar to C<aio_readdir>, but the C<$flags> argument allows to tune 868Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
662behaviour and output format. In case of an error, C<$entries> will be 869tune behaviour and output format. In case of an error, C<$entries> will be
663C<undef>. 870C<undef>.
664 871
665The flags are a combination of the following constants, ORed together (the 872The flags are a combination of the following constants, ORed together (the
666flags will also be passed to the callback, possibly modified): 873flags will also be passed to the callback, possibly modified):
667 874
668=over 4 875=over 4
669 876
670=item IO::AIO::READDIR_DENTS 877=item IO::AIO::READDIR_DENTS
671 878
672When this flag is off, then the callback gets an arrayref with of names 879When this flag is off, then the callback gets an arrayref consisting of
673only (as with C<aio_readdir>), otherwise it gets an arrayref with 880names only (as with C<aio_readdir>), otherwise it gets an arrayref with
674C<[$name, $type, $inode]> arrayrefs, each describing a single directory 881C<[$name, $type, $inode]> arrayrefs, each describing a single directory
675entry in more detail. 882entry in more detail.
676 883
677C<$name> is the name of the entry. 884C<$name> is the name of the entry.
678 885
691systems that do not deliver the inode information. 898systems that do not deliver the inode information.
692 899
693=item IO::AIO::READDIR_DIRS_FIRST 900=item IO::AIO::READDIR_DIRS_FIRST
694 901
695When this flag is set, then the names will be returned in an order where 902When this flag is set, then the names will be returned in an order where
696likely directories come first. This is useful when you need to quickly 903likely directories come first, in optimal stat order. This is useful when
697find directories, or you want to find all directories while avoiding to 904you need to quickly find directories, or you want to find all directories
698stat() each entry. 905while avoiding to stat() each entry.
699 906
700If the system returns type information in readdir, then this is used 907If the system returns type information in readdir, then this is used
701to find directories directly. Otherwise, likely directories are files 908to find directories directly. Otherwise, likely directories are names
702beginning with ".", or otherwise files with no dots, of which files with 909beginning with ".", or otherwise names with no dots, of which names with
703short names are tried first. 910short names are tried first.
704 911
705=item IO::AIO::READDIR_STAT_ORDER 912=item IO::AIO::READDIR_STAT_ORDER
706 913
707When this flag is set, then the names will be returned in an order 914When this flag is set, then the names will be returned in an order
714 921
715=item IO::AIO::READDIR_FOUND_UNKNOWN 922=item IO::AIO::READDIR_FOUND_UNKNOWN
716 923
717This flag should not be set when calling C<aio_readdirx>. Instead, it 924This flag should not be set when calling C<aio_readdirx>. Instead, it
718is being set by C<aio_readdirx>, when any of the C<$type>'s found were 925is being set by C<aio_readdirx>, when any of the C<$type>'s found were
719C<IO::AIO::DT_UNKNOWN>. The absense of this flag therefore indicates that all 926C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
720C<$type>'s are known, which can be used to speed up some algorithms. 927C<$type>'s are known, which can be used to speed up some algorithms.
721 928
722=back 929=back
723 930
724 931
932=item aio_slurp $pathname, $offset, $length, $data, $callback->($status)
933
934Opens, reads and closes the given file. The data is put into C<$data>,
935which is resized as required.
936
937If C<$offset> is negative, then it is counted from the end of the file.
938
939If C<$length> is zero, then the remaining length of the file is
940used. Also, in this case, the same limitations to modifying C<$data> apply
941as when IO::AIO::mmap is used, i.e. it must only be modified in-place
942with C<substr>. If the size of the file is known, specifying a non-zero
943C<$length> results in a performance advantage.
944
945This request is similar to the older C<aio_load> request, but since it is
946a single request, it might be more efficient to use.
947
948Example: load F</etc/passwd> into C<$passwd>.
949
950 my $passwd;
951 aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
952 $_[0] >= 0
953 or die "/etc/passwd: $!\n";
954
955 printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
956 print $passwd;
957 };
958 IO::AIO::flush;
959
960
725=item aio_load $path, $data, $callback->($status) 961=item aio_load $pathname, $data, $callback->($status)
726 962
727This is a composite request that tries to fully load the given file into 963This is a composite request that tries to fully load the given file into
728memory. Status is the same as with aio_read. 964memory. Status is the same as with aio_read.
965
966Using C<aio_slurp> might be more efficient, as it is a single request.
729 967
730=cut 968=cut
731 969
732sub aio_load($$;$) { 970sub aio_load($$;$) {
733 my ($path, undef, $cb) = @_; 971 my ($path, undef, $cb) = @_;
753=item aio_copy $srcpath, $dstpath, $callback->($status) 991=item aio_copy $srcpath, $dstpath, $callback->($status)
754 992
755Try to copy the I<file> (directories not supported as either source or 993Try to copy the I<file> (directories not supported as either source or
756destination) from C<$srcpath> to C<$dstpath> and call the callback with 994destination) from C<$srcpath> to C<$dstpath> and call the callback with
757a status of C<0> (ok) or C<-1> (error, see C<$!>). 995a status of C<0> (ok) or C<-1> (error, see C<$!>).
996
997Existing destination files will be truncated.
758 998
759This is a composite request that creates the destination file with 999This is a composite request that creates the destination file with
760mode 0200 and copies the contents of the source file into it using 1000mode 0200 and copies the contents of the source file into it using
761C<aio_sendfile>, followed by restoring atime, mtime, access mode and 1001C<aio_sendfile>, followed by restoring atime, mtime, access mode and
762uid/gid, in that order. 1002uid/gid, in that order.
852 if ($_[0] && $! == EXDEV) { 1092 if ($_[0] && $! == EXDEV) {
853 aioreq_pri $pri; 1093 aioreq_pri $pri;
854 add $grp aio_copy $src, $dst, sub { 1094 add $grp aio_copy $src, $dst, sub {
855 $grp->result ($_[0]); 1095 $grp->result ($_[0]);
856 1096
857 if (!$_[0]) { 1097 unless ($_[0]) {
858 aioreq_pri $pri; 1098 aioreq_pri $pri;
859 add $grp aio_unlink $src; 1099 add $grp aio_unlink $src;
860 } 1100 }
861 }; 1101 };
862 } else { 1102 } else {
865 }; 1105 };
866 1106
867 $grp 1107 $grp
868} 1108}
869 1109
870=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 1110=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
871 1111
872Scans a directory (similar to C<aio_readdir>) but additionally tries to 1112Scans a directory (similar to C<aio_readdir>) but additionally tries to
873efficiently separate the entries of directory C<$path> into two sets of 1113efficiently separate the entries of directory C<$path> into two sets of
874names, directories you can recurse into (directories), and ones you cannot 1114names, directories you can recurse into (directories), and ones you cannot
875recurse into (everything else, including symlinks to directories). 1115recurse into (everything else, including symlinks to directories).
876 1116
877C<aio_scandir> is a composite request that creates of many sub requests_ 1117C<aio_scandir> is a composite request that generates many sub requests.
878C<$maxreq> specifies the maximum number of outstanding aio requests that 1118C<$maxreq> specifies the maximum number of outstanding aio requests that
879this function generates. If it is C<< <= 0 >>, then a suitable default 1119this function generates. If it is C<< <= 0 >>, then a suitable default
880will be chosen (currently 4). 1120will be chosen (currently 4).
881 1121
882On error, the callback is called without arguments, otherwise it receives 1122On error, the callback is called without arguments, otherwise it receives
906Then entries will be sorted into likely directories a non-initial dot 1146Then entries will be sorted into likely directories a non-initial dot
907currently) and likely non-directories (see C<aio_readdirx>). Then every 1147currently) and likely non-directories (see C<aio_readdirx>). Then every
908entry plus an appended C</.> will be C<stat>'ed, likely directories first, 1148entry plus an appended C</.> will be C<stat>'ed, likely directories first,
909in order of their inode numbers. If that succeeds, it assumes that the 1149in order of their inode numbers. If that succeeds, it assumes that the
910entry is a directory or a symlink to directory (which will be checked 1150entry is a directory or a symlink to directory (which will be checked
911seperately). This is often faster than stat'ing the entry itself because 1151separately). This is often faster than stat'ing the entry itself because
912filesystems might detect the type of the entry without reading the inode 1152filesystems might detect the type of the entry without reading the inode
913data (e.g. ext2fs filetype feature), even on systems that cannot return 1153data (e.g. ext2fs filetype feature), even on systems that cannot return
914the filetype information on readdir. 1154the filetype information on readdir.
915 1155
916If the known number of directories (link count - 2) has been reached, the 1156If the known number of directories (link count - 2) has been reached, the
932 1172
933 my $grp = aio_group $cb; 1173 my $grp = aio_group $cb;
934 1174
935 $maxreq = 4 if $maxreq <= 0; 1175 $maxreq = 4 if $maxreq <= 0;
936 1176
937 # stat once 1177 # get a wd object
938 aioreq_pri $pri; 1178 aioreq_pri $pri;
939 add $grp aio_stat $path, sub { 1179 add $grp aio_wd $path, sub {
1180 $_[0]
940 return $grp->result () if $_[0]; 1181 or return $grp->result ();
941 my $now = time;
942 my $hash1 = join ":", (stat _)[0,1,3,7,9];
943 1182
944 # read the directory entries 1183 my $wd = [shift, "."];
1184
1185 # stat once
945 aioreq_pri $pri; 1186 aioreq_pri $pri;
946 add $grp aio_readdirx $path, READDIR_DIRS_FIRST, sub { 1187 add $grp aio_stat $wd, sub {
947 my $entries = shift
948 or return $grp->result (); 1188 return $grp->result () if $_[0];
1189 my $now = time;
1190 my $hash1 = join ":", (stat _)[0,1,3,7,9];
949 1191
950 # stat the dir another time 1192 # read the directory entries
951 aioreq_pri $pri; 1193 aioreq_pri $pri;
1194 add $grp aio_readdirx $wd, READDIR_DIRS_FIRST, sub {
1195 my $entries = shift
1196 or return $grp->result ();
1197
1198 # stat the dir another time
1199 aioreq_pri $pri;
952 add $grp aio_stat $path, sub { 1200 add $grp aio_stat $wd, sub {
953 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 1201 my $hash2 = join ":", (stat _)[0,1,3,7,9];
954 1202
955 my $ndirs; 1203 my $ndirs;
956 1204
957 # take the slow route if anything looks fishy 1205 # take the slow route if anything looks fishy
958 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 1206 if ($hash1 ne $hash2 or (stat _)[9] == $now) {
959 $ndirs = -1; 1207 $ndirs = -1;
960 } else { 1208 } else {
961 # if nlink == 2, we are finished 1209 # if nlink == 2, we are finished
962 # for non-posix-fs's, we rely on nlink < 2 1210 # for non-posix-fs's, we rely on nlink < 2
963 $ndirs = (stat _)[3] - 2 1211 $ndirs = (stat _)[3] - 2
964 or return $grp->result ([], $entries); 1212 or return $grp->result ([], $entries);
965 } 1213 }
966 1214
967 my (@dirs, @nondirs); 1215 my (@dirs, @nondirs);
968 1216
969 my $statgrp = add $grp aio_group sub { 1217 my $statgrp = add $grp aio_group sub {
970 $grp->result (\@dirs, \@nondirs); 1218 $grp->result (\@dirs, \@nondirs);
971 }; 1219 };
972 1220
973 limit $statgrp $maxreq; 1221 limit $statgrp $maxreq;
974 feed $statgrp sub { 1222 feed $statgrp sub {
975 return unless @$entries; 1223 return unless @$entries;
976 my $entry = shift @$entries; 1224 my $entry = shift @$entries;
977 1225
978 aioreq_pri $pri; 1226 aioreq_pri $pri;
1227 $wd->[1] = "$entry/.";
979 add $statgrp aio_stat "$path/$entry/.", sub { 1228 add $statgrp aio_stat $wd, sub {
980 if ($_[0] < 0) { 1229 if ($_[0] < 0) {
981 push @nondirs, $entry; 1230 push @nondirs, $entry;
982 } else { 1231 } else {
983 # need to check for real directory 1232 # need to check for real directory
984 aioreq_pri $pri; 1233 aioreq_pri $pri;
1234 $wd->[1] = $entry;
985 add $statgrp aio_lstat "$path/$entry", sub { 1235 add $statgrp aio_lstat $wd, sub {
986 if (-d _) { 1236 if (-d _) {
987 push @dirs, $entry; 1237 push @dirs, $entry;
988 1238
989 unless (--$ndirs) { 1239 unless (--$ndirs) {
990 push @nondirs, @$entries; 1240 push @nondirs, @$entries;
991 feed $statgrp; 1241 feed $statgrp;
1242 }
1243 } else {
1244 push @nondirs, $entry;
992 } 1245 }
993 } else {
994 push @nondirs, $entry;
995 } 1246 }
996 } 1247 }
997 } 1248 };
998 }; 1249 };
999 }; 1250 };
1000 }; 1251 };
1001 }; 1252 };
1002 }; 1253 };
1003 1254
1004 $grp 1255 $grp
1005} 1256}
1006 1257
1007=item aio_rmtree $path, $callback->($status) 1258=item aio_rmtree $pathname, $callback->($status)
1008 1259
1009Delete a directory tree starting (and including) C<$path>, return the 1260Delete a directory tree starting (and including) C<$path>, return the
1010status of the final C<rmdir> only. This is a composite request that 1261status of the final C<rmdir> only. This is a composite request that
1011uses C<aio_scandir> to recurse into and rmdir directories, and unlink 1262uses C<aio_scandir> to recurse into and rmdir directories, and unlink
1012everything else. 1263everything else.
1013 1264
1014=cut 1265=cut
1015 1266
1037 }; 1288 };
1038 1289
1039 $grp 1290 $grp
1040} 1291}
1041 1292
1293=item aio_fcntl $fh, $cmd, $arg, $callback->($status)
1294
1295=item aio_ioctl $fh, $request, $buf, $callback->($status)
1296
1297These work just like the C<fcntl> and C<ioctl> built-in functions, except
1298they execute asynchronously and pass the return value to the callback.
1299
1300Both calls can be used for a lot of things, some of which make more sense
1301to run asynchronously in their own thread, while some others make less
1302sense. For example, calls that block waiting for external events, such
1303as locking, will also lock down an I/O thread while it is waiting, which
1304can deadlock the whole I/O system. At the same time, there might be no
1305alternative to using a thread to wait.
1306
1307So in general, you should only use these calls for things that do
1308(filesystem) I/O, not for things that wait for other events (network,
1309other processes), although if you are careful and know what you are doing,
1310you still can.
1311
1312The following constants are available (missing ones are, as usual C<0>):
1313
1314C<F_DUPFD_CLOEXEC>,
1315
1316C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>,
1317
1318C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>.
1319
1320C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>,
1321C<FS_IOC_FIEMAP>.
1322
1323C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>,
1324C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>.
1325
1326C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>,
1327C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>,
1328C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>,
1329C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>,
1330C<FS_FL_USER_MODIFIABLE>.
1331
1332C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>,
1333C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>,
1334C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>,
1335C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>,
1336
1042=item aio_sync $callback->($status) 1337=item aio_sync $callback->($status)
1043 1338
1044Asynchronously call sync and call the callback when finished. 1339Asynchronously call sync and call the callback when finished.
1045 1340
1046=item aio_fsync $fh, $callback->($status) 1341=item aio_fsync $fh, $callback->($status)
1053Asynchronously call fdatasync on the given filehandle and call the 1348Asynchronously call fdatasync on the given filehandle and call the
1054callback with the fdatasync result code. 1349callback with the fdatasync result code.
1055 1350
1056If this call isn't available because your OS lacks it or it couldn't be 1351If this call isn't available because your OS lacks it or it couldn't be
1057detected, it will be emulated by calling C<fsync> instead. 1352detected, it will be emulated by calling C<fsync> instead.
1353
1354=item aio_syncfs $fh, $callback->($status)
1355
1356Asynchronously call the syncfs syscall to sync the filesystem associated
1357to the given filehandle and call the callback with the syncfs result
1358code. If syncfs is not available, calls sync(), but returns C<-1> and sets
1359errno to C<ENOSYS> nevertheless.
1058 1360
1059=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 1361=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
1060 1362
1061Sync the data portion of the file specified by C<$offset> and C<$length> 1363Sync the data portion of the file specified by C<$offset> and C<$length>
1062to disk (but NOT the metadata), by calling the Linux-specific 1364to disk (but NOT the metadata), by calling the Linux-specific
1066C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>, 1368C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
1067C<IO::AIO::SYNC_FILE_RANGE_WRITE> and 1369C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
1068C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range 1370C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
1069manpage for details. 1371manpage for details.
1070 1372
1071=item aio_pathsync $path, $callback->($status) 1373=item aio_pathsync $pathname, $callback->($status)
1072 1374
1073This request tries to open, fsync and close the given path. This is a 1375This request tries to open, fsync and close the given path. This is a
1074composite request intended to sync directories after directory operations 1376composite request intended to sync directories after directory operations
1075(E.g. rename). This might not work on all operating systems or have any 1377(E.g. rename). This might not work on all operating systems or have any
1076specific effect, but usually it makes sure that directory changes get 1378specific effect, but usually it makes sure that directory changes get
1107 }; 1409 };
1108 1410
1109 $grp 1411 $grp
1110} 1412}
1111 1413
1112=item aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 1414=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
1113 1415
1114This is a rather advanced IO::AIO call, which only works on mmap(2)ed 1416This is a rather advanced IO::AIO call, which only works on mmap(2)ed
1115scalars (see the C<IO::AIO::mmap> function, although it also works on data 1417scalars (see the C<IO::AIO::mmap> function, although it also works on data
1116scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the 1418scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
1117scalar must only be modified in-place while an aio operation is pending on 1419scalar must only be modified in-place while an aio operation is pending on
1119 1421
1120It calls the C<msync> function of your OS, if available, with the memory 1422It calls the C<msync> function of your OS, if available, with the memory
1121area starting at C<$offset> in the string and ending C<$length> bytes 1423area starting at C<$offset> in the string and ending C<$length> bytes
1122later. If C<$length> is negative, counts from the end, and if C<$length> 1424later. If C<$length> is negative, counts from the end, and if C<$length>
1123is C<undef>, then it goes till the end of the string. The flags can be 1425is C<undef>, then it goes till the end of the string. The flags can be
1124a combination of C<IO::AIO::MS_ASYNC>, C<IO::AIO::MS_INVALIDATE> and 1426either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional
1125C<IO::AIO::MS_SYNC>. 1427C<IO::AIO::MS_INVALIDATE>.
1126 1428
1127=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 1429=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
1128 1430
1129This is a rather advanced IO::AIO call, which works best on mmap(2)ed 1431This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1130scalars. 1432scalars.
1131 1433
1132It touches (reads or writes) all memory pages in the specified 1434It touches (reads or writes) all memory pages in the specified
1133range inside the scalar. All caveats and parameters are the same 1435range inside the scalar. All caveats and parameters are the same
1134as for C<aio_msync>, above, except for flags, which must be either 1436as for C<aio_msync>, above, except for flags, which must be either
1135C<0> (which reads all pages and ensures they are instantiated) or 1437C<0> (which reads all pages and ensures they are instantiated) or
1136C<IO::AIO::MT_MODIFY>, which modifies the memory page s(by reading and 1438C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and
1137writing an octet from it, which dirties the page). 1439writing an octet from it, which dirties the page).
1138 1440
1139=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status) 1441=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1140 1442
1141This is a rather advanced IO::AIO call, which works best on mmap(2)ed 1443This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1172documented under L<MISCELLANEOUS FUNCTIONS>. 1474documented under L<MISCELLANEOUS FUNCTIONS>.
1173 1475
1174Example: asynchronously lock all current and future pages into memory. 1476Example: asynchronously lock all current and future pages into memory.
1175 1477
1176 aio_mlockall IO::AIO::MCL_FUTURE; 1478 aio_mlockall IO::AIO::MCL_FUTURE;
1479
1480=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
1481
1482Queries the extents of the given file (by calling the Linux C<FIEMAP>
1483ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If
1484the ioctl is not available on your OS, then this request will fail with
1485C<ENOSYS>.
1486
1487C<$start> is the starting offset to query extents for, C<$length> is the
1488size of the range to query - if it is C<undef>, then the whole file will
1489be queried.
1490
1491C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or
1492C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also
1493exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query
1494the data portion.
1495
1496C<$count> is the maximum number of extent records to return. If it is
1497C<undef>, then IO::AIO queries all extents of the range. As a very special
1498case, if it is C<0>, then the callback receives the number of extents
1499instead of the extents themselves (which is unreliable, see below).
1500
1501If an error occurs, the callback receives no arguments. The special
1502C<errno> value C<IO::AIO::EBADR> is available to test for flag errors.
1503
1504Otherwise, the callback receives an array reference with extent
1505structures. Each extent structure is an array reference itself, with the
1506following members:
1507
1508 [$logical, $physical, $length, $flags]
1509
1510Flags is any combination of the following flag values (typically either C<0>
1511or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)):
1512
1513C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>,
1514C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>,
1515C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>,
1516C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>,
1517C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or
1518C<IO::AIO::FIEMAP_EXTENT_SHARED>.
1519
1520At the time of this writing (Linux 3.2), this request is unreliable unless
1521C<$count> is C<undef>, as the kernel has all sorts of bugs preventing
1522it to return all extents of a range for files with a large number of
1523extents. The code (only) works around all these issues if C<$count> is
1524C<undef>.
1177 1525
1178=item aio_group $callback->(...) 1526=item aio_group $callback->(...)
1179 1527
1180This is a very special aio request: Instead of doing something, it is a 1528This is a very special aio request: Instead of doing something, it is a
1181container for other aio requests, which is useful if you want to bundle 1529container for other aio requests, which is useful if you want to bundle
1218like sleep and file handle readable/writable, the overhead this creates is 1566like sleep and file handle readable/writable, the overhead this creates is
1219immense (it blocks a thread for a long time) so do not use this function 1567immense (it blocks a thread for a long time) so do not use this function
1220except to put your application under artificial I/O pressure. 1568except to put your application under artificial I/O pressure.
1221 1569
1222=back 1570=back
1571
1572
1573=head2 IO::AIO::WD - multiple working directories
1574
1575Your process only has one current working directory, which is used by all
1576threads. This makes it hard to use relative paths (some other component
1577could call C<chdir> at any time, and it is hard to control when the path
1578will be used by IO::AIO).
1579
1580One solution for this is to always use absolute paths. This usually works,
1581but can be quite slow (the kernel has to walk the whole path on every
1582access), and can also be a hassle to implement.
1583
1584Newer POSIX systems have a number of functions (openat, fdopendir,
1585futimensat and so on) that make it possible to specify working directories
1586per operation.
1587
1588For portability, and because the clowns who "designed", or shall I write,
1589perpetrated this new interface were obviously half-drunk, this abstraction
1590cannot be perfect, though.
1591
1592IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
1593object. This object stores the canonicalised, absolute version of the
1594path, and on systems that allow it, also a directory file descriptor.
1595
1596Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
1597or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
1598object and a pathname instead (or the IO::AIO::WD object alone, which
1599gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
1600IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
1601to that IO::AIO::WD object.
1602
1603For example, to get a wd object for F</etc> and then stat F<passwd>
1604inside, you would write:
1605
1606 aio_wd "/etc", sub {
1607 my $etcdir = shift;
1608
1609 # although $etcdir can be undef on error, there is generally no reason
1610 # to check for errors here, as aio_stat will fail with ENOENT
1611 # when $etcdir is undef.
1612
1613 aio_stat [$etcdir, "passwd"], sub {
1614 # yay
1615 };
1616 };
1617
1618The fact that C<aio_wd> is a request and not a normal function shows that
1619creating an IO::AIO::WD object is itself a potentially blocking operation,
1620which is why it is done asynchronously.
1621
1622To stat the directory obtained with C<aio_wd> above, one could write
1623either of the following three request calls:
1624
1625 aio_lstat "/etc" , sub { ... # pathname as normal string
1626 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1627 aio_lstat $wd , sub { ... # shorthand for the previous
1628
1629As with normal pathnames, IO::AIO keeps a copy of the working directory
1630object and the pathname string, so you could write the following without
1631causing any issues due to C<$path> getting reused:
1632
1633 my $path = [$wd, undef];
1634
1635 for my $name (qw(abc def ghi)) {
1636 $path->[1] = $name;
1637 aio_stat $path, sub {
1638 # ...
1639 };
1640 }
1641
1642There are some caveats: when directories get renamed (or deleted), the
1643pathname string doesn't change, so will point to the new directory (or
1644nowhere at all), while the directory fd, if available on the system,
1645will still point to the original directory. Most functions accepting a
1646pathname will use the directory fd on newer systems, and the string on
1647older systems. Some functions (such as C<aio_realpath>) will always rely on
1648the string form of the pathname.
1649
1650So this functionality is mainly useful to get some protection against
1651C<chdir>, to easily get an absolute path out of a relative path for future
1652reference, and to speed up doing many operations in the same directory
1653(e.g. when stat'ing all files in a directory).
1654
1655The following functions implement this working directory abstraction:
1656
1657=over 4
1658
1659=item aio_wd $pathname, $callback->($wd)
1660
1661Asynchonously canonicalise the given pathname and convert it to an
1662IO::AIO::WD object representing it. If possible and supported on the
1663system, also open a directory fd to speed up pathname resolution relative
1664to this working directory.
1665
1666If something goes wrong, then C<undef> is passwd to the callback instead
1667of a working directory object and C<$!> is set appropriately. Since
1668passing C<undef> as working directory component of a pathname fails the
1669request with C<ENOENT>, there is often no need for error checking in the
1670C<aio_wd> callback, as future requests using the value will fail in the
1671expected way.
1672
1673=item IO::AIO::CWD
1674
1675This is a compiletime constant (object) that represents the process
1676current working directory.
1677
1678Specifying this object as working directory object for a pathname is as if
1679the pathname would be specified directly, without a directory object. For
1680example, these calls are functionally identical:
1681
1682 aio_stat "somefile", sub { ... };
1683 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1684
1685=back
1686
1687To recover the path associated with an IO::AIO::WD object, you can use
1688C<aio_realpath>:
1689
1690 aio_realpath $wd, sub {
1691 warn "path is $_[0]\n";
1692 };
1693
1694Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir>
1695sometimes, fall back to using an absolue path.
1223 1696
1224=head2 IO::AIO::REQ CLASS 1697=head2 IO::AIO::REQ CLASS
1225 1698
1226All non-aggregate C<aio_*> functions return an object of this class when 1699All non-aggregate C<aio_*> functions return an object of this class when
1227called in non-void context. 1700called in non-void context.
1345 1818
1346Sets a feeder/generator on this group: every group can have an attached 1819Sets a feeder/generator on this group: every group can have an attached
1347generator that generates requests if idle. The idea behind this is that, 1820generator that generates requests if idle. The idea behind this is that,
1348although you could just queue as many requests as you want in a group, 1821although you could just queue as many requests as you want in a group,
1349this might starve other requests for a potentially long time. For example, 1822this might starve other requests for a potentially long time. For example,
1350C<aio_scandir> might generate hundreds of thousands C<aio_stat> requests, 1823C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
1351delaying any later requests for a long time. 1824requests, delaying any later requests for a long time.
1352 1825
1353To avoid this, and allow incremental generation of requests, you can 1826To avoid this, and allow incremental generation of requests, you can
1354instead a group and set a feeder on it that generates those requests. The 1827instead a group and set a feeder on it that generates those requests. The
1355feed callback will be called whenever there are few enough (see C<limit>, 1828feed callback will be called whenever there are few enough (see C<limit>,
1356below) requests active in the group itself and is expected to queue more 1829below) requests active in the group itself and is expected to queue more
1405 1878
1406See C<poll_cb> for an example. 1879See C<poll_cb> for an example.
1407 1880
1408=item IO::AIO::poll_cb 1881=item IO::AIO::poll_cb
1409 1882
1410Process some outstanding events on the result pipe. You have to call this 1883Process some requests that have reached the result phase (i.e. they have
1411regularly. Returns C<0> if all events could be processed, or C<-1> if it 1884been executed but the results are not yet reported). You have to call
1412returned earlier for whatever reason. Returns immediately when no events 1885this "regularly" to finish outstanding requests.
1413are outstanding. The amount of events processed depends on the settings of
1414C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
1415 1886
1887Returns C<0> if all events could be processed (or there were no
1888events to process), or C<-1> if it returned earlier for whatever
1889reason. Returns immediately when no events are outstanding. The amount
1890of events processed depends on the settings of C<IO::AIO::max_poll_req>,
1891C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>.
1892
1416If not all requests were processed for whatever reason, the filehandle 1893If not all requests were processed for whatever reason, the poll file
1417will still be ready when C<poll_cb> returns, so normally you don't have to 1894descriptor will still be ready when C<poll_cb> returns, so normally you
1418do anything special to have it called later. 1895don't have to do anything special to have it called later.
1896
1897Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
1898ready, it can be beneficial to call this function from loops which submit
1899a lot of requests, to make sure the results get processed when they become
1900available and not just when the loop is finished and the event loop takes
1901over again. This function returns very fast when there are no outstanding
1902requests.
1419 1903
1420Example: Install an Event watcher that automatically calls 1904Example: Install an Event watcher that automatically calls
1421IO::AIO::poll_cb with high priority (more examples can be found in the 1905IO::AIO::poll_cb with high priority (more examples can be found in the
1422SYNOPSIS section, at the top of this document): 1906SYNOPSIS section, at the top of this document):
1423 1907
1425 poll => 'r', async => 1, 1909 poll => 'r', async => 1,
1426 cb => \&IO::AIO::poll_cb); 1910 cb => \&IO::AIO::poll_cb);
1427 1911
1428=item IO::AIO::poll_wait 1912=item IO::AIO::poll_wait
1429 1913
1430If there are any outstanding requests and none of them in the result 1914Wait until either at least one request is in the result phase or no
1431phase, wait till the result filehandle becomes ready for reading (simply 1915requests are outstanding anymore.
1432does a C<select> on the filehandle. This is useful if you want to 1916
1433synchronously wait for some requests to finish). 1917This is useful if you want to synchronously wait for some requests to
1918become ready, without actually handling them.
1434 1919
1435See C<nreqs> for an example. 1920See C<nreqs> for an example.
1436 1921
1437=item IO::AIO::poll 1922=item IO::AIO::poll
1438 1923
1525 2010
1526Under normal circumstances you don't need to call this function. 2011Under normal circumstances you don't need to call this function.
1527 2012
1528=item IO::AIO::max_idle $nthreads 2013=item IO::AIO::max_idle $nthreads
1529 2014
1530Limit the number of threads (default: 4) that are allowed to idle (i.e., 2015Limit the number of threads (default: 4) that are allowed to idle
1531threads that did not get a request to process within 10 seconds). That 2016(i.e., threads that did not get a request to process within the idle
1532means if a thread becomes idle while C<$nthreads> other threads are also 2017timeout (default: 10 seconds). That means if a thread becomes idle while
1533idle, it will free its resources and exit. 2018C<$nthreads> other threads are also idle, it will free its resources and
2019exit.
1534 2020
1535This is useful when you allow a large number of threads (e.g. 100 or 1000) 2021This is useful when you allow a large number of threads (e.g. 100 or 1000)
1536to allow for extremely high load situations, but want to free resources 2022to allow for extremely high load situations, but want to free resources
1537under normal circumstances (1000 threads can easily consume 30MB of RAM). 2023under normal circumstances (1000 threads can easily consume 30MB of RAM).
1538 2024
1539The default is probably ok in most situations, especially if thread 2025The default is probably ok in most situations, especially if thread
1540creation is fast. If thread creation is very slow on your system you might 2026creation is fast. If thread creation is very slow on your system you might
1541want to use larger values. 2027want to use larger values.
1542 2028
2029=item IO::AIO::idle_timeout $seconds
2030
2031Sets the minimum idle timeout (default 10) after which worker threads are
2032allowed to exit. SEe C<IO::AIO::max_idle>.
2033
1543=item IO::AIO::max_outstanding $maxreqs 2034=item IO::AIO::max_outstanding $maxreqs
2035
2036Sets the maximum number of outstanding requests to C<$nreqs>. If
2037you do queue up more than this number of requests, the next call to
2038C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
2039C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
2040longer exceeded.
2041
2042In other words, this setting does not enforce a queue limit, but can be
2043used to make poll functions block if the limit is exceeded.
1544 2044
1545This is a very bad function to use in interactive programs because it 2045This is a very bad function to use in interactive programs because it
1546blocks, and a bad way to reduce concurrency because it is inexact: Better 2046blocks, and a bad way to reduce concurrency because it is inexact: Better
1547use an C<aio_group> together with a feed callback. 2047use an C<aio_group> together with a feed callback.
1548 2048
1549Sets the maximum number of outstanding requests to C<$nreqs>. If you 2049Its main use is in scripts without an event loop - when you want to stat
1550do queue up more than this number of requests, the next call to the 2050a lot of files, you can write something like this:
1551C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
1552function will block until the limit is no longer exceeded.
1553 2051
1554The default value is very large, so there is no practical limit on the 2052 IO::AIO::max_outstanding 32;
1555number of outstanding requests.
1556 2053
1557You can still queue as many requests as you want. Therefore, 2054 for my $path (...) {
1558C<max_outstanding> is mainly useful in simple scripts (with low values) or 2055 aio_stat $path , ...;
1559as a stop gap to shield against fatal memory overflow (with large values). 2056 IO::AIO::poll_cb;
2057 }
2058
2059 IO::AIO::flush;
2060
2061The call to C<poll_cb> inside the loop will normally return instantly, but
2062as soon as more thna C<32> reqeusts are in-flight, it will block until
2063some requests have been handled. This keeps the loop from pushing a large
2064number of C<aio_stat> requests onto the queue.
2065
2066The default value for C<max_outstanding> is very large, so there is no
2067practical limit on the number of outstanding requests.
1560 2068
1561=back 2069=back
1562 2070
1563=head3 STATISTICAL INFORMATION 2071=head3 STATISTICAL INFORMATION
1564 2072
1586 2094
1587=back 2095=back
1588 2096
1589=head3 MISCELLANEOUS FUNCTIONS 2097=head3 MISCELLANEOUS FUNCTIONS
1590 2098
1591IO::AIO implements some functions that might be useful, but are not 2099IO::AIO implements some functions that are useful when you want to use
1592asynchronous. 2100some "Advanced I/O" function not available to in Perl, without going the
2101"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*>
2102counterpart.
1593 2103
1594=over 4 2104=over 4
2105
2106=item $numfd = IO::AIO::get_fdlimit
2107
2108This function is I<EXPERIMENTAL> and subject to change.
2109
2110Tries to find the current file descriptor limit and returns it, or
2111C<undef> and sets C<$!> in case of an error. The limit is one larger than
2112the highest valid file descriptor number.
2113
2114=item IO::AIO::min_fdlimit [$numfd]
2115
2116This function is I<EXPERIMENTAL> and subject to change.
2117
2118Try to increase the current file descriptor limit(s) to at least C<$numfd>
2119by changing the soft or hard file descriptor resource limit. If C<$numfd>
2120is missing, it will try to set a very high limit, although this is not
2121recommended when you know the actual minimum that you require.
2122
2123If the limit cannot be raised enough, the function makes a best-effort
2124attempt to increase the limit as much as possible, using various
2125tricks, while still failing. You can query the resulting limit using
2126C<IO::AIO::get_fdlimit>.
2127
2128If an error occurs, returns C<undef> and sets C<$!>, otherwise returns
2129true.
1595 2130
1596=item IO::AIO::sendfile $ofh, $ifh, $offset, $count 2131=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
1597 2132
1598Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>, 2133Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
1599but is blocking (this makes most sense if you know the input data is 2134but is blocking (this makes most sense if you know the input data is
1604 2139
1605=item IO::AIO::fadvise $fh, $offset, $len, $advice 2140=item IO::AIO::fadvise $fh, $offset, $len, $advice
1606 2141
1607Simply calls the C<posix_fadvise> function (see its 2142Simply calls the C<posix_fadvise> function (see its
1608manpage for details). The following advice constants are 2143manpage for details). The following advice constants are
1609avaiable: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>, 2144available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
1610C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>, 2145C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
1611C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>. 2146C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
1612 2147
1613On systems that do not implement C<posix_fadvise>, this function returns 2148On systems that do not implement C<posix_fadvise>, this function returns
1614ENOSYS, otherwise the return value of C<posix_fadvise>. 2149ENOSYS, otherwise the return value of C<posix_fadvise>.
1615 2150
1616=item IO::AIO::madvise $scalar, $offset, $len, $advice 2151=item IO::AIO::madvise $scalar, $offset, $len, $advice
1617 2152
1618Simply calls the C<posix_madvise> function (see its 2153Simply calls the C<posix_madvise> function (see its
1619manpage for details). The following advice constants are 2154manpage for details). The following advice constants are
1620avaiable: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>, 2155available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
1621C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, C<IO::AIO::MADV_DONTNEED>. 2156C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>,
2157C<IO::AIO::MADV_DONTNEED>.
2158
2159If C<$offset> is negative, counts from the end. If C<$length> is negative,
2160the remaining length of the C<$scalar> is used. If possible, C<$length>
2161will be reduced to fit into the C<$scalar>.
1622 2162
1623On systems that do not implement C<posix_madvise>, this function returns 2163On systems that do not implement C<posix_madvise>, this function returns
1624ENOSYS, otherwise the return value of C<posix_madvise>. 2164ENOSYS, otherwise the return value of C<posix_madvise>.
1625 2165
1626=item IO::AIO::mprotect $scalar, $offset, $len, $protect 2166=item IO::AIO::mprotect $scalar, $offset, $len, $protect
1627 2167
1628Simply calls the C<mprotect> function on the preferably AIO::mmap'ed 2168Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
1629$scalar (see its manpage for details). The following protect 2169$scalar (see its manpage for details). The following protect
1630constants are avaiable: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>, 2170constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
1631C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>. 2171C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
2172
2173If C<$offset> is negative, counts from the end. If C<$length> is negative,
2174the remaining length of the C<$scalar> is used. If possible, C<$length>
2175will be reduced to fit into the C<$scalar>.
1632 2176
1633On systems that do not implement C<mprotect>, this function returns 2177On systems that do not implement C<mprotect>, this function returns
1634ENOSYS, otherwise the return value of C<mprotect>. 2178ENOSYS, otherwise the return value of C<mprotect>.
1635 2179
1636=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset] 2180=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1637 2181
1638Memory-maps a file (or anonymous memory range) and attaches it to the 2182Memory-maps a file (or anonymous memory range) and attaches it to the
1639given C<$scalar>, which will act like a string scalar. 2183given C<$scalar>, which will act like a string scalar. Returns true on
2184success, and false otherwise.
1640 2185
2186The scalar must exist, but its contents do not matter - this means you
2187cannot use a nonexistant array or hash element. When in doubt, C<undef>
2188the scalar first.
2189
1641The only operations allowed on the scalar are C<substr>/C<vec> that don't 2190The only operations allowed on the mmapped scalar are C<substr>/C<vec>,
1642change the string length, and most read-only operations such as copying it 2191which don't change the string length, and most read-only operations such
1643or searching it with regexes and so on. 2192as copying it or searching it with regexes and so on.
1644 2193
1645Anything else is unsafe and will, at best, result in memory leaks. 2194Anything else is unsafe and will, at best, result in memory leaks.
1646 2195
1647The memory map associated with the C<$scalar> is automatically removed 2196The memory map associated with the C<$scalar> is automatically removed
1648when the C<$scalar> is destroyed, or when the C<IO::AIO::mmap> or 2197when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap>
1649C<IO::AIO::munmap> functions are called. 2198or C<IO::AIO::munmap> functions are called on it.
1650 2199
1651This calls the C<mmap>(2) function internally. See your system's manual 2200This calls the C<mmap>(2) function internally. See your system's manual
1652page for details on the C<$length>, C<$prot> and C<$flags> parameters. 2201page for details on the C<$length>, C<$prot> and C<$flags> parameters.
1653 2202
1654The C<$length> must be larger than zero and smaller than the actual 2203The C<$length> must be larger than zero and smaller than the actual
1655filesize. 2204filesize.
1656 2205
1657C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>, 2206C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
1658C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>, 2207C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
1659 2208
1660C<$flags> can be a combination of C<IO::AIO::MAP_SHARED> or 2209C<$flags> can be a combination of
1661C<IO::AIO::MAP_PRIVATE>, or a number of system-specific flags (when 2210C<IO::AIO::MAP_SHARED> or
1662not available, the are defined as 0): C<IO::AIO::MAP_ANONYMOUS> 2211C<IO::AIO::MAP_PRIVATE>,
2212or a number of system-specific flags (when not available, the are C<0>):
1663(which is set to C<MAP_ANON> if your system only provides this 2213C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant),
1664constant), C<IO::AIO::MAP_HUGETLB>, C<IO::AIO::MAP_LOCKED>, 2214C<IO::AIO::MAP_LOCKED>,
1665C<IO::AIO::MAP_NORESERVE>, C<IO::AIO::MAP_POPULATE> or 2215C<IO::AIO::MAP_NORESERVE>,
2216C<IO::AIO::MAP_POPULATE>,
1666C<IO::AIO::MAP_NONBLOCK> 2217C<IO::AIO::MAP_NONBLOCK>,
2218C<IO::AIO::MAP_FIXED>,
2219C<IO::AIO::MAP_GROWSDOWN>,
2220C<IO::AIO::MAP_32BIT>,
2221C<IO::AIO::MAP_HUGETLB> or
2222C<IO::AIO::MAP_STACK>.
1667 2223
1668If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed. 2224If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
1669 2225
1670C<$offset> is the offset from the start of the file - it generally must be 2226C<$offset> is the offset from the start of the file - it generally must be
1671a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>. 2227a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
1696 2252
1697Calls the C<munlockall> function. 2253Calls the C<munlockall> function.
1698 2254
1699On systems that do not implement C<munlockall>, this function returns 2255On systems that do not implement C<munlockall>, this function returns
1700ENOSYS, otherwise the return value of C<munlockall>. 2256ENOSYS, otherwise the return value of C<munlockall>.
2257
2258=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
2259
2260Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or
2261C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they
2262should be the file offset.
2263
2264C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might
2265silently corrupt the data in this case.
2266
2267The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>,
2268C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and
2269C<IO::AIO::SPLICE_F_GIFT>.
2270
2271See the C<splice(2)> manpage for details.
2272
2273=item IO::AIO::tee $r_fh, $w_fh, $length, $flags
2274
2275Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the
2276description for C<IO::AIO::splice> above for details.
2277
2278=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
2279
2280Attempts to query or change the pipe buffer size. Obviously works only
2281on pipes, and currently works only on GNU/Linux systems, and fails with
2282C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer
2283size on other systems, drop me a note.
2284
2285=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
2286
2287This is a direct interface to the Linux L<pipe2(2)> system call. If
2288C<$flags> is missing or C<0>, then this should be the same as a call to
2289perl's built-in C<pipe> function and create a new pipe, and works on
2290systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe
2291(..., 4096, O_BINARY)>.
2292
2293If C<$flags> is non-zero, it tries to invoke the pipe2 system call with
2294the given flags (Linux 2.6.27, glibc 2.9).
2295
2296On success, the read and write file handles are returned.
2297
2298On error, nothing will be returned. If the pipe2 syscall is missing and
2299C<$flags> is non-zero, fails with C<ENOSYS>.
2300
2301Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the
2302time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and
2303C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported.
1701 2304
1702=back 2305=back
1703 2306
1704=cut 2307=cut
1705 2308
1740 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => 2343 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
1741 \&IO::AIO::poll_cb); 2344 \&IO::AIO::poll_cb);
1742 2345
1743=head2 FORK BEHAVIOUR 2346=head2 FORK BEHAVIOUR
1744 2347
1745This module should do "the right thing" when the process using it forks: 2348Usage of pthreads in a program changes the semantics of fork
2349considerably. Specifically, only async-safe functions can be called after
2350fork. Perl doesn't know about this, so in general, you cannot call fork
2351with defined behaviour in perl if pthreads are involved. IO::AIO uses
2352pthreads, so this applies, but many other extensions and (for inexplicable
2353reasons) perl itself often is linked against pthreads, so this limitation
2354applies to quite a lot of perls.
1746 2355
1747Before the fork, IO::AIO enters a quiescent state where no requests 2356This module no longer tries to fight your OS, or POSIX. That means IO::AIO
1748can be added in other threads and no results will be processed. After 2357only works in the process that loaded it. Forking is fully supported, but
1749the fork the parent simply leaves the quiescent state and continues 2358using IO::AIO in the child is not.
1750request/result processing, while the child frees the request/result queue
1751(so that the requests started before the fork will only be handled in the
1752parent). Threads will be started on demand until the limit set in the
1753parent process has been reached again.
1754 2359
1755In short: the parent will, after a short pause, continue as if fork had 2360You might get around by not I<using> IO::AIO before (or after)
1756not been called, while the child will act as if IO::AIO has not been used 2361forking. You could also try to call the L<IO::AIO::reinit> function in the
1757yet. 2362child:
2363
2364=over 4
2365
2366=item IO::AIO::reinit
2367
2368Abandons all current requests and I/O threads and simply reinitialises all
2369data structures. This is not an operation supported by any standards, but
2370happens to work on GNU/Linux and some newer BSD systems.
2371
2372The only reasonable use for this function is to call it after forking, if
2373C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
2374the process will result in undefined behaviour. Calling it at any time
2375will also result in any undefined (by POSIX) behaviour.
2376
2377=back
1758 2378
1759=head2 MEMORY USAGE 2379=head2 MEMORY USAGE
1760 2380
1761Per-request usage: 2381Per-request usage:
1762 2382

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines