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 1.111 by root, Mon Aug 13 12:08:13 2007 UTC vs.
Revision 1.302 by root, Wed Apr 3 03:03:53 2019 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
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
… …
26 $req->cancel; # cancel request if still in queue 26 $req->cancel; # cancel request if still in queue
27 27
28 my $grp = aio_group sub { print "all stats done\n" }; 28 my $grp = aio_group sub { print "all stats done\n" };
29 add $grp aio_stat "..." for ...; 29 add $grp aio_stat "..." for ...;
30 30
31 # AnyEvent integration
32 open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
33 my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
34
35 # Event integration
36 Event->io (fd => IO::AIO::poll_fileno,
37 poll => 'r',
38 cb => \&IO::AIO::poll_cb);
39
40 # Glib/Gtk2 integration
41 add_watch Glib::IO IO::AIO::poll_fileno,
42 in => sub { IO::AIO::poll_cb; 1 };
43
44 # Tk integration
45 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
46 readable => \&IO::AIO::poll_cb);
47
48 # Danga::Socket integration
49 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
50 \&IO::AIO::poll_cb);
51
52=head1 DESCRIPTION 31=head1 DESCRIPTION
53 32
54This module implements asynchronous I/O using whatever means your 33This module implements asynchronous I/O using whatever means your
55operating system supports. 34operating system supports. It is implemented as an interface to C<libeio>
35(L<http://software.schmorp.de/pkg/libeio.html>).
56 36
57Asynchronous means that operations that can normally block your program 37Asynchronous means that operations that can normally block your program
58(e.g. reading from disk) will be done asynchronously: the operation 38(e.g. reading from disk) will be done asynchronously: the operation
59will still block, but you can do something else in the meantime. This 39will still block, but you can do something else in the meantime. This
60is extremely useful for programs that need to stay interactive even 40is extremely useful for programs that need to stay interactive even
… …
64on a RAID volume or over NFS when you do a number of stat operations 44on a RAID volume or over NFS when you do a number of stat operations
65concurrently. 45concurrently.
66 46
67While most of this works on all types of file descriptors (for 47While most of this works on all types of file descriptors (for
68example sockets), using these functions on file descriptors that 48example sockets), using these functions on file descriptors that
69support nonblocking operation (again, sockets, pipes etc.) is very 49support nonblocking operation (again, sockets, pipes etc.) is
70inefficient. Use an event loop for that (such as the L<Event|Event> 50very inefficient. Use an event loop for that (such as the L<EV>
71module): IO::AIO will naturally fit into such an event loop itself. 51module): IO::AIO will naturally fit into such an event loop itself.
72 52
73In this version, a number of threads are started that execute your 53In this version, a number of threads are started that execute your
74requests and signal their completion. You don't need thread support 54requests and signal their completion. You don't need thread support
75in perl, and the threads created by this module will not be visible 55in perl, and the threads created by this module will not be visible
… …
78not 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
79files currently, for example), and they would only support aio_read and 59files currently, for example), and they would only support aio_read and
80aio_write, so the remaining functionality would have to be implemented 60aio_write, so the remaining functionality would have to be implemented
81using threads anyway. 61using threads anyway.
82 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
83Although the module will work in the presence of other (Perl-) threads, 67Although the module will work in the presence of other (Perl-) threads,
84it is currently not reentrant in any way, so use appropriate locking 68it is currently not reentrant in any way, so use appropriate locking
85yourself, 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
86call C<poll_cb> (or other C<aio_> functions) recursively. 70call C<poll_cb> (or other C<aio_> functions) recursively.
87 71
88=head2 EXAMPLE 72=head2 EXAMPLE
89 73
90This is a simple example that uses the Event module and loads 74This is a simple example that uses the EV module and loads
91F</etc/passwd> asynchronously: 75F</etc/passwd> asynchronously:
92 76
93 use Fcntl;
94 use Event; 77 use EV;
95 use IO::AIO; 78 use IO::AIO;
96 79
97 # register the IO::AIO callback with Event 80 # register the IO::AIO callback with EV
98 Event->io (fd => IO::AIO::poll_fileno, 81 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
99 poll => 'r',
100 cb => \&IO::AIO::poll_cb);
101 82
102 # queue the request to open /etc/passwd 83 # queue the request to open /etc/passwd
103 aio_open "/etc/passwd", O_RDONLY, 0, sub { 84 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
104 my $fh = shift 85 my $fh = shift
105 or die "error while opening: $!"; 86 or die "error while opening: $!";
106 87
107 # stat'ing filehandles is generally non-blocking 88 # stat'ing filehandles is generally non-blocking
108 my $size = -s $fh; 89 my $size = -s $fh;
… …
117 98
118 # file contents now in $contents 99 # file contents now in $contents
119 print $contents; 100 print $contents;
120 101
121 # exit event loop and program 102 # exit event loop and program
122 Event::unloop; 103 EV::break;
123 }; 104 };
124 }; 105 };
125 106
126 # possibly queue up other requests, or open GUI windows, 107 # possibly queue up other requests, or open GUI windows,
127 # check for sockets etc. etc. 108 # check for sockets etc. etc.
128 109
129 # process events as long as there are some: 110 # process events as long as there are some:
130 Event::loop; 111 EV::run;
131 112
132=head1 REQUEST ANATOMY AND LIFETIME 113=head1 REQUEST ANATOMY AND LIFETIME
133 114
134Every 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
135directly visible to Perl. 116directly visible to Perl.
… …
183 164
184=cut 165=cut
185 166
186package IO::AIO; 167package IO::AIO;
187 168
188no warnings; 169use Carp ();
189use strict 'vars'; 170
171use common::sense;
190 172
191use base 'Exporter'; 173use base 'Exporter';
192 174
193BEGIN { 175BEGIN {
194 our $VERSION = '2.4'; 176 our $VERSION = 4.72;
195 177
196 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat 178 our @AIO_REQ = qw(aio_sendfile aio_seek aio_read aio_write aio_open aio_close
197 aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink 179 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx
198 aio_readlink aio_fsync aio_fdatasync aio_readahead aio_rename aio_link 180 aio_scandir aio_symlink aio_readlink aio_realpath aio_fcntl aio_ioctl
181 aio_sync aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range
182 aio_pathsync aio_readahead aio_fiemap aio_allocate
183 aio_rename aio_rename2 aio_link aio_move aio_copy aio_group
199 aio_move aio_copy aio_group aio_nop aio_mknod aio_load aio_rmtree aio_mkdir 184 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown
200 aio_chown aio_chmod aio_utime aio_truncate); 185 aio_chmod aio_utime aio_truncate
186 aio_msync aio_mtouch aio_mlock aio_mlockall
187 aio_statvfs
188 aio_slurp
189 aio_wd);
190
201 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice aio_block)); 191 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
202 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 192 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
203 min_parallel max_parallel max_idle 193 min_parallel max_parallel max_idle idle_timeout
204 nreqs nready npending nthreads 194 nreqs nready npending nthreads
205 max_poll_time max_poll_reqs); 195 max_poll_time max_poll_reqs
196 sendfile fadvise madvise
197 mmap munmap mremap munlock munlockall);
198
199 push @AIO_REQ, qw(aio_busy); # not exported
206 200
207 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 201 @IO::AIO::GRP::ISA = 'IO::AIO::REQ';
208 202
209 require XSLoader; 203 require XSLoader;
210 XSLoader::load ("IO::AIO", $VERSION); 204 XSLoader::load ("IO::AIO", $VERSION);
211} 205}
212 206
213=head1 FUNCTIONS 207=head1 FUNCTIONS
214 208
215=head2 AIO REQUEST FUNCTIONS 209=head2 QUICK OVERVIEW
210
211This section simply lists the prototypes most of the functions for
212quick reference. See the following sections for function-by-function
213documentation.
214
215 aio_wd $pathname, $callback->($wd)
216 aio_open $pathname, $flags, $mode, $callback->($fh)
217 aio_close $fh, $callback->($status)
218 aio_seek $fh,$offset,$whence, $callback->($offs)
219 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
220 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
221 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
222 aio_readahead $fh,$offset,$length, $callback->($retval)
223 aio_stat $fh_or_path, $callback->($status)
224 aio_lstat $fh, $callback->($status)
225 aio_statvfs $fh_or_path, $callback->($statvfs)
226 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
227 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
228 aio_chmod $fh_or_path, $mode, $callback->($status)
229 aio_truncate $fh_or_path, $offset, $callback->($status)
230 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
231 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
232 aio_unlink $pathname, $callback->($status)
233 aio_mknod $pathname, $mode, $dev, $callback->($status)
234 aio_link $srcpath, $dstpath, $callback->($status)
235 aio_symlink $srcpath, $dstpath, $callback->($status)
236 aio_readlink $pathname, $callback->($link)
237 aio_realpath $pathname, $callback->($path)
238 aio_rename $srcpath, $dstpath, $callback->($status)
239 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
240 aio_mkdir $pathname, $mode, $callback->($status)
241 aio_rmdir $pathname, $callback->($status)
242 aio_readdir $pathname, $callback->($entries)
243 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
244 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
245 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
246 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
247 aio_load $pathname, $data, $callback->($status)
248 aio_copy $srcpath, $dstpath, $callback->($status)
249 aio_move $srcpath, $dstpath, $callback->($status)
250 aio_rmtree $pathname, $callback->($status)
251 aio_fcntl $fh, $cmd, $arg, $callback->($status)
252 aio_ioctl $fh, $request, $buf, $callback->($status)
253 aio_sync $callback->($status)
254 aio_syncfs $fh, $callback->($status)
255 aio_fsync $fh, $callback->($status)
256 aio_fdatasync $fh, $callback->($status)
257 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
258 aio_pathsync $pathname, $callback->($status)
259 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
260 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
261 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
262 aio_mlockall $flags, $callback->($status)
263 aio_group $callback->(...)
264 aio_nop $callback->()
265
266 $prev_pri = aioreq_pri [$pri]
267 aioreq_nice $pri_adjust
268
269 IO::AIO::poll_wait
270 IO::AIO::poll_cb
271 IO::AIO::poll
272 IO::AIO::flush
273 IO::AIO::max_poll_reqs $nreqs
274 IO::AIO::max_poll_time $seconds
275 IO::AIO::min_parallel $nthreads
276 IO::AIO::max_parallel $nthreads
277 IO::AIO::max_idle $nthreads
278 IO::AIO::idle_timeout $seconds
279 IO::AIO::max_outstanding $maxreqs
280 IO::AIO::nreqs
281 IO::AIO::nready
282 IO::AIO::npending
283 IO::AIO::reinit
284
285 $nfd = IO::AIO::get_fdlimit [EXPERIMENTAL]
286 IO::AIO::min_fdlimit $nfd [EXPERIMENTAL]
287
288 IO::AIO::sendfile $ofh, $ifh, $offset, $count
289 IO::AIO::fadvise $fh, $offset, $len, $advice
290
291 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
292 IO::AIO::munmap $scalar
293 IO::AIO::mremap $scalar, $new_length, $flags[, $new_address]
294 IO::AIO::madvise $scalar, $offset, $length, $advice
295 IO::AIO::mprotect $scalar, $offset, $length, $protect
296 IO::AIO::munlock $scalar, $offset = 0, $length = undef
297 IO::AIO::munlockall
298
299 # stat extensions
300 $counter = IO::AIO::st_gen
301 $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime
302 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
303 $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec
304 $seconds = IO::AIO::st_btimesec
305 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
306
307 # very much unportable syscalls
308 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
309 IO::AIO::tee $r_fh, $w_fh, $length, $flags
310 $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
311 ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
312 $fh = IO::AIO::memfd_create $pathname[, $flags]
313 $fh = IO::AIO::eventfd [$initval, [$flags]]
314 $fh = IO::AIO::timerfd_create $clockid[, $flags]
315 ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
316 ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
317
318=head2 API NOTES
216 319
217All the C<aio_*> calls are more or less thin wrappers around the syscall 320All the C<aio_*> calls are more or less thin wrappers around the syscall
218with the same name (sans C<aio_>). The arguments are similar or identical, 321with the same name (sans C<aio_>). The arguments are similar or identical,
219and they all accept an additional (and optional) C<$callback> argument 322and they all accept an additional (and optional) C<$callback> argument
220which must be a code reference. This code reference will get called with 323which must be a code reference. This code reference will be called after
221the syscall return code (e.g. most syscalls return C<-1> on error, unlike 324the syscall has been executed in an asynchronous fashion. The results
222perl, which usually delivers "false") as it's sole argument when the given 325of the request will be passed as arguments to the callback (and, if an
223syscall has been executed asynchronously. 326error occured, in C<$!>) - for most requests the syscall return code (e.g.
327most syscalls return C<-1> on error, unlike perl, which usually delivers
328"false").
329
330Some requests (such as C<aio_readdir>) pass the actual results and
331communicate failures by passing C<undef>.
224 332
225All functions expecting a filehandle keep a copy of the filehandle 333All functions expecting a filehandle keep a copy of the filehandle
226internally until the request has finished. 334internally until the request has finished.
227 335
228All functions return request objects of type L<IO::AIO::REQ> that allow 336All functions return request objects of type L<IO::AIO::REQ> that allow
229further manipulation of those requests while they are in-flight. 337further manipulation of those requests while they are in-flight.
230 338
231The pathnames you pass to these routines I<must> be absolute and 339The pathnames you pass to these routines I<should> be absolute. The
232encoded as octets. The reason for the former is that at the time the 340reason for this is that at the time the request is being executed, the
233request is being executed, the current working directory could have 341current working directory could have changed. Alternatively, you can
234changed. Alternatively, you can make sure that you never change the 342make sure that you never change the current working directory anywhere
235current working directory anywhere in the program and then use relative 343in the program and then use relative paths. You can also take advantage
236paths. 344of IO::AIOs working directory abstraction, that lets you specify paths
345relative to some previously-opened "working directory object" - see the
346description of the C<IO::AIO::WD> class later in this document.
237 347
238To encode pathnames as octets, either make sure you either: a) always pass 348To encode pathnames as octets, either make sure you either: a) always pass
239in filenames you got from outside (command line, readdir etc.) without 349in filenames you got from outside (command line, readdir etc.) without
240tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode 350tinkering, b) are in your native filesystem encoding, c) use the Encode
241your pathnames to the locale (or other) encoding in effect in the user 351module and encode your pathnames to the locale (or other) encoding in
242environment, d) use Glib::filename_from_unicode on unicode filenames or e) 352effect in the user environment, d) use Glib::filename_from_unicode on
243use something else to ensure your scalar has the correct contents. 353unicode filenames or e) use something else to ensure your scalar has the
354correct contents.
244 355
245This works, btw. independent of the internal UTF-8 bit, which IO::AIO 356This works, btw. independent of the internal UTF-8 bit, which IO::AIO
246handles correctly wether it is set or not. 357handles correctly whether it is set or not.
358
359=head2 AIO REQUEST FUNCTIONS
247 360
248=over 4 361=over 4
249 362
250=item $prev_pri = aioreq_pri [$pri] 363=item $prev_pri = aioreq_pri [$pri]
251 364
… …
281 394
282 395
283=item aio_open $pathname, $flags, $mode, $callback->($fh) 396=item aio_open $pathname, $flags, $mode, $callback->($fh)
284 397
285Asynchronously open or create a file and call the callback with a newly 398Asynchronously open or create a file and call the callback with a newly
286created filehandle for the file. 399created filehandle for the file (or C<undef> in case of an error).
287 400
288The pathname passed to C<aio_open> must be absolute. See API NOTES, above, 401The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
289for an explanation. 402for an explanation.
290 403
291The C<$flags> argument is a bitmask. See the C<Fcntl> module for a 404The C<$flags> argument is a bitmask. See the C<Fcntl> module for a
… …
298by the umask in effect then the request is being executed, so better never 411by the umask in effect then the request is being executed, so better never
299change the umask. 412change the umask.
300 413
301Example: 414Example:
302 415
303 aio_open "/etc/passwd", O_RDONLY, 0, sub { 416 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
304 if ($_[0]) { 417 if ($_[0]) {
305 print "open successful, fh is $_[0]\n"; 418 print "open successful, fh is $_[0]\n";
306 ... 419 ...
307 } else { 420 } else {
308 die "open failed: $!\n"; 421 die "open failed: $!\n";
309 } 422 }
310 }; 423 };
311 424
425In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>,
426C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the
427following POSIX and non-POSIX constants are available (missing ones on
428your system are, as usual, C<0>):
429
430C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>,
431C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>,
432C<O_RSYNC>, C<O_SYNC>, C<O_PATH>, C<O_TMPFILE>, C<O_TTY_INIT> and C<O_ACCMODE>.
433
312 434
313=item aio_close $fh, $callback->($status) 435=item aio_close $fh, $callback->($status)
314 436
315Asynchronously close a file and call the callback with the result 437Asynchronously close a file and call the callback with the result
316code. I<WARNING:> although accepted, you should not pass in a perl 438code.
317filehandle here, as perl will likely close the file descriptor another
318time when the filehandle is destroyed. Normally, you can safely call perls
319C<close> or just let filehandles go out of scope.
320 439
321This is supposed to be a bug in the API, so that might change. It's 440Unfortunately, you can't do this to perl. Perl I<insists> very strongly on
322therefore best to avoid this function. 441closing the file descriptor associated with the filehandle itself.
323 442
443Therefore, C<aio_close> will not close the filehandle - instead it will
444use dup2 to overwrite the file descriptor with the write-end of a pipe
445(the pipe fd will be created on demand and will be cached).
446
447Or in other words: the file descriptor will be closed, but it will not be
448free for reuse until the perl filehandle is closed.
449
450=cut
451
452=item aio_seek $fh, $offset, $whence, $callback->($offs)
453
454Seeks the filehandle to the new C<$offset>, similarly to perl's
455C<sysseek>. The C<$whence> can use the traditional values (C<0> for
456C<IO::AIO::SEEK_SET>, C<1> for C<IO::AIO::SEEK_CUR> or C<2> for
457C<IO::AIO::SEEK_END>).
458
459The resulting absolute offset will be passed to the callback, or C<-1> in
460case of an error.
461
462In theory, the C<$whence> constants could be different than the
463corresponding values from L<Fcntl>, but perl guarantees they are the same,
464so don't panic.
465
466As a GNU/Linux (and maybe Solaris) extension, also the constants
467C<IO::AIO::SEEK_DATA> and C<IO::AIO::SEEK_HOLE> are available, if they
468could be found. No guarantees about suitability for use in C<aio_seek> or
469Perl's C<sysseek> can be made though, although I would naively assume they
470"just work".
324 471
325=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 472=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
326 473
327=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 474=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
328 475
329Reads or writes C<$length> bytes from the specified C<$fh> and C<$offset> 476Reads or writes C<$length> bytes from or to the specified C<$fh> and
330into the scalar given by C<$data> and offset C<$dataoffset> and calls the 477C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> and
331callback without the actual number of bytes read (or -1 on error, just 478calls the callback with the actual number of bytes transferred (or -1 on
332like the syscall). 479error, just like the syscall).
333 480
481C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to
482offset plus the actual number of bytes read.
483
334If C<$offset> is undefined, then the current file offset will be used (and 484If C<$offset> is undefined, then the current file descriptor offset will
335updated), otherwise the file offset will not be changed by these calls. 485be used (and updated), otherwise the file descriptor offset will not be
486changed by these calls.
336 487
337If C<$length> is undefined in C<aio_write>, use the remaining length of C<$data>. 488If C<$length> is undefined in C<aio_write>, use the remaining length of
489C<$data>.
338 490
339If C<$dataoffset> is less than zero, it will be counted from the end of 491If C<$dataoffset> is less than zero, it will be counted from the end of
340C<$data>. 492C<$data>.
341 493
342The C<$data> scalar I<MUST NOT> be modified in any way while the request 494The C<$data> scalar I<MUST NOT> be modified in any way while the request
… …
356 508
357Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 509Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts
358reading at byte offset C<$in_offset>, and starts writing at the current 510reading at byte offset C<$in_offset>, and starts writing at the current
359file offset of C<$out_fh>. Because of that, it is not safe to issue more 511file offset of C<$out_fh>. Because of that, it is not safe to issue more
360than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each 512than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each
361other. 513other. The same C<$in_fh> works fine though, as this function does not
514move or use the file offset of C<$in_fh>.
362 515
516Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than
517are written, and there is no way to find out how many more bytes have been
518read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the
519number of bytes written to C<$out_fh>. Only if the result value equals
520C<$length> one can assume that C<$length> bytes have been read.
521
522Unlike with other C<aio_> functions, it makes a lot of sense to use
523C<aio_sendfile> on non-blocking sockets, as long as one end (typically
524the C<$in_fh>) is a file - the file I/O will then be asynchronous, while
525the socket I/O will be non-blocking. Note, however, that you can run
526into a trap where C<aio_sendfile> reads some data with readahead, then
527fails to write all data, and when the socket is ready the next time, the
528data in the cache is already lost, forcing C<aio_sendfile> to again hit
529the disk. Explicit C<aio_read> + C<aio_write> let's you better control
530resource usage.
531
363This call tries to make use of a native C<sendfile> syscall to provide 532This call tries to make use of a native C<sendfile>-like syscall to
364zero-copy operation. For this to work, C<$out_fh> should refer to a 533provide zero-copy operation. For this to work, C<$out_fh> should refer to
365socket, and C<$in_fh> should refer to mmap'able file. 534a socket, and C<$in_fh> should refer to an mmap'able file.
366 535
367If the native sendfile call fails or is not implemented, it will be 536If a native sendfile cannot be found or it fails with C<ENOSYS>,
368emulated, so you can call C<aio_sendfile> on any type of filehandle 537C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or
538C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any
369regardless of the limitations of the operating system. 539type of filehandle regardless of the limitations of the operating system.
370 540
371Please note, however, that C<aio_sendfile> can read more bytes from 541As native sendfile syscalls (as practically any non-POSIX interface hacked
372C<$in_fh> than are written, and there is no way to find out how many 542together in a hurry to improve benchmark numbers) tend to be rather buggy
373bytes have been read from C<aio_sendfile> alone, as C<aio_sendfile> only 543on many systems, this implementation tries to work around some known bugs
374provides the number of bytes written to C<$out_fh>. Only if the result 544in Linux and FreeBSD kernels (probably others, too), but that might fail,
375value equals C<$length> one can assume that C<$length> bytes have been 545so you really really should check the return value of C<aio_sendfile> -
376read. 546fewer bytes than expected might have been transferred.
377 547
378 548
379=item aio_readahead $fh,$offset,$length, $callback->($retval) 549=item aio_readahead $fh,$offset,$length, $callback->($retval)
380 550
381C<aio_readahead> populates the page cache with data from a file so that 551C<aio_readahead> populates the page cache with data from a file so that
… …
385whole pages, so that offset is effectively rounded down to a page boundary 555whole pages, so that offset is effectively rounded down to a page boundary
386and bytes are read up to the next page boundary greater than or equal to 556and bytes are read up to the next page boundary greater than or equal to
387(off-set+length). C<aio_readahead> does not read beyond the end of the 557(off-set+length). C<aio_readahead> does not read beyond the end of the
388file. The current file offset of the file is left unchanged. 558file. The current file offset of the file is left unchanged.
389 559
390If that syscall doesn't exist (likely if your OS isn't Linux) it will be 560If that syscall doesn't exist (likely if your kernel isn't Linux) it will
391emulated by simply reading the data, which would have a similar effect. 561be emulated by simply reading the data, which would have a similar effect.
392 562
393 563
394=item aio_stat $fh_or_path, $callback->($status) 564=item aio_stat $fh_or_path, $callback->($status)
395 565
396=item aio_lstat $fh, $callback->($status) 566=item aio_lstat $fh, $callback->($status)
397 567
398Works like perl's C<stat> or C<lstat> in void context. The callback will 568Works almost exactly like perl's C<stat> or C<lstat> in void context. The
399be called after the stat and the results will be available using C<stat _> 569callback will be called after the stat and the results will be available
400or C<-s _> etc... 570using C<stat _> or C<-s _> and other tests (with the exception of C<-B>
571and C<-T>).
401 572
402The pathname passed to C<aio_stat> must be absolute. See API NOTES, above, 573The pathname passed to C<aio_stat> must be absolute. See API NOTES, above,
403for an explanation. 574for an explanation.
404 575
405Currently, the stats are always 64-bit-stats, i.e. instead of returning an 576Currently, the stats are always 64-bit-stats, i.e. instead of returning an
406error when stat'ing a large file, the results will be silently truncated 577error when stat'ing a large file, the results will be silently truncated
407unless perl itself is compiled with large file support. 578unless perl itself is compiled with large file support.
579
580To help interpret the mode and dev/rdev stat values, IO::AIO offers the
581following constants and functions (if not implemented, the constants will
582be C<0> and the functions will either C<croak> or fall back on traditional
583behaviour).
584
585C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>,
586C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>,
587C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>.
588
589To access higher resolution stat timestamps, see L<SUBSECOND STAT TIME
590ACCESS>.
408 591
409Example: Print the length of F</etc/passwd>: 592Example: Print the length of F</etc/passwd>:
410 593
411 aio_stat "/etc/passwd", sub { 594 aio_stat "/etc/passwd", sub {
412 $_[0] and die "stat failed: $!"; 595 $_[0] and die "stat failed: $!";
413 print "size is ", -s _, "\n"; 596 print "size is ", -s _, "\n";
414 }; 597 };
415 598
416 599
600=item aio_statvfs $fh_or_path, $callback->($statvfs)
601
602Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on
603whether a file handle or path was passed.
604
605On success, the callback is passed a hash reference with the following
606members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>,
607C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef>
608is passed.
609
610The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and
611C<ST_NOSUID>.
612
613The following non-POSIX IO::AIO::ST_* flag masks are defined to
614their correct value when available, or to C<0> on systems that do
615not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>,
616C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>,
617C<ST_NODIRATIME> and C<ST_RELATIME>.
618
619Example: stat C</wd> and dump out the data if successful.
620
621 aio_statvfs "/wd", sub {
622 my $f = $_[0]
623 or die "statvfs: $!";
624
625 use Data::Dumper;
626 say Dumper $f;
627 };
628
629 # result:
630 {
631 bsize => 1024,
632 bfree => 4333064312,
633 blocks => 10253828096,
634 files => 2050765568,
635 flag => 4096,
636 favail => 2042092649,
637 bavail => 4333064312,
638 ffree => 2042092649,
639 namemax => 255,
640 frsize => 1024,
641 fsid => 1810
642 }
643
417=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 644=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
418 645
419Works like perl's C<utime> function (including the special case of $atime 646Works like perl's C<utime> function (including the special case of $atime
420and $mtime being undef). Fractional times are supported if the underlying 647and $mtime being undef). Fractional times are supported if the underlying
421syscalls support them. 648syscalls support them.
422 649
423When called with a pathname, uses utimes(2) if available, otherwise 650When called with a pathname, uses utimensat(2) or utimes(2) if available,
424utime(2). If called on a file descriptor, uses futimes(2) if available, 651otherwise utime(2). If called on a file descriptor, uses futimens(2)
425otherwise returns ENOSYS, so this is not portable. 652or futimes(2) if available, otherwise returns ENOSYS, so this is not
653portable.
426 654
427Examples: 655Examples:
428 656
429 # set atime and mtime to current time (basically touch(1)): 657 # set atime and mtime to current time (basically touch(1)):
430 aio_utime "path", undef, undef; 658 aio_utime "path", undef, undef;
… …
448=item aio_truncate $fh_or_path, $offset, $callback->($status) 676=item aio_truncate $fh_or_path, $offset, $callback->($status)
449 677
450Works like truncate(2) or ftruncate(2). 678Works like truncate(2) or ftruncate(2).
451 679
452 680
681=item aio_allocate $fh, $mode, $offset, $len, $callback->($status)
682
683Allocates or frees disk space according to the C<$mode> argument. See the
684linux C<fallocate> documentation for details.
685
686C<$mode> is usually C<0> or C<IO::AIO::FALLOC_FL_KEEP_SIZE> to allocate
687space, or C<IO::AIO::FALLOC_FL_PUNCH_HOLE | IO::AIO::FALLOC_FL_KEEP_SIZE>,
688to deallocate a file range.
689
690IO::AIO also supports C<FALLOC_FL_COLLAPSE_RANGE>, to remove a range
691(without leaving a hole), C<FALLOC_FL_ZERO_RANGE>, to zero a range,
692C<FALLOC_FL_INSERT_RANGE> to insert a range and C<FALLOC_FL_UNSHARE_RANGE>
693to unshare shared blocks (see your L<fallocate(2)> manpage).
694
695The file system block size used by C<fallocate> is presumably the
696C<f_bsize> returned by C<statvfs>, but different filesystems and filetypes
697can dictate other limitations.
698
699If C<fallocate> isn't available or cannot be emulated (currently no
700emulation will be attempted), passes C<-1> and sets C<$!> to C<ENOSYS>.
701
702
453=item aio_chmod $fh_or_path, $mode, $callback->($status) 703=item aio_chmod $fh_or_path, $mode, $callback->($status)
454 704
455Works like perl's C<chmod> function. 705Works like perl's C<chmod> function.
456 706
457 707
… …
459 709
460Asynchronously unlink (delete) a file and call the callback with the 710Asynchronously unlink (delete) a file and call the callback with the
461result code. 711result code.
462 712
463 713
464=item aio_mknod $path, $mode, $dev, $callback->($status) 714=item aio_mknod $pathname, $mode, $dev, $callback->($status)
465 715
466[EXPERIMENTAL] 716[EXPERIMENTAL]
467 717
468Asynchronously create a device node (or fifo). See mknod(2). 718Asynchronously create a device node (or fifo). See mknod(2).
469 719
470The only (POSIX-) portable way of calling this function is: 720The only (POSIX-) portable way of calling this function is:
471 721
472 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 722 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
473 723
724See C<aio_stat> for info about some potentially helpful extra constants
725and functions.
474 726
475=item aio_link $srcpath, $dstpath, $callback->($status) 727=item aio_link $srcpath, $dstpath, $callback->($status)
476 728
477Asynchronously create a new link to the existing object at C<$srcpath> at 729Asynchronously create a new link to the existing object at C<$srcpath> at
478the path C<$dstpath> and call the callback with the result code. 730the path C<$dstpath> and call the callback with the result code.
… …
482 734
483Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 735Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
484the path C<$dstpath> and call the callback with the result code. 736the path C<$dstpath> and call the callback with the result code.
485 737
486 738
487=item aio_readlink $path, $callback->($link) 739=item aio_readlink $pathname, $callback->($link)
488 740
489Asynchronously read the symlink specified by C<$path> and pass it to 741Asynchronously read the symlink specified by C<$path> and pass it to
490the callback. If an error occurs, nothing or undef gets passed to the 742the callback. If an error occurs, nothing or undef gets passed to the
491callback. 743callback.
492 744
493 745
746=item aio_realpath $pathname, $callback->($path)
747
748Asynchronously make the path absolute and resolve any symlinks in
749C<$path>. The resulting path only consists of directories (same as
750L<Cwd::realpath>).
751
752This request can be used to get the absolute path of the current working
753directory by passing it a path of F<.> (a single dot).
754
755
494=item aio_rename $srcpath, $dstpath, $callback->($status) 756=item aio_rename $srcpath, $dstpath, $callback->($status)
495 757
496Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 758Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
497rename(2) and call the callback with the result code. 759rename(2) and call the callback with the result code.
760
761On systems that support the AIO::WD working directory abstraction
762natively, the case C<[$wd, "."]> as C<$srcpath> is specialcased - instead
763of failing, C<rename> is called on the absolute path of C<$wd>.
764
765
766=item aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
767
768Basically a version of C<aio_rename> with an additional C<$flags>
769argument. Calling this with C<$flags=0> is the same as calling
770C<aio_rename>.
771
772Non-zero flags are currently only supported on GNU/Linux systems that
773support renameat2. Other systems fail with C<ENOSYS> in this case.
774
775The following constants are available (missing ones are, as usual C<0>),
776see renameat2(2) for details:
777
778C<IO::AIO::RENAME_NOREPLACE>, C<IO::AIO::RENAME_EXCHANGE>
779and C<IO::AIO::RENAME_WHITEOUT>.
498 780
499 781
500=item aio_mkdir $pathname, $mode, $callback->($status) 782=item aio_mkdir $pathname, $mode, $callback->($status)
501 783
502Asynchronously mkdir (create) a directory and call the callback with 784Asynchronously mkdir (create) a directory and call the callback with
… …
507=item aio_rmdir $pathname, $callback->($status) 789=item aio_rmdir $pathname, $callback->($status)
508 790
509Asynchronously rmdir (delete) a directory and call the callback with the 791Asynchronously rmdir (delete) a directory and call the callback with the
510result code. 792result code.
511 793
794On systems that support the AIO::WD working directory abstraction
795natively, the case C<[$wd, "."]> is specialcased - instead of failing,
796C<rmdir> is called on the absolute path of C<$wd>.
797
512 798
513=item aio_readdir $pathname, $callback->($entries) 799=item aio_readdir $pathname, $callback->($entries)
514 800
515Unlike the POSIX call of the same name, C<aio_readdir> reads an entire 801Unlike the POSIX call of the same name, C<aio_readdir> reads an entire
516directory (i.e. opendir + readdir + closedir). The entries will not be 802directory (i.e. opendir + readdir + closedir). The entries will not be
517sorted, and will B<NOT> include the C<.> and C<..> entries. 803sorted, and will B<NOT> include the C<.> and C<..> entries.
518 804
519The callback a single argument which is either C<undef> or an array-ref 805The callback is passed a single argument which is either C<undef> or an
520with the filenames. 806array-ref with the filenames.
521 807
522 808
809=item aio_readdirx $pathname, $flags, $callback->($entries, $flags)
810
811Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to
812tune behaviour and output format. In case of an error, C<$entries> will be
813C<undef>.
814
815The flags are a combination of the following constants, ORed together (the
816flags will also be passed to the callback, possibly modified):
817
818=over 4
819
820=item IO::AIO::READDIR_DENTS
821
822Normally the callback gets an arrayref consisting of names only (as
823with C<aio_readdir>). If this flag is set, then the callback gets an
824arrayref with C<[$name, $type, $inode]> arrayrefs, each describing a
825single directory entry in more detail:
826
827C<$name> is the name of the entry.
828
829C<$type> is one of the C<IO::AIO::DT_xxx> constants:
830
831C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>,
832C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>,
833C<IO::AIO::DT_WHT>.
834
835C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need
836to know, you have to run stat yourself. Also, for speed/memory reasons,
837the C<$type> scalars are read-only: you must not modify them.
838
839C<$inode> is the inode number (which might not be exact on systems with 64
840bit inode numbers and 32 bit perls). This field has unspecified content on
841systems that do not deliver the inode information.
842
843=item IO::AIO::READDIR_DIRS_FIRST
844
845When this flag is set, then the names will be returned in an order where
846likely directories come first, in optimal stat order. This is useful when
847you need to quickly find directories, or you want to find all directories
848while avoiding to stat() each entry.
849
850If the system returns type information in readdir, then this is used
851to find directories directly. Otherwise, likely directories are names
852beginning with ".", or otherwise names with no dots, of which names with
853short names are tried first.
854
855=item IO::AIO::READDIR_STAT_ORDER
856
857When this flag is set, then the names will be returned in an order
858suitable for stat()'ing each one. That is, when you plan to stat() most or
859all files in the given directory, then the returned order will likely be
860faster.
861
862If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified,
863then the likely dirs come first, resulting in a less optimal stat order
864for stat'ing all entries, but likely a more optimal order for finding
865subdirectories.
866
867=item IO::AIO::READDIR_FOUND_UNKNOWN
868
869This flag should not be set when calling C<aio_readdirx>. Instead, it
870is being set by C<aio_readdirx>, when any of the C<$type>'s found were
871C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all
872C<$type>'s are known, which can be used to speed up some algorithms.
873
874=back
875
876
877=item aio_slurp $pathname, $offset, $length, $data, $callback->($status)
878
879Opens, reads and closes the given file. The data is put into C<$data>,
880which is resized as required.
881
882If C<$offset> is negative, then it is counted from the end of the file.
883
884If C<$length> is zero, then the remaining length of the file is
885used. Also, in this case, the same limitations to modifying C<$data> apply
886as when IO::AIO::mmap is used, i.e. it must only be modified in-place
887with C<substr>. If the size of the file is known, specifying a non-zero
888C<$length> results in a performance advantage.
889
890This request is similar to the older C<aio_load> request, but since it is
891a single request, it might be more efficient to use.
892
893Example: load F</etc/passwd> into C<$passwd>.
894
895 my $passwd;
896 aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
897 $_[0] >= 0
898 or die "/etc/passwd: $!\n";
899
900 printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
901 print $passwd;
902 };
903 IO::AIO::flush;
904
905
523=item aio_load $path, $data, $callback->($status) 906=item aio_load $pathname, $data, $callback->($status)
524 907
525This is a composite request that tries to fully load the given file into 908This is a composite request that tries to fully load the given file into
526memory. Status is the same as with aio_read. 909memory. Status is the same as with aio_read.
527 910
911Using C<aio_slurp> might be more efficient, as it is a single request.
912
528=cut 913=cut
529 914
530sub aio_load($$;$) { 915sub aio_load($$;$) {
531 aio_block {
532 my ($path, undef, $cb) = @_; 916 my ($path, undef, $cb) = @_;
533 my $data = \$_[1]; 917 my $data = \$_[1];
534 918
535 my $pri = aioreq_pri; 919 my $pri = aioreq_pri;
536 my $grp = aio_group $cb; 920 my $grp = aio_group $cb;
921
922 aioreq_pri $pri;
923 add $grp aio_open $path, O_RDONLY, 0, sub {
924 my $fh = shift
925 or return $grp->result (-1);
537 926
538 aioreq_pri $pri; 927 aioreq_pri $pri;
539 add $grp aio_open $path, O_RDONLY, 0, sub {
540 my $fh = shift
541 or return $grp->result (-1);
542
543 aioreq_pri $pri;
544 add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub { 928 add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub {
545 $grp->result ($_[0]); 929 $grp->result ($_[0]);
546 };
547 }; 930 };
548
549 $grp
550 } 931 };
932
933 $grp
551} 934}
552 935
553=item aio_copy $srcpath, $dstpath, $callback->($status) 936=item aio_copy $srcpath, $dstpath, $callback->($status)
554 937
555Try to copy the I<file> (directories not supported as either source or 938Try to copy the I<file> (directories not supported as either source or
556destination) from C<$srcpath> to C<$dstpath> and call the callback with 939destination) from C<$srcpath> to C<$dstpath> and call the callback with
557the C<0> (error) or C<-1> ok. 940a status of C<0> (ok) or C<-1> (error, see C<$!>).
558 941
942Existing destination files will be truncated.
943
559This is a composite request that it creates the destination file with 944This is a composite request that creates the destination file with
560mode 0200 and copies the contents of the source file into it using 945mode 0200 and copies the contents of the source file into it using
561C<aio_sendfile>, followed by restoring atime, mtime, access mode and 946C<aio_sendfile>, followed by restoring atime, mtime, access mode and
562uid/gid, in that order. 947uid/gid, in that order.
563 948
564If an error occurs, the partial destination file will be unlinked, if 949If an error occurs, the partial destination file will be unlinked, if
… …
566errors are being ignored. 951errors are being ignored.
567 952
568=cut 953=cut
569 954
570sub aio_copy($$;$) { 955sub aio_copy($$;$) {
571 aio_block {
572 my ($src, $dst, $cb) = @_; 956 my ($src, $dst, $cb) = @_;
573 957
574 my $pri = aioreq_pri; 958 my $pri = aioreq_pri;
575 my $grp = aio_group $cb; 959 my $grp = aio_group $cb;
576 960
577 aioreq_pri $pri; 961 aioreq_pri $pri;
578 add $grp aio_open $src, O_RDONLY, 0, sub { 962 add $grp aio_open $src, O_RDONLY, 0, sub {
579 if (my $src_fh = $_[0]) { 963 if (my $src_fh = $_[0]) {
580 my @stat = stat $src_fh; 964 my @stat = stat $src_fh; # hmm, might block over nfs?
581 965
582 aioreq_pri $pri; 966 aioreq_pri $pri;
583 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub { 967 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub {
584 if (my $dst_fh = $_[0]) { 968 if (my $dst_fh = $_[0]) {
585 aioreq_pri $pri; 969 aioreq_pri $pri;
586 add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub { 970 add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub {
587 if ($_[0] == $stat[7]) { 971 if ($_[0] == $stat[7]) {
588 $grp->result (0); 972 $grp->result (0);
589 close $src_fh; 973 close $src_fh;
590 974
591 # those should not normally block. should. should.
592 utime $stat[8], $stat[9], $dst;
593 chmod $stat[2] & 07777, $dst_fh;
594 chown $stat[4], $stat[5], $dst_fh;
595 close $dst_fh;
596 } else { 975 my $ch = sub {
597 $grp->result (-1);
598 close $src_fh;
599 close $dst_fh;
600
601 aioreq $pri; 976 aioreq_pri $pri;
977 add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub {
978 aioreq_pri $pri;
979 add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub {
980 aioreq_pri $pri;
602 add $grp aio_unlink $dst; 981 add $grp aio_close $dst_fh;
982 }
983 };
603 } 984 };
985
986 aioreq_pri $pri;
987 add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub {
988 if ($_[0] < 0 && $! == ENOSYS) {
989 aioreq_pri $pri;
990 add $grp aio_utime $dst, $stat[8], $stat[9], $ch;
991 } else {
992 $ch->();
993 }
994 };
995 } else {
996 $grp->result (-1);
997 close $src_fh;
998 close $dst_fh;
999
1000 aioreq $pri;
1001 add $grp aio_unlink $dst;
604 }; 1002 }
605 } else {
606 $grp->result (-1);
607 } 1003 };
1004 } else {
1005 $grp->result (-1);
608 }, 1006 }
609
610 } else {
611 $grp->result (-1);
612 } 1007 },
1008
1009 } else {
1010 $grp->result (-1);
613 }; 1011 }
614
615 $grp
616 } 1012 };
1013
1014 $grp
617} 1015}
618 1016
619=item aio_move $srcpath, $dstpath, $callback->($status) 1017=item aio_move $srcpath, $dstpath, $callback->($status)
620 1018
621Try to move the I<file> (directories not supported as either source or 1019Try to move the I<file> (directories not supported as either source or
622destination) from C<$srcpath> to C<$dstpath> and call the callback with 1020destination) from C<$srcpath> to C<$dstpath> and call the callback with
623the C<0> (error) or C<-1> ok. 1021a status of C<0> (ok) or C<-1> (error, see C<$!>).
624 1022
625This is a composite request that tries to rename(2) the file first. If 1023This is a composite request that tries to rename(2) the file first; if
626rename files with C<EXDEV>, it copies the file with C<aio_copy> and, if 1024rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if
627that is successful, unlinking the C<$srcpath>. 1025that is successful, unlinks the C<$srcpath>.
628 1026
629=cut 1027=cut
630 1028
631sub aio_move($$;$) { 1029sub aio_move($$;$) {
632 aio_block {
633 my ($src, $dst, $cb) = @_; 1030 my ($src, $dst, $cb) = @_;
634 1031
635 my $pri = aioreq_pri; 1032 my $pri = aioreq_pri;
636 my $grp = aio_group $cb; 1033 my $grp = aio_group $cb;
637 1034
638 aioreq_pri $pri; 1035 aioreq_pri $pri;
639 add $grp aio_rename $src, $dst, sub { 1036 add $grp aio_rename $src, $dst, sub {
640 if ($_[0] && $! == EXDEV) { 1037 if ($_[0] && $! == EXDEV) {
641 aioreq_pri $pri; 1038 aioreq_pri $pri;
642 add $grp aio_copy $src, $dst, sub { 1039 add $grp aio_copy $src, $dst, sub {
643 $grp->result ($_[0]);
644
645 if (!$_[0]) {
646 aioreq_pri $pri;
647 add $grp aio_unlink $src;
648 }
649 };
650 } else {
651 $grp->result ($_[0]); 1040 $grp->result ($_[0]);
1041
1042 unless ($_[0]) {
1043 aioreq_pri $pri;
1044 add $grp aio_unlink $src;
1045 }
652 } 1046 };
1047 } else {
1048 $grp->result ($_[0]);
653 }; 1049 }
654
655 $grp
656 } 1050 };
1051
1052 $grp
657} 1053}
658 1054
659=item aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 1055=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
660 1056
661Scans a directory (similar to C<aio_readdir>) but additionally tries to 1057Scans a directory (similar to C<aio_readdir>) but additionally tries to
662efficiently separate the entries of directory C<$path> into two sets of 1058efficiently separate the entries of directory C<$path> into two sets of
663names, directories you can recurse into (directories), and ones you cannot 1059names, directories you can recurse into (directories), and ones you cannot
664recurse into (everything else, including symlinks to directories). 1060recurse into (everything else, including symlinks to directories).
665 1061
666C<aio_scandir> is a composite request that creates of many sub requests_ 1062C<aio_scandir> is a composite request that generates many sub requests.
667C<$maxreq> specifies the maximum number of outstanding aio requests that 1063C<$maxreq> specifies the maximum number of outstanding aio requests that
668this function generates. If it is C<< <= 0 >>, then a suitable default 1064this function generates. If it is C<< <= 0 >>, then a suitable default
669will be chosen (currently 4). 1065will be chosen (currently 4).
670 1066
671On error, the callback is called without arguments, otherwise it receives 1067On error, the callback is called without arguments, otherwise it receives
… …
681 1077
682Implementation notes. 1078Implementation notes.
683 1079
684The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can. 1080The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can.
685 1081
1082If readdir returns file type information, then this is used directly to
1083find directories.
1084
686After reading the directory, the modification time, size etc. of the 1085Otherwise, after reading the directory, the modification time, size etc.
687directory before and after the readdir is checked, and if they match (and 1086of the directory before and after the readdir is checked, and if they
688isn't the current time), the link count will be used to decide how many 1087match (and isn't the current time), the link count will be used to decide
689entries are directories (if >= 2). Otherwise, no knowledge of the number 1088how many entries are directories (if >= 2). Otherwise, no knowledge of the
690of subdirectories will be assumed. 1089number of subdirectories will be assumed.
691 1090
692Then entries will be sorted into likely directories (everything without 1091Then entries will be sorted into likely directories a non-initial dot
693a non-initial dot currently) and likely non-directories (everything 1092currently) and likely non-directories (see C<aio_readdirx>). Then every
694else). Then every entry plus an appended C</.> will be C<stat>'ed, 1093entry plus an appended C</.> will be C<stat>'ed, likely directories first,
695likely directories first. If that succeeds, it assumes that the entry 1094in order of their inode numbers. If that succeeds, it assumes that the
696is a directory or a symlink to directory (which will be checked 1095entry is a directory or a symlink to directory (which will be checked
697seperately). This is often faster than stat'ing the entry itself because 1096separately). This is often faster than stat'ing the entry itself because
698filesystems might detect the type of the entry without reading the inode 1097filesystems might detect the type of the entry without reading the inode
699data (e.g. ext2fs filetype feature). 1098data (e.g. ext2fs filetype feature), even on systems that cannot return
1099the filetype information on readdir.
700 1100
701If the known number of directories (link count - 2) has been reached, the 1101If the known number of directories (link count - 2) has been reached, the
702rest of the entries is assumed to be non-directories. 1102rest of the entries is assumed to be non-directories.
703 1103
704This only works with certainty on POSIX (= UNIX) filesystems, which 1104This only works with certainty on POSIX (= UNIX) filesystems, which
… …
709directory counting heuristic. 1109directory counting heuristic.
710 1110
711=cut 1111=cut
712 1112
713sub aio_scandir($$;$) { 1113sub aio_scandir($$;$) {
714 aio_block {
715 my ($path, $maxreq, $cb) = @_; 1114 my ($path, $maxreq, $cb) = @_;
716 1115
717 my $pri = aioreq_pri; 1116 my $pri = aioreq_pri;
718 1117
719 my $grp = aio_group $cb; 1118 my $grp = aio_group $cb;
720 1119
721 $maxreq = 4 if $maxreq <= 0; 1120 $maxreq = 4 if $maxreq <= 0;
1121
1122 # get a wd object
1123 aioreq_pri $pri;
1124 add $grp aio_wd $path, sub {
1125 $_[0]
1126 or return $grp->result ();
1127
1128 my $wd = [shift, "."];
722 1129
723 # stat once 1130 # stat once
724 aioreq_pri $pri; 1131 aioreq_pri $pri;
725 add $grp aio_stat $path, sub { 1132 add $grp aio_stat $wd, sub {
726 return $grp->result () if $_[0]; 1133 return $grp->result () if $_[0];
727 my $now = time; 1134 my $now = time;
728 my $hash1 = join ":", (stat _)[0,1,3,7,9]; 1135 my $hash1 = join ":", (stat _)[0,1,3,7,9];
1136 my $rdxflags = READDIR_DIRS_FIRST;
1137
1138 if ((stat _)[3] < 2) {
1139 # at least one non-POSIX filesystem exists
1140 # that returns useful DT_type values: btrfs,
1141 # so optimise for this here by requesting dents
1142 $rdxflags |= READDIR_DENTS;
1143 }
729 1144
730 # read the directory entries 1145 # read the directory entries
731 aioreq_pri $pri; 1146 aioreq_pri $pri;
732 add $grp aio_readdir $path, sub { 1147 add $grp aio_readdirx $wd, $rdxflags, sub {
733 my $entries = shift 1148 my ($entries, $flags) = @_
734 or return $grp->result (); 1149 or return $grp->result ();
1150
1151 if ($rdxflags & READDIR_DENTS) {
1152 # if we requested type values, see if we can use them directly.
1153
1154 # if there were any DT_UNKNOWN entries then we assume we
1155 # don't know. alternatively, we could assume that if we get
1156 # one DT_DIR, then all directories are indeed marked with
1157 # DT_DIR, but this seems not required for btrfs, and this
1158 # is basically the "btrfs can't get it's act together" code
1159 # branch.
1160 unless ($flags & READDIR_FOUND_UNKNOWN) {
1161 # now we have valid DT_ information for all entries,
1162 # so use it as an optimisation without further stat's.
1163 # they must also all be at the beginning of @$entries
1164 # by now.
1165
1166 my $dirs;
1167
1168 if (@$entries) {
1169 for (0 .. $#$entries) {
1170 if ($entries->[$_][1] != DT_DIR) {
1171 # splice out directories
1172 $dirs = [splice @$entries, 0, $_];
1173 last;
1174 }
1175 }
1176
1177 # if we didn't find any non-dir, then all entries are dirs
1178 unless ($dirs) {
1179 ($dirs, $entries) = ($entries, []);
1180 }
1181 } else {
1182 # directory is empty, so there are no sbdirs
1183 $dirs = [];
1184 }
1185
1186 # either splice'd the directories out or the dir was empty.
1187 # convert dents to filenames
1188 $_ = $_->[0] for @$dirs;
1189 $_ = $_->[0] for @$entries;
1190
1191 return $grp->result ($dirs, $entries);
1192 }
1193
1194 # cannot use, so return to our old ways
1195 # by pretending we only scanned for names.
1196 $_ = $_->[0] for @$entries;
1197 }
735 1198
736 # stat the dir another time 1199 # stat the dir another time
737 aioreq_pri $pri; 1200 aioreq_pri $pri;
738 add $grp aio_stat $path, sub { 1201 add $grp aio_stat $wd, sub {
739 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 1202 my $hash2 = join ":", (stat _)[0,1,3,7,9];
740 1203
741 my $ndirs; 1204 my $ndirs;
742 1205
743 # take the slow route if anything looks fishy 1206 # take the slow route if anything looks fishy
744 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 1207 if ($hash1 ne $hash2 or (stat _)[9] == $now) {
745 $ndirs = -1; 1208 $ndirs = -1;
746 } else { 1209 } else {
747 # if nlink == 2, we are finished 1210 # if nlink == 2, we are finished
748 # on non-posix-fs's, we rely on nlink < 2 1211 # for non-posix-fs's, we rely on nlink < 2
749 $ndirs = (stat _)[3] - 2 1212 $ndirs = (stat _)[3] - 2
750 or return $grp->result ([], $entries); 1213 or return $grp->result ([], $entries);
751 } 1214 }
752 1215
753 # sort into likely dirs and likely nondirs
754 # dirs == files without ".", short entries first
755 $entries = [map $_->[0],
756 sort { $b->[1] cmp $a->[1] }
757 map [$_, sprintf "%s%04d", (/.\./ ? "1" : "0"), length],
758 @$entries];
759
760 my (@dirs, @nondirs); 1216 my (@dirs, @nondirs);
761 1217
762 my $statgrp = add $grp aio_group sub { 1218 my $statgrp = add $grp aio_group sub {
763 $grp->result (\@dirs, \@nondirs); 1219 $grp->result (\@dirs, \@nondirs);
764 }; 1220 };
765 1221
766 limit $statgrp $maxreq; 1222 limit $statgrp $maxreq;
767 feed $statgrp sub { 1223 feed $statgrp sub {
768 return unless @$entries; 1224 return unless @$entries;
769 my $entry = pop @$entries; 1225 my $entry = shift @$entries;
770 1226
771 aioreq_pri $pri; 1227 aioreq_pri $pri;
1228 $wd->[1] = "$entry/.";
772 add $statgrp aio_stat "$path/$entry/.", sub { 1229 add $statgrp aio_stat $wd, sub {
773 if ($_[0] < 0) { 1230 if ($_[0] < 0) {
774 push @nondirs, $entry; 1231 push @nondirs, $entry;
775 } else { 1232 } else {
776 # need to check for real directory 1233 # need to check for real directory
777 aioreq_pri $pri; 1234 aioreq_pri $pri;
1235 $wd->[1] = $entry;
778 add $statgrp aio_lstat "$path/$entry", sub { 1236 add $statgrp aio_lstat $wd, sub {
779 if (-d _) { 1237 if (-d _) {
780 push @dirs, $entry; 1238 push @dirs, $entry;
781 1239
782 unless (--$ndirs) { 1240 unless (--$ndirs) {
783 push @nondirs, @$entries; 1241 push @nondirs, @$entries;
… …
791 }; 1249 };
792 }; 1250 };
793 }; 1251 };
794 }; 1252 };
795 }; 1253 };
796
797 $grp
798 } 1254 };
1255
1256 $grp
799} 1257}
800 1258
801=item aio_rmtree $path, $callback->($status) 1259=item aio_rmtree $pathname, $callback->($status)
802 1260
803Delete a directory tree starting (and including) C<$path>, return the 1261Delete a directory tree starting (and including) C<$path>, return the
804status of the final C<rmdir> only. This is a composite request that 1262status of the final C<rmdir> only. This is a composite request that
805uses C<aio_scandir> to recurse into and rmdir directories, and unlink 1263uses C<aio_scandir> to recurse into and rmdir directories, and unlink
806everything else. 1264everything else.
807 1265
808=cut 1266=cut
809 1267
810sub aio_rmtree; 1268sub aio_rmtree;
811sub aio_rmtree($;$) { 1269sub aio_rmtree($;$) {
812 aio_block {
813 my ($path, $cb) = @_; 1270 my ($path, $cb) = @_;
814 1271
815 my $pri = aioreq_pri; 1272 my $pri = aioreq_pri;
816 my $grp = aio_group $cb; 1273 my $grp = aio_group $cb;
817 1274
818 aioreq_pri $pri; 1275 aioreq_pri $pri;
819 add $grp aio_scandir $path, 0, sub { 1276 add $grp aio_scandir $path, 0, sub {
820 my ($dirs, $nondirs) = @_; 1277 my ($dirs, $nondirs) = @_;
821 1278
822 my $dirgrp = aio_group sub { 1279 my $dirgrp = aio_group sub {
823 add $grp aio_rmdir $path, sub { 1280 add $grp aio_rmdir $path, sub {
824 $grp->result ($_[0]); 1281 $grp->result ($_[0]);
825 };
826 }; 1282 };
827
828 (aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
829 (aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
830
831 add $grp $dirgrp;
832 }; 1283 };
833 1284
834 $grp 1285 (aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs;
1286 (aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs;
1287
1288 add $grp $dirgrp;
835 } 1289 };
1290
1291 $grp
836} 1292}
1293
1294=item aio_fcntl $fh, $cmd, $arg, $callback->($status)
1295
1296=item aio_ioctl $fh, $request, $buf, $callback->($status)
1297
1298These work just like the C<fcntl> and C<ioctl> built-in functions, except
1299they execute asynchronously and pass the return value to the callback.
1300
1301Both calls can be used for a lot of things, some of which make more sense
1302to run asynchronously in their own thread, while some others make less
1303sense. For example, calls that block waiting for external events, such
1304as locking, will also lock down an I/O thread while it is waiting, which
1305can deadlock the whole I/O system. At the same time, there might be no
1306alternative to using a thread to wait.
1307
1308So in general, you should only use these calls for things that do
1309(filesystem) I/O, not for things that wait for other events (network,
1310other processes), although if you are careful and know what you are doing,
1311you still can.
1312
1313The following constants are available (missing ones are, as usual C<0>):
1314
1315C<F_DUPFD_CLOEXEC>,
1316
1317C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>,
1318
1319C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>.
1320
1321C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>,
1322C<FS_IOC_FIEMAP>.
1323
1324C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>,
1325C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>.
1326
1327C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>,
1328C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>,
1329C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>,
1330C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>,
1331C<FS_FL_USER_MODIFIABLE>.
1332
1333C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>,
1334C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>,
1335C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>,
1336C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>,
1337
1338=item aio_sync $callback->($status)
1339
1340Asynchronously call sync and call the callback when finished.
837 1341
838=item aio_fsync $fh, $callback->($status) 1342=item aio_fsync $fh, $callback->($status)
839 1343
840Asynchronously call fsync on the given filehandle and call the callback 1344Asynchronously call fsync on the given filehandle and call the callback
841with the fsync result code. 1345with the fsync result code.
… …
845Asynchronously call fdatasync on the given filehandle and call the 1349Asynchronously call fdatasync on the given filehandle and call the
846callback with the fdatasync result code. 1350callback with the fdatasync result code.
847 1351
848If this call isn't available because your OS lacks it or it couldn't be 1352If this call isn't available because your OS lacks it or it couldn't be
849detected, it will be emulated by calling C<fsync> instead. 1353detected, it will be emulated by calling C<fsync> instead.
1354
1355=item aio_syncfs $fh, $callback->($status)
1356
1357Asynchronously call the syncfs syscall to sync the filesystem associated
1358to the given filehandle and call the callback with the syncfs result
1359code. If syncfs is not available, calls sync(), but returns C<-1> and sets
1360errno to C<ENOSYS> nevertheless.
1361
1362=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
1363
1364Sync the data portion of the file specified by C<$offset> and C<$length>
1365to disk (but NOT the metadata), by calling the Linux-specific
1366sync_file_range call. If sync_file_range is not available or it returns
1367ENOSYS, then fdatasync or fsync is being substituted.
1368
1369C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>,
1370C<IO::AIO::SYNC_FILE_RANGE_WRITE> and
1371C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range
1372manpage for details.
1373
1374=item aio_pathsync $pathname, $callback->($status)
1375
1376This request tries to open, fsync and close the given path. This is a
1377composite request intended to sync directories after directory operations
1378(E.g. rename). This might not work on all operating systems or have any
1379specific effect, but usually it makes sure that directory changes get
1380written to disc. It works for anything that can be opened for read-only,
1381not just directories.
1382
1383Future versions of this function might fall back to other methods when
1384C<fsync> on the directory fails (such as calling C<sync>).
1385
1386Passes C<0> when everything went ok, and C<-1> on error.
1387
1388=cut
1389
1390sub aio_pathsync($;$) {
1391 my ($path, $cb) = @_;
1392
1393 my $pri = aioreq_pri;
1394 my $grp = aio_group $cb;
1395
1396 aioreq_pri $pri;
1397 add $grp aio_open $path, O_RDONLY, 0, sub {
1398 my ($fh) = @_;
1399 if ($fh) {
1400 aioreq_pri $pri;
1401 add $grp aio_fsync $fh, sub {
1402 $grp->result ($_[0]);
1403
1404 aioreq_pri $pri;
1405 add $grp aio_close $fh;
1406 };
1407 } else {
1408 $grp->result (-1);
1409 }
1410 };
1411
1412 $grp
1413}
1414
1415=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
1416
1417This is a rather advanced IO::AIO call, which only works on mmap(2)ed
1418scalars (see the C<IO::AIO::mmap> function, although it also works on data
1419scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the
1420scalar must only be modified in-place while an aio operation is pending on
1421it).
1422
1423It calls the C<msync> function of your OS, if available, with the memory
1424area starting at C<$offset> in the string and ending C<$length> bytes
1425later. If C<$length> is negative, counts from the end, and if C<$length>
1426is C<undef>, then it goes till the end of the string. The flags can be
1427either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional
1428C<IO::AIO::MS_INVALIDATE>.
1429
1430=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
1431
1432This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1433scalars.
1434
1435It touches (reads or writes) all memory pages in the specified
1436range inside the scalar. All caveats and parameters are the same
1437as for C<aio_msync>, above, except for flags, which must be either
1438C<0> (which reads all pages and ensures they are instantiated) or
1439C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and
1440writing an octet from it, which dirties the page).
1441
1442=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1443
1444This is a rather advanced IO::AIO call, which works best on mmap(2)ed
1445scalars.
1446
1447It reads in all the pages of the underlying storage into memory (if any)
1448and locks them, so they are not getting swapped/paged out or removed.
1449
1450If C<$length> is undefined, then the scalar will be locked till the end.
1451
1452On systems that do not implement C<mlock>, this function returns C<-1>
1453and sets errno to C<ENOSYS>.
1454
1455Note that the corresponding C<munlock> is synchronous and is
1456documented under L<MISCELLANEOUS FUNCTIONS>.
1457
1458Example: open a file, mmap and mlock it - both will be undone when
1459C<$data> gets destroyed.
1460
1461 open my $fh, "<", $path or die "$path: $!";
1462 my $data;
1463 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1464 aio_mlock $data; # mlock in background
1465
1466=item aio_mlockall $flags, $callback->($status)
1467
1468Calls the C<mlockall> function with the given C<$flags> (a
1469combination of C<IO::AIO::MCL_CURRENT>, C<IO::AIO::MCL_FUTURE> and
1470C<IO::AIO::MCL_ONFAULT>).
1471
1472On systems that do not implement C<mlockall>, this function returns C<-1>
1473and sets errno to C<ENOSYS>. Similarly, flag combinations not supported
1474by the system result in a return value of C<-1> with errno being set to
1475C<EINVAL>.
1476
1477Note that the corresponding C<munlockall> is synchronous and is
1478documented under L<MISCELLANEOUS FUNCTIONS>.
1479
1480Example: asynchronously lock all current and future pages into memory.
1481
1482 aio_mlockall IO::AIO::MCL_FUTURE;
1483
1484=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
1485
1486Queries the extents of the given file (by calling the Linux C<FIEMAP>
1487ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If
1488the ioctl is not available on your OS, then this request will fail with
1489C<ENOSYS>.
1490
1491C<$start> is the starting offset to query extents for, C<$length> is the
1492size of the range to query - if it is C<undef>, then the whole file will
1493be queried.
1494
1495C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or
1496C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also
1497exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query
1498the data portion.
1499
1500C<$count> is the maximum number of extent records to return. If it is
1501C<undef>, then IO::AIO queries all extents of the range. As a very special
1502case, if it is C<0>, then the callback receives the number of extents
1503instead of the extents themselves (which is unreliable, see below).
1504
1505If an error occurs, the callback receives no arguments. The special
1506C<errno> value C<IO::AIO::EBADR> is available to test for flag errors.
1507
1508Otherwise, the callback receives an array reference with extent
1509structures. Each extent structure is an array reference itself, with the
1510following members:
1511
1512 [$logical, $physical, $length, $flags]
1513
1514Flags is any combination of the following flag values (typically either C<0>
1515or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)):
1516
1517C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>,
1518C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>,
1519C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>,
1520C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>,
1521C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or
1522C<IO::AIO::FIEMAP_EXTENT_SHARED>.
1523
1524At the time of this writing (Linux 3.2), this request is unreliable unless
1525C<$count> is C<undef>, as the kernel has all sorts of bugs preventing
1526it to return all extents of a range for files with a large number of
1527extents. The code (only) works around all these issues if C<$count> is
1528C<undef>.
850 1529
851=item aio_group $callback->(...) 1530=item aio_group $callback->(...)
852 1531
853This is a very special aio request: Instead of doing something, it is a 1532This is a very special aio request: Instead of doing something, it is a
854container for other aio requests, which is useful if you want to bundle 1533container for other aio requests, which is useful if you want to bundle
… …
892immense (it blocks a thread for a long time) so do not use this function 1571immense (it blocks a thread for a long time) so do not use this function
893except to put your application under artificial I/O pressure. 1572except to put your application under artificial I/O pressure.
894 1573
895=back 1574=back
896 1575
1576
1577=head2 IO::AIO::WD - multiple working directories
1578
1579Your process only has one current working directory, which is used by all
1580threads. This makes it hard to use relative paths (some other component
1581could call C<chdir> at any time, and it is hard to control when the path
1582will be used by IO::AIO).
1583
1584One solution for this is to always use absolute paths. This usually works,
1585but can be quite slow (the kernel has to walk the whole path on every
1586access), and can also be a hassle to implement.
1587
1588Newer POSIX systems have a number of functions (openat, fdopendir,
1589futimensat and so on) that make it possible to specify working directories
1590per operation.
1591
1592For portability, and because the clowns who "designed", or shall I write,
1593perpetrated this new interface were obviously half-drunk, this abstraction
1594cannot be perfect, though.
1595
1596IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD
1597object. This object stores the canonicalised, absolute version of the
1598path, and on systems that allow it, also a directory file descriptor.
1599
1600Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat>
1601or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD
1602object and a pathname instead (or the IO::AIO::WD object alone, which
1603gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the
1604IO::AIO::WD object is ignored, otherwise the pathname is resolved relative
1605to that IO::AIO::WD object.
1606
1607For example, to get a wd object for F</etc> and then stat F<passwd>
1608inside, you would write:
1609
1610 aio_wd "/etc", sub {
1611 my $etcdir = shift;
1612
1613 # although $etcdir can be undef on error, there is generally no reason
1614 # to check for errors here, as aio_stat will fail with ENOENT
1615 # when $etcdir is undef.
1616
1617 aio_stat [$etcdir, "passwd"], sub {
1618 # yay
1619 };
1620 };
1621
1622The fact that C<aio_wd> is a request and not a normal function shows that
1623creating an IO::AIO::WD object is itself a potentially blocking operation,
1624which is why it is done asynchronously.
1625
1626To stat the directory obtained with C<aio_wd> above, one could write
1627either of the following three request calls:
1628
1629 aio_lstat "/etc" , sub { ... # pathname as normal string
1630 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1631 aio_lstat $wd , sub { ... # shorthand for the previous
1632
1633As with normal pathnames, IO::AIO keeps a copy of the working directory
1634object and the pathname string, so you could write the following without
1635causing any issues due to C<$path> getting reused:
1636
1637 my $path = [$wd, undef];
1638
1639 for my $name (qw(abc def ghi)) {
1640 $path->[1] = $name;
1641 aio_stat $path, sub {
1642 # ...
1643 };
1644 }
1645
1646There are some caveats: when directories get renamed (or deleted), the
1647pathname string doesn't change, so will point to the new directory (or
1648nowhere at all), while the directory fd, if available on the system,
1649will still point to the original directory. Most functions accepting a
1650pathname will use the directory fd on newer systems, and the string on
1651older systems. Some functions (such as C<aio_realpath>) will always rely on
1652the string form of the pathname.
1653
1654So this functionality is mainly useful to get some protection against
1655C<chdir>, to easily get an absolute path out of a relative path for future
1656reference, and to speed up doing many operations in the same directory
1657(e.g. when stat'ing all files in a directory).
1658
1659The following functions implement this working directory abstraction:
1660
1661=over 4
1662
1663=item aio_wd $pathname, $callback->($wd)
1664
1665Asynchonously canonicalise the given pathname and convert it to an
1666IO::AIO::WD object representing it. If possible and supported on the
1667system, also open a directory fd to speed up pathname resolution relative
1668to this working directory.
1669
1670If something goes wrong, then C<undef> is passwd to the callback instead
1671of a working directory object and C<$!> is set appropriately. Since
1672passing C<undef> as working directory component of a pathname fails the
1673request with C<ENOENT>, there is often no need for error checking in the
1674C<aio_wd> callback, as future requests using the value will fail in the
1675expected way.
1676
1677=item IO::AIO::CWD
1678
1679This is a compiletime constant (object) that represents the process
1680current working directory.
1681
1682Specifying this object as working directory object for a pathname is as if
1683the pathname would be specified directly, without a directory object. For
1684example, these calls are functionally identical:
1685
1686 aio_stat "somefile", sub { ... };
1687 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1688
1689=back
1690
1691To recover the path associated with an IO::AIO::WD object, you can use
1692C<aio_realpath>:
1693
1694 aio_realpath $wd, sub {
1695 warn "path is $_[0]\n";
1696 };
1697
1698Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir>
1699sometimes, fall back to using an absolue path.
1700
897=head2 IO::AIO::REQ CLASS 1701=head2 IO::AIO::REQ CLASS
898 1702
899All non-aggregate C<aio_*> functions return an object of this class when 1703All non-aggregate C<aio_*> functions return an object of this class when
900called in non-void context. 1704called in non-void context.
901 1705
… …
904=item cancel $req 1708=item cancel $req
905 1709
906Cancels the request, if possible. Has the effect of skipping execution 1710Cancels the request, if possible. Has the effect of skipping execution
907when entering the B<execute> state and skipping calling the callback when 1711when entering the B<execute> state and skipping calling the callback when
908entering the the B<result> state, but will leave the request otherwise 1712entering the the B<result> state, but will leave the request otherwise
909untouched. That means that requests that currently execute will not be 1713untouched (with the exception of readdir). That means that requests that
910stopped and resources held by the request will not be freed prematurely. 1714currently execute will not be stopped and resources held by the request
1715will not be freed prematurely.
911 1716
912=item cb $req $callback->(...) 1717=item cb $req $callback->(...)
913 1718
914Replace (or simply set) the callback registered to the request. 1719Replace (or simply set) the callback registered to the request.
915 1720
… …
966Their lifetime, simplified, looks like this: when they are empty, they 1771Their lifetime, simplified, looks like this: when they are empty, they
967will finish very quickly. If they contain only requests that are in the 1772will finish very quickly. If they contain only requests that are in the
968C<done> state, they will also finish. Otherwise they will continue to 1773C<done> state, they will also finish. Otherwise they will continue to
969exist. 1774exist.
970 1775
971That means after creating a group you have some time to add requests. And 1776That means after creating a group you have some time to add requests
972in the callbacks of those requests, you can add further requests to the 1777(precisely before the callback has been invoked, which is only done within
973group. And only when all those requests have finished will the the group 1778the C<poll_cb>). And in the callbacks of those requests, you can add
974itself finish. 1779further requests to the group. And only when all those requests have
1780finished will the the group itself finish.
975 1781
976=over 4 1782=over 4
977 1783
978=item add $grp ... 1784=item add $grp ...
979 1785
… …
988=item $grp->cancel_subs 1794=item $grp->cancel_subs
989 1795
990Cancel all subrequests and clears any feeder, but not the group request 1796Cancel all subrequests and clears any feeder, but not the group request
991itself. Useful when you queued a lot of events but got a result early. 1797itself. Useful when you queued a lot of events but got a result early.
992 1798
1799The group request will finish normally (you cannot add requests to the
1800group).
1801
993=item $grp->result (...) 1802=item $grp->result (...)
994 1803
995Set the result value(s) that will be passed to the group callback when all 1804Set the result value(s) that will be passed to the group callback when all
996subrequests have finished and set thre groups errno to the current value 1805subrequests have finished and set the groups errno to the current value
997of errno (just like calling C<errno> without an error number). By default, 1806of errno (just like calling C<errno> without an error number). By default,
998no argument will be passed and errno is zero. 1807no argument will be passed and errno is zero.
999 1808
1000=item $grp->errno ([$errno]) 1809=item $grp->errno ([$errno])
1001 1810
… …
1012=item feed $grp $callback->($grp) 1821=item feed $grp $callback->($grp)
1013 1822
1014Sets a feeder/generator on this group: every group can have an attached 1823Sets a feeder/generator on this group: every group can have an attached
1015generator that generates requests if idle. The idea behind this is that, 1824generator that generates requests if idle. The idea behind this is that,
1016although you could just queue as many requests as you want in a group, 1825although you could just queue as many requests as you want in a group,
1017this might starve other requests for a potentially long time. For 1826this might starve other requests for a potentially long time. For example,
1018example, C<aio_scandir> might generate hundreds of thousands C<aio_stat> 1827C<aio_scandir> might generate hundreds of thousands of C<aio_stat>
1019requests, delaying any later requests for a long time. 1828requests, delaying any later requests for a long time.
1020 1829
1021To avoid this, and allow incremental generation of requests, you can 1830To avoid this, and allow incremental generation of requests, you can
1022instead a group and set a feeder on it that generates those requests. The 1831instead a group and set a feeder on it that generates those requests. The
1023feed callback will be called whenever there are few enough (see C<limit>, 1832feed callback will be called whenever there are few enough (see C<limit>,
… …
1028not impose any limits). 1837not impose any limits).
1029 1838
1030If the feed does not queue more requests when called, it will be 1839If the feed does not queue more requests when called, it will be
1031automatically removed from the group. 1840automatically removed from the group.
1032 1841
1033If the feed limit is C<0>, it will be set to C<2> automatically. 1842If the feed limit is C<0> when this method is called, it will be set to
1843C<2> automatically.
1034 1844
1035Example: 1845Example:
1036 1846
1037 # stat all files in @files, but only ever use four aio requests concurrently: 1847 # stat all files in @files, but only ever use four aio requests concurrently:
1038 1848
… …
1050Sets the feeder limit for the group: The feeder will be called whenever 1860Sets the feeder limit for the group: The feeder will be called whenever
1051the group contains less than this many requests. 1861the group contains less than this many requests.
1052 1862
1053Setting the limit to C<0> will pause the feeding process. 1863Setting the limit to C<0> will pause the feeding process.
1054 1864
1865The default value for the limit is C<0>, but note that setting a feeder
1866automatically bumps it up to C<2>.
1867
1055=back 1868=back
1056 1869
1870
1057=head2 SUPPORT FUNCTIONS 1871=head2 SUPPORT FUNCTIONS
1058 1872
1059=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION 1873=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
1060 1874
1061=over 4 1875=over 4
1062 1876
1063=item $fileno = IO::AIO::poll_fileno 1877=item $fileno = IO::AIO::poll_fileno
1064 1878
1065Return the I<request result pipe file descriptor>. This filehandle must be 1879Return the I<request result pipe file descriptor>. This filehandle must be
1066polled for reading by some mechanism outside this module (e.g. Event or 1880polled for reading by some mechanism outside this module (e.g. EV, Glib,
1067select, see below or the SYNOPSIS). If the pipe becomes readable you have 1881select and so on, see below or the SYNOPSIS). If the pipe becomes readable
1068to call C<poll_cb> to check the results. 1882you have to call C<poll_cb> to check the results.
1069 1883
1070See C<poll_cb> for an example. 1884See C<poll_cb> for an example.
1071 1885
1072=item IO::AIO::poll_cb 1886=item IO::AIO::poll_cb
1073 1887
1074Process some outstanding events on the result pipe. You have to call this 1888Process some requests that have reached the result phase (i.e. they have
1075regularly. Returns the number of events processed. Returns immediately 1889been executed but the results are not yet reported). You have to call
1076when no events are outstanding. The amount of events processed depends on 1890this "regularly" to finish outstanding requests.
1077the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
1078 1891
1892Returns C<0> if all events could be processed (or there were no
1893events to process), or C<-1> if it returned earlier for whatever
1894reason. Returns immediately when no events are outstanding. The amount
1895of events processed depends on the settings of C<IO::AIO::max_poll_req>,
1896C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>.
1897
1079If not all requests were processed for whatever reason, the filehandle 1898If not all requests were processed for whatever reason, the poll file
1080will still be ready when C<poll_cb> returns. 1899descriptor will still be ready when C<poll_cb> returns, so normally you
1900don't have to do anything special to have it called later.
1901
1902Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes
1903ready, it can be beneficial to call this function from loops which submit
1904a lot of requests, to make sure the results get processed when they become
1905available and not just when the loop is finished and the event loop takes
1906over again. This function returns very fast when there are no outstanding
1907requests.
1081 1908
1082Example: Install an Event watcher that automatically calls 1909Example: Install an Event watcher that automatically calls
1083IO::AIO::poll_cb with high priority: 1910IO::AIO::poll_cb with high priority (more examples can be found in the
1911SYNOPSIS section, at the top of this document):
1084 1912
1085 Event->io (fd => IO::AIO::poll_fileno, 1913 Event->io (fd => IO::AIO::poll_fileno,
1086 poll => 'r', async => 1, 1914 poll => 'r', async => 1,
1087 cb => \&IO::AIO::poll_cb); 1915 cb => \&IO::AIO::poll_cb);
1916
1917=item IO::AIO::poll_wait
1918
1919Wait until either at least one request is in the result phase or no
1920requests are outstanding anymore.
1921
1922This is useful if you want to synchronously wait for some requests to
1923become ready, without actually handling them.
1924
1925See C<nreqs> for an example.
1926
1927=item IO::AIO::poll
1928
1929Waits until some requests have been handled.
1930
1931Returns the number of requests processed, but is otherwise strictly
1932equivalent to:
1933
1934 IO::AIO::poll_wait, IO::AIO::poll_cb
1935
1936=item IO::AIO::flush
1937
1938Wait till all outstanding AIO requests have been handled.
1939
1940Strictly equivalent to:
1941
1942 IO::AIO::poll_wait, IO::AIO::poll_cb
1943 while IO::AIO::nreqs;
1944
1945This function can be useful at program aborts, to make sure outstanding
1946I/O has been done (C<IO::AIO> uses an C<END> block which already calls
1947this function on normal exits), or when you are merely using C<IO::AIO>
1948for its more advanced functions, rather than for async I/O, e.g.:
1949
1950 my ($dirs, $nondirs);
1951 IO::AIO::aio_scandir "/tmp", 0, sub { ($dirs, $nondirs) = @_ };
1952 IO::AIO::flush;
1953 # $dirs, $nondirs are now set
1088 1954
1089=item IO::AIO::max_poll_reqs $nreqs 1955=item IO::AIO::max_poll_reqs $nreqs
1090 1956
1091=item IO::AIO::max_poll_time $seconds 1957=item IO::AIO::max_poll_time $seconds
1092 1958
… …
1117 # use a low priority so other tasks have priority 1983 # use a low priority so other tasks have priority
1118 Event->io (fd => IO::AIO::poll_fileno, 1984 Event->io (fd => IO::AIO::poll_fileno,
1119 poll => 'r', nice => 1, 1985 poll => 'r', nice => 1,
1120 cb => &IO::AIO::poll_cb); 1986 cb => &IO::AIO::poll_cb);
1121 1987
1122=item IO::AIO::poll_wait
1123
1124If there are any outstanding requests and none of them in the result
1125phase, wait till the result filehandle becomes ready for reading (simply
1126does a C<select> on the filehandle. This is useful if you want to
1127synchronously wait for some requests to finish).
1128
1129See C<nreqs> for an example.
1130
1131=item IO::AIO::poll
1132
1133Waits until some requests have been handled.
1134
1135Returns the number of requests processed, but is otherwise strictly
1136equivalent to:
1137
1138 IO::AIO::poll_wait, IO::AIO::poll_cb
1139
1140=item IO::AIO::flush
1141
1142Wait till all outstanding AIO requests have been handled.
1143
1144Strictly equivalent to:
1145
1146 IO::AIO::poll_wait, IO::AIO::poll_cb
1147 while IO::AIO::nreqs;
1148
1149=back 1988=back
1989
1150 1990
1151=head3 CONTROLLING THE NUMBER OF THREADS 1991=head3 CONTROLLING THE NUMBER OF THREADS
1152 1992
1153=over 1993=over
1154 1994
… …
1186 2026
1187Under normal circumstances you don't need to call this function. 2027Under normal circumstances you don't need to call this function.
1188 2028
1189=item IO::AIO::max_idle $nthreads 2029=item IO::AIO::max_idle $nthreads
1190 2030
1191Limit the number of threads (default: 4) that are allowed to idle (i.e., 2031Limit the number of threads (default: 4) that are allowed to idle
1192threads that did not get a request to process within 10 seconds). That 2032(i.e., threads that did not get a request to process within the idle
1193means if a thread becomes idle while C<$nthreads> other threads are also 2033timeout (default: 10 seconds). That means if a thread becomes idle while
1194idle, it will free its resources and exit. 2034C<$nthreads> other threads are also idle, it will free its resources and
2035exit.
1195 2036
1196This is useful when you allow a large number of threads (e.g. 100 or 1000) 2037This is useful when you allow a large number of threads (e.g. 100 or 1000)
1197to allow for extremely high load situations, but want to free resources 2038to allow for extremely high load situations, but want to free resources
1198under normal circumstances (1000 threads can easily consume 30MB of RAM). 2039under normal circumstances (1000 threads can easily consume 30MB of RAM).
1199 2040
1200The default is probably ok in most situations, especially if thread 2041The default is probably ok in most situations, especially if thread
1201creation is fast. If thread creation is very slow on your system you might 2042creation is fast. If thread creation is very slow on your system you might
1202want to use larger values. 2043want to use larger values.
1203 2044
2045=item IO::AIO::idle_timeout $seconds
2046
2047Sets the minimum idle timeout (default 10) after which worker threads are
2048allowed to exit. SEe C<IO::AIO::max_idle>.
2049
1204=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 2050=item IO::AIO::max_outstanding $maxreqs
2051
2052Sets the maximum number of outstanding requests to C<$nreqs>. If
2053you do queue up more than this number of requests, the next call to
2054C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as
2055C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no
2056longer exceeded.
2057
2058In other words, this setting does not enforce a queue limit, but can be
2059used to make poll functions block if the limit is exceeded.
1205 2060
1206This is a very bad function to use in interactive programs because it 2061This is a very bad function to use in interactive programs because it
1207blocks, and a bad way to reduce concurrency because it is inexact: Better 2062blocks, and a bad way to reduce concurrency because it is inexact: Better
1208use an C<aio_group> together with a feed callback. 2063use an C<aio_group> together with a feed callback.
1209 2064
1210Sets the maximum number of outstanding requests to C<$nreqs>. If you 2065Its main use is in scripts without an event loop - when you want to stat
1211to queue up more than this number of requests, the next call to the 2066a lot of files, you can write something like this:
1212C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>)
1213function will block until the limit is no longer exceeded.
1214 2067
1215The default value is very large, so there is no practical limit on the 2068 IO::AIO::max_outstanding 32;
1216number of outstanding requests.
1217 2069
1218You can still queue as many requests as you want. Therefore, 2070 for my $path (...) {
1219C<max_oustsanding> is mainly useful in simple scripts (with low values) or 2071 aio_stat $path , ...;
1220as a stop gap to shield against fatal memory overflow (with large values). 2072 IO::AIO::poll_cb;
2073 }
2074
2075 IO::AIO::flush;
2076
2077The call to C<poll_cb> inside the loop will normally return instantly, but
2078as soon as more thna C<32> reqeusts are in-flight, it will block until
2079some requests have been handled. This keeps the loop from pushing a large
2080number of C<aio_stat> requests onto the queue.
2081
2082The default value for C<max_outstanding> is very large, so there is no
2083practical limit on the number of outstanding requests.
1221 2084
1222=back 2085=back
2086
1223 2087
1224=head3 STATISTICAL INFORMATION 2088=head3 STATISTICAL INFORMATION
1225 2089
1226=over 2090=over
1227 2091
… …
1245Returns the number of requests currently in the pending state (executed, 2109Returns the number of requests currently in the pending state (executed,
1246but not yet processed by poll_cb). 2110but not yet processed by poll_cb).
1247 2111
1248=back 2112=back
1249 2113
2114
2115=head3 SUBSECOND STAT TIME ACCESS
2116
2117Both C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> functions can
2118generally find access/modification and change times with subsecond time
2119accuracy of the system supports it, but perl's built-in functions only
2120return the integer part.
2121
2122The following functions return the timestamps of the most recent
2123stat with subsecond precision on most systems and work both after
2124C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> calls. Their return
2125value is only meaningful after a successful C<stat>/C<lstat> call, or
2126during/after a successful C<aio_stat>/C<aio_lstat> callback.
2127
2128This is similar to the L<Time::HiRes> C<stat> functions, but can return
2129full resolution without rounding and work with standard perl C<stat>,
2130alleviating the need to call the special C<Time::HiRes> functions, which
2131do not act like their perl counterparts.
2132
2133On operating systems or file systems where subsecond time resolution is
2134not supported or could not be detected, a fractional part of C<0> is
2135returned, so it is always safe to call these functions.
2136
2137=over 4
2138
2139=item $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime
2140
2141Return the access, modication, change or birth time, respectively,
2142including fractional part. Due to the limited precision of floating point,
2143the accuracy on most platforms is only a bit better than milliseconds
2144for times around now - see the I<nsec> function family, below, for full
2145accuracy.
2146
2147File birth time is only available when the OS and perl support it (on
2148FreeBSD and NetBSD at the time of this writing, although support is
2149adaptive, so if your OS/perl gains support, IO::AIO can take advantage of
2150it). On systems where it isn't available, C<0> is currently returned, but
2151this might change to C<undef> in a future version.
2152
2153=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
2154
2155Returns access, modification, change and birth time all in one go, and
2156maybe more times in the future version.
2157
2158=item $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec
2159
2160Return the fractional access, modifcation, change or birth time, in nanoseconds,
2161as an integer in the range C<0> to C<999999999>.
2162
2163Note that no accessors are provided for access, modification and
2164change times - you need to get those from C<stat _> if required (C<int
2165IO::AIO::st_atime> and so on will I<not> generally give you the correct
2166value).
2167
2168=item $seconds = IO::AIO::st_btimesec
2169
2170The (integral) seconds part of the file birth time, if available.
2171
2172=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
2173
2174Like the functions above, but returns all four times in one go (and maybe
2175more in future versions).
2176
2177=item $counter = IO::AIO::st_gen
2178
2179Returns the generation counter (in practice this is just a random number)
2180of the file. This is only available on platforms which have this member in
2181their C<struct stat> (most BSDs at the time of this writing) and generally
2182only to the root usert. If unsupported, C<0> is returned, but this might
2183change to C<undef> in a future version.
2184
2185=back
2186
2187Example: print the high resolution modification time of F</etc>, using
2188C<stat>, and C<IO::AIO::aio_stat>.
2189
2190 if (stat "/etc") {
2191 printf "stat(/etc) mtime: %f\n", IO::AIO::st_mtime;
2192 }
2193
2194 IO::AIO::aio_stat "/etc", sub {
2195 $_[0]
2196 and return;
2197
2198 printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::st_mtimensec;
2199 };
2200
2201 IO::AIO::flush;
2202
2203Output of the awbove on my system, showing reduced and full accuracy:
2204
2205 stat(/etc) mtime: 1534043702.020808
2206 aio_stat(/etc) mtime: 1534043702.020807792
2207
2208
2209=head3 MISCELLANEOUS FUNCTIONS
2210
2211IO::AIO implements some functions that are useful when you want to use
2212some "Advanced I/O" function not available to in Perl, without going the
2213"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*>
2214counterpart.
2215
2216=over 4
2217
2218=item $numfd = IO::AIO::get_fdlimit
2219
2220This function is I<EXPERIMENTAL> and subject to change.
2221
2222Tries to find the current file descriptor limit and returns it, or
2223C<undef> and sets C<$!> in case of an error. The limit is one larger than
2224the highest valid file descriptor number.
2225
2226=item IO::AIO::min_fdlimit [$numfd]
2227
2228This function is I<EXPERIMENTAL> and subject to change.
2229
2230Try to increase the current file descriptor limit(s) to at least C<$numfd>
2231by changing the soft or hard file descriptor resource limit. If C<$numfd>
2232is missing, it will try to set a very high limit, although this is not
2233recommended when you know the actual minimum that you require.
2234
2235If the limit cannot be raised enough, the function makes a best-effort
2236attempt to increase the limit as much as possible, using various
2237tricks, while still failing. You can query the resulting limit using
2238C<IO::AIO::get_fdlimit>.
2239
2240If an error occurs, returns C<undef> and sets C<$!>, otherwise returns
2241true.
2242
2243=item IO::AIO::sendfile $ofh, $ifh, $offset, $count
2244
2245Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>,
2246but is blocking (this makes most sense if you know the input data is
2247likely cached already and the output filehandle is set to non-blocking
2248operations).
2249
2250Returns the number of bytes copied, or C<-1> on error.
2251
2252=item IO::AIO::fadvise $fh, $offset, $len, $advice
2253
2254Simply calls the C<posix_fadvise> function (see its
2255manpage for details). The following advice constants are
2256available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>,
2257C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>,
2258C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>.
2259
2260On systems that do not implement C<posix_fadvise>, this function returns
2261ENOSYS, otherwise the return value of C<posix_fadvise>.
2262
2263=item IO::AIO::madvise $scalar, $offset, $len, $advice
2264
2265Simply calls the C<posix_madvise> function (see its
2266manpage for details). The following advice constants are
2267available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>,
2268C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>,
2269C<IO::AIO::MADV_DONTNEED>.
2270
2271If C<$offset> is negative, counts from the end. If C<$length> is negative,
2272the remaining length of the C<$scalar> is used. If possible, C<$length>
2273will be reduced to fit into the C<$scalar>.
2274
2275On systems that do not implement C<posix_madvise>, this function returns
2276ENOSYS, otherwise the return value of C<posix_madvise>.
2277
2278=item IO::AIO::mprotect $scalar, $offset, $len, $protect
2279
2280Simply calls the C<mprotect> function on the preferably AIO::mmap'ed
2281$scalar (see its manpage for details). The following protect
2282constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>,
2283C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>.
2284
2285If C<$offset> is negative, counts from the end. If C<$length> is negative,
2286the remaining length of the C<$scalar> is used. If possible, C<$length>
2287will be reduced to fit into the C<$scalar>.
2288
2289On systems that do not implement C<mprotect>, this function returns
2290ENOSYS, otherwise the return value of C<mprotect>.
2291
2292=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
2293
2294Memory-maps a file (or anonymous memory range) and attaches it to the
2295given C<$scalar>, which will act like a string scalar. Returns true on
2296success, and false otherwise.
2297
2298The scalar must exist, but its contents do not matter - this means you
2299cannot use a nonexistant array or hash element. When in doubt, C<undef>
2300the scalar first.
2301
2302The only operations allowed on the mmapped scalar are C<substr>/C<vec>,
2303which don't change the string length, and most read-only operations such
2304as copying it or searching it with regexes and so on.
2305
2306Anything else is unsafe and will, at best, result in memory leaks.
2307
2308The memory map associated with the C<$scalar> is automatically removed
2309when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap>
2310or C<IO::AIO::munmap> functions are called on it.
2311
2312This calls the C<mmap>(2) function internally. See your system's manual
2313page for details on the C<$length>, C<$prot> and C<$flags> parameters.
2314
2315The C<$length> must be larger than zero and smaller than the actual
2316filesize.
2317
2318C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>,
2319C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>,
2320
2321C<$flags> can be a combination of
2322C<IO::AIO::MAP_SHARED> or
2323C<IO::AIO::MAP_PRIVATE>,
2324or a number of system-specific flags (when not available, the are C<0>):
2325C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant),
2326C<IO::AIO::MAP_LOCKED>,
2327C<IO::AIO::MAP_NORESERVE>,
2328C<IO::AIO::MAP_POPULATE>,
2329C<IO::AIO::MAP_NONBLOCK>,
2330C<IO::AIO::MAP_FIXED>,
2331C<IO::AIO::MAP_GROWSDOWN>,
2332C<IO::AIO::MAP_32BIT>,
2333C<IO::AIO::MAP_HUGETLB> or
2334C<IO::AIO::MAP_STACK>.
2335
2336If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed.
2337
2338C<$offset> is the offset from the start of the file - it generally must be
2339a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>.
2340
2341Example:
2342
2343 use Digest::MD5;
2344 use IO::AIO;
2345
2346 open my $fh, "<verybigfile"
2347 or die "$!";
2348
2349 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
2350 or die "verybigfile: $!";
2351
2352 my $fast_md5 = md5 $data;
2353
2354=item IO::AIO::munmap $scalar
2355
2356Removes a previous mmap and undefines the C<$scalar>.
2357
2358=item IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[, $new_address = 0]
2359
2360Calls the Linux-specific mremap(2) system call. The C<$scalar> must have
2361been mapped by C<IO::AIO::mmap>, and C<$flags> must currently either be
2362C<0> or C<IO::AIO::MREMAP_MAYMOVE>.
2363
2364Returns true if successful, and false otherwise. If the underlying mmapped
2365region has changed address, then the true value has the numerical value
2366C<1>, otherwise it has the numerical value C<0>:
2367
2368 my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE
2369 or die "mremap: $!";
2370
2371 if ($success*1) {
2372 warn "scalar has chanegd address in memory\n";
2373 }
2374
2375C<IO::AIO::MREMAP_FIXED> and the C<$new_address> argument are currently
2376implemented, but not supported and might go away in a future version.
2377
2378On systems where this call is not supported or is not emulated, this call
2379returns falls and sets C<$!> to C<ENOSYS>.
2380
2381=item IO::AIO::mlockall $flags
2382
2383Calls the C<eio_mlockall_sync> function, which is like C<aio_mlockall>,
2384but is blocking.
2385
2386=item IO::AIO::munlock $scalar, $offset = 0, $length = undef
2387
2388Calls the C<munlock> function, undoing the effects of a previous
2389C<aio_mlock> call (see its description for details).
2390
2391=item IO::AIO::munlockall
2392
2393Calls the C<munlockall> function.
2394
2395On systems that do not implement C<munlockall>, this function returns
2396ENOSYS, otherwise the return value of C<munlockall>.
2397
2398=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
2399
2400Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or
2401C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they
2402should be the file offset.
2403
2404C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might
2405silently corrupt the data in this case.
2406
2407The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>,
2408C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and
2409C<IO::AIO::SPLICE_F_GIFT>.
2410
2411See the C<splice(2)> manpage for details.
2412
2413=item IO::AIO::tee $r_fh, $w_fh, $length, $flags
2414
2415Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the
2416description for C<IO::AIO::splice> above for details.
2417
2418=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
2419
2420Attempts to query or change the pipe buffer size. Obviously works only
2421on pipes, and currently works only on GNU/Linux systems, and fails with
2422C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer
2423size on other systems, drop me a note.
2424
2425=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
2426
2427This is a direct interface to the Linux L<pipe2(2)> system call. If
2428C<$flags> is missing or C<0>, then this should be the same as a call to
2429perl's built-in C<pipe> function and create a new pipe, and works on
2430systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe
2431(..., 4096, O_BINARY)>.
2432
2433If C<$flags> is non-zero, it tries to invoke the pipe2 system call with
2434the given flags (Linux 2.6.27, glibc 2.9).
2435
2436On success, the read and write file handles are returned.
2437
2438On error, nothing will be returned. If the pipe2 syscall is missing and
2439C<$flags> is non-zero, fails with C<ENOSYS>.
2440
2441Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the
2442time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and
2443C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported.
2444
2445Example: create a pipe race-free w.r.t. threads and fork:
2446
2447 my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
2448 or die "pipe2: $!\n";
2449
2450=item $fh = IO::AIO::memfd_create $pathname[, $flags]
2451
2452This is a direct interface to the Linux L<memfd_create(2)> system
2453call. The (unhelpful) default for C<$flags> is C<0>, but your default
2454should be C<IO::AIO::MFD_CLOEXEC>.
2455
2456On success, the new memfd filehandle is returned, otherwise returns
2457C<undef>. If the memfd_create syscall is missing, fails with C<ENOSYS>.
2458
2459Please refer to L<memfd_create(2)> for more info on this call.
2460
2461The following C<$flags> values are available: C<IO::AIO::MFD_CLOEXEC>,
2462C<IO::AIO::MFD_ALLOW_SEALING> and C<IO::AIO::MFD_HUGETLB>.
2463
2464Example: create a new memfd.
2465
2466 my $fh = IO::AIO::memfd_create "somenameforprocfd", IO::AIO::MFD_CLOEXEC
2467 or die "m,emfd_create: $!\n";
2468=item $fh = IO::AIO::eventfd [$initval, [$flags]]
2469
2470This is a direct interface to the Linux L<eventfd(2)> system call. The
2471(unhelpful) defaults for C<$initval> and C<$flags> are C<0> for both.
2472
2473On success, the new eventfd filehandle is returned, otherwise returns
2474C<undef>. If the eventfd syscall is missing, fails with C<ENOSYS>.
2475
2476Please refer to L<eventfd(2)> for more info on this call.
2477
2478The following symbol flag values are available: C<IO::AIO::EFD_CLOEXEC>,
2479C<IO::AIO::EFD_NONBLOCK> and C<IO::AIO::EFD_SEMAPHORE> (Linux 2.6.30).
2480
2481Example: create a new eventfd filehandle:
2482
2483 $fh = IO::AIO::eventfd 0, IO::AIO::EFD_CLOEXEC
2484 or die "eventfd: $!\n";
2485
2486=item $fh = IO::AIO::timerfd_create $clockid[, $flags]
2487
2488This is a direct interface to the Linux L<timerfd_create(2)> system
2489call. The (unhelpful) default for C<$flags> is C<0>, but your default
2490should be C<IO::AIO::TFD_CLOEXEC>.
2491
2492On success, the new timerfd filehandle is returned, otherwise returns
2493C<undef>. If the timerfd_create syscall is missing, fails with C<ENOSYS>.
2494
2495Please refer to L<timerfd_create(2)> for more info on this call.
2496
2497The following C<$clockid> values are
2498available: C<IO::AIO::CLOCK_REALTIME>, C<IO::AIO::CLOCK_MONOTONIC>
2499C<IO::AIO::CLOCK_CLOCK_BOOTTIME> (Linux 3.15)
2500C<IO::AIO::CLOCK_CLOCK_REALTIME_ALARM> (Linux 3.11) and
2501C<IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM> (Linux 3.11).
2502
2503The following C<$flags> values are available (Linux
25042.6.27): C<IO::AIO::TFD_NONBLOCK> and C<IO::AIO::TFD_CLOEXEC>.
2505
2506Example: create a new timerfd and set it to one-second repeated alarms,
2507then wait for two alarms:
2508
2509 my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC
2510 or die "timerfd_create: $!\n";
2511
2512 defined IO::AIO::timerfd_settime $fh, 0, 1, 1
2513 or die "timerfd_settime: $!\n";
2514
2515 for (1..2) {
2516 8 == sysread $fh, my $buf, 8
2517 or die "timerfd read failure\n";
2518
2519 printf "number of expirations (likely 1): %d\n",
2520 unpack "Q", $buf;
2521 }
2522
2523=item ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
2524
2525This is a direct interface to the Linux L<timerfd_settime(2)> system
2526call. Please refer to its manpage for more info on this call.
2527
2528The new itimerspec is specified using two (possibly fractional) second
2529values, C<$new_interval> and C<$new_value>).
2530
2531On success, the current interval and value are returned (as per
2532C<timerfd_gettime>). On failure, the empty list is returned.
2533
2534The following C<$flags> values are
2535available: C<IO::AIO::TFD_TIMER_ABSTIME> and
2536C<IO::AIO::TFD_TIMER_CANCEL_ON_SET>.
2537
2538See C<IO::AIO::timerfd_create> for a full example.
2539
2540=item ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
2541
2542This is a direct interface to the Linux L<timerfd_gettime(2)> system
2543call. Please refer to its manpage for more info on this call.
2544
2545On success, returns the current values of interval and value for the given
2546timerfd (as potentially fractional second values). On failure, the empty
2547list is returned.
2548
2549=back
2550
1250=cut 2551=cut
1251 2552
1252# support function to convert a fd into a perl filehandle
1253sub _fd2fh {
1254 return undef if $_[0] < 0;
1255
1256 # try to generate nice filehandles
1257 my $sym = "IO::AIO::fd#$_[0]";
1258 local *$sym;
1259
1260 open *$sym, "+<&=$_[0]" # usually works under any unix
1261 or open *$sym, "<&=$_[0]" # cygwin needs this
1262 or open *$sym, ">&=$_[0]" # or this
1263 or return undef;
1264
1265 *$sym
1266}
1267
1268min_parallel 8; 2553min_parallel 8;
1269 2554
1270END { flush } 2555END { flush }
1271 2556
12721; 25571;
1273 2558
2559=head1 EVENT LOOP INTEGRATION
2560
2561It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO
2562automatically into many event loops:
2563
2564 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
2565 use AnyEvent::AIO;
2566
2567You can also integrate IO::AIO manually into many event loops, here are
2568some examples of how to do this:
2569
2570 # EV integration
2571 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
2572
2573 # Event integration
2574 Event->io (fd => IO::AIO::poll_fileno,
2575 poll => 'r',
2576 cb => \&IO::AIO::poll_cb);
2577
2578 # Glib/Gtk2 integration
2579 add_watch Glib::IO IO::AIO::poll_fileno,
2580 in => sub { IO::AIO::poll_cb; 1 };
2581
2582 # Tk integration
2583 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
2584 readable => \&IO::AIO::poll_cb);
2585
2586 # Danga::Socket integration
2587 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
2588 \&IO::AIO::poll_cb);
2589
1274=head2 FORK BEHAVIOUR 2590=head2 FORK BEHAVIOUR
1275 2591
1276This module should do "the right thing" when the process using it forks: 2592Usage of pthreads in a program changes the semantics of fork
2593considerably. Specifically, only async-safe functions can be called after
2594fork. Perl doesn't know about this, so in general, you cannot call fork
2595with defined behaviour in perl if pthreads are involved. IO::AIO uses
2596pthreads, so this applies, but many other extensions and (for inexplicable
2597reasons) perl itself often is linked against pthreads, so this limitation
2598applies to quite a lot of perls.
1277 2599
1278Before the fork, IO::AIO enters a quiescent state where no requests 2600This module no longer tries to fight your OS, or POSIX. That means IO::AIO
1279can be added in other threads and no results will be processed. After 2601only works in the process that loaded it. Forking is fully supported, but
1280the fork the parent simply leaves the quiescent state and continues 2602using IO::AIO in the child is not.
1281request/result processing, while the child frees the request/result queue
1282(so that the requests started before the fork will only be handled in the
1283parent). Threads will be started on demand until the limit set in the
1284parent process has been reached again.
1285 2603
1286In short: the parent will, after a short pause, continue as if fork had 2604You might get around by not I<using> IO::AIO before (or after)
1287not been called, while the child will act as if IO::AIO has not been used 2605forking. You could also try to call the L<IO::AIO::reinit> function in the
1288yet. 2606child:
2607
2608=over 4
2609
2610=item IO::AIO::reinit
2611
2612Abandons all current requests and I/O threads and simply reinitialises all
2613data structures. This is not an operation supported by any standards, but
2614happens to work on GNU/Linux and some newer BSD systems.
2615
2616The only reasonable use for this function is to call it after forking, if
2617C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in
2618the process will result in undefined behaviour. Calling it at any time
2619will also result in any undefined (by POSIX) behaviour.
2620
2621=back
2622
2623=head2 LINUX-SPECIFIC CALLS
2624
2625When a call is documented as "linux-specific" then this means it
2626originated on GNU/Linux. C<IO::AIO> will usually try to autodetect the
2627availability and compatibility of such calls regardless of the platform
2628it is compiled on, so platforms such as FreeBSD which often implement
2629these calls will work. When in doubt, call them and see if they fail wth
2630C<ENOSYS>.
1289 2631
1290=head2 MEMORY USAGE 2632=head2 MEMORY USAGE
1291 2633
1292Per-request usage: 2634Per-request usage:
1293 2635
… …
1306temporary buffers, and each thread requires a stack and other data 2648temporary buffers, and each thread requires a stack and other data
1307structures (usually around 16k-128k, depending on the OS). 2649structures (usually around 16k-128k, depending on the OS).
1308 2650
1309=head1 KNOWN BUGS 2651=head1 KNOWN BUGS
1310 2652
1311Known bugs will be fixed in the next release. 2653Known bugs will be fixed in the next release :)
2654
2655=head1 KNOWN ISSUES
2656
2657Calls that try to "import" foreign memory areas (such as C<IO::AIO::mmap>
2658or C<IO::AIO::aio_slurp>) do not work with generic lvalues, such as
2659non-created hash slots or other scalars I didn't think of. It's best to
2660avoid such and either use scalar variables or making sure that the scalar
2661exists (e.g. by storing C<undef>) and isn't "funny" (e.g. tied).
2662
2663I am not sure anything can be done about this, so this is considered a
2664known issue, rather than a bug.
1312 2665
1313=head1 SEE ALSO 2666=head1 SEE ALSO
1314 2667
1315L<Coro::AIO>. 2668L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a
2669more natural syntax.
1316 2670
1317=head1 AUTHOR 2671=head1 AUTHOR
1318 2672
1319 Marc Lehmann <schmorp@schmorp.de> 2673 Marc Lehmann <schmorp@schmorp.de>
1320 http://home.schmorp.de/ 2674 http://home.schmorp.de/

Diff Legend

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