ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/IO-AIO/README
(Generate patch)

Comparing IO-AIO/README (file contents):
Revision 1.13 by root, Tue Aug 30 15:45:10 2005 UTC vs.
Revision 1.63 by root, Mon Mar 4 10:28:38 2019 UTC

1NAME 1NAME
2 IO::AIO - Asynchronous Input/Output 2 IO::AIO - Asynchronous/Advanced Input/Output
3 3
4SYNOPSIS 4SYNOPSIS
5 use IO::AIO; 5 use IO::AIO;
6 6
7 aio_open "/etc/passwd", O_RDONLY, 0, sub { 7 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
8 my ($fh) = @_; 8 my $fh = shift
9 or die "/etc/passwd: $!";
9 ... 10 ...
10 }; 11 };
11 12
12 aio_unlink "/tmp/file", sub { }; 13 aio_unlink "/tmp/file", sub { };
13 14
14 aio_read $fh, 30000, 1024, $buffer, 0, sub { 15 aio_read $fh, 30000, 1024, $buffer, 0, sub {
15 $_[0] > 0 or die "read error: $!"; 16 $_[0] > 0 or die "read error: $!";
16 }; 17 };
17 18
18 # Event 19 # version 2+ has request and group objects
19 Event->io (fd => IO::AIO::poll_fileno, 20 use IO::AIO 2;
20 poll => 'r',
21 cb => \&IO::AIO::poll_cb);
22 21
23 # Glib/Gtk2 22 aioreq_pri 4; # give next request a very high priority
24 add_watch Glib::IO IO::AIO::poll_fileno, 23 my $req = aio_unlink "/tmp/file", sub { };
25 in => sub { IO::AIO::poll_cb; 1 }; 24 $req->cancel; # cancel request if still in queue
26 25
27 # Tk 26 my $grp = aio_group sub { print "all stats done\n" };
28 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "", 27 add $grp aio_stat "..." for ...;
29 readable => \&IO::AIO::poll_cb);
30
31 # Danga::Socket
32 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
33 \&IO::AIO::poll_cb);
34 28
35DESCRIPTION 29DESCRIPTION
36 This module implements asynchronous I/O using whatever means your 30 This module implements asynchronous I/O using whatever means your
37 operating system supports. 31 operating system supports. It is implemented as an interface to "libeio"
32 (<http://software.schmorp.de/pkg/libeio.html>).
38 33
34 Asynchronous means that operations that can normally block your program
35 (e.g. reading from disk) will be done asynchronously: the operation will
36 still block, but you can do something else in the meantime. This is
37 extremely useful for programs that need to stay interactive even when
38 doing heavy I/O (GUI programs, high performance network servers etc.),
39 but can also be used to easily do operations in parallel that are
40 normally done sequentially, e.g. stat'ing many files, which is much
41 faster on a RAID volume or over NFS when you do a number of stat
42 operations concurrently.
43
44 While most of this works on all types of file descriptors (for example
45 sockets), using these functions on file descriptors that support
46 nonblocking operation (again, sockets, pipes etc.) is very inefficient.
47 Use an event loop for that (such as the EV module): IO::AIO will
48 naturally fit into such an event loop itself.
49
39 Currently, a number of threads are started that execute your read/writes 50 In this version, a number of threads are started that execute your
40 and signal their completion. You don't need thread support in your libc 51 requests and signal their completion. You don't need thread support in
41 or perl, and the threads created by this module will not be visible to 52 perl, and the threads created by this module will not be visible to
42 the pthreads library. In the future, this module might make use of the 53 perl. In the future, this module might make use of the native aio
43 native aio functions available on many operating systems. However, they 54 functions available on many operating systems. However, they are often
44 are often not well-supported (Linux doesn't allow them on normal files 55 not well-supported or restricted (GNU/Linux doesn't allow them on normal
45 currently, for example), and they would only support aio_read and 56 files currently, for example), and they would only support aio_read and
46 aio_write, so the remaining functionality would have to be implemented 57 aio_write, so the remaining functionality would have to be implemented
47 using threads anyway. 58 using threads anyway.
48 59
60 In addition to asynchronous I/O, this module also exports some rather
61 arcane interfaces, such as "madvise" or linux's "splice" system call,
62 which is why the "A" in "AIO" can also mean *advanced*.
63
49 Although the module will work with in the presence of other threads, it 64 Although the module will work in the presence of other (Perl-) threads,
50 is currently not reentrant, so use appropriate locking yourself, always 65 it is currently not reentrant in any way, so use appropriate locking
51 call "poll_cb" from within the same thread, or never call "poll_cb" (or 66 yourself, always call "poll_cb" from within the same thread, or never
52 other "aio_" functions) recursively. 67 call "poll_cb" (or other "aio_" functions) recursively.
68
69 EXAMPLE
70 This is a simple example that uses the EV module and loads /etc/passwd
71 asynchronously:
72
73 use EV;
74 use IO::AIO;
75
76 # register the IO::AIO callback with EV
77 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
78
79 # queue the request to open /etc/passwd
80 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
81 my $fh = shift
82 or die "error while opening: $!";
83
84 # stat'ing filehandles is generally non-blocking
85 my $size = -s $fh;
86
87 # queue a request to read the file
88 my $contents;
89 aio_read $fh, 0, $size, $contents, 0, sub {
90 $_[0] == $size
91 or die "short read: $!";
92
93 close $fh;
94
95 # file contents now in $contents
96 print $contents;
97
98 # exit event loop and program
99 EV::break;
100 };
101 };
102
103 # possibly queue up other requests, or open GUI windows,
104 # check for sockets etc. etc.
105
106 # process events as long as there are some:
107 EV::run;
108
109REQUEST ANATOMY AND LIFETIME
110 Every "aio_*" function creates a request. which is a C data structure
111 not directly visible to Perl.
112
113 If called in non-void context, every request function returns a Perl
114 object representing the request. In void context, nothing is returned,
115 which saves a bit of memory.
116
117 The perl object is a fairly standard ref-to-hash object. The hash
118 contents are not used by IO::AIO so you are free to store anything you
119 like in it.
120
121 During their existance, aio requests travel through the following
122 states, in order:
123
124 ready
125 Immediately after a request is created it is put into the ready
126 state, waiting for a thread to execute it.
127
128 execute
129 A thread has accepted the request for processing and is currently
130 executing it (e.g. blocking in read).
131
132 pending
133 The request has been executed and is waiting for result processing.
134
135 While request submission and execution is fully asynchronous, result
136 processing is not and relies on the perl interpreter calling
137 "poll_cb" (or another function with the same effect).
138
139 result
140 The request results are processed synchronously by "poll_cb".
141
142 The "poll_cb" function will process all outstanding aio requests by
143 calling their callbacks, freeing memory associated with them and
144 managing any groups they are contained in.
145
146 done
147 Request has reached the end of its lifetime and holds no resources
148 anymore (except possibly for the Perl object, but its connection to
149 the actual aio request is severed and calling its methods will
150 either do nothing or result in a runtime error).
53 151
54FUNCTIONS 152FUNCTIONS
55 AIO FUNCTIONS 153 QUICK OVERVIEW
154 This section simply lists the prototypes most of the functions for quick
155 reference. See the following sections for function-by-function
156 documentation.
157
158 aio_wd $pathname, $callback->($wd)
159 aio_open $pathname, $flags, $mode, $callback->($fh)
160 aio_close $fh, $callback->($status)
161 aio_seek $fh,$offset,$whence, $callback->($offs)
162 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
163 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
164 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
165 aio_readahead $fh,$offset,$length, $callback->($retval)
166 aio_stat $fh_or_path, $callback->($status)
167 aio_lstat $fh, $callback->($status)
168 aio_statvfs $fh_or_path, $callback->($statvfs)
169 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
170 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
171 aio_chmod $fh_or_path, $mode, $callback->($status)
172 aio_truncate $fh_or_path, $offset, $callback->($status)
173 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
174 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
175 aio_unlink $pathname, $callback->($status)
176 aio_mknod $pathname, $mode, $dev, $callback->($status)
177 aio_link $srcpath, $dstpath, $callback->($status)
178 aio_symlink $srcpath, $dstpath, $callback->($status)
179 aio_readlink $pathname, $callback->($link)
180 aio_realpath $pathname, $callback->($path)
181 aio_rename $srcpath, $dstpath, $callback->($status)
182 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
183 aio_mkdir $pathname, $mode, $callback->($status)
184 aio_rmdir $pathname, $callback->($status)
185 aio_readdir $pathname, $callback->($entries)
186 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
187 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
188 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
189 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
190 aio_load $pathname, $data, $callback->($status)
191 aio_copy $srcpath, $dstpath, $callback->($status)
192 aio_move $srcpath, $dstpath, $callback->($status)
193 aio_rmtree $pathname, $callback->($status)
194 aio_fcntl $fh, $cmd, $arg, $callback->($status)
195 aio_ioctl $fh, $request, $buf, $callback->($status)
196 aio_sync $callback->($status)
197 aio_syncfs $fh, $callback->($status)
198 aio_fsync $fh, $callback->($status)
199 aio_fdatasync $fh, $callback->($status)
200 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
201 aio_pathsync $pathname, $callback->($status)
202 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
203 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
204 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
205 aio_mlockall $flags, $callback->($status)
206 aio_group $callback->(...)
207 aio_nop $callback->()
208
209 $prev_pri = aioreq_pri [$pri]
210 aioreq_nice $pri_adjust
211
212 IO::AIO::poll_wait
213 IO::AIO::poll_cb
214 IO::AIO::poll
215 IO::AIO::flush
216 IO::AIO::max_poll_reqs $nreqs
217 IO::AIO::max_poll_time $seconds
218 IO::AIO::min_parallel $nthreads
219 IO::AIO::max_parallel $nthreads
220 IO::AIO::max_idle $nthreads
221 IO::AIO::idle_timeout $seconds
222 IO::AIO::max_outstanding $maxreqs
223 IO::AIO::nreqs
224 IO::AIO::nready
225 IO::AIO::npending
226 $nfd = IO::AIO::get_fdlimit [EXPERIMENTAL]
227 IO::AIO::min_fdlimit $nfd [EXPERIMENTAL]
228
229 IO::AIO::sendfile $ofh, $ifh, $offset, $count
230 IO::AIO::fadvise $fh, $offset, $len, $advice
231 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
232 IO::AIO::munmap $scalar
233 IO::AIO::mremap $scalar, $new_length, $flags[, $new_address]
234 IO::AIO::madvise $scalar, $offset, $length, $advice
235 IO::AIO::mprotect $scalar, $offset, $length, $protect
236 IO::AIO::munlock $scalar, $offset = 0, $length = undef
237 IO::AIO::munlockall
238
239 API NOTES
56 All the "aio_*" calls are more or less thin wrappers around the syscall 240 All the "aio_*" calls are more or less thin wrappers around the syscall
57 with the same name (sans "aio_"). The arguments are similar or 241 with the same name (sans "aio_"). The arguments are similar or
58 identical, and they all accept an additional (and optional) $callback 242 identical, and they all accept an additional (and optional) $callback
59 argument which must be a code reference. This code reference will get 243 argument which must be a code reference. This code reference will be
60 called with the syscall return code (e.g. most syscalls return -1 on 244 called after the syscall has been executed in an asynchronous fashion.
61 error, unlike perl, which usually delivers "false") as it's sole 245 The results of the request will be passed as arguments to the callback
62 argument when the given syscall has been executed asynchronously. 246 (and, if an error occured, in $!) - for most requests the syscall return
247 code (e.g. most syscalls return -1 on error, unlike perl, which usually
248 delivers "false").
249
250 Some requests (such as "aio_readdir") pass the actual results and
251 communicate failures by passing "undef".
63 252
64 All functions expecting a filehandle keep a copy of the filehandle 253 All functions expecting a filehandle keep a copy of the filehandle
65 internally until the request has finished. 254 internally until the request has finished.
66 255
256 All functions return request objects of type IO::AIO::REQ that allow
257 further manipulation of those requests while they are in-flight.
258
67 The pathnames you pass to these routines *must* be absolute and encoded 259 The pathnames you pass to these routines *should* be absolute. The
68 in byte form. The reason for the former is that at the time the request 260 reason for this is that at the time the request is being executed, the
69 is being executed, the current working directory could have changed. 261 current working directory could have changed. Alternatively, you can
70 Alternatively, you can make sure that you never change the current 262 make sure that you never change the current working directory anywhere
71 working directory. 263 in the program and then use relative paths. You can also take advantage
264 of IO::AIOs working directory abstraction, that lets you specify paths
265 relative to some previously-opened "working directory object" - see the
266 description of the "IO::AIO::WD" class later in this document.
72 267
73 To encode pathnames to byte form, either make sure you either: a) always 268 To encode pathnames as octets, either make sure you either: a) always
74 pass in filenames you got from outside (command line, readdir etc.), b) 269 pass in filenames you got from outside (command line, readdir etc.)
75 are ASCII or ISO 8859-1, c) use the Encode module and encode your 270 without tinkering, b) are in your native filesystem encoding, c) use the
76 pathnames to the locale (or other) encoding in effect in the user 271 Encode module and encode your pathnames to the locale (or other)
77 environment, d) use Glib::filename_from_unicode on unicode filenames or 272 encoding in effect in the user environment, d) use
78 e) use something else. 273 Glib::filename_from_unicode on unicode filenames or e) use something
274 else to ensure your scalar has the correct contents.
275
276 This works, btw. independent of the internal UTF-8 bit, which IO::AIO
277 handles correctly whether it is set or not.
278
279 AIO REQUEST FUNCTIONS
280 $prev_pri = aioreq_pri [$pri]
281 Returns the priority value that would be used for the next request
282 and, if $pri is given, sets the priority for the next aio request.
283
284 The default priority is 0, the minimum and maximum priorities are -4
285 and 4, respectively. Requests with higher priority will be serviced
286 first.
287
288 The priority will be reset to 0 after each call to one of the
289 "aio_*" functions.
290
291 Example: open a file with low priority, then read something from it
292 with higher priority so the read request is serviced before other
293 low priority open requests (potentially spamming the cache):
294
295 aioreq_pri -3;
296 aio_open ..., sub {
297 return unless $_[0];
298
299 aioreq_pri -2;
300 aio_read $_[0], ..., sub {
301 ...
302 };
303 };
304
305 aioreq_nice $pri_adjust
306 Similar to "aioreq_pri", but subtracts the given value from the
307 current priority, so the effect is cumulative.
79 308
80 aio_open $pathname, $flags, $mode, $callback->($fh) 309 aio_open $pathname, $flags, $mode, $callback->($fh)
81 Asynchronously open or create a file and call the callback with a 310 Asynchronously open or create a file and call the callback with a
82 newly created filehandle for the file. 311 newly created filehandle for the file (or "undef" in case of an
312 error).
83 313
84 The pathname passed to "aio_open" must be absolute. See API NOTES, 314 The pathname passed to "aio_open" must be absolute. See API NOTES,
85 above, for an explanation. 315 above, for an explanation.
86 316
87 The $flags argument is a bitmask. See the "Fcntl" module for a list. 317 The $flags argument is a bitmask. See the "Fcntl" module for a list.
88 They are the same as used by "sysopen". 318 They are the same as used by "sysopen".
89 319
90 Likewise, $mode specifies the mode of the newly created file, if it 320 Likewise, $mode specifies the mode of the newly created file, if it
91 didn't exist and "O_CREAT" has been given, just like perl's 321 didn't exist and "O_CREAT" has been given, just like perl's
92 "sysopen", except that it is mandatory (i.e. use 0 if you don't 322 "sysopen", except that it is mandatory (i.e. use 0 if you don't
93 create new files, and 0666 or 0777 if you do). 323 create new files, and 0666 or 0777 if you do). Note that the $mode
324 will be modified by the umask in effect then the request is being
325 executed, so better never change the umask.
94 326
95 Example: 327 Example:
96 328
97 aio_open "/etc/passwd", O_RDONLY, 0, sub { 329 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
98 if ($_[0]) { 330 if ($_[0]) {
99 print "open successful, fh is $_[0]\n"; 331 print "open successful, fh is $_[0]\n";
100 ... 332 ...
101 } else { 333 } else {
102 die "open failed: $!\n"; 334 die "open failed: $!\n";
103 } 335 }
104 }; 336 };
105 337
338 In addition to all the common open modes/flags ("O_RDONLY",
339 "O_WRONLY", "O_RDWR", "O_CREAT", "O_TRUNC", "O_EXCL" and
340 "O_APPEND"), the following POSIX and non-POSIX constants are
341 available (missing ones on your system are, as usual, 0):
342
343 "O_ASYNC", "O_DIRECT", "O_NOATIME", "O_CLOEXEC", "O_NOCTTY",
344 "O_NOFOLLOW", "O_NONBLOCK", "O_EXEC", "O_SEARCH", "O_DIRECTORY",
345 "O_DSYNC", "O_RSYNC", "O_SYNC", "O_PATH", "O_TMPFILE", "O_TTY_INIT"
346 and "O_ACCMODE".
347
106 aio_close $fh, $callback->($status) 348 aio_close $fh, $callback->($status)
107 Asynchronously close a file and call the callback with the result 349 Asynchronously close a file and call the callback with the result
108 code. *WARNING:* although accepted, you should not pass in a perl 350 code.
109 filehandle here, as perl will likely close the file descriptor
110 another time when the filehandle is destroyed. Normally, you can
111 safely call perls "close" or just let filehandles go out of scope.
112 351
113 This is supposed to be a bug in the API, so that might change. It's 352 Unfortunately, you can't do this to perl. Perl *insists* very
114 therefore best to avoid this function. 353 strongly on closing the file descriptor associated with the
354 filehandle itself.
355
356 Therefore, "aio_close" will not close the filehandle - instead it
357 will use dup2 to overwrite the file descriptor with the write-end of
358 a pipe (the pipe fd will be created on demand and will be cached).
359
360 Or in other words: the file descriptor will be closed, but it will
361 not be free for reuse until the perl filehandle is closed.
362
363 aio_seek $fh, $offset, $whence, $callback->($offs)
364 Seeks the filehandle to the new $offset, similarly to perl's
365 "sysseek". The $whence can use the traditional values (0 for
366 "IO::AIO::SEEK_SET", 1 for "IO::AIO::SEEK_CUR" or 2 for
367 "IO::AIO::SEEK_END").
368
369 The resulting absolute offset will be passed to the callback, or -1
370 in case of an error.
371
372 In theory, the $whence constants could be different than the
373 corresponding values from Fcntl, but perl guarantees they are the
374 same, so don't panic.
375
376 As a GNU/Linux (and maybe Solaris) extension, also the constants
377 "IO::AIO::SEEK_DATA" and "IO::AIO::SEEK_HOLE" are available, if they
378 could be found. No guarantees about suitability for use in
379 "aio_seek" or Perl's "sysseek" can be made though, although I would
380 naively assume they "just work".
115 381
116 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 382 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
117 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 383 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
118 Reads or writes "length" bytes from the specified "fh" and "offset" 384 Reads or writes $length bytes from or to the specified $fh and
119 into the scalar given by "data" and offset "dataoffset" and calls 385 $offset into the scalar given by $data and offset $dataoffset and
120 the callback without the actual number of bytes read (or -1 on 386 calls the callback with the actual number of bytes transferred (or
121 error, just like the syscall). 387 -1 on error, just like the syscall).
388
389 "aio_read" will, like "sysread", shrink or grow the $data scalar to
390 offset plus the actual number of bytes read.
391
392 If $offset is undefined, then the current file descriptor offset
393 will be used (and updated), otherwise the file descriptor offset
394 will not be changed by these calls.
395
396 If $length is undefined in "aio_write", use the remaining length of
397 $data.
398
399 If $dataoffset is less than zero, it will be counted from the end of
400 $data.
122 401
123 The $data scalar *MUST NOT* be modified in any way while the request 402 The $data scalar *MUST NOT* be modified in any way while the request
124 is outstanding. Modifying it can result in segfaults or WW3 (if the 403 is outstanding. Modifying it can result in segfaults or World War
125 necessary/optional hardware is installed). 404 III (if the necessary/optional hardware is installed).
126 405
127 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at 406 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at
128 offset 0 within the scalar: 407 offset 0 within the scalar:
129 408
130 aio_read $fh, 7, 15, $buffer, 0, sub { 409 aio_read $fh, 7, 15, $buffer, 0, sub {
… …
135 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 414 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
136 Tries to copy $length bytes from $in_fh to $out_fh. It starts 415 Tries to copy $length bytes from $in_fh to $out_fh. It starts
137 reading at byte offset $in_offset, and starts writing at the current 416 reading at byte offset $in_offset, and starts writing at the current
138 file offset of $out_fh. Because of that, it is not safe to issue 417 file offset of $out_fh. Because of that, it is not safe to issue
139 more than one "aio_sendfile" per $out_fh, as they will interfere 418 more than one "aio_sendfile" per $out_fh, as they will interfere
140 with each other. 419 with each other. The same $in_fh works fine though, as this function
420 does not move or use the file offset of $in_fh.
141 421
422 Please note that "aio_sendfile" can read more bytes from $in_fh than
423 are written, and there is no way to find out how many more bytes
424 have been read from "aio_sendfile" alone, as "aio_sendfile" only
425 provides the number of bytes written to $out_fh. Only if the result
426 value equals $length one can assume that $length bytes have been
427 read.
428
429 Unlike with other "aio_" functions, it makes a lot of sense to use
430 "aio_sendfile" on non-blocking sockets, as long as one end
431 (typically the $in_fh) is a file - the file I/O will then be
432 asynchronous, while the socket I/O will be non-blocking. Note,
433 however, that you can run into a trap where "aio_sendfile" reads
434 some data with readahead, then fails to write all data, and when the
435 socket is ready the next time, the data in the cache is already
436 lost, forcing "aio_sendfile" to again hit the disk. Explicit
437 "aio_read" + "aio_write" let's you better control resource usage.
438
142 This call tries to make use of a native "sendfile" syscall to 439 This call tries to make use of a native "sendfile"-like syscall to
143 provide zero-copy operation. For this to work, $out_fh should refer 440 provide zero-copy operation. For this to work, $out_fh should refer
144 to a socket, and $in_fh should refer to mmap'able file. 441 to a socket, and $in_fh should refer to an mmap'able file.
145 442
146 If the native sendfile call fails or is not implemented, it will be 443 If a native sendfile cannot be found or it fails with "ENOSYS",
147 emulated, so you can call "aio_sendfile" on any type of filehandle 444 "EINVAL", "ENOTSUP", "EOPNOTSUPP", "EAFNOSUPPORT", "EPROTOTYPE" or
148 regardless of the limitations of the operating system. 445 "ENOTSOCK", it will be emulated, so you can call "aio_sendfile" on
446 any type of filehandle regardless of the limitations of the
447 operating system.
149 448
150 Please note, however, that "aio_sendfile" can read more bytes from 449 As native sendfile syscalls (as practically any non-POSIX interface
151 $in_fh than are written, and there is no way to find out how many 450 hacked together in a hurry to improve benchmark numbers) tend to be
152 bytes have been read from "aio_sendfile" alone, as "aio_sendfile" 451 rather buggy on many systems, this implementation tries to work
153 only provides the number of bytes written to $out_fh. Only if the 452 around some known bugs in Linux and FreeBSD kernels (probably
154 result value equals $length one can assume that $length bytes have 453 others, too), but that might fail, so you really really should check
155 been read. 454 the return value of "aio_sendfile" - fewer bytes than expected might
455 have been transferred.
156 456
157 aio_readahead $fh,$offset,$length, $callback->($retval) 457 aio_readahead $fh,$offset,$length, $callback->($retval)
158 "aio_readahead" populates the page cache with data from a file so 458 "aio_readahead" populates the page cache with data from a file so
159 that subsequent reads from that file will not block on disk I/O. The 459 that subsequent reads from that file will not block on disk I/O. The
160 $offset argument specifies the starting point from which data is to 460 $offset argument specifies the starting point from which data is to
… …
163 to a page boundary and bytes are read up to the next page boundary 463 to a page boundary and bytes are read up to the next page boundary
164 greater than or equal to (off-set+length). "aio_readahead" does not 464 greater than or equal to (off-set+length). "aio_readahead" does not
165 read beyond the end of the file. The current file offset of the file 465 read beyond the end of the file. The current file offset of the file
166 is left unchanged. 466 is left unchanged.
167 467
168 If that syscall doesn't exist (likely if your OS isn't Linux) it 468 If that syscall doesn't exist (likely if your kernel isn't Linux) it
169 will be emulated by simply reading the data, which would have a 469 will be emulated by simply reading the data, which would have a
170 similar effect. 470 similar effect.
171 471
172 aio_stat $fh_or_path, $callback->($status) 472 aio_stat $fh_or_path, $callback->($status)
173 aio_lstat $fh, $callback->($status) 473 aio_lstat $fh, $callback->($status)
174 Works like perl's "stat" or "lstat" in void context. The callback 474 Works almost exactly like perl's "stat" or "lstat" in void context.
175 will be called after the stat and the results will be available 475 The callback will be called after the stat and the results will be
176 using "stat _" or "-s _" etc... 476 available using "stat _" or "-s _" and other tests (with the
477 exception of "-B" and "-T").
177 478
178 The pathname passed to "aio_stat" must be absolute. See API NOTES, 479 The pathname passed to "aio_stat" must be absolute. See API NOTES,
179 above, for an explanation. 480 above, for an explanation.
180 481
181 Currently, the stats are always 64-bit-stats, i.e. instead of 482 Currently, the stats are always 64-bit-stats, i.e. instead of
182 returning an error when stat'ing a large file, the results will be 483 returning an error when stat'ing a large file, the results will be
183 silently truncated unless perl itself is compiled with large file 484 silently truncated unless perl itself is compiled with large file
184 support. 485 support.
185 486
487 To help interpret the mode and dev/rdev stat values, IO::AIO offers
488 the following constants and functions (if not implemented, the
489 constants will be 0 and the functions will either "croak" or fall
490 back on traditional behaviour).
491
492 "S_IFMT", "S_IFIFO", "S_IFCHR", "S_IFBLK", "S_IFLNK", "S_IFREG",
493 "S_IFDIR", "S_IFWHT", "S_IFSOCK", "IO::AIO::major $dev_t",
494 "IO::AIO::minor $dev_t", "IO::AIO::makedev $major, $minor".
495
496 To access higher resolution stat timestamps, see "SUBSECOND STAT
497 TIME ACCESS".
498
186 Example: Print the length of /etc/passwd: 499 Example: Print the length of /etc/passwd:
187 500
188 aio_stat "/etc/passwd", sub { 501 aio_stat "/etc/passwd", sub {
189 $_[0] and die "stat failed: $!"; 502 $_[0] and die "stat failed: $!";
190 print "size is ", -s _, "\n"; 503 print "size is ", -s _, "\n";
191 }; 504 };
192 505
506 aio_statvfs $fh_or_path, $callback->($statvfs)
507 Works like the POSIX "statvfs" or "fstatvfs" syscalls, depending on
508 whether a file handle or path was passed.
509
510 On success, the callback is passed a hash reference with the
511 following members: "bsize", "frsize", "blocks", "bfree", "bavail",
512 "files", "ffree", "favail", "fsid", "flag" and "namemax". On
513 failure, "undef" is passed.
514
515 The following POSIX IO::AIO::ST_* constants are defined: "ST_RDONLY"
516 and "ST_NOSUID".
517
518 The following non-POSIX IO::AIO::ST_* flag masks are defined to
519 their correct value when available, or to 0 on systems that do not
520 support them: "ST_NODEV", "ST_NOEXEC", "ST_SYNCHRONOUS",
521 "ST_MANDLOCK", "ST_WRITE", "ST_APPEND", "ST_IMMUTABLE",
522 "ST_NOATIME", "ST_NODIRATIME" and "ST_RELATIME".
523
524 Example: stat "/wd" and dump out the data if successful.
525
526 aio_statvfs "/wd", sub {
527 my $f = $_[0]
528 or die "statvfs: $!";
529
530 use Data::Dumper;
531 say Dumper $f;
532 };
533
534 # result:
535 {
536 bsize => 1024,
537 bfree => 4333064312,
538 blocks => 10253828096,
539 files => 2050765568,
540 flag => 4096,
541 favail => 2042092649,
542 bavail => 4333064312,
543 ffree => 2042092649,
544 namemax => 255,
545 frsize => 1024,
546 fsid => 1810
547 }
548
549 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
550 Works like perl's "utime" function (including the special case of
551 $atime and $mtime being undef). Fractional times are supported if
552 the underlying syscalls support them.
553
554 When called with a pathname, uses utimensat(2) or utimes(2) if
555 available, otherwise utime(2). If called on a file descriptor, uses
556 futimens(2) or futimes(2) if available, otherwise returns ENOSYS, so
557 this is not portable.
558
559 Examples:
560
561 # set atime and mtime to current time (basically touch(1)):
562 aio_utime "path", undef, undef;
563 # set atime to current time and mtime to beginning of the epoch:
564 aio_utime "path", time, undef; # undef==0
565
566 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
567 Works like perl's "chown" function, except that "undef" for either
568 $uid or $gid is being interpreted as "do not change" (but -1 can
569 also be used).
570
571 Examples:
572
573 # same as "chown root path" in the shell:
574 aio_chown "path", 0, -1;
575 # same as above:
576 aio_chown "path", 0, undef;
577
578 aio_truncate $fh_or_path, $offset, $callback->($status)
579 Works like truncate(2) or ftruncate(2).
580
581 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
582 Allocates or frees disk space according to the $mode argument. See
583 the linux "fallocate" documentation for details.
584
585 $mode is usually 0 or "IO::AIO::FALLOC_FL_KEEP_SIZE" to allocate
586 space, or "IO::AIO::FALLOC_FL_PUNCH_HOLE |
587 IO::AIO::FALLOC_FL_KEEP_SIZE", to deallocate a file range.
588
589 IO::AIO also supports "FALLOC_FL_COLLAPSE_RANGE", to remove a range
590 (without leaving a hole), "FALLOC_FL_ZERO_RANGE", to zero a range,
591 "FALLOC_FL_INSERT_RANGE" to insert a range and
592 "FALLOC_FL_UNSHARE_RANGE" to unshare shared blocks (see your
593 fallocate(2) manpage).
594
595 The file system block size used by "fallocate" is presumably the
596 "f_bsize" returned by "statvfs", but different filesystems and
597 filetypes can dictate other limitations.
598
599 If "fallocate" isn't available or cannot be emulated (currently no
600 emulation will be attempted), passes -1 and sets $! to "ENOSYS".
601
602 aio_chmod $fh_or_path, $mode, $callback->($status)
603 Works like perl's "chmod" function.
604
193 aio_unlink $pathname, $callback->($status) 605 aio_unlink $pathname, $callback->($status)
194 Asynchronously unlink (delete) a file and call the callback with the 606 Asynchronously unlink (delete) a file and call the callback with the
195 result code. 607 result code.
196 608
609 aio_mknod $pathname, $mode, $dev, $callback->($status)
610 [EXPERIMENTAL]
611
612 Asynchronously create a device node (or fifo). See mknod(2).
613
614 The only (POSIX-) portable way of calling this function is:
615
616 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
617
618 See "aio_stat" for info about some potentially helpful extra
619 constants and functions.
620
621 aio_link $srcpath, $dstpath, $callback->($status)
622 Asynchronously create a new link to the existing object at $srcpath
623 at the path $dstpath and call the callback with the result code.
624
625 aio_symlink $srcpath, $dstpath, $callback->($status)
626 Asynchronously create a new symbolic link to the existing object at
627 $srcpath at the path $dstpath and call the callback with the result
628 code.
629
630 aio_readlink $pathname, $callback->($link)
631 Asynchronously read the symlink specified by $path and pass it to
632 the callback. If an error occurs, nothing or undef gets passed to
633 the callback.
634
635 aio_realpath $pathname, $callback->($path)
636 Asynchronously make the path absolute and resolve any symlinks in
637 $path. The resulting path only consists of directories (same as
638 Cwd::realpath).
639
640 This request can be used to get the absolute path of the current
641 working directory by passing it a path of . (a single dot).
642
643 aio_rename $srcpath, $dstpath, $callback->($status)
644 Asynchronously rename the object at $srcpath to $dstpath, just as
645 rename(2) and call the callback with the result code.
646
647 On systems that support the AIO::WD working directory abstraction
648 natively, the case "[$wd, "."]" as $srcpath is specialcased -
649 instead of failing, "rename" is called on the absolute path of $wd.
650
651 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
652 Basically a version of "aio_rename" with an additional $flags
653 argument. Calling this with "$flags=0" is the same as calling
654 "aio_rename".
655
656 Non-zero flags are currently only supported on GNU/Linux systems
657 that support renameat2. Other systems fail with "ENOSYS" in this
658 case.
659
660 The following constants are available (missing ones are, as usual
661 0), see renameat2(2) for details:
662
663 "IO::AIO::RENAME_NOREPLACE", "IO::AIO::RENAME_EXCHANGE" and
664 "IO::AIO::RENAME_WHITEOUT".
665
666 aio_mkdir $pathname, $mode, $callback->($status)
667 Asynchronously mkdir (create) a directory and call the callback with
668 the result code. $mode will be modified by the umask at the time the
669 request is executed, so do not change your umask.
670
197 aio_rmdir $pathname, $callback->($status) 671 aio_rmdir $pathname, $callback->($status)
198 Asynchronously rmdir (delete) a directory and call the callback with 672 Asynchronously rmdir (delete) a directory and call the callback with
199 the result code. 673 the result code.
200 674
675 On systems that support the AIO::WD working directory abstraction
676 natively, the case "[$wd, "."]" is specialcased - instead of
677 failing, "rmdir" is called on the absolute path of $wd.
678
201 aio_readdir $pathname $callback->($entries) 679 aio_readdir $pathname, $callback->($entries)
202 Unlike the POSIX call of the same name, "aio_readdir" reads an 680 Unlike the POSIX call of the same name, "aio_readdir" reads an
203 entire directory (i.e. opendir + readdir + closedir). The entries 681 entire directory (i.e. opendir + readdir + closedir). The entries
204 will not be sorted, and will NOT include the "." and ".." entries. 682 will not be sorted, and will NOT include the "." and ".." entries.
205 683
206 The callback a single argument which is either "undef" or an 684 The callback is passed a single argument which is either "undef" or
207 array-ref with the filenames. 685 an array-ref with the filenames.
208 686
687 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
688 Quite similar to "aio_readdir", but the $flags argument allows one
689 to tune behaviour and output format. In case of an error, $entries
690 will be "undef".
691
692 The flags are a combination of the following constants, ORed
693 together (the flags will also be passed to the callback, possibly
694 modified):
695
696 IO::AIO::READDIR_DENTS
697 Normally the callback gets an arrayref consisting of names only
698 (as with "aio_readdir"). If this flag is set, then the callback
699 gets an arrayref with "[$name, $type, $inode]" arrayrefs, each
700 describing a single directory entry in more detail:
701
702 $name is the name of the entry.
703
704 $type is one of the "IO::AIO::DT_xxx" constants:
705
706 "IO::AIO::DT_UNKNOWN", "IO::AIO::DT_FIFO", "IO::AIO::DT_CHR",
707 "IO::AIO::DT_DIR", "IO::AIO::DT_BLK", "IO::AIO::DT_REG",
708 "IO::AIO::DT_LNK", "IO::AIO::DT_SOCK", "IO::AIO::DT_WHT".
709
710 "IO::AIO::DT_UNKNOWN" means just that: readdir does not know. If
711 you need to know, you have to run stat yourself. Also, for
712 speed/memory reasons, the $type scalars are read-only: you must
713 not modify them.
714
715 $inode is the inode number (which might not be exact on systems
716 with 64 bit inode numbers and 32 bit perls). This field has
717 unspecified content on systems that do not deliver the inode
718 information.
719
720 IO::AIO::READDIR_DIRS_FIRST
721 When this flag is set, then the names will be returned in an
722 order where likely directories come first, in optimal stat
723 order. This is useful when you need to quickly find directories,
724 or you want to find all directories while avoiding to stat()
725 each entry.
726
727 If the system returns type information in readdir, then this is
728 used to find directories directly. Otherwise, likely directories
729 are names beginning with ".", or otherwise names with no dots,
730 of which names with short names are tried first.
731
732 IO::AIO::READDIR_STAT_ORDER
733 When this flag is set, then the names will be returned in an
734 order suitable for stat()'ing each one. That is, when you plan
735 to stat() most or all files in the given directory, then the
736 returned order will likely be faster.
737
738 If both this flag and "IO::AIO::READDIR_DIRS_FIRST" are
739 specified, then the likely dirs come first, resulting in a less
740 optimal stat order for stat'ing all entries, but likely a more
741 optimal order for finding subdirectories.
742
743 IO::AIO::READDIR_FOUND_UNKNOWN
744 This flag should not be set when calling "aio_readdirx".
745 Instead, it is being set by "aio_readdirx", when any of the
746 $type's found were "IO::AIO::DT_UNKNOWN". The absence of this
747 flag therefore indicates that all $type's are known, which can
748 be used to speed up some algorithms.
749
750 aio_slurp $pathname, $offset, $length, $data, $callback->($status)
751 Opens, reads and closes the given file. The data is put into $data,
752 which is resized as required.
753
754 If $offset is negative, then it is counted from the end of the file.
755
756 If $length is zero, then the remaining length of the file is used.
757 Also, in this case, the same limitations to modifying $data apply as
758 when IO::AIO::mmap is used, i.e. it must only be modified in-place
759 with "substr". If the size of the file is known, specifying a
760 non-zero $length results in a performance advantage.
761
762 This request is similar to the older "aio_load" request, but since
763 it is a single request, it might be more efficient to use.
764
765 Example: load /etc/passwd into $passwd.
766
767 my $passwd;
768 aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
769 $_[0] >= 0
770 or die "/etc/passwd: $!\n";
771
772 printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
773 print $passwd;
774 };
775 IO::AIO::flush;
776
777 aio_load $pathname, $data, $callback->($status)
778 This is a composite request that tries to fully load the given file
779 into memory. Status is the same as with aio_read.
780
781 Using "aio_slurp" might be more efficient, as it is a single
782 request.
783
784 aio_copy $srcpath, $dstpath, $callback->($status)
785 Try to copy the *file* (directories not supported as either source
786 or destination) from $srcpath to $dstpath and call the callback with
787 a status of 0 (ok) or -1 (error, see $!).
788
789 Existing destination files will be truncated.
790
791 This is a composite request that creates the destination file with
792 mode 0200 and copies the contents of the source file into it using
793 "aio_sendfile", followed by restoring atime, mtime, access mode and
794 uid/gid, in that order.
795
796 If an error occurs, the partial destination file will be unlinked,
797 if possible, except when setting atime, mtime, access mode and
798 uid/gid, where errors are being ignored.
799
800 aio_move $srcpath, $dstpath, $callback->($status)
801 Try to move the *file* (directories not supported as either source
802 or destination) from $srcpath to $dstpath and call the callback with
803 a status of 0 (ok) or -1 (error, see $!).
804
805 This is a composite request that tries to rename(2) the file first;
806 if rename fails with "EXDEV", it copies the file with "aio_copy"
807 and, if that is successful, unlinks the $srcpath.
808
209 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 809 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
210 Scans a directory (similar to "aio_readdir") and tries to separate 810 Scans a directory (similar to "aio_readdir") but additionally tries
211 the entries of directory $path into two sets of names, ones you can 811 to efficiently separate the entries of directory $path into two sets
212 recurse into (directories), and ones you cannot recurse into 812 of names, directories you can recurse into (directories), and ones
213 (everything else). 813 you cannot recurse into (everything else, including symlinks to
814 directories).
214 815
215 "aio_scandir" is a composite request that consists of many 816 "aio_scandir" is a composite request that generates many sub
216 aio-primitives. $maxreq specifies the maximum number of outstanding 817 requests. $maxreq specifies the maximum number of outstanding aio
217 aio requests that this function generates. If it is "<= 0", then a 818 requests that this function generates. If it is "<= 0", then a
218 suitable default will be chosen (currently 8). 819 suitable default will be chosen (currently 4).
219 820
220 On error, the callback is called without arguments, otherwise it 821 On error, the callback is called without arguments, otherwise it
221 receives two array-refs with path-relative entry names. 822 receives two array-refs with path-relative entry names.
222 823
223 Example: 824 Example:
… …
231 Implementation notes. 832 Implementation notes.
232 833
233 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry 834 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry
234 can. 835 can.
235 836
837 If readdir returns file type information, then this is used directly
838 to find directories.
839
236 After reading the directory, the modification time, size etc. of the 840 Otherwise, after reading the directory, the modification time, size
237 directory before and after the readdir is checked, and if they 841 etc. of the directory before and after the readdir is checked, and
238 match, the link count will be used to decide how many entries are 842 if they match (and isn't the current time), the link count will be
239 directories (if >= 2). Otherwise, no knowledge of the number of 843 used to decide how many entries are directories (if >= 2).
240 subdirectories will be assumed. 844 Otherwise, no knowledge of the number of subdirectories will be
845 assumed.
241 846
242 Then entires will be sorted into likely directories (everything 847 Then entries will be sorted into likely directories a non-initial
243 without a non-initial dot) and likely non-directories (everything 848 dot currently) and likely non-directories (see "aio_readdirx"). Then
244 else). Then every entry + "/." will be "stat"'ed, likely directories 849 every entry plus an appended "/." will be "stat"'ed, likely
245 first. This is often faster because filesystems might detect the 850 directories first, in order of their inode numbers. If that
851 succeeds, it assumes that the entry is a directory or a symlink to
852 directory (which will be checked separately). This is often faster
853 than stat'ing the entry itself because filesystems might detect the
246 type of the entry without reading the inode data (e.g. ext2s 854 type of the entry without reading the inode data (e.g. ext2fs
247 filetype feature). If that succeeds, it assumes that the entry is a 855 filetype feature), even on systems that cannot return the filetype
248 directory or a symlink to directory (which will be checked 856 information on readdir.
249 seperately).
250 857
251 If the known number of directories has been reached, the rest of the 858 If the known number of directories (link count - 2) has been
252 entries is assumed to be non-directories. 859 reached, the rest of the entries is assumed to be non-directories.
860
861 This only works with certainty on POSIX (= UNIX) filesystems, which
862 fortunately are the vast majority of filesystems around.
863
864 It will also likely work on non-POSIX filesystems with reduced
865 efficiency as those tend to return 0 or 1 as link counts, which
866 disables the directory counting heuristic.
867
868 aio_rmtree $pathname, $callback->($status)
869 Delete a directory tree starting (and including) $path, return the
870 status of the final "rmdir" only. This is a composite request that
871 uses "aio_scandir" to recurse into and rmdir directories, and unlink
872 everything else.
873
874 aio_fcntl $fh, $cmd, $arg, $callback->($status)
875 aio_ioctl $fh, $request, $buf, $callback->($status)
876 These work just like the "fcntl" and "ioctl" built-in functions,
877 except they execute asynchronously and pass the return value to the
878 callback.
879
880 Both calls can be used for a lot of things, some of which make more
881 sense to run asynchronously in their own thread, while some others
882 make less sense. For example, calls that block waiting for external
883 events, such as locking, will also lock down an I/O thread while it
884 is waiting, which can deadlock the whole I/O system. At the same
885 time, there might be no alternative to using a thread to wait.
886
887 So in general, you should only use these calls for things that do
888 (filesystem) I/O, not for things that wait for other events
889 (network, other processes), although if you are careful and know
890 what you are doing, you still can.
891
892 The following constants are available (missing ones are, as usual
893 0):
894
895 "F_DUPFD_CLOEXEC",
896
897 "F_OFD_GETLK", "F_OFD_SETLK", "F_OFD_GETLKW",
898
899 "FIFREEZE", "FITHAW", "FITRIM", "FICLONE", "FICLONERANGE",
900 "FIDEDUPERANGE".
901
902 "FS_IOC_GETFLAGS", "FS_IOC_SETFLAGS", "FS_IOC_GETVERSION",
903 "FS_IOC_SETVERSION", "FS_IOC_FIEMAP".
904
905 "FS_IOC_FSGETXATTR", "FS_IOC_FSSETXATTR",
906 "FS_IOC_SET_ENCRYPTION_POLICY", "FS_IOC_GET_ENCRYPTION_PWSALT",
907 "FS_IOC_GET_ENCRYPTION_POLICY", "FS_KEY_DESCRIPTOR_SIZE".
908
909 "FS_SECRM_FL", "FS_UNRM_FL", "FS_COMPR_FL", "FS_SYNC_FL",
910 "FS_IMMUTABLE_FL", "FS_APPEND_FL", "FS_NODUMP_FL", "FS_NOATIME_FL",
911 "FS_DIRTY_FL", "FS_COMPRBLK_FL", "FS_NOCOMP_FL", "FS_ENCRYPT_FL",
912 "FS_BTREE_FL", "FS_INDEX_FL", "FS_JOURNAL_DATA_FL", "FS_NOTAIL_FL",
913 "FS_DIRSYNC_FL", "FS_TOPDIR_FL", "FS_FL_USER_MODIFIABLE".
914
915 "FS_XFLAG_REALTIME", "FS_XFLAG_PREALLOC", "FS_XFLAG_IMMUTABLE",
916 "FS_XFLAG_APPEND", "FS_XFLAG_SYNC", "FS_XFLAG_NOATIME",
917 "FS_XFLAG_NODUMP", "FS_XFLAG_RTINHERIT", "FS_XFLAG_PROJINHERIT",
918 "FS_XFLAG_NOSYMLINKS", "FS_XFLAG_EXTSIZE", "FS_XFLAG_EXTSZINHERIT",
919 "FS_XFLAG_NODEFRAG", "FS_XFLAG_FILESTREAM", "FS_XFLAG_DAX",
920 "FS_XFLAG_HASATTR",
921
922 aio_sync $callback->($status)
923 Asynchronously call sync and call the callback when finished.
253 924
254 aio_fsync $fh, $callback->($status) 925 aio_fsync $fh, $callback->($status)
255 Asynchronously call fsync on the given filehandle and call the 926 Asynchronously call fsync on the given filehandle and call the
256 callback with the fsync result code. 927 callback with the fsync result code.
257 928
… …
260 callback with the fdatasync result code. 931 callback with the fdatasync result code.
261 932
262 If this call isn't available because your OS lacks it or it couldn't 933 If this call isn't available because your OS lacks it or it couldn't
263 be detected, it will be emulated by calling "fsync" instead. 934 be detected, it will be emulated by calling "fsync" instead.
264 935
936 aio_syncfs $fh, $callback->($status)
937 Asynchronously call the syncfs syscall to sync the filesystem
938 associated to the given filehandle and call the callback with the
939 syncfs result code. If syncfs is not available, calls sync(), but
940 returns -1 and sets errno to "ENOSYS" nevertheless.
941
942 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
943 Sync the data portion of the file specified by $offset and $length
944 to disk (but NOT the metadata), by calling the Linux-specific
945 sync_file_range call. If sync_file_range is not available or it
946 returns ENOSYS, then fdatasync or fsync is being substituted.
947
948 $flags can be a combination of
949 "IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE",
950 "IO::AIO::SYNC_FILE_RANGE_WRITE" and
951 "IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER": refer to the sync_file_range
952 manpage for details.
953
954 aio_pathsync $pathname, $callback->($status)
955 This request tries to open, fsync and close the given path. This is
956 a composite request intended to sync directories after directory
957 operations (E.g. rename). This might not work on all operating
958 systems or have any specific effect, but usually it makes sure that
959 directory changes get written to disc. It works for anything that
960 can be opened for read-only, not just directories.
961
962 Future versions of this function might fall back to other methods
963 when "fsync" on the directory fails (such as calling "sync").
964
965 Passes 0 when everything went ok, and -1 on error.
966
967 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC,
968 $callback->($status)
969 This is a rather advanced IO::AIO call, which only works on
970 mmap(2)ed scalars (see the "IO::AIO::mmap" function, although it
971 also works on data scalars managed by the Sys::Mmap or Mmap modules,
972 note that the scalar must only be modified in-place while an aio
973 operation is pending on it).
974
975 It calls the "msync" function of your OS, if available, with the
976 memory area starting at $offset in the string and ending $length
977 bytes later. If $length is negative, counts from the end, and if
978 $length is "undef", then it goes till the end of the string. The
979 flags can be either "IO::AIO::MS_ASYNC" or "IO::AIO::MS_SYNC", plus
980 an optional "IO::AIO::MS_INVALIDATE".
981
982 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0,
983 $callback->($status)
984 This is a rather advanced IO::AIO call, which works best on
985 mmap(2)ed scalars.
986
987 It touches (reads or writes) all memory pages in the specified range
988 inside the scalar. All caveats and parameters are the same as for
989 "aio_msync", above, except for flags, which must be either 0 (which
990 reads all pages and ensures they are instantiated) or
991 "IO::AIO::MT_MODIFY", which modifies the memory pages (by reading
992 and writing an octet from it, which dirties the page).
993
994 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
995 This is a rather advanced IO::AIO call, which works best on
996 mmap(2)ed scalars.
997
998 It reads in all the pages of the underlying storage into memory (if
999 any) and locks them, so they are not getting swapped/paged out or
1000 removed.
1001
1002 If $length is undefined, then the scalar will be locked till the
1003 end.
1004
1005 On systems that do not implement "mlock", this function returns -1
1006 and sets errno to "ENOSYS".
1007
1008 Note that the corresponding "munlock" is synchronous and is
1009 documented under "MISCELLANEOUS FUNCTIONS".
1010
1011 Example: open a file, mmap and mlock it - both will be undone when
1012 $data gets destroyed.
1013
1014 open my $fh, "<", $path or die "$path: $!";
1015 my $data;
1016 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1017 aio_mlock $data; # mlock in background
1018
1019 aio_mlockall $flags, $callback->($status)
1020 Calls the "mlockall" function with the given $flags (a combination
1021 of "IO::AIO::MCL_CURRENT", "IO::AIO::MCL_FUTURE" and
1022 "IO::AIO::MCL_ONFAULT").
1023
1024 On systems that do not implement "mlockall", this function returns
1025 -1 and sets errno to "ENOSYS". Similarly, flag combinations not
1026 supported by the system result in a return value of -1 with errno
1027 being set to "EINVAL".
1028
1029 Note that the corresponding "munlockall" is synchronous and is
1030 documented under "MISCELLANEOUS FUNCTIONS".
1031
1032 Example: asynchronously lock all current and future pages into
1033 memory.
1034
1035 aio_mlockall IO::AIO::MCL_FUTURE;
1036
1037 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
1038 Queries the extents of the given file (by calling the Linux "FIEMAP"
1039 ioctl, see <http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for
1040 details). If the ioctl is not available on your OS, then this
1041 request will fail with "ENOSYS".
1042
1043 $start is the starting offset to query extents for, $length is the
1044 size of the range to query - if it is "undef", then the whole file
1045 will be queried.
1046
1047 $flags is a combination of flags ("IO::AIO::FIEMAP_FLAG_SYNC" or
1048 "IO::AIO::FIEMAP_FLAG_XATTR" - "IO::AIO::FIEMAP_FLAGS_COMPAT" is
1049 also exported), and is normally 0 or "IO::AIO::FIEMAP_FLAG_SYNC" to
1050 query the data portion.
1051
1052 $count is the maximum number of extent records to return. If it is
1053 "undef", then IO::AIO queries all extents of the range. As a very
1054 special case, if it is 0, then the callback receives the number of
1055 extents instead of the extents themselves (which is unreliable, see
1056 below).
1057
1058 If an error occurs, the callback receives no arguments. The special
1059 "errno" value "IO::AIO::EBADR" is available to test for flag errors.
1060
1061 Otherwise, the callback receives an array reference with extent
1062 structures. Each extent structure is an array reference itself, with
1063 the following members:
1064
1065 [$logical, $physical, $length, $flags]
1066
1067 Flags is any combination of the following flag values (typically
1068 either 0 or "IO::AIO::FIEMAP_EXTENT_LAST" (1)):
1069
1070 "IO::AIO::FIEMAP_EXTENT_LAST", "IO::AIO::FIEMAP_EXTENT_UNKNOWN",
1071 "IO::AIO::FIEMAP_EXTENT_DELALLOC", "IO::AIO::FIEMAP_EXTENT_ENCODED",
1072 "IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED",
1073 "IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED",
1074 "IO::AIO::FIEMAP_EXTENT_DATA_INLINE",
1075 "IO::AIO::FIEMAP_EXTENT_DATA_TAIL",
1076 "IO::AIO::FIEMAP_EXTENT_UNWRITTEN", "IO::AIO::FIEMAP_EXTENT_MERGED"
1077 or "IO::AIO::FIEMAP_EXTENT_SHARED".
1078
1079 At the time of this writing (Linux 3.2), this request is unreliable
1080 unless $count is "undef", as the kernel has all sorts of bugs
1081 preventing it to return all extents of a range for files with a
1082 large number of extents. The code (only) works around all these
1083 issues if $count is "undef".
1084
1085 aio_group $callback->(...)
1086 This is a very special aio request: Instead of doing something, it
1087 is a container for other aio requests, which is useful if you want
1088 to bundle many requests into a single, composite, request with a
1089 definite callback and the ability to cancel the whole request with
1090 its subrequests.
1091
1092 Returns an object of class IO::AIO::GRP. See its documentation below
1093 for more info.
1094
1095 Example:
1096
1097 my $grp = aio_group sub {
1098 print "all stats done\n";
1099 };
1100
1101 add $grp
1102 (aio_stat ...),
1103 (aio_stat ...),
1104 ...;
1105
1106 aio_nop $callback->()
1107 This is a special request - it does nothing in itself and is only
1108 used for side effects, such as when you want to add a dummy request
1109 to a group so that finishing the requests in the group depends on
1110 executing the given code.
1111
1112 While this request does nothing, it still goes through the execution
1113 phase and still requires a worker thread. Thus, the callback will
1114 not be executed immediately but only after other requests in the
1115 queue have entered their execution phase. This can be used to
1116 measure request latency.
1117
1118 IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
1119 Mainly used for debugging and benchmarking, this aio request puts
1120 one of the request workers to sleep for the given time.
1121
1122 While it is theoretically handy to have simple I/O scheduling
1123 requests like sleep and file handle readable/writable, the overhead
1124 this creates is immense (it blocks a thread for a long time) so do
1125 not use this function except to put your application under
1126 artificial I/O pressure.
1127
1128 IO::AIO::WD - multiple working directories
1129 Your process only has one current working directory, which is used by
1130 all threads. This makes it hard to use relative paths (some other
1131 component could call "chdir" at any time, and it is hard to control when
1132 the path will be used by IO::AIO).
1133
1134 One solution for this is to always use absolute paths. This usually
1135 works, but can be quite slow (the kernel has to walk the whole path on
1136 every access), and can also be a hassle to implement.
1137
1138 Newer POSIX systems have a number of functions (openat, fdopendir,
1139 futimensat and so on) that make it possible to specify working
1140 directories per operation.
1141
1142 For portability, and because the clowns who "designed", or shall I
1143 write, perpetrated this new interface were obviously half-drunk, this
1144 abstraction cannot be perfect, though.
1145
1146 IO::AIO allows you to convert directory paths into a so-called
1147 IO::AIO::WD object. This object stores the canonicalised, absolute
1148 version of the path, and on systems that allow it, also a directory file
1149 descriptor.
1150
1151 Everywhere where a pathname is accepted by IO::AIO (e.g. in "aio_stat"
1152 or "aio_unlink"), one can specify an array reference with an IO::AIO::WD
1153 object and a pathname instead (or the IO::AIO::WD object alone, which
1154 gets interpreted as "[$wd, "."]"). If the pathname is absolute, the
1155 IO::AIO::WD object is ignored, otherwise the pathname is resolved
1156 relative to that IO::AIO::WD object.
1157
1158 For example, to get a wd object for /etc and then stat passwd inside,
1159 you would write:
1160
1161 aio_wd "/etc", sub {
1162 my $etcdir = shift;
1163
1164 # although $etcdir can be undef on error, there is generally no reason
1165 # to check for errors here, as aio_stat will fail with ENOENT
1166 # when $etcdir is undef.
1167
1168 aio_stat [$etcdir, "passwd"], sub {
1169 # yay
1170 };
1171 };
1172
1173 The fact that "aio_wd" is a request and not a normal function shows that
1174 creating an IO::AIO::WD object is itself a potentially blocking
1175 operation, which is why it is done asynchronously.
1176
1177 To stat the directory obtained with "aio_wd" above, one could write
1178 either of the following three request calls:
1179
1180 aio_lstat "/etc" , sub { ... # pathname as normal string
1181 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1182 aio_lstat $wd , sub { ... # shorthand for the previous
1183
1184 As with normal pathnames, IO::AIO keeps a copy of the working directory
1185 object and the pathname string, so you could write the following without
1186 causing any issues due to $path getting reused:
1187
1188 my $path = [$wd, undef];
1189
1190 for my $name (qw(abc def ghi)) {
1191 $path->[1] = $name;
1192 aio_stat $path, sub {
1193 # ...
1194 };
1195 }
1196
1197 There are some caveats: when directories get renamed (or deleted), the
1198 pathname string doesn't change, so will point to the new directory (or
1199 nowhere at all), while the directory fd, if available on the system,
1200 will still point to the original directory. Most functions accepting a
1201 pathname will use the directory fd on newer systems, and the string on
1202 older systems. Some functions (such as "aio_realpath") will always rely
1203 on the string form of the pathname.
1204
1205 So this functionality is mainly useful to get some protection against
1206 "chdir", to easily get an absolute path out of a relative path for
1207 future reference, and to speed up doing many operations in the same
1208 directory (e.g. when stat'ing all files in a directory).
1209
1210 The following functions implement this working directory abstraction:
1211
1212 aio_wd $pathname, $callback->($wd)
1213 Asynchonously canonicalise the given pathname and convert it to an
1214 IO::AIO::WD object representing it. If possible and supported on the
1215 system, also open a directory fd to speed up pathname resolution
1216 relative to this working directory.
1217
1218 If something goes wrong, then "undef" is passwd to the callback
1219 instead of a working directory object and $! is set appropriately.
1220 Since passing "undef" as working directory component of a pathname
1221 fails the request with "ENOENT", there is often no need for error
1222 checking in the "aio_wd" callback, as future requests using the
1223 value will fail in the expected way.
1224
1225 IO::AIO::CWD
1226 This is a compiletime constant (object) that represents the process
1227 current working directory.
1228
1229 Specifying this object as working directory object for a pathname is
1230 as if the pathname would be specified directly, without a directory
1231 object. For example, these calls are functionally identical:
1232
1233 aio_stat "somefile", sub { ... };
1234 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1235
1236 To recover the path associated with an IO::AIO::WD object, you can use
1237 "aio_realpath":
1238
1239 aio_realpath $wd, sub {
1240 warn "path is $_[0]\n";
1241 };
1242
1243 Currently, "aio_statvfs" always, and "aio_rename" and "aio_rmdir"
1244 sometimes, fall back to using an absolue path.
1245
1246 IO::AIO::REQ CLASS
1247 All non-aggregate "aio_*" functions return an object of this class when
1248 called in non-void context.
1249
1250 cancel $req
1251 Cancels the request, if possible. Has the effect of skipping
1252 execution when entering the execute state and skipping calling the
1253 callback when entering the the result state, but will leave the
1254 request otherwise untouched (with the exception of readdir). That
1255 means that requests that currently execute will not be stopped and
1256 resources held by the request will not be freed prematurely.
1257
1258 cb $req $callback->(...)
1259 Replace (or simply set) the callback registered to the request.
1260
1261 IO::AIO::GRP CLASS
1262 This class is a subclass of IO::AIO::REQ, so all its methods apply to
1263 objects of this class, too.
1264
1265 A IO::AIO::GRP object is a special request that can contain multiple
1266 other aio requests.
1267
1268 You create one by calling the "aio_group" constructing function with a
1269 callback that will be called when all contained requests have entered
1270 the "done" state:
1271
1272 my $grp = aio_group sub {
1273 print "all requests are done\n";
1274 };
1275
1276 You add requests by calling the "add" method with one or more
1277 "IO::AIO::REQ" objects:
1278
1279 $grp->add (aio_unlink "...");
1280
1281 add $grp aio_stat "...", sub {
1282 $_[0] or return $grp->result ("error");
1283
1284 # add another request dynamically, if first succeeded
1285 add $grp aio_open "...", sub {
1286 $grp->result ("ok");
1287 };
1288 };
1289
1290 This makes it very easy to create composite requests (see the source of
1291 "aio_move" for an application) that work and feel like simple requests.
1292
1293 * The IO::AIO::GRP objects will be cleaned up during calls to
1294 "IO::AIO::poll_cb", just like any other request.
1295
1296 * They can be canceled like any other request. Canceling will cancel
1297 not only the request itself, but also all requests it contains.
1298
1299 * They can also can also be added to other IO::AIO::GRP objects.
1300
1301 * You must not add requests to a group from within the group callback
1302 (or any later time).
1303
1304 Their lifetime, simplified, looks like this: when they are empty, they
1305 will finish very quickly. If they contain only requests that are in the
1306 "done" state, they will also finish. Otherwise they will continue to
1307 exist.
1308
1309 That means after creating a group you have some time to add requests
1310 (precisely before the callback has been invoked, which is only done
1311 within the "poll_cb"). And in the callbacks of those requests, you can
1312 add further requests to the group. And only when all those requests have
1313 finished will the the group itself finish.
1314
1315 add $grp ...
1316 $grp->add (...)
1317 Add one or more requests to the group. Any type of IO::AIO::REQ can
1318 be added, including other groups, as long as you do not create
1319 circular dependencies.
1320
1321 Returns all its arguments.
1322
1323 $grp->cancel_subs
1324 Cancel all subrequests and clears any feeder, but not the group
1325 request itself. Useful when you queued a lot of events but got a
1326 result early.
1327
1328 The group request will finish normally (you cannot add requests to
1329 the group).
1330
1331 $grp->result (...)
1332 Set the result value(s) that will be passed to the group callback
1333 when all subrequests have finished and set the groups errno to the
1334 current value of errno (just like calling "errno" without an error
1335 number). By default, no argument will be passed and errno is zero.
1336
1337 $grp->errno ([$errno])
1338 Sets the group errno value to $errno, or the current value of errno
1339 when the argument is missing.
1340
1341 Every aio request has an associated errno value that is restored
1342 when the callback is invoked. This method lets you change this value
1343 from its default (0).
1344
1345 Calling "result" will also set errno, so make sure you either set $!
1346 before the call to "result", or call c<errno> after it.
1347
1348 feed $grp $callback->($grp)
1349 Sets a feeder/generator on this group: every group can have an
1350 attached generator that generates requests if idle. The idea behind
1351 this is that, although you could just queue as many requests as you
1352 want in a group, this might starve other requests for a potentially
1353 long time. For example, "aio_scandir" might generate hundreds of
1354 thousands of "aio_stat" requests, delaying any later requests for a
1355 long time.
1356
1357 To avoid this, and allow incremental generation of requests, you can
1358 instead a group and set a feeder on it that generates those
1359 requests. The feed callback will be called whenever there are few
1360 enough (see "limit", below) requests active in the group itself and
1361 is expected to queue more requests.
1362
1363 The feed callback can queue as many requests as it likes (i.e. "add"
1364 does not impose any limits).
1365
1366 If the feed does not queue more requests when called, it will be
1367 automatically removed from the group.
1368
1369 If the feed limit is 0 when this method is called, it will be set to
1370 2 automatically.
1371
1372 Example:
1373
1374 # stat all files in @files, but only ever use four aio requests concurrently:
1375
1376 my $grp = aio_group sub { print "finished\n" };
1377 limit $grp 4;
1378 feed $grp sub {
1379 my $file = pop @files
1380 or return;
1381
1382 add $grp aio_stat $file, sub { ... };
1383 };
1384
1385 limit $grp $num
1386 Sets the feeder limit for the group: The feeder will be called
1387 whenever the group contains less than this many requests.
1388
1389 Setting the limit to 0 will pause the feeding process.
1390
1391 The default value for the limit is 0, but note that setting a feeder
1392 automatically bumps it up to 2.
1393
265 SUPPORT FUNCTIONS 1394 SUPPORT FUNCTIONS
1395 EVENT PROCESSING AND EVENT LOOP INTEGRATION
266 $fileno = IO::AIO::poll_fileno 1396 $fileno = IO::AIO::poll_fileno
267 Return the *request result pipe file descriptor*. This filehandle 1397 Return the *request result pipe file descriptor*. This filehandle
268 must be polled for reading by some mechanism outside this module 1398 must be polled for reading by some mechanism outside this module
269 (e.g. Event or select, see below or the SYNOPSIS). If the pipe 1399 (e.g. EV, Glib, select and so on, see below or the SYNOPSIS). If the
270 becomes readable you have to call "poll_cb" to check the results. 1400 pipe becomes readable you have to call "poll_cb" to check the
1401 results.
271 1402
272 See "poll_cb" for an example. 1403 See "poll_cb" for an example.
273 1404
274 IO::AIO::poll_cb 1405 IO::AIO::poll_cb
275 Process all outstanding events on the result pipe. You have to call 1406 Process some requests that have reached the result phase (i.e. they
276 this regularly. Returns the number of events processed. Returns 1407 have been executed but the results are not yet reported). You have
1408 to call this "regularly" to finish outstanding requests.
1409
1410 Returns 0 if all events could be processed (or there were no events
1411 to process), or -1 if it returned earlier for whatever reason.
277 immediately when no events are outstanding. 1412 Returns immediately when no events are outstanding. The amount of
1413 events processed depends on the settings of "IO::AIO::max_poll_req",
1414 "IO::AIO::max_poll_time" and "IO::AIO::max_outstanding".
1415
1416 If not all requests were processed for whatever reason, the poll
1417 file descriptor will still be ready when "poll_cb" returns, so
1418 normally you don't have to do anything special to have it called
1419 later.
1420
1421 Apart from calling "IO::AIO::poll_cb" when the event filehandle
1422 becomes ready, it can be beneficial to call this function from loops
1423 which submit a lot of requests, to make sure the results get
1424 processed when they become available and not just when the loop is
1425 finished and the event loop takes over again. This function returns
1426 very fast when there are no outstanding requests.
278 1427
279 Example: Install an Event watcher that automatically calls 1428 Example: Install an Event watcher that automatically calls
280 IO::AIO::poll_cb with high priority: 1429 IO::AIO::poll_cb with high priority (more examples can be found in
1430 the SYNOPSIS section, at the top of this document):
281 1431
282 Event->io (fd => IO::AIO::poll_fileno, 1432 Event->io (fd => IO::AIO::poll_fileno,
283 poll => 'r', async => 1, 1433 poll => 'r', async => 1,
284 cb => \&IO::AIO::poll_cb); 1434 cb => \&IO::AIO::poll_cb);
285 1435
286 IO::AIO::poll_wait 1436 IO::AIO::poll_wait
287 Wait till the result filehandle becomes ready for reading (simply 1437 Wait until either at least one request is in the result phase or no
288 does a "select" on the filehandle. This is useful if you want to 1438 requests are outstanding anymore.
289 synchronously wait for some requests to finish). 1439
1440 This is useful if you want to synchronously wait for some requests
1441 to become ready, without actually handling them.
290 1442
291 See "nreqs" for an example. 1443 See "nreqs" for an example.
292 1444
1445 IO::AIO::poll
1446 Waits until some requests have been handled.
1447
1448 Returns the number of requests processed, but is otherwise strictly
1449 equivalent to:
1450
1451 IO::AIO::poll_wait, IO::AIO::poll_cb
1452
293 IO::AIO::nreqs 1453 IO::AIO::flush
294 Returns the number of requests currently outstanding (i.e. for which 1454 Wait till all outstanding AIO requests have been handled.
295 their callback has not been invoked yet).
296 1455
297 Example: wait till there are no outstanding requests anymore: 1456 Strictly equivalent to:
298 1457
299 IO::AIO::poll_wait, IO::AIO::poll_cb 1458 IO::AIO::poll_wait, IO::AIO::poll_cb
300 while IO::AIO::nreqs; 1459 while IO::AIO::nreqs;
301 1460
1461 This function can be useful at program aborts, to make sure
1462 outstanding I/O has been done ("IO::AIO" uses an "END" block which
1463 already calls this function on normal exits), or when you are merely
1464 using "IO::AIO" for its more advanced functions, rather than for
1465 async I/O, e.g.:
1466
1467 my ($dirs, $nondirs);
1468 IO::AIO::aio_scandir "/tmp", 0, sub { ($dirs, $nondirs) = @_ };
302 IO::AIO::flush 1469 IO::AIO::flush;
303 Wait till all outstanding AIO requests have been handled. 1470 # $dirs, $nondirs are now set
304 1471
305 Strictly equivalent to: 1472 IO::AIO::max_poll_reqs $nreqs
1473 IO::AIO::max_poll_time $seconds
1474 These set the maximum number of requests (default 0, meaning
1475 infinity) that are being processed by "IO::AIO::poll_cb" in one
1476 call, respectively the maximum amount of time (default 0, meaning
1477 infinity) spent in "IO::AIO::poll_cb" to process requests (more
1478 correctly the mininum amount of time "poll_cb" is allowed to use).
306 1479
307 IO::AIO::poll_wait, IO::AIO::poll_cb 1480 Setting "max_poll_time" to a non-zero value creates an overhead of
308 while IO::AIO::nreqs; 1481 one syscall per request processed, which is not normally a problem
1482 unless your callbacks are really really fast or your OS is really
1483 really slow (I am not mentioning Solaris here). Using
1484 "max_poll_reqs" incurs no overhead.
309 1485
310 IO::AIO::poll 1486 Setting these is useful if you want to ensure some level of
311 Waits until some requests have been handled. 1487 interactiveness when perl is not fast enough to process all requests
1488 in time.
312 1489
313 Strictly equivalent to: 1490 For interactive programs, values such as 0.01 to 0.1 should be fine.
314 1491
315 IO::AIO::poll_wait, IO::AIO::poll_cb 1492 Example: Install an Event watcher that automatically calls
316 if IO::AIO::nreqs; 1493 IO::AIO::poll_cb with low priority, to ensure that other parts of
1494 the program get the CPU sometimes even under high AIO load.
317 1495
1496 # try not to spend much more than 0.1s in poll_cb
1497 IO::AIO::max_poll_time 0.1;
1498
1499 # use a low priority so other tasks have priority
1500 Event->io (fd => IO::AIO::poll_fileno,
1501 poll => 'r', nice => 1,
1502 cb => &IO::AIO::poll_cb);
1503
1504 CONTROLLING THE NUMBER OF THREADS
318 IO::AIO::min_parallel $nthreads 1505 IO::AIO::min_parallel $nthreads
319 Set the minimum number of AIO threads to $nthreads. The current 1506 Set the minimum number of AIO threads to $nthreads. The current
320 default is 4, which means four asynchronous operations can be done 1507 default is 8, which means eight asynchronous operations can execute
321 at one time (the number of outstanding operations, however, is 1508 concurrently at any one time (the number of outstanding requests,
322 unlimited). 1509 however, is unlimited).
323 1510
324 IO::AIO starts threads only on demand, when an AIO request is queued 1511 IO::AIO starts threads only on demand, when an AIO request is queued
325 and no free thread exists. 1512 and no free thread exists. Please note that queueing up a hundred
1513 requests can create demand for a hundred threads, even if it turns
1514 out that everything is in the cache and could have been processed
1515 faster by a single thread.
326 1516
327 It is recommended to keep the number of threads low, as some Linux 1517 It is recommended to keep the number of threads relatively low, as
328 kernel versions will scale negatively with the number of threads 1518 some Linux kernel versions will scale negatively with the number of
329 (higher parallelity => MUCH higher latency). With current Linux 2.6 1519 threads (higher parallelity => MUCH higher latency). With current
330 versions, 4-32 threads should be fine. 1520 Linux 2.6 versions, 4-32 threads should be fine.
331 1521
332 Under most circumstances you don't need to call this function, as 1522 Under most circumstances you don't need to call this function, as
333 the module selects a default that is suitable for low to moderate 1523 the module selects a default that is suitable for low to moderate
334 load. 1524 load.
335 1525
… …
345 ensure that all threads are killed and that there are no outstanding 1535 ensure that all threads are killed and that there are no outstanding
346 requests. 1536 requests.
347 1537
348 Under normal circumstances you don't need to call this function. 1538 Under normal circumstances you don't need to call this function.
349 1539
1540 IO::AIO::max_idle $nthreads
1541 Limit the number of threads (default: 4) that are allowed to idle
1542 (i.e., threads that did not get a request to process within the idle
1543 timeout (default: 10 seconds). That means if a thread becomes idle
1544 while $nthreads other threads are also idle, it will free its
1545 resources and exit.
1546
1547 This is useful when you allow a large number of threads (e.g. 100 or
1548 1000) to allow for extremely high load situations, but want to free
1549 resources under normal circumstances (1000 threads can easily
1550 consume 30MB of RAM).
1551
1552 The default is probably ok in most situations, especially if thread
1553 creation is fast. If thread creation is very slow on your system you
1554 might want to use larger values.
1555
1556 IO::AIO::idle_timeout $seconds
1557 Sets the minimum idle timeout (default 10) after which worker
1558 threads are allowed to exit. SEe "IO::AIO::max_idle".
1559
350 $oldnreqs = IO::AIO::max_outstanding $nreqs 1560 IO::AIO::max_outstanding $maxreqs
351 Sets the maximum number of outstanding requests to $nreqs. If you 1561 Sets the maximum number of outstanding requests to $nreqs. If you do
352 try to queue up more than this number of requests, the caller will 1562 queue up more than this number of requests, the next call to
1563 "IO::AIO::poll_cb" (and other functions calling "poll_cb", such as
1564 "IO::AIO::flush" or "IO::AIO::poll") will block until the limit is
1565 no longer exceeded.
1566
1567 In other words, this setting does not enforce a queue limit, but can
1568 be used to make poll functions block if the limit is exceeded.
1569
1570 This is a very bad function to use in interactive programs because
1571 it blocks, and a bad way to reduce concurrency because it is
1572 inexact: Better use an "aio_group" together with a feed callback.
1573
1574 Its main use is in scripts without an event loop - when you want to
1575 stat a lot of files, you can write something like this:
1576
1577 IO::AIO::max_outstanding 32;
1578
1579 for my $path (...) {
1580 aio_stat $path , ...;
1581 IO::AIO::poll_cb;
1582 }
1583
1584 IO::AIO::flush;
1585
1586 The call to "poll_cb" inside the loop will normally return
1587 instantly, but as soon as more thna 32 reqeusts are in-flight, it
353 block until some requests have been handled. 1588 will block until some requests have been handled. This keeps the
1589 loop from pushing a large number of "aio_stat" requests onto the
1590 queue.
354 1591
355 The default is very large, so normally there is no practical limit. 1592 The default value for "max_outstanding" is very large, so there is
356 If you queue up many requests in a loop it often improves speed if 1593 no practical limit on the number of outstanding requests.
357 you set this to a relatively low number, such as 100.
358 1594
359 Under normal circumstances you don't need to call this function. 1595 STATISTICAL INFORMATION
1596 IO::AIO::nreqs
1597 Returns the number of requests currently in the ready, execute or
1598 pending states (i.e. for which their callback has not been invoked
1599 yet).
1600
1601 Example: wait till there are no outstanding requests anymore:
1602
1603 IO::AIO::poll_wait, IO::AIO::poll_cb
1604 while IO::AIO::nreqs;
1605
1606 IO::AIO::nready
1607 Returns the number of requests currently in the ready state (not yet
1608 executed).
1609
1610 IO::AIO::npending
1611 Returns the number of requests currently in the pending state
1612 (executed, but not yet processed by poll_cb).
1613
1614 SUBSECOND STAT TIME ACCESS
1615 Both "aio_stat"/"aio_lstat" and perl's "stat"/"lstat" functions can
1616 generally find access/modification and change times with subsecond time
1617 accuracy of the system supports it, but perl's built-in functions only
1618 return the integer part.
1619
1620 The following functions return the timestamps of the most recent stat
1621 with subsecond precision on most systems and work both after
1622 "aio_stat"/"aio_lstat" and perl's "stat"/"lstat" calls. Their return
1623 value is only meaningful after a successful "stat"/"lstat" call, or
1624 during/after a successful "aio_stat"/"aio_lstat" callback.
1625
1626 This is similar to the Time::HiRes "stat" functions, but can return full
1627 resolution without rounding and work with standard perl "stat",
1628 alleviating the need to call the special "Time::HiRes" functions, which
1629 do not act like their perl counterparts.
1630
1631 On operating systems or file systems where subsecond time resolution is
1632 not supported or could not be detected, a fractional part of 0 is
1633 returned, so it is always safe to call these functions.
1634
1635 $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime,
1636 IO::AIO::st_btime
1637 Return the access, modication, change or birth time, respectively,
1638 including fractional part. Due to the limited precision of floating
1639 point, the accuracy on most platforms is only a bit better than
1640 milliseconds for times around now - see the *nsec* function family,
1641 below, for full accuracy.
1642
1643 File birth time is only available when the OS and perl support it
1644 (on FreeBSD and NetBSD at the time of this writing, although support
1645 is adaptive, so if your OS/perl gains support, IO::AIO can take
1646 avdantage of it). On systems where it isn't available, 0 is
1647 currently returned, but this might change to "undef" in a future
1648 version.
1649
1650 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
1651 Returns access, modification, change and birth time all in one go,
1652 and maybe more times in the future version.
1653
1654 $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec,
1655 IO::AIO::st_ctimensec, IO::AIO::st_btimensec
1656 Return the fractional access, modifcation, change or birth time, in
1657 nanoseconds, as an integer in the range 0 to 999999999.
1658
1659 Note that no accessors are provided for access, modification and
1660 change times - you need to get those from "stat _" if required ("int
1661 IO::AIO::st_atime" and so on will *not* generally give you the
1662 correct value).
1663
1664 $seconds = IO::AIO::st_btimesec
1665 The (integral) seconds part of the file birth time, if available.
1666
1667 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
1668 Like the functions above, but returns all four times in one go (and
1669 maybe more in future versions).
1670
1671 $counter = IO::AIO::st_gen
1672 Returns the generation counter (in practice this is just a random
1673 number) of the file. This is only available on platforms which have
1674 this member in their "struct stat" (most BSDs at the time of this
1675 writing) and generally only to the root usert. If unsupported, 0 is
1676 returned, but this might change to "undef" in a future version.
1677
1678 Example: print the high resolution modification time of /etc, using
1679 "stat", and "IO::AIO::aio_stat".
1680
1681 if (stat "/etc") {
1682 printf "stat(/etc) mtime: %f\n", IO::AIO::st_mtime;
1683 }
1684
1685 IO::AIO::aio_stat "/etc", sub {
1686 $_[0]
1687 and return;
1688
1689 printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::st_mtimensec;
1690 };
1691
1692 IO::AIO::flush;
1693
1694 Output of the awbove on my system, showing reduced and full accuracy:
1695
1696 stat(/etc) mtime: 1534043702.020808
1697 aio_stat(/etc) mtime: 1534043702.020807792
1698
1699 MISCELLANEOUS FUNCTIONS
1700 IO::AIO implements some functions that are useful when you want to use
1701 some "Advanced I/O" function not available to in Perl, without going the
1702 "Asynchronous I/O" route. Many of these have an asynchronous "aio_*"
1703 counterpart.
1704
1705 $numfd = IO::AIO::get_fdlimit
1706 This function is *EXPERIMENTAL* and subject to change.
1707
1708 Tries to find the current file descriptor limit and returns it, or
1709 "undef" and sets $! in case of an error. The limit is one larger
1710 than the highest valid file descriptor number.
1711
1712 IO::AIO::min_fdlimit [$numfd]
1713 This function is *EXPERIMENTAL* and subject to change.
1714
1715 Try to increase the current file descriptor limit(s) to at least
1716 $numfd by changing the soft or hard file descriptor resource limit.
1717 If $numfd is missing, it will try to set a very high limit, although
1718 this is not recommended when you know the actual minimum that you
1719 require.
1720
1721 If the limit cannot be raised enough, the function makes a
1722 best-effort attempt to increase the limit as much as possible, using
1723 various tricks, while still failing. You can query the resulting
1724 limit using "IO::AIO::get_fdlimit".
1725
1726 If an error occurs, returns "undef" and sets $!, otherwise returns
1727 true.
1728
1729 IO::AIO::sendfile $ofh, $ifh, $offset, $count
1730 Calls the "eio_sendfile_sync" function, which is like
1731 "aio_sendfile", but is blocking (this makes most sense if you know
1732 the input data is likely cached already and the output filehandle is
1733 set to non-blocking operations).
1734
1735 Returns the number of bytes copied, or -1 on error.
1736
1737 IO::AIO::fadvise $fh, $offset, $len, $advice
1738 Simply calls the "posix_fadvise" function (see its manpage for
1739 details). The following advice constants are available:
1740 "IO::AIO::FADV_NORMAL", "IO::AIO::FADV_SEQUENTIAL",
1741 "IO::AIO::FADV_RANDOM", "IO::AIO::FADV_NOREUSE",
1742 "IO::AIO::FADV_WILLNEED", "IO::AIO::FADV_DONTNEED".
1743
1744 On systems that do not implement "posix_fadvise", this function
1745 returns ENOSYS, otherwise the return value of "posix_fadvise".
1746
1747 IO::AIO::madvise $scalar, $offset, $len, $advice
1748 Simply calls the "posix_madvise" function (see its manpage for
1749 details). The following advice constants are available:
1750 "IO::AIO::MADV_NORMAL", "IO::AIO::MADV_SEQUENTIAL",
1751 "IO::AIO::MADV_RANDOM", "IO::AIO::MADV_WILLNEED",
1752 "IO::AIO::MADV_DONTNEED".
1753
1754 If $offset is negative, counts from the end. If $length is negative,
1755 the remaining length of the $scalar is used. If possible, $length
1756 will be reduced to fit into the $scalar.
1757
1758 On systems that do not implement "posix_madvise", this function
1759 returns ENOSYS, otherwise the return value of "posix_madvise".
1760
1761 IO::AIO::mprotect $scalar, $offset, $len, $protect
1762 Simply calls the "mprotect" function on the preferably AIO::mmap'ed
1763 $scalar (see its manpage for details). The following protect
1764 constants are available: "IO::AIO::PROT_NONE", "IO::AIO::PROT_READ",
1765 "IO::AIO::PROT_WRITE", "IO::AIO::PROT_EXEC".
1766
1767 If $offset is negative, counts from the end. If $length is negative,
1768 the remaining length of the $scalar is used. If possible, $length
1769 will be reduced to fit into the $scalar.
1770
1771 On systems that do not implement "mprotect", this function returns
1772 ENOSYS, otherwise the return value of "mprotect".
1773
1774 IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1775 Memory-maps a file (or anonymous memory range) and attaches it to
1776 the given $scalar, which will act like a string scalar. Returns true
1777 on success, and false otherwise.
1778
1779 The scalar must exist, but its contents do not matter - this means
1780 you cannot use a nonexistant array or hash element. When in doubt,
1781 "undef" the scalar first.
1782
1783 The only operations allowed on the mmapped scalar are
1784 "substr"/"vec", which don't change the string length, and most
1785 read-only operations such as copying it or searching it with regexes
1786 and so on.
1787
1788 Anything else is unsafe and will, at best, result in memory leaks.
1789
1790 The memory map associated with the $scalar is automatically removed
1791 when the $scalar is undef'd or destroyed, or when the
1792 "IO::AIO::mmap" or "IO::AIO::munmap" functions are called on it.
1793
1794 This calls the "mmap"(2) function internally. See your system's
1795 manual page for details on the $length, $prot and $flags parameters.
1796
1797 The $length must be larger than zero and smaller than the actual
1798 filesize.
1799
1800 $prot is a combination of "IO::AIO::PROT_NONE",
1801 "IO::AIO::PROT_EXEC", "IO::AIO::PROT_READ" and/or
1802 "IO::AIO::PROT_WRITE",
1803
1804 $flags can be a combination of "IO::AIO::MAP_SHARED" or
1805 "IO::AIO::MAP_PRIVATE", or a number of system-specific flags (when
1806 not available, the are 0): "IO::AIO::MAP_ANONYMOUS" (which is set to
1807 "MAP_ANON" if your system only provides this constant),
1808 "IO::AIO::MAP_LOCKED", "IO::AIO::MAP_NORESERVE",
1809 "IO::AIO::MAP_POPULATE", "IO::AIO::MAP_NONBLOCK",
1810 "IO::AIO::MAP_FIXED", "IO::AIO::MAP_GROWSDOWN",
1811 "IO::AIO::MAP_32BIT", "IO::AIO::MAP_HUGETLB" or
1812 "IO::AIO::MAP_STACK".
1813
1814 If $fh is "undef", then a file descriptor of -1 is passed.
1815
1816 $offset is the offset from the start of the file - it generally must
1817 be a multiple of "IO::AIO::PAGESIZE" and defaults to 0.
1818
1819 Example:
1820
1821 use Digest::MD5;
1822 use IO::AIO;
1823
1824 open my $fh, "<verybigfile"
1825 or die "$!";
1826
1827 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1828 or die "verybigfile: $!";
1829
1830 my $fast_md5 = md5 $data;
1831
1832 IO::AIO::munmap $scalar
1833 Removes a previous mmap and undefines the $scalar.
1834
1835 IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[,
1836 $new_address = 0]
1837 Calls the Linux-specific mremap(2) system call. The $scalar must
1838 have been mapped by "IO::AIO::mmap", and $flags must currently
1839 either be 0 or "IO::AIO::MREMAP_MAYMOVE".
1840
1841 Returns true if successful, and false otherwise. If the underlying
1842 mmapped region has changed address, then the true value has the
1843 numerical value 1, otherwise it has the numerical value 0:
1844
1845 my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE
1846 or die "mremap: $!";
1847
1848 if ($success*1) {
1849 warn "scalar has chanegd address in memory\n";
1850 }
1851
1852 "IO::AIO::MREMAP_FIXED" and the $new_address argument are currently
1853 implemented, but not supported and might go away in a future
1854 version.
1855
1856 On systems where this call is not supported or is not emulated, this
1857 call returns falls and sets $! to "ENOSYS".
1858
1859 IO::AIO::mlockall $flags
1860 Calls the "eio_mlockall_sync" function, which is like
1861 "aio_mlockall", but is blocking.
1862
1863 IO::AIO::munlock $scalar, $offset = 0, $length = undef
1864 Calls the "munlock" function, undoing the effects of a previous
1865 "aio_mlock" call (see its description for details).
1866
1867 IO::AIO::munlockall
1868 Calls the "munlockall" function.
1869
1870 On systems that do not implement "munlockall", this function returns
1871 ENOSYS, otherwise the return value of "munlockall".
1872
1873 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
1874 Calls the GNU/Linux splice(2) syscall, if available. If $r_off or
1875 $w_off are "undef", then "NULL" is passed for these, otherwise they
1876 should be the file offset.
1877
1878 $r_fh and $w_fh should not refer to the same file, as splice might
1879 silently corrupt the data in this case.
1880
1881 The following symbol flag values are available:
1882 "IO::AIO::SPLICE_F_MOVE", "IO::AIO::SPLICE_F_NONBLOCK",
1883 "IO::AIO::SPLICE_F_MORE" and "IO::AIO::SPLICE_F_GIFT".
1884
1885 See the splice(2) manpage for details.
1886
1887 IO::AIO::tee $r_fh, $w_fh, $length, $flags
1888 Calls the GNU/Linux tee(2) syscall, see its manpage and the
1889 description for "IO::AIO::splice" above for details.
1890
1891 $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
1892 Attempts to query or change the pipe buffer size. Obviously works
1893 only on pipes, and currently works only on GNU/Linux systems, and
1894 fails with -1/"ENOSYS" everywhere else. If anybody knows how to
1895 influence pipe buffer size on other systems, drop me a note.
1896
1897 ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
1898 This is a direct interface to the Linux pipe2(2) system call. If
1899 $flags is missing or 0, then this should be the same as a call to
1900 perl's built-in "pipe" function and create a new pipe, and works on
1901 systems that lack the pipe2 syscall. On win32, this case invokes
1902 "_pipe (..., 4096, O_BINARY)".
1903
1904 If $flags is non-zero, it tries to invoke the pipe2 system call with
1905 the given flags (Linux 2.6.27, glibc 2.9).
1906
1907 On success, the read and write file handles are returned.
1908
1909 On error, nothing will be returned. If the pipe2 syscall is missing
1910 and $flags is non-zero, fails with "ENOSYS".
1911
1912 Please refer to pipe2(2) for more info on the $flags, but at the
1913 time of this writing, "IO::AIO::O_CLOEXEC", "IO::AIO::O_NONBLOCK"
1914 and "IO::AIO::O_DIRECT" (Linux 3.4, for packet-based pipes) were
1915 supported.
1916
1917 Example: create a pipe race-free w.r.t. threads and fork:
1918
1919 my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
1920 or die "pipe2: $!\n";
1921
1922 $fh = IO::AIO::eventfd [$initval, [$flags]]
1923 This is a direct interface to the Linux eventfd(2) system call. The
1924 (unhelpful) defaults for $initval and $flags are 0 for both.
1925
1926 On success, the new eventfd filehandle is returned, otherwise
1927 returns "undef". If the eventfd syscall is missing, fails with
1928 "ENOSYS".
1929
1930 Please refer to eventfd(2) for more info on this call.
1931
1932 The following symbol flag values are available:
1933 "IO::AIO::EFD_CLOEXEC", "IO::AIO::EFD_NONBLOCK" and
1934 "IO::AIO::EFD_SEMAPHORE" (Linux 2.6.30).
1935
1936 Example: create a new eventfd filehandle:
1937
1938 $fh = IO::AIO::eventfd 0, IO::AIO::O_CLOEXEC
1939 or die "eventfd: $!\n";
1940
1941 $fh = IO::AIO::timerfd_create $clockid[, $flags]
1942 This is a direct interface to the Linux timerfd_create(2) system
1943 call. The (unhelpful) default for $flags is 0.
1944
1945 On success, the new timerfd filehandle is returned, otherwise
1946 returns "undef". If the eventfd syscall is missing, fails with
1947 "ENOSYS".
1948
1949 Please refer to timerfd_create(2) for more info on this call.
1950
1951 The following $clockid values are available:
1952 "IO::AIO::CLOCK_REALTIME", "IO::AIO::CLOCK_MONOTONIC"
1953 "IO::AIO::CLOCK_CLOCK_BOOTTIME" (Linux 3.15)
1954 "IO::AIO::CLOCK_CLOCK_REALTIME_ALARM" (Linux 3.11) and
1955 "IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM" (Linux 3.11).
1956
1957 The following $flags values are available (Linux 2.6.27):
1958 "IO::AIO::TFD_NONBLOCK" and "IO::AIO::TFD_CLOEXEC".
1959
1960 Example: create a new timerfd and set it to one-second repeated
1961 alarms, then wait for two alarms:
1962
1963 my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC
1964 or die "timerfd_create: $!\n";
1965
1966 defined IO::AIO::timerfd_settime $fh, 0, 1, 1
1967 or die "timerfd_settime: $!\n";
1968
1969 for (1..2) {
1970 8 == sysread $fh, my $buf, 8
1971 or die "timerfd read failure\n";
1972
1973 printf "number of expirations (likely 1): %d\n",
1974 unpack "Q", $buf;
1975 }
1976
1977 ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags,
1978 $new_interval, $nbw_value
1979 This is a direct interface to the Linux timerfd_settime(2) system
1980 call. Please refer to its manpage for more info on this call.
1981
1982 The new itimerspec is specified using two (possibly fractional)
1983 second values, $new_interval and $new_value).
1984
1985 On success, the current interval and value are returned (as per
1986 "timerfd_gettime"). On failure, the empty list is returned.
1987
1988 The following $flags values are available:
1989 "IO::AIO::TFD_TIMER_ABSTIME" and "IO::AIO::TFD_TIMER_CANCEL_ON_SET".
1990
1991 See "IO::AIO::timerfd_create" for a full example.
1992
1993 ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
1994 This is a direct interface to the Linux timerfd_gettime(2) system
1995 call. Please refer to its manpage for more info on this call.
1996
1997 On success, returns the current values of interval and value for the
1998 given timerfd (as potentially fractional second values). On failure,
1999 the empty list is returned.
2000
2001EVENT LOOP INTEGRATION
2002 It is recommended to use AnyEvent::AIO to integrate IO::AIO
2003 automatically into many event loops:
2004
2005 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
2006 use AnyEvent::AIO;
2007
2008 You can also integrate IO::AIO manually into many event loops, here are
2009 some examples of how to do this:
2010
2011 # EV integration
2012 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
2013
2014 # Event integration
2015 Event->io (fd => IO::AIO::poll_fileno,
2016 poll => 'r',
2017 cb => \&IO::AIO::poll_cb);
2018
2019 # Glib/Gtk2 integration
2020 add_watch Glib::IO IO::AIO::poll_fileno,
2021 in => sub { IO::AIO::poll_cb; 1 };
2022
2023 # Tk integration
2024 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
2025 readable => \&IO::AIO::poll_cb);
2026
2027 # Danga::Socket integration
2028 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
2029 \&IO::AIO::poll_cb);
360 2030
361 FORK BEHAVIOUR 2031 FORK BEHAVIOUR
362 Before the fork, IO::AIO enters a quiescent state where no requests can 2032 Usage of pthreads in a program changes the semantics of fork
363 be added in other threads and no results will be processed. After the 2033 considerably. Specifically, only async-safe functions can be called
364 fork the parent simply leaves the quiescent state and continues 2034 after fork. Perl doesn't know about this, so in general, you cannot call
365 request/result processing, while the child clears the request/result 2035 fork with defined behaviour in perl if pthreads are involved. IO::AIO
366 queue (so the requests started before the fork will only be handled in 2036 uses pthreads, so this applies, but many other extensions and (for
367 the parent). Threats will be started on demand until the limit ste in 2037 inexplicable reasons) perl itself often is linked against pthreads, so
368 the parent process has been reached again. 2038 this limitation applies to quite a lot of perls.
2039
2040 This module no longer tries to fight your OS, or POSIX. That means
2041 IO::AIO only works in the process that loaded it. Forking is fully
2042 supported, but using IO::AIO in the child is not.
2043
2044 You might get around by not *using* IO::AIO before (or after) forking.
2045 You could also try to call the IO::AIO::reinit function in the child:
2046
2047 IO::AIO::reinit
2048 Abandons all current requests and I/O threads and simply
2049 reinitialises all data structures. This is not an operation
2050 supported by any standards, but happens to work on GNU/Linux and
2051 some newer BSD systems.
2052
2053 The only reasonable use for this function is to call it after
2054 forking, if "IO::AIO" was used in the parent. Calling it while
2055 IO::AIO is active in the process will result in undefined behaviour.
2056 Calling it at any time will also result in any undefined (by POSIX)
2057 behaviour.
2058
2059 LINUX-SPECIFIC CALLS
2060 When a call is documented as "linux-specific" then this means it
2061 originated on GNU/Linux. "IO::AIO" will usually try to autodetect the
2062 availability and compatibility of such calls regardless of the platform
2063 it is compiled on, so platforms such as FreeBSD which often implement
2064 these calls will work. When in doubt, call them and see if they fail wth
2065 "ENOSYS".
2066
2067 MEMORY USAGE
2068 Per-request usage:
2069
2070 Each aio request uses - depending on your architecture - around 100-200
2071 bytes of memory. In addition, stat requests need a stat buffer (possibly
2072 a few hundred bytes), readdir requires a result buffer and so on. Perl
2073 scalars and other data passed into aio requests will also be locked and
2074 will consume memory till the request has entered the done state.
2075
2076 This is not awfully much, so queuing lots of requests is not usually a
2077 problem.
2078
2079 Per-thread usage:
2080
2081 In the execution phase, some aio requests require more memory for
2082 temporary buffers, and each thread requires a stack and other data
2083 structures (usually around 16k-128k, depending on the OS).
2084
2085KNOWN BUGS
2086 Known bugs will be fixed in the next release :)
2087
2088KNOWN ISSUES
2089 Calls that try to "import" foreign memory areas (such as "IO::AIO::mmap"
2090 or "IO::AIO::aio_slurp") do not work with generic lvalues, such as
2091 non-created hash slots or other scalars I didn't think of. It's best to
2092 avoid such and either use scalar variables or making sure that the
2093 scalar exists (e.g. by storing "undef") and isn't "funny" (e.g. tied).
2094
2095 I am not sure anything can be done about this, so this is considered a
2096 known issue, rather than a bug.
369 2097
370SEE ALSO 2098SEE ALSO
371 Coro, Linux::AIO. 2099 AnyEvent::AIO for easy integration into event loops, Coro::AIO for a
2100 more natural syntax.
372 2101
373AUTHOR 2102AUTHOR
374 Marc Lehmann <schmorp@schmorp.de> 2103 Marc Lehmann <schmorp@schmorp.de>
375 http://home.schmorp.de/ 2104 http://home.schmorp.de/
376 2105

Diff Legend

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