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

Comparing IO-AIO/README (file contents):
Revision 1.18 by root, Thu Oct 26 16:28:33 2006 UTC vs.
Revision 1.66 by root, Tue Dec 29 15:20:12 2020 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
… …
23 $req->cancel; # cancel request if still in queue 24 $req->cancel; # cancel request if still in queue
24 25
25 my $grp = aio_group sub { print "all stats done\n" }; 26 my $grp = aio_group sub { print "all stats done\n" };
26 add $grp aio_stat "..." for ...; 27 add $grp aio_stat "..." for ...;
27 28
28 # AnyEvent integration
29 open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
30 my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
31
32 # Event integration
33 Event->io (fd => IO::AIO::poll_fileno,
34 poll => 'r',
35 cb => \&IO::AIO::poll_cb);
36
37 # Glib/Gtk2 integration
38 add_watch Glib::IO IO::AIO::poll_fileno,
39 in => sub { IO::AIO::poll_cb; 1 };
40
41 # Tk integration
42 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
43 readable => \&IO::AIO::poll_cb);
44
45 # Danga::Socket integration
46 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
47 \&IO::AIO::poll_cb);
48
49DESCRIPTION 29DESCRIPTION
50 This module implements asynchronous I/O using whatever means your 30 This module implements asynchronous I/O using whatever means your
51 operating system supports. 31 operating system supports. It is implemented as an interface to "libeio"
32 (<http://software.schmorp.de/pkg/libeio.html>).
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.
52 49
53 In this version, a number of threads are started that execute your 50 In this version, a number of threads are started that execute your
54 requests and signal their completion. You don't need thread support in 51 requests and signal their completion. You don't need thread support in
55 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
56 perl. In the future, this module might make use of the native aio 53 perl. In the future, this module might make use of the native aio
57 functions available on many operating systems. However, they are often 54 functions available on many operating systems. However, they are often
58 not well-supported or restricted (Linux doesn't allow them on normal 55 not well-supported or restricted (GNU/Linux doesn't allow them on normal
59 files currently, for example), and they would only support aio_read and 56 files currently, for example), and they would only support aio_read and
60 aio_write, so the remaining functionality would have to be implemented 57 aio_write, so the remaining functionality would have to be implemented
61 using threads anyway. 58 using threads anyway.
62 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
63 Although the module will work with in the presence of other (Perl-) 64 Although the module will work in the presence of other (Perl-) threads,
64 threads, it is currently not reentrant in any way, so use appropriate 65 it is currently not reentrant in any way, so use appropriate locking
65 locking yourself, always call "poll_cb" from within the same thread, or 66 yourself, always call "poll_cb" from within the same thread, or never
66 never call "poll_cb" (or 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;
67 108
68REQUEST ANATOMY AND LIFETIME 109REQUEST ANATOMY AND LIFETIME
69 Every "aio_*" function creates a request. which is a C data structure 110 Every "aio_*" function creates a request. which is a C data structure
70 not directly visible to Perl. 111 not directly visible to Perl.
71 112
… …
107 anymore (except possibly for the Perl object, but its connection to 148 anymore (except possibly for the Perl object, but its connection to
108 the actual aio request is severed and calling its methods will 149 the actual aio request is severed and calling its methods will
109 either do nothing or result in a runtime error). 150 either do nothing or result in a runtime error).
110 151
111FUNCTIONS 152FUNCTIONS
112 AIO FUNCTIONS 153 QUICK OVERVIEW
113 All the "aio_*" calls are more or less thin wrappers around the 154 This section simply lists the prototypes most of the functions for quick
114 syscall with the same name (sans "aio_"). The arguments are similar 155 reference. See the following sections for function-by-function
115 or identical, and they all accept an additional (and optional) 156 documentation.
116 $callback argument which must be a code reference. This code
117 reference will get called with the syscall return code (e.g. most
118 syscalls return -1 on error, unlike perl, which usually delivers
119 "false") as it's sole argument when the given syscall has been
120 executed asynchronously.
121 157
122 All functions expecting a filehandle keep a copy of the filehandle 158 aio_wd $pathname, $callback->($wd)
123 internally until the request has finished.
124
125 All requests return objects of type IO::AIO::REQ that allow further
126 manipulation of those requests while they are in-flight.
127
128 The pathnames you pass to these routines *must* be absolute and
129 encoded in byte form. The reason for the former is that at the time
130 the request is being executed, the current working directory could
131 have changed. Alternatively, you can make sure that you never change
132 the current working directory.
133
134 To encode pathnames to byte form, either make sure you either: a)
135 always pass in filenames you got from outside (command line, readdir
136 etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and
137 encode your pathnames to the locale (or other) encoding in effect in
138 the user environment, d) use Glib::filename_from_unicode on unicode
139 filenames or e) use something else.
140
141 $prev_pri = aioreq_pri [$pri]
142 Returns the priority value that would be used for the next
143 request and, if $pri is given, sets the priority for the next
144 aio request.
145
146 The default priority is 0, the minimum and maximum priorities
147 are -4 and 4, respectively. Requests with higher priority will
148 be serviced first.
149
150 The priority will be reset to 0 after each call to one of the
151 "aio_*" functions.
152
153 Example: open a file with low priority, then read something from
154 it with higher priority so the read request is serviced before
155 other low priority open requests (potentially spamming the
156 cache):
157
158 aioreq_pri -3;
159 aio_open ..., sub {
160 return unless $_[0];
161
162 aioreq_pri -2;
163 aio_read $_[0], ..., sub {
164 ...
165 };
166 };
167
168 aioreq_nice $pri_adjust
169 Similar to "aioreq_pri", but subtracts the given value from the
170 current priority, so effects are cumulative.
171
172 aio_open $pathname, $flags, $mode, $callback->($fh) 159 aio_open $pathname, $flags, $mode, $callback->($fh)
173 Asynchronously open or create a file and call the callback with
174 a newly created filehandle for the file.
175
176 The pathname passed to "aio_open" must be absolute. See API
177 NOTES, above, for an explanation.
178
179 The $flags argument is a bitmask. See the "Fcntl" module for a
180 list. They are the same as used by "sysopen".
181
182 Likewise, $mode specifies the mode of the newly created file, if
183 it didn't exist and "O_CREAT" has been given, just like perl's
184 "sysopen", except that it is mandatory (i.e. use 0 if you don't
185 create new files, and 0666 or 0777 if you do).
186
187 Example:
188
189 aio_open "/etc/passwd", O_RDONLY, 0, sub {
190 if ($_[0]) {
191 print "open successful, fh is $_[0]\n";
192 ...
193 } else {
194 die "open failed: $!\n";
195 }
196 };
197
198 aio_close $fh, $callback->($status) 160 aio_close $fh, $callback->($status)
199 Asynchronously close a file and call the callback with the 161 aio_seek $fh,$offset,$whence, $callback->($offs)
200 result code. *WARNING:* although accepted, you should not pass
201 in a perl filehandle here, as perl will likely close the file
202 descriptor another time when the filehandle is destroyed.
203 Normally, you can safely call perls "close" or just let
204 filehandles go out of scope.
205
206 This is supposed to be a bug in the API, so that might change.
207 It's therefore best to avoid this function.
208
209 aio_read $fh,$offset,$length, $data,$dataoffset, 162 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
210 $callback->($retval)
211 aio_write $fh,$offset,$length, $data,$dataoffset, 163 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
212 $callback->($retval)
213 Reads or writes "length" bytes from the specified "fh" and
214 "offset" into the scalar given by "data" and offset "dataoffset"
215 and calls the callback without the actual number of bytes read
216 (or -1 on error, just like the syscall).
217
218 The $data scalar *MUST NOT* be modified in any way while the
219 request is outstanding. Modifying it can result in segfaults or
220 WW3 (if the necessary/optional hardware is installed).
221
222 Example: Read 15 bytes at offset 7 into scalar $buffer, starting
223 at offset 0 within the scalar:
224
225 aio_read $fh, 7, 15, $buffer, 0, sub {
226 $_[0] > 0 or die "read error: $!";
227 print "read $_[0] bytes: <$buffer>\n";
228 };
229
230 aio_move $srcpath, $dstpath, $callback->($status)
231 Try to move the *file* (directories not supported as either
232 source or destination) from $srcpath to $dstpath and call the
233 callback with the 0 (error) or -1 ok.
234
235 This is a composite request that tries to rename(2) the file
236 first. If rename files with "EXDEV", it creates the destination
237 file with mode 0200 and copies the contents of the source file
238 into it using "aio_sendfile", followed by restoring atime,
239 mtime, access mode and uid/gid, in that order, and unlinking the
240 $srcpath.
241
242 If an error occurs, the partial destination file will be
243 unlinked, if possible, except when setting atime, mtime, access
244 mode and uid/gid, where errors are being ignored.
245
246 aio_sendfile $out_fh, $in_fh, $in_offset, $length, 164 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
247 $callback->($retval)
248 Tries to copy $length bytes from $in_fh to $out_fh. It starts
249 reading at byte offset $in_offset, and starts writing at the
250 current file offset of $out_fh. Because of that, it is not safe
251 to issue more than one "aio_sendfile" per $out_fh, as they will
252 interfere with each other.
253
254 This call tries to make use of a native "sendfile" syscall to
255 provide zero-copy operation. For this to work, $out_fh should
256 refer to a socket, and $in_fh should refer to mmap'able file.
257
258 If the native sendfile call fails or is not implemented, it will
259 be emulated, so you can call "aio_sendfile" on any type of
260 filehandle regardless of the limitations of the operating
261 system.
262
263 Please note, however, that "aio_sendfile" can read more bytes
264 from $in_fh than are written, and there is no way to find out
265 how many bytes have been read from "aio_sendfile" alone, as
266 "aio_sendfile" only provides the number of bytes written to
267 $out_fh. Only if the result value equals $length one can assume
268 that $length bytes have been read.
269
270 aio_readahead $fh,$offset,$length, $callback->($retval) 165 aio_readahead $fh,$offset,$length, $callback->($retval)
271 "aio_readahead" populates the page cache with data from a file
272 so that subsequent reads from that file will not block on disk
273 I/O. The $offset argument specifies the starting point from
274 which data is to be read and $length specifies the number of
275 bytes to be read. I/O is performed in whole pages, so that
276 offset is effectively rounded down to a page boundary and bytes
277 are read up to the next page boundary greater than or equal to
278 (off-set+length). "aio_readahead" does not read beyond the end
279 of the file. The current file offset of the file is left
280 unchanged.
281
282 If that syscall doesn't exist (likely if your OS isn't Linux) it
283 will be emulated by simply reading the data, which would have a
284 similar effect.
285
286 aio_stat $fh_or_path, $callback->($status) 166 aio_stat $fh_or_path, $callback->($status)
287 aio_lstat $fh, $callback->($status) 167 aio_lstat $fh, $callback->($status)
288 Works like perl's "stat" or "lstat" in void context. The 168 aio_statvfs $fh_or_path, $callback->($statvfs)
289 callback will be called after the stat and the results will be 169 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
290 available using "stat _" or "-s _" etc... 170 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
291 171 aio_chmod $fh_or_path, $mode, $callback->($status)
292 The pathname passed to "aio_stat" must be absolute. See API 172 aio_truncate $fh_or_path, $offset, $callback->($status)
293 NOTES, above, for an explanation. 173 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
294 174 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
295 Currently, the stats are always 64-bit-stats, i.e. instead of
296 returning an error when stat'ing a large file, the results will
297 be silently truncated unless perl itself is compiled with large
298 file support.
299
300 Example: Print the length of /etc/passwd:
301
302 aio_stat "/etc/passwd", sub {
303 $_[0] and die "stat failed: $!";
304 print "size is ", -s _, "\n";
305 };
306
307 aio_unlink $pathname, $callback->($status) 175 aio_unlink $pathname, $callback->($status)
308 Asynchronously unlink (delete) a file and call the callback with 176 aio_mknod $pathname, $mode, $dev, $callback->($status)
309 the result code.
310
311 aio_link $srcpath, $dstpath, $callback->($status) 177 aio_link $srcpath, $dstpath, $callback->($status)
312 Asynchronously create a new link to the existing object at
313 $srcpath at the path $dstpath and call the callback with the
314 result code.
315
316 aio_symlink $srcpath, $dstpath, $callback->($status) 178 aio_symlink $srcpath, $dstpath, $callback->($status)
317 Asynchronously create a new symbolic link to the existing object 179 aio_readlink $pathname, $callback->($link)
318 at $srcpath at the path $dstpath and call the callback with the 180 aio_realpath $pathname, $callback->($path)
319 result code.
320
321 aio_rename $srcpath, $dstpath, $callback->($status) 181 aio_rename $srcpath, $dstpath, $callback->($status)
322 Asynchronously rename the object at $srcpath to $dstpath, just 182 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
323 as rename(2) and call the callback with the result code. 183 aio_mkdir $pathname, $mode, $callback->($status)
324
325 aio_rmdir $pathname, $callback->($status) 184 aio_rmdir $pathname, $callback->($status)
326 Asynchronously rmdir (delete) a directory and call the callback
327 with the result code.
328
329 aio_readdir $pathname, $callback->($entries) 185 aio_readdir $pathname, $callback->($entries)
330 Unlike the POSIX call of the same name, "aio_readdir" reads an 186 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
331 entire directory (i.e. opendir + readdir + closedir). The 187 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
332 entries will not be sorted, and will NOT include the "." and 188 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
333 ".." entries.
334
335 The callback a single argument which is either "undef" or an
336 array-ref with the filenames.
337
338 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 189 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
339 Scans a directory (similar to "aio_readdir") but additionally 190 aio_load $pathname, $data, $callback->($status)
340 tries to efficiently separate the entries of directory $path 191 aio_copy $srcpath, $dstpath, $callback->($status)
341 into two sets of names, directories you can recurse into 192 aio_move $srcpath, $dstpath, $callback->($status)
342 (directories), and ones you cannot recurse into (everything 193 aio_rmtree $pathname, $callback->($status)
343 else, including symlinks to directories). 194 aio_fcntl $fh, $cmd, $arg, $callback->($status)
344 195 aio_ioctl $fh, $request, $buf, $callback->($status)
345 "aio_scandir" is a composite request that creates of many sub 196 aio_sync $callback->($status)
346 requests_ $maxreq specifies the maximum number of outstanding 197 aio_syncfs $fh, $callback->($status)
347 aio requests that this function generates. If it is "<= 0", then
348 a suitable default will be chosen (currently 6).
349
350 On error, the callback is called without arguments, otherwise it
351 receives two array-refs with path-relative entry names.
352
353 Example:
354
355 aio_scandir $dir, 0, sub {
356 my ($dirs, $nondirs) = @_;
357 print "real directories: @$dirs\n";
358 print "everything else: @$nondirs\n";
359 };
360
361 Implementation notes.
362
363 The "aio_readdir" cannot be avoided, but "stat()"'ing every
364 entry can.
365
366 After reading the directory, the modification time, size etc. of
367 the directory before and after the readdir is checked, and if
368 they match (and isn't the current time), the link count will be
369 used to decide how many entries are directories (if >= 2).
370 Otherwise, no knowledge of the number of subdirectories will be
371 assumed.
372
373 Then entries will be sorted into likely directories (everything
374 without a non-initial dot currently) and likely non-directories
375 (everything else). Then every entry plus an appended "/." will
376 be "stat"'ed, likely directories first. If that succeeds, it
377 assumes that the entry is a directory or a symlink to directory
378 (which will be checked seperately). This is often faster than
379 stat'ing the entry itself because filesystems might detect the
380 type of the entry without reading the inode data (e.g. ext2fs
381 filetype feature).
382
383 If the known number of directories (link count - 2) has been
384 reached, the rest of the entries is assumed to be
385 non-directories.
386
387 This only works with certainty on POSIX (= UNIX) filesystems,
388 which fortunately are the vast majority of filesystems around.
389
390 It will also likely work on non-POSIX filesystems with reduced
391 efficiency as those tend to return 0 or 1 as link counts, which
392 disables the directory counting heuristic.
393
394 aio_fsync $fh, $callback->($status) 198 aio_fsync $fh, $callback->($status)
395 Asynchronously call fsync on the given filehandle and call the
396 callback with the fsync result code.
397
398 aio_fdatasync $fh, $callback->($status) 199 aio_fdatasync $fh, $callback->($status)
399 Asynchronously call fdatasync on the given filehandle and call 200 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
400 the callback with the fdatasync result code. 201 aio_pathsync $pathname, $callback->($status)
401 202 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
402 If this call isn't available because your OS lacks it or it 203 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
403 couldn't be detected, it will be emulated by calling "fsync" 204 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
404 instead. 205 aio_mlockall $flags, $callback->($status)
405
406 aio_group $callback->(...) 206 aio_group $callback->(...)
407 This is a very special aio request: Instead of doing something,
408 it is a container for other aio requests, which is useful if you
409 want to bundle many requests into a single, composite, request
410 with a definite callback and the ability to cancel the whole
411 request with its subrequests.
412
413 Returns an object of class IO::AIO::GRP. See its documentation
414 below for more info.
415
416 Example:
417
418 my $grp = aio_group sub {
419 print "all stats done\n";
420 };
421
422 add $grp
423 (aio_stat ...),
424 (aio_stat ...),
425 ...;
426
427 aio_nop $callback->() 207 aio_nop $callback->()
428 This is a special request - it does nothing in itself and is
429 only used for side effects, such as when you want to add a dummy
430 request to a group so that finishing the requests in the group
431 depends on executing the given code.
432 208
433 While this request does nothing, it still goes through the 209 $prev_pri = aioreq_pri [$pri]
434 execution phase and still requires a worker thread. Thus, the 210 aioreq_nice $pri_adjust
435 callback will not be executed immediately but only after other
436 requests in the queue have entered their execution phase. This
437 can be used to measure request latency.
438 211
439 IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED* 212 IO::AIO::poll_wait
440 Mainly used for debugging and benchmarking, this aio request 213 IO::AIO::poll_cb
441 puts one of the request workers to sleep for the given time. 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 IO::AIO::reinit
442 227
443 While it is theoretically handy to have simple I/O scheduling 228 $nfd = IO::AIO::get_fdlimit
444 requests like sleep and file handle readable/writable, the 229 IO::AIO::min_fdlimit $nfd
445 overhead this creates is immense (it blocks a thread for a long
446 time) so do not use this function except to put your application
447 under artificial I/O pressure.
448 230
449 IO::AIO::REQ CLASS 231 IO::AIO::sendfile $ofh, $ifh, $offset, $count
450 All non-aggregate "aio_*" functions return an object of this class 232 IO::AIO::fadvise $fh, $offset, $len, $advice
451 when called in non-void context.
452 233
453 cancel $req 234 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
454 Cancels the request, if possible. Has the effect of skipping 235 IO::AIO::munmap $scalar
455 execution when entering the execute state and skipping calling 236 IO::AIO::mremap $scalar, $new_length, $flags[, $new_address]
456 the callback when entering the the result state, but will leave 237 IO::AIO::madvise $scalar, $offset, $length, $advice
457 the request otherwise untouched. That means that requests that 238 IO::AIO::mprotect $scalar, $offset, $length, $protect
458 currently execute will not be stopped and resources held by the 239 IO::AIO::munlock $scalar, $offset = 0, $length = undef
459 request will not be freed prematurely. 240 IO::AIO::munlockall
460 241
461 cb $req $callback->(...) 242 # stat extensions
462 Replace (or simply set) the callback registered to the request. 243 $counter = IO::AIO::st_gen
244 $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime
245 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
246 $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec
247 $seconds = IO::AIO::st_btimesec
248 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
463 249
464 IO::AIO::GRP CLASS 250 # very much unportable syscalls
465 This class is a subclass of IO::AIO::REQ, so all its methods apply 251 IO::AIO::accept4 $r_fh, $sockaddr, $sockaddr_len, $flags
466 to objects of this class, too. 252 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
253 IO::AIO::tee $r_fh, $w_fh, $length, $flags
254 $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
255 ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
256 $fh = IO::AIO::memfd_create $pathname[, $flags]
257 $fh = IO::AIO::eventfd [$initval, [$flags]]
258 $fh = IO::AIO::timerfd_create $clockid[, $flags]
259 ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
260 ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
467 261
468 A IO::AIO::GRP object is a special request that can contain multiple 262 API NOTES
469 other aio requests. 263 All the "aio_*" calls are more or less thin wrappers around the syscall
264 with the same name (sans "aio_"). The arguments are similar or
265 identical, and they all accept an additional (and optional) $callback
266 argument which must be a code reference. This code reference will be
267 called after the syscall has been executed in an asynchronous fashion.
268 The results of the request will be passed as arguments to the callback
269 (and, if an error occured, in $!) - for most requests the syscall return
270 code (e.g. most syscalls return -1 on error, unlike perl, which usually
271 delivers "false").
470 272
471 You create one by calling the "aio_group" constructing function with 273 Some requests (such as "aio_readdir") pass the actual results and
472 a callback that will be called when all contained requests have 274 communicate failures by passing "undef".
473 entered the "done" state:
474 275
475 my $grp = aio_group sub { 276 All functions expecting a filehandle keep a copy of the filehandle
476 print "all requests are done\n"; 277 internally until the request has finished.
477 };
478 278
479 You add requests by calling the "add" method with one or more 279 All functions return request objects of type IO::AIO::REQ that allow
480 "IO::AIO::REQ" objects: 280 further manipulation of those requests while they are in-flight.
481 281
482 $grp->add (aio_unlink "..."); 282 The pathnames you pass to these routines *should* be absolute. The
283 reason for this is that at the time the request is being executed, the
284 current working directory could have changed. Alternatively, you can
285 make sure that you never change the current working directory anywhere
286 in the program and then use relative paths. You can also take advantage
287 of IO::AIOs working directory abstraction, that lets you specify paths
288 relative to some previously-opened "working directory object" - see the
289 description of the "IO::AIO::WD" class later in this document.
483 290
484 add $grp aio_stat "...", sub { 291 To encode pathnames as octets, either make sure you either: a) always
485 $_[0] or return $grp->result ("error"); 292 pass in filenames you got from outside (command line, readdir etc.)
293 without tinkering, b) are in your native filesystem encoding, c) use the
294 Encode module and encode your pathnames to the locale (or other)
295 encoding in effect in the user environment, d) use
296 Glib::filename_from_unicode on unicode filenames or e) use something
297 else to ensure your scalar has the correct contents.
486 298
487 # add another request dynamically, if first succeeded 299 This works, btw. independent of the internal UTF-8 bit, which IO::AIO
300 handles correctly whether it is set or not.
301
302 AIO REQUEST FUNCTIONS
303 $prev_pri = aioreq_pri [$pri]
304 Returns the priority value that would be used for the next request
305 and, if $pri is given, sets the priority for the next aio request.
306
307 The default priority is 0, the minimum and maximum priorities are -4
308 and 4, respectively. Requests with higher priority will be serviced
309 first.
310
311 The priority will be reset to 0 after each call to one of the
312 "aio_*" functions.
313
314 Example: open a file with low priority, then read something from it
315 with higher priority so the read request is serviced before other
316 low priority open requests (potentially spamming the cache):
317
318 aioreq_pri -3;
488 add $grp aio_open "...", sub { 319 aio_open ..., sub {
489 $grp->result ("ok"); 320 return unless $_[0];
321
322 aioreq_pri -2;
323 aio_read $_[0], ..., sub {
324 ...
490 }; 325 };
491 }; 326 };
492 327
328 aioreq_nice $pri_adjust
329 Similar to "aioreq_pri", but subtracts the given value from the
330 current priority, so the effect is cumulative.
331
332 aio_open $pathname, $flags, $mode, $callback->($fh)
333 Asynchronously open or create a file and call the callback with a
334 newly created filehandle for the file (or "undef" in case of an
335 error).
336
337 The pathname passed to "aio_open" must be absolute. See API NOTES,
338 above, for an explanation.
339
340 The $flags argument is a bitmask. See the "Fcntl" module for a list.
341 They are the same as used by "sysopen".
342
343 Likewise, $mode specifies the mode of the newly created file, if it
344 didn't exist and "O_CREAT" has been given, just like perl's
345 "sysopen", except that it is mandatory (i.e. use 0 if you don't
346 create new files, and 0666 or 0777 if you do). Note that the $mode
347 will be modified by the umask in effect then the request is being
348 executed, so better never change the umask.
349
350 Example:
351
352 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
353 if ($_[0]) {
354 print "open successful, fh is $_[0]\n";
355 ...
356 } else {
357 die "open failed: $!\n";
358 }
359 };
360
361 In addition to all the common open modes/flags ("O_RDONLY",
362 "O_WRONLY", "O_RDWR", "O_CREAT", "O_TRUNC", "O_EXCL" and
363 "O_APPEND"), the following POSIX and non-POSIX constants are
364 available (missing ones on your system are, as usual, 0):
365
366 "O_ASYNC", "O_DIRECT", "O_NOATIME", "O_CLOEXEC", "O_NOCTTY",
367 "O_NOFOLLOW", "O_NONBLOCK", "O_EXEC", "O_SEARCH", "O_DIRECTORY",
368 "O_DSYNC", "O_RSYNC", "O_SYNC", "O_PATH", "O_TMPFILE", "O_TTY_INIT"
369 and "O_ACCMODE".
370
371 aio_close $fh, $callback->($status)
372 Asynchronously close a file and call the callback with the result
373 code.
374
375 Unfortunately, you can't do this to perl. Perl *insists* very
376 strongly on closing the file descriptor associated with the
377 filehandle itself.
378
379 Therefore, "aio_close" will not close the filehandle - instead it
380 will use dup2 to overwrite the file descriptor with the write-end of
381 a pipe (the pipe fd will be created on demand and will be cached).
382
383 Or in other words: the file descriptor will be closed, but it will
384 not be free for reuse until the perl filehandle is closed.
385
386 aio_seek $fh, $offset, $whence, $callback->($offs)
387 Seeks the filehandle to the new $offset, similarly to perl's
388 "sysseek". The $whence can use the traditional values (0 for
389 "IO::AIO::SEEK_SET", 1 for "IO::AIO::SEEK_CUR" or 2 for
390 "IO::AIO::SEEK_END").
391
392 The resulting absolute offset will be passed to the callback, or -1
393 in case of an error.
394
395 In theory, the $whence constants could be different than the
396 corresponding values from Fcntl, but perl guarantees they are the
397 same, so don't panic.
398
399 As a GNU/Linux (and maybe Solaris) extension, also the constants
400 "IO::AIO::SEEK_DATA" and "IO::AIO::SEEK_HOLE" are available, if they
401 could be found. No guarantees about suitability for use in
402 "aio_seek" or Perl's "sysseek" can be made though, although I would
403 naively assume they "just work".
404
405 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
406 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
407 Reads or writes $length bytes from or to the specified $fh and
408 $offset into the scalar given by $data and offset $dataoffset and
409 calls the callback with the actual number of bytes transferred (or
410 -1 on error, just like the syscall).
411
412 "aio_read" will, like "sysread", shrink or grow the $data scalar to
413 offset plus the actual number of bytes read.
414
415 If $offset is undefined, then the current file descriptor offset
416 will be used (and updated), otherwise the file descriptor offset
417 will not be changed by these calls.
418
419 If $length is undefined in "aio_write", use the remaining length of
420 $data.
421
422 If $dataoffset is less than zero, it will be counted from the end of
423 $data.
424
425 The $data scalar *MUST NOT* be modified in any way while the request
426 is outstanding. Modifying it can result in segfaults or World War
427 III (if the necessary/optional hardware is installed).
428
429 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at
430 offset 0 within the scalar:
431
432 aio_read $fh, 7, 15, $buffer, 0, sub {
433 $_[0] > 0 or die "read error: $!";
434 print "read $_[0] bytes: <$buffer>\n";
435 };
436
437 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
438 Tries to copy $length bytes from $in_fh to $out_fh. It starts
439 reading at byte offset $in_offset, and starts writing at the current
440 file offset of $out_fh. Because of that, it is not safe to issue
441 more than one "aio_sendfile" per $out_fh, as they will interfere
442 with each other. The same $in_fh works fine though, as this function
443 does not move or use the file offset of $in_fh.
444
445 Please note that "aio_sendfile" can read more bytes from $in_fh than
446 are written, and there is no way to find out how many more bytes
447 have been read from "aio_sendfile" alone, as "aio_sendfile" only
448 provides the number of bytes written to $out_fh. Only if the result
449 value equals $length one can assume that $length bytes have been
450 read.
451
452 Unlike with other "aio_" functions, it makes a lot of sense to use
453 "aio_sendfile" on non-blocking sockets, as long as one end
454 (typically the $in_fh) is a file - the file I/O will then be
455 asynchronous, while the socket I/O will be non-blocking. Note,
456 however, that you can run into a trap where "aio_sendfile" reads
457 some data with readahead, then fails to write all data, and when the
458 socket is ready the next time, the data in the cache is already
459 lost, forcing "aio_sendfile" to again hit the disk. Explicit
460 "aio_read" + "aio_write" let's you better control resource usage.
461
462 This call tries to make use of a native "sendfile"-like syscall to
463 provide zero-copy operation. For this to work, $out_fh should refer
464 to a socket, and $in_fh should refer to an mmap'able file.
465
466 If a native sendfile cannot be found or it fails with "ENOSYS",
467 "EINVAL", "ENOTSUP", "EOPNOTSUPP", "EAFNOSUPPORT", "EPROTOTYPE" or
468 "ENOTSOCK", it will be emulated, so you can call "aio_sendfile" on
469 any type of filehandle regardless of the limitations of the
470 operating system.
471
472 As native sendfile syscalls (as practically any non-POSIX interface
473 hacked together in a hurry to improve benchmark numbers) tend to be
474 rather buggy on many systems, this implementation tries to work
475 around some known bugs in Linux and FreeBSD kernels (probably
476 others, too), but that might fail, so you really really should check
477 the return value of "aio_sendfile" - fewer bytes than expected might
478 have been transferred.
479
480 aio_readahead $fh,$offset,$length, $callback->($retval)
481 "aio_readahead" populates the page cache with data from a file so
482 that subsequent reads from that file will not block on disk I/O. The
483 $offset argument specifies the starting point from which data is to
484 be read and $length specifies the number of bytes to be read. I/O is
485 performed in whole pages, so that offset is effectively rounded down
486 to a page boundary and bytes are read up to the next page boundary
487 greater than or equal to (off-set+length). "aio_readahead" does not
488 read beyond the end of the file. The current file offset of the file
489 is left unchanged.
490
491 If that syscall doesn't exist (likely if your kernel isn't Linux) it
492 will be emulated by simply reading the data, which would have a
493 similar effect.
494
495 aio_stat $fh_or_path, $callback->($status)
496 aio_lstat $fh, $callback->($status)
497 Works almost exactly like perl's "stat" or "lstat" in void context.
498 The callback will be called after the stat and the results will be
499 available using "stat _" or "-s _" and other tests (with the
500 exception of "-B" and "-T").
501
502 The pathname passed to "aio_stat" must be absolute. See API NOTES,
503 above, for an explanation.
504
505 Currently, the stats are always 64-bit-stats, i.e. instead of
506 returning an error when stat'ing a large file, the results will be
507 silently truncated unless perl itself is compiled with large file
508 support.
509
510 To help interpret the mode and dev/rdev stat values, IO::AIO offers
511 the following constants and functions (if not implemented, the
512 constants will be 0 and the functions will either "croak" or fall
513 back on traditional behaviour).
514
515 "S_IFMT", "S_IFIFO", "S_IFCHR", "S_IFBLK", "S_IFLNK", "S_IFREG",
516 "S_IFDIR", "S_IFWHT", "S_IFSOCK", "IO::AIO::major $dev_t",
517 "IO::AIO::minor $dev_t", "IO::AIO::makedev $major, $minor".
518
519 To access higher resolution stat timestamps, see "SUBSECOND STAT
520 TIME ACCESS".
521
522 Example: Print the length of /etc/passwd:
523
524 aio_stat "/etc/passwd", sub {
525 $_[0] and die "stat failed: $!";
526 print "size is ", -s _, "\n";
527 };
528
529 aio_statvfs $fh_or_path, $callback->($statvfs)
530 Works like the POSIX "statvfs" or "fstatvfs" syscalls, depending on
531 whether a file handle or path was passed.
532
533 On success, the callback is passed a hash reference with the
534 following members: "bsize", "frsize", "blocks", "bfree", "bavail",
535 "files", "ffree", "favail", "fsid", "flag" and "namemax". On
536 failure, "undef" is passed.
537
538 The following POSIX IO::AIO::ST_* constants are defined: "ST_RDONLY"
539 and "ST_NOSUID".
540
541 The following non-POSIX IO::AIO::ST_* flag masks are defined to
542 their correct value when available, or to 0 on systems that do not
543 support them: "ST_NODEV", "ST_NOEXEC", "ST_SYNCHRONOUS",
544 "ST_MANDLOCK", "ST_WRITE", "ST_APPEND", "ST_IMMUTABLE",
545 "ST_NOATIME", "ST_NODIRATIME" and "ST_RELATIME".
546
547 Example: stat "/wd" and dump out the data if successful.
548
549 aio_statvfs "/wd", sub {
550 my $f = $_[0]
551 or die "statvfs: $!";
552
553 use Data::Dumper;
554 say Dumper $f;
555 };
556
557 # result:
558 {
559 bsize => 1024,
560 bfree => 4333064312,
561 blocks => 10253828096,
562 files => 2050765568,
563 flag => 4096,
564 favail => 2042092649,
565 bavail => 4333064312,
566 ffree => 2042092649,
567 namemax => 255,
568 frsize => 1024,
569 fsid => 1810
570 }
571
572 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
573 Works like perl's "utime" function (including the special case of
574 $atime and $mtime being undef). Fractional times are supported if
575 the underlying syscalls support them.
576
577 When called with a pathname, uses utimensat(2) or utimes(2) if
578 available, otherwise utime(2). If called on a file descriptor, uses
579 futimens(2) or futimes(2) if available, otherwise returns ENOSYS, so
580 this is not portable.
581
582 Examples:
583
584 # set atime and mtime to current time (basically touch(1)):
585 aio_utime "path", undef, undef;
586 # set atime to current time and mtime to beginning of the epoch:
587 aio_utime "path", time, undef; # undef==0
588
589 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
590 Works like perl's "chown" function, except that "undef" for either
591 $uid or $gid is being interpreted as "do not change" (but -1 can
592 also be used).
593
594 Examples:
595
596 # same as "chown root path" in the shell:
597 aio_chown "path", 0, -1;
598 # same as above:
599 aio_chown "path", 0, undef;
600
601 aio_truncate $fh_or_path, $offset, $callback->($status)
602 Works like truncate(2) or ftruncate(2).
603
604 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
605 Allocates or frees disk space according to the $mode argument. See
606 the linux "fallocate" documentation for details.
607
608 $mode is usually 0 or "IO::AIO::FALLOC_FL_KEEP_SIZE" to allocate
609 space, or "IO::AIO::FALLOC_FL_PUNCH_HOLE |
610 IO::AIO::FALLOC_FL_KEEP_SIZE", to deallocate a file range.
611
612 IO::AIO also supports "FALLOC_FL_COLLAPSE_RANGE", to remove a range
613 (without leaving a hole), "FALLOC_FL_ZERO_RANGE", to zero a range,
614 "FALLOC_FL_INSERT_RANGE" to insert a range and
615 "FALLOC_FL_UNSHARE_RANGE" to unshare shared blocks (see your
616 fallocate(2) manpage).
617
618 The file system block size used by "fallocate" is presumably the
619 "f_bsize" returned by "statvfs", but different filesystems and
620 filetypes can dictate other limitations.
621
622 If "fallocate" isn't available or cannot be emulated (currently no
623 emulation will be attempted), passes -1 and sets $! to "ENOSYS".
624
625 aio_chmod $fh_or_path, $mode, $callback->($status)
626 Works like perl's "chmod" function.
627
628 aio_unlink $pathname, $callback->($status)
629 Asynchronously unlink (delete) a file and call the callback with the
630 result code.
631
632 aio_mknod $pathname, $mode, $dev, $callback->($status)
633 [EXPERIMENTAL]
634
635 Asynchronously create a device node (or fifo). See mknod(2).
636
637 The only (POSIX-) portable way of calling this function is:
638
639 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
640
641 See "aio_stat" for info about some potentially helpful extra
642 constants and functions.
643
644 aio_link $srcpath, $dstpath, $callback->($status)
645 Asynchronously create a new link to the existing object at $srcpath
646 at the path $dstpath and call the callback with the result code.
647
648 aio_symlink $srcpath, $dstpath, $callback->($status)
649 Asynchronously create a new symbolic link to the existing object at
650 $srcpath at the path $dstpath and call the callback with the result
651 code.
652
653 aio_readlink $pathname, $callback->($link)
654 Asynchronously read the symlink specified by $path and pass it to
655 the callback. If an error occurs, nothing or undef gets passed to
656 the callback.
657
658 aio_realpath $pathname, $callback->($path)
659 Asynchronously make the path absolute and resolve any symlinks in
660 $path. The resulting path only consists of directories (same as
661 Cwd::realpath).
662
663 This request can be used to get the absolute path of the current
664 working directory by passing it a path of . (a single dot).
665
666 aio_rename $srcpath, $dstpath, $callback->($status)
667 Asynchronously rename the object at $srcpath to $dstpath, just as
668 rename(2) and call the callback with the result code.
669
670 On systems that support the AIO::WD working directory abstraction
671 natively, the case "[$wd, "."]" as $srcpath is specialcased -
672 instead of failing, "rename" is called on the absolute path of $wd.
673
674 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
675 Basically a version of "aio_rename" with an additional $flags
676 argument. Calling this with "$flags=0" is the same as calling
677 "aio_rename".
678
679 Non-zero flags are currently only supported on GNU/Linux systems
680 that support renameat2. Other systems fail with "ENOSYS" in this
681 case.
682
683 The following constants are available (missing ones are, as usual
684 0), see renameat2(2) for details:
685
686 "IO::AIO::RENAME_NOREPLACE", "IO::AIO::RENAME_EXCHANGE" and
687 "IO::AIO::RENAME_WHITEOUT".
688
689 aio_mkdir $pathname, $mode, $callback->($status)
690 Asynchronously mkdir (create) a directory and call the callback with
691 the result code. $mode will be modified by the umask at the time the
692 request is executed, so do not change your umask.
693
694 aio_rmdir $pathname, $callback->($status)
695 Asynchronously rmdir (delete) a directory and call the callback with
696 the result code.
697
698 On systems that support the AIO::WD working directory abstraction
699 natively, the case "[$wd, "."]" is specialcased - instead of
700 failing, "rmdir" is called on the absolute path of $wd.
701
702 aio_readdir $pathname, $callback->($entries)
703 Unlike the POSIX call of the same name, "aio_readdir" reads an
704 entire directory (i.e. opendir + readdir + closedir). The entries
705 will not be sorted, and will NOT include the "." and ".." entries.
706
707 The callback is passed a single argument which is either "undef" or
708 an array-ref with the filenames.
709
710 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
711 Quite similar to "aio_readdir", but the $flags argument allows one
712 to tune behaviour and output format. In case of an error, $entries
713 will be "undef".
714
715 The flags are a combination of the following constants, ORed
716 together (the flags will also be passed to the callback, possibly
717 modified):
718
719 IO::AIO::READDIR_DENTS
720 Normally the callback gets an arrayref consisting of names only
721 (as with "aio_readdir"). If this flag is set, then the callback
722 gets an arrayref with "[$name, $type, $inode]" arrayrefs, each
723 describing a single directory entry in more detail:
724
725 $name is the name of the entry.
726
727 $type is one of the "IO::AIO::DT_xxx" constants:
728
729 "IO::AIO::DT_UNKNOWN", "IO::AIO::DT_FIFO", "IO::AIO::DT_CHR",
730 "IO::AIO::DT_DIR", "IO::AIO::DT_BLK", "IO::AIO::DT_REG",
731 "IO::AIO::DT_LNK", "IO::AIO::DT_SOCK", "IO::AIO::DT_WHT".
732
733 "IO::AIO::DT_UNKNOWN" means just that: readdir does not know. If
734 you need to know, you have to run stat yourself. Also, for
735 speed/memory reasons, the $type scalars are read-only: you must
736 not modify them.
737
738 $inode is the inode number (which might not be exact on systems
739 with 64 bit inode numbers and 32 bit perls). This field has
740 unspecified content on systems that do not deliver the inode
741 information.
742
743 IO::AIO::READDIR_DIRS_FIRST
744 When this flag is set, then the names will be returned in an
745 order where likely directories come first, in optimal stat
746 order. This is useful when you need to quickly find directories,
747 or you want to find all directories while avoiding to stat()
748 each entry.
749
750 If the system returns type information in readdir, then this is
751 used to find directories directly. Otherwise, likely directories
752 are names beginning with ".", or otherwise names with no dots,
753 of which names with short names are tried first.
754
755 IO::AIO::READDIR_STAT_ORDER
756 When this flag is set, then the names will be returned in an
757 order suitable for stat()'ing each one. That is, when you plan
758 to stat() most or all files in the given directory, then the
759 returned order will likely be faster.
760
761 If both this flag and "IO::AIO::READDIR_DIRS_FIRST" are
762 specified, then the likely dirs come first, resulting in a less
763 optimal stat order for stat'ing all entries, but likely a more
764 optimal order for finding subdirectories.
765
766 IO::AIO::READDIR_FOUND_UNKNOWN
767 This flag should not be set when calling "aio_readdirx".
768 Instead, it is being set by "aio_readdirx", when any of the
769 $type's found were "IO::AIO::DT_UNKNOWN". The absence of this
770 flag therefore indicates that all $type's are known, which can
771 be used to speed up some algorithms.
772
773 aio_slurp $pathname, $offset, $length, $data, $callback->($status)
774 Opens, reads and closes the given file. The data is put into $data,
775 which is resized as required.
776
777 If $offset is negative, then it is counted from the end of the file.
778
779 If $length is zero, then the remaining length of the file is used.
780 Also, in this case, the same limitations to modifying $data apply as
781 when IO::AIO::mmap is used, i.e. it must only be modified in-place
782 with "substr". If the size of the file is known, specifying a
783 non-zero $length results in a performance advantage.
784
785 This request is similar to the older "aio_load" request, but since
786 it is a single request, it might be more efficient to use.
787
788 Example: load /etc/passwd into $passwd.
789
790 my $passwd;
791 aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
792 $_[0] >= 0
793 or die "/etc/passwd: $!\n";
794
795 printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
796 print $passwd;
797 };
798 IO::AIO::flush;
799
800 aio_load $pathname, $data, $callback->($status)
801 This is a composite request that tries to fully load the given file
802 into memory. Status is the same as with aio_read.
803
804 Using "aio_slurp" might be more efficient, as it is a single
805 request.
806
807 aio_copy $srcpath, $dstpath, $callback->($status)
808 Try to copy the *file* (directories not supported as either source
809 or destination) from $srcpath to $dstpath and call the callback with
810 a status of 0 (ok) or -1 (error, see $!).
811
812 Existing destination files will be truncated.
813
814 This is a composite request that creates the destination file with
815 mode 0200 and copies the contents of the source file into it using
816 "aio_sendfile", followed by restoring atime, mtime, access mode and
817 uid/gid, in that order.
818
819 If an error occurs, the partial destination file will be unlinked,
820 if possible, except when setting atime, mtime, access mode and
821 uid/gid, where errors are being ignored.
822
823 aio_move $srcpath, $dstpath, $callback->($status)
824 Try to move the *file* (directories not supported as either source
825 or destination) from $srcpath to $dstpath and call the callback with
826 a status of 0 (ok) or -1 (error, see $!).
827
828 This is a composite request that tries to rename(2) the file first;
829 if rename fails with "EXDEV", it copies the file with "aio_copy"
830 and, if that is successful, unlinks the $srcpath.
831
832 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
833 Scans a directory (similar to "aio_readdir") but additionally tries
834 to efficiently separate the entries of directory $path into two sets
835 of names, directories you can recurse into (directories), and ones
836 you cannot recurse into (everything else, including symlinks to
837 directories).
838
839 "aio_scandir" is a composite request that generates many sub
840 requests. $maxreq specifies the maximum number of outstanding aio
841 requests that this function generates. If it is "<= 0", then a
842 suitable default will be chosen (currently 4).
843
844 On error, the callback is called without arguments, otherwise it
845 receives two array-refs with path-relative entry names.
846
847 Example:
848
849 aio_scandir $dir, 0, sub {
850 my ($dirs, $nondirs) = @_;
851 print "real directories: @$dirs\n";
852 print "everything else: @$nondirs\n";
853 };
854
855 Implementation notes.
856
857 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry
858 can.
859
860 If readdir returns file type information, then this is used directly
861 to find directories.
862
863 Otherwise, after reading the directory, the modification time, size
864 etc. of the directory before and after the readdir is checked, and
865 if they match (and isn't the current time), the link count will be
866 used to decide how many entries are directories (if >= 2).
867 Otherwise, no knowledge of the number of subdirectories will be
868 assumed.
869
870 Then entries will be sorted into likely directories a non-initial
871 dot currently) and likely non-directories (see "aio_readdirx"). Then
872 every entry plus an appended "/." will be "stat"'ed, likely
873 directories first, in order of their inode numbers. If that
874 succeeds, it assumes that the entry is a directory or a symlink to
875 directory (which will be checked separately). This is often faster
876 than stat'ing the entry itself because filesystems might detect the
877 type of the entry without reading the inode data (e.g. ext2fs
878 filetype feature), even on systems that cannot return the filetype
879 information on readdir.
880
881 If the known number of directories (link count - 2) has been
882 reached, the rest of the entries is assumed to be non-directories.
883
884 This only works with certainty on POSIX (= UNIX) filesystems, which
885 fortunately are the vast majority of filesystems around.
886
887 It will also likely work on non-POSIX filesystems with reduced
888 efficiency as those tend to return 0 or 1 as link counts, which
889 disables the directory counting heuristic.
890
891 aio_rmtree $pathname, $callback->($status)
892 Delete a directory tree starting (and including) $path, return the
893 status of the final "rmdir" only. This is a composite request that
894 uses "aio_scandir" to recurse into and rmdir directories, and unlink
895 everything else.
896
897 aio_fcntl $fh, $cmd, $arg, $callback->($status)
898 aio_ioctl $fh, $request, $buf, $callback->($status)
899 These work just like the "fcntl" and "ioctl" built-in functions,
900 except they execute asynchronously and pass the return value to the
901 callback.
902
903 Both calls can be used for a lot of things, some of which make more
904 sense to run asynchronously in their own thread, while some others
905 make less sense. For example, calls that block waiting for external
906 events, such as locking, will also lock down an I/O thread while it
907 is waiting, which can deadlock the whole I/O system. At the same
908 time, there might be no alternative to using a thread to wait.
909
910 So in general, you should only use these calls for things that do
911 (filesystem) I/O, not for things that wait for other events
912 (network, other processes), although if you are careful and know
913 what you are doing, you still can.
914
915 The following constants are available and can be used for normal
916 "ioctl" and "fcntl" as well (missing ones are, as usual 0):
917
918 "F_DUPFD_CLOEXEC",
919
920 "F_OFD_GETLK", "F_OFD_SETLK", "F_OFD_GETLKW",
921
922 "FIFREEZE", "FITHAW", "FITRIM", "FICLONE", "FICLONERANGE",
923 "FIDEDUPERANGE".
924
925 "F_ADD_SEALS", "F_GET_SEALS", "F_SEAL_SEAL", "F_SEAL_SHRINK",
926 "F_SEAL_GROW" and "F_SEAL_WRITE".
927
928 "FS_IOC_GETFLAGS", "FS_IOC_SETFLAGS", "FS_IOC_GETVERSION",
929 "FS_IOC_SETVERSION", "FS_IOC_FIEMAP".
930
931 "FS_IOC_FSGETXATTR", "FS_IOC_FSSETXATTR",
932 "FS_IOC_SET_ENCRYPTION_POLICY", "FS_IOC_GET_ENCRYPTION_PWSALT",
933 "FS_IOC_GET_ENCRYPTION_POLICY", "FS_KEY_DESCRIPTOR_SIZE".
934
935 "FS_SECRM_FL", "FS_UNRM_FL", "FS_COMPR_FL", "FS_SYNC_FL",
936 "FS_IMMUTABLE_FL", "FS_APPEND_FL", "FS_NODUMP_FL", "FS_NOATIME_FL",
937 "FS_DIRTY_FL", "FS_COMPRBLK_FL", "FS_NOCOMP_FL", "FS_ENCRYPT_FL",
938 "FS_BTREE_FL", "FS_INDEX_FL", "FS_JOURNAL_DATA_FL", "FS_NOTAIL_FL",
939 "FS_DIRSYNC_FL", "FS_TOPDIR_FL", "FS_FL_USER_MODIFIABLE".
940
941 "FS_XFLAG_REALTIME", "FS_XFLAG_PREALLOC", "FS_XFLAG_IMMUTABLE",
942 "FS_XFLAG_APPEND", "FS_XFLAG_SYNC", "FS_XFLAG_NOATIME",
943 "FS_XFLAG_NODUMP", "FS_XFLAG_RTINHERIT", "FS_XFLAG_PROJINHERIT",
944 "FS_XFLAG_NOSYMLINKS", "FS_XFLAG_EXTSIZE", "FS_XFLAG_EXTSZINHERIT",
945 "FS_XFLAG_NODEFRAG", "FS_XFLAG_FILESTREAM", "FS_XFLAG_DAX",
946 "FS_XFLAG_HASATTR",
947
948 aio_sync $callback->($status)
949 Asynchronously call sync and call the callback when finished.
950
951 aio_fsync $fh, $callback->($status)
952 Asynchronously call fsync on the given filehandle and call the
953 callback with the fsync result code.
954
955 aio_fdatasync $fh, $callback->($status)
956 Asynchronously call fdatasync on the given filehandle and call the
957 callback with the fdatasync result code.
958
959 If this call isn't available because your OS lacks it or it couldn't
960 be detected, it will be emulated by calling "fsync" instead.
961
962 aio_syncfs $fh, $callback->($status)
963 Asynchronously call the syncfs syscall to sync the filesystem
964 associated to the given filehandle and call the callback with the
965 syncfs result code. If syncfs is not available, calls sync(), but
966 returns -1 and sets errno to "ENOSYS" nevertheless.
967
968 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
969 Sync the data portion of the file specified by $offset and $length
970 to disk (but NOT the metadata), by calling the Linux-specific
971 sync_file_range call. If sync_file_range is not available or it
972 returns ENOSYS, then fdatasync or fsync is being substituted.
973
974 $flags can be a combination of
975 "IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE",
976 "IO::AIO::SYNC_FILE_RANGE_WRITE" and
977 "IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER": refer to the sync_file_range
978 manpage for details.
979
980 aio_pathsync $pathname, $callback->($status)
981 This request tries to open, fsync and close the given path. This is
982 a composite request intended to sync directories after directory
983 operations (E.g. rename). This might not work on all operating
984 systems or have any specific effect, but usually it makes sure that
985 directory changes get written to disc. It works for anything that
986 can be opened for read-only, not just directories.
987
988 Future versions of this function might fall back to other methods
989 when "fsync" on the directory fails (such as calling "sync").
990
991 Passes 0 when everything went ok, and -1 on error.
992
993 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC,
994 $callback->($status)
995 This is a rather advanced IO::AIO call, which only works on
996 mmap(2)ed scalars (see the "IO::AIO::mmap" function, although it
997 also works on data scalars managed by the Sys::Mmap or Mmap modules,
998 note that the scalar must only be modified in-place while an aio
999 operation is pending on it).
1000
1001 It calls the "msync" function of your OS, if available, with the
1002 memory area starting at $offset in the string and ending $length
1003 bytes later. If $length is negative, counts from the end, and if
1004 $length is "undef", then it goes till the end of the string. The
1005 flags can be either "IO::AIO::MS_ASYNC" or "IO::AIO::MS_SYNC", plus
1006 an optional "IO::AIO::MS_INVALIDATE".
1007
1008 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0,
1009 $callback->($status)
1010 This is a rather advanced IO::AIO call, which works best on
1011 mmap(2)ed scalars.
1012
1013 It touches (reads or writes) all memory pages in the specified range
1014 inside the scalar. All caveats and parameters are the same as for
1015 "aio_msync", above, except for flags, which must be either 0 (which
1016 reads all pages and ensures they are instantiated) or
1017 "IO::AIO::MT_MODIFY", which modifies the memory pages (by reading
1018 and writing an octet from it, which dirties the page).
1019
1020 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1021 This is a rather advanced IO::AIO call, which works best on
1022 mmap(2)ed scalars.
1023
1024 It reads in all the pages of the underlying storage into memory (if
1025 any) and locks them, so they are not getting swapped/paged out or
1026 removed.
1027
1028 If $length is undefined, then the scalar will be locked till the
1029 end.
1030
1031 On systems that do not implement "mlock", this function returns -1
1032 and sets errno to "ENOSYS".
1033
1034 Note that the corresponding "munlock" is synchronous and is
1035 documented under "MISCELLANEOUS FUNCTIONS".
1036
1037 Example: open a file, mmap and mlock it - both will be undone when
1038 $data gets destroyed.
1039
1040 open my $fh, "<", $path or die "$path: $!";
1041 my $data;
1042 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1043 aio_mlock $data; # mlock in background
1044
1045 aio_mlockall $flags, $callback->($status)
1046 Calls the "mlockall" function with the given $flags (a combination
1047 of "IO::AIO::MCL_CURRENT", "IO::AIO::MCL_FUTURE" and
1048 "IO::AIO::MCL_ONFAULT").
1049
1050 On systems that do not implement "mlockall", this function returns
1051 -1 and sets errno to "ENOSYS". Similarly, flag combinations not
1052 supported by the system result in a return value of -1 with errno
1053 being set to "EINVAL".
1054
1055 Note that the corresponding "munlockall" is synchronous and is
1056 documented under "MISCELLANEOUS FUNCTIONS".
1057
1058 Example: asynchronously lock all current and future pages into
1059 memory.
1060
1061 aio_mlockall IO::AIO::MCL_FUTURE;
1062
1063 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
1064 Queries the extents of the given file (by calling the Linux "FIEMAP"
1065 ioctl, see <http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for
1066 details). If the ioctl is not available on your OS, then this
1067 request will fail with "ENOSYS".
1068
1069 $start is the starting offset to query extents for, $length is the
1070 size of the range to query - if it is "undef", then the whole file
1071 will be queried.
1072
1073 $flags is a combination of flags ("IO::AIO::FIEMAP_FLAG_SYNC" or
1074 "IO::AIO::FIEMAP_FLAG_XATTR" - "IO::AIO::FIEMAP_FLAGS_COMPAT" is
1075 also exported), and is normally 0 or "IO::AIO::FIEMAP_FLAG_SYNC" to
1076 query the data portion.
1077
1078 $count is the maximum number of extent records to return. If it is
1079 "undef", then IO::AIO queries all extents of the range. As a very
1080 special case, if it is 0, then the callback receives the number of
1081 extents instead of the extents themselves (which is unreliable, see
1082 below).
1083
1084 If an error occurs, the callback receives no arguments. The special
1085 "errno" value "IO::AIO::EBADR" is available to test for flag errors.
1086
1087 Otherwise, the callback receives an array reference with extent
1088 structures. Each extent structure is an array reference itself, with
1089 the following members:
1090
1091 [$logical, $physical, $length, $flags]
1092
1093 Flags is any combination of the following flag values (typically
1094 either 0 or "IO::AIO::FIEMAP_EXTENT_LAST" (1)):
1095
1096 "IO::AIO::FIEMAP_EXTENT_LAST", "IO::AIO::FIEMAP_EXTENT_UNKNOWN",
1097 "IO::AIO::FIEMAP_EXTENT_DELALLOC", "IO::AIO::FIEMAP_EXTENT_ENCODED",
1098 "IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED",
1099 "IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED",
1100 "IO::AIO::FIEMAP_EXTENT_DATA_INLINE",
1101 "IO::AIO::FIEMAP_EXTENT_DATA_TAIL",
1102 "IO::AIO::FIEMAP_EXTENT_UNWRITTEN", "IO::AIO::FIEMAP_EXTENT_MERGED"
1103 or "IO::AIO::FIEMAP_EXTENT_SHARED".
1104
1105 At the time of this writing (Linux 3.2), this request is unreliable
1106 unless $count is "undef", as the kernel has all sorts of bugs
1107 preventing it to return all extents of a range for files with a
1108 large number of extents. The code (only) works around all these
1109 issues if $count is "undef".
1110
1111 aio_group $callback->(...)
1112 This is a very special aio request: Instead of doing something, it
1113 is a container for other aio requests, which is useful if you want
1114 to bundle many requests into a single, composite, request with a
1115 definite callback and the ability to cancel the whole request with
1116 its subrequests.
1117
1118 Returns an object of class IO::AIO::GRP. See its documentation below
1119 for more info.
1120
1121 Example:
1122
1123 my $grp = aio_group sub {
1124 print "all stats done\n";
1125 };
1126
1127 add $grp
1128 (aio_stat ...),
1129 (aio_stat ...),
1130 ...;
1131
1132 aio_nop $callback->()
1133 This is a special request - it does nothing in itself and is only
1134 used for side effects, such as when you want to add a dummy request
1135 to a group so that finishing the requests in the group depends on
1136 executing the given code.
1137
1138 While this request does nothing, it still goes through the execution
1139 phase and still requires a worker thread. Thus, the callback will
1140 not be executed immediately but only after other requests in the
1141 queue have entered their execution phase. This can be used to
1142 measure request latency.
1143
1144 IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
1145 Mainly used for debugging and benchmarking, this aio request puts
1146 one of the request workers to sleep for the given time.
1147
1148 While it is theoretically handy to have simple I/O scheduling
1149 requests like sleep and file handle readable/writable, the overhead
1150 this creates is immense (it blocks a thread for a long time) so do
1151 not use this function except to put your application under
1152 artificial I/O pressure.
1153
1154 IO::AIO::WD - multiple working directories
1155 Your process only has one current working directory, which is used by
1156 all threads. This makes it hard to use relative paths (some other
1157 component could call "chdir" at any time, and it is hard to control when
1158 the path will be used by IO::AIO).
1159
1160 One solution for this is to always use absolute paths. This usually
1161 works, but can be quite slow (the kernel has to walk the whole path on
1162 every access), and can also be a hassle to implement.
1163
1164 Newer POSIX systems have a number of functions (openat, fdopendir,
1165 futimensat and so on) that make it possible to specify working
1166 directories per operation.
1167
1168 For portability, and because the clowns who "designed", or shall I
1169 write, perpetrated this new interface were obviously half-drunk, this
1170 abstraction cannot be perfect, though.
1171
1172 IO::AIO allows you to convert directory paths into a so-called
1173 IO::AIO::WD object. This object stores the canonicalised, absolute
1174 version of the path, and on systems that allow it, also a directory file
1175 descriptor.
1176
1177 Everywhere where a pathname is accepted by IO::AIO (e.g. in "aio_stat"
1178 or "aio_unlink"), one can specify an array reference with an IO::AIO::WD
1179 object and a pathname instead (or the IO::AIO::WD object alone, which
1180 gets interpreted as "[$wd, "."]"). If the pathname is absolute, the
1181 IO::AIO::WD object is ignored, otherwise the pathname is resolved
1182 relative to that IO::AIO::WD object.
1183
1184 For example, to get a wd object for /etc and then stat passwd inside,
1185 you would write:
1186
1187 aio_wd "/etc", sub {
1188 my $etcdir = shift;
1189
1190 # although $etcdir can be undef on error, there is generally no reason
1191 # to check for errors here, as aio_stat will fail with ENOENT
1192 # when $etcdir is undef.
1193
1194 aio_stat [$etcdir, "passwd"], sub {
1195 # yay
1196 };
1197 };
1198
1199 The fact that "aio_wd" is a request and not a normal function shows that
1200 creating an IO::AIO::WD object is itself a potentially blocking
1201 operation, which is why it is done asynchronously.
1202
1203 To stat the directory obtained with "aio_wd" above, one could write
1204 either of the following three request calls:
1205
1206 aio_lstat "/etc" , sub { ... # pathname as normal string
1207 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1208 aio_lstat $wd , sub { ... # shorthand for the previous
1209
1210 As with normal pathnames, IO::AIO keeps a copy of the working directory
1211 object and the pathname string, so you could write the following without
1212 causing any issues due to $path getting reused:
1213
1214 my $path = [$wd, undef];
1215
1216 for my $name (qw(abc def ghi)) {
1217 $path->[1] = $name;
1218 aio_stat $path, sub {
1219 # ...
1220 };
1221 }
1222
1223 There are some caveats: when directories get renamed (or deleted), the
1224 pathname string doesn't change, so will point to the new directory (or
1225 nowhere at all), while the directory fd, if available on the system,
1226 will still point to the original directory. Most functions accepting a
1227 pathname will use the directory fd on newer systems, and the string on
1228 older systems. Some functions (such as "aio_realpath") will always rely
1229 on the string form of the pathname.
1230
1231 So this functionality is mainly useful to get some protection against
1232 "chdir", to easily get an absolute path out of a relative path for
1233 future reference, and to speed up doing many operations in the same
1234 directory (e.g. when stat'ing all files in a directory).
1235
1236 The following functions implement this working directory abstraction:
1237
1238 aio_wd $pathname, $callback->($wd)
1239 Asynchonously canonicalise the given pathname and convert it to an
1240 IO::AIO::WD object representing it. If possible and supported on the
1241 system, also open a directory fd to speed up pathname resolution
1242 relative to this working directory.
1243
1244 If something goes wrong, then "undef" is passwd to the callback
1245 instead of a working directory object and $! is set appropriately.
1246 Since passing "undef" as working directory component of a pathname
1247 fails the request with "ENOENT", there is often no need for error
1248 checking in the "aio_wd" callback, as future requests using the
1249 value will fail in the expected way.
1250
1251 IO::AIO::CWD
1252 This is a compile time constant (object) that represents the process
1253 current working directory.
1254
1255 Specifying this object as working directory object for a pathname is
1256 as if the pathname would be specified directly, without a directory
1257 object. For example, these calls are functionally identical:
1258
1259 aio_stat "somefile", sub { ... };
1260 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1261
1262 To recover the path associated with an IO::AIO::WD object, you can use
1263 "aio_realpath":
1264
1265 aio_realpath $wd, sub {
1266 warn "path is $_[0]\n";
1267 };
1268
1269 Currently, "aio_statvfs" always, and "aio_rename" and "aio_rmdir"
1270 sometimes, fall back to using an absolue path.
1271
1272 IO::AIO::REQ CLASS
1273 All non-aggregate "aio_*" functions return an object of this class when
1274 called in non-void context.
1275
1276 cancel $req
1277 Cancels the request, if possible. Has the effect of skipping
1278 execution when entering the execute state and skipping calling the
1279 callback when entering the the result state, but will leave the
1280 request otherwise untouched (with the exception of readdir). That
1281 means that requests that currently execute will not be stopped and
1282 resources held by the request will not be freed prematurely.
1283
1284 cb $req $callback->(...)
1285 Replace (or simply set) the callback registered to the request.
1286
1287 IO::AIO::GRP CLASS
1288 This class is a subclass of IO::AIO::REQ, so all its methods apply to
1289 objects of this class, too.
1290
1291 A IO::AIO::GRP object is a special request that can contain multiple
1292 other aio requests.
1293
1294 You create one by calling the "aio_group" constructing function with a
1295 callback that will be called when all contained requests have entered
1296 the "done" state:
1297
1298 my $grp = aio_group sub {
1299 print "all requests are done\n";
1300 };
1301
1302 You add requests by calling the "add" method with one or more
1303 "IO::AIO::REQ" objects:
1304
1305 $grp->add (aio_unlink "...");
1306
1307 add $grp aio_stat "...", sub {
1308 $_[0] or return $grp->result ("error");
1309
1310 # add another request dynamically, if first succeeded
1311 add $grp aio_open "...", sub {
1312 $grp->result ("ok");
1313 };
1314 };
1315
493 This makes it very easy to create composite requests (see the source 1316 This makes it very easy to create composite requests (see the source of
494 of "aio_move" for an application) that work and feel like simple 1317 "aio_move" for an application) that work and feel like simple requests.
1318
1319 * The IO::AIO::GRP objects will be cleaned up during calls to
1320 "IO::AIO::poll_cb", just like any other request.
1321
1322 * They can be canceled like any other request. Canceling will cancel
1323 not only the request itself, but also all requests it contains.
1324
1325 * They can also can also be added to other IO::AIO::GRP objects.
1326
1327 * You must not add requests to a group from within the group callback
1328 (or any later time).
1329
1330 Their lifetime, simplified, looks like this: when they are empty, they
1331 will finish very quickly. If they contain only requests that are in the
1332 "done" state, they will also finish. Otherwise they will continue to
1333 exist.
1334
1335 That means after creating a group you have some time to add requests
1336 (precisely before the callback has been invoked, which is only done
1337 within the "poll_cb"). And in the callbacks of those requests, you can
1338 add further requests to the group. And only when all those requests have
1339 finished will the the group itself finish.
1340
1341 add $grp ...
1342 $grp->add (...)
1343 Add one or more requests to the group. Any type of IO::AIO::REQ can
1344 be added, including other groups, as long as you do not create
1345 circular dependencies.
1346
1347 Returns all its arguments.
1348
1349 $grp->cancel_subs
1350 Cancel all subrequests and clears any feeder, but not the group
1351 request itself. Useful when you queued a lot of events but got a
1352 result early.
1353
1354 The group request will finish normally (you cannot add requests to
1355 the group).
1356
1357 $grp->result (...)
1358 Set the result value(s) that will be passed to the group callback
1359 when all subrequests have finished and set the groups errno to the
1360 current value of errno (just like calling "errno" without an error
1361 number). By default, no argument will be passed and errno is zero.
1362
1363 $grp->errno ([$errno])
1364 Sets the group errno value to $errno, or the current value of errno
1365 when the argument is missing.
1366
1367 Every aio request has an associated errno value that is restored
1368 when the callback is invoked. This method lets you change this value
1369 from its default (0).
1370
1371 Calling "result" will also set errno, so make sure you either set $!
1372 before the call to "result", or call c<errno> after it.
1373
1374 feed $grp $callback->($grp)
1375 Sets a feeder/generator on this group: every group can have an
1376 attached generator that generates requests if idle. The idea behind
1377 this is that, although you could just queue as many requests as you
1378 want in a group, this might starve other requests for a potentially
1379 long time. For example, "aio_scandir" might generate hundreds of
1380 thousands of "aio_stat" requests, delaying any later requests for a
1381 long time.
1382
1383 To avoid this, and allow incremental generation of requests, you can
1384 instead a group and set a feeder on it that generates those
1385 requests. The feed callback will be called whenever there are few
1386 enough (see "limit", below) requests active in the group itself and
1387 is expected to queue more requests.
1388
1389 The feed callback can queue as many requests as it likes (i.e. "add"
1390 does not impose any limits).
1391
1392 If the feed does not queue more requests when called, it will be
1393 automatically removed from the group.
1394
1395 If the feed limit is 0 when this method is called, it will be set to
1396 2 automatically.
1397
1398 Example:
1399
1400 # stat all files in @files, but only ever use four aio requests concurrently:
1401
1402 my $grp = aio_group sub { print "finished\n" };
1403 limit $grp 4;
1404 feed $grp sub {
1405 my $file = pop @files
1406 or return;
1407
1408 add $grp aio_stat $file, sub { ... };
1409 };
1410
1411 limit $grp $num
1412 Sets the feeder limit for the group: The feeder will be called
1413 whenever the group contains less than this many requests.
1414
1415 Setting the limit to 0 will pause the feeding process.
1416
1417 The default value for the limit is 0, but note that setting a feeder
1418 automatically bumps it up to 2.
1419
1420 SUPPORT FUNCTIONS
1421 EVENT PROCESSING AND EVENT LOOP INTEGRATION
1422 $fileno = IO::AIO::poll_fileno
1423 Return the *request result pipe file descriptor*. This filehandle
1424 must be polled for reading by some mechanism outside this module
1425 (e.g. EV, Glib, select and so on, see below or the SYNOPSIS). If the
1426 pipe becomes readable you have to call "poll_cb" to check the
1427 results.
1428
1429 See "poll_cb" for an example.
1430
1431 IO::AIO::poll_cb
1432 Process some requests that have reached the result phase (i.e. they
1433 have been executed but the results are not yet reported). You have
1434 to call this "regularly" to finish outstanding requests.
1435
1436 Returns 0 if all events could be processed (or there were no events
1437 to process), or -1 if it returned earlier for whatever reason.
1438 Returns immediately when no events are outstanding. The amount of
1439 events processed depends on the settings of "IO::AIO::max_poll_req",
1440 "IO::AIO::max_poll_time" and "IO::AIO::max_outstanding".
1441
1442 If not all requests were processed for whatever reason, the poll
1443 file descriptor will still be ready when "poll_cb" returns, so
1444 normally you don't have to do anything special to have it called
1445 later.
1446
1447 Apart from calling "IO::AIO::poll_cb" when the event filehandle
1448 becomes ready, it can be beneficial to call this function from loops
1449 which submit a lot of requests, to make sure the results get
1450 processed when they become available and not just when the loop is
1451 finished and the event loop takes over again. This function returns
1452 very fast when there are no outstanding requests.
1453
1454 Example: Install an Event watcher that automatically calls
1455 IO::AIO::poll_cb with high priority (more examples can be found in
1456 the SYNOPSIS section, at the top of this document):
1457
1458 Event->io (fd => IO::AIO::poll_fileno,
1459 poll => 'r', async => 1,
1460 cb => \&IO::AIO::poll_cb);
1461
1462 IO::AIO::poll_wait
1463 Wait until either at least one request is in the result phase or no
1464 requests are outstanding anymore.
1465
1466 This is useful if you want to synchronously wait for some requests
1467 to become ready, without actually handling them.
1468
1469 See "nreqs" for an example.
1470
1471 IO::AIO::poll
1472 Waits until some requests have been handled.
1473
1474 Returns the number of requests processed, but is otherwise strictly
1475 equivalent to:
1476
1477 IO::AIO::poll_wait, IO::AIO::poll_cb
1478
1479 IO::AIO::flush
1480 Wait till all outstanding AIO requests have been handled.
1481
1482 Strictly equivalent to:
1483
1484 IO::AIO::poll_wait, IO::AIO::poll_cb
1485 while IO::AIO::nreqs;
1486
1487 This function can be useful at program aborts, to make sure
1488 outstanding I/O has been done ("IO::AIO" uses an "END" block which
1489 already calls this function on normal exits), or when you are merely
1490 using "IO::AIO" for its more advanced functions, rather than for
1491 async I/O, e.g.:
1492
1493 my ($dirs, $nondirs);
1494 IO::AIO::aio_scandir "/tmp", 0, sub { ($dirs, $nondirs) = @_ };
1495 IO::AIO::flush;
1496 # $dirs, $nondirs are now set
1497
1498 IO::AIO::max_poll_reqs $nreqs
1499 IO::AIO::max_poll_time $seconds
1500 These set the maximum number of requests (default 0, meaning
1501 infinity) that are being processed by "IO::AIO::poll_cb" in one
1502 call, respectively the maximum amount of time (default 0, meaning
1503 infinity) spent in "IO::AIO::poll_cb" to process requests (more
1504 correctly the mininum amount of time "poll_cb" is allowed to use).
1505
1506 Setting "max_poll_time" to a non-zero value creates an overhead of
1507 one syscall per request processed, which is not normally a problem
1508 unless your callbacks are really really fast or your OS is really
1509 really slow (I am not mentioning Solaris here). Using
1510 "max_poll_reqs" incurs no overhead.
1511
1512 Setting these is useful if you want to ensure some level of
1513 interactiveness when perl is not fast enough to process all requests
1514 in time.
1515
1516 For interactive programs, values such as 0.01 to 0.1 should be fine.
1517
1518 Example: Install an Event watcher that automatically calls
1519 IO::AIO::poll_cb with low priority, to ensure that other parts of
1520 the program get the CPU sometimes even under high AIO load.
1521
1522 # try not to spend much more than 0.1s in poll_cb
1523 IO::AIO::max_poll_time 0.1;
1524
1525 # use a low priority so other tasks have priority
1526 Event->io (fd => IO::AIO::poll_fileno,
1527 poll => 'r', nice => 1,
1528 cb => &IO::AIO::poll_cb);
1529
1530 CONTROLLING THE NUMBER OF THREADS
1531 IO::AIO::min_parallel $nthreads
1532 Set the minimum number of AIO threads to $nthreads. The current
1533 default is 8, which means eight asynchronous operations can execute
1534 concurrently at any one time (the number of outstanding requests,
1535 however, is unlimited).
1536
1537 IO::AIO starts threads only on demand, when an AIO request is queued
1538 and no free thread exists. Please note that queueing up a hundred
1539 requests can create demand for a hundred threads, even if it turns
1540 out that everything is in the cache and could have been processed
1541 faster by a single thread.
1542
1543 It is recommended to keep the number of threads relatively low, as
1544 some Linux kernel versions will scale negatively with the number of
1545 threads (higher parallelity => MUCH higher latency). With current
1546 Linux 2.6 versions, 4-32 threads should be fine.
1547
1548 Under most circumstances you don't need to call this function, as
1549 the module selects a default that is suitable for low to moderate
1550 load.
1551
1552 IO::AIO::max_parallel $nthreads
1553 Sets the maximum number of AIO threads to $nthreads. If more than
1554 the specified number of threads are currently running, this function
1555 kills them. This function blocks until the limit is reached.
1556
1557 While $nthreads are zero, aio requests get queued but not executed
1558 until the number of threads has been increased again.
1559
1560 This module automatically runs "max_parallel 0" at program end, to
1561 ensure that all threads are killed and that there are no outstanding
495 requests. 1562 requests.
496 1563
497 * The IO::AIO::GRP objects will be cleaned up during calls to
498 "IO::AIO::poll_cb", just like any other request.
499 * They can be canceled like any other request. Canceling will cancel
500 not only the request itself, but also all requests it contains.
501 * They can also can also be added to other IO::AIO::GRP objects.
502 * You must not add requests to a group from within the group
503 callback (or any later time).
504
505 Their lifetime, simplified, looks like this: when they are empty,
506 they will finish very quickly. If they contain only requests that
507 are in the "done" state, they will also finish. Otherwise they will
508 continue to exist.
509
510 That means after creating a group you have some time to add
511 requests. And in the callbacks of those requests, you can add
512 further requests to the group. And only when all those requests have
513 finished will the the group itself finish.
514
515 add $grp ...
516 $grp->add (...)
517 Add one or more requests to the group. Any type of IO::AIO::REQ
518 can be added, including other groups, as long as you do not
519 create circular dependencies.
520
521 Returns all its arguments.
522
523 $grp->cancel_subs
524 Cancel all subrequests and clears any feeder, but not the group
525 request itself. Useful when you queued a lot of events but got a
526 result early.
527
528 $grp->result (...)
529 Set the result value(s) that will be passed to the group
530 callback when all subrequests have finished and set thre groups
531 errno to the current value of errno (just like calling "errno"
532 without an error number). By default, no argument will be passed
533 and errno is zero.
534
535 $grp->errno ([$errno])
536 Sets the group errno value to $errno, or the current value of
537 errno when the argument is missing.
538
539 Every aio request has an associated errno value that is restored
540 when the callback is invoked. This method lets you change this
541 value from its default (0).
542
543 Calling "result" will also set errno, so make sure you either
544 set $! before the call to "result", or call c<errno> after it.
545
546 feed $grp $callback->($grp)
547 Sets a feeder/generator on this group: every group can have an
548 attached generator that generates requests if idle. The idea
549 behind this is that, although you could just queue as many
550 requests as you want in a group, this might starve other
551 requests for a potentially long time. For example, "aio_scandir"
552 might generate hundreds of thousands "aio_stat" requests,
553 delaying any later requests for a long time.
554
555 To avoid this, and allow incremental generation of requests, you
556 can instead a group and set a feeder on it that generates those
557 requests. The feed callback will be called whenever there are
558 few enough (see "limit", below) requests active in the group
559 itself and is expected to queue more requests.
560
561 The feed callback can queue as many requests as it likes (i.e.
562 "add" does not impose any limits).
563
564 If the feed does not queue more requests when called, it will be
565 automatically removed from the group.
566
567 If the feed limit is 0, it will be set to 2 automatically.
568
569 Example:
570
571 # stat all files in @files, but only ever use four aio requests concurrently:
572
573 my $grp = aio_group sub { print "finished\n" };
574 limit $grp 4;
575 feed $grp sub {
576 my $file = pop @files
577 or return;
578
579 add $grp aio_stat $file, sub { ... };
580 };
581
582 limit $grp $num
583 Sets the feeder limit for the group: The feeder will be called
584 whenever the group contains less than this many requests.
585
586 Setting the limit to 0 will pause the feeding process.
587
588 SUPPORT FUNCTIONS
589 $fileno = IO::AIO::poll_fileno
590 Return the *request result pipe file descriptor*. This
591 filehandle must be polled for reading by some mechanism outside
592 this module (e.g. Event or select, see below or the SYNOPSIS).
593 If the pipe becomes readable you have to call "poll_cb" to check
594 the results.
595
596 See "poll_cb" for an example.
597
598 IO::AIO::poll_cb
599 Process all outstanding events on the result pipe. You have to
600 call this regularly. Returns the number of events processed.
601 Returns immediately when no events are outstanding.
602
603 If not all requests were processed for whatever reason, the
604 filehandle will still be ready when "poll_cb" returns.
605
606 Example: Install an Event watcher that automatically calls
607 IO::AIO::poll_cb with high priority:
608
609 Event->io (fd => IO::AIO::poll_fileno,
610 poll => 'r', async => 1,
611 cb => \&IO::AIO::poll_cb);
612
613 IO::AIO::poll_some $max_requests
614 Similar to "poll_cb", but only processes up to $max_requests
615 requests at a time.
616
617 Useful if you want to ensure some level of interactiveness when
618 perl is not fast enough to process all requests in time.
619
620 Example: Install an Event watcher that automatically calls
621 IO::AIO::poll_some with low priority, to ensure that other parts
622 of the program get the CPU sometimes even under high AIO load.
623
624 Event->io (fd => IO::AIO::poll_fileno,
625 poll => 'r', nice => 1,
626 cb => sub { IO::AIO::poll_some 256 });
627
628 IO::AIO::poll_wait
629 Wait till the result filehandle becomes ready for reading
630 (simply does a "select" on the filehandle. This is useful if you
631 want to synchronously wait for some requests to finish).
632
633 See "nreqs" for an example.
634
635 IO::AIO::nreqs
636 Returns the number of requests currently in the ready, execute
637 or pending states (i.e. for which their callback has not been
638 invoked yet).
639
640 Example: wait till there are no outstanding requests anymore:
641
642 IO::AIO::poll_wait, IO::AIO::poll_cb
643 while IO::AIO::nreqs;
644
645 IO::AIO::nready
646 Returns the number of requests currently in the ready state (not
647 yet executed).
648
649 IO::AIO::npending
650 Returns the number of requests currently in the pending state
651 (executed, but not yet processed by poll_cb).
652
653 IO::AIO::flush
654 Wait till all outstanding AIO requests have been handled.
655
656 Strictly equivalent to:
657
658 IO::AIO::poll_wait, IO::AIO::poll_cb
659 while IO::AIO::nreqs;
660
661 IO::AIO::poll
662 Waits until some requests have been handled.
663
664 Strictly equivalent to:
665
666 IO::AIO::poll_wait, IO::AIO::poll_cb
667 if IO::AIO::nreqs;
668
669 IO::AIO::min_parallel $nthreads
670 Set the minimum number of AIO threads to $nthreads. The current
671 default is 8, which means eight asynchronous operations can
672 execute concurrently at any one time (the number of outstanding
673 requests, however, is unlimited).
674
675 IO::AIO starts threads only on demand, when an AIO request is
676 queued and no free thread exists.
677
678 It is recommended to keep the number of threads relatively low,
679 as some Linux kernel versions will scale negatively with the
680 number of threads (higher parallelity => MUCH higher latency).
681 With current Linux 2.6 versions, 4-32 threads should be fine.
682
683 Under most circumstances you don't need to call this function,
684 as the module selects a default that is suitable for low to
685 moderate load.
686
687 IO::AIO::max_parallel $nthreads
688 Sets the maximum number of AIO threads to $nthreads. If more
689 than the specified number of threads are currently running, this
690 function kills them. This function blocks until the limit is
691 reached.
692
693 While $nthreads are zero, aio requests get queued but not
694 executed until the number of threads has been increased again.
695
696 This module automatically runs "max_parallel 0" at program end,
697 to ensure that all threads are killed and that there are no
698 outstanding requests.
699
700 Under normal circumstances you don't need to call this function. 1564 Under normal circumstances you don't need to call this function.
701 1565
1566 IO::AIO::max_idle $nthreads
1567 Limit the number of threads (default: 4) that are allowed to idle
1568 (i.e., threads that did not get a request to process within the idle
1569 timeout (default: 10 seconds). That means if a thread becomes idle
1570 while $nthreads other threads are also idle, it will free its
1571 resources and exit.
1572
1573 This is useful when you allow a large number of threads (e.g. 100 or
1574 1000) to allow for extremely high load situations, but want to free
1575 resources under normal circumstances (1000 threads can easily
1576 consume 30MB of RAM).
1577
1578 The default is probably ok in most situations, especially if thread
1579 creation is fast. If thread creation is very slow on your system you
1580 might want to use larger values.
1581
1582 IO::AIO::idle_timeout $seconds
1583 Sets the minimum idle timeout (default 10) after which worker
1584 threads are allowed to exit. SEe "IO::AIO::max_idle".
1585
702 $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 1586 IO::AIO::max_outstanding $maxreqs
703 This is a very bad function to use in interactive programs
704 because it blocks, and a bad way to reduce concurrency because
705 it is inexact: Better use an "aio_group" together with a feed
706 callback.
707
708 Sets the maximum number of outstanding requests to $nreqs. If 1587 Sets the maximum number of outstanding requests to $nreqs. If you do
709 you to queue up more than this number of requests, the next call 1588 queue up more than this number of requests, the next call to
710 to the "poll_cb" (and "poll_some" and other functions calling 1589 "IO::AIO::poll_cb" (and other functions calling "poll_cb", such as
711 "poll_cb") function will block until the limit is no longer 1590 "IO::AIO::flush" or "IO::AIO::poll") will block until the limit is
712 exceeded. 1591 no longer exceeded.
713 1592
714 The default value is very large, so there is no practical limit 1593 In other words, this setting does not enforce a queue limit, but can
1594 be used to make poll functions block if the limit is exceeded.
1595
1596 This is a very bad function to use in interactive programs because
1597 it blocks, and a bad way to reduce concurrency because it is
1598 inexact: Better use an "aio_group" together with a feed callback.
1599
1600 Its main use is in scripts without an event loop - when you want to
1601 stat a lot of files, you can write something like this:
1602
1603 IO::AIO::max_outstanding 32;
1604
1605 for my $path (...) {
1606 aio_stat $path , ...;
1607 IO::AIO::poll_cb;
1608 }
1609
1610 IO::AIO::flush;
1611
1612 The call to "poll_cb" inside the loop will normally return
1613 instantly, but as soon as more thna 32 reqeusts are in-flight, it
1614 will block until some requests have been handled. This keeps the
1615 loop from pushing a large number of "aio_stat" requests onto the
1616 queue.
1617
1618 The default value for "max_outstanding" is very large, so there is
715 on the number of outstanding requests. 1619 no practical limit on the number of outstanding requests.
716 1620
717 You can still queue as many requests as you want. Therefore, 1621 STATISTICAL INFORMATION
718 "max_oustsanding" is mainly useful in simple scripts (with low 1622 IO::AIO::nreqs
719 values) or as a stop gap to shield against fatal memory overflow 1623 Returns the number of requests currently in the ready, execute or
720 (with large values). 1624 pending states (i.e. for which their callback has not been invoked
1625 yet).
1626
1627 Example: wait till there are no outstanding requests anymore:
1628
1629 IO::AIO::poll_wait, IO::AIO::poll_cb
1630 while IO::AIO::nreqs;
1631
1632 IO::AIO::nready
1633 Returns the number of requests currently in the ready state (not yet
1634 executed).
1635
1636 IO::AIO::npending
1637 Returns the number of requests currently in the pending state
1638 (executed, but not yet processed by poll_cb).
1639
1640 SUBSECOND STAT TIME ACCESS
1641 Both "aio_stat"/"aio_lstat" and perl's "stat"/"lstat" functions can
1642 generally find access/modification and change times with subsecond time
1643 accuracy of the system supports it, but perl's built-in functions only
1644 return the integer part.
1645
1646 The following functions return the timestamps of the most recent stat
1647 with subsecond precision on most systems and work both after
1648 "aio_stat"/"aio_lstat" and perl's "stat"/"lstat" calls. Their return
1649 value is only meaningful after a successful "stat"/"lstat" call, or
1650 during/after a successful "aio_stat"/"aio_lstat" callback.
1651
1652 This is similar to the Time::HiRes "stat" functions, but can return full
1653 resolution without rounding and work with standard perl "stat",
1654 alleviating the need to call the special "Time::HiRes" functions, which
1655 do not act like their perl counterparts.
1656
1657 On operating systems or file systems where subsecond time resolution is
1658 not supported or could not be detected, a fractional part of 0 is
1659 returned, so it is always safe to call these functions.
1660
1661 $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime,
1662 IO::AIO::st_btime
1663 Return the access, modication, change or birth time, respectively,
1664 including fractional part. Due to the limited precision of floating
1665 point, the accuracy on most platforms is only a bit better than
1666 milliseconds for times around now - see the *nsec* function family,
1667 below, for full accuracy.
1668
1669 File birth time is only available when the OS and perl support it
1670 (on FreeBSD and NetBSD at the time of this writing, although support
1671 is adaptive, so if your OS/perl gains support, IO::AIO can take
1672 advantage of it). On systems where it isn't available, 0 is
1673 currently returned, but this might change to "undef" in a future
1674 version.
1675
1676 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
1677 Returns access, modification, change and birth time all in one go,
1678 and maybe more times in the future version.
1679
1680 $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec,
1681 IO::AIO::st_ctimensec, IO::AIO::st_btimensec
1682 Return the fractional access, modifcation, change or birth time, in
1683 nanoseconds, as an integer in the range 0 to 999999999.
1684
1685 Note that no accessors are provided for access, modification and
1686 change times - you need to get those from "stat _" if required ("int
1687 IO::AIO::st_atime" and so on will *not* generally give you the
1688 correct value).
1689
1690 $seconds = IO::AIO::st_btimesec
1691 The (integral) seconds part of the file birth time, if available.
1692
1693 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
1694 Like the functions above, but returns all four times in one go (and
1695 maybe more in future versions).
1696
1697 $counter = IO::AIO::st_gen
1698 Returns the generation counter (in practice this is just a random
1699 number) of the file. This is only available on platforms which have
1700 this member in their "struct stat" (most BSDs at the time of this
1701 writing) and generally only to the root usert. If unsupported, 0 is
1702 returned, but this might change to "undef" in a future version.
1703
1704 Example: print the high resolution modification time of /etc, using
1705 "stat", and "IO::AIO::aio_stat".
1706
1707 if (stat "/etc") {
1708 printf "stat(/etc) mtime: %f\n", IO::AIO::st_mtime;
1709 }
1710
1711 IO::AIO::aio_stat "/etc", sub {
1712 $_[0]
1713 and return;
1714
1715 printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::st_mtimensec;
1716 };
1717
1718 IO::AIO::flush;
1719
1720 Output of the awbove on my system, showing reduced and full accuracy:
1721
1722 stat(/etc) mtime: 1534043702.020808
1723 aio_stat(/etc) mtime: 1534043702.020807792
1724
1725 MISCELLANEOUS FUNCTIONS
1726 IO::AIO implements some functions that are useful when you want to use
1727 some "Advanced I/O" function not available to in Perl, without going the
1728 "Asynchronous I/O" route. Many of these have an asynchronous "aio_*"
1729 counterpart.
1730
1731 $numfd = IO::AIO::get_fdlimit
1732 Tries to find the current file descriptor limit and returns it, or
1733 "undef" and sets $! in case of an error. The limit is one larger
1734 than the highest valid file descriptor number.
1735
1736 IO::AIO::min_fdlimit [$numfd]
1737 Try to increase the current file descriptor limit(s) to at least
1738 $numfd by changing the soft or hard file descriptor resource limit.
1739 If $numfd is missing, it will try to set a very high limit, although
1740 this is not recommended when you know the actual minimum that you
1741 require.
1742
1743 If the limit cannot be raised enough, the function makes a
1744 best-effort attempt to increase the limit as much as possible, using
1745 various tricks, while still failing. You can query the resulting
1746 limit using "IO::AIO::get_fdlimit".
1747
1748 If an error occurs, returns "undef" and sets $!, otherwise returns
1749 true.
1750
1751 IO::AIO::sendfile $ofh, $ifh, $offset, $count
1752 Calls the "eio_sendfile_sync" function, which is like
1753 "aio_sendfile", but is blocking (this makes most sense if you know
1754 the input data is likely cached already and the output filehandle is
1755 set to non-blocking operations).
1756
1757 Returns the number of bytes copied, or -1 on error.
1758
1759 IO::AIO::fadvise $fh, $offset, $len, $advice
1760 Simply calls the "posix_fadvise" function (see its manpage for
1761 details). The following advice constants are available:
1762 "IO::AIO::FADV_NORMAL", "IO::AIO::FADV_SEQUENTIAL",
1763 "IO::AIO::FADV_RANDOM", "IO::AIO::FADV_NOREUSE",
1764 "IO::AIO::FADV_WILLNEED", "IO::AIO::FADV_DONTNEED".
1765
1766 On systems that do not implement "posix_fadvise", this function
1767 returns ENOSYS, otherwise the return value of "posix_fadvise".
1768
1769 IO::AIO::madvise $scalar, $offset, $len, $advice
1770 Simply calls the "posix_madvise" function (see its manpage for
1771 details). The following advice constants are available:
1772 "IO::AIO::MADV_NORMAL", "IO::AIO::MADV_SEQUENTIAL",
1773 "IO::AIO::MADV_RANDOM", "IO::AIO::MADV_WILLNEED",
1774 "IO::AIO::MADV_DONTNEED".
1775
1776 If $offset is negative, counts from the end. If $length is negative,
1777 the remaining length of the $scalar is used. If possible, $length
1778 will be reduced to fit into the $scalar.
1779
1780 On systems that do not implement "posix_madvise", this function
1781 returns ENOSYS, otherwise the return value of "posix_madvise".
1782
1783 IO::AIO::mprotect $scalar, $offset, $len, $protect
1784 Simply calls the "mprotect" function on the preferably AIO::mmap'ed
1785 $scalar (see its manpage for details). The following protect
1786 constants are available: "IO::AIO::PROT_NONE", "IO::AIO::PROT_READ",
1787 "IO::AIO::PROT_WRITE", "IO::AIO::PROT_EXEC".
1788
1789 If $offset is negative, counts from the end. If $length is negative,
1790 the remaining length of the $scalar is used. If possible, $length
1791 will be reduced to fit into the $scalar.
1792
1793 On systems that do not implement "mprotect", this function returns
1794 ENOSYS, otherwise the return value of "mprotect".
1795
1796 IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1797 Memory-maps a file (or anonymous memory range) and attaches it to
1798 the given $scalar, which will act like a string scalar. Returns true
1799 on success, and false otherwise.
1800
1801 The scalar must exist, but its contents do not matter - this means
1802 you cannot use a nonexistant array or hash element. When in doubt,
1803 "undef" the scalar first.
1804
1805 The only operations allowed on the mmapped scalar are
1806 "substr"/"vec", which don't change the string length, and most
1807 read-only operations such as copying it or searching it with regexes
1808 and so on.
1809
1810 Anything else is unsafe and will, at best, result in memory leaks.
1811
1812 The memory map associated with the $scalar is automatically removed
1813 when the $scalar is undef'd or destroyed, or when the
1814 "IO::AIO::mmap" or "IO::AIO::munmap" functions are called on it.
1815
1816 This calls the "mmap"(2) function internally. See your system's
1817 manual page for details on the $length, $prot and $flags parameters.
1818
1819 The $length must be larger than zero and smaller than the actual
1820 filesize.
1821
1822 $prot is a combination of "IO::AIO::PROT_NONE",
1823 "IO::AIO::PROT_EXEC", "IO::AIO::PROT_READ" and/or
1824 "IO::AIO::PROT_WRITE",
1825
1826 $flags can be a combination of "IO::AIO::MAP_SHARED" or
1827 "IO::AIO::MAP_PRIVATE", or a number of system-specific flags (when
1828 not available, the are 0): "IO::AIO::MAP_ANONYMOUS" (which is set to
1829 "MAP_ANON" if your system only provides this constant),
1830 "IO::AIO::MAP_LOCKED", "IO::AIO::MAP_NORESERVE",
1831 "IO::AIO::MAP_POPULATE", "IO::AIO::MAP_NONBLOCK",
1832 "IO::AIO::MAP_FIXED", "IO::AIO::MAP_GROWSDOWN",
1833 "IO::AIO::MAP_32BIT", "IO::AIO::MAP_HUGETLB" or
1834 "IO::AIO::MAP_STACK".
1835
1836 If $fh is "undef", then a file descriptor of -1 is passed.
1837
1838 $offset is the offset from the start of the file - it generally must
1839 be a multiple of "IO::AIO::PAGESIZE" and defaults to 0.
1840
1841 Example:
1842
1843 use Digest::MD5;
1844 use IO::AIO;
1845
1846 open my $fh, "<verybigfile"
1847 or die "$!";
1848
1849 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1850 or die "verybigfile: $!";
1851
1852 my $fast_md5 = md5 $data;
1853
1854 IO::AIO::munmap $scalar
1855 Removes a previous mmap and undefines the $scalar.
1856
1857 IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[,
1858 $new_address = 0]
1859 Calls the Linux-specific mremap(2) system call. The $scalar must
1860 have been mapped by "IO::AIO::mmap", and $flags must currently
1861 either be 0 or "IO::AIO::MREMAP_MAYMOVE".
1862
1863 Returns true if successful, and false otherwise. If the underlying
1864 mmapped region has changed address, then the true value has the
1865 numerical value 1, otherwise it has the numerical value 0:
1866
1867 my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE
1868 or die "mremap: $!";
1869
1870 if ($success*1) {
1871 warn "scalar has chanegd address in memory\n";
1872 }
1873
1874 "IO::AIO::MREMAP_FIXED" and the $new_address argument are currently
1875 implemented, but not supported and might go away in a future
1876 version.
1877
1878 On systems where this call is not supported or is not emulated, this
1879 call returns falls and sets $! to "ENOSYS".
1880
1881 IO::AIO::mlockall $flags
1882 Calls the "eio_mlockall_sync" function, which is like
1883 "aio_mlockall", but is blocking.
1884
1885 IO::AIO::munlock $scalar, $offset = 0, $length = undef
1886 Calls the "munlock" function, undoing the effects of a previous
1887 "aio_mlock" call (see its description for details).
1888
1889 IO::AIO::munlockall
1890 Calls the "munlockall" function.
1891
1892 On systems that do not implement "munlockall", this function returns
1893 ENOSYS, otherwise the return value of "munlockall".
1894
1895 $fh = IO::AIO::accept4 $r_fh, $sockaddr, $sockaddr_maxlen, $flags
1896 Uses the GNU/Linux accept4(2) syscall, if available, to accept a
1897 socket and return the new file handle on success, or sets $! and
1898 returns "undef" on error.
1899
1900 The remote name of the new socket will be stored in $sockaddr, which
1901 will be extended to allow for at least $sockaddr_maxlen octets. If
1902 the socket name does not fit into $sockaddr_maxlen octets, this is
1903 signaled by returning a longer string in $sockaddr, which might or
1904 might not be truncated.
1905
1906 To accept name-less sockets, use "undef" for $sockaddr and 0 for
1907 $sockaddr_maxlen.
1908
1909 The main reasons to use this syscall rather than portable accept(2)
1910 are that you can specify "SOCK_NONBLOCK" and/or "SOCK_CLOEXEC" flags
1911 and you can accept name-less sockets by specifying 0 for
1912 $sockaddr_maxlen, which is sadly not possible with perl's interface
1913 to "accept".
1914
1915 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
1916 Calls the GNU/Linux splice(2) syscall, if available. If $r_off or
1917 $w_off are "undef", then "NULL" is passed for these, otherwise they
1918 should be the file offset.
1919
1920 $r_fh and $w_fh should not refer to the same file, as splice might
1921 silently corrupt the data in this case.
1922
1923 The following symbol flag values are available:
1924 "IO::AIO::SPLICE_F_MOVE", "IO::AIO::SPLICE_F_NONBLOCK",
1925 "IO::AIO::SPLICE_F_MORE" and "IO::AIO::SPLICE_F_GIFT".
1926
1927 See the splice(2) manpage for details.
1928
1929 IO::AIO::tee $r_fh, $w_fh, $length, $flags
1930 Calls the GNU/Linux tee(2) syscall, see its manpage and the
1931 description for "IO::AIO::splice" above for details.
1932
1933 $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
1934 Attempts to query or change the pipe buffer size. Obviously works
1935 only on pipes, and currently works only on GNU/Linux systems, and
1936 fails with -1/"ENOSYS" everywhere else. If anybody knows how to
1937 influence pipe buffer size on other systems, drop me a note.
1938
1939 ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
1940 This is a direct interface to the Linux pipe2(2) system call. If
1941 $flags is missing or 0, then this should be the same as a call to
1942 perl's built-in "pipe" function and create a new pipe, and works on
1943 systems that lack the pipe2 syscall. On win32, this case invokes
1944 "_pipe (..., 4096, O_BINARY)".
1945
1946 If $flags is non-zero, it tries to invoke the pipe2 system call with
1947 the given flags (Linux 2.6.27, glibc 2.9).
1948
1949 On success, the read and write file handles are returned.
1950
1951 On error, nothing will be returned. If the pipe2 syscall is missing
1952 and $flags is non-zero, fails with "ENOSYS".
1953
1954 Please refer to pipe2(2) for more info on the $flags, but at the
1955 time of this writing, "IO::AIO::O_CLOEXEC", "IO::AIO::O_NONBLOCK"
1956 and "IO::AIO::O_DIRECT" (Linux 3.4, for packet-based pipes) were
1957 supported.
1958
1959 Example: create a pipe race-free w.r.t. threads and fork:
1960
1961 my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
1962 or die "pipe2: $!\n";
1963
1964 $fh = IO::AIO::memfd_create $pathname[, $flags]
1965 This is a direct interface to the Linux memfd_create(2) system call.
1966 The (unhelpful) default for $flags is 0, but your default should be
1967 "IO::AIO::MFD_CLOEXEC".
1968
1969 On success, the new memfd filehandle is returned, otherwise returns
1970 "undef". If the memfd_create syscall is missing, fails with
1971 "ENOSYS".
1972
1973 Please refer to memfd_create(2) for more info on this call.
1974
1975 The following $flags values are available: "IO::AIO::MFD_CLOEXEC",
1976 "IO::AIO::MFD_ALLOW_SEALING" and "IO::AIO::MFD_HUGETLB".
1977
1978 Example: create a new memfd.
1979
1980 my $fh = IO::AIO::memfd_create "somenameforprocfd", IO::AIO::MFD_CLOEXEC
1981 or die "memfd_create: $!\n";
1982
1983 $fh = IO::AIO::pidfd_open $pid[, $flags]
1984 This is an interface to the Linux pidfd_open(2) system call. The
1985 default for $flags is 0.
1986
1987 On success, a new pidfd filehandle is returned (that is already set
1988 to close-on-exec), otherwise returns "undef". If the syscall is
1989 missing, fails with "ENOSYS".
1990
1991 Example: open pid 6341 as pidfd.
1992
1993 my $fh = IO::AIO::pidfd_open 6341
1994 or die "pidfd_open: $!\n";
1995
1996 $status = IO::AIO::pidfd_send_signal $pidfh, $signal[, $siginfo[,
1997 $flags]]
1998 This is an interface to the Linux pidfd_send_signal system call. The
1999 default for $siginfo is "undef" and the default for $flags is 0.
2000
2001 Returns the system call status. If the syscall is missing, fails
2002 with "ENOSYS".
2003
2004 When specified, $siginfo must be a reference to a hash with one or
2005 more of the following members:
2006
2007 code - the "si_code" member
2008 pid - the "si_pid" member
2009 uid - the "si_uid" member
2010 value_int - the "si_value.sival_int" member
2011 value_ptr - the "si_value.sival_ptr" member, specified as an integer
2012
2013 Example: send a SIGKILL to the specified process.
2014
2015 my $status = IO::AIO::pidfd_send_signal $pidfh, 9, undef
2016 and die "pidfd_send_signal: $!\n";
2017
2018 Example: send a SIGKILL to the specified process with extra data.
2019
2020 my $status = IO::AIO::pidfd_send_signal $pidfh, 9, { code => -1, value_int => 7 }
2021 and die "pidfd_send_signal: $!\n";
2022
2023 $fh = IO::AIO::pidfd_getfd $pidfh, $targetfd[, $flags]
2024 This is an interface to the Linux pidfd_getfd system call. The
2025 default for $flags is 0.
2026
2027 On success, returns a dup'ed copy of the target file descriptor
2028 (specified as an integer) returned (that is already set to
2029 close-on-exec), otherwise returns "undef". If the syscall is
2030 missing, fails with "ENOSYS".
2031
2032 Example: get a copy of standard error of another process and print
2033 soemthing to it.
2034
2035 my $errfh = IO::AIO::pidfd_getfd $pidfh, 2
2036 or die "pidfd_getfd: $!\n";
2037 print $errfh "stderr\n";
2038
2039 $fh = IO::AIO::eventfd [$initval, [$flags]]
2040 This is a direct interface to the Linux eventfd(2) system call. The
2041 (unhelpful) defaults for $initval and $flags are 0 for both.
2042
2043 On success, the new eventfd filehandle is returned, otherwise
2044 returns "undef". If the eventfd syscall is missing, fails with
2045 "ENOSYS".
2046
2047 Please refer to eventfd(2) for more info on this call.
2048
2049 The following symbol flag values are available:
2050 "IO::AIO::EFD_CLOEXEC", "IO::AIO::EFD_NONBLOCK" and
2051 "IO::AIO::EFD_SEMAPHORE" (Linux 2.6.30).
2052
2053 Example: create a new eventfd filehandle:
2054
2055 $fh = IO::AIO::eventfd 0, IO::AIO::EFD_CLOEXEC
2056 or die "eventfd: $!\n";
2057
2058 $fh = IO::AIO::timerfd_create $clockid[, $flags]
2059 This is a direct interface to the Linux timerfd_create(2) system
2060 call. The (unhelpful) default for $flags is 0, but your default
2061 should be "IO::AIO::TFD_CLOEXEC".
2062
2063 On success, the new timerfd filehandle is returned, otherwise
2064 returns "undef". If the timerfd_create syscall is missing, fails
2065 with "ENOSYS".
2066
2067 Please refer to timerfd_create(2) for more info on this call.
2068
2069 The following $clockid values are available:
2070 "IO::AIO::CLOCK_REALTIME", "IO::AIO::CLOCK_MONOTONIC"
2071 "IO::AIO::CLOCK_CLOCK_BOOTTIME" (Linux 3.15)
2072 "IO::AIO::CLOCK_CLOCK_REALTIME_ALARM" (Linux 3.11) and
2073 "IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM" (Linux 3.11).
2074
2075 The following $flags values are available (Linux 2.6.27):
2076 "IO::AIO::TFD_NONBLOCK" and "IO::AIO::TFD_CLOEXEC".
2077
2078 Example: create a new timerfd and set it to one-second repeated
2079 alarms, then wait for two alarms:
2080
2081 my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC
2082 or die "timerfd_create: $!\n";
2083
2084 defined IO::AIO::timerfd_settime $fh, 0, 1, 1
2085 or die "timerfd_settime: $!\n";
2086
2087 for (1..2) {
2088 8 == sysread $fh, my $buf, 8
2089 or die "timerfd read failure\n";
2090
2091 printf "number of expirations (likely 1): %d\n",
2092 unpack "Q", $buf;
2093 }
2094
2095 ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags,
2096 $new_interval, $nbw_value
2097 This is a direct interface to the Linux timerfd_settime(2) system
2098 call. Please refer to its manpage for more info on this call.
2099
2100 The new itimerspec is specified using two (possibly fractional)
2101 second values, $new_interval and $new_value).
2102
2103 On success, the current interval and value are returned (as per
2104 "timerfd_gettime"). On failure, the empty list is returned.
2105
2106 The following $flags values are available:
2107 "IO::AIO::TFD_TIMER_ABSTIME" and "IO::AIO::TFD_TIMER_CANCEL_ON_SET".
2108
2109 See "IO::AIO::timerfd_create" for a full example.
2110
2111 ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
2112 This is a direct interface to the Linux timerfd_gettime(2) system
2113 call. Please refer to its manpage for more info on this call.
2114
2115 On success, returns the current values of interval and value for the
2116 given timerfd (as potentially fractional second values). On failure,
2117 the empty list is returned.
2118
2119EVENT LOOP INTEGRATION
2120 It is recommended to use AnyEvent::AIO to integrate IO::AIO
2121 automatically into many event loops:
2122
2123 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
2124 use AnyEvent::AIO;
2125
2126 You can also integrate IO::AIO manually into many event loops, here are
2127 some examples of how to do this:
2128
2129 # EV integration
2130 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
2131
2132 # Event integration
2133 Event->io (fd => IO::AIO::poll_fileno,
2134 poll => 'r',
2135 cb => \&IO::AIO::poll_cb);
2136
2137 # Glib/Gtk2 integration
2138 add_watch Glib::IO IO::AIO::poll_fileno,
2139 in => sub { IO::AIO::poll_cb; 1 };
2140
2141 # Tk integration
2142 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
2143 readable => \&IO::AIO::poll_cb);
2144
2145 # Danga::Socket integration
2146 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
2147 \&IO::AIO::poll_cb);
721 2148
722 FORK BEHAVIOUR 2149 FORK BEHAVIOUR
723 This module should do "the right thing" when the process using it 2150 Usage of pthreads in a program changes the semantics of fork
724 forks: 2151 considerably. Specifically, only async-safe functions can be called
2152 after fork. Perl doesn't know about this, so in general, you cannot call
2153 fork with defined behaviour in perl if pthreads are involved. IO::AIO
2154 uses pthreads, so this applies, but many other extensions and (for
2155 inexplicable reasons) perl itself often is linked against pthreads, so
2156 this limitation applies to quite a lot of perls.
725 2157
726 Before the fork, IO::AIO enters a quiescent state where no requests 2158 This module no longer tries to fight your OS, or POSIX. That means
727 can be added in other threads and no results will be processed. 2159 IO::AIO only works in the process that loaded it. Forking is fully
728 After the fork the parent simply leaves the quiescent state and 2160 supported, but using IO::AIO in the child is not.
729 continues request/result processing, while the child frees the
730 request/result queue (so that the requests started before the fork
731 will only be handled in the parent). Threads will be started on
732 demand until the limit set in the parent process has been reached
733 again.
734 2161
735 In short: the parent will, after a short pause, continue as if fork 2162 You might get around by not *using* IO::AIO before (or after) forking.
736 had not been called, while the child will act as if IO::AIO has not 2163 You could also try to call the IO::AIO::reinit function in the child:
737 been used yet. 2164
2165 IO::AIO::reinit
2166 Abandons all current requests and I/O threads and simply
2167 reinitialises all data structures. This is not an operation
2168 supported by any standards, but happens to work on GNU/Linux and
2169 some newer BSD systems.
2170
2171 The only reasonable use for this function is to call it after
2172 forking, if "IO::AIO" was used in the parent. Calling it while
2173 IO::AIO is active in the process will result in undefined behaviour.
2174 Calling it at any time will also result in any undefined (by POSIX)
2175 behaviour.
2176
2177 LINUX-SPECIFIC CALLS
2178 When a call is documented as "linux-specific" then this means it
2179 originated on GNU/Linux. "IO::AIO" will usually try to autodetect the
2180 availability and compatibility of such calls regardless of the platform
2181 it is compiled on, so platforms such as FreeBSD which often implement
2182 these calls will work. When in doubt, call them and see if they fail wth
2183 "ENOSYS".
738 2184
739 MEMORY USAGE 2185 MEMORY USAGE
740 Per-request usage: 2186 Per-request usage:
741 2187
742 Each aio request uses - depending on your architecture - around 2188 Each aio request uses - depending on your architecture - around 100-200
743 100-200 bytes of memory. In addition, stat requests need a stat 2189 bytes of memory. In addition, stat requests need a stat buffer (possibly
744 buffer (possibly a few hundred bytes), readdir requires a result 2190 a few hundred bytes), readdir requires a result buffer and so on. Perl
745 buffer and so on. Perl scalars and other data passed into aio 2191 scalars and other data passed into aio requests will also be locked and
746 requests will also be locked and will consume memory till the 2192 will consume memory till the request has entered the done state.
747 request has entered the done state.
748 2193
749 This is now awfully much, so queuing lots of requests is not usually 2194 This is not awfully much, so queuing lots of requests is not usually a
750 a problem. 2195 problem.
751 2196
752 Per-thread usage: 2197 Per-thread usage:
753 2198
754 In the execution phase, some aio requests require more memory for 2199 In the execution phase, some aio requests require more memory for
755 temporary buffers, and each thread requires a stack and other data 2200 temporary buffers, and each thread requires a stack and other data
756 structures (usually around 16k-128k, depending on the OS). 2201 structures (usually around 16k-128k, depending on the OS).
757 2202
758KNOWN BUGS 2203KNOWN BUGS
759 Known bugs will be fixed in the next release. 2204 Known bugs will be fixed in the next release :)
2205
2206KNOWN ISSUES
2207 Calls that try to "import" foreign memory areas (such as "IO::AIO::mmap"
2208 or "IO::AIO::aio_slurp") do not work with generic lvalues, such as
2209 non-created hash slots or other scalars I didn't think of. It's best to
2210 avoid such and either use scalar variables or making sure that the
2211 scalar exists (e.g. by storing "undef") and isn't "funny" (e.g. tied).
2212
2213 I am not sure anything can be done about this, so this is considered a
2214 known issue, rather than a bug.
760 2215
761SEE ALSO 2216SEE ALSO
762 Coro::AIO. 2217 AnyEvent::AIO for easy integration into event loops, Coro::AIO for a
2218 more natural syntax and IO::FDPass for file descriptor passing.
763 2219
764AUTHOR 2220AUTHOR
765 Marc Lehmann <schmorp@schmorp.de> 2221 Marc Lehmann <schmorp@schmorp.de>
766 http://home.schmorp.de/ 2222 http://home.schmorp.de/
767 2223

Diff Legend

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