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

Comparing IO-AIO/README (file contents):
Revision 1.23 by root, Mon Jan 22 15:59:52 2007 UTC vs.
Revision 1.54 by root, Sun Jan 6 11:48:14 2013 UTC

… …
2 IO::AIO - Asynchronous Input/Output 2 IO::AIO - Asynchronous 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 = shift 8 my $fh = shift
9 or die "/etc/passwd: $!"; 9 or die "/etc/passwd: $!";
10 ... 10 ...
11 }; 11 };
12 12
… …
24 $req->cancel; # cancel request if still in queue 24 $req->cancel; # cancel request if still in queue
25 25
26 my $grp = aio_group sub { print "all stats done\n" }; 26 my $grp = aio_group sub { print "all stats done\n" };
27 add $grp aio_stat "..." for ...; 27 add $grp aio_stat "..." for ...;
28 28
29 # AnyEvent integration
30 open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
31 my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
32
33 # Event integration
34 Event->io (fd => IO::AIO::poll_fileno,
35 poll => 'r',
36 cb => \&IO::AIO::poll_cb);
37
38 # Glib/Gtk2 integration
39 add_watch Glib::IO IO::AIO::poll_fileno,
40 in => sub { IO::AIO::poll_cb; 1 };
41
42 # Tk integration
43 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
44 readable => \&IO::AIO::poll_cb);
45
46 # Danga::Socket integration
47 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
48 \&IO::AIO::poll_cb);
49
50DESCRIPTION 29DESCRIPTION
51 This module implements asynchronous I/O using whatever means your 30 This module implements asynchronous I/O using whatever means your
52 operating system supports. 31 operating system supports. It is implemented as an interface to "libeio"
32 (<http://software.schmorp.de/pkg/libeio.html>).
53 33
54 Asynchronous means that operations that can normally block your program 34 Asynchronous means that operations that can normally block your program
55 (e.g. reading from disk) will be done asynchronously: the operation will 35 (e.g. reading from disk) will be done asynchronously: the operation will
56 still block, but you can do something else in the meantime. This is 36 still block, but you can do something else in the meantime. This is
57 extremely useful for programs that need to stay interactive even when 37 extremely useful for programs that need to stay interactive even when
… …
61 faster on a RAID volume or over NFS when you do a number of stat 41 faster on a RAID volume or over NFS when you do a number of stat
62 operations concurrently. 42 operations concurrently.
63 43
64 While most of this works on all types of file descriptors (for example 44 While most of this works on all types of file descriptors (for example
65 sockets), using these functions on file descriptors that support 45 sockets), using these functions on file descriptors that support
66 nonblocking operation (again, sockets, pipes etc.) is very inefficient 46 nonblocking operation (again, sockets, pipes etc.) is very inefficient.
67 or might not work (aio_read fails on sockets/pipes/fifos). Use an event
68 loop for that (such as the Event module): IO::AIO will naturally fit 47 Use an event loop for that (such as the EV module): IO::AIO will
69 into such an event loop itself. 48 naturally fit into such an event loop itself.
70 49
71 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
72 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
73 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
74 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
… …
76 not well-supported or restricted (GNU/Linux doesn't allow them on normal 55 not well-supported or restricted (GNU/Linux doesn't allow them on normal
77 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
78 aio_write, so the remaining functionality would have to be implemented 57 aio_write, so the remaining functionality would have to be implemented
79 using threads anyway. 58 using threads anyway.
80 59
81 Although the module will work with in the presence of other (Perl-) 60 Although the module will work in the presence of other (Perl-) threads,
82 threads, it is currently not reentrant in any way, so use appropriate 61 it is currently not reentrant in any way, so use appropriate locking
83 locking yourself, always call "poll_cb" from within the same thread, or 62 yourself, always call "poll_cb" from within the same thread, or never
84 never call "poll_cb" (or other "aio_" functions) recursively. 63 call "poll_cb" (or other "aio_" functions) recursively.
85 64
86 EXAMPLE 65 EXAMPLE
87 This is a simple example that uses the Event module and loads 66 This is a simple example that uses the EV module and loads /etc/passwd
88 /etc/passwd asynchronously: 67 asynchronously:
89 68
90 use Fcntl; 69 use Fcntl;
91 use Event; 70 use EV;
92 use IO::AIO; 71 use IO::AIO;
93 72
94 # register the IO::AIO callback with Event 73 # register the IO::AIO callback with EV
95 Event->io (fd => IO::AIO::poll_fileno, 74 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
96 poll => 'r',
97 cb => \&IO::AIO::poll_cb);
98 75
99 # queue the request to open /etc/passwd 76 # queue the request to open /etc/passwd
100 aio_open "/etc/passwd", O_RDONLY, 0, sub { 77 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
101 my $fh = shift 78 my $fh = shift
102 or die "error while opening: $!"; 79 or die "error while opening: $!";
103 80
104 # stat'ing filehandles is generally non-blocking 81 # stat'ing filehandles is generally non-blocking
105 my $size = -s $fh; 82 my $size = -s $fh;
… …
114 91
115 # file contents now in $contents 92 # file contents now in $contents
116 print $contents; 93 print $contents;
117 94
118 # exit event loop and program 95 # exit event loop and program
119 Event::unloop; 96 EV::unloop;
120 }; 97 };
121 }; 98 };
122 99
123 # possibly queue up other requests, or open GUI windows, 100 # possibly queue up other requests, or open GUI windows,
124 # check for sockets etc. etc. 101 # check for sockets etc. etc.
125 102
126 # process events as long as there are some: 103 # process events as long as there are some:
127 Event::loop; 104 EV::loop;
128 105
129REQUEST ANATOMY AND LIFETIME 106REQUEST ANATOMY AND LIFETIME
130 Every "aio_*" function creates a request. which is a C data structure 107 Every "aio_*" function creates a request. which is a C data structure
131 not directly visible to Perl. 108 not directly visible to Perl.
132 109
… …
168 anymore (except possibly for the Perl object, but its connection to 145 anymore (except possibly for the Perl object, but its connection to
169 the actual aio request is severed and calling its methods will 146 the actual aio request is severed and calling its methods will
170 either do nothing or result in a runtime error). 147 either do nothing or result in a runtime error).
171 148
172FUNCTIONS 149FUNCTIONS
173 AIO REQUEST FUNCTIONS 150 QUICK OVERVIEW
151 This section simply lists the prototypes most of the functions for quick
152 reference. See the following sections for function-by-function
153 documentation.
154
155 aio_wd $pathname, $callback->($wd)
156 aio_open $pathname, $flags, $mode, $callback->($fh)
157 aio_close $fh, $callback->($status)
158 aio_seek $fh,$offset,$whence, $callback->($offs)
159 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
160 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
161 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
162 aio_readahead $fh,$offset,$length, $callback->($retval)
163 aio_stat $fh_or_path, $callback->($status)
164 aio_lstat $fh, $callback->($status)
165 aio_statvfs $fh_or_path, $callback->($statvfs)
166 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
167 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
168 aio_chmod $fh_or_path, $mode, $callback->($status)
169 aio_truncate $fh_or_path, $offset, $callback->($status)
170 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
171 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
172 aio_unlink $pathname, $callback->($status)
173 aio_mknod $pathname, $mode, $dev, $callback->($status)
174 aio_link $srcpath, $dstpath, $callback->($status)
175 aio_symlink $srcpath, $dstpath, $callback->($status)
176 aio_readlink $pathname, $callback->($link)
177 aio_realpath $pathname, $callback->($link)
178 aio_rename $srcpath, $dstpath, $callback->($status)
179 aio_mkdir $pathname, $mode, $callback->($status)
180 aio_rmdir $pathname, $callback->($status)
181 aio_readdir $pathname, $callback->($entries)
182 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
183 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
184 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
185 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
186 aio_load $pathname, $data, $callback->($status)
187 aio_copy $srcpath, $dstpath, $callback->($status)
188 aio_move $srcpath, $dstpath, $callback->($status)
189 aio_rmtree $pathname, $callback->($status)
190 aio_sync $callback->($status)
191 aio_syncfs $fh, $callback->($status)
192 aio_fsync $fh, $callback->($status)
193 aio_fdatasync $fh, $callback->($status)
194 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
195 aio_pathsync $pathname, $callback->($status)
196 aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
197 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
198 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
199 aio_mlockall $flags, $callback->($status)
200 aio_group $callback->(...)
201 aio_nop $callback->()
202
203 $prev_pri = aioreq_pri [$pri]
204 aioreq_nice $pri_adjust
205
206 IO::AIO::poll_wait
207 IO::AIO::poll_cb
208 IO::AIO::poll
209 IO::AIO::flush
210 IO::AIO::max_poll_reqs $nreqs
211 IO::AIO::max_poll_time $seconds
212 IO::AIO::min_parallel $nthreads
213 IO::AIO::max_parallel $nthreads
214 IO::AIO::max_idle $nthreads
215 IO::AIO::idle_timeout $seconds
216 IO::AIO::max_outstanding $maxreqs
217 IO::AIO::nreqs
218 IO::AIO::nready
219 IO::AIO::npending
220
221 IO::AIO::sendfile $ofh, $ifh, $offset, $count
222 IO::AIO::fadvise $fh, $offset, $len, $advice
223 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
224 IO::AIO::munmap $scalar
225 IO::AIO::madvise $scalar, $offset, $length, $advice
226 IO::AIO::mprotect $scalar, $offset, $length, $protect
227 IO::AIO::munlock $scalar, $offset = 0, $length = undef
228 IO::AIO::munlockall
229
230 API NOTES
174 All the "aio_*" calls are more or less thin wrappers around the syscall 231 All the "aio_*" calls are more or less thin wrappers around the syscall
175 with the same name (sans "aio_"). The arguments are similar or 232 with the same name (sans "aio_"). The arguments are similar or
176 identical, and they all accept an additional (and optional) $callback 233 identical, and they all accept an additional (and optional) $callback
177 argument which must be a code reference. This code reference will get 234 argument which must be a code reference. This code reference will be
178 called with the syscall return code (e.g. most syscalls return -1 on 235 called after the syscall has been executed in an asynchronous fashion.
179 error, unlike perl, which usually delivers "false") as it's sole 236 The results of the request will be passed as arguments to the callback
180 argument when the given syscall has been executed asynchronously. 237 (and, if an error occured, in $!) - for most requests the syscall return
238 code (e.g. most syscalls return -1 on error, unlike perl, which usually
239 delivers "false").
240
241 Some requests (such as "aio_readdir") pass the actual results and
242 communicate failures by passing "undef".
181 243
182 All functions expecting a filehandle keep a copy of the filehandle 244 All functions expecting a filehandle keep a copy of the filehandle
183 internally until the request has finished. 245 internally until the request has finished.
184 246
185 All functions return request objects of type IO::AIO::REQ that allow 247 All functions return request objects of type IO::AIO::REQ that allow
186 further manipulation of those requests while they are in-flight. 248 further manipulation of those requests while they are in-flight.
187 249
188 The pathnames you pass to these routines *must* be absolute and encoded 250 The pathnames you pass to these routines *should* be absolute. The
189 as octets. The reason for the former is that at the time the request is 251 reason for this is that at the time the request is being executed, the
190 being executed, the current working directory could have changed. 252 current working directory could have changed. Alternatively, you can
191 Alternatively, you can make sure that you never change the current 253 make sure that you never change the current working directory anywhere
192 working directory anywhere in the program and then use relative paths. 254 in the program and then use relative paths. You can also take advantage
255 of IO::AIOs working directory abstraction, that lets you specify paths
256 relative to some previously-opened "working directory object" - see the
257 description of the "IO::AIO::WD" class later in this document.
193 258
194 To encode pathnames as octets, either make sure you either: a) always 259 To encode pathnames as octets, either make sure you either: a) always
195 pass in filenames you got from outside (command line, readdir etc.) 260 pass in filenames you got from outside (command line, readdir etc.)
196 without tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module 261 without tinkering, b) are in your native filesystem encoding, c) use the
197 and encode your pathnames to the locale (or other) encoding in effect in 262 Encode module and encode your pathnames to the locale (or other)
198 the user environment, d) use Glib::filename_from_unicode on unicode 263 encoding in effect in the user environment, d) use
199 filenames or e) use something else to ensure your scalar has the correct 264 Glib::filename_from_unicode on unicode filenames or e) use something
200 contents. 265 else to ensure your scalar has the correct contents.
201 266
202 This works, btw. independent of the internal UTF-8 bit, which IO::AIO 267 This works, btw. independent of the internal UTF-8 bit, which IO::AIO
203 handles correctly wether it is set or not. 268 handles correctly whether it is set or not.
204 269
270 AIO REQUEST FUNCTIONS
205 $prev_pri = aioreq_pri [$pri] 271 $prev_pri = aioreq_pri [$pri]
206 Returns the priority value that would be used for the next request 272 Returns the priority value that would be used for the next request
207 and, if $pri is given, sets the priority for the next aio request. 273 and, if $pri is given, sets the priority for the next aio request.
208 274
209 The default priority is 0, the minimum and maximum priorities are -4 275 The default priority is 0, the minimum and maximum priorities are -4
… …
231 Similar to "aioreq_pri", but subtracts the given value from the 297 Similar to "aioreq_pri", but subtracts the given value from the
232 current priority, so the effect is cumulative. 298 current priority, so the effect is cumulative.
233 299
234 aio_open $pathname, $flags, $mode, $callback->($fh) 300 aio_open $pathname, $flags, $mode, $callback->($fh)
235 Asynchronously open or create a file and call the callback with a 301 Asynchronously open or create a file and call the callback with a
236 newly created filehandle for the file. 302 newly created filehandle for the file (or "undef" in case of an
303 error).
237 304
238 The pathname passed to "aio_open" must be absolute. See API NOTES, 305 The pathname passed to "aio_open" must be absolute. See API NOTES,
239 above, for an explanation. 306 above, for an explanation.
240 307
241 The $flags argument is a bitmask. See the "Fcntl" module for a list. 308 The $flags argument is a bitmask. See the "Fcntl" module for a list.
… …
248 will be modified by the umask in effect then the request is being 315 will be modified by the umask in effect then the request is being
249 executed, so better never change the umask. 316 executed, so better never change the umask.
250 317
251 Example: 318 Example:
252 319
253 aio_open "/etc/passwd", O_RDONLY, 0, sub { 320 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
254 if ($_[0]) { 321 if ($_[0]) {
255 print "open successful, fh is $_[0]\n"; 322 print "open successful, fh is $_[0]\n";
256 ... 323 ...
257 } else { 324 } else {
258 die "open failed: $!\n"; 325 die "open failed: $!\n";
259 } 326 }
260 }; 327 };
261 328
329 In addition to all the common open modes/flags ("O_RDONLY",
330 "O_WRONLY", "O_RDWR", "O_CREAT", "O_TRUNC", "O_EXCL" and
331 "O_APPEND"), the following POSIX and non-POSIX constants are
332 available (missing ones on your system are, as usual, 0):
333
334 "O_ASYNC", "O_DIRECT", "O_NOATIME", "O_CLOEXEC", "O_NOCTTY",
335 "O_NOFOLLOW", "O_NONBLOCK", "O_EXEC", "O_SEARCH", "O_DIRECTORY",
336 "O_DSYNC", "O_RSYNC", "O_SYNC" and "O_TTY_INIT".
337
262 aio_close $fh, $callback->($status) 338 aio_close $fh, $callback->($status)
263 Asynchronously close a file and call the callback with the result 339 Asynchronously close a file and call the callback with the result
264 code. *WARNING:* although accepted, you should not pass in a perl 340 code.
265 filehandle here, as perl will likely close the file descriptor
266 another time when the filehandle is destroyed. Normally, you can
267 safely call perls "close" or just let filehandles go out of scope.
268 341
269 This is supposed to be a bug in the API, so that might change. It's 342 Unfortunately, you can't do this to perl. Perl *insists* very
270 therefore best to avoid this function. 343 strongly on closing the file descriptor associated with the
344 filehandle itself.
345
346 Therefore, "aio_close" will not close the filehandle - instead it
347 will use dup2 to overwrite the file descriptor with the write-end of
348 a pipe (the pipe fd will be created on demand and will be cached).
349
350 Or in other words: the file descriptor will be closed, but it will
351 not be free for reuse until the perl filehandle is closed.
352
353 aio_seek $fh, $offset, $whence, $callback->($offs)
354 Seeks the filehandle to the new $offset, similarly to perl's
355 "sysseek". The $whence can use the traditional values (0 for
356 "IO::AIO::SEEK_SET", 1 for "IO::AIO::SEEK_CUR" or 2 for
357 "IO::AIO::SEEK_END").
358
359 The resulting absolute offset will be passed to the callback, or -1
360 in case of an error.
361
362 In theory, the $whence constants could be different than the
363 corresponding values from Fcntl, but perl guarantees they are the
364 same, so don't panic.
365
366 As a GNU/Linux (and maybe Solaris) extension, also the constants
367 "IO::AIO::SEEK_DATA" and "IO::AIO::SEEK_HOLE" are available, if they
368 could be found. No guarantees about suitability for use in
369 "aio_seek" or Perl's "sysseek" can be made though, although I would
370 naively assume they "just work".
271 371
272 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 372 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
273 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 373 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
274 Reads or writes "length" bytes from the specified "fh" and "offset" 374 Reads or writes $length bytes from or to the specified $fh and
275 into the scalar given by "data" and offset "dataoffset" and calls 375 $offset into the scalar given by $data and offset $dataoffset and
276 the callback without the actual number of bytes read (or -1 on 376 calls the callback without the actual number of bytes read (or -1 on
277 error, just like the syscall). 377 error, just like the syscall).
278 378
379 "aio_read" will, like "sysread", shrink or grow the $data scalar to
380 offset plus the actual number of bytes read.
381
382 If $offset is undefined, then the current file descriptor offset
383 will be used (and updated), otherwise the file descriptor offset
384 will not be changed by these calls.
385
386 If $length is undefined in "aio_write", use the remaining length of
387 $data.
388
389 If $dataoffset is less than zero, it will be counted from the end of
390 $data.
391
279 The $data scalar *MUST NOT* be modified in any way while the request 392 The $data scalar *MUST NOT* be modified in any way while the request
280 is outstanding. Modifying it can result in segfaults or WW3 (if the 393 is outstanding. Modifying it can result in segfaults or World War
281 necessary/optional hardware is installed). 394 III (if the necessary/optional hardware is installed).
282 395
283 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at 396 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at
284 offset 0 within the scalar: 397 offset 0 within the scalar:
285 398
286 aio_read $fh, 7, 15, $buffer, 0, sub { 399 aio_read $fh, 7, 15, $buffer, 0, sub {
… …
291 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 404 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
292 Tries to copy $length bytes from $in_fh to $out_fh. It starts 405 Tries to copy $length bytes from $in_fh to $out_fh. It starts
293 reading at byte offset $in_offset, and starts writing at the current 406 reading at byte offset $in_offset, and starts writing at the current
294 file offset of $out_fh. Because of that, it is not safe to issue 407 file offset of $out_fh. Because of that, it is not safe to issue
295 more than one "aio_sendfile" per $out_fh, as they will interfere 408 more than one "aio_sendfile" per $out_fh, as they will interfere
296 with each other. 409 with each other. The same $in_fh works fine though, as this function
410 does not move or use the file offset of $in_fh.
297 411
412 Please note that "aio_sendfile" can read more bytes from $in_fh than
413 are written, and there is no way to find out how many more bytes
414 have been read from "aio_sendfile" alone, as "aio_sendfile" only
415 provides the number of bytes written to $out_fh. Only if the result
416 value equals $length one can assume that $length bytes have been
417 read.
418
419 Unlike with other "aio_" functions, it makes a lot of sense to use
420 "aio_sendfile" on non-blocking sockets, as long as one end
421 (typically the $in_fh) is a file - the file I/O will then be
422 asynchronous, while the socket I/O will be non-blocking. Note,
423 however, that you can run into a trap where "aio_sendfile" reads
424 some data with readahead, then fails to write all data, and when the
425 socket is ready the next time, the data in the cache is already
426 lost, forcing "aio_sendfile" to again hit the disk. Explicit
427 "aio_read" + "aio_write" let's you better control resource usage.
428
298 This call tries to make use of a native "sendfile" syscall to 429 This call tries to make use of a native "sendfile"-like syscall to
299 provide zero-copy operation. For this to work, $out_fh should refer 430 provide zero-copy operation. For this to work, $out_fh should refer
300 to a socket, and $in_fh should refer to mmap'able file. 431 to a socket, and $in_fh should refer to an mmap'able file.
301 432
302 If the native sendfile call fails or is not implemented, it will be 433 If a native sendfile cannot be found or it fails with "ENOSYS",
303 emulated, so you can call "aio_sendfile" on any type of filehandle 434 "EINVAL", "ENOTSUP", "EOPNOTSUPP", "EAFNOSUPPORT", "EPROTOTYPE" or
304 regardless of the limitations of the operating system. 435 "ENOTSOCK", it will be emulated, so you can call "aio_sendfile" on
436 any type of filehandle regardless of the limitations of the
437 operating system.
305 438
306 Please note, however, that "aio_sendfile" can read more bytes from 439 As native sendfile syscalls (as practically any non-POSIX interface
307 $in_fh than are written, and there is no way to find out how many 440 hacked together in a hurry to improve benchmark numbers) tend to be
308 bytes have been read from "aio_sendfile" alone, as "aio_sendfile" 441 rather buggy on many systems, this implementation tries to work
309 only provides the number of bytes written to $out_fh. Only if the 442 around some known bugs in Linux and FreeBSD kernels (probably
310 result value equals $length one can assume that $length bytes have 443 others, too), but that might fail, so you really really should check
311 been read. 444 the return value of "aio_sendfile" - fewre bytes than expected might
445 have been transferred.
312 446
313 aio_readahead $fh,$offset,$length, $callback->($retval) 447 aio_readahead $fh,$offset,$length, $callback->($retval)
314 "aio_readahead" populates the page cache with data from a file so 448 "aio_readahead" populates the page cache with data from a file so
315 that subsequent reads from that file will not block on disk I/O. The 449 that subsequent reads from that file will not block on disk I/O. The
316 $offset argument specifies the starting point from which data is to 450 $offset argument specifies the starting point from which data is to
… …
337 Currently, the stats are always 64-bit-stats, i.e. instead of 471 Currently, the stats are always 64-bit-stats, i.e. instead of
338 returning an error when stat'ing a large file, the results will be 472 returning an error when stat'ing a large file, the results will be
339 silently truncated unless perl itself is compiled with large file 473 silently truncated unless perl itself is compiled with large file
340 support. 474 support.
341 475
476 To help interpret the mode and dev/rdev stat values, IO::AIO offers
477 the following constants and functions (if not implemented, the
478 constants will be 0 and the functions will either "croak" or fall
479 back on traditional behaviour).
480
481 "S_IFMT", "S_IFIFO", "S_IFCHR", "S_IFBLK", "S_IFLNK", "S_IFREG",
482 "S_IFDIR", "S_IFWHT", "S_IFSOCK", "IO::AIO::major $dev_t",
483 "IO::AIO::minor $dev_t", "IO::AIO::makedev $major, $minor".
484
342 Example: Print the length of /etc/passwd: 485 Example: Print the length of /etc/passwd:
343 486
344 aio_stat "/etc/passwd", sub { 487 aio_stat "/etc/passwd", sub {
345 $_[0] and die "stat failed: $!"; 488 $_[0] and die "stat failed: $!";
346 print "size is ", -s _, "\n"; 489 print "size is ", -s _, "\n";
347 }; 490 };
348 491
492 aio_statvfs $fh_or_path, $callback->($statvfs)
493 Works like the POSIX "statvfs" or "fstatvfs" syscalls, depending on
494 whether a file handle or path was passed.
495
496 On success, the callback is passed a hash reference with the
497 following members: "bsize", "frsize", "blocks", "bfree", "bavail",
498 "files", "ffree", "favail", "fsid", "flag" and "namemax". On
499 failure, "undef" is passed.
500
501 The following POSIX IO::AIO::ST_* constants are defined: "ST_RDONLY"
502 and "ST_NOSUID".
503
504 The following non-POSIX IO::AIO::ST_* flag masks are defined to
505 their correct value when available, or to 0 on systems that do not
506 support them: "ST_NODEV", "ST_NOEXEC", "ST_SYNCHRONOUS",
507 "ST_MANDLOCK", "ST_WRITE", "ST_APPEND", "ST_IMMUTABLE",
508 "ST_NOATIME", "ST_NODIRATIME" and "ST_RELATIME".
509
510 Example: stat "/wd" and dump out the data if successful.
511
512 aio_statvfs "/wd", sub {
513 my $f = $_[0]
514 or die "statvfs: $!";
515
516 use Data::Dumper;
517 say Dumper $f;
518 };
519
520 # result:
521 {
522 bsize => 1024,
523 bfree => 4333064312,
524 blocks => 10253828096,
525 files => 2050765568,
526 flag => 4096,
527 favail => 2042092649,
528 bavail => 4333064312,
529 ffree => 2042092649,
530 namemax => 255,
531 frsize => 1024,
532 fsid => 1810
533 }
534
535 Here is a (likely partial) list of fsid values used by Linux - it is
536 safe to hardcode these when the $^O is "linux":
537
538 0x0000adf5 adfs
539 0x0000adff affs
540 0x5346414f afs
541 0x09041934 anon-inode filesystem
542 0x00000187 autofs
543 0x42465331 befs
544 0x1badface bfs
545 0x42494e4d binfmt_misc
546 0x9123683e btrfs
547 0x0027e0eb cgroupfs
548 0xff534d42 cifs
549 0x73757245 coda
550 0x012ff7b7 coh
551 0x28cd3d45 cramfs
552 0x453dcd28 cramfs-wend (wrong endianness)
553 0x64626720 debugfs
554 0x00001373 devfs
555 0x00001cd1 devpts
556 0x0000f15f ecryptfs
557 0x00414a53 efs
558 0x0000137d ext
559 0x0000ef53 ext2/ext3
560 0x0000ef51 ext2
561 0x00004006 fat
562 0x65735546 fuseblk
563 0x65735543 fusectl
564 0x0bad1dea futexfs
565 0x01161970 gfs2
566 0x47504653 gpfs
567 0x00004244 hfs
568 0xf995e849 hpfs
569 0x958458f6 hugetlbfs
570 0x2bad1dea inotifyfs
571 0x00009660 isofs
572 0x000072b6 jffs2
573 0x3153464a jfs
574 0x6b414653 k-afs
575 0x0bd00bd0 lustre
576 0x0000137f minix
577 0x0000138f minix 30 char names
578 0x00002468 minix v2
579 0x00002478 minix v2 30 char names
580 0x00004d5a minix v3
581 0x19800202 mqueue
582 0x00004d44 msdos
583 0x0000564c novell
584 0x00006969 nfs
585 0x6e667364 nfsd
586 0x00003434 nilfs
587 0x5346544e ntfs
588 0x00009fa1 openprom
589 0x7461636F ocfs2
590 0x00009fa0 proc
591 0x6165676c pstorefs
592 0x0000002f qnx4
593 0x858458f6 ramfs
594 0x52654973 reiserfs
595 0x00007275 romfs
596 0x67596969 rpc_pipefs
597 0x73636673 securityfs
598 0xf97cff8c selinux
599 0x0000517b smb
600 0x534f434b sockfs
601 0x73717368 squashfs
602 0x62656572 sysfs
603 0x012ff7b6 sysv2
604 0x012ff7b5 sysv4
605 0x01021994 tmpfs
606 0x15013346 udf
607 0x00011954 ufs
608 0x54190100 ufs byteswapped
609 0x00009fa2 usbdevfs
610 0x01021997 v9fs
611 0xa501fcf5 vxfs
612 0xabba1974 xenfs
613 0x012ff7b4 xenix
614 0x58465342 xfs
615 0x012fd16d xia
616
617 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
618 Works like perl's "utime" function (including the special case of
619 $atime and $mtime being undef). Fractional times are supported if
620 the underlying syscalls support them.
621
622 When called with a pathname, uses utimes(2) if available, otherwise
623 utime(2). If called on a file descriptor, uses futimes(2) if
624 available, otherwise returns ENOSYS, so this is not portable.
625
626 Examples:
627
628 # set atime and mtime to current time (basically touch(1)):
629 aio_utime "path", undef, undef;
630 # set atime to current time and mtime to beginning of the epoch:
631 aio_utime "path", time, undef; # undef==0
632
633 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
634 Works like perl's "chown" function, except that "undef" for either
635 $uid or $gid is being interpreted as "do not change" (but -1 can
636 also be used).
637
638 Examples:
639
640 # same as "chown root path" in the shell:
641 aio_chown "path", 0, -1;
642 # same as above:
643 aio_chown "path", 0, undef;
644
645 aio_truncate $fh_or_path, $offset, $callback->($status)
646 Works like truncate(2) or ftruncate(2).
647
648 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
649 Allocates or freed disk space according to the $mode argument. See
650 the linux "fallocate" docuemntation for details.
651
652 $mode can currently be 0 or "IO::AIO::FALLOC_FL_KEEP_SIZE" to
653 allocate space, or "IO::AIO::FALLOC_FL_PUNCH_HOLE |
654 IO::AIO::FALLOC_FL_KEEP_SIZE", to deallocate a file range.
655
656 The file system block size used by "fallocate" is presumably the
657 "f_bsize" returned by "statvfs".
658
659 If "fallocate" isn't available or cannot be emulated (currently no
660 emulation will be attempted), passes -1 and sets $! to "ENOSYS".
661
662 aio_chmod $fh_or_path, $mode, $callback->($status)
663 Works like perl's "chmod" function.
664
349 aio_unlink $pathname, $callback->($status) 665 aio_unlink $pathname, $callback->($status)
350 Asynchronously unlink (delete) a file and call the callback with the 666 Asynchronously unlink (delete) a file and call the callback with the
351 result code. 667 result code.
352 668
353 aio_mknod $path, $mode, $dev, $callback->($status) 669 aio_mknod $pathname, $mode, $dev, $callback->($status)
354 [EXPERIMENTAL] 670 [EXPERIMENTAL]
355 671
356 Asynchronously create a device node (or fifo). See mknod(2). 672 Asynchronously create a device node (or fifo). See mknod(2).
357 673
358 The only (POSIX-) portable way of calling this function is: 674 The only (POSIX-) portable way of calling this function is:
359 675
360 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 676 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
677
678 See "aio_stat" for info about some potentially helpful extra
679 constants and functions.
361 680
362 aio_link $srcpath, $dstpath, $callback->($status) 681 aio_link $srcpath, $dstpath, $callback->($status)
363 Asynchronously create a new link to the existing object at $srcpath 682 Asynchronously create a new link to the existing object at $srcpath
364 at the path $dstpath and call the callback with the result code. 683 at the path $dstpath and call the callback with the result code.
365 684
366 aio_symlink $srcpath, $dstpath, $callback->($status) 685 aio_symlink $srcpath, $dstpath, $callback->($status)
367 Asynchronously create a new symbolic link to the existing object at 686 Asynchronously create a new symbolic link to the existing object at
368 $srcpath at the path $dstpath and call the callback with the result 687 $srcpath at the path $dstpath and call the callback with the result
369 code. 688 code.
370 689
371 aio_readlink $path, $callback->($link) 690 aio_readlink $pathname, $callback->($link)
372 Asynchronously read the symlink specified by $path and pass it to 691 Asynchronously read the symlink specified by $path and pass it to
373 the callback. If an error occurs, nothing or undef gets passed to 692 the callback. If an error occurs, nothing or undef gets passed to
374 the callback. 693 the callback.
375 694
695 aio_realpath $pathname, $callback->($path)
696 Asynchronously make the path absolute and resolve any symlinks in
697 $path. The resulting path only consists of directories (same as
698 Cwd::realpath).
699
700 This request can be used to get the absolute path of the current
701 working directory by passing it a path of . (a single dot).
702
376 aio_rename $srcpath, $dstpath, $callback->($status) 703 aio_rename $srcpath, $dstpath, $callback->($status)
377 Asynchronously rename the object at $srcpath to $dstpath, just as 704 Asynchronously rename the object at $srcpath to $dstpath, just as
378 rename(2) and call the callback with the result code. 705 rename(2) and call the callback with the result code.
706
707 On systems that support the AIO::WD working directory abstraction
708 natively, the case "[$wd, "."]" as $srcpath is specialcased -
709 instead of failing, "rename" is called on the absolute path of $wd.
379 710
380 aio_mkdir $pathname, $mode, $callback->($status) 711 aio_mkdir $pathname, $mode, $callback->($status)
381 Asynchronously mkdir (create) a directory and call the callback with 712 Asynchronously mkdir (create) a directory and call the callback with
382 the result code. $mode will be modified by the umask at the time the 713 the result code. $mode will be modified by the umask at the time the
383 request is executed, so do not change your umask. 714 request is executed, so do not change your umask.
384 715
385 aio_rmdir $pathname, $callback->($status) 716 aio_rmdir $pathname, $callback->($status)
386 Asynchronously rmdir (delete) a directory and call the callback with 717 Asynchronously rmdir (delete) a directory and call the callback with
387 the result code. 718 the result code.
388 719
720 On systems that support the AIO::WD working directory abstraction
721 natively, the case "[$wd, "."]" is specialcased - instead of
722 failing, "rmdir" is called on the absolute path of $wd.
723
389 aio_readdir $pathname, $callback->($entries) 724 aio_readdir $pathname, $callback->($entries)
390 Unlike the POSIX call of the same name, "aio_readdir" reads an 725 Unlike the POSIX call of the same name, "aio_readdir" reads an
391 entire directory (i.e. opendir + readdir + closedir). The entries 726 entire directory (i.e. opendir + readdir + closedir). The entries
392 will not be sorted, and will NOT include the "." and ".." entries. 727 will not be sorted, and will NOT include the "." and ".." entries.
393 728
394 The callback a single argument which is either "undef" or an 729 The callback is passed a single argument which is either "undef" or
395 array-ref with the filenames. 730 an array-ref with the filenames.
396 731
732 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
733 Quite similar to "aio_readdir", but the $flags argument allows one
734 to tune behaviour and output format. In case of an error, $entries
735 will be "undef".
736
737 The flags are a combination of the following constants, ORed
738 together (the flags will also be passed to the callback, possibly
739 modified):
740
741 IO::AIO::READDIR_DENTS
742 When this flag is off, then the callback gets an arrayref
743 consisting of names only (as with "aio_readdir"), otherwise it
744 gets an arrayref with "[$name, $type, $inode]" arrayrefs, each
745 describing a single directory entry in more detail.
746
747 $name is the name of the entry.
748
749 $type is one of the "IO::AIO::DT_xxx" constants:
750
751 "IO::AIO::DT_UNKNOWN", "IO::AIO::DT_FIFO", "IO::AIO::DT_CHR",
752 "IO::AIO::DT_DIR", "IO::AIO::DT_BLK", "IO::AIO::DT_REG",
753 "IO::AIO::DT_LNK", "IO::AIO::DT_SOCK", "IO::AIO::DT_WHT".
754
755 "IO::AIO::DT_UNKNOWN" means just that: readdir does not know. If
756 you need to know, you have to run stat yourself. Also, for speed
757 reasons, the $type scalars are read-only: you can not modify
758 them.
759
760 $inode is the inode number (which might not be exact on systems
761 with 64 bit inode numbers and 32 bit perls). This field has
762 unspecified content on systems that do not deliver the inode
763 information.
764
765 IO::AIO::READDIR_DIRS_FIRST
766 When this flag is set, then the names will be returned in an
767 order where likely directories come first, in optimal stat
768 order. This is useful when you need to quickly find directories,
769 or you want to find all directories while avoiding to stat()
770 each entry.
771
772 If the system returns type information in readdir, then this is
773 used to find directories directly. Otherwise, likely directories
774 are names beginning with ".", or otherwise names with no dots,
775 of which names with short names are tried first.
776
777 IO::AIO::READDIR_STAT_ORDER
778 When this flag is set, then the names will be returned in an
779 order suitable for stat()'ing each one. That is, when you plan
780 to stat() all files in the given directory, then the returned
781 order will likely be fastest.
782
783 If both this flag and "IO::AIO::READDIR_DIRS_FIRST" are
784 specified, then the likely dirs come first, resulting in a less
785 optimal stat order.
786
787 IO::AIO::READDIR_FOUND_UNKNOWN
788 This flag should not be set when calling "aio_readdirx".
789 Instead, it is being set by "aio_readdirx", when any of the
790 $type's found were "IO::AIO::DT_UNKNOWN". The absence of this
791 flag therefore indicates that all $type's are known, which can
792 be used to speed up some algorithms.
793
397 aio_load $path, $data, $callback->($status) 794 aio_load $pathname, $data, $callback->($status)
398 This is a composite request that tries to fully load the given file 795 This is a composite request that tries to fully load the given file
399 into memory. Status is the same as with aio_read. 796 into memory. Status is the same as with aio_read.
400 797
401 aio_copy $srcpath, $dstpath, $callback->($status) 798 aio_copy $srcpath, $dstpath, $callback->($status)
402 Try to copy the *file* (directories not supported as either source 799 Try to copy the *file* (directories not supported as either source
403 or destination) from $srcpath to $dstpath and call the callback with 800 or destination) from $srcpath to $dstpath and call the callback with
404 the 0 (error) or -1 ok. 801 a status of 0 (ok) or -1 (error, see $!).
405 802
406 This is a composite request that it creates the destination file 803 This is a composite request that creates the destination file with
407 with mode 0200 and copies the contents of the source file into it 804 mode 0200 and copies the contents of the source file into it using
408 using "aio_sendfile", followed by restoring atime, mtime, access 805 "aio_sendfile", followed by restoring atime, mtime, access mode and
409 mode and uid/gid, in that order. 806 uid/gid, in that order.
410 807
411 If an error occurs, the partial destination file will be unlinked, 808 If an error occurs, the partial destination file will be unlinked,
412 if possible, except when setting atime, mtime, access mode and 809 if possible, except when setting atime, mtime, access mode and
413 uid/gid, where errors are being ignored. 810 uid/gid, where errors are being ignored.
414 811
415 aio_move $srcpath, $dstpath, $callback->($status) 812 aio_move $srcpath, $dstpath, $callback->($status)
416 Try to move the *file* (directories not supported as either source 813 Try to move the *file* (directories not supported as either source
417 or destination) from $srcpath to $dstpath and call the callback with 814 or destination) from $srcpath to $dstpath and call the callback with
418 the 0 (error) or -1 ok. 815 a status of 0 (ok) or -1 (error, see $!).
419 816
420 This is a composite request that tries to rename(2) the file first. 817 This is a composite request that tries to rename(2) the file first;
421 If rename files with "EXDEV", it copies the file with "aio_copy" 818 if rename fails with "EXDEV", it copies the file with "aio_copy"
422 and, if that is successful, unlinking the $srcpath. 819 and, if that is successful, unlinks the $srcpath.
423 820
424 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 821 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
425 Scans a directory (similar to "aio_readdir") but additionally tries 822 Scans a directory (similar to "aio_readdir") but additionally tries
426 to efficiently separate the entries of directory $path into two sets 823 to efficiently separate the entries of directory $path into two sets
427 of names, directories you can recurse into (directories), and ones 824 of names, directories you can recurse into (directories), and ones
428 you cannot recurse into (everything else, including symlinks to 825 you cannot recurse into (everything else, including symlinks to
429 directories). 826 directories).
… …
447 Implementation notes. 844 Implementation notes.
448 845
449 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry 846 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry
450 can. 847 can.
451 848
849 If readdir returns file type information, then this is used directly
850 to find directories.
851
452 After reading the directory, the modification time, size etc. of the 852 Otherwise, after reading the directory, the modification time, size
453 directory before and after the readdir is checked, and if they match 853 etc. of the directory before and after the readdir is checked, and
454 (and isn't the current time), the link count will be used to decide 854 if they match (and isn't the current time), the link count will be
455 how many entries are directories (if >= 2). Otherwise, no knowledge 855 used to decide how many entries are directories (if >= 2).
456 of the number of subdirectories will be assumed. 856 Otherwise, no knowledge of the number of subdirectories will be
857 assumed.
457 858
458 Then entries will be sorted into likely directories (everything 859 Then entries will be sorted into likely directories a non-initial
459 without a non-initial dot currently) and likely non-directories 860 dot currently) and likely non-directories (see "aio_readdirx"). Then
460 (everything else). Then every entry plus an appended "/." will be 861 every entry plus an appended "/." will be "stat"'ed, likely
461 "stat"'ed, likely directories first. If that succeeds, it assumes 862 directories first, in order of their inode numbers. If that
462 that the entry is a directory or a symlink to directory (which will 863 succeeds, it assumes that the entry is a directory or a symlink to
463 be checked seperately). This is often faster than stat'ing the entry 864 directory (which will be checked separately). This is often faster
464 itself because filesystems might detect the type of the entry 865 than stat'ing the entry itself because filesystems might detect the
465 without reading the inode data (e.g. ext2fs filetype feature). 866 type of the entry without reading the inode data (e.g. ext2fs
867 filetype feature), even on systems that cannot return the filetype
868 information on readdir.
466 869
467 If the known number of directories (link count - 2) has been 870 If the known number of directories (link count - 2) has been
468 reached, the rest of the entries is assumed to be non-directories. 871 reached, the rest of the entries is assumed to be non-directories.
469 872
470 This only works with certainty on POSIX (= UNIX) filesystems, which 873 This only works with certainty on POSIX (= UNIX) filesystems, which
… …
472 875
473 It will also likely work on non-POSIX filesystems with reduced 876 It will also likely work on non-POSIX filesystems with reduced
474 efficiency as those tend to return 0 or 1 as link counts, which 877 efficiency as those tend to return 0 or 1 as link counts, which
475 disables the directory counting heuristic. 878 disables the directory counting heuristic.
476 879
477 aio_rmtree $path, $callback->($status) 880 aio_rmtree $pathname, $callback->($status)
478 Delete a directory tree starting (and including) $path, return the 881 Delete a directory tree starting (and including) $path, return the
479 status of the final "rmdir" only. This is a composite request that 882 status of the final "rmdir" only. This is a composite request that
480 uses "aio_scandir" to recurse into and rmdir directories, and unlink 883 uses "aio_scandir" to recurse into and rmdir directories, and unlink
481 everything else. 884 everything else.
482 885
886 aio_sync $callback->($status)
887 Asynchronously call sync and call the callback when finished.
888
483 aio_fsync $fh, $callback->($status) 889 aio_fsync $fh, $callback->($status)
484 Asynchronously call fsync on the given filehandle and call the 890 Asynchronously call fsync on the given filehandle and call the
485 callback with the fsync result code. 891 callback with the fsync result code.
486 892
487 aio_fdatasync $fh, $callback->($status) 893 aio_fdatasync $fh, $callback->($status)
488 Asynchronously call fdatasync on the given filehandle and call the 894 Asynchronously call fdatasync on the given filehandle and call the
489 callback with the fdatasync result code. 895 callback with the fdatasync result code.
490 896
491 If this call isn't available because your OS lacks it or it couldn't 897 If this call isn't available because your OS lacks it or it couldn't
492 be detected, it will be emulated by calling "fsync" instead. 898 be detected, it will be emulated by calling "fsync" instead.
899
900 aio_syncfs $fh, $callback->($status)
901 Asynchronously call the syncfs syscall to sync the filesystem
902 associated to the given filehandle and call the callback with the
903 syncfs result code. If syncfs is not available, calls sync(), but
904 returns -1 and sets errno to "ENOSYS" nevertheless.
905
906 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
907 Sync the data portion of the file specified by $offset and $length
908 to disk (but NOT the metadata), by calling the Linux-specific
909 sync_file_range call. If sync_file_range is not available or it
910 returns ENOSYS, then fdatasync or fsync is being substituted.
911
912 $flags can be a combination of
913 "IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE",
914 "IO::AIO::SYNC_FILE_RANGE_WRITE" and
915 "IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER": refer to the sync_file_range
916 manpage for details.
917
918 aio_pathsync $pathname, $callback->($status)
919 This request tries to open, fsync and close the given path. This is
920 a composite request intended to sync directories after directory
921 operations (E.g. rename). This might not work on all operating
922 systems or have any specific effect, but usually it makes sure that
923 directory changes get written to disc. It works for anything that
924 can be opened for read-only, not just directories.
925
926 Future versions of this function might fall back to other methods
927 when "fsync" on the directory fails (such as calling "sync").
928
929 Passes 0 when everything went ok, and -1 on error.
930
931 aio_msync $scalar, $offset = 0, $length = undef, flags = 0,
932 $callback->($status)
933 This is a rather advanced IO::AIO call, which only works on
934 mmap(2)ed scalars (see the "IO::AIO::mmap" function, although it
935 also works on data scalars managed by the Sys::Mmap or Mmap modules,
936 note that the scalar must only be modified in-place while an aio
937 operation is pending on it).
938
939 It calls the "msync" function of your OS, if available, with the
940 memory area starting at $offset in the string and ending $length
941 bytes later. If $length is negative, counts from the end, and if
942 $length is "undef", then it goes till the end of the string. The
943 flags can be a combination of "IO::AIO::MS_ASYNC",
944 "IO::AIO::MS_INVALIDATE" and "IO::AIO::MS_SYNC".
945
946 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0,
947 $callback->($status)
948 This is a rather advanced IO::AIO call, which works best on
949 mmap(2)ed scalars.
950
951 It touches (reads or writes) all memory pages in the specified range
952 inside the scalar. All caveats and parameters are the same as for
953 "aio_msync", above, except for flags, which must be either 0 (which
954 reads all pages and ensures they are instantiated) or
955 "IO::AIO::MT_MODIFY", which modifies the memory pages (by reading
956 and writing an octet from it, which dirties the page).
957
958 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
959 This is a rather advanced IO::AIO call, which works best on
960 mmap(2)ed scalars.
961
962 It reads in all the pages of the underlying storage into memory (if
963 any) and locks them, so they are not getting swapped/paged out or
964 removed.
965
966 If $length is undefined, then the scalar will be locked till the
967 end.
968
969 On systems that do not implement "mlock", this function returns -1
970 and sets errno to "ENOSYS".
971
972 Note that the corresponding "munlock" is synchronous and is
973 documented under "MISCELLANEOUS FUNCTIONS".
974
975 Example: open a file, mmap and mlock it - both will be undone when
976 $data gets destroyed.
977
978 open my $fh, "<", $path or die "$path: $!";
979 my $data;
980 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
981 aio_mlock $data; # mlock in background
982
983 aio_mlockall $flags, $callback->($status)
984 Calls the "mlockall" function with the given $flags (a combination
985 of "IO::AIO::MCL_CURRENT" and "IO::AIO::MCL_FUTURE").
986
987 On systems that do not implement "mlockall", this function returns
988 -1 and sets errno to "ENOSYS".
989
990 Note that the corresponding "munlockall" is synchronous and is
991 documented under "MISCELLANEOUS FUNCTIONS".
992
993 Example: asynchronously lock all current and future pages into
994 memory.
995
996 aio_mlockall IO::AIO::MCL_FUTURE;
997
998 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
999 Queries the extents of the given file (by calling the Linux "FIEMAP"
1000 ioctl, see <http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for
1001 details). If the ioctl is not available on your OS, then this
1002 request will fail with "ENOSYS".
1003
1004 $start is the starting offset to query extents for, $length is the
1005 size of the range to query - if it is "undef", then the whole file
1006 will be queried.
1007
1008 $flags is a combination of flags ("IO::AIO::FIEMAP_FLAG_SYNC" or
1009 "IO::AIO::FIEMAP_FLAG_XATTR" - "IO::AIO::FIEMAP_FLAGS_COMPAT" is
1010 also exported), and is normally 0 or "IO::AIO::FIEMAP_FLAG_SYNC" to
1011 query the data portion.
1012
1013 $count is the maximum number of extent records to return. If it is
1014 "undef", then IO::AIO queries all extents of the range. As a very
1015 special case, if it is 0, then the callback receives the number of
1016 extents instead of the extents themselves (which is unreliable, see
1017 below).
1018
1019 If an error occurs, the callback receives no arguments. The special
1020 "errno" value "IO::AIO::EBADR" is available to test for flag errors.
1021
1022 Otherwise, the callback receives an array reference with extent
1023 structures. Each extent structure is an array reference itself, with
1024 the following members:
1025
1026 [$logical, $physical, $length, $flags]
1027
1028 Flags is any combination of the following flag values (typically
1029 either 0 or "IO::AIO::FIEMAP_EXTENT_LAST" (1)):
1030
1031 "IO::AIO::FIEMAP_EXTENT_LAST", "IO::AIO::FIEMAP_EXTENT_UNKNOWN",
1032 "IO::AIO::FIEMAP_EXTENT_DELALLOC", "IO::AIO::FIEMAP_EXTENT_ENCODED",
1033 "IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED",
1034 "IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED",
1035 "IO::AIO::FIEMAP_EXTENT_DATA_INLINE",
1036 "IO::AIO::FIEMAP_EXTENT_DATA_TAIL",
1037 "IO::AIO::FIEMAP_EXTENT_UNWRITTEN", "IO::AIO::FIEMAP_EXTENT_MERGED"
1038 or "IO::AIO::FIEMAP_EXTENT_SHARED".
1039
1040 At the time of this writing (Linux 3.2), this requets is unreliable
1041 unless $count is "undef", as the kernel has all sorts of bugs
1042 preventing it to return all extents of a range for files with large
1043 number of extents. The code works around all these issues if $count
1044 is undef.
493 1045
494 aio_group $callback->(...) 1046 aio_group $callback->(...)
495 This is a very special aio request: Instead of doing something, it 1047 This is a very special aio request: Instead of doing something, it
496 is a container for other aio requests, which is useful if you want 1048 is a container for other aio requests, which is useful if you want
497 to bundle many requests into a single, composite, request with a 1049 to bundle many requests into a single, composite, request with a
… …
532 requests like sleep and file handle readable/writable, the overhead 1084 requests like sleep and file handle readable/writable, the overhead
533 this creates is immense (it blocks a thread for a long time) so do 1085 this creates is immense (it blocks a thread for a long time) so do
534 not use this function except to put your application under 1086 not use this function except to put your application under
535 artificial I/O pressure. 1087 artificial I/O pressure.
536 1088
1089 IO::AIO::WD - multiple working directories
1090 Your process only has one current working directory, which is used by
1091 all threads. This makes it hard to use relative paths (some other
1092 component could call "chdir" at any time, and it is hard to control when
1093 the path will be used by IO::AIO).
1094
1095 One solution for this is to always use absolute paths. This usually
1096 works, but can be quite slow (the kernel has to walk the whole path on
1097 every access), and can also be a hassle to implement.
1098
1099 Newer POSIX systems have a number of functions (openat, fdopendir,
1100 futimensat and so on) that make it possible to specify working
1101 directories per operation.
1102
1103 For portability, and because the clowns who "designed", or shall I
1104 write, perpetrated this new interface were obviously half-drunk, this
1105 abstraction cannot be perfect, though.
1106
1107 IO::AIO allows you to convert directory paths into a so-called
1108 IO::AIO::WD object. This object stores the canonicalised, absolute
1109 version of the path, and on systems that allow it, also a directory file
1110 descriptor.
1111
1112 Everywhere where a pathname is accepted by IO::AIO (e.g. in "aio_stat"
1113 or "aio_unlink"), one can specify an array reference with an IO::AIO::WD
1114 object and a pathname instead (or the IO::AIO::WD object alone, which
1115 gets interpreted as "[$wd, "."]"). If the pathname is absolute, the
1116 IO::AIO::WD object is ignored, otherwise the pathname is resolved
1117 relative to that IO::AIO::WD object.
1118
1119 For example, to get a wd object for /etc and then stat passwd inside,
1120 you would write:
1121
1122 aio_wd "/etc", sub {
1123 my $etcdir = shift;
1124
1125 # although $etcdir can be undef on error, there is generally no reason
1126 # to check for errors here, as aio_stat will fail with ENOENT
1127 # when $etcdir is undef.
1128
1129 aio_stat [$etcdir, "passwd"], sub {
1130 # yay
1131 };
1132 };
1133
1134 That "aio_wd" is a request and not a normal function shows that creating
1135 an IO::AIO::WD object is itself a potentially blocking operation, which
1136 is why it is done asynchronously.
1137
1138 To stat the directory obtained with "aio_wd" above, one could write
1139 either of the following three request calls:
1140
1141 aio_lstat "/etc" , sub { ... # pathname as normal string
1142 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1143 aio_lstat $wd , sub { ... # shorthand for the previous
1144
1145 As with normal pathnames, IO::AIO keeps a copy of the working directory
1146 object and the pathname string, so you could write the following without
1147 causing any issues due to $path getting reused:
1148
1149 my $path = [$wd, undef];
1150
1151 for my $name (qw(abc def ghi)) {
1152 $path->[1] = $name;
1153 aio_stat $path, sub {
1154 # ...
1155 };
1156 }
1157
1158 There are some caveats: when directories get renamed (or deleted), the
1159 pathname string doesn't change, so will point to the new directory (or
1160 nowhere at all), while the directory fd, if available on the system,
1161 will still point to the original directory. Most functions accepting a
1162 pathname will use the directory fd on newer systems, and the string on
1163 older systems. Some functions (such as realpath) will always rely on the
1164 string form of the pathname.
1165
1166 So this functionality is mainly useful to get some protection against
1167 "chdir", to easily get an absolute path out of a relative path for
1168 future reference, and to speed up doing many operations in the same
1169 directory (e.g. when stat'ing all files in a directory).
1170
1171 The following functions implement this working directory abstraction:
1172
1173 aio_wd $pathname, $callback->($wd)
1174 Asynchonously canonicalise the given pathname and convert it to an
1175 IO::AIO::WD object representing it. If possible and supported on the
1176 system, also open a directory fd to speed up pathname resolution
1177 relative to this working directory.
1178
1179 If something goes wrong, then "undef" is passwd to the callback
1180 instead of a working directory object and $! is set appropriately.
1181 Since passing "undef" as working directory component of a pathname
1182 fails the request with "ENOENT", there is often no need for error
1183 checking in the "aio_wd" callback, as future requests using the
1184 value will fail in the expected way.
1185
1186 If this call isn't available because your OS lacks it or it couldn't
1187 be detected, it will be emulated by calling "fsync" instead.
1188
1189 IO::AIO::CWD
1190 This is a compiletime constant (object) that represents the process
1191 current working directory.
1192
1193 Specifying this object as working directory object for a pathname is
1194 as if the pathname would be specified directly, without a directory
1195 object. For example, these calls are functionally identical:
1196
1197 aio_stat "somefile", sub { ... };
1198 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1199
1200 To recover the path associated with an IO::AIO::WD object, you can use
1201 "aio_realpath":
1202
1203 aio_realpath $wd, sub {
1204 warn "path is $_[0]\n";
1205 };
1206
1207 Currently, "aio_statvfs" always, and "aio_rename" and "aio_rmdir"
1208 sometimes, fall back to using an absolue path.
1209
537 IO::AIO::REQ CLASS 1210 IO::AIO::REQ CLASS
538 All non-aggregate "aio_*" functions return an object of this class when 1211 All non-aggregate "aio_*" functions return an object of this class when
539 called in non-void context. 1212 called in non-void context.
540 1213
541 cancel $req 1214 cancel $req
542 Cancels the request, if possible. Has the effect of skipping 1215 Cancels the request, if possible. Has the effect of skipping
543 execution when entering the execute state and skipping calling the 1216 execution when entering the execute state and skipping calling the
544 callback when entering the the result state, but will leave the 1217 callback when entering the the result state, but will leave the
545 request otherwise untouched. That means that requests that currently 1218 request otherwise untouched (with the exception of readdir). That
546 execute will not be stopped and resources held by the request will 1219 means that requests that currently execute will not be stopped and
547 not be freed prematurely. 1220 resources held by the request will not be freed prematurely.
548 1221
549 cb $req $callback->(...) 1222 cb $req $callback->(...)
550 Replace (or simply set) the callback registered to the request. 1223 Replace (or simply set) the callback registered to the request.
551 1224
552 IO::AIO::GRP CLASS 1225 IO::AIO::GRP CLASS
… …
579 }; 1252 };
580 1253
581 This makes it very easy to create composite requests (see the source of 1254 This makes it very easy to create composite requests (see the source of
582 "aio_move" for an application) that work and feel like simple requests. 1255 "aio_move" for an application) that work and feel like simple requests.
583 1256
584 * The IO::AIO::GRP objects will be cleaned up during calls to 1257 * The IO::AIO::GRP objects will be cleaned up during calls to
585 "IO::AIO::poll_cb", just like any other request. 1258 "IO::AIO::poll_cb", just like any other request.
1259
586 * They can be canceled like any other request. Canceling will cancel not 1260 * They can be canceled like any other request. Canceling will cancel
587 only the request itself, but also all requests it contains. 1261 not only the request itself, but also all requests it contains.
1262
588 * They can also can also be added to other IO::AIO::GRP objects. 1263 * They can also can also be added to other IO::AIO::GRP objects.
1264
589 * You must not add requests to a group from within the group callback 1265 * You must not add requests to a group from within the group callback
590 (or any later time). 1266 (or any later time).
591 1267
592 Their lifetime, simplified, looks like this: when they are empty, they 1268 Their lifetime, simplified, looks like this: when they are empty, they
593 will finish very quickly. If they contain only requests that are in the 1269 will finish very quickly. If they contain only requests that are in the
594 "done" state, they will also finish. Otherwise they will continue to 1270 "done" state, they will also finish. Otherwise they will continue to
595 exist. 1271 exist.
596 1272
597 That means after creating a group you have some time to add requests. 1273 That means after creating a group you have some time to add requests
598 And in the callbacks of those requests, you can add further requests to 1274 (precisely before the callback has been invoked, which is only done
599 the group. And only when all those requests have finished will the the 1275 within the "poll_cb"). And in the callbacks of those requests, you can
600 group itself finish. 1276 add further requests to the group. And only when all those requests have
1277 finished will the the group itself finish.
601 1278
602 add $grp ... 1279 add $grp ...
603 $grp->add (...) 1280 $grp->add (...)
604 Add one or more requests to the group. Any type of IO::AIO::REQ can 1281 Add one or more requests to the group. Any type of IO::AIO::REQ can
605 be added, including other groups, as long as you do not create 1282 be added, including other groups, as long as you do not create
… …
610 $grp->cancel_subs 1287 $grp->cancel_subs
611 Cancel all subrequests and clears any feeder, but not the group 1288 Cancel all subrequests and clears any feeder, but not the group
612 request itself. Useful when you queued a lot of events but got a 1289 request itself. Useful when you queued a lot of events but got a
613 result early. 1290 result early.
614 1291
1292 The group request will finish normally (you cannot add requests to
1293 the group).
1294
615 $grp->result (...) 1295 $grp->result (...)
616 Set the result value(s) that will be passed to the group callback 1296 Set the result value(s) that will be passed to the group callback
617 when all subrequests have finished and set thre groups errno to the 1297 when all subrequests have finished and set the groups errno to the
618 current value of errno (just like calling "errno" without an error 1298 current value of errno (just like calling "errno" without an error
619 number). By default, no argument will be passed and errno is zero. 1299 number). By default, no argument will be passed and errno is zero.
620 1300
621 $grp->errno ([$errno]) 1301 $grp->errno ([$errno])
622 Sets the group errno value to $errno, or the current value of errno 1302 Sets the group errno value to $errno, or the current value of errno
… …
633 Sets a feeder/generator on this group: every group can have an 1313 Sets a feeder/generator on this group: every group can have an
634 attached generator that generates requests if idle. The idea behind 1314 attached generator that generates requests if idle. The idea behind
635 this is that, although you could just queue as many requests as you 1315 this is that, although you could just queue as many requests as you
636 want in a group, this might starve other requests for a potentially 1316 want in a group, this might starve other requests for a potentially
637 long time. For example, "aio_scandir" might generate hundreds of 1317 long time. For example, "aio_scandir" might generate hundreds of
638 thousands "aio_stat" requests, delaying any later requests for a 1318 thousands of "aio_stat" requests, delaying any later requests for a
639 long time. 1319 long time.
640 1320
641 To avoid this, and allow incremental generation of requests, you can 1321 To avoid this, and allow incremental generation of requests, you can
642 instead a group and set a feeder on it that generates those 1322 instead a group and set a feeder on it that generates those
643 requests. The feed callback will be called whenever there are few 1323 requests. The feed callback will be called whenever there are few
… …
648 does not impose any limits). 1328 does not impose any limits).
649 1329
650 If the feed does not queue more requests when called, it will be 1330 If the feed does not queue more requests when called, it will be
651 automatically removed from the group. 1331 automatically removed from the group.
652 1332
653 If the feed limit is 0, it will be set to 2 automatically. 1333 If the feed limit is 0 when this method is called, it will be set to
1334 2 automatically.
654 1335
655 Example: 1336 Example:
656 1337
657 # stat all files in @files, but only ever use four aio requests concurrently: 1338 # stat all files in @files, but only ever use four aio requests concurrently:
658 1339
… …
669 Sets the feeder limit for the group: The feeder will be called 1350 Sets the feeder limit for the group: The feeder will be called
670 whenever the group contains less than this many requests. 1351 whenever the group contains less than this many requests.
671 1352
672 Setting the limit to 0 will pause the feeding process. 1353 Setting the limit to 0 will pause the feeding process.
673 1354
1355 The default value for the limit is 0, but note that setting a feeder
1356 automatically bumps it up to 2.
1357
674 SUPPORT FUNCTIONS 1358 SUPPORT FUNCTIONS
675 EVENT PROCESSING AND EVENT LOOP INTEGRATION 1359 EVENT PROCESSING AND EVENT LOOP INTEGRATION
676 $fileno = IO::AIO::poll_fileno 1360 $fileno = IO::AIO::poll_fileno
677 Return the *request result pipe file descriptor*. This filehandle 1361 Return the *request result pipe file descriptor*. This filehandle
678 must be polled for reading by some mechanism outside this module 1362 must be polled for reading by some mechanism outside this module
679 (e.g. Event or select, see below or the SYNOPSIS). If the pipe 1363 (e.g. EV, Glib, select and so on, see below or the SYNOPSIS). If the
680 becomes readable you have to call "poll_cb" to check the results. 1364 pipe becomes readable you have to call "poll_cb" to check the
1365 results.
681 1366
682 See "poll_cb" for an example. 1367 See "poll_cb" for an example.
683 1368
684 IO::AIO::poll_cb 1369 IO::AIO::poll_cb
685 Process some outstanding events on the result pipe. You have to call 1370 Process some requests that have reached the result phase (i.e. they
686 this regularly. Returns the number of events processed. Returns 1371 have been executed but the results are not yet reported). You have
1372 to call this "regularly" to finish outstanding requests.
1373
1374 Returns 0 if all events could be processed (or there were no events
1375 to process), or -1 if it returned earlier for whatever reason.
687 immediately when no events are outstanding. The amount of events 1376 Returns immediately when no events are outstanding. The amount of
688 processed depends on the settings of "IO::AIO::max_poll_req" and 1377 events processed depends on the settings of "IO::AIO::max_poll_req",
689 "IO::AIO::max_poll_time". 1378 "IO::AIO::max_poll_time" and "IO::AIO::max_outstanding".
690 1379
691 If not all requests were processed for whatever reason, the 1380 If not all requests were processed for whatever reason, the poll
692 filehandle will still be ready when "poll_cb" returns. 1381 file descriptor will still be ready when "poll_cb" returns, so
1382 normally you don't have to do anything special to have it called
1383 later.
1384
1385 Apart from calling "IO::AIO::poll_cb" when the event filehandle
1386 becomes ready, it can be beneficial to call this function from loops
1387 which submit a lot of requests, to make sure the results get
1388 processed when they become available and not just when the loop is
1389 finished and the event loop takes over again. This function returns
1390 very fast when there are no outstanding requests.
693 1391
694 Example: Install an Event watcher that automatically calls 1392 Example: Install an Event watcher that automatically calls
695 IO::AIO::poll_cb with high priority: 1393 IO::AIO::poll_cb with high priority (more examples can be found in
1394 the SYNOPSIS section, at the top of this document):
696 1395
697 Event->io (fd => IO::AIO::poll_fileno, 1396 Event->io (fd => IO::AIO::poll_fileno,
698 poll => 'r', async => 1, 1397 poll => 'r', async => 1,
699 cb => \&IO::AIO::poll_cb); 1398 cb => \&IO::AIO::poll_cb);
1399
1400 IO::AIO::poll_wait
1401 Wait until either at least one request is in the result phase or no
1402 requests are outstanding anymore.
1403
1404 This is useful if you want to synchronously wait for some requests
1405 to become ready, without actually handling them.
1406
1407 See "nreqs" for an example.
1408
1409 IO::AIO::poll
1410 Waits until some requests have been handled.
1411
1412 Returns the number of requests processed, but is otherwise strictly
1413 equivalent to:
1414
1415 IO::AIO::poll_wait, IO::AIO::poll_cb
1416
1417 IO::AIO::flush
1418 Wait till all outstanding AIO requests have been handled.
1419
1420 Strictly equivalent to:
1421
1422 IO::AIO::poll_wait, IO::AIO::poll_cb
1423 while IO::AIO::nreqs;
700 1424
701 IO::AIO::max_poll_reqs $nreqs 1425 IO::AIO::max_poll_reqs $nreqs
702 IO::AIO::max_poll_time $seconds 1426 IO::AIO::max_poll_time $seconds
703 These set the maximum number of requests (default 0, meaning 1427 These set the maximum number of requests (default 0, meaning
704 infinity) that are being processed by "IO::AIO::poll_cb" in one 1428 infinity) that are being processed by "IO::AIO::poll_cb" in one
… …
728 # use a low priority so other tasks have priority 1452 # use a low priority so other tasks have priority
729 Event->io (fd => IO::AIO::poll_fileno, 1453 Event->io (fd => IO::AIO::poll_fileno,
730 poll => 'r', nice => 1, 1454 poll => 'r', nice => 1,
731 cb => &IO::AIO::poll_cb); 1455 cb => &IO::AIO::poll_cb);
732 1456
733 IO::AIO::poll_wait
734 If there are any outstanding requests and none of them in the result
735 phase, wait till the result filehandle becomes ready for reading
736 (simply does a "select" on the filehandle. This is useful if you
737 want to synchronously wait for some requests to finish).
738
739 See "nreqs" for an example.
740
741 IO::AIO::poll
742 Waits until some requests have been handled.
743
744 Returns the number of requests processed, but is otherwise strictly
745 equivalent to:
746
747 IO::AIO::poll_wait, IO::AIO::poll_cb
748
749 IO::AIO::flush
750 Wait till all outstanding AIO requests have been handled.
751
752 Strictly equivalent to:
753
754 IO::AIO::poll_wait, IO::AIO::poll_cb
755 while IO::AIO::nreqs;
756
757 CONTROLLING THE NUMBER OF THREADS 1457 CONTROLLING THE NUMBER OF THREADS
758 IO::AIO::min_parallel $nthreads 1458 IO::AIO::min_parallel $nthreads
759 Set the minimum number of AIO threads to $nthreads. The current 1459 Set the minimum number of AIO threads to $nthreads. The current
760 default is 8, which means eight asynchronous operations can execute 1460 default is 8, which means eight asynchronous operations can execute
761 concurrently at any one time (the number of outstanding requests, 1461 concurrently at any one time (the number of outstanding requests,
… …
790 1490
791 Under normal circumstances you don't need to call this function. 1491 Under normal circumstances you don't need to call this function.
792 1492
793 IO::AIO::max_idle $nthreads 1493 IO::AIO::max_idle $nthreads
794 Limit the number of threads (default: 4) that are allowed to idle 1494 Limit the number of threads (default: 4) that are allowed to idle
795 (i.e., threads that did not get a request to process within 10 1495 (i.e., threads that did not get a request to process within the idle
796 seconds). That means if a thread becomes idle while $nthreads other 1496 timeout (default: 10 seconds). That means if a thread becomes idle
797 threads are also idle, it will free its resources and exit. 1497 while $nthreads other threads are also idle, it will free its
1498 resources and exit.
798 1499
799 This is useful when you allow a large number of threads (e.g. 100 or 1500 This is useful when you allow a large number of threads (e.g. 100 or
800 1000) to allow for extremely high load situations, but want to free 1501 1000) to allow for extremely high load situations, but want to free
801 resources under normal circumstances (1000 threads can easily 1502 resources under normal circumstances (1000 threads can easily
802 consume 30MB of RAM). 1503 consume 30MB of RAM).
803 1504
804 The default is probably ok in most situations, especially if thread 1505 The default is probably ok in most situations, especially if thread
805 creation is fast. If thread creation is very slow on your system you 1506 creation is fast. If thread creation is very slow on your system you
806 might want to use larger values. 1507 might want to use larger values.
807 1508
1509 IO::AIO::idle_timeout $seconds
1510 Sets the minimum idle timeout (default 10) after which worker
1511 threads are allowed to exit. SEe "IO::AIO::max_idle".
1512
808 $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 1513 IO::AIO::max_outstanding $maxreqs
1514 Sets the maximum number of outstanding requests to $nreqs. If you do
1515 queue up more than this number of requests, the next call to
1516 "IO::AIO::poll_cb" (and other functions calling "poll_cb", such as
1517 "IO::AIO::flush" or "IO::AIO::poll") will block until the limit is
1518 no longer exceeded.
1519
1520 In other words, this setting does not enforce a queue limit, but can
1521 be used to make poll functions block if the limit is exceeded.
1522
809 This is a very bad function to use in interactive programs because 1523 This is a very bad function to use in interactive programs because
810 it blocks, and a bad way to reduce concurrency because it is 1524 it blocks, and a bad way to reduce concurrency because it is
811 inexact: Better use an "aio_group" together with a feed callback. 1525 inexact: Better use an "aio_group" together with a feed callback.
812 1526
813 Sets the maximum number of outstanding requests to $nreqs. If you to 1527 It's main use is in scripts without an event loop - when you want to
814 queue up more than this number of requests, the next call to the 1528 stat a lot of files, you can write somehting like this:
815 "poll_cb" (and "poll_some" and other functions calling "poll_cb")
816 function will block until the limit is no longer exceeded.
817 1529
818 The default value is very large, so there is no practical limit on 1530 IO::AIO::max_outstanding 32;
1531
1532 for my $path (...) {
1533 aio_stat $path , ...;
1534 IO::AIO::poll_cb;
1535 }
1536
1537 IO::AIO::flush;
1538
1539 The call to "poll_cb" inside the loop will normally return
1540 instantly, but as soon as more thna 32 reqeusts are in-flight, it
1541 will block until some requests have been handled. This keeps the
1542 loop from pushing a large number of "aio_stat" requests onto the
1543 queue.
1544
1545 The default value for "max_outstanding" is very large, so there is
819 the number of outstanding requests. 1546 no practical limit on the number of outstanding requests.
820
821 You can still queue as many requests as you want. Therefore,
822 "max_oustsanding" is mainly useful in simple scripts (with low
823 values) or as a stop gap to shield against fatal memory overflow
824 (with large values).
825 1547
826 STATISTICAL INFORMATION 1548 STATISTICAL INFORMATION
827 IO::AIO::nreqs 1549 IO::AIO::nreqs
828 Returns the number of requests currently in the ready, execute or 1550 Returns the number of requests currently in the ready, execute or
829 pending states (i.e. for which their callback has not been invoked 1551 pending states (i.e. for which their callback has not been invoked
… …
840 1562
841 IO::AIO::npending 1563 IO::AIO::npending
842 Returns the number of requests currently in the pending state 1564 Returns the number of requests currently in the pending state
843 (executed, but not yet processed by poll_cb). 1565 (executed, but not yet processed by poll_cb).
844 1566
1567 MISCELLANEOUS FUNCTIONS
1568 IO::AIO implements some functions that might be useful, but are not
1569 asynchronous.
1570
1571 IO::AIO::sendfile $ofh, $ifh, $offset, $count
1572 Calls the "eio_sendfile_sync" function, which is like
1573 "aio_sendfile", but is blocking (this makes most sense if you know
1574 the input data is likely cached already and the output filehandle is
1575 set to non-blocking operations).
1576
1577 Returns the number of bytes copied, or -1 on error.
1578
1579 IO::AIO::fadvise $fh, $offset, $len, $advice
1580 Simply calls the "posix_fadvise" function (see its manpage for
1581 details). The following advice constants are available:
1582 "IO::AIO::FADV_NORMAL", "IO::AIO::FADV_SEQUENTIAL",
1583 "IO::AIO::FADV_RANDOM", "IO::AIO::FADV_NOREUSE",
1584 "IO::AIO::FADV_WILLNEED", "IO::AIO::FADV_DONTNEED".
1585
1586 On systems that do not implement "posix_fadvise", this function
1587 returns ENOSYS, otherwise the return value of "posix_fadvise".
1588
1589 IO::AIO::madvise $scalar, $offset, $len, $advice
1590 Simply calls the "posix_madvise" function (see its manpage for
1591 details). The following advice constants are available:
1592 "IO::AIO::MADV_NORMAL", "IO::AIO::MADV_SEQUENTIAL",
1593 "IO::AIO::MADV_RANDOM", "IO::AIO::MADV_WILLNEED",
1594 "IO::AIO::MADV_DONTNEED".
1595
1596 On systems that do not implement "posix_madvise", this function
1597 returns ENOSYS, otherwise the return value of "posix_madvise".
1598
1599 IO::AIO::mprotect $scalar, $offset, $len, $protect
1600 Simply calls the "mprotect" function on the preferably AIO::mmap'ed
1601 $scalar (see its manpage for details). The following protect
1602 constants are available: "IO::AIO::PROT_NONE", "IO::AIO::PROT_READ",
1603 "IO::AIO::PROT_WRITE", "IO::AIO::PROT_EXEC".
1604
1605 On systems that do not implement "mprotect", this function returns
1606 ENOSYS, otherwise the return value of "mprotect".
1607
1608 IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1609 Memory-maps a file (or anonymous memory range) and attaches it to
1610 the given $scalar, which will act like a string scalar. Returns true
1611 on success, and false otherwise.
1612
1613 The only operations allowed on the scalar are "substr"/"vec" that
1614 don't change the string length, and most read-only operations such
1615 as copying it or searching it with regexes and so on.
1616
1617 Anything else is unsafe and will, at best, result in memory leaks.
1618
1619 The memory map associated with the $scalar is automatically removed
1620 when the $scalar is destroyed, or when the "IO::AIO::mmap" or
1621 "IO::AIO::munmap" functions are called.
1622
1623 This calls the "mmap"(2) function internally. See your system's
1624 manual page for details on the $length, $prot and $flags parameters.
1625
1626 The $length must be larger than zero and smaller than the actual
1627 filesize.
1628
1629 $prot is a combination of "IO::AIO::PROT_NONE",
1630 "IO::AIO::PROT_EXEC", "IO::AIO::PROT_READ" and/or
1631 "IO::AIO::PROT_WRITE",
1632
1633 $flags can be a combination of "IO::AIO::MAP_SHARED" or
1634 "IO::AIO::MAP_PRIVATE", or a number of system-specific flags (when
1635 not available, the are defined as 0): "IO::AIO::MAP_ANONYMOUS"
1636 (which is set to "MAP_ANON" if your system only provides this
1637 constant), "IO::AIO::MAP_HUGETLB", "IO::AIO::MAP_LOCKED",
1638 "IO::AIO::MAP_NORESERVE", "IO::AIO::MAP_POPULATE" or
1639 "IO::AIO::MAP_NONBLOCK"
1640
1641 If $fh is "undef", then a file descriptor of -1 is passed.
1642
1643 $offset is the offset from the start of the file - it generally must
1644 be a multiple of "IO::AIO::PAGESIZE" and defaults to 0.
1645
1646 Example:
1647
1648 use Digest::MD5;
1649 use IO::AIO;
1650
1651 open my $fh, "<verybigfile"
1652 or die "$!";
1653
1654 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1655 or die "verybigfile: $!";
1656
1657 my $fast_md5 = md5 $data;
1658
1659 IO::AIO::munmap $scalar
1660 Removes a previous mmap and undefines the $scalar.
1661
1662 IO::AIO::munlock $scalar, $offset = 0, $length = undef
1663 Calls the "munlock" function, undoing the effects of a previous
1664 "aio_mlock" call (see its description for details).
1665
1666 IO::AIO::munlockall
1667 Calls the "munlockall" function.
1668
1669 On systems that do not implement "munlockall", this function returns
1670 ENOSYS, otherwise the return value of "munlockall".
1671
1672 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
1673 Calls the GNU/Linux splice(2) syscall, if available. If $r_off or
1674 $w_off are "undef", then "NULL" is passed for these, otherwise they
1675 should be the file offset.
1676
1677 $r_fh and $w_fh should not refer to the same file, as splice might
1678 silently corrupt the data in this case.
1679
1680 The following symbol flag values are available:
1681 "IO::AIO::SPLICE_F_MOVE", "IO::AIO::SPLICE_F_NONBLOCK",
1682 "IO::AIO::SPLICE_F_MORE" and "IO::AIO::SPLICE_F_GIFT".
1683
1684 See the splice(2) manpage for details.
1685
1686 IO::AIO::tee $r_fh, $w_fh, $length, $flags
1687 Calls the GNU/Linux tee(2) syscall, see it's manpage and the
1688 description for "IO::AIO::splice" above for details.
1689
1690EVENT LOOP INTEGRATION
1691 It is recommended to use AnyEvent::AIO to integrate IO::AIO
1692 automatically into many event loops:
1693
1694 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
1695 use AnyEvent::AIO;
1696
1697 You can also integrate IO::AIO manually into many event loops, here are
1698 some examples of how to do this:
1699
1700 # EV integration
1701 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
1702
1703 # Event integration
1704 Event->io (fd => IO::AIO::poll_fileno,
1705 poll => 'r',
1706 cb => \&IO::AIO::poll_cb);
1707
1708 # Glib/Gtk2 integration
1709 add_watch Glib::IO IO::AIO::poll_fileno,
1710 in => sub { IO::AIO::poll_cb; 1 };
1711
1712 # Tk integration
1713 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
1714 readable => \&IO::AIO::poll_cb);
1715
1716 # Danga::Socket integration
1717 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
1718 \&IO::AIO::poll_cb);
1719
845 FORK BEHAVIOUR 1720 FORK BEHAVIOUR
846 This module should do "the right thing" when the process using it forks: 1721 Usage of pthreads in a program changes the semantics of fork
1722 considerably. Specifically, only async-safe functions can be called
1723 after fork. Perl doesn't know about this, so in general, you cannot call
1724 fork with defined behaviour in perl if pthreads are involved. IO::AIO
1725 uses pthreads, so this applies, but many other extensions and (for
1726 inexplicable reasons) perl itself often is linked against pthreads, so
1727 this limitation applies to quite a lot of perls.
847 1728
848 Before the fork, IO::AIO enters a quiescent state where no requests can 1729 This module no longer tries to fight your OS, or POSIX. That means
849 be added in other threads and no results will be processed. After the 1730 IO::AIO only works in the process that loaded it. Forking is fully
850 fork the parent simply leaves the quiescent state and continues 1731 supported, but using IO::AIO in the child is not.
851 request/result processing, while the child frees the request/result
852 queue (so that the requests started before the fork will only be handled
853 in the parent). Threads will be started on demand until the limit set in
854 the parent process has been reached again.
855 1732
856 In short: the parent will, after a short pause, continue as if fork had 1733 You might get around by not *using* IO::AIO before (or after) forking.
857 not been called, while the child will act as if IO::AIO has not been 1734 You could also try to call the IO::AIO::reinit function in the child:
858 used yet. 1735
1736 IO::AIO::reinit
1737 Abandons all current requests and I/O threads and simply
1738 reinitialises all data structures. This is not an operation
1739 supported by any standards, but happens to work on GNU/Linux and
1740 some newer BSD systems.
1741
1742 The only reasonable use for this function is to call it after
1743 forking, if "IO::AIO" was used in the parent. Calling it while
1744 IO::AIO is active in the process will result in undefined behaviour.
1745 Calling it at any time will also result in any undefined (by POSIX)
1746 behaviour.
859 1747
860 MEMORY USAGE 1748 MEMORY USAGE
861 Per-request usage: 1749 Per-request usage:
862 1750
863 Each aio request uses - depending on your architecture - around 100-200 1751 Each aio request uses - depending on your architecture - around 100-200
864 bytes of memory. In addition, stat requests need a stat buffer (possibly 1752 bytes of memory. In addition, stat requests need a stat buffer (possibly
865 a few hundred bytes), readdir requires a result buffer and so on. Perl 1753 a few hundred bytes), readdir requires a result buffer and so on. Perl
866 scalars and other data passed into aio requests will also be locked and 1754 scalars and other data passed into aio requests will also be locked and
867 will consume memory till the request has entered the done state. 1755 will consume memory till the request has entered the done state.
868 1756
869 This is now awfully much, so queuing lots of requests is not usually a 1757 This is not awfully much, so queuing lots of requests is not usually a
870 problem. 1758 problem.
871 1759
872 Per-thread usage: 1760 Per-thread usage:
873 1761
874 In the execution phase, some aio requests require more memory for 1762 In the execution phase, some aio requests require more memory for
… …
877 1765
878KNOWN BUGS 1766KNOWN BUGS
879 Known bugs will be fixed in the next release. 1767 Known bugs will be fixed in the next release.
880 1768
881SEE ALSO 1769SEE ALSO
882 Coro::AIO. 1770 AnyEvent::AIO for easy integration into event loops, Coro::AIO for a
1771 more natural syntax.
883 1772
884AUTHOR 1773AUTHOR
885 Marc Lehmann <schmorp@schmorp.de> 1774 Marc Lehmann <schmorp@schmorp.de>
886 http://home.schmorp.de/ 1775 http://home.schmorp.de/
887 1776

Diff Legend

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