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

Comparing IO-AIO/README (file contents):
Revision 1.21 by root, Fri Dec 22 04:05:50 2006 UTC vs.
Revision 1.57 by root, Mon Jan 18 11:53:09 2016 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;
91 use Event; 69 use EV;
92 use IO::AIO; 70 use IO::AIO;
93 71
94 # register the IO::AIO callback with Event 72 # register the IO::AIO callback with EV
95 Event->io (fd => IO::AIO::poll_fileno, 73 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 74
99 # queue the request to open /etc/passwd 75 # queue the request to open /etc/passwd
100 aio_open "/etc/passwd", O_RDONLY, 0, sub { 76 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
101 my $fh = shift 77 my $fh = shift
102 or die "error while opening: $!"; 78 or die "error while opening: $!";
103 79
104 # stat'ing filehandles is generally non-blocking 80 # stat'ing filehandles is generally non-blocking
105 my $size = -s $fh; 81 my $size = -s $fh;
… …
114 90
115 # file contents now in $contents 91 # file contents now in $contents
116 print $contents; 92 print $contents;
117 93
118 # exit event loop and program 94 # exit event loop and program
119 Event::unloop; 95 EV::break;
120 }; 96 };
121 }; 97 };
122 98
123 # possibly queue up other requests, or open GUI windows, 99 # possibly queue up other requests, or open GUI windows,
124 # check for sockets etc. etc. 100 # check for sockets etc. etc.
125 101
126 # process events as long as there are some: 102 # process events as long as there are some:
127 Event::loop; 103 EV::run;
128 104
129REQUEST ANATOMY AND LIFETIME 105REQUEST ANATOMY AND LIFETIME
130 Every "aio_*" function creates a request. which is a C data structure 106 Every "aio_*" function creates a request. which is a C data structure
131 not directly visible to Perl. 107 not directly visible to Perl.
132 108
… …
168 anymore (except possibly for the Perl object, but its connection to 144 anymore (except possibly for the Perl object, but its connection to
169 the actual aio request is severed and calling its methods will 145 the actual aio request is severed and calling its methods will
170 either do nothing or result in a runtime error). 146 either do nothing or result in a runtime error).
171 147
172FUNCTIONS 148FUNCTIONS
173 AIO REQUEST FUNCTIONS 149 QUICK OVERVIEW
150 This section simply lists the prototypes most of the functions for quick
151 reference. See the following sections for function-by-function
152 documentation.
153
154 aio_wd $pathname, $callback->($wd)
155 aio_open $pathname, $flags, $mode, $callback->($fh)
156 aio_close $fh, $callback->($status)
157 aio_seek $fh,$offset,$whence, $callback->($offs)
158 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
159 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
160 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
161 aio_readahead $fh,$offset,$length, $callback->($retval)
162 aio_stat $fh_or_path, $callback->($status)
163 aio_lstat $fh, $callback->($status)
164 aio_statvfs $fh_or_path, $callback->($statvfs)
165 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
166 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
167 aio_chmod $fh_or_path, $mode, $callback->($status)
168 aio_truncate $fh_or_path, $offset, $callback->($status)
169 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
170 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
171 aio_unlink $pathname, $callback->($status)
172 aio_mknod $pathname, $mode, $dev, $callback->($status)
173 aio_link $srcpath, $dstpath, $callback->($status)
174 aio_symlink $srcpath, $dstpath, $callback->($status)
175 aio_readlink $pathname, $callback->($link)
176 aio_realpath $pathname, $callback->($path)
177 aio_rename $srcpath, $dstpath, $callback->($status)
178 aio_mkdir $pathname, $mode, $callback->($status)
179 aio_rmdir $pathname, $callback->($status)
180 aio_readdir $pathname, $callback->($entries)
181 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
182 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
183 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
184 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
185 aio_load $pathname, $data, $callback->($status)
186 aio_copy $srcpath, $dstpath, $callback->($status)
187 aio_move $srcpath, $dstpath, $callback->($status)
188 aio_rmtree $pathname, $callback->($status)
189 aio_sync $callback->($status)
190 aio_syncfs $fh, $callback->($status)
191 aio_fsync $fh, $callback->($status)
192 aio_fdatasync $fh, $callback->($status)
193 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
194 aio_pathsync $pathname, $callback->($status)
195 aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
196 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
197 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
198 aio_mlockall $flags, $callback->($status)
199 aio_group $callback->(...)
200 aio_nop $callback->()
201
202 $prev_pri = aioreq_pri [$pri]
203 aioreq_nice $pri_adjust
204
205 IO::AIO::poll_wait
206 IO::AIO::poll_cb
207 IO::AIO::poll
208 IO::AIO::flush
209 IO::AIO::max_poll_reqs $nreqs
210 IO::AIO::max_poll_time $seconds
211 IO::AIO::min_parallel $nthreads
212 IO::AIO::max_parallel $nthreads
213 IO::AIO::max_idle $nthreads
214 IO::AIO::idle_timeout $seconds
215 IO::AIO::max_outstanding $maxreqs
216 IO::AIO::nreqs
217 IO::AIO::nready
218 IO::AIO::npending
219
220 IO::AIO::sendfile $ofh, $ifh, $offset, $count
221 IO::AIO::fadvise $fh, $offset, $len, $advice
222 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
223 IO::AIO::munmap $scalar
224 IO::AIO::madvise $scalar, $offset, $length, $advice
225 IO::AIO::mprotect $scalar, $offset, $length, $protect
226 IO::AIO::munlock $scalar, $offset = 0, $length = undef
227 IO::AIO::munlockall
228
229 API NOTES
174 All the "aio_*" calls are more or less thin wrappers around the syscall 230 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 231 with the same name (sans "aio_"). The arguments are similar or
176 identical, and they all accept an additional (and optional) $callback 232 identical, and they all accept an additional (and optional) $callback
177 argument which must be a code reference. This code reference will get 233 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 234 called after the syscall has been executed in an asynchronous fashion.
179 error, unlike perl, which usually delivers "false") as it's sole 235 The results of the request will be passed as arguments to the callback
180 argument when the given syscall has been executed asynchronously. 236 (and, if an error occured, in $!) - for most requests the syscall return
237 code (e.g. most syscalls return -1 on error, unlike perl, which usually
238 delivers "false").
239
240 Some requests (such as "aio_readdir") pass the actual results and
241 communicate failures by passing "undef".
181 242
182 All functions expecting a filehandle keep a copy of the filehandle 243 All functions expecting a filehandle keep a copy of the filehandle
183 internally until the request has finished. 244 internally until the request has finished.
184 245
185 All functions return request objects of type IO::AIO::REQ that allow 246 All functions return request objects of type IO::AIO::REQ that allow
186 further manipulation of those requests while they are in-flight. 247 further manipulation of those requests while they are in-flight.
187 248
188 The pathnames you pass to these routines *must* be absolute and encoded 249 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 250 reason for this is that at the time the request is being executed, the
190 being executed, the current working directory could have changed. 251 current working directory could have changed. Alternatively, you can
191 Alternatively, you can make sure that you never change the current 252 make sure that you never change the current working directory anywhere
192 working directory anywhere in the program and then use relative paths. 253 in the program and then use relative paths. You can also take advantage
254 of IO::AIOs working directory abstraction, that lets you specify paths
255 relative to some previously-opened "working directory object" - see the
256 description of the "IO::AIO::WD" class later in this document.
193 257
194 To encode pathnames as octets, either make sure you either: a) always 258 To encode pathnames as octets, either make sure you either: a) always
195 pass in filenames you got from outside (command line, readdir etc.) 259 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 260 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 261 Encode module and encode your pathnames to the locale (or other)
198 the user environment, d) use Glib::filename_from_unicode on unicode 262 encoding in effect in the user environment, d) use
199 filenames or e) use something else to ensure your scalar has the correct 263 Glib::filename_from_unicode on unicode filenames or e) use something
200 contents. 264 else to ensure your scalar has the correct contents.
201 265
202 This works, btw. independent of the internal UTF-8 bit, which IO::AIO 266 This works, btw. independent of the internal UTF-8 bit, which IO::AIO
203 handles correctly wether it is set or not. 267 handles correctly whether it is set or not.
204 268
269 AIO REQUEST FUNCTIONS
205 $prev_pri = aioreq_pri [$pri] 270 $prev_pri = aioreq_pri [$pri]
206 Returns the priority value that would be used for the next request 271 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. 272 and, if $pri is given, sets the priority for the next aio request.
208 273
209 The default priority is 0, the minimum and maximum priorities are -4 274 The default priority is 0, the minimum and maximum priorities are -4
… …
231 Similar to "aioreq_pri", but subtracts the given value from the 296 Similar to "aioreq_pri", but subtracts the given value from the
232 current priority, so the effect is cumulative. 297 current priority, so the effect is cumulative.
233 298
234 aio_open $pathname, $flags, $mode, $callback->($fh) 299 aio_open $pathname, $flags, $mode, $callback->($fh)
235 Asynchronously open or create a file and call the callback with a 300 Asynchronously open or create a file and call the callback with a
236 newly created filehandle for the file. 301 newly created filehandle for the file (or "undef" in case of an
302 error).
237 303
238 The pathname passed to "aio_open" must be absolute. See API NOTES, 304 The pathname passed to "aio_open" must be absolute. See API NOTES,
239 above, for an explanation. 305 above, for an explanation.
240 306
241 The $flags argument is a bitmask. See the "Fcntl" module for a list. 307 The $flags argument is a bitmask. See the "Fcntl" module for a list.
242 They are the same as used by "sysopen". 308 They are the same as used by "sysopen".
243 309
244 Likewise, $mode specifies the mode of the newly created file, if it 310 Likewise, $mode specifies the mode of the newly created file, if it
245 didn't exist and "O_CREAT" has been given, just like perl's 311 didn't exist and "O_CREAT" has been given, just like perl's
246 "sysopen", except that it is mandatory (i.e. use 0 if you don't 312 "sysopen", except that it is mandatory (i.e. use 0 if you don't
247 create new files, and 0666 or 0777 if you do). 313 create new files, and 0666 or 0777 if you do). Note that the $mode
314 will be modified by the umask in effect then the request is being
315 executed, so better never change the umask.
248 316
249 Example: 317 Example:
250 318
251 aio_open "/etc/passwd", O_RDONLY, 0, sub { 319 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
252 if ($_[0]) { 320 if ($_[0]) {
253 print "open successful, fh is $_[0]\n"; 321 print "open successful, fh is $_[0]\n";
254 ... 322 ...
255 } else { 323 } else {
256 die "open failed: $!\n"; 324 die "open failed: $!\n";
257 } 325 }
258 }; 326 };
259 327
328 In addition to all the common open modes/flags ("O_RDONLY",
329 "O_WRONLY", "O_RDWR", "O_CREAT", "O_TRUNC", "O_EXCL" and
330 "O_APPEND"), the following POSIX and non-POSIX constants are
331 available (missing ones on your system are, as usual, 0):
332
333 "O_ASYNC", "O_DIRECT", "O_NOATIME", "O_CLOEXEC", "O_NOCTTY",
334 "O_NOFOLLOW", "O_NONBLOCK", "O_EXEC", "O_SEARCH", "O_DIRECTORY",
335 "O_DSYNC", "O_RSYNC", "O_SYNC", "O_PATH", "O_TMPFILE", and
336 "O_TTY_INIT".
337
260 aio_close $fh, $callback->($status) 338 aio_close $fh, $callback->($status)
261 Asynchronously close a file and call the callback with the result 339 Asynchronously close a file and call the callback with the result
262 code. *WARNING:* although accepted, you should not pass in a perl 340 code.
263 filehandle here, as perl will likely close the file descriptor
264 another time when the filehandle is destroyed. Normally, you can
265 safely call perls "close" or just let filehandles go out of scope.
266 341
267 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
268 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".
269 371
270 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 372 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
271 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 373 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
272 Reads or writes "length" bytes from the specified "fh" and "offset" 374 Reads or writes $length bytes from or to the specified $fh and
273 into the scalar given by "data" and offset "dataoffset" and calls 375 $offset into the scalar given by $data and offset $dataoffset and
274 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
275 error, just like the syscall). 377 error, just like the syscall).
276 378
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
277 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
278 is outstanding. Modifying it can result in segfaults or WW3 (if the 393 is outstanding. Modifying it can result in segfaults or World War
279 necessary/optional hardware is installed). 394 III (if the necessary/optional hardware is installed).
280 395
281 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
282 offset 0 within the scalar: 397 offset 0 within the scalar:
283 398
284 aio_read $fh, 7, 15, $buffer, 0, sub { 399 aio_read $fh, 7, 15, $buffer, 0, sub {
… …
289 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 404 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
290 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
291 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
292 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
293 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
294 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.
295 411
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
296 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
297 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
298 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.
299 432
300 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",
301 emulated, so you can call "aio_sendfile" on any type of filehandle 434 "EINVAL", "ENOTSUP", "EOPNOTSUPP", "EAFNOSUPPORT", "EPROTOTYPE" or
302 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.
303 438
304 Please note, however, that "aio_sendfile" can read more bytes from 439 As native sendfile syscalls (as practically any non-POSIX interface
305 $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
306 bytes have been read from "aio_sendfile" alone, as "aio_sendfile" 441 rather buggy on many systems, this implementation tries to work
307 only provides the number of bytes written to $out_fh. Only if the 442 around some known bugs in Linux and FreeBSD kernels (probably
308 result value equals $length one can assume that $length bytes have 443 others, too), but that might fail, so you really really should check
309 been read. 444 the return value of "aio_sendfile" - fewre bytes than expected might
445 have been transferred.
310 446
311 aio_readahead $fh,$offset,$length, $callback->($retval) 447 aio_readahead $fh,$offset,$length, $callback->($retval)
312 "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
313 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
314 $offset argument specifies the starting point from which data is to 450 $offset argument specifies the starting point from which data is to
… …
335 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
336 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
337 silently truncated unless perl itself is compiled with large file 473 silently truncated unless perl itself is compiled with large file
338 support. 474 support.
339 475
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
340 Example: Print the length of /etc/passwd: 485 Example: Print the length of /etc/passwd:
341 486
342 aio_stat "/etc/passwd", sub { 487 aio_stat "/etc/passwd", sub {
343 $_[0] and die "stat failed: $!"; 488 $_[0] and die "stat failed: $!";
344 print "size is ", -s _, "\n"; 489 print "size is ", -s _, "\n";
345 }; 490 };
346 491
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 - send me updates!) list of fsid values
536 used by Linux - it is safe to hardcode these when $^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/ext4
560 0x0000ef51 ext2
561 0xf2f52010 f2fs
562 0x00004006 fat
563 0x65735546 fuseblk
564 0x65735543 fusectl
565 0x0bad1dea futexfs
566 0x01161970 gfs2
567 0x47504653 gpfs
568 0x00004244 hfs
569 0xf995e849 hpfs
570 0x00c0ffee hostfs
571 0x958458f6 hugetlbfs
572 0x2bad1dea inotifyfs
573 0x00009660 isofs
574 0x000072b6 jffs2
575 0x3153464a jfs
576 0x6b414653 k-afs
577 0x0bd00bd0 lustre
578 0x0000137f minix
579 0x0000138f minix 30 char names
580 0x00002468 minix v2
581 0x00002478 minix v2 30 char names
582 0x00004d5a minix v3
583 0x19800202 mqueue
584 0x00004d44 msdos
585 0x0000564c novell
586 0x00006969 nfs
587 0x6e667364 nfsd
588 0x00003434 nilfs
589 0x5346544e ntfs
590 0x00009fa1 openprom
591 0x7461636F ocfs2
592 0x00009fa0 proc
593 0x6165676c pstorefs
594 0x0000002f qnx4
595 0x68191122 qnx6
596 0x858458f6 ramfs
597 0x52654973 reiserfs
598 0x00007275 romfs
599 0x67596969 rpc_pipefs
600 0x73636673 securityfs
601 0xf97cff8c selinux
602 0x0000517b smb
603 0x534f434b sockfs
604 0x73717368 squashfs
605 0x62656572 sysfs
606 0x012ff7b6 sysv2
607 0x012ff7b5 sysv4
608 0x01021994 tmpfs
609 0x15013346 udf
610 0x00011954 ufs
611 0x54190100 ufs byteswapped
612 0x00009fa2 usbdevfs
613 0x01021997 v9fs
614 0xa501fcf5 vxfs
615 0xabba1974 xenfs
616 0x012ff7b4 xenix
617 0x58465342 xfs
618 0x012fd16d xia
619
620 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
621 Works like perl's "utime" function (including the special case of
622 $atime and $mtime being undef). Fractional times are supported if
623 the underlying syscalls support them.
624
625 When called with a pathname, uses utimes(2) if available, otherwise
626 utime(2). If called on a file descriptor, uses futimes(2) if
627 available, otherwise returns ENOSYS, so this is not portable.
628
629 Examples:
630
631 # set atime and mtime to current time (basically touch(1)):
632 aio_utime "path", undef, undef;
633 # set atime to current time and mtime to beginning of the epoch:
634 aio_utime "path", time, undef; # undef==0
635
636 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
637 Works like perl's "chown" function, except that "undef" for either
638 $uid or $gid is being interpreted as "do not change" (but -1 can
639 also be used).
640
641 Examples:
642
643 # same as "chown root path" in the shell:
644 aio_chown "path", 0, -1;
645 # same as above:
646 aio_chown "path", 0, undef;
647
648 aio_truncate $fh_or_path, $offset, $callback->($status)
649 Works like truncate(2) or ftruncate(2).
650
651 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
652 Allocates or frees disk space according to the $mode argument. See
653 the linux "fallocate" documentation for details.
654
655 $mode is usually 0 or "IO::AIO::FALLOC_FL_KEEP_SIZE" to allocate
656 space, or "IO::AIO::FALLOC_FL_PUNCH_HOLE |
657 IO::AIO::FALLOC_FL_KEEP_SIZE", to deallocate a file range.
658
659 IO::AIO also supports "FALLOC_FL_COLLAPSE_RANGE", to remove a range
660 (without leaving a hole) and "FALLOC_FL_ZERO_RANGE", to zero a range
661 (see your fallocate(2) manpage).
662
663 The file system block size used by "fallocate" is presumably the
664 "f_bsize" returned by "statvfs".
665
666 If "fallocate" isn't available or cannot be emulated (currently no
667 emulation will be attempted), passes -1 and sets $! to "ENOSYS".
668
669 aio_chmod $fh_or_path, $mode, $callback->($status)
670 Works like perl's "chmod" function.
671
347 aio_unlink $pathname, $callback->($status) 672 aio_unlink $pathname, $callback->($status)
348 Asynchronously unlink (delete) a file and call the callback with the 673 Asynchronously unlink (delete) a file and call the callback with the
349 result code. 674 result code.
350 675
351 aio_mknod $path, $mode, $dev, $callback->($status) 676 aio_mknod $pathname, $mode, $dev, $callback->($status)
352 [EXPERIMENTAL] 677 [EXPERIMENTAL]
353 678
354 Asynchronously create a device node (or fifo). See mknod(2). 679 Asynchronously create a device node (or fifo). See mknod(2).
355 680
356 The only (POSIX-) portable way of calling this function is: 681 The only (POSIX-) portable way of calling this function is:
357 682
358 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 683 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
684
685 See "aio_stat" for info about some potentially helpful extra
686 constants and functions.
359 687
360 aio_link $srcpath, $dstpath, $callback->($status) 688 aio_link $srcpath, $dstpath, $callback->($status)
361 Asynchronously create a new link to the existing object at $srcpath 689 Asynchronously create a new link to the existing object at $srcpath
362 at the path $dstpath and call the callback with the result code. 690 at the path $dstpath and call the callback with the result code.
363 691
364 aio_symlink $srcpath, $dstpath, $callback->($status) 692 aio_symlink $srcpath, $dstpath, $callback->($status)
365 Asynchronously create a new symbolic link to the existing object at 693 Asynchronously create a new symbolic link to the existing object at
366 $srcpath at the path $dstpath and call the callback with the result 694 $srcpath at the path $dstpath and call the callback with the result
367 code. 695 code.
368 696
369 aio_readlink $path, $callback->($link) 697 aio_readlink $pathname, $callback->($link)
370 Asynchronously read the symlink specified by $path and pass it to 698 Asynchronously read the symlink specified by $path and pass it to
371 the callback. If an error occurs, nothing or undef gets passed to 699 the callback. If an error occurs, nothing or undef gets passed to
372 the callback. 700 the callback.
373 701
702 aio_realpath $pathname, $callback->($path)
703 Asynchronously make the path absolute and resolve any symlinks in
704 $path. The resulting path only consists of directories (same as
705 Cwd::realpath).
706
707 This request can be used to get the absolute path of the current
708 working directory by passing it a path of . (a single dot).
709
374 aio_rename $srcpath, $dstpath, $callback->($status) 710 aio_rename $srcpath, $dstpath, $callback->($status)
375 Asynchronously rename the object at $srcpath to $dstpath, just as 711 Asynchronously rename the object at $srcpath to $dstpath, just as
376 rename(2) and call the callback with the result code. 712 rename(2) and call the callback with the result code.
377 713
714 On systems that support the AIO::WD working directory abstraction
715 natively, the case "[$wd, "."]" as $srcpath is specialcased -
716 instead of failing, "rename" is called on the absolute path of $wd.
717
718 aio_mkdir $pathname, $mode, $callback->($status)
719 Asynchronously mkdir (create) a directory and call the callback with
720 the result code. $mode will be modified by the umask at the time the
721 request is executed, so do not change your umask.
722
378 aio_rmdir $pathname, $callback->($status) 723 aio_rmdir $pathname, $callback->($status)
379 Asynchronously rmdir (delete) a directory and call the callback with 724 Asynchronously rmdir (delete) a directory and call the callback with
380 the result code. 725 the result code.
726
727 On systems that support the AIO::WD working directory abstraction
728 natively, the case "[$wd, "."]" is specialcased - instead of
729 failing, "rmdir" is called on the absolute path of $wd.
381 730
382 aio_readdir $pathname, $callback->($entries) 731 aio_readdir $pathname, $callback->($entries)
383 Unlike the POSIX call of the same name, "aio_readdir" reads an 732 Unlike the POSIX call of the same name, "aio_readdir" reads an
384 entire directory (i.e. opendir + readdir + closedir). The entries 733 entire directory (i.e. opendir + readdir + closedir). The entries
385 will not be sorted, and will NOT include the "." and ".." entries. 734 will not be sorted, and will NOT include the "." and ".." entries.
386 735
387 The callback a single argument which is either "undef" or an 736 The callback is passed a single argument which is either "undef" or
388 array-ref with the filenames. 737 an array-ref with the filenames.
738
739 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
740 Quite similar to "aio_readdir", but the $flags argument allows one
741 to tune behaviour and output format. In case of an error, $entries
742 will be "undef".
743
744 The flags are a combination of the following constants, ORed
745 together (the flags will also be passed to the callback, possibly
746 modified):
747
748 IO::AIO::READDIR_DENTS
749 When this flag is off, then the callback gets an arrayref
750 consisting of names only (as with "aio_readdir"), otherwise it
751 gets an arrayref with "[$name, $type, $inode]" arrayrefs, each
752 describing a single directory entry in more detail.
753
754 $name is the name of the entry.
755
756 $type is one of the "IO::AIO::DT_xxx" constants:
757
758 "IO::AIO::DT_UNKNOWN", "IO::AIO::DT_FIFO", "IO::AIO::DT_CHR",
759 "IO::AIO::DT_DIR", "IO::AIO::DT_BLK", "IO::AIO::DT_REG",
760 "IO::AIO::DT_LNK", "IO::AIO::DT_SOCK", "IO::AIO::DT_WHT".
761
762 "IO::AIO::DT_UNKNOWN" means just that: readdir does not know. If
763 you need to know, you have to run stat yourself. Also, for speed
764 reasons, the $type scalars are read-only: you can not modify
765 them.
766
767 $inode is the inode number (which might not be exact on systems
768 with 64 bit inode numbers and 32 bit perls). This field has
769 unspecified content on systems that do not deliver the inode
770 information.
771
772 IO::AIO::READDIR_DIRS_FIRST
773 When this flag is set, then the names will be returned in an
774 order where likely directories come first, in optimal stat
775 order. This is useful when you need to quickly find directories,
776 or you want to find all directories while avoiding to stat()
777 each entry.
778
779 If the system returns type information in readdir, then this is
780 used to find directories directly. Otherwise, likely directories
781 are names beginning with ".", or otherwise names with no dots,
782 of which names with short names are tried first.
783
784 IO::AIO::READDIR_STAT_ORDER
785 When this flag is set, then the names will be returned in an
786 order suitable for stat()'ing each one. That is, when you plan
787 to stat() all files in the given directory, then the returned
788 order will likely be fastest.
789
790 If both this flag and "IO::AIO::READDIR_DIRS_FIRST" are
791 specified, then the likely dirs come first, resulting in a less
792 optimal stat order.
793
794 IO::AIO::READDIR_FOUND_UNKNOWN
795 This flag should not be set when calling "aio_readdirx".
796 Instead, it is being set by "aio_readdirx", when any of the
797 $type's found were "IO::AIO::DT_UNKNOWN". The absence of this
798 flag therefore indicates that all $type's are known, which can
799 be used to speed up some algorithms.
800
801 aio_load $pathname, $data, $callback->($status)
802 This is a composite request that tries to fully load the given file
803 into memory. Status is the same as with aio_read.
389 804
390 aio_copy $srcpath, $dstpath, $callback->($status) 805 aio_copy $srcpath, $dstpath, $callback->($status)
391 Try to copy the *file* (directories not supported as either source 806 Try to copy the *file* (directories not supported as either source
392 or destination) from $srcpath to $dstpath and call the callback with 807 or destination) from $srcpath to $dstpath and call the callback with
393 the 0 (error) or -1 ok. 808 a status of 0 (ok) or -1 (error, see $!).
394 809
395 This is a composite request that it creates the destination file 810 This is a composite request that creates the destination file with
396 with mode 0200 and copies the contents of the source file into it 811 mode 0200 and copies the contents of the source file into it using
397 using "aio_sendfile", followed by restoring atime, mtime, access 812 "aio_sendfile", followed by restoring atime, mtime, access mode and
398 mode and uid/gid, in that order. 813 uid/gid, in that order.
399 814
400 If an error occurs, the partial destination file will be unlinked, 815 If an error occurs, the partial destination file will be unlinked,
401 if possible, except when setting atime, mtime, access mode and 816 if possible, except when setting atime, mtime, access mode and
402 uid/gid, where errors are being ignored. 817 uid/gid, where errors are being ignored.
403 818
404 aio_move $srcpath, $dstpath, $callback->($status) 819 aio_move $srcpath, $dstpath, $callback->($status)
405 Try to move the *file* (directories not supported as either source 820 Try to move the *file* (directories not supported as either source
406 or destination) from $srcpath to $dstpath and call the callback with 821 or destination) from $srcpath to $dstpath and call the callback with
407 the 0 (error) or -1 ok. 822 a status of 0 (ok) or -1 (error, see $!).
408 823
409 This is a composite request that tries to rename(2) the file first. 824 This is a composite request that tries to rename(2) the file first;
410 If rename files with "EXDEV", it copies the file with "aio_copy" 825 if rename fails with "EXDEV", it copies the file with "aio_copy"
411 and, if that is successful, unlinking the $srcpath. 826 and, if that is successful, unlinks the $srcpath.
412 827
413 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 828 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
414 Scans a directory (similar to "aio_readdir") but additionally tries 829 Scans a directory (similar to "aio_readdir") but additionally tries
415 to efficiently separate the entries of directory $path into two sets 830 to efficiently separate the entries of directory $path into two sets
416 of names, directories you can recurse into (directories), and ones 831 of names, directories you can recurse into (directories), and ones
417 you cannot recurse into (everything else, including symlinks to 832 you cannot recurse into (everything else, including symlinks to
418 directories). 833 directories).
… …
436 Implementation notes. 851 Implementation notes.
437 852
438 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry 853 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry
439 can. 854 can.
440 855
856 If readdir returns file type information, then this is used directly
857 to find directories.
858
441 After reading the directory, the modification time, size etc. of the 859 Otherwise, after reading the directory, the modification time, size
442 directory before and after the readdir is checked, and if they match 860 etc. of the directory before and after the readdir is checked, and
443 (and isn't the current time), the link count will be used to decide 861 if they match (and isn't the current time), the link count will be
444 how many entries are directories (if >= 2). Otherwise, no knowledge 862 used to decide how many entries are directories (if >= 2).
445 of the number of subdirectories will be assumed. 863 Otherwise, no knowledge of the number of subdirectories will be
864 assumed.
446 865
447 Then entries will be sorted into likely directories (everything 866 Then entries will be sorted into likely directories a non-initial
448 without a non-initial dot currently) and likely non-directories 867 dot currently) and likely non-directories (see "aio_readdirx"). Then
449 (everything else). Then every entry plus an appended "/." will be 868 every entry plus an appended "/." will be "stat"'ed, likely
450 "stat"'ed, likely directories first. If that succeeds, it assumes 869 directories first, in order of their inode numbers. If that
451 that the entry is a directory or a symlink to directory (which will 870 succeeds, it assumes that the entry is a directory or a symlink to
452 be checked seperately). This is often faster than stat'ing the entry 871 directory (which will be checked separately). This is often faster
453 itself because filesystems might detect the type of the entry 872 than stat'ing the entry itself because filesystems might detect the
454 without reading the inode data (e.g. ext2fs filetype feature). 873 type of the entry without reading the inode data (e.g. ext2fs
874 filetype feature), even on systems that cannot return the filetype
875 information on readdir.
455 876
456 If the known number of directories (link count - 2) has been 877 If the known number of directories (link count - 2) has been
457 reached, the rest of the entries is assumed to be non-directories. 878 reached, the rest of the entries is assumed to be non-directories.
458 879
459 This only works with certainty on POSIX (= UNIX) filesystems, which 880 This only works with certainty on POSIX (= UNIX) filesystems, which
… …
461 882
462 It will also likely work on non-POSIX filesystems with reduced 883 It will also likely work on non-POSIX filesystems with reduced
463 efficiency as those tend to return 0 or 1 as link counts, which 884 efficiency as those tend to return 0 or 1 as link counts, which
464 disables the directory counting heuristic. 885 disables the directory counting heuristic.
465 886
887 aio_rmtree $pathname, $callback->($status)
888 Delete a directory tree starting (and including) $path, return the
889 status of the final "rmdir" only. This is a composite request that
890 uses "aio_scandir" to recurse into and rmdir directories, and unlink
891 everything else.
892
893 aio_sync $callback->($status)
894 Asynchronously call sync and call the callback when finished.
895
466 aio_fsync $fh, $callback->($status) 896 aio_fsync $fh, $callback->($status)
467 Asynchronously call fsync on the given filehandle and call the 897 Asynchronously call fsync on the given filehandle and call the
468 callback with the fsync result code. 898 callback with the fsync result code.
469 899
470 aio_fdatasync $fh, $callback->($status) 900 aio_fdatasync $fh, $callback->($status)
471 Asynchronously call fdatasync on the given filehandle and call the 901 Asynchronously call fdatasync on the given filehandle and call the
472 callback with the fdatasync result code. 902 callback with the fdatasync result code.
473 903
474 If this call isn't available because your OS lacks it or it couldn't 904 If this call isn't available because your OS lacks it or it couldn't
475 be detected, it will be emulated by calling "fsync" instead. 905 be detected, it will be emulated by calling "fsync" instead.
906
907 aio_syncfs $fh, $callback->($status)
908 Asynchronously call the syncfs syscall to sync the filesystem
909 associated to the given filehandle and call the callback with the
910 syncfs result code. If syncfs is not available, calls sync(), but
911 returns -1 and sets errno to "ENOSYS" nevertheless.
912
913 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
914 Sync the data portion of the file specified by $offset and $length
915 to disk (but NOT the metadata), by calling the Linux-specific
916 sync_file_range call. If sync_file_range is not available or it
917 returns ENOSYS, then fdatasync or fsync is being substituted.
918
919 $flags can be a combination of
920 "IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE",
921 "IO::AIO::SYNC_FILE_RANGE_WRITE" and
922 "IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER": refer to the sync_file_range
923 manpage for details.
924
925 aio_pathsync $pathname, $callback->($status)
926 This request tries to open, fsync and close the given path. This is
927 a composite request intended to sync directories after directory
928 operations (E.g. rename). This might not work on all operating
929 systems or have any specific effect, but usually it makes sure that
930 directory changes get written to disc. It works for anything that
931 can be opened for read-only, not just directories.
932
933 Future versions of this function might fall back to other methods
934 when "fsync" on the directory fails (such as calling "sync").
935
936 Passes 0 when everything went ok, and -1 on error.
937
938 aio_msync $scalar, $offset = 0, $length = undef, flags = 0,
939 $callback->($status)
940 This is a rather advanced IO::AIO call, which only works on
941 mmap(2)ed scalars (see the "IO::AIO::mmap" function, although it
942 also works on data scalars managed by the Sys::Mmap or Mmap modules,
943 note that the scalar must only be modified in-place while an aio
944 operation is pending on it).
945
946 It calls the "msync" function of your OS, if available, with the
947 memory area starting at $offset in the string and ending $length
948 bytes later. If $length is negative, counts from the end, and if
949 $length is "undef", then it goes till the end of the string. The
950 flags can be a combination of "IO::AIO::MS_ASYNC",
951 "IO::AIO::MS_INVALIDATE" and "IO::AIO::MS_SYNC".
952
953 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0,
954 $callback->($status)
955 This is a rather advanced IO::AIO call, which works best on
956 mmap(2)ed scalars.
957
958 It touches (reads or writes) all memory pages in the specified range
959 inside the scalar. All caveats and parameters are the same as for
960 "aio_msync", above, except for flags, which must be either 0 (which
961 reads all pages and ensures they are instantiated) or
962 "IO::AIO::MT_MODIFY", which modifies the memory pages (by reading
963 and writing an octet from it, which dirties the page).
964
965 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
966 This is a rather advanced IO::AIO call, which works best on
967 mmap(2)ed scalars.
968
969 It reads in all the pages of the underlying storage into memory (if
970 any) and locks them, so they are not getting swapped/paged out or
971 removed.
972
973 If $length is undefined, then the scalar will be locked till the
974 end.
975
976 On systems that do not implement "mlock", this function returns -1
977 and sets errno to "ENOSYS".
978
979 Note that the corresponding "munlock" is synchronous and is
980 documented under "MISCELLANEOUS FUNCTIONS".
981
982 Example: open a file, mmap and mlock it - both will be undone when
983 $data gets destroyed.
984
985 open my $fh, "<", $path or die "$path: $!";
986 my $data;
987 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
988 aio_mlock $data; # mlock in background
989
990 aio_mlockall $flags, $callback->($status)
991 Calls the "mlockall" function with the given $flags (a combination
992 of "IO::AIO::MCL_CURRENT" and "IO::AIO::MCL_FUTURE").
993
994 On systems that do not implement "mlockall", this function returns
995 -1 and sets errno to "ENOSYS".
996
997 Note that the corresponding "munlockall" is synchronous and is
998 documented under "MISCELLANEOUS FUNCTIONS".
999
1000 Example: asynchronously lock all current and future pages into
1001 memory.
1002
1003 aio_mlockall IO::AIO::MCL_FUTURE;
1004
1005 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
1006 Queries the extents of the given file (by calling the Linux "FIEMAP"
1007 ioctl, see <http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for
1008 details). If the ioctl is not available on your OS, then this
1009 request will fail with "ENOSYS".
1010
1011 $start is the starting offset to query extents for, $length is the
1012 size of the range to query - if it is "undef", then the whole file
1013 will be queried.
1014
1015 $flags is a combination of flags ("IO::AIO::FIEMAP_FLAG_SYNC" or
1016 "IO::AIO::FIEMAP_FLAG_XATTR" - "IO::AIO::FIEMAP_FLAGS_COMPAT" is
1017 also exported), and is normally 0 or "IO::AIO::FIEMAP_FLAG_SYNC" to
1018 query the data portion.
1019
1020 $count is the maximum number of extent records to return. If it is
1021 "undef", then IO::AIO queries all extents of the range. As a very
1022 special case, if it is 0, then the callback receives the number of
1023 extents instead of the extents themselves (which is unreliable, see
1024 below).
1025
1026 If an error occurs, the callback receives no arguments. The special
1027 "errno" value "IO::AIO::EBADR" is available to test for flag errors.
1028
1029 Otherwise, the callback receives an array reference with extent
1030 structures. Each extent structure is an array reference itself, with
1031 the following members:
1032
1033 [$logical, $physical, $length, $flags]
1034
1035 Flags is any combination of the following flag values (typically
1036 either 0 or "IO::AIO::FIEMAP_EXTENT_LAST" (1)):
1037
1038 "IO::AIO::FIEMAP_EXTENT_LAST", "IO::AIO::FIEMAP_EXTENT_UNKNOWN",
1039 "IO::AIO::FIEMAP_EXTENT_DELALLOC", "IO::AIO::FIEMAP_EXTENT_ENCODED",
1040 "IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED",
1041 "IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED",
1042 "IO::AIO::FIEMAP_EXTENT_DATA_INLINE",
1043 "IO::AIO::FIEMAP_EXTENT_DATA_TAIL",
1044 "IO::AIO::FIEMAP_EXTENT_UNWRITTEN", "IO::AIO::FIEMAP_EXTENT_MERGED"
1045 or "IO::AIO::FIEMAP_EXTENT_SHARED".
1046
1047 At the time of this writing (Linux 3.2), this requets is unreliable
1048 unless $count is "undef", as the kernel has all sorts of bugs
1049 preventing it to return all extents of a range for files with large
1050 number of extents. The code works around all these issues if $count
1051 is undef.
476 1052
477 aio_group $callback->(...) 1053 aio_group $callback->(...)
478 This is a very special aio request: Instead of doing something, it 1054 This is a very special aio request: Instead of doing something, it
479 is a container for other aio requests, which is useful if you want 1055 is a container for other aio requests, which is useful if you want
480 to bundle many requests into a single, composite, request with a 1056 to bundle many requests into a single, composite, request with a
… …
515 requests like sleep and file handle readable/writable, the overhead 1091 requests like sleep and file handle readable/writable, the overhead
516 this creates is immense (it blocks a thread for a long time) so do 1092 this creates is immense (it blocks a thread for a long time) so do
517 not use this function except to put your application under 1093 not use this function except to put your application under
518 artificial I/O pressure. 1094 artificial I/O pressure.
519 1095
1096 IO::AIO::WD - multiple working directories
1097 Your process only has one current working directory, which is used by
1098 all threads. This makes it hard to use relative paths (some other
1099 component could call "chdir" at any time, and it is hard to control when
1100 the path will be used by IO::AIO).
1101
1102 One solution for this is to always use absolute paths. This usually
1103 works, but can be quite slow (the kernel has to walk the whole path on
1104 every access), and can also be a hassle to implement.
1105
1106 Newer POSIX systems have a number of functions (openat, fdopendir,
1107 futimensat and so on) that make it possible to specify working
1108 directories per operation.
1109
1110 For portability, and because the clowns who "designed", or shall I
1111 write, perpetrated this new interface were obviously half-drunk, this
1112 abstraction cannot be perfect, though.
1113
1114 IO::AIO allows you to convert directory paths into a so-called
1115 IO::AIO::WD object. This object stores the canonicalised, absolute
1116 version of the path, and on systems that allow it, also a directory file
1117 descriptor.
1118
1119 Everywhere where a pathname is accepted by IO::AIO (e.g. in "aio_stat"
1120 or "aio_unlink"), one can specify an array reference with an IO::AIO::WD
1121 object and a pathname instead (or the IO::AIO::WD object alone, which
1122 gets interpreted as "[$wd, "."]"). If the pathname is absolute, the
1123 IO::AIO::WD object is ignored, otherwise the pathname is resolved
1124 relative to that IO::AIO::WD object.
1125
1126 For example, to get a wd object for /etc and then stat passwd inside,
1127 you would write:
1128
1129 aio_wd "/etc", sub {
1130 my $etcdir = shift;
1131
1132 # although $etcdir can be undef on error, there is generally no reason
1133 # to check for errors here, as aio_stat will fail with ENOENT
1134 # when $etcdir is undef.
1135
1136 aio_stat [$etcdir, "passwd"], sub {
1137 # yay
1138 };
1139 };
1140
1141 The fact that "aio_wd" is a request and not a normal function shows that
1142 creating an IO::AIO::WD object is itself a potentially blocking
1143 operation, which is why it is done asynchronously.
1144
1145 To stat the directory obtained with "aio_wd" above, one could write
1146 either of the following three request calls:
1147
1148 aio_lstat "/etc" , sub { ... # pathname as normal string
1149 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1150 aio_lstat $wd , sub { ... # shorthand for the previous
1151
1152 As with normal pathnames, IO::AIO keeps a copy of the working directory
1153 object and the pathname string, so you could write the following without
1154 causing any issues due to $path getting reused:
1155
1156 my $path = [$wd, undef];
1157
1158 for my $name (qw(abc def ghi)) {
1159 $path->[1] = $name;
1160 aio_stat $path, sub {
1161 # ...
1162 };
1163 }
1164
1165 There are some caveats: when directories get renamed (or deleted), the
1166 pathname string doesn't change, so will point to the new directory (or
1167 nowhere at all), while the directory fd, if available on the system,
1168 will still point to the original directory. Most functions accepting a
1169 pathname will use the directory fd on newer systems, and the string on
1170 older systems. Some functions (such as realpath) will always rely on the
1171 string form of the pathname.
1172
1173 So this functionality is mainly useful to get some protection against
1174 "chdir", to easily get an absolute path out of a relative path for
1175 future reference, and to speed up doing many operations in the same
1176 directory (e.g. when stat'ing all files in a directory).
1177
1178 The following functions implement this working directory abstraction:
1179
1180 aio_wd $pathname, $callback->($wd)
1181 Asynchonously canonicalise the given pathname and convert it to an
1182 IO::AIO::WD object representing it. If possible and supported on the
1183 system, also open a directory fd to speed up pathname resolution
1184 relative to this working directory.
1185
1186 If something goes wrong, then "undef" is passwd to the callback
1187 instead of a working directory object and $! is set appropriately.
1188 Since passing "undef" as working directory component of a pathname
1189 fails the request with "ENOENT", there is often no need for error
1190 checking in the "aio_wd" callback, as future requests using the
1191 value will fail in the expected way.
1192
1193 IO::AIO::CWD
1194 This is a compiletime constant (object) that represents the process
1195 current working directory.
1196
1197 Specifying this object as working directory object for a pathname is
1198 as if the pathname would be specified directly, without a directory
1199 object. For example, these calls are functionally identical:
1200
1201 aio_stat "somefile", sub { ... };
1202 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1203
1204 To recover the path associated with an IO::AIO::WD object, you can use
1205 "aio_realpath":
1206
1207 aio_realpath $wd, sub {
1208 warn "path is $_[0]\n";
1209 };
1210
1211 Currently, "aio_statvfs" always, and "aio_rename" and "aio_rmdir"
1212 sometimes, fall back to using an absolue path.
1213
520 IO::AIO::REQ CLASS 1214 IO::AIO::REQ CLASS
521 All non-aggregate "aio_*" functions return an object of this class when 1215 All non-aggregate "aio_*" functions return an object of this class when
522 called in non-void context. 1216 called in non-void context.
523 1217
524 cancel $req 1218 cancel $req
525 Cancels the request, if possible. Has the effect of skipping 1219 Cancels the request, if possible. Has the effect of skipping
526 execution when entering the execute state and skipping calling the 1220 execution when entering the execute state and skipping calling the
527 callback when entering the the result state, but will leave the 1221 callback when entering the the result state, but will leave the
528 request otherwise untouched. That means that requests that currently 1222 request otherwise untouched (with the exception of readdir). That
529 execute will not be stopped and resources held by the request will 1223 means that requests that currently execute will not be stopped and
530 not be freed prematurely. 1224 resources held by the request will not be freed prematurely.
531 1225
532 cb $req $callback->(...) 1226 cb $req $callback->(...)
533 Replace (or simply set) the callback registered to the request. 1227 Replace (or simply set) the callback registered to the request.
534 1228
535 IO::AIO::GRP CLASS 1229 IO::AIO::GRP CLASS
… …
562 }; 1256 };
563 1257
564 This makes it very easy to create composite requests (see the source of 1258 This makes it very easy to create composite requests (see the source of
565 "aio_move" for an application) that work and feel like simple requests. 1259 "aio_move" for an application) that work and feel like simple requests.
566 1260
567 * The IO::AIO::GRP objects will be cleaned up during calls to 1261 * The IO::AIO::GRP objects will be cleaned up during calls to
568 "IO::AIO::poll_cb", just like any other request. 1262 "IO::AIO::poll_cb", just like any other request.
1263
569 * They can be canceled like any other request. Canceling will cancel not 1264 * They can be canceled like any other request. Canceling will cancel
570 only the request itself, but also all requests it contains. 1265 not only the request itself, but also all requests it contains.
1266
571 * They can also can also be added to other IO::AIO::GRP objects. 1267 * They can also can also be added to other IO::AIO::GRP objects.
1268
572 * You must not add requests to a group from within the group callback 1269 * You must not add requests to a group from within the group callback
573 (or any later time). 1270 (or any later time).
574 1271
575 Their lifetime, simplified, looks like this: when they are empty, they 1272 Their lifetime, simplified, looks like this: when they are empty, they
576 will finish very quickly. If they contain only requests that are in the 1273 will finish very quickly. If they contain only requests that are in the
577 "done" state, they will also finish. Otherwise they will continue to 1274 "done" state, they will also finish. Otherwise they will continue to
578 exist. 1275 exist.
579 1276
580 That means after creating a group you have some time to add requests. 1277 That means after creating a group you have some time to add requests
581 And in the callbacks of those requests, you can add further requests to 1278 (precisely before the callback has been invoked, which is only done
582 the group. And only when all those requests have finished will the the 1279 within the "poll_cb"). And in the callbacks of those requests, you can
583 group itself finish. 1280 add further requests to the group. And only when all those requests have
1281 finished will the the group itself finish.
584 1282
585 add $grp ... 1283 add $grp ...
586 $grp->add (...) 1284 $grp->add (...)
587 Add one or more requests to the group. Any type of IO::AIO::REQ can 1285 Add one or more requests to the group. Any type of IO::AIO::REQ can
588 be added, including other groups, as long as you do not create 1286 be added, including other groups, as long as you do not create
… …
593 $grp->cancel_subs 1291 $grp->cancel_subs
594 Cancel all subrequests and clears any feeder, but not the group 1292 Cancel all subrequests and clears any feeder, but not the group
595 request itself. Useful when you queued a lot of events but got a 1293 request itself. Useful when you queued a lot of events but got a
596 result early. 1294 result early.
597 1295
1296 The group request will finish normally (you cannot add requests to
1297 the group).
1298
598 $grp->result (...) 1299 $grp->result (...)
599 Set the result value(s) that will be passed to the group callback 1300 Set the result value(s) that will be passed to the group callback
600 when all subrequests have finished and set thre groups errno to the 1301 when all subrequests have finished and set the groups errno to the
601 current value of errno (just like calling "errno" without an error 1302 current value of errno (just like calling "errno" without an error
602 number). By default, no argument will be passed and errno is zero. 1303 number). By default, no argument will be passed and errno is zero.
603 1304
604 $grp->errno ([$errno]) 1305 $grp->errno ([$errno])
605 Sets the group errno value to $errno, or the current value of errno 1306 Sets the group errno value to $errno, or the current value of errno
… …
616 Sets a feeder/generator on this group: every group can have an 1317 Sets a feeder/generator on this group: every group can have an
617 attached generator that generates requests if idle. The idea behind 1318 attached generator that generates requests if idle. The idea behind
618 this is that, although you could just queue as many requests as you 1319 this is that, although you could just queue as many requests as you
619 want in a group, this might starve other requests for a potentially 1320 want in a group, this might starve other requests for a potentially
620 long time. For example, "aio_scandir" might generate hundreds of 1321 long time. For example, "aio_scandir" might generate hundreds of
621 thousands "aio_stat" requests, delaying any later requests for a 1322 thousands of "aio_stat" requests, delaying any later requests for a
622 long time. 1323 long time.
623 1324
624 To avoid this, and allow incremental generation of requests, you can 1325 To avoid this, and allow incremental generation of requests, you can
625 instead a group and set a feeder on it that generates those 1326 instead a group and set a feeder on it that generates those
626 requests. The feed callback will be called whenever there are few 1327 requests. The feed callback will be called whenever there are few
… …
631 does not impose any limits). 1332 does not impose any limits).
632 1333
633 If the feed does not queue more requests when called, it will be 1334 If the feed does not queue more requests when called, it will be
634 automatically removed from the group. 1335 automatically removed from the group.
635 1336
636 If the feed limit is 0, it will be set to 2 automatically. 1337 If the feed limit is 0 when this method is called, it will be set to
1338 2 automatically.
637 1339
638 Example: 1340 Example:
639 1341
640 # stat all files in @files, but only ever use four aio requests concurrently: 1342 # stat all files in @files, but only ever use four aio requests concurrently:
641 1343
… …
652 Sets the feeder limit for the group: The feeder will be called 1354 Sets the feeder limit for the group: The feeder will be called
653 whenever the group contains less than this many requests. 1355 whenever the group contains less than this many requests.
654 1356
655 Setting the limit to 0 will pause the feeding process. 1357 Setting the limit to 0 will pause the feeding process.
656 1358
1359 The default value for the limit is 0, but note that setting a feeder
1360 automatically bumps it up to 2.
1361
657 SUPPORT FUNCTIONS 1362 SUPPORT FUNCTIONS
658 EVENT PROCESSING AND EVENT LOOP INTEGRATION 1363 EVENT PROCESSING AND EVENT LOOP INTEGRATION
659 $fileno = IO::AIO::poll_fileno 1364 $fileno = IO::AIO::poll_fileno
660 Return the *request result pipe file descriptor*. This filehandle 1365 Return the *request result pipe file descriptor*. This filehandle
661 must be polled for reading by some mechanism outside this module 1366 must be polled for reading by some mechanism outside this module
662 (e.g. Event or select, see below or the SYNOPSIS). If the pipe 1367 (e.g. EV, Glib, select and so on, see below or the SYNOPSIS). If the
663 becomes readable you have to call "poll_cb" to check the results. 1368 pipe becomes readable you have to call "poll_cb" to check the
1369 results.
664 1370
665 See "poll_cb" for an example. 1371 See "poll_cb" for an example.
666 1372
667 IO::AIO::poll_cb 1373 IO::AIO::poll_cb
668 Process some outstanding events on the result pipe. You have to call 1374 Process some requests that have reached the result phase (i.e. they
669 this regularly. Returns the number of events processed. Returns 1375 have been executed but the results are not yet reported). You have
1376 to call this "regularly" to finish outstanding requests.
1377
1378 Returns 0 if all events could be processed (or there were no events
1379 to process), or -1 if it returned earlier for whatever reason.
670 immediately when no events are outstanding. The amount of events 1380 Returns immediately when no events are outstanding. The amount of
671 processed depends on the settings of "IO::AIO::max_poll_req" and 1381 events processed depends on the settings of "IO::AIO::max_poll_req",
672 "IO::AIO::max_poll_time". 1382 "IO::AIO::max_poll_time" and "IO::AIO::max_outstanding".
673 1383
674 If not all requests were processed for whatever reason, the 1384 If not all requests were processed for whatever reason, the poll
675 filehandle will still be ready when "poll_cb" returns. 1385 file descriptor will still be ready when "poll_cb" returns, so
1386 normally you don't have to do anything special to have it called
1387 later.
1388
1389 Apart from calling "IO::AIO::poll_cb" when the event filehandle
1390 becomes ready, it can be beneficial to call this function from loops
1391 which submit a lot of requests, to make sure the results get
1392 processed when they become available and not just when the loop is
1393 finished and the event loop takes over again. This function returns
1394 very fast when there are no outstanding requests.
676 1395
677 Example: Install an Event watcher that automatically calls 1396 Example: Install an Event watcher that automatically calls
678 IO::AIO::poll_cb with high priority: 1397 IO::AIO::poll_cb with high priority (more examples can be found in
1398 the SYNOPSIS section, at the top of this document):
679 1399
680 Event->io (fd => IO::AIO::poll_fileno, 1400 Event->io (fd => IO::AIO::poll_fileno,
681 poll => 'r', async => 1, 1401 poll => 'r', async => 1,
682 cb => \&IO::AIO::poll_cb); 1402 cb => \&IO::AIO::poll_cb);
1403
1404 IO::AIO::poll_wait
1405 Wait until either at least one request is in the result phase or no
1406 requests are outstanding anymore.
1407
1408 This is useful if you want to synchronously wait for some requests
1409 to become ready, without actually handling them.
1410
1411 See "nreqs" for an example.
1412
1413 IO::AIO::poll
1414 Waits until some requests have been handled.
1415
1416 Returns the number of requests processed, but is otherwise strictly
1417 equivalent to:
1418
1419 IO::AIO::poll_wait, IO::AIO::poll_cb
1420
1421 IO::AIO::flush
1422 Wait till all outstanding AIO requests have been handled.
1423
1424 Strictly equivalent to:
1425
1426 IO::AIO::poll_wait, IO::AIO::poll_cb
1427 while IO::AIO::nreqs;
683 1428
684 IO::AIO::max_poll_reqs $nreqs 1429 IO::AIO::max_poll_reqs $nreqs
685 IO::AIO::max_poll_time $seconds 1430 IO::AIO::max_poll_time $seconds
686 These set the maximum number of requests (default 0, meaning 1431 These set the maximum number of requests (default 0, meaning
687 infinity) that are being processed by "IO::AIO::poll_cb" in one 1432 infinity) that are being processed by "IO::AIO::poll_cb" in one
… …
711 # use a low priority so other tasks have priority 1456 # use a low priority so other tasks have priority
712 Event->io (fd => IO::AIO::poll_fileno, 1457 Event->io (fd => IO::AIO::poll_fileno,
713 poll => 'r', nice => 1, 1458 poll => 'r', nice => 1,
714 cb => &IO::AIO::poll_cb); 1459 cb => &IO::AIO::poll_cb);
715 1460
716 IO::AIO::poll_wait
717 If there are any outstanding requests and none of them in the result
718 phase, wait till the result filehandle becomes ready for reading
719 (simply does a "select" on the filehandle. This is useful if you
720 want to synchronously wait for some requests to finish).
721
722 See "nreqs" for an example.
723
724 IO::AIO::poll
725 Waits until some requests have been handled.
726
727 Returns the number of requests processed, but is otherwise strictly
728 equivalent to:
729
730 IO::AIO::poll_wait, IO::AIO::poll_cb
731
732 IO::AIO::flush
733 Wait till all outstanding AIO requests have been handled.
734
735 Strictly equivalent to:
736
737 IO::AIO::poll_wait, IO::AIO::poll_cb
738 while IO::AIO::nreqs;
739
740 CONTROLLING THE NUMBER OF THREADS 1461 CONTROLLING THE NUMBER OF THREADS
741 IO::AIO::min_parallel $nthreads 1462 IO::AIO::min_parallel $nthreads
742 Set the minimum number of AIO threads to $nthreads. The current 1463 Set the minimum number of AIO threads to $nthreads. The current
743 default is 8, which means eight asynchronous operations can execute 1464 default is 8, which means eight asynchronous operations can execute
744 concurrently at any one time (the number of outstanding requests, 1465 concurrently at any one time (the number of outstanding requests,
… …
773 1494
774 Under normal circumstances you don't need to call this function. 1495 Under normal circumstances you don't need to call this function.
775 1496
776 IO::AIO::max_idle $nthreads 1497 IO::AIO::max_idle $nthreads
777 Limit the number of threads (default: 4) that are allowed to idle 1498 Limit the number of threads (default: 4) that are allowed to idle
778 (i.e., threads that did not get a request to process within 10 1499 (i.e., threads that did not get a request to process within the idle
779 seconds). That means if a thread becomes idle while $nthreads other 1500 timeout (default: 10 seconds). That means if a thread becomes idle
780 threads are also idle, it will free its resources and exit. 1501 while $nthreads other threads are also idle, it will free its
1502 resources and exit.
781 1503
782 This is useful when you allow a large number of threads (e.g. 100 or 1504 This is useful when you allow a large number of threads (e.g. 100 or
783 1000) to allow for extremely high load situations, but want to free 1505 1000) to allow for extremely high load situations, but want to free
784 resources under normal circumstances (1000 threads can easily 1506 resources under normal circumstances (1000 threads can easily
785 consume 30MB of RAM). 1507 consume 30MB of RAM).
786 1508
787 The default is probably ok in most situations, especially if thread 1509 The default is probably ok in most situations, especially if thread
788 creation is fast. If thread creation is very slow on your system you 1510 creation is fast. If thread creation is very slow on your system you
789 might want to use larger values. 1511 might want to use larger values.
790 1512
1513 IO::AIO::idle_timeout $seconds
1514 Sets the minimum idle timeout (default 10) after which worker
1515 threads are allowed to exit. SEe "IO::AIO::max_idle".
1516
791 $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 1517 IO::AIO::max_outstanding $maxreqs
1518 Sets the maximum number of outstanding requests to $nreqs. If you do
1519 queue up more than this number of requests, the next call to
1520 "IO::AIO::poll_cb" (and other functions calling "poll_cb", such as
1521 "IO::AIO::flush" or "IO::AIO::poll") will block until the limit is
1522 no longer exceeded.
1523
1524 In other words, this setting does not enforce a queue limit, but can
1525 be used to make poll functions block if the limit is exceeded.
1526
792 This is a very bad function to use in interactive programs because 1527 This is a very bad function to use in interactive programs because
793 it blocks, and a bad way to reduce concurrency because it is 1528 it blocks, and a bad way to reduce concurrency because it is
794 inexact: Better use an "aio_group" together with a feed callback. 1529 inexact: Better use an "aio_group" together with a feed callback.
795 1530
796 Sets the maximum number of outstanding requests to $nreqs. If you to 1531 Its main use is in scripts without an event loop - when you want to
797 queue up more than this number of requests, the next call to the 1532 stat a lot of files, you can write somehting like this:
798 "poll_cb" (and "poll_some" and other functions calling "poll_cb")
799 function will block until the limit is no longer exceeded.
800 1533
801 The default value is very large, so there is no practical limit on 1534 IO::AIO::max_outstanding 32;
1535
1536 for my $path (...) {
1537 aio_stat $path , ...;
1538 IO::AIO::poll_cb;
1539 }
1540
1541 IO::AIO::flush;
1542
1543 The call to "poll_cb" inside the loop will normally return
1544 instantly, but as soon as more thna 32 reqeusts are in-flight, it
1545 will block until some requests have been handled. This keeps the
1546 loop from pushing a large number of "aio_stat" requests onto the
1547 queue.
1548
1549 The default value for "max_outstanding" is very large, so there is
802 the number of outstanding requests. 1550 no practical limit on the number of outstanding requests.
803
804 You can still queue as many requests as you want. Therefore,
805 "max_oustsanding" is mainly useful in simple scripts (with low
806 values) or as a stop gap to shield against fatal memory overflow
807 (with large values).
808 1551
809 STATISTICAL INFORMATION 1552 STATISTICAL INFORMATION
810 IO::AIO::nreqs 1553 IO::AIO::nreqs
811 Returns the number of requests currently in the ready, execute or 1554 Returns the number of requests currently in the ready, execute or
812 pending states (i.e. for which their callback has not been invoked 1555 pending states (i.e. for which their callback has not been invoked
… …
823 1566
824 IO::AIO::npending 1567 IO::AIO::npending
825 Returns the number of requests currently in the pending state 1568 Returns the number of requests currently in the pending state
826 (executed, but not yet processed by poll_cb). 1569 (executed, but not yet processed by poll_cb).
827 1570
1571 MISCELLANEOUS FUNCTIONS
1572 IO::AIO implements some functions that are useful when you want to use
1573 some "Advanced I/O" function not available to in Perl, without going the
1574 "Asynchronous I/O" route. Many of these have an asynchronous "aio_*"
1575 counterpart.
1576
1577 IO::AIO::sendfile $ofh, $ifh, $offset, $count
1578 Calls the "eio_sendfile_sync" function, which is like
1579 "aio_sendfile", but is blocking (this makes most sense if you know
1580 the input data is likely cached already and the output filehandle is
1581 set to non-blocking operations).
1582
1583 Returns the number of bytes copied, or -1 on error.
1584
1585 IO::AIO::fadvise $fh, $offset, $len, $advice
1586 Simply calls the "posix_fadvise" function (see its manpage for
1587 details). The following advice constants are available:
1588 "IO::AIO::FADV_NORMAL", "IO::AIO::FADV_SEQUENTIAL",
1589 "IO::AIO::FADV_RANDOM", "IO::AIO::FADV_NOREUSE",
1590 "IO::AIO::FADV_WILLNEED", "IO::AIO::FADV_DONTNEED".
1591
1592 On systems that do not implement "posix_fadvise", this function
1593 returns ENOSYS, otherwise the return value of "posix_fadvise".
1594
1595 IO::AIO::madvise $scalar, $offset, $len, $advice
1596 Simply calls the "posix_madvise" function (see its manpage for
1597 details). The following advice constants are available:
1598 "IO::AIO::MADV_NORMAL", "IO::AIO::MADV_SEQUENTIAL",
1599 "IO::AIO::MADV_RANDOM", "IO::AIO::MADV_WILLNEED",
1600 "IO::AIO::MADV_DONTNEED".
1601
1602 On systems that do not implement "posix_madvise", this function
1603 returns ENOSYS, otherwise the return value of "posix_madvise".
1604
1605 IO::AIO::mprotect $scalar, $offset, $len, $protect
1606 Simply calls the "mprotect" function on the preferably AIO::mmap'ed
1607 $scalar (see its manpage for details). The following protect
1608 constants are available: "IO::AIO::PROT_NONE", "IO::AIO::PROT_READ",
1609 "IO::AIO::PROT_WRITE", "IO::AIO::PROT_EXEC".
1610
1611 On systems that do not implement "mprotect", this function returns
1612 ENOSYS, otherwise the return value of "mprotect".
1613
1614 IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1615 Memory-maps a file (or anonymous memory range) and attaches it to
1616 the given $scalar, which will act like a string scalar. Returns true
1617 on success, and false otherwise.
1618
1619 The only operations allowed on the scalar are "substr"/"vec" that
1620 don't change the string length, and most read-only operations such
1621 as copying it or searching it with regexes and so on.
1622
1623 Anything else is unsafe and will, at best, result in memory leaks.
1624
1625 The memory map associated with the $scalar is automatically removed
1626 when the $scalar is destroyed, or when the "IO::AIO::mmap" or
1627 "IO::AIO::munmap" functions are called.
1628
1629 This calls the "mmap"(2) function internally. See your system's
1630 manual page for details on the $length, $prot and $flags parameters.
1631
1632 The $length must be larger than zero and smaller than the actual
1633 filesize.
1634
1635 $prot is a combination of "IO::AIO::PROT_NONE",
1636 "IO::AIO::PROT_EXEC", "IO::AIO::PROT_READ" and/or
1637 "IO::AIO::PROT_WRITE",
1638
1639 $flags can be a combination of "IO::AIO::MAP_SHARED" or
1640 "IO::AIO::MAP_PRIVATE", or a number of system-specific flags (when
1641 not available, the are 0): "IO::AIO::MAP_ANONYMOUS" (which is set to
1642 "MAP_ANON" if your system only provides this constant),
1643 "IO::AIO::MAP_HUGETLB", "IO::AIO::MAP_LOCKED",
1644 "IO::AIO::MAP_NORESERVE", "IO::AIO::MAP_POPULATE",
1645 "IO::AIO::MAP_NONBLOCK", "IO::AIO::MAP_FIXED",
1646 "IO::AIO::MAP_GROWSDOWN", "IO::AIO::MAP_32BIT",
1647 "IO::AIO::MAP_HUGETLB" or "IO::AIO::MAP_STACK".
1648
1649 If $fh is "undef", then a file descriptor of -1 is passed.
1650
1651 $offset is the offset from the start of the file - it generally must
1652 be a multiple of "IO::AIO::PAGESIZE" and defaults to 0.
1653
1654 Example:
1655
1656 use Digest::MD5;
1657 use IO::AIO;
1658
1659 open my $fh, "<verybigfile"
1660 or die "$!";
1661
1662 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1663 or die "verybigfile: $!";
1664
1665 my $fast_md5 = md5 $data;
1666
1667 IO::AIO::munmap $scalar
1668 Removes a previous mmap and undefines the $scalar.
1669
1670 IO::AIO::munlock $scalar, $offset = 0, $length = undef
1671 Calls the "munlock" function, undoing the effects of a previous
1672 "aio_mlock" call (see its description for details).
1673
1674 IO::AIO::munlockall
1675 Calls the "munlockall" function.
1676
1677 On systems that do not implement "munlockall", this function returns
1678 ENOSYS, otherwise the return value of "munlockall".
1679
1680 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
1681 Calls the GNU/Linux splice(2) syscall, if available. If $r_off or
1682 $w_off are "undef", then "NULL" is passed for these, otherwise they
1683 should be the file offset.
1684
1685 $r_fh and $w_fh should not refer to the same file, as splice might
1686 silently corrupt the data in this case.
1687
1688 The following symbol flag values are available:
1689 "IO::AIO::SPLICE_F_MOVE", "IO::AIO::SPLICE_F_NONBLOCK",
1690 "IO::AIO::SPLICE_F_MORE" and "IO::AIO::SPLICE_F_GIFT".
1691
1692 See the splice(2) manpage for details.
1693
1694 IO::AIO::tee $r_fh, $w_fh, $length, $flags
1695 Calls the GNU/Linux tee(2) syscall, see its manpage and the
1696 description for "IO::AIO::splice" above for details.
1697
1698 $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
1699 Attempts to query or change the pipe buffer size. Obviously works
1700 only on pipes, and currently works only on GNU/Linux systems, and
1701 fails with -1/"ENOSYS" everywhere else. If anybody knows how to
1702 influence pipe buffer size on other systems, drop me a note.
1703
1704 ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
1705 This is a direct interface to the Linux pipe2(2) system call. If
1706 $flags is missing or 0, then this should be the same as a call to
1707 perl's built-in "pipe" function and create a new pipe, and works on
1708 systems that lack the pipe2 syscall. On win32, this case invokes
1709 "_pipe (..., 4096, O_BINARY)".
1710
1711 If $flags is non-zero, it tries to invoke the pipe2 system call with
1712 the given flags (Linux 2.6.27, glibc 2.9).
1713
1714 On success, the read and write file handles are returned.
1715
1716 On error, nothing will be returned. If the pipe2 syscall is missing
1717 and $flags is non-zero, fails with "ENOSYS".
1718
1719 Please refer to pipe2(2) for more info on the $flags, but at the
1720 time of this writing, "IO::AIO::O_CLOEXEC", "IO::AIO::O_NONBLOCK"
1721 and "IO::AIO::O_DIRECT" (Linux 3.4, for packet-based pipes) were
1722 supported.
1723
1724EVENT LOOP INTEGRATION
1725 It is recommended to use AnyEvent::AIO to integrate IO::AIO
1726 automatically into many event loops:
1727
1728 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
1729 use AnyEvent::AIO;
1730
1731 You can also integrate IO::AIO manually into many event loops, here are
1732 some examples of how to do this:
1733
1734 # EV integration
1735 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
1736
1737 # Event integration
1738 Event->io (fd => IO::AIO::poll_fileno,
1739 poll => 'r',
1740 cb => \&IO::AIO::poll_cb);
1741
1742 # Glib/Gtk2 integration
1743 add_watch Glib::IO IO::AIO::poll_fileno,
1744 in => sub { IO::AIO::poll_cb; 1 };
1745
1746 # Tk integration
1747 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
1748 readable => \&IO::AIO::poll_cb);
1749
1750 # Danga::Socket integration
1751 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
1752 \&IO::AIO::poll_cb);
1753
828 FORK BEHAVIOUR 1754 FORK BEHAVIOUR
829 This module should do "the right thing" when the process using it forks: 1755 Usage of pthreads in a program changes the semantics of fork
1756 considerably. Specifically, only async-safe functions can be called
1757 after fork. Perl doesn't know about this, so in general, you cannot call
1758 fork with defined behaviour in perl if pthreads are involved. IO::AIO
1759 uses pthreads, so this applies, but many other extensions and (for
1760 inexplicable reasons) perl itself often is linked against pthreads, so
1761 this limitation applies to quite a lot of perls.
830 1762
831 Before the fork, IO::AIO enters a quiescent state where no requests can 1763 This module no longer tries to fight your OS, or POSIX. That means
832 be added in other threads and no results will be processed. After the 1764 IO::AIO only works in the process that loaded it. Forking is fully
833 fork the parent simply leaves the quiescent state and continues 1765 supported, but using IO::AIO in the child is not.
834 request/result processing, while the child frees the request/result
835 queue (so that the requests started before the fork will only be handled
836 in the parent). Threads will be started on demand until the limit set in
837 the parent process has been reached again.
838 1766
839 In short: the parent will, after a short pause, continue as if fork had 1767 You might get around by not *using* IO::AIO before (or after) forking.
840 not been called, while the child will act as if IO::AIO has not been 1768 You could also try to call the IO::AIO::reinit function in the child:
841 used yet. 1769
1770 IO::AIO::reinit
1771 Abandons all current requests and I/O threads and simply
1772 reinitialises all data structures. This is not an operation
1773 supported by any standards, but happens to work on GNU/Linux and
1774 some newer BSD systems.
1775
1776 The only reasonable use for this function is to call it after
1777 forking, if "IO::AIO" was used in the parent. Calling it while
1778 IO::AIO is active in the process will result in undefined behaviour.
1779 Calling it at any time will also result in any undefined (by POSIX)
1780 behaviour.
842 1781
843 MEMORY USAGE 1782 MEMORY USAGE
844 Per-request usage: 1783 Per-request usage:
845 1784
846 Each aio request uses - depending on your architecture - around 100-200 1785 Each aio request uses - depending on your architecture - around 100-200
847 bytes of memory. In addition, stat requests need a stat buffer (possibly 1786 bytes of memory. In addition, stat requests need a stat buffer (possibly
848 a few hundred bytes), readdir requires a result buffer and so on. Perl 1787 a few hundred bytes), readdir requires a result buffer and so on. Perl
849 scalars and other data passed into aio requests will also be locked and 1788 scalars and other data passed into aio requests will also be locked and
850 will consume memory till the request has entered the done state. 1789 will consume memory till the request has entered the done state.
851 1790
852 This is now awfully much, so queuing lots of requests is not usually a 1791 This is not awfully much, so queuing lots of requests is not usually a
853 problem. 1792 problem.
854 1793
855 Per-thread usage: 1794 Per-thread usage:
856 1795
857 In the execution phase, some aio requests require more memory for 1796 In the execution phase, some aio requests require more memory for
… …
860 1799
861KNOWN BUGS 1800KNOWN BUGS
862 Known bugs will be fixed in the next release. 1801 Known bugs will be fixed in the next release.
863 1802
864SEE ALSO 1803SEE ALSO
865 Coro::AIO. 1804 AnyEvent::AIO for easy integration into event loops, Coro::AIO for a
1805 more natural syntax.
866 1806
867AUTHOR 1807AUTHOR
868 Marc Lehmann <schmorp@schmorp.de> 1808 Marc Lehmann <schmorp@schmorp.de>
869 http://home.schmorp.de/ 1809 http://home.schmorp.de/
870 1810

Diff Legend

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