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

Comparing IO-AIO/README (file contents):
Revision 1.20 by root, Tue Oct 31 00:45:41 2006 UTC vs.
Revision 1.64 by root, Wed Apr 3 03:03:53 2019 UTC

1NAME 1NAME
2 IO::AIO - Asynchronous Input/Output 2 IO::AIO - Asynchronous/Advanced Input/Output
3 3
4SYNOPSIS 4SYNOPSIS
5 use IO::AIO; 5 use IO::AIO;
6 6
7 aio_open "/etc/passwd", O_RDONLY, 0, sub { 7 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
8 my ($fh) = @_; 8 my $fh = shift
9 or die "/etc/passwd: $!";
9 ... 10 ...
10 }; 11 };
11 12
12 aio_unlink "/tmp/file", sub { }; 13 aio_unlink "/tmp/file", sub { };
13 14
… …
23 $req->cancel; # cancel request if still in queue 24 $req->cancel; # cancel request if still in queue
24 25
25 my $grp = aio_group sub { print "all stats done\n" }; 26 my $grp = aio_group sub { print "all stats done\n" };
26 add $grp aio_stat "..." for ...; 27 add $grp aio_stat "..." for ...;
27 28
28 # AnyEvent integration
29 open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!";
30 my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb });
31
32 # Event integration
33 Event->io (fd => IO::AIO::poll_fileno,
34 poll => 'r',
35 cb => \&IO::AIO::poll_cb);
36
37 # Glib/Gtk2 integration
38 add_watch Glib::IO IO::AIO::poll_fileno,
39 in => sub { IO::AIO::poll_cb; 1 };
40
41 # Tk integration
42 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
43 readable => \&IO::AIO::poll_cb);
44
45 # Danga::Socket integration
46 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
47 \&IO::AIO::poll_cb);
48
49DESCRIPTION 29DESCRIPTION
50 This module implements asynchronous I/O using whatever means your 30 This module implements asynchronous I/O using whatever means your
51 operating system supports. 31 operating system supports. It is implemented as an interface to "libeio"
32 (<http://software.schmorp.de/pkg/libeio.html>).
52 33
53 Asynchronous means that operations that can normally block your program 34 Asynchronous means that operations that can normally block your program
54 (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
55 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
56 extremely useful for programs that need to stay interactive even when 37 extremely useful for programs that need to stay interactive even when
… …
60 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
61 operations concurrently. 42 operations concurrently.
62 43
63 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
64 sockets), using these functions on file descriptors that support 45 sockets), using these functions on file descriptors that support
65 nonblocking operation (again, sockets, pipes etc.) is very inefficient 46 nonblocking operation (again, sockets, pipes etc.) is very inefficient.
66 or might not work (aio_read fails on sockets/pipes/fifos). Use an event
67 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
68 into such an event loop itself. 48 naturally fit into such an event loop itself.
69 49
70 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
71 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
72 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
73 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
… …
75 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
76 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
77 aio_write, so the remaining functionality would have to be implemented 57 aio_write, so the remaining functionality would have to be implemented
78 using threads anyway. 58 using threads anyway.
79 59
60 In addition to asynchronous I/O, this module also exports some rather
61 arcane interfaces, such as "madvise" or linux's "splice" system call,
62 which is why the "A" in "AIO" can also mean *advanced*.
63
80 Although the module will work with in the presence of other (Perl-) 64 Although the module will work in the presence of other (Perl-) threads,
81 threads, it is currently not reentrant in any way, so use appropriate 65 it is currently not reentrant in any way, so use appropriate locking
82 locking yourself, always call "poll_cb" from within the same thread, or 66 yourself, always call "poll_cb" from within the same thread, or never
83 never call "poll_cb" (or other "aio_" functions) recursively. 67 call "poll_cb" (or other "aio_" functions) recursively.
84 68
85 EXAMPLE 69 EXAMPLE
86 This is a simple example that uses the Event module and loads 70 This is a simple example that uses the EV module and loads /etc/passwd
87 /etc/passwd asynchronously: 71 asynchronously:
88 72
89 use Fcntl;
90 use Event; 73 use EV;
91 use IO::AIO; 74 use IO::AIO;
92 75
93 # register the IO::AIO callback with Event 76 # register the IO::AIO callback with EV
94 Event->io (fd => IO::AIO::poll_fileno, 77 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
95 poll => 'r',
96 cb => \&IO::AIO::poll_cb);
97 78
98 # queue the request to open /etc/passwd 79 # queue the request to open /etc/passwd
99 aio_open "/etc/passwd", O_RDONLY, 0, sub { 80 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
100 my $fh = $_[0] 81 my $fh = shift
101 or die "error while opening: $!"; 82 or die "error while opening: $!";
102 83
103 # stat'ing filehandles is generally non-blocking 84 # stat'ing filehandles is generally non-blocking
104 my $size = -s $fh; 85 my $size = -s $fh;
105 86
… …
113 94
114 # file contents now in $contents 95 # file contents now in $contents
115 print $contents; 96 print $contents;
116 97
117 # exit event loop and program 98 # exit event loop and program
118 Event::unloop; 99 EV::break;
119 }; 100 };
120 }; 101 };
121 102
122 # possibly queue up other requests, or open GUI windows, 103 # possibly queue up other requests, or open GUI windows,
123 # check for sockets etc. etc. 104 # check for sockets etc. etc.
124 105
125 # process events as long as there are some: 106 # process events as long as there are some:
126 Event::loop; 107 EV::run;
127 108
128REQUEST ANATOMY AND LIFETIME 109REQUEST ANATOMY AND LIFETIME
129 Every "aio_*" function creates a request. which is a C data structure 110 Every "aio_*" function creates a request. which is a C data structure
130 not directly visible to Perl. 111 not directly visible to Perl.
131 112
… …
167 anymore (except possibly for the Perl object, but its connection to 148 anymore (except possibly for the Perl object, but its connection to
168 the actual aio request is severed and calling its methods will 149 the actual aio request is severed and calling its methods will
169 either do nothing or result in a runtime error). 150 either do nothing or result in a runtime error).
170 151
171FUNCTIONS 152FUNCTIONS
172 AIO REQUEST FUNCTIONS 153 QUICK OVERVIEW
154 This section simply lists the prototypes most of the functions for quick
155 reference. See the following sections for function-by-function
156 documentation.
157
158 aio_wd $pathname, $callback->($wd)
159 aio_open $pathname, $flags, $mode, $callback->($fh)
160 aio_close $fh, $callback->($status)
161 aio_seek $fh,$offset,$whence, $callback->($offs)
162 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
163 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
164 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
165 aio_readahead $fh,$offset,$length, $callback->($retval)
166 aio_stat $fh_or_path, $callback->($status)
167 aio_lstat $fh, $callback->($status)
168 aio_statvfs $fh_or_path, $callback->($statvfs)
169 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
170 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
171 aio_chmod $fh_or_path, $mode, $callback->($status)
172 aio_truncate $fh_or_path, $offset, $callback->($status)
173 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
174 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
175 aio_unlink $pathname, $callback->($status)
176 aio_mknod $pathname, $mode, $dev, $callback->($status)
177 aio_link $srcpath, $dstpath, $callback->($status)
178 aio_symlink $srcpath, $dstpath, $callback->($status)
179 aio_readlink $pathname, $callback->($link)
180 aio_realpath $pathname, $callback->($path)
181 aio_rename $srcpath, $dstpath, $callback->($status)
182 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
183 aio_mkdir $pathname, $mode, $callback->($status)
184 aio_rmdir $pathname, $callback->($status)
185 aio_readdir $pathname, $callback->($entries)
186 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
187 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
188 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
189 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
190 aio_load $pathname, $data, $callback->($status)
191 aio_copy $srcpath, $dstpath, $callback->($status)
192 aio_move $srcpath, $dstpath, $callback->($status)
193 aio_rmtree $pathname, $callback->($status)
194 aio_fcntl $fh, $cmd, $arg, $callback->($status)
195 aio_ioctl $fh, $request, $buf, $callback->($status)
196 aio_sync $callback->($status)
197 aio_syncfs $fh, $callback->($status)
198 aio_fsync $fh, $callback->($status)
199 aio_fdatasync $fh, $callback->($status)
200 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
201 aio_pathsync $pathname, $callback->($status)
202 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status)
203 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
204 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
205 aio_mlockall $flags, $callback->($status)
206 aio_group $callback->(...)
207 aio_nop $callback->()
208
209 $prev_pri = aioreq_pri [$pri]
210 aioreq_nice $pri_adjust
211
212 IO::AIO::poll_wait
213 IO::AIO::poll_cb
214 IO::AIO::poll
215 IO::AIO::flush
216 IO::AIO::max_poll_reqs $nreqs
217 IO::AIO::max_poll_time $seconds
218 IO::AIO::min_parallel $nthreads
219 IO::AIO::max_parallel $nthreads
220 IO::AIO::max_idle $nthreads
221 IO::AIO::idle_timeout $seconds
222 IO::AIO::max_outstanding $maxreqs
223 IO::AIO::nreqs
224 IO::AIO::nready
225 IO::AIO::npending
226 IO::AIO::reinit
227
228 $nfd = IO::AIO::get_fdlimit [EXPERIMENTAL]
229 IO::AIO::min_fdlimit $nfd [EXPERIMENTAL]
230
231 IO::AIO::sendfile $ofh, $ifh, $offset, $count
232 IO::AIO::fadvise $fh, $offset, $len, $advice
233
234 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]]
235 IO::AIO::munmap $scalar
236 IO::AIO::mremap $scalar, $new_length, $flags[, $new_address]
237 IO::AIO::madvise $scalar, $offset, $length, $advice
238 IO::AIO::mprotect $scalar, $offset, $length, $protect
239 IO::AIO::munlock $scalar, $offset = 0, $length = undef
240 IO::AIO::munlockall
241
242 # stat extensions
243 $counter = IO::AIO::st_gen
244 $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime
245 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
246 $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec
247 $seconds = IO::AIO::st_btimesec
248 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
249
250 # very much unportable syscalls
251 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
252 IO::AIO::tee $r_fh, $w_fh, $length, $flags
253 $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
254 ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
255 $fh = IO::AIO::memfd_create $pathname[, $flags]
256 $fh = IO::AIO::eventfd [$initval, [$flags]]
257 $fh = IO::AIO::timerfd_create $clockid[, $flags]
258 ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value
259 ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
260
261 API NOTES
173 All the "aio_*" calls are more or less thin wrappers around the syscall 262 All the "aio_*" calls are more or less thin wrappers around the syscall
174 with the same name (sans "aio_"). The arguments are similar or 263 with the same name (sans "aio_"). The arguments are similar or
175 identical, and they all accept an additional (and optional) $callback 264 identical, and they all accept an additional (and optional) $callback
176 argument which must be a code reference. This code reference will get 265 argument which must be a code reference. This code reference will be
177 called with the syscall return code (e.g. most syscalls return -1 on 266 called after the syscall has been executed in an asynchronous fashion.
178 error, unlike perl, which usually delivers "false") as it's sole 267 The results of the request will be passed as arguments to the callback
179 argument when the given syscall has been executed asynchronously. 268 (and, if an error occured, in $!) - for most requests the syscall return
269 code (e.g. most syscalls return -1 on error, unlike perl, which usually
270 delivers "false").
271
272 Some requests (such as "aio_readdir") pass the actual results and
273 communicate failures by passing "undef".
180 274
181 All functions expecting a filehandle keep a copy of the filehandle 275 All functions expecting a filehandle keep a copy of the filehandle
182 internally until the request has finished. 276 internally until the request has finished.
183 277
184 All functions return request objects of type IO::AIO::REQ that allow 278 All functions return request objects of type IO::AIO::REQ that allow
185 further manipulation of those requests while they are in-flight. 279 further manipulation of those requests while they are in-flight.
186 280
187 The pathnames you pass to these routines *must* be absolute and encoded 281 The pathnames you pass to these routines *should* be absolute. The
188 as octets. The reason for the former is that at the time the request is 282 reason for this is that at the time the request is being executed, the
189 being executed, the current working directory could have changed. 283 current working directory could have changed. Alternatively, you can
190 Alternatively, you can make sure that you never change the current 284 make sure that you never change the current working directory anywhere
191 working directory anywhere in the program and then use relative paths. 285 in the program and then use relative paths. You can also take advantage
286 of IO::AIOs working directory abstraction, that lets you specify paths
287 relative to some previously-opened "working directory object" - see the
288 description of the "IO::AIO::WD" class later in this document.
192 289
193 To encode pathnames as octets, either make sure you either: a) always 290 To encode pathnames as octets, either make sure you either: a) always
194 pass in filenames you got from outside (command line, readdir etc.) 291 pass in filenames you got from outside (command line, readdir etc.)
195 without tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module 292 without tinkering, b) are in your native filesystem encoding, c) use the
196 and encode your pathnames to the locale (or other) encoding in effect in 293 Encode module and encode your pathnames to the locale (or other)
197 the user environment, d) use Glib::filename_from_unicode on unicode 294 encoding in effect in the user environment, d) use
198 filenames or e) use something else to ensure your scalar has the correct 295 Glib::filename_from_unicode on unicode filenames or e) use something
199 contents. 296 else to ensure your scalar has the correct contents.
200 297
201 This works, btw. independent of the internal UTF-8 bit, which IO::AIO 298 This works, btw. independent of the internal UTF-8 bit, which IO::AIO
202 handles correctly wether it is set or not. 299 handles correctly whether it is set or not.
203 300
301 AIO REQUEST FUNCTIONS
204 $prev_pri = aioreq_pri [$pri] 302 $prev_pri = aioreq_pri [$pri]
205 Returns the priority value that would be used for the next request 303 Returns the priority value that would be used for the next request
206 and, if $pri is given, sets the priority for the next aio request. 304 and, if $pri is given, sets the priority for the next aio request.
207 305
208 The default priority is 0, the minimum and maximum priorities are -4 306 The default priority is 0, the minimum and maximum priorities are -4
… …
230 Similar to "aioreq_pri", but subtracts the given value from the 328 Similar to "aioreq_pri", but subtracts the given value from the
231 current priority, so the effect is cumulative. 329 current priority, so the effect is cumulative.
232 330
233 aio_open $pathname, $flags, $mode, $callback->($fh) 331 aio_open $pathname, $flags, $mode, $callback->($fh)
234 Asynchronously open or create a file and call the callback with a 332 Asynchronously open or create a file and call the callback with a
235 newly created filehandle for the file. 333 newly created filehandle for the file (or "undef" in case of an
334 error).
236 335
237 The pathname passed to "aio_open" must be absolute. See API NOTES, 336 The pathname passed to "aio_open" must be absolute. See API NOTES,
238 above, for an explanation. 337 above, for an explanation.
239 338
240 The $flags argument is a bitmask. See the "Fcntl" module for a list. 339 The $flags argument is a bitmask. See the "Fcntl" module for a list.
241 They are the same as used by "sysopen". 340 They are the same as used by "sysopen".
242 341
243 Likewise, $mode specifies the mode of the newly created file, if it 342 Likewise, $mode specifies the mode of the newly created file, if it
244 didn't exist and "O_CREAT" has been given, just like perl's 343 didn't exist and "O_CREAT" has been given, just like perl's
245 "sysopen", except that it is mandatory (i.e. use 0 if you don't 344 "sysopen", except that it is mandatory (i.e. use 0 if you don't
246 create new files, and 0666 or 0777 if you do). 345 create new files, and 0666 or 0777 if you do). Note that the $mode
346 will be modified by the umask in effect then the request is being
347 executed, so better never change the umask.
247 348
248 Example: 349 Example:
249 350
250 aio_open "/etc/passwd", O_RDONLY, 0, sub { 351 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
251 if ($_[0]) { 352 if ($_[0]) {
252 print "open successful, fh is $_[0]\n"; 353 print "open successful, fh is $_[0]\n";
253 ... 354 ...
254 } else { 355 } else {
255 die "open failed: $!\n"; 356 die "open failed: $!\n";
256 } 357 }
257 }; 358 };
258 359
360 In addition to all the common open modes/flags ("O_RDONLY",
361 "O_WRONLY", "O_RDWR", "O_CREAT", "O_TRUNC", "O_EXCL" and
362 "O_APPEND"), the following POSIX and non-POSIX constants are
363 available (missing ones on your system are, as usual, 0):
364
365 "O_ASYNC", "O_DIRECT", "O_NOATIME", "O_CLOEXEC", "O_NOCTTY",
366 "O_NOFOLLOW", "O_NONBLOCK", "O_EXEC", "O_SEARCH", "O_DIRECTORY",
367 "O_DSYNC", "O_RSYNC", "O_SYNC", "O_PATH", "O_TMPFILE", "O_TTY_INIT"
368 and "O_ACCMODE".
369
259 aio_close $fh, $callback->($status) 370 aio_close $fh, $callback->($status)
260 Asynchronously close a file and call the callback with the result 371 Asynchronously close a file and call the callback with the result
261 code. *WARNING:* although accepted, you should not pass in a perl 372 code.
262 filehandle here, as perl will likely close the file descriptor
263 another time when the filehandle is destroyed. Normally, you can
264 safely call perls "close" or just let filehandles go out of scope.
265 373
266 This is supposed to be a bug in the API, so that might change. It's 374 Unfortunately, you can't do this to perl. Perl *insists* very
267 therefore best to avoid this function. 375 strongly on closing the file descriptor associated with the
376 filehandle itself.
377
378 Therefore, "aio_close" will not close the filehandle - instead it
379 will use dup2 to overwrite the file descriptor with the write-end of
380 a pipe (the pipe fd will be created on demand and will be cached).
381
382 Or in other words: the file descriptor will be closed, but it will
383 not be free for reuse until the perl filehandle is closed.
384
385 aio_seek $fh, $offset, $whence, $callback->($offs)
386 Seeks the filehandle to the new $offset, similarly to perl's
387 "sysseek". The $whence can use the traditional values (0 for
388 "IO::AIO::SEEK_SET", 1 for "IO::AIO::SEEK_CUR" or 2 for
389 "IO::AIO::SEEK_END").
390
391 The resulting absolute offset will be passed to the callback, or -1
392 in case of an error.
393
394 In theory, the $whence constants could be different than the
395 corresponding values from Fcntl, but perl guarantees they are the
396 same, so don't panic.
397
398 As a GNU/Linux (and maybe Solaris) extension, also the constants
399 "IO::AIO::SEEK_DATA" and "IO::AIO::SEEK_HOLE" are available, if they
400 could be found. No guarantees about suitability for use in
401 "aio_seek" or Perl's "sysseek" can be made though, although I would
402 naively assume they "just work".
268 403
269 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 404 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
270 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 405 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
271 Reads or writes "length" bytes from the specified "fh" and "offset" 406 Reads or writes $length bytes from or to the specified $fh and
272 into the scalar given by "data" and offset "dataoffset" and calls 407 $offset into the scalar given by $data and offset $dataoffset and
273 the callback without the actual number of bytes read (or -1 on 408 calls the callback with the actual number of bytes transferred (or
274 error, just like the syscall). 409 -1 on error, just like the syscall).
410
411 "aio_read" will, like "sysread", shrink or grow the $data scalar to
412 offset plus the actual number of bytes read.
413
414 If $offset is undefined, then the current file descriptor offset
415 will be used (and updated), otherwise the file descriptor offset
416 will not be changed by these calls.
417
418 If $length is undefined in "aio_write", use the remaining length of
419 $data.
420
421 If $dataoffset is less than zero, it will be counted from the end of
422 $data.
275 423
276 The $data scalar *MUST NOT* be modified in any way while the request 424 The $data scalar *MUST NOT* be modified in any way while the request
277 is outstanding. Modifying it can result in segfaults or WW3 (if the 425 is outstanding. Modifying it can result in segfaults or World War
278 necessary/optional hardware is installed). 426 III (if the necessary/optional hardware is installed).
279 427
280 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at 428 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at
281 offset 0 within the scalar: 429 offset 0 within the scalar:
282 430
283 aio_read $fh, 7, 15, $buffer, 0, sub { 431 aio_read $fh, 7, 15, $buffer, 0, sub {
… …
288 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 436 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
289 Tries to copy $length bytes from $in_fh to $out_fh. It starts 437 Tries to copy $length bytes from $in_fh to $out_fh. It starts
290 reading at byte offset $in_offset, and starts writing at the current 438 reading at byte offset $in_offset, and starts writing at the current
291 file offset of $out_fh. Because of that, it is not safe to issue 439 file offset of $out_fh. Because of that, it is not safe to issue
292 more than one "aio_sendfile" per $out_fh, as they will interfere 440 more than one "aio_sendfile" per $out_fh, as they will interfere
293 with each other. 441 with each other. The same $in_fh works fine though, as this function
442 does not move or use the file offset of $in_fh.
294 443
444 Please note that "aio_sendfile" can read more bytes from $in_fh than
445 are written, and there is no way to find out how many more bytes
446 have been read from "aio_sendfile" alone, as "aio_sendfile" only
447 provides the number of bytes written to $out_fh. Only if the result
448 value equals $length one can assume that $length bytes have been
449 read.
450
451 Unlike with other "aio_" functions, it makes a lot of sense to use
452 "aio_sendfile" on non-blocking sockets, as long as one end
453 (typically the $in_fh) is a file - the file I/O will then be
454 asynchronous, while the socket I/O will be non-blocking. Note,
455 however, that you can run into a trap where "aio_sendfile" reads
456 some data with readahead, then fails to write all data, and when the
457 socket is ready the next time, the data in the cache is already
458 lost, forcing "aio_sendfile" to again hit the disk. Explicit
459 "aio_read" + "aio_write" let's you better control resource usage.
460
295 This call tries to make use of a native "sendfile" syscall to 461 This call tries to make use of a native "sendfile"-like syscall to
296 provide zero-copy operation. For this to work, $out_fh should refer 462 provide zero-copy operation. For this to work, $out_fh should refer
297 to a socket, and $in_fh should refer to mmap'able file. 463 to a socket, and $in_fh should refer to an mmap'able file.
298 464
299 If the native sendfile call fails or is not implemented, it will be 465 If a native sendfile cannot be found or it fails with "ENOSYS",
300 emulated, so you can call "aio_sendfile" on any type of filehandle 466 "EINVAL", "ENOTSUP", "EOPNOTSUPP", "EAFNOSUPPORT", "EPROTOTYPE" or
301 regardless of the limitations of the operating system. 467 "ENOTSOCK", it will be emulated, so you can call "aio_sendfile" on
468 any type of filehandle regardless of the limitations of the
469 operating system.
302 470
303 Please note, however, that "aio_sendfile" can read more bytes from 471 As native sendfile syscalls (as practically any non-POSIX interface
304 $in_fh than are written, and there is no way to find out how many 472 hacked together in a hurry to improve benchmark numbers) tend to be
305 bytes have been read from "aio_sendfile" alone, as "aio_sendfile" 473 rather buggy on many systems, this implementation tries to work
306 only provides the number of bytes written to $out_fh. Only if the 474 around some known bugs in Linux and FreeBSD kernels (probably
307 result value equals $length one can assume that $length bytes have 475 others, too), but that might fail, so you really really should check
308 been read. 476 the return value of "aio_sendfile" - fewer bytes than expected might
477 have been transferred.
309 478
310 aio_readahead $fh,$offset,$length, $callback->($retval) 479 aio_readahead $fh,$offset,$length, $callback->($retval)
311 "aio_readahead" populates the page cache with data from a file so 480 "aio_readahead" populates the page cache with data from a file so
312 that subsequent reads from that file will not block on disk I/O. The 481 that subsequent reads from that file will not block on disk I/O. The
313 $offset argument specifies the starting point from which data is to 482 $offset argument specifies the starting point from which data is to
… …
316 to a page boundary and bytes are read up to the next page boundary 485 to a page boundary and bytes are read up to the next page boundary
317 greater than or equal to (off-set+length). "aio_readahead" does not 486 greater than or equal to (off-set+length). "aio_readahead" does not
318 read beyond the end of the file. The current file offset of the file 487 read beyond the end of the file. The current file offset of the file
319 is left unchanged. 488 is left unchanged.
320 489
321 If that syscall doesn't exist (likely if your OS isn't Linux) it 490 If that syscall doesn't exist (likely if your kernel isn't Linux) it
322 will be emulated by simply reading the data, which would have a 491 will be emulated by simply reading the data, which would have a
323 similar effect. 492 similar effect.
324 493
325 aio_stat $fh_or_path, $callback->($status) 494 aio_stat $fh_or_path, $callback->($status)
326 aio_lstat $fh, $callback->($status) 495 aio_lstat $fh, $callback->($status)
327 Works like perl's "stat" or "lstat" in void context. The callback 496 Works almost exactly like perl's "stat" or "lstat" in void context.
328 will be called after the stat and the results will be available 497 The callback will be called after the stat and the results will be
329 using "stat _" or "-s _" etc... 498 available using "stat _" or "-s _" and other tests (with the
499 exception of "-B" and "-T").
330 500
331 The pathname passed to "aio_stat" must be absolute. See API NOTES, 501 The pathname passed to "aio_stat" must be absolute. See API NOTES,
332 above, for an explanation. 502 above, for an explanation.
333 503
334 Currently, the stats are always 64-bit-stats, i.e. instead of 504 Currently, the stats are always 64-bit-stats, i.e. instead of
335 returning an error when stat'ing a large file, the results will be 505 returning an error when stat'ing a large file, the results will be
336 silently truncated unless perl itself is compiled with large file 506 silently truncated unless perl itself is compiled with large file
337 support. 507 support.
338 508
509 To help interpret the mode and dev/rdev stat values, IO::AIO offers
510 the following constants and functions (if not implemented, the
511 constants will be 0 and the functions will either "croak" or fall
512 back on traditional behaviour).
513
514 "S_IFMT", "S_IFIFO", "S_IFCHR", "S_IFBLK", "S_IFLNK", "S_IFREG",
515 "S_IFDIR", "S_IFWHT", "S_IFSOCK", "IO::AIO::major $dev_t",
516 "IO::AIO::minor $dev_t", "IO::AIO::makedev $major, $minor".
517
518 To access higher resolution stat timestamps, see "SUBSECOND STAT
519 TIME ACCESS".
520
339 Example: Print the length of /etc/passwd: 521 Example: Print the length of /etc/passwd:
340 522
341 aio_stat "/etc/passwd", sub { 523 aio_stat "/etc/passwd", sub {
342 $_[0] and die "stat failed: $!"; 524 $_[0] and die "stat failed: $!";
343 print "size is ", -s _, "\n"; 525 print "size is ", -s _, "\n";
344 }; 526 };
345 527
528 aio_statvfs $fh_or_path, $callback->($statvfs)
529 Works like the POSIX "statvfs" or "fstatvfs" syscalls, depending on
530 whether a file handle or path was passed.
531
532 On success, the callback is passed a hash reference with the
533 following members: "bsize", "frsize", "blocks", "bfree", "bavail",
534 "files", "ffree", "favail", "fsid", "flag" and "namemax". On
535 failure, "undef" is passed.
536
537 The following POSIX IO::AIO::ST_* constants are defined: "ST_RDONLY"
538 and "ST_NOSUID".
539
540 The following non-POSIX IO::AIO::ST_* flag masks are defined to
541 their correct value when available, or to 0 on systems that do not
542 support them: "ST_NODEV", "ST_NOEXEC", "ST_SYNCHRONOUS",
543 "ST_MANDLOCK", "ST_WRITE", "ST_APPEND", "ST_IMMUTABLE",
544 "ST_NOATIME", "ST_NODIRATIME" and "ST_RELATIME".
545
546 Example: stat "/wd" and dump out the data if successful.
547
548 aio_statvfs "/wd", sub {
549 my $f = $_[0]
550 or die "statvfs: $!";
551
552 use Data::Dumper;
553 say Dumper $f;
554 };
555
556 # result:
557 {
558 bsize => 1024,
559 bfree => 4333064312,
560 blocks => 10253828096,
561 files => 2050765568,
562 flag => 4096,
563 favail => 2042092649,
564 bavail => 4333064312,
565 ffree => 2042092649,
566 namemax => 255,
567 frsize => 1024,
568 fsid => 1810
569 }
570
571 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
572 Works like perl's "utime" function (including the special case of
573 $atime and $mtime being undef). Fractional times are supported if
574 the underlying syscalls support them.
575
576 When called with a pathname, uses utimensat(2) or utimes(2) if
577 available, otherwise utime(2). If called on a file descriptor, uses
578 futimens(2) or futimes(2) if available, otherwise returns ENOSYS, so
579 this is not portable.
580
581 Examples:
582
583 # set atime and mtime to current time (basically touch(1)):
584 aio_utime "path", undef, undef;
585 # set atime to current time and mtime to beginning of the epoch:
586 aio_utime "path", time, undef; # undef==0
587
588 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
589 Works like perl's "chown" function, except that "undef" for either
590 $uid or $gid is being interpreted as "do not change" (but -1 can
591 also be used).
592
593 Examples:
594
595 # same as "chown root path" in the shell:
596 aio_chown "path", 0, -1;
597 # same as above:
598 aio_chown "path", 0, undef;
599
600 aio_truncate $fh_or_path, $offset, $callback->($status)
601 Works like truncate(2) or ftruncate(2).
602
603 aio_allocate $fh, $mode, $offset, $len, $callback->($status)
604 Allocates or frees disk space according to the $mode argument. See
605 the linux "fallocate" documentation for details.
606
607 $mode is usually 0 or "IO::AIO::FALLOC_FL_KEEP_SIZE" to allocate
608 space, or "IO::AIO::FALLOC_FL_PUNCH_HOLE |
609 IO::AIO::FALLOC_FL_KEEP_SIZE", to deallocate a file range.
610
611 IO::AIO also supports "FALLOC_FL_COLLAPSE_RANGE", to remove a range
612 (without leaving a hole), "FALLOC_FL_ZERO_RANGE", to zero a range,
613 "FALLOC_FL_INSERT_RANGE" to insert a range and
614 "FALLOC_FL_UNSHARE_RANGE" to unshare shared blocks (see your
615 fallocate(2) manpage).
616
617 The file system block size used by "fallocate" is presumably the
618 "f_bsize" returned by "statvfs", but different filesystems and
619 filetypes can dictate other limitations.
620
621 If "fallocate" isn't available or cannot be emulated (currently no
622 emulation will be attempted), passes -1 and sets $! to "ENOSYS".
623
624 aio_chmod $fh_or_path, $mode, $callback->($status)
625 Works like perl's "chmod" function.
626
346 aio_unlink $pathname, $callback->($status) 627 aio_unlink $pathname, $callback->($status)
347 Asynchronously unlink (delete) a file and call the callback with the 628 Asynchronously unlink (delete) a file and call the callback with the
348 result code. 629 result code.
349 630
350 aio_mknod $path, $mode, $dev, $callback->($status) 631 aio_mknod $pathname, $mode, $dev, $callback->($status)
351 [EXPERIMENTAL] 632 [EXPERIMENTAL]
352 633
353 Asynchronously create a device node (or fifo). See mknod(2). 634 Asynchronously create a device node (or fifo). See mknod(2).
354 635
355 The only (POSIX-) portable way of calling this function is: 636 The only (POSIX-) portable way of calling this function is:
356 637
357 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 638 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
639
640 See "aio_stat" for info about some potentially helpful extra
641 constants and functions.
358 642
359 aio_link $srcpath, $dstpath, $callback->($status) 643 aio_link $srcpath, $dstpath, $callback->($status)
360 Asynchronously create a new link to the existing object at $srcpath 644 Asynchronously create a new link to the existing object at $srcpath
361 at the path $dstpath and call the callback with the result code. 645 at the path $dstpath and call the callback with the result code.
362 646
363 aio_symlink $srcpath, $dstpath, $callback->($status) 647 aio_symlink $srcpath, $dstpath, $callback->($status)
364 Asynchronously create a new symbolic link to the existing object at 648 Asynchronously create a new symbolic link to the existing object at
365 $srcpath at the path $dstpath and call the callback with the result 649 $srcpath at the path $dstpath and call the callback with the result
366 code. 650 code.
367 651
368 aio_readlink $path, $callback->($link) 652 aio_readlink $pathname, $callback->($link)
369 Asynchronously read the symlink specified by $path and pass it to 653 Asynchronously read the symlink specified by $path and pass it to
370 the callback. If an error occurs, nothing or undef gets passed to 654 the callback. If an error occurs, nothing or undef gets passed to
371 the callback. 655 the callback.
372 656
657 aio_realpath $pathname, $callback->($path)
658 Asynchronously make the path absolute and resolve any symlinks in
659 $path. The resulting path only consists of directories (same as
660 Cwd::realpath).
661
662 This request can be used to get the absolute path of the current
663 working directory by passing it a path of . (a single dot).
664
373 aio_rename $srcpath, $dstpath, $callback->($status) 665 aio_rename $srcpath, $dstpath, $callback->($status)
374 Asynchronously rename the object at $srcpath to $dstpath, just as 666 Asynchronously rename the object at $srcpath to $dstpath, just as
375 rename(2) and call the callback with the result code. 667 rename(2) and call the callback with the result code.
376 668
669 On systems that support the AIO::WD working directory abstraction
670 natively, the case "[$wd, "."]" as $srcpath is specialcased -
671 instead of failing, "rename" is called on the absolute path of $wd.
672
673 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status)
674 Basically a version of "aio_rename" with an additional $flags
675 argument. Calling this with "$flags=0" is the same as calling
676 "aio_rename".
677
678 Non-zero flags are currently only supported on GNU/Linux systems
679 that support renameat2. Other systems fail with "ENOSYS" in this
680 case.
681
682 The following constants are available (missing ones are, as usual
683 0), see renameat2(2) for details:
684
685 "IO::AIO::RENAME_NOREPLACE", "IO::AIO::RENAME_EXCHANGE" and
686 "IO::AIO::RENAME_WHITEOUT".
687
688 aio_mkdir $pathname, $mode, $callback->($status)
689 Asynchronously mkdir (create) a directory and call the callback with
690 the result code. $mode will be modified by the umask at the time the
691 request is executed, so do not change your umask.
692
377 aio_rmdir $pathname, $callback->($status) 693 aio_rmdir $pathname, $callback->($status)
378 Asynchronously rmdir (delete) a directory and call the callback with 694 Asynchronously rmdir (delete) a directory and call the callback with
379 the result code. 695 the result code.
696
697 On systems that support the AIO::WD working directory abstraction
698 natively, the case "[$wd, "."]" is specialcased - instead of
699 failing, "rmdir" is called on the absolute path of $wd.
380 700
381 aio_readdir $pathname, $callback->($entries) 701 aio_readdir $pathname, $callback->($entries)
382 Unlike the POSIX call of the same name, "aio_readdir" reads an 702 Unlike the POSIX call of the same name, "aio_readdir" reads an
383 entire directory (i.e. opendir + readdir + closedir). The entries 703 entire directory (i.e. opendir + readdir + closedir). The entries
384 will not be sorted, and will NOT include the "." and ".." entries. 704 will not be sorted, and will NOT include the "." and ".." entries.
385 705
386 The callback a single argument which is either "undef" or an 706 The callback is passed a single argument which is either "undef" or
387 array-ref with the filenames. 707 an array-ref with the filenames.
708
709 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
710 Quite similar to "aio_readdir", but the $flags argument allows one
711 to tune behaviour and output format. In case of an error, $entries
712 will be "undef".
713
714 The flags are a combination of the following constants, ORed
715 together (the flags will also be passed to the callback, possibly
716 modified):
717
718 IO::AIO::READDIR_DENTS
719 Normally the callback gets an arrayref consisting of names only
720 (as with "aio_readdir"). If this flag is set, then the callback
721 gets an arrayref with "[$name, $type, $inode]" arrayrefs, each
722 describing a single directory entry in more detail:
723
724 $name is the name of the entry.
725
726 $type is one of the "IO::AIO::DT_xxx" constants:
727
728 "IO::AIO::DT_UNKNOWN", "IO::AIO::DT_FIFO", "IO::AIO::DT_CHR",
729 "IO::AIO::DT_DIR", "IO::AIO::DT_BLK", "IO::AIO::DT_REG",
730 "IO::AIO::DT_LNK", "IO::AIO::DT_SOCK", "IO::AIO::DT_WHT".
731
732 "IO::AIO::DT_UNKNOWN" means just that: readdir does not know. If
733 you need to know, you have to run stat yourself. Also, for
734 speed/memory reasons, the $type scalars are read-only: you must
735 not modify them.
736
737 $inode is the inode number (which might not be exact on systems
738 with 64 bit inode numbers and 32 bit perls). This field has
739 unspecified content on systems that do not deliver the inode
740 information.
741
742 IO::AIO::READDIR_DIRS_FIRST
743 When this flag is set, then the names will be returned in an
744 order where likely directories come first, in optimal stat
745 order. This is useful when you need to quickly find directories,
746 or you want to find all directories while avoiding to stat()
747 each entry.
748
749 If the system returns type information in readdir, then this is
750 used to find directories directly. Otherwise, likely directories
751 are names beginning with ".", or otherwise names with no dots,
752 of which names with short names are tried first.
753
754 IO::AIO::READDIR_STAT_ORDER
755 When this flag is set, then the names will be returned in an
756 order suitable for stat()'ing each one. That is, when you plan
757 to stat() most or all files in the given directory, then the
758 returned order will likely be faster.
759
760 If both this flag and "IO::AIO::READDIR_DIRS_FIRST" are
761 specified, then the likely dirs come first, resulting in a less
762 optimal stat order for stat'ing all entries, but likely a more
763 optimal order for finding subdirectories.
764
765 IO::AIO::READDIR_FOUND_UNKNOWN
766 This flag should not be set when calling "aio_readdirx".
767 Instead, it is being set by "aio_readdirx", when any of the
768 $type's found were "IO::AIO::DT_UNKNOWN". The absence of this
769 flag therefore indicates that all $type's are known, which can
770 be used to speed up some algorithms.
771
772 aio_slurp $pathname, $offset, $length, $data, $callback->($status)
773 Opens, reads and closes the given file. The data is put into $data,
774 which is resized as required.
775
776 If $offset is negative, then it is counted from the end of the file.
777
778 If $length is zero, then the remaining length of the file is used.
779 Also, in this case, the same limitations to modifying $data apply as
780 when IO::AIO::mmap is used, i.e. it must only be modified in-place
781 with "substr". If the size of the file is known, specifying a
782 non-zero $length results in a performance advantage.
783
784 This request is similar to the older "aio_load" request, but since
785 it is a single request, it might be more efficient to use.
786
787 Example: load /etc/passwd into $passwd.
788
789 my $passwd;
790 aio_slurp "/etc/passwd", 0, 0, $passwd, sub {
791 $_[0] >= 0
792 or die "/etc/passwd: $!\n";
793
794 printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd;
795 print $passwd;
796 };
797 IO::AIO::flush;
798
799 aio_load $pathname, $data, $callback->($status)
800 This is a composite request that tries to fully load the given file
801 into memory. Status is the same as with aio_read.
802
803 Using "aio_slurp" might be more efficient, as it is a single
804 request.
388 805
389 aio_copy $srcpath, $dstpath, $callback->($status) 806 aio_copy $srcpath, $dstpath, $callback->($status)
390 Try to copy the *file* (directories not supported as either source 807 Try to copy the *file* (directories not supported as either source
391 or destination) from $srcpath to $dstpath and call the callback with 808 or destination) from $srcpath to $dstpath and call the callback with
392 the 0 (error) or -1 ok. 809 a status of 0 (ok) or -1 (error, see $!).
393 810
811 Existing destination files will be truncated.
812
394 This is a composite request that it creates the destination file 813 This is a composite request that creates the destination file with
395 with mode 0200 and copies the contents of the source file into it 814 mode 0200 and copies the contents of the source file into it using
396 using "aio_sendfile", followed by restoring atime, mtime, access 815 "aio_sendfile", followed by restoring atime, mtime, access mode and
397 mode and uid/gid, in that order. 816 uid/gid, in that order.
398 817
399 If an error occurs, the partial destination file will be unlinked, 818 If an error occurs, the partial destination file will be unlinked,
400 if possible, except when setting atime, mtime, access mode and 819 if possible, except when setting atime, mtime, access mode and
401 uid/gid, where errors are being ignored. 820 uid/gid, where errors are being ignored.
402 821
403 aio_move $srcpath, $dstpath, $callback->($status) 822 aio_move $srcpath, $dstpath, $callback->($status)
404 Try to move the *file* (directories not supported as either source 823 Try to move the *file* (directories not supported as either source
405 or destination) from $srcpath to $dstpath and call the callback with 824 or destination) from $srcpath to $dstpath and call the callback with
406 the 0 (error) or -1 ok. 825 a status of 0 (ok) or -1 (error, see $!).
407 826
408 This is a composite request that tries to rename(2) the file first. 827 This is a composite request that tries to rename(2) the file first;
409 If rename files with "EXDEV", it copies the file with "aio_copy" 828 if rename fails with "EXDEV", it copies the file with "aio_copy"
410 and, if that is successful, unlinking the $srcpath. 829 and, if that is successful, unlinks the $srcpath.
411 830
412 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) 831 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
413 Scans a directory (similar to "aio_readdir") but additionally tries 832 Scans a directory (similar to "aio_readdir") but additionally tries
414 to efficiently separate the entries of directory $path into two sets 833 to efficiently separate the entries of directory $path into two sets
415 of names, directories you can recurse into (directories), and ones 834 of names, directories you can recurse into (directories), and ones
416 you cannot recurse into (everything else, including symlinks to 835 you cannot recurse into (everything else, including symlinks to
417 directories). 836 directories).
418 837
419 "aio_scandir" is a composite request that creates of many sub 838 "aio_scandir" is a composite request that generates many sub
420 requests_ $maxreq specifies the maximum number of outstanding aio 839 requests. $maxreq specifies the maximum number of outstanding aio
421 requests that this function generates. If it is "<= 0", then a 840 requests that this function generates. If it is "<= 0", then a
422 suitable default will be chosen (currently 4). 841 suitable default will be chosen (currently 4).
423 842
424 On error, the callback is called without arguments, otherwise it 843 On error, the callback is called without arguments, otherwise it
425 receives two array-refs with path-relative entry names. 844 receives two array-refs with path-relative entry names.
… …
435 Implementation notes. 854 Implementation notes.
436 855
437 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry 856 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry
438 can. 857 can.
439 858
859 If readdir returns file type information, then this is used directly
860 to find directories.
861
440 After reading the directory, the modification time, size etc. of the 862 Otherwise, after reading the directory, the modification time, size
441 directory before and after the readdir is checked, and if they match 863 etc. of the directory before and after the readdir is checked, and
442 (and isn't the current time), the link count will be used to decide 864 if they match (and isn't the current time), the link count will be
443 how many entries are directories (if >= 2). Otherwise, no knowledge 865 used to decide how many entries are directories (if >= 2).
444 of the number of subdirectories will be assumed. 866 Otherwise, no knowledge of the number of subdirectories will be
867 assumed.
445 868
446 Then entries will be sorted into likely directories (everything 869 Then entries will be sorted into likely directories a non-initial
447 without a non-initial dot currently) and likely non-directories 870 dot currently) and likely non-directories (see "aio_readdirx"). Then
448 (everything else). Then every entry plus an appended "/." will be 871 every entry plus an appended "/." will be "stat"'ed, likely
449 "stat"'ed, likely directories first. If that succeeds, it assumes 872 directories first, in order of their inode numbers. If that
450 that the entry is a directory or a symlink to directory (which will 873 succeeds, it assumes that the entry is a directory or a symlink to
451 be checked seperately). This is often faster than stat'ing the entry 874 directory (which will be checked separately). This is often faster
452 itself because filesystems might detect the type of the entry 875 than stat'ing the entry itself because filesystems might detect the
453 without reading the inode data (e.g. ext2fs filetype feature). 876 type of the entry without reading the inode data (e.g. ext2fs
877 filetype feature), even on systems that cannot return the filetype
878 information on readdir.
454 879
455 If the known number of directories (link count - 2) has been 880 If the known number of directories (link count - 2) has been
456 reached, the rest of the entries is assumed to be non-directories. 881 reached, the rest of the entries is assumed to be non-directories.
457 882
458 This only works with certainty on POSIX (= UNIX) filesystems, which 883 This only works with certainty on POSIX (= UNIX) filesystems, which
… …
460 885
461 It will also likely work on non-POSIX filesystems with reduced 886 It will also likely work on non-POSIX filesystems with reduced
462 efficiency as those tend to return 0 or 1 as link counts, which 887 efficiency as those tend to return 0 or 1 as link counts, which
463 disables the directory counting heuristic. 888 disables the directory counting heuristic.
464 889
890 aio_rmtree $pathname, $callback->($status)
891 Delete a directory tree starting (and including) $path, return the
892 status of the final "rmdir" only. This is a composite request that
893 uses "aio_scandir" to recurse into and rmdir directories, and unlink
894 everything else.
895
896 aio_fcntl $fh, $cmd, $arg, $callback->($status)
897 aio_ioctl $fh, $request, $buf, $callback->($status)
898 These work just like the "fcntl" and "ioctl" built-in functions,
899 except they execute asynchronously and pass the return value to the
900 callback.
901
902 Both calls can be used for a lot of things, some of which make more
903 sense to run asynchronously in their own thread, while some others
904 make less sense. For example, calls that block waiting for external
905 events, such as locking, will also lock down an I/O thread while it
906 is waiting, which can deadlock the whole I/O system. At the same
907 time, there might be no alternative to using a thread to wait.
908
909 So in general, you should only use these calls for things that do
910 (filesystem) I/O, not for things that wait for other events
911 (network, other processes), although if you are careful and know
912 what you are doing, you still can.
913
914 The following constants are available (missing ones are, as usual
915 0):
916
917 "F_DUPFD_CLOEXEC",
918
919 "F_OFD_GETLK", "F_OFD_SETLK", "F_OFD_GETLKW",
920
921 "FIFREEZE", "FITHAW", "FITRIM", "FICLONE", "FICLONERANGE",
922 "FIDEDUPERANGE".
923
924 "FS_IOC_GETFLAGS", "FS_IOC_SETFLAGS", "FS_IOC_GETVERSION",
925 "FS_IOC_SETVERSION", "FS_IOC_FIEMAP".
926
927 "FS_IOC_FSGETXATTR", "FS_IOC_FSSETXATTR",
928 "FS_IOC_SET_ENCRYPTION_POLICY", "FS_IOC_GET_ENCRYPTION_PWSALT",
929 "FS_IOC_GET_ENCRYPTION_POLICY", "FS_KEY_DESCRIPTOR_SIZE".
930
931 "FS_SECRM_FL", "FS_UNRM_FL", "FS_COMPR_FL", "FS_SYNC_FL",
932 "FS_IMMUTABLE_FL", "FS_APPEND_FL", "FS_NODUMP_FL", "FS_NOATIME_FL",
933 "FS_DIRTY_FL", "FS_COMPRBLK_FL", "FS_NOCOMP_FL", "FS_ENCRYPT_FL",
934 "FS_BTREE_FL", "FS_INDEX_FL", "FS_JOURNAL_DATA_FL", "FS_NOTAIL_FL",
935 "FS_DIRSYNC_FL", "FS_TOPDIR_FL", "FS_FL_USER_MODIFIABLE".
936
937 "FS_XFLAG_REALTIME", "FS_XFLAG_PREALLOC", "FS_XFLAG_IMMUTABLE",
938 "FS_XFLAG_APPEND", "FS_XFLAG_SYNC", "FS_XFLAG_NOATIME",
939 "FS_XFLAG_NODUMP", "FS_XFLAG_RTINHERIT", "FS_XFLAG_PROJINHERIT",
940 "FS_XFLAG_NOSYMLINKS", "FS_XFLAG_EXTSIZE", "FS_XFLAG_EXTSZINHERIT",
941 "FS_XFLAG_NODEFRAG", "FS_XFLAG_FILESTREAM", "FS_XFLAG_DAX",
942 "FS_XFLAG_HASATTR",
943
944 aio_sync $callback->($status)
945 Asynchronously call sync and call the callback when finished.
946
465 aio_fsync $fh, $callback->($status) 947 aio_fsync $fh, $callback->($status)
466 Asynchronously call fsync on the given filehandle and call the 948 Asynchronously call fsync on the given filehandle and call the
467 callback with the fsync result code. 949 callback with the fsync result code.
468 950
469 aio_fdatasync $fh, $callback->($status) 951 aio_fdatasync $fh, $callback->($status)
470 Asynchronously call fdatasync on the given filehandle and call the 952 Asynchronously call fdatasync on the given filehandle and call the
471 callback with the fdatasync result code. 953 callback with the fdatasync result code.
472 954
473 If this call isn't available because your OS lacks it or it couldn't 955 If this call isn't available because your OS lacks it or it couldn't
474 be detected, it will be emulated by calling "fsync" instead. 956 be detected, it will be emulated by calling "fsync" instead.
957
958 aio_syncfs $fh, $callback->($status)
959 Asynchronously call the syncfs syscall to sync the filesystem
960 associated to the given filehandle and call the callback with the
961 syncfs result code. If syncfs is not available, calls sync(), but
962 returns -1 and sets errno to "ENOSYS" nevertheless.
963
964 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
965 Sync the data portion of the file specified by $offset and $length
966 to disk (but NOT the metadata), by calling the Linux-specific
967 sync_file_range call. If sync_file_range is not available or it
968 returns ENOSYS, then fdatasync or fsync is being substituted.
969
970 $flags can be a combination of
971 "IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE",
972 "IO::AIO::SYNC_FILE_RANGE_WRITE" and
973 "IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER": refer to the sync_file_range
974 manpage for details.
975
976 aio_pathsync $pathname, $callback->($status)
977 This request tries to open, fsync and close the given path. This is
978 a composite request intended to sync directories after directory
979 operations (E.g. rename). This might not work on all operating
980 systems or have any specific effect, but usually it makes sure that
981 directory changes get written to disc. It works for anything that
982 can be opened for read-only, not just directories.
983
984 Future versions of this function might fall back to other methods
985 when "fsync" on the directory fails (such as calling "sync").
986
987 Passes 0 when everything went ok, and -1 on error.
988
989 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC,
990 $callback->($status)
991 This is a rather advanced IO::AIO call, which only works on
992 mmap(2)ed scalars (see the "IO::AIO::mmap" function, although it
993 also works on data scalars managed by the Sys::Mmap or Mmap modules,
994 note that the scalar must only be modified in-place while an aio
995 operation is pending on it).
996
997 It calls the "msync" function of your OS, if available, with the
998 memory area starting at $offset in the string and ending $length
999 bytes later. If $length is negative, counts from the end, and if
1000 $length is "undef", then it goes till the end of the string. The
1001 flags can be either "IO::AIO::MS_ASYNC" or "IO::AIO::MS_SYNC", plus
1002 an optional "IO::AIO::MS_INVALIDATE".
1003
1004 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0,
1005 $callback->($status)
1006 This is a rather advanced IO::AIO call, which works best on
1007 mmap(2)ed scalars.
1008
1009 It touches (reads or writes) all memory pages in the specified range
1010 inside the scalar. All caveats and parameters are the same as for
1011 "aio_msync", above, except for flags, which must be either 0 (which
1012 reads all pages and ensures they are instantiated) or
1013 "IO::AIO::MT_MODIFY", which modifies the memory pages (by reading
1014 and writing an octet from it, which dirties the page).
1015
1016 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
1017 This is a rather advanced IO::AIO call, which works best on
1018 mmap(2)ed scalars.
1019
1020 It reads in all the pages of the underlying storage into memory (if
1021 any) and locks them, so they are not getting swapped/paged out or
1022 removed.
1023
1024 If $length is undefined, then the scalar will be locked till the
1025 end.
1026
1027 On systems that do not implement "mlock", this function returns -1
1028 and sets errno to "ENOSYS".
1029
1030 Note that the corresponding "munlock" is synchronous and is
1031 documented under "MISCELLANEOUS FUNCTIONS".
1032
1033 Example: open a file, mmap and mlock it - both will be undone when
1034 $data gets destroyed.
1035
1036 open my $fh, "<", $path or die "$path: $!";
1037 my $data;
1038 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
1039 aio_mlock $data; # mlock in background
1040
1041 aio_mlockall $flags, $callback->($status)
1042 Calls the "mlockall" function with the given $flags (a combination
1043 of "IO::AIO::MCL_CURRENT", "IO::AIO::MCL_FUTURE" and
1044 "IO::AIO::MCL_ONFAULT").
1045
1046 On systems that do not implement "mlockall", this function returns
1047 -1 and sets errno to "ENOSYS". Similarly, flag combinations not
1048 supported by the system result in a return value of -1 with errno
1049 being set to "EINVAL".
1050
1051 Note that the corresponding "munlockall" is synchronous and is
1052 documented under "MISCELLANEOUS FUNCTIONS".
1053
1054 Example: asynchronously lock all current and future pages into
1055 memory.
1056
1057 aio_mlockall IO::AIO::MCL_FUTURE;
1058
1059 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
1060 Queries the extents of the given file (by calling the Linux "FIEMAP"
1061 ioctl, see <http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for
1062 details). If the ioctl is not available on your OS, then this
1063 request will fail with "ENOSYS".
1064
1065 $start is the starting offset to query extents for, $length is the
1066 size of the range to query - if it is "undef", then the whole file
1067 will be queried.
1068
1069 $flags is a combination of flags ("IO::AIO::FIEMAP_FLAG_SYNC" or
1070 "IO::AIO::FIEMAP_FLAG_XATTR" - "IO::AIO::FIEMAP_FLAGS_COMPAT" is
1071 also exported), and is normally 0 or "IO::AIO::FIEMAP_FLAG_SYNC" to
1072 query the data portion.
1073
1074 $count is the maximum number of extent records to return. If it is
1075 "undef", then IO::AIO queries all extents of the range. As a very
1076 special case, if it is 0, then the callback receives the number of
1077 extents instead of the extents themselves (which is unreliable, see
1078 below).
1079
1080 If an error occurs, the callback receives no arguments. The special
1081 "errno" value "IO::AIO::EBADR" is available to test for flag errors.
1082
1083 Otherwise, the callback receives an array reference with extent
1084 structures. Each extent structure is an array reference itself, with
1085 the following members:
1086
1087 [$logical, $physical, $length, $flags]
1088
1089 Flags is any combination of the following flag values (typically
1090 either 0 or "IO::AIO::FIEMAP_EXTENT_LAST" (1)):
1091
1092 "IO::AIO::FIEMAP_EXTENT_LAST", "IO::AIO::FIEMAP_EXTENT_UNKNOWN",
1093 "IO::AIO::FIEMAP_EXTENT_DELALLOC", "IO::AIO::FIEMAP_EXTENT_ENCODED",
1094 "IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED",
1095 "IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED",
1096 "IO::AIO::FIEMAP_EXTENT_DATA_INLINE",
1097 "IO::AIO::FIEMAP_EXTENT_DATA_TAIL",
1098 "IO::AIO::FIEMAP_EXTENT_UNWRITTEN", "IO::AIO::FIEMAP_EXTENT_MERGED"
1099 or "IO::AIO::FIEMAP_EXTENT_SHARED".
1100
1101 At the time of this writing (Linux 3.2), this request is unreliable
1102 unless $count is "undef", as the kernel has all sorts of bugs
1103 preventing it to return all extents of a range for files with a
1104 large number of extents. The code (only) works around all these
1105 issues if $count is "undef".
475 1106
476 aio_group $callback->(...) 1107 aio_group $callback->(...)
477 This is a very special aio request: Instead of doing something, it 1108 This is a very special aio request: Instead of doing something, it
478 is a container for other aio requests, which is useful if you want 1109 is a container for other aio requests, which is useful if you want
479 to bundle many requests into a single, composite, request with a 1110 to bundle many requests into a single, composite, request with a
… …
514 requests like sleep and file handle readable/writable, the overhead 1145 requests like sleep and file handle readable/writable, the overhead
515 this creates is immense (it blocks a thread for a long time) so do 1146 this creates is immense (it blocks a thread for a long time) so do
516 not use this function except to put your application under 1147 not use this function except to put your application under
517 artificial I/O pressure. 1148 artificial I/O pressure.
518 1149
1150 IO::AIO::WD - multiple working directories
1151 Your process only has one current working directory, which is used by
1152 all threads. This makes it hard to use relative paths (some other
1153 component could call "chdir" at any time, and it is hard to control when
1154 the path will be used by IO::AIO).
1155
1156 One solution for this is to always use absolute paths. This usually
1157 works, but can be quite slow (the kernel has to walk the whole path on
1158 every access), and can also be a hassle to implement.
1159
1160 Newer POSIX systems have a number of functions (openat, fdopendir,
1161 futimensat and so on) that make it possible to specify working
1162 directories per operation.
1163
1164 For portability, and because the clowns who "designed", or shall I
1165 write, perpetrated this new interface were obviously half-drunk, this
1166 abstraction cannot be perfect, though.
1167
1168 IO::AIO allows you to convert directory paths into a so-called
1169 IO::AIO::WD object. This object stores the canonicalised, absolute
1170 version of the path, and on systems that allow it, also a directory file
1171 descriptor.
1172
1173 Everywhere where a pathname is accepted by IO::AIO (e.g. in "aio_stat"
1174 or "aio_unlink"), one can specify an array reference with an IO::AIO::WD
1175 object and a pathname instead (or the IO::AIO::WD object alone, which
1176 gets interpreted as "[$wd, "."]"). If the pathname is absolute, the
1177 IO::AIO::WD object is ignored, otherwise the pathname is resolved
1178 relative to that IO::AIO::WD object.
1179
1180 For example, to get a wd object for /etc and then stat passwd inside,
1181 you would write:
1182
1183 aio_wd "/etc", sub {
1184 my $etcdir = shift;
1185
1186 # although $etcdir can be undef on error, there is generally no reason
1187 # to check for errors here, as aio_stat will fail with ENOENT
1188 # when $etcdir is undef.
1189
1190 aio_stat [$etcdir, "passwd"], sub {
1191 # yay
1192 };
1193 };
1194
1195 The fact that "aio_wd" is a request and not a normal function shows that
1196 creating an IO::AIO::WD object is itself a potentially blocking
1197 operation, which is why it is done asynchronously.
1198
1199 To stat the directory obtained with "aio_wd" above, one could write
1200 either of the following three request calls:
1201
1202 aio_lstat "/etc" , sub { ... # pathname as normal string
1203 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1204 aio_lstat $wd , sub { ... # shorthand for the previous
1205
1206 As with normal pathnames, IO::AIO keeps a copy of the working directory
1207 object and the pathname string, so you could write the following without
1208 causing any issues due to $path getting reused:
1209
1210 my $path = [$wd, undef];
1211
1212 for my $name (qw(abc def ghi)) {
1213 $path->[1] = $name;
1214 aio_stat $path, sub {
1215 # ...
1216 };
1217 }
1218
1219 There are some caveats: when directories get renamed (or deleted), the
1220 pathname string doesn't change, so will point to the new directory (or
1221 nowhere at all), while the directory fd, if available on the system,
1222 will still point to the original directory. Most functions accepting a
1223 pathname will use the directory fd on newer systems, and the string on
1224 older systems. Some functions (such as "aio_realpath") will always rely
1225 on the string form of the pathname.
1226
1227 So this functionality is mainly useful to get some protection against
1228 "chdir", to easily get an absolute path out of a relative path for
1229 future reference, and to speed up doing many operations in the same
1230 directory (e.g. when stat'ing all files in a directory).
1231
1232 The following functions implement this working directory abstraction:
1233
1234 aio_wd $pathname, $callback->($wd)
1235 Asynchonously canonicalise the given pathname and convert it to an
1236 IO::AIO::WD object representing it. If possible and supported on the
1237 system, also open a directory fd to speed up pathname resolution
1238 relative to this working directory.
1239
1240 If something goes wrong, then "undef" is passwd to the callback
1241 instead of a working directory object and $! is set appropriately.
1242 Since passing "undef" as working directory component of a pathname
1243 fails the request with "ENOENT", there is often no need for error
1244 checking in the "aio_wd" callback, as future requests using the
1245 value will fail in the expected way.
1246
1247 IO::AIO::CWD
1248 This is a compiletime constant (object) that represents the process
1249 current working directory.
1250
1251 Specifying this object as working directory object for a pathname is
1252 as if the pathname would be specified directly, without a directory
1253 object. For example, these calls are functionally identical:
1254
1255 aio_stat "somefile", sub { ... };
1256 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1257
1258 To recover the path associated with an IO::AIO::WD object, you can use
1259 "aio_realpath":
1260
1261 aio_realpath $wd, sub {
1262 warn "path is $_[0]\n";
1263 };
1264
1265 Currently, "aio_statvfs" always, and "aio_rename" and "aio_rmdir"
1266 sometimes, fall back to using an absolue path.
1267
519 IO::AIO::REQ CLASS 1268 IO::AIO::REQ CLASS
520 All non-aggregate "aio_*" functions return an object of this class when 1269 All non-aggregate "aio_*" functions return an object of this class when
521 called in non-void context. 1270 called in non-void context.
522 1271
523 cancel $req 1272 cancel $req
524 Cancels the request, if possible. Has the effect of skipping 1273 Cancels the request, if possible. Has the effect of skipping
525 execution when entering the execute state and skipping calling the 1274 execution when entering the execute state and skipping calling the
526 callback when entering the the result state, but will leave the 1275 callback when entering the the result state, but will leave the
527 request otherwise untouched. That means that requests that currently 1276 request otherwise untouched (with the exception of readdir). That
528 execute will not be stopped and resources held by the request will 1277 means that requests that currently execute will not be stopped and
529 not be freed prematurely. 1278 resources held by the request will not be freed prematurely.
530 1279
531 cb $req $callback->(...) 1280 cb $req $callback->(...)
532 Replace (or simply set) the callback registered to the request. 1281 Replace (or simply set) the callback registered to the request.
533 1282
534 IO::AIO::GRP CLASS 1283 IO::AIO::GRP CLASS
… …
561 }; 1310 };
562 1311
563 This makes it very easy to create composite requests (see the source of 1312 This makes it very easy to create composite requests (see the source of
564 "aio_move" for an application) that work and feel like simple requests. 1313 "aio_move" for an application) that work and feel like simple requests.
565 1314
566 * The IO::AIO::GRP objects will be cleaned up during calls to 1315 * The IO::AIO::GRP objects will be cleaned up during calls to
567 "IO::AIO::poll_cb", just like any other request. 1316 "IO::AIO::poll_cb", just like any other request.
1317
568 * They can be canceled like any other request. Canceling will cancel not 1318 * They can be canceled like any other request. Canceling will cancel
569 only the request itself, but also all requests it contains. 1319 not only the request itself, but also all requests it contains.
1320
570 * They can also can also be added to other IO::AIO::GRP objects. 1321 * They can also can also be added to other IO::AIO::GRP objects.
1322
571 * You must not add requests to a group from within the group callback 1323 * You must not add requests to a group from within the group callback
572 (or any later time). 1324 (or any later time).
573 1325
574 Their lifetime, simplified, looks like this: when they are empty, they 1326 Their lifetime, simplified, looks like this: when they are empty, they
575 will finish very quickly. If they contain only requests that are in the 1327 will finish very quickly. If they contain only requests that are in the
576 "done" state, they will also finish. Otherwise they will continue to 1328 "done" state, they will also finish. Otherwise they will continue to
577 exist. 1329 exist.
578 1330
579 That means after creating a group you have some time to add requests. 1331 That means after creating a group you have some time to add requests
580 And in the callbacks of those requests, you can add further requests to 1332 (precisely before the callback has been invoked, which is only done
581 the group. And only when all those requests have finished will the the 1333 within the "poll_cb"). And in the callbacks of those requests, you can
582 group itself finish. 1334 add further requests to the group. And only when all those requests have
1335 finished will the the group itself finish.
583 1336
584 add $grp ... 1337 add $grp ...
585 $grp->add (...) 1338 $grp->add (...)
586 Add one or more requests to the group. Any type of IO::AIO::REQ can 1339 Add one or more requests to the group. Any type of IO::AIO::REQ can
587 be added, including other groups, as long as you do not create 1340 be added, including other groups, as long as you do not create
… …
592 $grp->cancel_subs 1345 $grp->cancel_subs
593 Cancel all subrequests and clears any feeder, but not the group 1346 Cancel all subrequests and clears any feeder, but not the group
594 request itself. Useful when you queued a lot of events but got a 1347 request itself. Useful when you queued a lot of events but got a
595 result early. 1348 result early.
596 1349
1350 The group request will finish normally (you cannot add requests to
1351 the group).
1352
597 $grp->result (...) 1353 $grp->result (...)
598 Set the result value(s) that will be passed to the group callback 1354 Set the result value(s) that will be passed to the group callback
599 when all subrequests have finished and set thre groups errno to the 1355 when all subrequests have finished and set the groups errno to the
600 current value of errno (just like calling "errno" without an error 1356 current value of errno (just like calling "errno" without an error
601 number). By default, no argument will be passed and errno is zero. 1357 number). By default, no argument will be passed and errno is zero.
602 1358
603 $grp->errno ([$errno]) 1359 $grp->errno ([$errno])
604 Sets the group errno value to $errno, or the current value of errno 1360 Sets the group errno value to $errno, or the current value of errno
… …
615 Sets a feeder/generator on this group: every group can have an 1371 Sets a feeder/generator on this group: every group can have an
616 attached generator that generates requests if idle. The idea behind 1372 attached generator that generates requests if idle. The idea behind
617 this is that, although you could just queue as many requests as you 1373 this is that, although you could just queue as many requests as you
618 want in a group, this might starve other requests for a potentially 1374 want in a group, this might starve other requests for a potentially
619 long time. For example, "aio_scandir" might generate hundreds of 1375 long time. For example, "aio_scandir" might generate hundreds of
620 thousands "aio_stat" requests, delaying any later requests for a 1376 thousands of "aio_stat" requests, delaying any later requests for a
621 long time. 1377 long time.
622 1378
623 To avoid this, and allow incremental generation of requests, you can 1379 To avoid this, and allow incremental generation of requests, you can
624 instead a group and set a feeder on it that generates those 1380 instead a group and set a feeder on it that generates those
625 requests. The feed callback will be called whenever there are few 1381 requests. The feed callback will be called whenever there are few
… …
630 does not impose any limits). 1386 does not impose any limits).
631 1387
632 If the feed does not queue more requests when called, it will be 1388 If the feed does not queue more requests when called, it will be
633 automatically removed from the group. 1389 automatically removed from the group.
634 1390
635 If the feed limit is 0, it will be set to 2 automatically. 1391 If the feed limit is 0 when this method is called, it will be set to
1392 2 automatically.
636 1393
637 Example: 1394 Example:
638 1395
639 # stat all files in @files, but only ever use four aio requests concurrently: 1396 # stat all files in @files, but only ever use four aio requests concurrently:
640 1397
… …
651 Sets the feeder limit for the group: The feeder will be called 1408 Sets the feeder limit for the group: The feeder will be called
652 whenever the group contains less than this many requests. 1409 whenever the group contains less than this many requests.
653 1410
654 Setting the limit to 0 will pause the feeding process. 1411 Setting the limit to 0 will pause the feeding process.
655 1412
1413 The default value for the limit is 0, but note that setting a feeder
1414 automatically bumps it up to 2.
1415
656 SUPPORT FUNCTIONS 1416 SUPPORT FUNCTIONS
657 EVENT PROCESSING AND EVENT LOOP INTEGRATION 1417 EVENT PROCESSING AND EVENT LOOP INTEGRATION
658 $fileno = IO::AIO::poll_fileno 1418 $fileno = IO::AIO::poll_fileno
659 Return the *request result pipe file descriptor*. This filehandle 1419 Return the *request result pipe file descriptor*. This filehandle
660 must be polled for reading by some mechanism outside this module 1420 must be polled for reading by some mechanism outside this module
661 (e.g. Event or select, see below or the SYNOPSIS). If the pipe 1421 (e.g. EV, Glib, select and so on, see below or the SYNOPSIS). If the
662 becomes readable you have to call "poll_cb" to check the results. 1422 pipe becomes readable you have to call "poll_cb" to check the
1423 results.
663 1424
664 See "poll_cb" for an example. 1425 See "poll_cb" for an example.
665 1426
666 IO::AIO::poll_cb 1427 IO::AIO::poll_cb
667 Process some outstanding events on the result pipe. You have to call 1428 Process some requests that have reached the result phase (i.e. they
668 this regularly. Returns the number of events processed. Returns 1429 have been executed but the results are not yet reported). You have
1430 to call this "regularly" to finish outstanding requests.
1431
1432 Returns 0 if all events could be processed (or there were no events
1433 to process), or -1 if it returned earlier for whatever reason.
669 immediately when no events are outstanding. The amount of events 1434 Returns immediately when no events are outstanding. The amount of
670 processed depends on the settings of "IO::AIO::max_poll_req" and 1435 events processed depends on the settings of "IO::AIO::max_poll_req",
671 "IO::AIO::max_poll_time". 1436 "IO::AIO::max_poll_time" and "IO::AIO::max_outstanding".
672 1437
673 If not all requests were processed for whatever reason, the 1438 If not all requests were processed for whatever reason, the poll
674 filehandle will still be ready when "poll_cb" returns. 1439 file descriptor will still be ready when "poll_cb" returns, so
1440 normally you don't have to do anything special to have it called
1441 later.
1442
1443 Apart from calling "IO::AIO::poll_cb" when the event filehandle
1444 becomes ready, it can be beneficial to call this function from loops
1445 which submit a lot of requests, to make sure the results get
1446 processed when they become available and not just when the loop is
1447 finished and the event loop takes over again. This function returns
1448 very fast when there are no outstanding requests.
675 1449
676 Example: Install an Event watcher that automatically calls 1450 Example: Install an Event watcher that automatically calls
677 IO::AIO::poll_cb with high priority: 1451 IO::AIO::poll_cb with high priority (more examples can be found in
1452 the SYNOPSIS section, at the top of this document):
678 1453
679 Event->io (fd => IO::AIO::poll_fileno, 1454 Event->io (fd => IO::AIO::poll_fileno,
680 poll => 'r', async => 1, 1455 poll => 'r', async => 1,
681 cb => \&IO::AIO::poll_cb); 1456 cb => \&IO::AIO::poll_cb);
1457
1458 IO::AIO::poll_wait
1459 Wait until either at least one request is in the result phase or no
1460 requests are outstanding anymore.
1461
1462 This is useful if you want to synchronously wait for some requests
1463 to become ready, without actually handling them.
1464
1465 See "nreqs" for an example.
1466
1467 IO::AIO::poll
1468 Waits until some requests have been handled.
1469
1470 Returns the number of requests processed, but is otherwise strictly
1471 equivalent to:
1472
1473 IO::AIO::poll_wait, IO::AIO::poll_cb
1474
1475 IO::AIO::flush
1476 Wait till all outstanding AIO requests have been handled.
1477
1478 Strictly equivalent to:
1479
1480 IO::AIO::poll_wait, IO::AIO::poll_cb
1481 while IO::AIO::nreqs;
1482
1483 This function can be useful at program aborts, to make sure
1484 outstanding I/O has been done ("IO::AIO" uses an "END" block which
1485 already calls this function on normal exits), or when you are merely
1486 using "IO::AIO" for its more advanced functions, rather than for
1487 async I/O, e.g.:
1488
1489 my ($dirs, $nondirs);
1490 IO::AIO::aio_scandir "/tmp", 0, sub { ($dirs, $nondirs) = @_ };
1491 IO::AIO::flush;
1492 # $dirs, $nondirs are now set
682 1493
683 IO::AIO::max_poll_reqs $nreqs 1494 IO::AIO::max_poll_reqs $nreqs
684 IO::AIO::max_poll_time $seconds 1495 IO::AIO::max_poll_time $seconds
685 These set the maximum number of requests (default 0, meaning 1496 These set the maximum number of requests (default 0, meaning
686 infinity) that are being processed by "IO::AIO::poll_cb" in one 1497 infinity) that are being processed by "IO::AIO::poll_cb" in one
… …
710 # use a low priority so other tasks have priority 1521 # use a low priority so other tasks have priority
711 Event->io (fd => IO::AIO::poll_fileno, 1522 Event->io (fd => IO::AIO::poll_fileno,
712 poll => 'r', nice => 1, 1523 poll => 'r', nice => 1,
713 cb => &IO::AIO::poll_cb); 1524 cb => &IO::AIO::poll_cb);
714 1525
715 IO::AIO::poll_wait
716 Wait till the result filehandle becomes ready for reading (simply
717 does a "select" on the filehandle. This is useful if you want to
718 synchronously wait for some requests to finish).
719
720 See "nreqs" for an example.
721
722 IO::AIO::poll
723 Waits until some requests have been handled.
724
725 Strictly equivalent to:
726
727 IO::AIO::poll_wait, IO::AIO::poll_cb
728 if IO::AIO::nreqs;
729
730 IO::AIO::flush
731 Wait till all outstanding AIO requests have been handled.
732
733 Strictly equivalent to:
734
735 IO::AIO::poll_wait, IO::AIO::poll_cb
736 while IO::AIO::nreqs;
737
738 CONTROLLING THE NUMBER OF THREADS 1526 CONTROLLING THE NUMBER OF THREADS
739 IO::AIO::min_parallel $nthreads 1527 IO::AIO::min_parallel $nthreads
740 Set the minimum number of AIO threads to $nthreads. The current 1528 Set the minimum number of AIO threads to $nthreads. The current
741 default is 8, which means eight asynchronous operations can execute 1529 default is 8, which means eight asynchronous operations can execute
742 concurrently at any one time (the number of outstanding requests, 1530 concurrently at any one time (the number of outstanding requests,
… …
771 1559
772 Under normal circumstances you don't need to call this function. 1560 Under normal circumstances you don't need to call this function.
773 1561
774 IO::AIO::max_idle $nthreads 1562 IO::AIO::max_idle $nthreads
775 Limit the number of threads (default: 4) that are allowed to idle 1563 Limit the number of threads (default: 4) that are allowed to idle
776 (i.e., threads that did not get a request to process within 10 1564 (i.e., threads that did not get a request to process within the idle
777 seconds). That means if a thread becomes idle while $nthreads other 1565 timeout (default: 10 seconds). That means if a thread becomes idle
778 threads are also idle, it will free its resources and exit. 1566 while $nthreads other threads are also idle, it will free its
1567 resources and exit.
779 1568
780 This is useful when you allow a large number of threads (e.g. 100 or 1569 This is useful when you allow a large number of threads (e.g. 100 or
781 1000) to allow for extremely high load situations, but want to free 1570 1000) to allow for extremely high load situations, but want to free
782 resources under normal circumstances (1000 threads can easily 1571 resources under normal circumstances (1000 threads can easily
783 consume 30MB of RAM). 1572 consume 30MB of RAM).
784 1573
785 The default is probably ok in most situations, especially if thread 1574 The default is probably ok in most situations, especially if thread
786 creation is fast. If thread creation is very slow on your system you 1575 creation is fast. If thread creation is very slow on your system you
787 might want to use larger values. 1576 might want to use larger values.
788 1577
1578 IO::AIO::idle_timeout $seconds
1579 Sets the minimum idle timeout (default 10) after which worker
1580 threads are allowed to exit. SEe "IO::AIO::max_idle".
1581
789 $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 1582 IO::AIO::max_outstanding $maxreqs
1583 Sets the maximum number of outstanding requests to $nreqs. If you do
1584 queue up more than this number of requests, the next call to
1585 "IO::AIO::poll_cb" (and other functions calling "poll_cb", such as
1586 "IO::AIO::flush" or "IO::AIO::poll") will block until the limit is
1587 no longer exceeded.
1588
1589 In other words, this setting does not enforce a queue limit, but can
1590 be used to make poll functions block if the limit is exceeded.
1591
790 This is a very bad function to use in interactive programs because 1592 This is a very bad function to use in interactive programs because
791 it blocks, and a bad way to reduce concurrency because it is 1593 it blocks, and a bad way to reduce concurrency because it is
792 inexact: Better use an "aio_group" together with a feed callback. 1594 inexact: Better use an "aio_group" together with a feed callback.
793 1595
794 Sets the maximum number of outstanding requests to $nreqs. If you to 1596 Its main use is in scripts without an event loop - when you want to
795 queue up more than this number of requests, the next call to the 1597 stat a lot of files, you can write something like this:
796 "poll_cb" (and "poll_some" and other functions calling "poll_cb")
797 function will block until the limit is no longer exceeded.
798 1598
799 The default value is very large, so there is no practical limit on 1599 IO::AIO::max_outstanding 32;
1600
1601 for my $path (...) {
1602 aio_stat $path , ...;
1603 IO::AIO::poll_cb;
1604 }
1605
1606 IO::AIO::flush;
1607
1608 The call to "poll_cb" inside the loop will normally return
1609 instantly, but as soon as more thna 32 reqeusts are in-flight, it
1610 will block until some requests have been handled. This keeps the
1611 loop from pushing a large number of "aio_stat" requests onto the
1612 queue.
1613
1614 The default value for "max_outstanding" is very large, so there is
800 the number of outstanding requests. 1615 no practical limit on the number of outstanding requests.
801
802 You can still queue as many requests as you want. Therefore,
803 "max_oustsanding" is mainly useful in simple scripts (with low
804 values) or as a stop gap to shield against fatal memory overflow
805 (with large values).
806 1616
807 STATISTICAL INFORMATION 1617 STATISTICAL INFORMATION
808 IO::AIO::nreqs 1618 IO::AIO::nreqs
809 Returns the number of requests currently in the ready, execute or 1619 Returns the number of requests currently in the ready, execute or
810 pending states (i.e. for which their callback has not been invoked 1620 pending states (i.e. for which their callback has not been invoked
… …
821 1631
822 IO::AIO::npending 1632 IO::AIO::npending
823 Returns the number of requests currently in the pending state 1633 Returns the number of requests currently in the pending state
824 (executed, but not yet processed by poll_cb). 1634 (executed, but not yet processed by poll_cb).
825 1635
1636 SUBSECOND STAT TIME ACCESS
1637 Both "aio_stat"/"aio_lstat" and perl's "stat"/"lstat" functions can
1638 generally find access/modification and change times with subsecond time
1639 accuracy of the system supports it, but perl's built-in functions only
1640 return the integer part.
1641
1642 The following functions return the timestamps of the most recent stat
1643 with subsecond precision on most systems and work both after
1644 "aio_stat"/"aio_lstat" and perl's "stat"/"lstat" calls. Their return
1645 value is only meaningful after a successful "stat"/"lstat" call, or
1646 during/after a successful "aio_stat"/"aio_lstat" callback.
1647
1648 This is similar to the Time::HiRes "stat" functions, but can return full
1649 resolution without rounding and work with standard perl "stat",
1650 alleviating the need to call the special "Time::HiRes" functions, which
1651 do not act like their perl counterparts.
1652
1653 On operating systems or file systems where subsecond time resolution is
1654 not supported or could not be detected, a fractional part of 0 is
1655 returned, so it is always safe to call these functions.
1656
1657 $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime,
1658 IO::AIO::st_btime
1659 Return the access, modication, change or birth time, respectively,
1660 including fractional part. Due to the limited precision of floating
1661 point, the accuracy on most platforms is only a bit better than
1662 milliseconds for times around now - see the *nsec* function family,
1663 below, for full accuracy.
1664
1665 File birth time is only available when the OS and perl support it
1666 (on FreeBSD and NetBSD at the time of this writing, although support
1667 is adaptive, so if your OS/perl gains support, IO::AIO can take
1668 advantage of it). On systems where it isn't available, 0 is
1669 currently returned, but this might change to "undef" in a future
1670 version.
1671
1672 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime
1673 Returns access, modification, change and birth time all in one go,
1674 and maybe more times in the future version.
1675
1676 $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec,
1677 IO::AIO::st_ctimensec, IO::AIO::st_btimensec
1678 Return the fractional access, modifcation, change or birth time, in
1679 nanoseconds, as an integer in the range 0 to 999999999.
1680
1681 Note that no accessors are provided for access, modification and
1682 change times - you need to get those from "stat _" if required ("int
1683 IO::AIO::st_atime" and so on will *not* generally give you the
1684 correct value).
1685
1686 $seconds = IO::AIO::st_btimesec
1687 The (integral) seconds part of the file birth time, if available.
1688
1689 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec
1690 Like the functions above, but returns all four times in one go (and
1691 maybe more in future versions).
1692
1693 $counter = IO::AIO::st_gen
1694 Returns the generation counter (in practice this is just a random
1695 number) of the file. This is only available on platforms which have
1696 this member in their "struct stat" (most BSDs at the time of this
1697 writing) and generally only to the root usert. If unsupported, 0 is
1698 returned, but this might change to "undef" in a future version.
1699
1700 Example: print the high resolution modification time of /etc, using
1701 "stat", and "IO::AIO::aio_stat".
1702
1703 if (stat "/etc") {
1704 printf "stat(/etc) mtime: %f\n", IO::AIO::st_mtime;
1705 }
1706
1707 IO::AIO::aio_stat "/etc", sub {
1708 $_[0]
1709 and return;
1710
1711 printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::st_mtimensec;
1712 };
1713
1714 IO::AIO::flush;
1715
1716 Output of the awbove on my system, showing reduced and full accuracy:
1717
1718 stat(/etc) mtime: 1534043702.020808
1719 aio_stat(/etc) mtime: 1534043702.020807792
1720
1721 MISCELLANEOUS FUNCTIONS
1722 IO::AIO implements some functions that are useful when you want to use
1723 some "Advanced I/O" function not available to in Perl, without going the
1724 "Asynchronous I/O" route. Many of these have an asynchronous "aio_*"
1725 counterpart.
1726
1727 $numfd = IO::AIO::get_fdlimit
1728 This function is *EXPERIMENTAL* and subject to change.
1729
1730 Tries to find the current file descriptor limit and returns it, or
1731 "undef" and sets $! in case of an error. The limit is one larger
1732 than the highest valid file descriptor number.
1733
1734 IO::AIO::min_fdlimit [$numfd]
1735 This function is *EXPERIMENTAL* and subject to change.
1736
1737 Try to increase the current file descriptor limit(s) to at least
1738 $numfd by changing the soft or hard file descriptor resource limit.
1739 If $numfd is missing, it will try to set a very high limit, although
1740 this is not recommended when you know the actual minimum that you
1741 require.
1742
1743 If the limit cannot be raised enough, the function makes a
1744 best-effort attempt to increase the limit as much as possible, using
1745 various tricks, while still failing. You can query the resulting
1746 limit using "IO::AIO::get_fdlimit".
1747
1748 If an error occurs, returns "undef" and sets $!, otherwise returns
1749 true.
1750
1751 IO::AIO::sendfile $ofh, $ifh, $offset, $count
1752 Calls the "eio_sendfile_sync" function, which is like
1753 "aio_sendfile", but is blocking (this makes most sense if you know
1754 the input data is likely cached already and the output filehandle is
1755 set to non-blocking operations).
1756
1757 Returns the number of bytes copied, or -1 on error.
1758
1759 IO::AIO::fadvise $fh, $offset, $len, $advice
1760 Simply calls the "posix_fadvise" function (see its manpage for
1761 details). The following advice constants are available:
1762 "IO::AIO::FADV_NORMAL", "IO::AIO::FADV_SEQUENTIAL",
1763 "IO::AIO::FADV_RANDOM", "IO::AIO::FADV_NOREUSE",
1764 "IO::AIO::FADV_WILLNEED", "IO::AIO::FADV_DONTNEED".
1765
1766 On systems that do not implement "posix_fadvise", this function
1767 returns ENOSYS, otherwise the return value of "posix_fadvise".
1768
1769 IO::AIO::madvise $scalar, $offset, $len, $advice
1770 Simply calls the "posix_madvise" function (see its manpage for
1771 details). The following advice constants are available:
1772 "IO::AIO::MADV_NORMAL", "IO::AIO::MADV_SEQUENTIAL",
1773 "IO::AIO::MADV_RANDOM", "IO::AIO::MADV_WILLNEED",
1774 "IO::AIO::MADV_DONTNEED".
1775
1776 If $offset is negative, counts from the end. If $length is negative,
1777 the remaining length of the $scalar is used. If possible, $length
1778 will be reduced to fit into the $scalar.
1779
1780 On systems that do not implement "posix_madvise", this function
1781 returns ENOSYS, otherwise the return value of "posix_madvise".
1782
1783 IO::AIO::mprotect $scalar, $offset, $len, $protect
1784 Simply calls the "mprotect" function on the preferably AIO::mmap'ed
1785 $scalar (see its manpage for details). The following protect
1786 constants are available: "IO::AIO::PROT_NONE", "IO::AIO::PROT_READ",
1787 "IO::AIO::PROT_WRITE", "IO::AIO::PROT_EXEC".
1788
1789 If $offset is negative, counts from the end. If $length is negative,
1790 the remaining length of the $scalar is used. If possible, $length
1791 will be reduced to fit into the $scalar.
1792
1793 On systems that do not implement "mprotect", this function returns
1794 ENOSYS, otherwise the return value of "mprotect".
1795
1796 IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1797 Memory-maps a file (or anonymous memory range) and attaches it to
1798 the given $scalar, which will act like a string scalar. Returns true
1799 on success, and false otherwise.
1800
1801 The scalar must exist, but its contents do not matter - this means
1802 you cannot use a nonexistant array or hash element. When in doubt,
1803 "undef" the scalar first.
1804
1805 The only operations allowed on the mmapped scalar are
1806 "substr"/"vec", which don't change the string length, and most
1807 read-only operations such as copying it or searching it with regexes
1808 and so on.
1809
1810 Anything else is unsafe and will, at best, result in memory leaks.
1811
1812 The memory map associated with the $scalar is automatically removed
1813 when the $scalar is undef'd or destroyed, or when the
1814 "IO::AIO::mmap" or "IO::AIO::munmap" functions are called on it.
1815
1816 This calls the "mmap"(2) function internally. See your system's
1817 manual page for details on the $length, $prot and $flags parameters.
1818
1819 The $length must be larger than zero and smaller than the actual
1820 filesize.
1821
1822 $prot is a combination of "IO::AIO::PROT_NONE",
1823 "IO::AIO::PROT_EXEC", "IO::AIO::PROT_READ" and/or
1824 "IO::AIO::PROT_WRITE",
1825
1826 $flags can be a combination of "IO::AIO::MAP_SHARED" or
1827 "IO::AIO::MAP_PRIVATE", or a number of system-specific flags (when
1828 not available, the are 0): "IO::AIO::MAP_ANONYMOUS" (which is set to
1829 "MAP_ANON" if your system only provides this constant),
1830 "IO::AIO::MAP_LOCKED", "IO::AIO::MAP_NORESERVE",
1831 "IO::AIO::MAP_POPULATE", "IO::AIO::MAP_NONBLOCK",
1832 "IO::AIO::MAP_FIXED", "IO::AIO::MAP_GROWSDOWN",
1833 "IO::AIO::MAP_32BIT", "IO::AIO::MAP_HUGETLB" or
1834 "IO::AIO::MAP_STACK".
1835
1836 If $fh is "undef", then a file descriptor of -1 is passed.
1837
1838 $offset is the offset from the start of the file - it generally must
1839 be a multiple of "IO::AIO::PAGESIZE" and defaults to 0.
1840
1841 Example:
1842
1843 use Digest::MD5;
1844 use IO::AIO;
1845
1846 open my $fh, "<verybigfile"
1847 or die "$!";
1848
1849 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1850 or die "verybigfile: $!";
1851
1852 my $fast_md5 = md5 $data;
1853
1854 IO::AIO::munmap $scalar
1855 Removes a previous mmap and undefines the $scalar.
1856
1857 IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[,
1858 $new_address = 0]
1859 Calls the Linux-specific mremap(2) system call. The $scalar must
1860 have been mapped by "IO::AIO::mmap", and $flags must currently
1861 either be 0 or "IO::AIO::MREMAP_MAYMOVE".
1862
1863 Returns true if successful, and false otherwise. If the underlying
1864 mmapped region has changed address, then the true value has the
1865 numerical value 1, otherwise it has the numerical value 0:
1866
1867 my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE
1868 or die "mremap: $!";
1869
1870 if ($success*1) {
1871 warn "scalar has chanegd address in memory\n";
1872 }
1873
1874 "IO::AIO::MREMAP_FIXED" and the $new_address argument are currently
1875 implemented, but not supported and might go away in a future
1876 version.
1877
1878 On systems where this call is not supported or is not emulated, this
1879 call returns falls and sets $! to "ENOSYS".
1880
1881 IO::AIO::mlockall $flags
1882 Calls the "eio_mlockall_sync" function, which is like
1883 "aio_mlockall", but is blocking.
1884
1885 IO::AIO::munlock $scalar, $offset = 0, $length = undef
1886 Calls the "munlock" function, undoing the effects of a previous
1887 "aio_mlock" call (see its description for details).
1888
1889 IO::AIO::munlockall
1890 Calls the "munlockall" function.
1891
1892 On systems that do not implement "munlockall", this function returns
1893 ENOSYS, otherwise the return value of "munlockall".
1894
1895 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
1896 Calls the GNU/Linux splice(2) syscall, if available. If $r_off or
1897 $w_off are "undef", then "NULL" is passed for these, otherwise they
1898 should be the file offset.
1899
1900 $r_fh and $w_fh should not refer to the same file, as splice might
1901 silently corrupt the data in this case.
1902
1903 The following symbol flag values are available:
1904 "IO::AIO::SPLICE_F_MOVE", "IO::AIO::SPLICE_F_NONBLOCK",
1905 "IO::AIO::SPLICE_F_MORE" and "IO::AIO::SPLICE_F_GIFT".
1906
1907 See the splice(2) manpage for details.
1908
1909 IO::AIO::tee $r_fh, $w_fh, $length, $flags
1910 Calls the GNU/Linux tee(2) syscall, see its manpage and the
1911 description for "IO::AIO::splice" above for details.
1912
1913 $actual_size = IO::AIO::pipesize $r_fh[, $new_size]
1914 Attempts to query or change the pipe buffer size. Obviously works
1915 only on pipes, and currently works only on GNU/Linux systems, and
1916 fails with -1/"ENOSYS" everywhere else. If anybody knows how to
1917 influence pipe buffer size on other systems, drop me a note.
1918
1919 ($rfh, $wfh) = IO::AIO::pipe2 [$flags]
1920 This is a direct interface to the Linux pipe2(2) system call. If
1921 $flags is missing or 0, then this should be the same as a call to
1922 perl's built-in "pipe" function and create a new pipe, and works on
1923 systems that lack the pipe2 syscall. On win32, this case invokes
1924 "_pipe (..., 4096, O_BINARY)".
1925
1926 If $flags is non-zero, it tries to invoke the pipe2 system call with
1927 the given flags (Linux 2.6.27, glibc 2.9).
1928
1929 On success, the read and write file handles are returned.
1930
1931 On error, nothing will be returned. If the pipe2 syscall is missing
1932 and $flags is non-zero, fails with "ENOSYS".
1933
1934 Please refer to pipe2(2) for more info on the $flags, but at the
1935 time of this writing, "IO::AIO::O_CLOEXEC", "IO::AIO::O_NONBLOCK"
1936 and "IO::AIO::O_DIRECT" (Linux 3.4, for packet-based pipes) were
1937 supported.
1938
1939 Example: create a pipe race-free w.r.t. threads and fork:
1940
1941 my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC
1942 or die "pipe2: $!\n";
1943
1944 $fh = IO::AIO::memfd_create $pathname[, $flags]
1945 This is a direct interface to the Linux memfd_create(2) system call.
1946 The (unhelpful) default for $flags is 0, but your default should be
1947 "IO::AIO::MFD_CLOEXEC".
1948
1949 On success, the new memfd filehandle is returned, otherwise returns
1950 "undef". If the memfd_create syscall is missing, fails with
1951 "ENOSYS".
1952
1953 Please refer to memfd_create(2) for more info on this call.
1954
1955 The following $flags values are available: "IO::AIO::MFD_CLOEXEC",
1956 "IO::AIO::MFD_ALLOW_SEALING" and "IO::AIO::MFD_HUGETLB".
1957
1958 Example: create a new memfd.
1959
1960 my $fh = IO::AIO::memfd_create "somenameforprocfd", IO::AIO::MFD_CLOEXEC
1961 or die "m,emfd_create: $!\n";
1962 =item $fh = IO::AIO::eventfd [$initval, [$flags]]
1963
1964 This is a direct interface to the Linux eventfd(2) system call. The
1965 (unhelpful) defaults for $initval and $flags are 0 for both.
1966
1967 On success, the new eventfd filehandle is returned, otherwise
1968 returns "undef". If the eventfd syscall is missing, fails with
1969 "ENOSYS".
1970
1971 Please refer to eventfd(2) for more info on this call.
1972
1973 The following symbol flag values are available:
1974 "IO::AIO::EFD_CLOEXEC", "IO::AIO::EFD_NONBLOCK" and
1975 "IO::AIO::EFD_SEMAPHORE" (Linux 2.6.30).
1976
1977 Example: create a new eventfd filehandle:
1978
1979 $fh = IO::AIO::eventfd 0, IO::AIO::EFD_CLOEXEC
1980 or die "eventfd: $!\n";
1981
1982 $fh = IO::AIO::timerfd_create $clockid[, $flags]
1983 This is a direct interface to the Linux timerfd_create(2) system
1984 call. The (unhelpful) default for $flags is 0, but your default
1985 should be "IO::AIO::TFD_CLOEXEC".
1986
1987 On success, the new timerfd filehandle is returned, otherwise
1988 returns "undef". If the timerfd_create syscall is missing, fails
1989 with "ENOSYS".
1990
1991 Please refer to timerfd_create(2) for more info on this call.
1992
1993 The following $clockid values are available:
1994 "IO::AIO::CLOCK_REALTIME", "IO::AIO::CLOCK_MONOTONIC"
1995 "IO::AIO::CLOCK_CLOCK_BOOTTIME" (Linux 3.15)
1996 "IO::AIO::CLOCK_CLOCK_REALTIME_ALARM" (Linux 3.11) and
1997 "IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM" (Linux 3.11).
1998
1999 The following $flags values are available (Linux 2.6.27):
2000 "IO::AIO::TFD_NONBLOCK" and "IO::AIO::TFD_CLOEXEC".
2001
2002 Example: create a new timerfd and set it to one-second repeated
2003 alarms, then wait for two alarms:
2004
2005 my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC
2006 or die "timerfd_create: $!\n";
2007
2008 defined IO::AIO::timerfd_settime $fh, 0, 1, 1
2009 or die "timerfd_settime: $!\n";
2010
2011 for (1..2) {
2012 8 == sysread $fh, my $buf, 8
2013 or die "timerfd read failure\n";
2014
2015 printf "number of expirations (likely 1): %d\n",
2016 unpack "Q", $buf;
2017 }
2018
2019 ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags,
2020 $new_interval, $nbw_value
2021 This is a direct interface to the Linux timerfd_settime(2) system
2022 call. Please refer to its manpage for more info on this call.
2023
2024 The new itimerspec is specified using two (possibly fractional)
2025 second values, $new_interval and $new_value).
2026
2027 On success, the current interval and value are returned (as per
2028 "timerfd_gettime"). On failure, the empty list is returned.
2029
2030 The following $flags values are available:
2031 "IO::AIO::TFD_TIMER_ABSTIME" and "IO::AIO::TFD_TIMER_CANCEL_ON_SET".
2032
2033 See "IO::AIO::timerfd_create" for a full example.
2034
2035 ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh
2036 This is a direct interface to the Linux timerfd_gettime(2) system
2037 call. Please refer to its manpage for more info on this call.
2038
2039 On success, returns the current values of interval and value for the
2040 given timerfd (as potentially fractional second values). On failure,
2041 the empty list is returned.
2042
2043EVENT LOOP INTEGRATION
2044 It is recommended to use AnyEvent::AIO to integrate IO::AIO
2045 automatically into many event loops:
2046
2047 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
2048 use AnyEvent::AIO;
2049
2050 You can also integrate IO::AIO manually into many event loops, here are
2051 some examples of how to do this:
2052
2053 # EV integration
2054 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
2055
2056 # Event integration
2057 Event->io (fd => IO::AIO::poll_fileno,
2058 poll => 'r',
2059 cb => \&IO::AIO::poll_cb);
2060
2061 # Glib/Gtk2 integration
2062 add_watch Glib::IO IO::AIO::poll_fileno,
2063 in => sub { IO::AIO::poll_cb; 1 };
2064
2065 # Tk integration
2066 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
2067 readable => \&IO::AIO::poll_cb);
2068
2069 # Danga::Socket integration
2070 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
2071 \&IO::AIO::poll_cb);
2072
826 FORK BEHAVIOUR 2073 FORK BEHAVIOUR
827 This module should do "the right thing" when the process using it forks: 2074 Usage of pthreads in a program changes the semantics of fork
2075 considerably. Specifically, only async-safe functions can be called
2076 after fork. Perl doesn't know about this, so in general, you cannot call
2077 fork with defined behaviour in perl if pthreads are involved. IO::AIO
2078 uses pthreads, so this applies, but many other extensions and (for
2079 inexplicable reasons) perl itself often is linked against pthreads, so
2080 this limitation applies to quite a lot of perls.
828 2081
829 Before the fork, IO::AIO enters a quiescent state where no requests can 2082 This module no longer tries to fight your OS, or POSIX. That means
830 be added in other threads and no results will be processed. After the 2083 IO::AIO only works in the process that loaded it. Forking is fully
831 fork the parent simply leaves the quiescent state and continues 2084 supported, but using IO::AIO in the child is not.
832 request/result processing, while the child frees the request/result
833 queue (so that the requests started before the fork will only be handled
834 in the parent). Threads will be started on demand until the limit set in
835 the parent process has been reached again.
836 2085
837 In short: the parent will, after a short pause, continue as if fork had 2086 You might get around by not *using* IO::AIO before (or after) forking.
838 not been called, while the child will act as if IO::AIO has not been 2087 You could also try to call the IO::AIO::reinit function in the child:
839 used yet. 2088
2089 IO::AIO::reinit
2090 Abandons all current requests and I/O threads and simply
2091 reinitialises all data structures. This is not an operation
2092 supported by any standards, but happens to work on GNU/Linux and
2093 some newer BSD systems.
2094
2095 The only reasonable use for this function is to call it after
2096 forking, if "IO::AIO" was used in the parent. Calling it while
2097 IO::AIO is active in the process will result in undefined behaviour.
2098 Calling it at any time will also result in any undefined (by POSIX)
2099 behaviour.
2100
2101 LINUX-SPECIFIC CALLS
2102 When a call is documented as "linux-specific" then this means it
2103 originated on GNU/Linux. "IO::AIO" will usually try to autodetect the
2104 availability and compatibility of such calls regardless of the platform
2105 it is compiled on, so platforms such as FreeBSD which often implement
2106 these calls will work. When in doubt, call them and see if they fail wth
2107 "ENOSYS".
840 2108
841 MEMORY USAGE 2109 MEMORY USAGE
842 Per-request usage: 2110 Per-request usage:
843 2111
844 Each aio request uses - depending on your architecture - around 100-200 2112 Each aio request uses - depending on your architecture - around 100-200
845 bytes of memory. In addition, stat requests need a stat buffer (possibly 2113 bytes of memory. In addition, stat requests need a stat buffer (possibly
846 a few hundred bytes), readdir requires a result buffer and so on. Perl 2114 a few hundred bytes), readdir requires a result buffer and so on. Perl
847 scalars and other data passed into aio requests will also be locked and 2115 scalars and other data passed into aio requests will also be locked and
848 will consume memory till the request has entered the done state. 2116 will consume memory till the request has entered the done state.
849 2117
850 This is now awfully much, so queuing lots of requests is not usually a 2118 This is not awfully much, so queuing lots of requests is not usually a
851 problem. 2119 problem.
852 2120
853 Per-thread usage: 2121 Per-thread usage:
854 2122
855 In the execution phase, some aio requests require more memory for 2123 In the execution phase, some aio requests require more memory for
856 temporary buffers, and each thread requires a stack and other data 2124 temporary buffers, and each thread requires a stack and other data
857 structures (usually around 16k-128k, depending on the OS). 2125 structures (usually around 16k-128k, depending on the OS).
858 2126
859KNOWN BUGS 2127KNOWN BUGS
860 Known bugs will be fixed in the next release. 2128 Known bugs will be fixed in the next release :)
2129
2130KNOWN ISSUES
2131 Calls that try to "import" foreign memory areas (such as "IO::AIO::mmap"
2132 or "IO::AIO::aio_slurp") do not work with generic lvalues, such as
2133 non-created hash slots or other scalars I didn't think of. It's best to
2134 avoid such and either use scalar variables or making sure that the
2135 scalar exists (e.g. by storing "undef") and isn't "funny" (e.g. tied).
2136
2137 I am not sure anything can be done about this, so this is considered a
2138 known issue, rather than a bug.
861 2139
862SEE ALSO 2140SEE ALSO
863 Coro::AIO. 2141 AnyEvent::AIO for easy integration into event loops, Coro::AIO for a
2142 more natural syntax.
864 2143
865AUTHOR 2144AUTHOR
866 Marc Lehmann <schmorp@schmorp.de> 2145 Marc Lehmann <schmorp@schmorp.de>
867 http://home.schmorp.de/ 2146 http://home.schmorp.de/
868 2147

Diff Legend

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