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

Comparing IO-AIO/README (file contents):
Revision 1.19 by root, Sun Oct 29 01:03:13 2006 UTC vs.
Revision 1.52 by root, Tue Apr 10 05:01:33 2012 UTC

… …
2 IO::AIO - Asynchronous Input/Output 2 IO::AIO - Asynchronous Input/Output
3 3
4SYNOPSIS 4SYNOPSIS
5 use IO::AIO; 5 use IO::AIO;
6 6
7 aio_open "/etc/passwd", O_RDONLY, 0, sub { 7 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
8 my ($fh) = @_; 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
… …
58 but can also be used to easily do operations in parallel that are 39 but can also be used to easily do operations in parallel that are
59 normally done sequentially, e.g. stat'ing many files, which is much 40 normally done sequentially, e.g. stat'ing many files, which is much
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 this works on all types of file descriptors (for example sockets), 44 While most of this works on all types of file descriptors (for example
64 using these functions on file descriptors that support nonblocking 45 sockets), using these functions on file descriptors that support
65 operation (again, sockets, pipes etc.) is very inefficient. Use an event 46 nonblocking operation (again, sockets, pipes etc.) is very inefficient.
66 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
67 into such an event loop itself. 48 naturally fit into such an event loop itself.
68 49
69 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
70 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
71 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
72 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
… …
74 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
75 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
76 aio_write, so the remaining functionality would have to be implemented 57 aio_write, so the remaining functionality would have to be implemented
77 using threads anyway. 58 using threads anyway.
78 59
79 Although the module will work with in the presence of other (Perl-) 60 Although the module will work in the presence of other (Perl-) threads,
80 threads, it is currently not reentrant in any way, so use appropriate 61 it is currently not reentrant in any way, so use appropriate locking
81 locking yourself, always call "poll_cb" from within the same thread, or 62 yourself, always call "poll_cb" from within the same thread, or never
82 never call "poll_cb" (or other "aio_" functions) recursively. 63 call "poll_cb" (or other "aio_" functions) recursively.
83 64
84 EXAMPLE 65 EXAMPLE
85 This is a simple example that uses the Event module and loads 66 This is a simple example that uses the EV module and loads /etc/passwd
86 /etc/passwd asynchronously: 67 asynchronously:
87 68
88 use Fcntl; 69 use Fcntl;
89 use Event; 70 use EV;
90 use IO::AIO; 71 use IO::AIO;
91 72
92 # register the IO::AIO callback with Event 73 # register the IO::AIO callback with EV
93 Event->io (fd => IO::AIO::poll_fileno, 74 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
94 poll => 'r',
95 cb => \&IO::AIO::poll_cb);
96 75
97 # queue the request to open /etc/passwd 76 # queue the request to open /etc/passwd
98 aio_open "/etc/passwd", O_RDONLY, 0, sub { 77 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
99 my $fh = $_[0] 78 my $fh = shift
100 or die "error while opening: $!"; 79 or die "error while opening: $!";
101 80
102 # stat'ing filehandles is generally non-blocking 81 # stat'ing filehandles is generally non-blocking
103 my $size = -s $fh; 82 my $size = -s $fh;
104 83
… …
112 91
113 # file contents now in $contents 92 # file contents now in $contents
114 print $contents; 93 print $contents;
115 94
116 # exit event loop and program 95 # exit event loop and program
117 Event::unloop; 96 EV::unloop;
118 }; 97 };
119 }; 98 };
120 99
121 # possibly queue up other requests, or open GUI windows, 100 # possibly queue up other requests, or open GUI windows,
122 # check for sockets etc. etc. 101 # check for sockets etc. etc.
123 102
124 # process events as long as there are some: 103 # process events as long as there are some:
125 Event::loop; 104 EV::loop;
126 105
127REQUEST ANATOMY AND LIFETIME 106REQUEST ANATOMY AND LIFETIME
128 Every "aio_*" function creates a request. which is a C data structure 107 Every "aio_*" function creates a request. which is a C data structure
129 not directly visible to Perl. 108 not directly visible to Perl.
130 109
… …
166 anymore (except possibly for the Perl object, but its connection to 145 anymore (except possibly for the Perl object, but its connection to
167 the actual aio request is severed and calling its methods will 146 the actual aio request is severed and calling its methods will
168 either do nothing or result in a runtime error). 147 either do nothing or result in a runtime error).
169 148
170FUNCTIONS 149FUNCTIONS
150 QUICK OVERVIEW
151 This section simply lists the prototypes of the most important functions
152 for quick reference. See the following sections for function-by-function
153 documentation.
154
155 aio_wd $pathname, $callback->($wd)
156 aio_open $pathname, $flags, $mode, $callback->($fh)
157 aio_close $fh, $callback->($status)
158 aio_seek $fh,$offset,$whence, $callback->($offs)
159 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
160 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
161 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
162 aio_readahead $fh,$offset,$length, $callback->($retval)
163 aio_stat $fh_or_path, $callback->($status)
164 aio_lstat $fh, $callback->($status)
165 aio_statvfs $fh_or_path, $callback->($statvfs)
166 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
167 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
168 aio_chmod $fh_or_path, $mode, $callback->($status)
169 aio_truncate $fh_or_path, $offset, $callback->($status)
170 aio_unlink $pathname, $callback->($status)
171 aio_mknod $pathname, $mode, $dev, $callback->($status)
172 aio_link $srcpath, $dstpath, $callback->($status)
173 aio_symlink $srcpath, $dstpath, $callback->($status)
174 aio_readlink $pathname, $callback->($link)
175 aio_realpath $pathname, $callback->($link)
176 aio_rename $srcpath, $dstpath, $callback->($status)
177 aio_mkdir $pathname, $mode, $callback->($status)
178 aio_rmdir $pathname, $callback->($status)
179 aio_readdir $pathname, $callback->($entries)
180 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
181 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST
182 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN
183 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
184 aio_load $pathname, $data, $callback->($status)
185 aio_copy $srcpath, $dstpath, $callback->($status)
186 aio_move $srcpath, $dstpath, $callback->($status)
187 aio_rmtree $pathname, $callback->($status)
188 aio_sync $callback->($status)
189 aio_syncfs $fh, $callback->($status)
190 aio_fsync $fh, $callback->($status)
191 aio_fdatasync $fh, $callback->($status)
192 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
193 aio_pathsync $pathname, $callback->($status)
194 aio_msync $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
195 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status)
196 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
197 aio_mlockall $flags, $callback->($status)
198 aio_group $callback->(...)
199 aio_nop $callback->()
200
201 $prev_pri = aioreq_pri [$pri]
202 aioreq_nice $pri_adjust
203
204 IO::AIO::poll_wait
205 IO::AIO::poll_cb
206 IO::AIO::poll
207 IO::AIO::flush
208 IO::AIO::max_poll_reqs $nreqs
209 IO::AIO::max_poll_time $seconds
210 IO::AIO::min_parallel $nthreads
211 IO::AIO::max_parallel $nthreads
212 IO::AIO::max_idle $nthreads
213 IO::AIO::idle_timeout $seconds
214 IO::AIO::max_outstanding $maxreqs
215 IO::AIO::nreqs
216 IO::AIO::nready
217 IO::AIO::npending
218
219 IO::AIO::sendfile $ofh, $ifh, $offset, $count
220 IO::AIO::fadvise $fh, $offset, $len, $advice
221 IO::AIO::madvise $scalar, $offset, $length, $advice
222 IO::AIO::mprotect $scalar, $offset, $length, $protect
223 IO::AIO::munlock $scalar, $offset = 0, $length = undef
224 IO::AIO::munlockall
225
226 API NOTES
227 All the "aio_*" calls are more or less thin wrappers around the syscall
228 with the same name (sans "aio_"). The arguments are similar or
229 identical, and they all accept an additional (and optional) $callback
230 argument which must be a code reference. This code reference will be
231 called after the syscall has been executed in an asynchronous fashion.
232 The results of the request will be passed as arguments to the callback
233 (and, if an error occured, in $!) - for most requests the syscall return
234 code (e.g. most syscalls return -1 on error, unlike perl, which usually
235 delivers "false").
236
237 Some requests (such as "aio_readdir") pass the actual results and
238 communicate failures by passing "undef".
239
240 All functions expecting a filehandle keep a copy of the filehandle
241 internally until the request has finished.
242
243 All functions return request objects of type IO::AIO::REQ that allow
244 further manipulation of those requests while they are in-flight.
245
246 The pathnames you pass to these routines *should* be absolute. The
247 reason for this is that at the time the request is being executed, the
248 current working directory could have changed. Alternatively, you can
249 make sure that you never change the current working directory anywhere
250 in the program and then use relative paths. You can also take advantage
251 of IO::AIOs working directory abstraction, that lets you specify paths
252 relative to some previously-opened "working directory object" - see the
253 description of the "IO::AIO::WD" class later in this document.
254
255 To encode pathnames as octets, either make sure you either: a) always
256 pass in filenames you got from outside (command line, readdir etc.)
257 without tinkering, b) are in your native filesystem encoding, c) use the
258 Encode module and encode your pathnames to the locale (or other)
259 encoding in effect in the user environment, d) use
260 Glib::filename_from_unicode on unicode filenames or e) use something
261 else to ensure your scalar has the correct contents.
262
263 This works, btw. independent of the internal UTF-8 bit, which IO::AIO
264 handles correctly whether it is set or not.
265
171 AIO REQUEST FUNCTIONS 266 AIO REQUEST FUNCTIONS
172 All the "aio_*" calls are more or less thin wrappers around the
173 syscall with the same name (sans "aio_"). The arguments are similar
174 or identical, and they all accept an additional (and optional)
175 $callback argument which must be a code reference. This code
176 reference will get called with the syscall return code (e.g. most
177 syscalls return -1 on error, unlike perl, which usually delivers
178 "false") as it's sole argument when the given syscall has been
179 executed asynchronously.
180
181 All functions expecting a filehandle keep a copy of the filehandle
182 internally until the request has finished.
183
184 All functions return request objects of type IO::AIO::REQ that allow
185 further manipulation of those requests while they are in-flight.
186
187 The pathnames you pass to these routines *must* be absolute and
188 encoded as octets. The reason for the former is that at the time the
189 request is being executed, the current working directory could have
190 changed. Alternatively, you can make sure that you never change the
191 current working directory anywhere in the program and then use
192 relative paths.
193
194 To encode pathnames as octets, either make sure you either: a)
195 always pass in filenames you got from outside (command line, readdir
196 etc.) without tinkering, b) are ASCII or ISO 8859-1, c) use the
197 Encode module and encode your pathnames to the locale (or other)
198 encoding in effect in the user environment, d) use
199 Glib::filename_from_unicode on unicode filenames or e) use something
200 else to ensure your scalar has the correct contents.
201
202 This works, btw. independent of the internal UTF-8 bit, which
203 IO::AIO handles correctly wether it is set or not.
204
205 $prev_pri = aioreq_pri [$pri] 267 $prev_pri = aioreq_pri [$pri]
206 Returns the priority value that would be used for the next 268 Returns the priority value that would be used for the next request
207 request and, if $pri is given, sets the priority for the next 269 and, if $pri is given, sets the priority for the next aio request.
208 aio request.
209 270
210 The default priority is 0, the minimum and maximum priorities 271 The default priority is 0, the minimum and maximum priorities are -4
211 are -4 and 4, respectively. Requests with higher priority will 272 and 4, respectively. Requests with higher priority will be serviced
212 be serviced first. 273 first.
213 274
214 The priority will be reset to 0 after each call to one of the 275 The priority will be reset to 0 after each call to one of the
215 "aio_*" functions. 276 "aio_*" functions.
216 277
217 Example: open a file with low priority, then read something from 278 Example: open a file with low priority, then read something from it
218 it with higher priority so the read request is serviced before 279 with higher priority so the read request is serviced before other
219 other low priority open requests (potentially spamming the 280 low priority open requests (potentially spamming the cache):
220 cache):
221 281
282 aioreq_pri -3;
283 aio_open ..., sub {
284 return unless $_[0];
285
222 aioreq_pri -3; 286 aioreq_pri -2;
223 aio_open ..., sub {
224 return unless $_[0];
225
226 aioreq_pri -2;
227 aio_read $_[0], ..., sub { 287 aio_read $_[0], ..., sub {
228 ...
229 };
230 };
231
232 aioreq_nice $pri_adjust
233 Similar to "aioreq_pri", but subtracts the given value from the
234 current priority, so the effect is cumulative.
235
236 aio_open $pathname, $flags, $mode, $callback->($fh)
237 Asynchronously open or create a file and call the callback with
238 a newly created filehandle for the file.
239
240 The pathname passed to "aio_open" must be absolute. See API
241 NOTES, above, for an explanation.
242
243 The $flags argument is a bitmask. See the "Fcntl" module for a
244 list. They are the same as used by "sysopen".
245
246 Likewise, $mode specifies the mode of the newly created file, if
247 it didn't exist and "O_CREAT" has been given, just like perl's
248 "sysopen", except that it is mandatory (i.e. use 0 if you don't
249 create new files, and 0666 or 0777 if you do).
250
251 Example:
252
253 aio_open "/etc/passwd", O_RDONLY, 0, sub {
254 if ($_[0]) {
255 print "open successful, fh is $_[0]\n";
256 ...
257 } else {
258 die "open failed: $!\n";
259 }
260 };
261
262 aio_close $fh, $callback->($status)
263 Asynchronously close a file and call the callback with the
264 result code. *WARNING:* although accepted, you should not pass
265 in a perl filehandle here, as perl will likely close the file
266 descriptor another time when the filehandle is destroyed.
267 Normally, you can safely call perls "close" or just let
268 filehandles go out of scope.
269
270 This is supposed to be a bug in the API, so that might change.
271 It's therefore best to avoid this function.
272
273 aio_read $fh,$offset,$length, $data,$dataoffset,
274 $callback->($retval)
275 aio_write $fh,$offset,$length, $data,$dataoffset,
276 $callback->($retval)
277 Reads or writes "length" bytes from the specified "fh" and
278 "offset" into the scalar given by "data" and offset "dataoffset"
279 and calls the callback without the actual number of bytes read
280 (or -1 on error, just like the syscall).
281
282 The $data scalar *MUST NOT* be modified in any way while the
283 request is outstanding. Modifying it can result in segfaults or
284 WW3 (if the necessary/optional hardware is installed).
285
286 Example: Read 15 bytes at offset 7 into scalar $buffer, starting
287 at offset 0 within the scalar:
288
289 aio_read $fh, 7, 15, $buffer, 0, sub {
290 $_[0] > 0 or die "read error: $!";
291 print "read $_[0] bytes: <$buffer>\n";
292 };
293
294 aio_sendfile $out_fh, $in_fh, $in_offset, $length,
295 $callback->($retval)
296 Tries to copy $length bytes from $in_fh to $out_fh. It starts
297 reading at byte offset $in_offset, and starts writing at the
298 current file offset of $out_fh. Because of that, it is not safe
299 to issue more than one "aio_sendfile" per $out_fh, as they will
300 interfere with each other.
301
302 This call tries to make use of a native "sendfile" syscall to
303 provide zero-copy operation. For this to work, $out_fh should
304 refer to a socket, and $in_fh should refer to mmap'able file.
305
306 If the native sendfile call fails or is not implemented, it will
307 be emulated, so you can call "aio_sendfile" on any type of
308 filehandle regardless of the limitations of the operating
309 system.
310
311 Please note, however, that "aio_sendfile" can read more bytes
312 from $in_fh than are written, and there is no way to find out
313 how many bytes have been read from "aio_sendfile" alone, as
314 "aio_sendfile" only provides the number of bytes written to
315 $out_fh. Only if the result value equals $length one can assume
316 that $length bytes have been read.
317
318 aio_readahead $fh,$offset,$length, $callback->($retval)
319 "aio_readahead" populates the page cache with data from a file
320 so that subsequent reads from that file will not block on disk
321 I/O. The $offset argument specifies the starting point from
322 which data is to be read and $length specifies the number of
323 bytes to be read. I/O is performed in whole pages, so that
324 offset is effectively rounded down to a page boundary and bytes
325 are read up to the next page boundary greater than or equal to
326 (off-set+length). "aio_readahead" does not read beyond the end
327 of the file. The current file offset of the file is left
328 unchanged.
329
330 If that syscall doesn't exist (likely if your OS isn't Linux) it
331 will be emulated by simply reading the data, which would have a
332 similar effect.
333
334 aio_stat $fh_or_path, $callback->($status)
335 aio_lstat $fh, $callback->($status)
336 Works like perl's "stat" or "lstat" in void context. The
337 callback will be called after the stat and the results will be
338 available using "stat _" or "-s _" etc...
339
340 The pathname passed to "aio_stat" must be absolute. See API
341 NOTES, above, for an explanation.
342
343 Currently, the stats are always 64-bit-stats, i.e. instead of
344 returning an error when stat'ing a large file, the results will
345 be silently truncated unless perl itself is compiled with large
346 file support.
347
348 Example: Print the length of /etc/passwd:
349
350 aio_stat "/etc/passwd", sub {
351 $_[0] and die "stat failed: $!";
352 print "size is ", -s _, "\n";
353 };
354
355 aio_unlink $pathname, $callback->($status)
356 Asynchronously unlink (delete) a file and call the callback with
357 the result code.
358
359 aio_mknod $path, $mode, $dev, $callback->($status)
360 [EXPERIMENTAL]
361
362 Asynchronously create a device node (or fifo). See mknod(2).
363
364 The only (POSIX-) portable way of calling this function is:
365
366 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...
367
368 aio_link $srcpath, $dstpath, $callback->($status)
369 Asynchronously create a new link to the existing object at
370 $srcpath at the path $dstpath and call the callback with the
371 result code.
372
373 aio_symlink $srcpath, $dstpath, $callback->($status)
374 Asynchronously create a new symbolic link to the existing object
375 at $srcpath at the path $dstpath and call the callback with the
376 result code.
377
378 aio_rename $srcpath, $dstpath, $callback->($status)
379 Asynchronously rename the object at $srcpath to $dstpath, just
380 as rename(2) and call the callback with the result code.
381
382 aio_rmdir $pathname, $callback->($status)
383 Asynchronously rmdir (delete) a directory and call the callback
384 with the result code.
385
386 aio_readdir $pathname, $callback->($entries)
387 Unlike the POSIX call of the same name, "aio_readdir" reads an
388 entire directory (i.e. opendir + readdir + closedir). The
389 entries will not be sorted, and will NOT include the "." and
390 ".." entries.
391
392 The callback a single argument which is either "undef" or an
393 array-ref with the filenames.
394
395 aio_copy $srcpath, $dstpath, $callback->($status)
396 Try to copy the *file* (directories not supported as either
397 source or destination) from $srcpath to $dstpath and call the
398 callback with the 0 (error) or -1 ok.
399
400 This is a composite request that it creates the destination file
401 with mode 0200 and copies the contents of the source file into
402 it using "aio_sendfile", followed by restoring atime, mtime,
403 access mode and uid/gid, in that order.
404
405 If an error occurs, the partial destination file will be
406 unlinked, if possible, except when setting atime, mtime, access
407 mode and uid/gid, where errors are being ignored.
408
409 aio_move $srcpath, $dstpath, $callback->($status)
410 Try to move the *file* (directories not supported as either
411 source or destination) from $srcpath to $dstpath and call the
412 callback with the 0 (error) or -1 ok.
413
414 This is a composite request that tries to rename(2) the file
415 first. If rename files with "EXDEV", it copies the file with
416 "aio_copy" and, if that is successful, unlinking the $srcpath.
417
418 aio_scandir $path, $maxreq, $callback->($dirs, $nondirs)
419 Scans a directory (similar to "aio_readdir") but additionally
420 tries to efficiently separate the entries of directory $path
421 into two sets of names, directories you can recurse into
422 (directories), and ones you cannot recurse into (everything
423 else, including symlinks to directories).
424
425 "aio_scandir" is a composite request that creates of many sub
426 requests_ $maxreq specifies the maximum number of outstanding
427 aio requests that this function generates. If it is "<= 0", then
428 a suitable default will be chosen (currently 4).
429
430 On error, the callback is called without arguments, otherwise it
431 receives two array-refs with path-relative entry names.
432
433 Example:
434
435 aio_scandir $dir, 0, sub {
436 my ($dirs, $nondirs) = @_;
437 print "real directories: @$dirs\n";
438 print "everything else: @$nondirs\n";
439 };
440
441 Implementation notes.
442
443 The "aio_readdir" cannot be avoided, but "stat()"'ing every
444 entry can.
445
446 After reading the directory, the modification time, size etc. of
447 the directory before and after the readdir is checked, and if
448 they match (and isn't the current time), the link count will be
449 used to decide how many entries are directories (if >= 2).
450 Otherwise, no knowledge of the number of subdirectories will be
451 assumed.
452
453 Then entries will be sorted into likely directories (everything
454 without a non-initial dot currently) and likely non-directories
455 (everything else). Then every entry plus an appended "/." will
456 be "stat"'ed, likely directories first. If that succeeds, it
457 assumes that the entry is a directory or a symlink to directory
458 (which will be checked seperately). This is often faster than
459 stat'ing the entry itself because filesystems might detect the
460 type of the entry without reading the inode data (e.g. ext2fs
461 filetype feature).
462
463 If the known number of directories (link count - 2) has been
464 reached, the rest of the entries is assumed to be
465 non-directories.
466
467 This only works with certainty on POSIX (= UNIX) filesystems,
468 which fortunately are the vast majority of filesystems around.
469
470 It will also likely work on non-POSIX filesystems with reduced
471 efficiency as those tend to return 0 or 1 as link counts, which
472 disables the directory counting heuristic.
473
474 aio_fsync $fh, $callback->($status)
475 Asynchronously call fsync on the given filehandle and call the
476 callback with the fsync result code.
477
478 aio_fdatasync $fh, $callback->($status)
479 Asynchronously call fdatasync on the given filehandle and call
480 the callback with the fdatasync result code.
481
482 If this call isn't available because your OS lacks it or it
483 couldn't be detected, it will be emulated by calling "fsync"
484 instead.
485
486 aio_group $callback->(...)
487 This is a very special aio request: Instead of doing something,
488 it is a container for other aio requests, which is useful if you
489 want to bundle many requests into a single, composite, request
490 with a definite callback and the ability to cancel the whole
491 request with its subrequests.
492
493 Returns an object of class IO::AIO::GRP. See its documentation
494 below for more info.
495
496 Example:
497
498 my $grp = aio_group sub {
499 print "all stats done\n";
500 };
501
502 add $grp
503 (aio_stat ...),
504 (aio_stat ...),
505 ...; 288 ...
506
507 aio_nop $callback->()
508 This is a special request - it does nothing in itself and is
509 only used for side effects, such as when you want to add a dummy
510 request to a group so that finishing the requests in the group
511 depends on executing the given code.
512
513 While this request does nothing, it still goes through the
514 execution phase and still requires a worker thread. Thus, the
515 callback will not be executed immediately but only after other
516 requests in the queue have entered their execution phase. This
517 can be used to measure request latency.
518
519 IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
520 Mainly used for debugging and benchmarking, this aio request
521 puts one of the request workers to sleep for the given time.
522
523 While it is theoretically handy to have simple I/O scheduling
524 requests like sleep and file handle readable/writable, the
525 overhead this creates is immense (it blocks a thread for a long
526 time) so do not use this function except to put your application
527 under artificial I/O pressure.
528
529 IO::AIO::REQ CLASS
530 All non-aggregate "aio_*" functions return an object of this class
531 when called in non-void context.
532
533 cancel $req
534 Cancels the request, if possible. Has the effect of skipping
535 execution when entering the execute state and skipping calling
536 the callback when entering the the result state, but will leave
537 the request otherwise untouched. That means that requests that
538 currently execute will not be stopped and resources held by the
539 request will not be freed prematurely.
540
541 cb $req $callback->(...)
542 Replace (or simply set) the callback registered to the request.
543
544 IO::AIO::GRP CLASS
545 This class is a subclass of IO::AIO::REQ, so all its methods apply
546 to objects of this class, too.
547
548 A IO::AIO::GRP object is a special request that can contain multiple
549 other aio requests.
550
551 You create one by calling the "aio_group" constructing function with
552 a callback that will be called when all contained requests have
553 entered the "done" state:
554
555 my $grp = aio_group sub {
556 print "all requests are done\n";
557 };
558
559 You add requests by calling the "add" method with one or more
560 "IO::AIO::REQ" objects:
561
562 $grp->add (aio_unlink "...");
563
564 add $grp aio_stat "...", sub {
565 $_[0] or return $grp->result ("error");
566
567 # add another request dynamically, if first succeeded
568 add $grp aio_open "...", sub {
569 $grp->result ("ok");
570 }; 289 };
571 }; 290 };
572 291
292 aioreq_nice $pri_adjust
293 Similar to "aioreq_pri", but subtracts the given value from the
294 current priority, so the effect is cumulative.
295
296 aio_open $pathname, $flags, $mode, $callback->($fh)
297 Asynchronously open or create a file and call the callback with a
298 newly created filehandle for the file.
299
300 The pathname passed to "aio_open" must be absolute. See API NOTES,
301 above, for an explanation.
302
303 The $flags argument is a bitmask. See the "Fcntl" module for a list.
304 They are the same as used by "sysopen".
305
306 Likewise, $mode specifies the mode of the newly created file, if it
307 didn't exist and "O_CREAT" has been given, just like perl's
308 "sysopen", except that it is mandatory (i.e. use 0 if you don't
309 create new files, and 0666 or 0777 if you do). Note that the $mode
310 will be modified by the umask in effect then the request is being
311 executed, so better never change the umask.
312
313 Example:
314
315 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub {
316 if ($_[0]) {
317 print "open successful, fh is $_[0]\n";
318 ...
319 } else {
320 die "open failed: $!\n";
321 }
322 };
323
324 In addition to all the common open modes/flags ("O_RDONLY",
325 "O_WRONLY", "O_RDWR", "O_CREAT", "O_TRUNC", "O_EXCL" and
326 "O_APPEND"), the following POSIX and non-POSIX constants are
327 available (missing ones on your system are, as usual, 0):
328
329 "O_ASYNC", "O_DIRECT", "O_NOATIME", "O_CLOEXEC", "O_NOCTTY",
330 "O_NOFOLLOW", "O_NONBLOCK", "O_EXEC", "O_SEARCH", "O_DIRECTORY",
331 "O_DSYNC", "O_RSYNC", "O_SYNC" and "O_TTY_INIT".
332
333 aio_close $fh, $callback->($status)
334 Asynchronously close a file and call the callback with the result
335 code.
336
337 Unfortunately, you can't do this to perl. Perl *insists* very
338 strongly on closing the file descriptor associated with the
339 filehandle itself.
340
341 Therefore, "aio_close" will not close the filehandle - instead it
342 will use dup2 to overwrite the file descriptor with the write-end of
343 a pipe (the pipe fd will be created on demand and will be cached).
344
345 Or in other words: the file descriptor will be closed, but it will
346 not be free for reuse until the perl filehandle is closed.
347
348 aio_seek $fh, $offset, $whence, $callback->($offs)
349 Seeks the filehandle to the new $offset, similarly to perl's
350 "sysseek". The $whence can use the traditional values (0 for
351 "IO::AIO::SEEK_SET", 1 for "IO::AIO::SEEK_CUR" or 2 for
352 "IO::AIO::SEEK_END").
353
354 The resulting absolute offset will be passed to the callback, or -1
355 in case of an error.
356
357 In theory, the $whence constants could be different than the
358 corresponding values from Fcntl, but perl guarantees they are the
359 same, so don't panic.
360
361 As a GNU/Linux (and maybe Solaris) extension, also the constants
362 "IO::AIO::SEEK_DATA" and "IO::AIO::SEEK_HOLE" are available, if they
363 could be found. No guarantees about suitability for use in
364 "aio_seek" or Perl's "sysseek" can be made though, although I would
365 naively assume they "just work".
366
367 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
368 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval)
369 Reads or writes $length bytes from or to the specified $fh and
370 $offset into the scalar given by $data and offset $dataoffset and
371 calls the callback without the actual number of bytes read (or -1 on
372 error, just like the syscall).
373
374 "aio_read" will, like "sysread", shrink or grow the $data scalar to
375 offset plus the actual number of bytes read.
376
377 If $offset is undefined, then the current file descriptor offset
378 will be used (and updated), otherwise the file descriptor offset
379 will not be changed by these calls.
380
381 If $length is undefined in "aio_write", use the remaining length of
382 $data.
383
384 If $dataoffset is less than zero, it will be counted from the end of
385 $data.
386
387 The $data scalar *MUST NOT* be modified in any way while the request
388 is outstanding. Modifying it can result in segfaults or World War
389 III (if the necessary/optional hardware is installed).
390
391 Example: Read 15 bytes at offset 7 into scalar $buffer, starting at
392 offset 0 within the scalar:
393
394 aio_read $fh, 7, 15, $buffer, 0, sub {
395 $_[0] > 0 or die "read error: $!";
396 print "read $_[0] bytes: <$buffer>\n";
397 };
398
399 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval)
400 Tries to copy $length bytes from $in_fh to $out_fh. It starts
401 reading at byte offset $in_offset, and starts writing at the current
402 file offset of $out_fh. Because of that, it is not safe to issue
403 more than one "aio_sendfile" per $out_fh, as they will interfere
404 with each other. The same $in_fh works fine though, as this function
405 does not move or use the file offset of $in_fh.
406
407 Please note that "aio_sendfile" can read more bytes from $in_fh than
408 are written, and there is no way to find out how many more bytes
409 have been read from "aio_sendfile" alone, as "aio_sendfile" only
410 provides the number of bytes written to $out_fh. Only if the result
411 value equals $length one can assume that $length bytes have been
412 read.
413
414 Unlike with other "aio_" functions, it makes a lot of sense to use
415 "aio_sendfile" on non-blocking sockets, as long as one end
416 (typically the $in_fh) is a file - the file I/O will then be
417 asynchronous, while the socket I/O will be non-blocking. Note,
418 however, that you can run into a trap where "aio_sendfile" reads
419 some data with readahead, then fails to write all data, and when the
420 socket is ready the next time, the data in the cache is already
421 lost, forcing "aio_sendfile" to again hit the disk. Explicit
422 "aio_read" + "aio_write" let's you better control resource usage.
423
424 This call tries to make use of a native "sendfile"-like syscall to
425 provide zero-copy operation. For this to work, $out_fh should refer
426 to a socket, and $in_fh should refer to an mmap'able file.
427
428 If a native sendfile cannot be found or it fails with "ENOSYS",
429 "EINVAL", "ENOTSUP", "EOPNOTSUPP", "EAFNOSUPPORT", "EPROTOTYPE" or
430 "ENOTSOCK", it will be emulated, so you can call "aio_sendfile" on
431 any type of filehandle regardless of the limitations of the
432 operating system.
433
434 As native sendfile syscalls (as practically any non-POSIX interface
435 hacked together in a hurry to improve benchmark numbers) tend to be
436 rather buggy on many systems, this implementation tries to work
437 around some known bugs in Linux and FreeBSD kernels (probably
438 others, too), but that might fail, so you really really should check
439 the return value of "aio_sendfile" - fewre bytes than expected might
440 have been transferred.
441
442 aio_readahead $fh,$offset,$length, $callback->($retval)
443 "aio_readahead" populates the page cache with data from a file so
444 that subsequent reads from that file will not block on disk I/O. The
445 $offset argument specifies the starting point from which data is to
446 be read and $length specifies the number of bytes to be read. I/O is
447 performed in whole pages, so that offset is effectively rounded down
448 to a page boundary and bytes are read up to the next page boundary
449 greater than or equal to (off-set+length). "aio_readahead" does not
450 read beyond the end of the file. The current file offset of the file
451 is left unchanged.
452
453 If that syscall doesn't exist (likely if your OS isn't Linux) it
454 will be emulated by simply reading the data, which would have a
455 similar effect.
456
457 aio_stat $fh_or_path, $callback->($status)
458 aio_lstat $fh, $callback->($status)
459 Works like perl's "stat" or "lstat" in void context. The callback
460 will be called after the stat and the results will be available
461 using "stat _" or "-s _" etc...
462
463 The pathname passed to "aio_stat" must be absolute. See API NOTES,
464 above, for an explanation.
465
466 Currently, the stats are always 64-bit-stats, i.e. instead of
467 returning an error when stat'ing a large file, the results will be
468 silently truncated unless perl itself is compiled with large file
469 support.
470
471 To help interpret the mode and dev/rdev stat values, IO::AIO offers
472 the following constants and functions (if not implemented, the
473 constants will be 0 and the functions will either "croak" or fall
474 back on traditional behaviour).
475
476 "S_IFMT", "S_IFIFO", "S_IFCHR", "S_IFBLK", "S_IFLNK", "S_IFREG",
477 "S_IFDIR", "S_IFWHT", "S_IFSOCK", "IO::AIO::major $dev_t",
478 "IO::AIO::minor $dev_t", "IO::AIO::makedev $major, $minor".
479
480 Example: Print the length of /etc/passwd:
481
482 aio_stat "/etc/passwd", sub {
483 $_[0] and die "stat failed: $!";
484 print "size is ", -s _, "\n";
485 };
486
487 aio_statvfs $fh_or_path, $callback->($statvfs)
488 Works like the POSIX "statvfs" or "fstatvfs" syscalls, depending on
489 whether a file handle or path was passed.
490
491 On success, the callback is passed a hash reference with the
492 following members: "bsize", "frsize", "blocks", "bfree", "bavail",
493 "files", "ffree", "favail", "fsid", "flag" and "namemax". On
494 failure, "undef" is passed.
495
496 The following POSIX IO::AIO::ST_* constants are defined: "ST_RDONLY"
497 and "ST_NOSUID".
498
499 The following non-POSIX IO::AIO::ST_* flag masks are defined to
500 their correct value when available, or to 0 on systems that do not
501 support them: "ST_NODEV", "ST_NOEXEC", "ST_SYNCHRONOUS",
502 "ST_MANDLOCK", "ST_WRITE", "ST_APPEND", "ST_IMMUTABLE",
503 "ST_NOATIME", "ST_NODIRATIME" and "ST_RELATIME".
504
505 Example: stat "/wd" and dump out the data if successful.
506
507 aio_statvfs "/wd", sub {
508 my $f = $_[0]
509 or die "statvfs: $!";
510
511 use Data::Dumper;
512 say Dumper $f;
513 };
514
515 # result:
516 {
517 bsize => 1024,
518 bfree => 4333064312,
519 blocks => 10253828096,
520 files => 2050765568,
521 flag => 4096,
522 favail => 2042092649,
523 bavail => 4333064312,
524 ffree => 2042092649,
525 namemax => 255,
526 frsize => 1024,
527 fsid => 1810
528 }
529
530 aio_utime $fh_or_path, $atime, $mtime, $callback->($status)
531 Works like perl's "utime" function (including the special case of
532 $atime and $mtime being undef). Fractional times are supported if
533 the underlying syscalls support them.
534
535 When called with a pathname, uses utimes(2) if available, otherwise
536 utime(2). If called on a file descriptor, uses futimes(2) if
537 available, otherwise returns ENOSYS, so this is not portable.
538
539 Examples:
540
541 # set atime and mtime to current time (basically touch(1)):
542 aio_utime "path", undef, undef;
543 # set atime to current time and mtime to beginning of the epoch:
544 aio_utime "path", time, undef; # undef==0
545
546 aio_chown $fh_or_path, $uid, $gid, $callback->($status)
547 Works like perl's "chown" function, except that "undef" for either
548 $uid or $gid is being interpreted as "do not change" (but -1 can
549 also be used).
550
551 Examples:
552
553 # same as "chown root path" in the shell:
554 aio_chown "path", 0, -1;
555 # same as above:
556 aio_chown "path", 0, undef;
557
558 aio_truncate $fh_or_path, $offset, $callback->($status)
559 Works like truncate(2) or ftruncate(2).
560
561 aio_chmod $fh_or_path, $mode, $callback->($status)
562 Works like perl's "chmod" function.
563
564 aio_unlink $pathname, $callback->($status)
565 Asynchronously unlink (delete) a file and call the callback with the
566 result code.
567
568 aio_mknod $pathname, $mode, $dev, $callback->($status)
569 [EXPERIMENTAL]
570
571 Asynchronously create a device node (or fifo). See mknod(2).
572
573 The only (POSIX-) portable way of calling this function is:
574
575 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ...
576
577 See "aio_stat" for info about some potentially helpful extra
578 constants and functions.
579
580 aio_link $srcpath, $dstpath, $callback->($status)
581 Asynchronously create a new link to the existing object at $srcpath
582 at the path $dstpath and call the callback with the result code.
583
584 aio_symlink $srcpath, $dstpath, $callback->($status)
585 Asynchronously create a new symbolic link to the existing object at
586 $srcpath at the path $dstpath and call the callback with the result
587 code.
588
589 aio_readlink $pathname, $callback->($link)
590 Asynchronously read the symlink specified by $path and pass it to
591 the callback. If an error occurs, nothing or undef gets passed to
592 the callback.
593
594 aio_realpath $pathname, $callback->($path)
595 Asynchronously make the path absolute and resolve any symlinks in
596 $path. The resulting path only consists of directories (Same as
597 Cwd::realpath).
598
599 This request can be used to get the absolute path of the current
600 working directory by passing it a path of . (a single dot).
601
602 aio_rename $srcpath, $dstpath, $callback->($status)
603 Asynchronously rename the object at $srcpath to $dstpath, just as
604 rename(2) and call the callback with the result code.
605
606 aio_mkdir $pathname, $mode, $callback->($status)
607 Asynchronously mkdir (create) a directory and call the callback with
608 the result code. $mode will be modified by the umask at the time the
609 request is executed, so do not change your umask.
610
611 aio_rmdir $pathname, $callback->($status)
612 Asynchronously rmdir (delete) a directory and call the callback with
613 the result code.
614
615 aio_readdir $pathname, $callback->($entries)
616 Unlike the POSIX call of the same name, "aio_readdir" reads an
617 entire directory (i.e. opendir + readdir + closedir). The entries
618 will not be sorted, and will NOT include the "." and ".." entries.
619
620 The callback is passed a single argument which is either "undef" or
621 an array-ref with the filenames.
622
623 aio_readdirx $pathname, $flags, $callback->($entries, $flags)
624 Quite similar to "aio_readdir", but the $flags argument allows one
625 to tune behaviour and output format. In case of an error, $entries
626 will be "undef".
627
628 The flags are a combination of the following constants, ORed
629 together (the flags will also be passed to the callback, possibly
630 modified):
631
632 IO::AIO::READDIR_DENTS
633 When this flag is off, then the callback gets an arrayref
634 consisting of names only (as with "aio_readdir"), otherwise it
635 gets an arrayref with "[$name, $type, $inode]" arrayrefs, each
636 describing a single directory entry in more detail.
637
638 $name is the name of the entry.
639
640 $type is one of the "IO::AIO::DT_xxx" constants:
641
642 "IO::AIO::DT_UNKNOWN", "IO::AIO::DT_FIFO", "IO::AIO::DT_CHR",
643 "IO::AIO::DT_DIR", "IO::AIO::DT_BLK", "IO::AIO::DT_REG",
644 "IO::AIO::DT_LNK", "IO::AIO::DT_SOCK", "IO::AIO::DT_WHT".
645
646 "IO::AIO::DT_UNKNOWN" means just that: readdir does not know. If
647 you need to know, you have to run stat yourself. Also, for speed
648 reasons, the $type scalars are read-only: you can not modify
649 them.
650
651 $inode is the inode number (which might not be exact on systems
652 with 64 bit inode numbers and 32 bit perls). This field has
653 unspecified content on systems that do not deliver the inode
654 information.
655
656 IO::AIO::READDIR_DIRS_FIRST
657 When this flag is set, then the names will be returned in an
658 order where likely directories come first, in optimal stat
659 order. This is useful when you need to quickly find directories,
660 or you want to find all directories while avoiding to stat()
661 each entry.
662
663 If the system returns type information in readdir, then this is
664 used to find directories directly. Otherwise, likely directories
665 are names beginning with ".", or otherwise names with no dots,
666 of which names with short names are tried first.
667
668 IO::AIO::READDIR_STAT_ORDER
669 When this flag is set, then the names will be returned in an
670 order suitable for stat()'ing each one. That is, when you plan
671 to stat() all files in the given directory, then the returned
672 order will likely be fastest.
673
674 If both this flag and "IO::AIO::READDIR_DIRS_FIRST" are
675 specified, then the likely dirs come first, resulting in a less
676 optimal stat order.
677
678 IO::AIO::READDIR_FOUND_UNKNOWN
679 This flag should not be set when calling "aio_readdirx".
680 Instead, it is being set by "aio_readdirx", when any of the
681 $type's found were "IO::AIO::DT_UNKNOWN". The absence of this
682 flag therefore indicates that all $type's are known, which can
683 be used to speed up some algorithms.
684
685 aio_load $pathname, $data, $callback->($status)
686 This is a composite request that tries to fully load the given file
687 into memory. Status is the same as with aio_read.
688
689 aio_copy $srcpath, $dstpath, $callback->($status)
690 Try to copy the *file* (directories not supported as either source
691 or destination) from $srcpath to $dstpath and call the callback with
692 a status of 0 (ok) or -1 (error, see $!).
693
694 This is a composite request that creates the destination file with
695 mode 0200 and copies the contents of the source file into it using
696 "aio_sendfile", followed by restoring atime, mtime, access mode and
697 uid/gid, in that order.
698
699 If an error occurs, the partial destination file will be unlinked,
700 if possible, except when setting atime, mtime, access mode and
701 uid/gid, where errors are being ignored.
702
703 aio_move $srcpath, $dstpath, $callback->($status)
704 Try to move the *file* (directories not supported as either source
705 or destination) from $srcpath to $dstpath and call the callback with
706 a status of 0 (ok) or -1 (error, see $!).
707
708 This is a composite request that tries to rename(2) the file first;
709 if rename fails with "EXDEV", it copies the file with "aio_copy"
710 and, if that is successful, unlinks the $srcpath.
711
712 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs)
713 Scans a directory (similar to "aio_readdir") but additionally tries
714 to efficiently separate the entries of directory $path into two sets
715 of names, directories you can recurse into (directories), and ones
716 you cannot recurse into (everything else, including symlinks to
717 directories).
718
719 "aio_scandir" is a composite request that creates of many sub
720 requests_ $maxreq specifies the maximum number of outstanding aio
721 requests that this function generates. If it is "<= 0", then a
722 suitable default will be chosen (currently 4).
723
724 On error, the callback is called without arguments, otherwise it
725 receives two array-refs with path-relative entry names.
726
727 Example:
728
729 aio_scandir $dir, 0, sub {
730 my ($dirs, $nondirs) = @_;
731 print "real directories: @$dirs\n";
732 print "everything else: @$nondirs\n";
733 };
734
735 Implementation notes.
736
737 The "aio_readdir" cannot be avoided, but "stat()"'ing every entry
738 can.
739
740 If readdir returns file type information, then this is used directly
741 to find directories.
742
743 Otherwise, after reading the directory, the modification time, size
744 etc. of the directory before and after the readdir is checked, and
745 if they match (and isn't the current time), the link count will be
746 used to decide how many entries are directories (if >= 2).
747 Otherwise, no knowledge of the number of subdirectories will be
748 assumed.
749
750 Then entries will be sorted into likely directories a non-initial
751 dot currently) and likely non-directories (see "aio_readdirx"). Then
752 every entry plus an appended "/." will be "stat"'ed, likely
753 directories first, in order of their inode numbers. If that
754 succeeds, it assumes that the entry is a directory or a symlink to
755 directory (which will be checked separately). This is often faster
756 than stat'ing the entry itself because filesystems might detect the
757 type of the entry without reading the inode data (e.g. ext2fs
758 filetype feature), even on systems that cannot return the filetype
759 information on readdir.
760
761 If the known number of directories (link count - 2) has been
762 reached, the rest of the entries is assumed to be non-directories.
763
764 This only works with certainty on POSIX (= UNIX) filesystems, which
765 fortunately are the vast majority of filesystems around.
766
767 It will also likely work on non-POSIX filesystems with reduced
768 efficiency as those tend to return 0 or 1 as link counts, which
769 disables the directory counting heuristic.
770
771 aio_rmtree $pathname, $callback->($status)
772 Delete a directory tree starting (and including) $path, return the
773 status of the final "rmdir" only. This is a composite request that
774 uses "aio_scandir" to recurse into and rmdir directories, and unlink
775 everything else.
776
777 aio_sync $callback->($status)
778 Asynchronously call sync and call the callback when finished.
779
780 aio_fsync $fh, $callback->($status)
781 Asynchronously call fsync on the given filehandle and call the
782 callback with the fsync result code.
783
784 aio_fdatasync $fh, $callback->($status)
785 Asynchronously call fdatasync on the given filehandle and call the
786 callback with the fdatasync result code.
787
788 If this call isn't available because your OS lacks it or it couldn't
789 be detected, it will be emulated by calling "fsync" instead.
790
791 aio_syncfs $fh, $callback->($status)
792 Asynchronously call the syncfs syscall to sync the filesystem
793 associated to the given filehandle and call the callback with the
794 syncfs result code. If syncfs is not available, calls sync(), but
795 returns -1 and sets errno to "ENOSYS" nevertheless.
796
797 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status)
798 Sync the data portion of the file specified by $offset and $length
799 to disk (but NOT the metadata), by calling the Linux-specific
800 sync_file_range call. If sync_file_range is not available or it
801 returns ENOSYS, then fdatasync or fsync is being substituted.
802
803 $flags can be a combination of
804 "IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE",
805 "IO::AIO::SYNC_FILE_RANGE_WRITE" and
806 "IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER": refer to the sync_file_range
807 manpage for details.
808
809 aio_pathsync $pathname, $callback->($status)
810 This request tries to open, fsync and close the given path. This is
811 a composite request intended to sync directories after directory
812 operations (E.g. rename). This might not work on all operating
813 systems or have any specific effect, but usually it makes sure that
814 directory changes get written to disc. It works for anything that
815 can be opened for read-only, not just directories.
816
817 Future versions of this function might fall back to other methods
818 when "fsync" on the directory fails (such as calling "sync").
819
820 Passes 0 when everything went ok, and -1 on error.
821
822 aio_msync $scalar, $offset = 0, $length = undef, flags = 0,
823 $callback->($status)
824 This is a rather advanced IO::AIO call, which only works on
825 mmap(2)ed scalars (see the "IO::AIO::mmap" function, although it
826 also works on data scalars managed by the Sys::Mmap or Mmap modules,
827 note that the scalar must only be modified in-place while an aio
828 operation is pending on it).
829
830 It calls the "msync" function of your OS, if available, with the
831 memory area starting at $offset in the string and ending $length
832 bytes later. If $length is negative, counts from the end, and if
833 $length is "undef", then it goes till the end of the string. The
834 flags can be a combination of "IO::AIO::MS_ASYNC",
835 "IO::AIO::MS_INVALIDATE" and "IO::AIO::MS_SYNC".
836
837 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0,
838 $callback->($status)
839 This is a rather advanced IO::AIO call, which works best on
840 mmap(2)ed scalars.
841
842 It touches (reads or writes) all memory pages in the specified range
843 inside the scalar. All caveats and parameters are the same as for
844 "aio_msync", above, except for flags, which must be either 0 (which
845 reads all pages and ensures they are instantiated) or
846 "IO::AIO::MT_MODIFY", which modifies the memory page s(by reading
847 and writing an octet from it, which dirties the page).
848
849 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status)
850 This is a rather advanced IO::AIO call, which works best on
851 mmap(2)ed scalars.
852
853 It reads in all the pages of the underlying storage into memory (if
854 any) and locks them, so they are not getting swapped/paged out or
855 removed.
856
857 If $length is undefined, then the scalar will be locked till the
858 end.
859
860 On systems that do not implement "mlock", this function returns -1
861 and sets errno to "ENOSYS".
862
863 Note that the corresponding "munlock" is synchronous and is
864 documented under "MISCELLANEOUS FUNCTIONS".
865
866 Example: open a file, mmap and mlock it - both will be undone when
867 $data gets destroyed.
868
869 open my $fh, "<", $path or die "$path: $!";
870 my $data;
871 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh;
872 aio_mlock $data; # mlock in background
873
874 aio_mlockall $flags, $callback->($status)
875 Calls the "mlockall" function with the given $flags (a combination
876 of "IO::AIO::MCL_CURRENT" and "IO::AIO::MCL_FUTURE").
877
878 On systems that do not implement "mlockall", this function returns
879 -1 and sets errno to "ENOSYS".
880
881 Note that the corresponding "munlockall" is synchronous and is
882 documented under "MISCELLANEOUS FUNCTIONS".
883
884 Example: asynchronously lock all current and future pages into
885 memory.
886
887 aio_mlockall IO::AIO::MCL_FUTURE;
888
889 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents)
890 Queries the extents of the given file (by calling the Linux FIEMAP
891 ioctl, see <http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for
892 details). If the "ioctl" is not available on your OS, then this
893 rquiest will fail with "ENOSYS".
894
895 $start is the starting offset to query extents for, $length is the
896 size of the range to query - if it is "undef", then the whole file
897 will be queried.
898
899 $flags is a combination of flags ("IO::AIO::FIEMAP_FLAG_SYNC" or
900 "IO::AIO::FIEMAP_FLAG_XATTR" - "IO::AIO::FIEMAP_FLAGS_COMPAT" is
901 also exported), and is normally 0 or "IO::AIO::FIEMAP_FLAG_SYNC" to
902 query the data portion.
903
904 $count is the maximum number of extent records to return. If it is
905 "undef", then IO::AIO queries all extents of the file. As a very
906 special case, if it is 0, then the callback receives the number of
907 extents instead of the extents themselves.
908
909 If an error occurs, the callback receives no arguments. The special
910 "errno" value "IO::AIO::EBADR" is available to test for flag errors.
911
912 Otherwise, the callback receives an array reference with extent
913 structures. Each extent structure is an array reference itself, with
914 the following members:
915
916 [$logical, $physical, $length, $flags]
917
918 Flags is any combination of the following flag values (typically
919 either 0 or "IO::AIO::FIEMAP_EXTENT_LAST"):
920
921 "IO::AIO::FIEMAP_EXTENT_LAST", "IO::AIO::FIEMAP_EXTENT_UNKNOWN",
922 "IO::AIO::FIEMAP_EXTENT_DELALLOC", "IO::AIO::FIEMAP_EXTENT_ENCODED",
923 "IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED",
924 "IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED",
925 "IO::AIO::FIEMAP_EXTENT_DATA_INLINE",
926 "IO::AIO::FIEMAP_EXTENT_DATA_TAIL",
927 "IO::AIO::FIEMAP_EXTENT_UNWRITTEN", "IO::AIO::FIEMAP_EXTENT_MERGED"
928 or "IO::AIO::FIEMAP_EXTENT_SHARED".
929
930 aio_group $callback->(...)
931 This is a very special aio request: Instead of doing something, it
932 is a container for other aio requests, which is useful if you want
933 to bundle many requests into a single, composite, request with a
934 definite callback and the ability to cancel the whole request with
935 its subrequests.
936
937 Returns an object of class IO::AIO::GRP. See its documentation below
938 for more info.
939
940 Example:
941
942 my $grp = aio_group sub {
943 print "all stats done\n";
944 };
945
946 add $grp
947 (aio_stat ...),
948 (aio_stat ...),
949 ...;
950
951 aio_nop $callback->()
952 This is a special request - it does nothing in itself and is only
953 used for side effects, such as when you want to add a dummy request
954 to a group so that finishing the requests in the group depends on
955 executing the given code.
956
957 While this request does nothing, it still goes through the execution
958 phase and still requires a worker thread. Thus, the callback will
959 not be executed immediately but only after other requests in the
960 queue have entered their execution phase. This can be used to
961 measure request latency.
962
963 IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED*
964 Mainly used for debugging and benchmarking, this aio request puts
965 one of the request workers to sleep for the given time.
966
967 While it is theoretically handy to have simple I/O scheduling
968 requests like sleep and file handle readable/writable, the overhead
969 this creates is immense (it blocks a thread for a long time) so do
970 not use this function except to put your application under
971 artificial I/O pressure.
972
973 IO::AIO::WD - multiple working directories
974 Your process only has one current working directory, which is used by
975 all threads. This makes it hard to use relative paths (some other
976 component could call "chdir" at any time, and it is hard to control when
977 the path will be used by IO::AIO).
978
979 One solution for this is to always use absolute paths. This usually
980 works, but can be quite slow (the kernel has to walk the whole path on
981 every access), and can also be a hassle to implement.
982
983 Newer POSIX systems have a number of functions (openat, fdopendir,
984 futimensat and so on) that make it possible to specify working
985 directories per operation.
986
987 For portability, and because the clowns who "designed", or shall I
988 write, perpetrated this new interface were obviously half-drunk, this
989 abstraction cannot be perfect, though.
990
991 IO::AIO allows you to convert directory paths into a so-called
992 IO::AIO::WD object. This object stores the canonicalised, absolute
993 version of the path, and on systems that allow it, also a directory file
994 descriptor.
995
996 Everywhere where a pathname is accepted by IO::AIO (e.g. in "aio_stat"
997 or "aio_unlink"), one can specify an array reference with an IO::AIO::WD
998 object and a pathname instead (or the IO::AIO::WD object alone, which
999 gets interpreted as "[$wd, "."]"). If the pathname is absolute, the
1000 IO::AIO::WD object is ignored, otherwise the pathname is resolved
1001 relative to that IO::AIO::WD object.
1002
1003 For example, to get a wd object for /etc and then stat passwd inside,
1004 you would write:
1005
1006 aio_wd "/etc", sub {
1007 my $etcdir = shift;
1008
1009 # although $etcdir can be undef on error, there is generally no reason
1010 # to check for errors here, as aio_stat will fail with ENOENT
1011 # when $etcdir is undef.
1012
1013 aio_stat [$etcdir, "passwd"], sub {
1014 # yay
1015 };
1016 };
1017
1018 That "aio_wd" is a request and not a normal function shows that creating
1019 an IO::AIO::WD object is itself a potentially blocking operation, which
1020 is why it is done asynchronously.
1021
1022 To stat the directory obtained with "aio_wd" above, one could write
1023 either of the following three request calls:
1024
1025 aio_lstat "/etc" , sub { ... # pathname as normal string
1026 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself)
1027 aio_lstat $wd , sub { ... # shorthand for the previous
1028
1029 As with normal pathnames, IO::AIO keeps a copy of the working directory
1030 object and the pathname string, so you could write the following without
1031 causing any issues due to $path getting reused:
1032
1033 my $path = [$wd, undef];
1034
1035 for my $name (qw(abc def ghi)) {
1036 $path->[1] = $name;
1037 aio_stat $path, sub {
1038 # ...
1039 };
1040 }
1041
1042 There are some caveats: when directories get renamed (or deleted), the
1043 pathname string doesn't change, so will point to the new directory (or
1044 nowhere at all), while the directory fd, if available on the system,
1045 will still point to the original directory. Most functions accepting a
1046 pathname will use the directory fd on newer systems, and the string on
1047 older systems. Some functions (such as realpath) will always rely on the
1048 string form of the pathname.
1049
1050 So this fucntionality is mainly useful to get some protection against
1051 "chdir", to easily get an absolute path out of a relative path for
1052 future reference, and to speed up doing many operations in the same
1053 directory (e.g. when stat'ing all files in a directory).
1054
1055 The following functions implement this working directory abstraction:
1056
1057 aio_wd $pathname, $callback->($wd)
1058 Asynchonously canonicalise the given pathname and convert it to an
1059 IO::AIO::WD object representing it. If possible and supported on the
1060 system, also open a directory fd to speed up pathname resolution
1061 relative to this working directory.
1062
1063 If something goes wrong, then "undef" is passwd to the callback
1064 instead of a working directory object and $! is set appropriately.
1065 Since passing "undef" as working directory component of a pathname
1066 fails the request with "ENOENT", there is often no need for error
1067 checking in the "aio_wd" callback, as future requests using the
1068 value will fail in the expected way.
1069
1070 If this call isn't available because your OS lacks it or it couldn't
1071 be detected, it will be emulated by calling "fsync" instead.
1072
1073 IO::AIO::CWD
1074 This is a compiletime constant (object) that represents the process
1075 current working directory.
1076
1077 Specifying this object as working directory object for a pathname is
1078 as if the pathname would be specified directly, without a directory
1079 object, e.g., these calls are functionally identical:
1080
1081 aio_stat "somefile", sub { ... };
1082 aio_stat [IO::AIO::CWD, "somefile"], sub { ... };
1083
1084 IO::AIO::REQ CLASS
1085 All non-aggregate "aio_*" functions return an object of this class when
1086 called in non-void context.
1087
1088 cancel $req
1089 Cancels the request, if possible. Has the effect of skipping
1090 execution when entering the execute state and skipping calling the
1091 callback when entering the the result state, but will leave the
1092 request otherwise untouched (with the exception of readdir). That
1093 means that requests that currently execute will not be stopped and
1094 resources held by the request will not be freed prematurely.
1095
1096 cb $req $callback->(...)
1097 Replace (or simply set) the callback registered to the request.
1098
1099 IO::AIO::GRP CLASS
1100 This class is a subclass of IO::AIO::REQ, so all its methods apply to
1101 objects of this class, too.
1102
1103 A IO::AIO::GRP object is a special request that can contain multiple
1104 other aio requests.
1105
1106 You create one by calling the "aio_group" constructing function with a
1107 callback that will be called when all contained requests have entered
1108 the "done" state:
1109
1110 my $grp = aio_group sub {
1111 print "all requests are done\n";
1112 };
1113
1114 You add requests by calling the "add" method with one or more
1115 "IO::AIO::REQ" objects:
1116
1117 $grp->add (aio_unlink "...");
1118
1119 add $grp aio_stat "...", sub {
1120 $_[0] or return $grp->result ("error");
1121
1122 # add another request dynamically, if first succeeded
1123 add $grp aio_open "...", sub {
1124 $grp->result ("ok");
1125 };
1126 };
1127
573 This makes it very easy to create composite requests (see the source 1128 This makes it very easy to create composite requests (see the source of
574 of "aio_move" for an application) that work and feel like simple 1129 "aio_move" for an application) that work and feel like simple requests.
575 requests.
576 1130
577 * The IO::AIO::GRP objects will be cleaned up during calls to 1131 * The IO::AIO::GRP objects will be cleaned up during calls to
578 "IO::AIO::poll_cb", just like any other request. 1132 "IO::AIO::poll_cb", just like any other request.
1133
579 * They can be canceled like any other request. Canceling will cancel 1134 * They can be canceled like any other request. Canceling will cancel
580 not only the request itself, but also all requests it contains. 1135 not only the request itself, but also all requests it contains.
1136
581 * They can also can also be added to other IO::AIO::GRP objects. 1137 * They can also can also be added to other IO::AIO::GRP objects.
1138
582 * You must not add requests to a group from within the group 1139 * You must not add requests to a group from within the group callback
583 callback (or any later time). 1140 (or any later time).
584 1141
585 Their lifetime, simplified, looks like this: when they are empty, 1142 Their lifetime, simplified, looks like this: when they are empty, they
586 they will finish very quickly. If they contain only requests that 1143 will finish very quickly. If they contain only requests that are in the
587 are in the "done" state, they will also finish. Otherwise they will 1144 "done" state, they will also finish. Otherwise they will continue to
588 continue to exist. 1145 exist.
589 1146
590 That means after creating a group you have some time to add 1147 That means after creating a group you have some time to add requests
1148 (precisely before the callback has been invoked, which is only done
591 requests. And in the callbacks of those requests, you can add 1149 within the "poll_cb"). And in the callbacks of those requests, you can
592 further requests to the group. And only when all those requests have 1150 add further requests to the group. And only when all those requests have
593 finished will the the group itself finish. 1151 finished will the the group itself finish.
594 1152
595 add $grp ... 1153 add $grp ...
596 $grp->add (...) 1154 $grp->add (...)
597 Add one or more requests to the group. Any type of IO::AIO::REQ 1155 Add one or more requests to the group. Any type of IO::AIO::REQ can
598 can be added, including other groups, as long as you do not 1156 be added, including other groups, as long as you do not create
599 create circular dependencies. 1157 circular dependencies.
600 1158
601 Returns all its arguments. 1159 Returns all its arguments.
602 1160
603 $grp->cancel_subs 1161 $grp->cancel_subs
604 Cancel all subrequests and clears any feeder, but not the group 1162 Cancel all subrequests and clears any feeder, but not the group
605 request itself. Useful when you queued a lot of events but got a 1163 request itself. Useful when you queued a lot of events but got a
606 result early. 1164 result early.
607 1165
1166 The group request will finish normally (you cannot add requests to
1167 the group).
1168
608 $grp->result (...) 1169 $grp->result (...)
609 Set the result value(s) that will be passed to the group 1170 Set the result value(s) that will be passed to the group callback
610 callback when all subrequests have finished and set thre groups 1171 when all subrequests have finished and set the groups errno to the
611 errno to the current value of errno (just like calling "errno" 1172 current value of errno (just like calling "errno" without an error
612 without an error number). By default, no argument will be passed 1173 number). By default, no argument will be passed and errno is zero.
613 and errno is zero.
614 1174
615 $grp->errno ([$errno]) 1175 $grp->errno ([$errno])
616 Sets the group errno value to $errno, or the current value of 1176 Sets the group errno value to $errno, or the current value of errno
617 errno when the argument is missing. 1177 when the argument is missing.
618 1178
619 Every aio request has an associated errno value that is restored 1179 Every aio request has an associated errno value that is restored
620 when the callback is invoked. This method lets you change this 1180 when the callback is invoked. This method lets you change this value
621 value from its default (0). 1181 from its default (0).
622 1182
623 Calling "result" will also set errno, so make sure you either 1183 Calling "result" will also set errno, so make sure you either set $!
624 set $! before the call to "result", or call c<errno> after it. 1184 before the call to "result", or call c<errno> after it.
625 1185
626 feed $grp $callback->($grp) 1186 feed $grp $callback->($grp)
627 Sets a feeder/generator on this group: every group can have an 1187 Sets a feeder/generator on this group: every group can have an
628 attached generator that generates requests if idle. The idea 1188 attached generator that generates requests if idle. The idea behind
629 behind this is that, although you could just queue as many 1189 this is that, although you could just queue as many requests as you
630 requests as you want in a group, this might starve other 1190 want in a group, this might starve other requests for a potentially
631 requests for a potentially long time. For example, "aio_scandir" 1191 long time. For example, "aio_scandir" might generate hundreds of
632 might generate hundreds of thousands "aio_stat" requests, 1192 thousands of "aio_stat" requests, delaying any later requests for a
633 delaying any later requests for a long time. 1193 long time.
634 1194
635 To avoid this, and allow incremental generation of requests, you 1195 To avoid this, and allow incremental generation of requests, you can
636 can instead a group and set a feeder on it that generates those 1196 instead a group and set a feeder on it that generates those
637 requests. The feed callback will be called whenever there are 1197 requests. The feed callback will be called whenever there are few
638 few enough (see "limit", below) requests active in the group 1198 enough (see "limit", below) requests active in the group itself and
639 itself and is expected to queue more requests. 1199 is expected to queue more requests.
640 1200
641 The feed callback can queue as many requests as it likes (i.e. 1201 The feed callback can queue as many requests as it likes (i.e. "add"
642 "add" does not impose any limits). 1202 does not impose any limits).
643 1203
644 If the feed does not queue more requests when called, it will be 1204 If the feed does not queue more requests when called, it will be
645 automatically removed from the group. 1205 automatically removed from the group.
646 1206
647 If the feed limit is 0, it will be set to 2 automatically. 1207 If the feed limit is 0 when this method is called, it will be set to
1208 2 automatically.
648 1209
649 Example: 1210 Example:
650 1211
651 # stat all files in @files, but only ever use four aio requests concurrently: 1212 # stat all files in @files, but only ever use four aio requests concurrently:
652 1213
653 my $grp = aio_group sub { print "finished\n" }; 1214 my $grp = aio_group sub { print "finished\n" };
654 limit $grp 4; 1215 limit $grp 4;
655 feed $grp sub { 1216 feed $grp sub {
656 my $file = pop @files 1217 my $file = pop @files
657 or return; 1218 or return;
658 1219
659 add $grp aio_stat $file, sub { ... }; 1220 add $grp aio_stat $file, sub { ... };
660 }; 1221 };
661 1222
662 limit $grp $num 1223 limit $grp $num
663 Sets the feeder limit for the group: The feeder will be called 1224 Sets the feeder limit for the group: The feeder will be called
664 whenever the group contains less than this many requests. 1225 whenever the group contains less than this many requests.
665 1226
666 Setting the limit to 0 will pause the feeding process. 1227 Setting the limit to 0 will pause the feeding process.
1228
1229 The default value for the limit is 0, but note that setting a feeder
1230 automatically bumps it up to 2.
667 1231
668 SUPPORT FUNCTIONS 1232 SUPPORT FUNCTIONS
669 EVENT PROCESSING AND EVENT LOOP INTEGRATION 1233 EVENT PROCESSING AND EVENT LOOP INTEGRATION
670 $fileno = IO::AIO::poll_fileno 1234 $fileno = IO::AIO::poll_fileno
671 Return the *request result pipe file descriptor*. This 1235 Return the *request result pipe file descriptor*. This filehandle
672 filehandle must be polled for reading by some mechanism outside 1236 must be polled for reading by some mechanism outside this module
673 this module (e.g. Event or select, see below or the SYNOPSIS). 1237 (e.g. EV, Glib, select and so on, see below or the SYNOPSIS). If the
674 If the pipe becomes readable you have to call "poll_cb" to check 1238 pipe becomes readable you have to call "poll_cb" to check the
675 the results. 1239 results.
676 1240
677 See "poll_cb" for an example. 1241 See "poll_cb" for an example.
678 1242
679 IO::AIO::poll_cb 1243 IO::AIO::poll_cb
680 Process some outstanding events on the result pipe. You have to 1244 Process some outstanding events on the result pipe. You have to call
681 call this regularly. Returns the number of events processed. 1245 this regularly. Returns 0 if all events could be processed (or there
1246 were no events to process), or -1 if it returned earlier for
682 Returns immediately when no events are outstanding. The amount 1247 whatever reason. Returns immediately when no events are outstanding.
683 of events processed depends on the settings of 1248 The amount of events processed depends on the settings of
684 "IO::AIO::max_poll_req" and "IO::AIO::max_poll_time". 1249 "IO::AIO::max_poll_req" and "IO::AIO::max_poll_time".
685 1250
686 If not all requests were processed for whatever reason, the 1251 If not all requests were processed for whatever reason, the
687 filehandle will still be ready when "poll_cb" returns. 1252 filehandle will still be ready when "poll_cb" returns, so normally
1253 you don't have to do anything special to have it called later.
688 1254
1255 Apart from calling "IO::AIO::poll_cb" when the event filehandle
1256 becomes ready, it can be beneficial to call this function from loops
1257 which submit a lot of requests, to make sure the results get
1258 processed when they become available and not just when the loop is
1259 finished and the event loop takes over again. This function returns
1260 very fast when there are no outstanding requests.
1261
689 Example: Install an Event watcher that automatically calls 1262 Example: Install an Event watcher that automatically calls
690 IO::AIO::poll_cb with high priority: 1263 IO::AIO::poll_cb with high priority (more examples can be found in
1264 the SYNOPSIS section, at the top of this document):
691 1265
692 Event->io (fd => IO::AIO::poll_fileno, 1266 Event->io (fd => IO::AIO::poll_fileno,
693 poll => 'r', async => 1, 1267 poll => 'r', async => 1,
694 cb => \&IO::AIO::poll_cb); 1268 cb => \&IO::AIO::poll_cb);
695 1269
696 IO::AIO::max_poll_reqs $nreqs
697 IO::AIO::max_poll_time $seconds
698 These set the maximum number of requests (default 0, meaning
699 infinity) that are being processed by "IO::AIO::poll_cb" in one
700 call, respectively the maximum amount of time (default 0,
701 meaning infinity) spent in "IO::AIO::poll_cb" to process
702 requests (more correctly the mininum amount of time "poll_cb" is
703 allowed to use).
704
705 Setting these is useful if you want to ensure some level of
706 interactiveness when perl is not fast enough to process all
707 requests in time.
708
709 For interactive programs, values such as 0.01 to 0.1 should be
710 fine.
711
712 Example: Install an Event watcher that automatically calls
713 IO::AIO::poll_some with low priority, to ensure that other parts
714 of the program get the CPU sometimes even under high AIO load.
715
716 # try not to spend much more than 0.1s in poll_cb
717 IO::AIO::max_poll_time 0.1;
718
719 # use a low priority so other tasks have priority
720 Event->io (fd => IO::AIO::poll_fileno,
721 poll => 'r', nice => 1,
722 cb => &IO::AIO::poll_cb);
723
724 IO::AIO::poll_wait 1270 IO::AIO::poll_wait
1271 If there are any outstanding requests and none of them in the result
725 Wait till the result filehandle becomes ready for reading 1272 phase, wait till the result filehandle becomes ready for reading
726 (simply does a "select" on the filehandle. This is useful if you 1273 (simply does a "select" on the filehandle. This is useful if you
727 want to synchronously wait for some requests to finish). 1274 want to synchronously wait for some requests to finish).
728 1275
729 See "nreqs" for an example. 1276 See "nreqs" for an example.
730 1277
731 IO::AIO::poll 1278 IO::AIO::poll
732 Waits until some requests have been handled. 1279 Waits until some requests have been handled.
733 1280
1281 Returns the number of requests processed, but is otherwise strictly
734 Strictly equivalent to: 1282 equivalent to:
735 1283
736 IO::AIO::poll_wait, IO::AIO::poll_cb 1284 IO::AIO::poll_wait, IO::AIO::poll_cb
737 if IO::AIO::nreqs;
738 1285
739 IO::AIO::flush 1286 IO::AIO::flush
740 Wait till all outstanding AIO requests have been handled. 1287 Wait till all outstanding AIO requests have been handled.
741 1288
742 Strictly equivalent to: 1289 Strictly equivalent to:
743 1290
744 IO::AIO::poll_wait, IO::AIO::poll_cb 1291 IO::AIO::poll_wait, IO::AIO::poll_cb
745 while IO::AIO::nreqs; 1292 while IO::AIO::nreqs;
1293
1294 IO::AIO::max_poll_reqs $nreqs
1295 IO::AIO::max_poll_time $seconds
1296 These set the maximum number of requests (default 0, meaning
1297 infinity) that are being processed by "IO::AIO::poll_cb" in one
1298 call, respectively the maximum amount of time (default 0, meaning
1299 infinity) spent in "IO::AIO::poll_cb" to process requests (more
1300 correctly the mininum amount of time "poll_cb" is allowed to use).
1301
1302 Setting "max_poll_time" to a non-zero value creates an overhead of
1303 one syscall per request processed, which is not normally a problem
1304 unless your callbacks are really really fast or your OS is really
1305 really slow (I am not mentioning Solaris here). Using
1306 "max_poll_reqs" incurs no overhead.
1307
1308 Setting these is useful if you want to ensure some level of
1309 interactiveness when perl is not fast enough to process all requests
1310 in time.
1311
1312 For interactive programs, values such as 0.01 to 0.1 should be fine.
1313
1314 Example: Install an Event watcher that automatically calls
1315 IO::AIO::poll_cb with low priority, to ensure that other parts of
1316 the program get the CPU sometimes even under high AIO load.
1317
1318 # try not to spend much more than 0.1s in poll_cb
1319 IO::AIO::max_poll_time 0.1;
1320
1321 # use a low priority so other tasks have priority
1322 Event->io (fd => IO::AIO::poll_fileno,
1323 poll => 'r', nice => 1,
1324 cb => &IO::AIO::poll_cb);
746 1325
747 CONTROLLING THE NUMBER OF THREADS 1326 CONTROLLING THE NUMBER OF THREADS
748 IO::AIO::min_parallel $nthreads 1327 IO::AIO::min_parallel $nthreads
749 Set the minimum number of AIO threads to $nthreads. The current 1328 Set the minimum number of AIO threads to $nthreads. The current
750 default is 8, which means eight asynchronous operations can 1329 default is 8, which means eight asynchronous operations can execute
751 execute concurrently at any one time (the number of outstanding 1330 concurrently at any one time (the number of outstanding requests,
752 requests, however, is unlimited). 1331 however, is unlimited).
753 1332
754 IO::AIO starts threads only on demand, when an AIO request is 1333 IO::AIO starts threads only on demand, when an AIO request is queued
755 queued and no free thread exists. Please note that queueing up a 1334 and no free thread exists. Please note that queueing up a hundred
756 hundred requests can create demand for a hundred threads, even 1335 requests can create demand for a hundred threads, even if it turns
757 if it turns out that everything is in the cache and could have 1336 out that everything is in the cache and could have been processed
758 been processed faster by a single thread. 1337 faster by a single thread.
759 1338
760 It is recommended to keep the number of threads relatively low, 1339 It is recommended to keep the number of threads relatively low, as
761 as some Linux kernel versions will scale negatively with the 1340 some Linux kernel versions will scale negatively with the number of
762 number of threads (higher parallelity => MUCH higher latency). 1341 threads (higher parallelity => MUCH higher latency). With current
763 With current Linux 2.6 versions, 4-32 threads should be fine. 1342 Linux 2.6 versions, 4-32 threads should be fine.
764 1343
765 Under most circumstances you don't need to call this function, 1344 Under most circumstances you don't need to call this function, as
766 as the module selects a default that is suitable for low to 1345 the module selects a default that is suitable for low to moderate
767 moderate load. 1346 load.
768 1347
769 IO::AIO::max_parallel $nthreads 1348 IO::AIO::max_parallel $nthreads
770 Sets the maximum number of AIO threads to $nthreads. If more 1349 Sets the maximum number of AIO threads to $nthreads. If more than
771 than the specified number of threads are currently running, this 1350 the specified number of threads are currently running, this function
772 function kills them. This function blocks until the limit is 1351 kills them. This function blocks until the limit is reached.
773 reached.
774 1352
775 While $nthreads are zero, aio requests get queued but not 1353 While $nthreads are zero, aio requests get queued but not executed
776 executed until the number of threads has been increased again. 1354 until the number of threads has been increased again.
777 1355
778 This module automatically runs "max_parallel 0" at program end, 1356 This module automatically runs "max_parallel 0" at program end, to
779 to ensure that all threads are killed and that there are no 1357 ensure that all threads are killed and that there are no outstanding
780 outstanding requests. 1358 requests.
781 1359
782 Under normal circumstances you don't need to call this function. 1360 Under normal circumstances you don't need to call this function.
783 1361
784 IO::AIO::max_idle $nthreads 1362 IO::AIO::max_idle $nthreads
785 Limit the number of threads (default: 4) that are allowed to 1363 Limit the number of threads (default: 4) that are allowed to idle
786 idle (i.e., threads that did not get a request to process within 1364 (i.e., threads that did not get a request to process within the idle
787 10 seconds). That means if a thread becomes idle while $nthreads 1365 timeout (default: 10 seconds). That means if a thread becomes idle
788 other threads are also idle, it will free its resources and 1366 while $nthreads other threads are also idle, it will free its
789 exit. 1367 resources and exit.
790 1368
791 This is useful when you allow a large number of threads (e.g. 1369 This is useful when you allow a large number of threads (e.g. 100 or
792 100 or 1000) to allow for extremely high load situations, but 1370 1000) to allow for extremely high load situations, but want to free
793 want to free resources under normal circumstances (1000 threads 1371 resources under normal circumstances (1000 threads can easily
794 can easily consume 30MB of RAM). 1372 consume 30MB of RAM).
795 1373
796 The default is probably ok in most situations, especially if 1374 The default is probably ok in most situations, especially if thread
797 thread creation is fast. If thread creation is very slow on your 1375 creation is fast. If thread creation is very slow on your system you
798 system you might want to use larger values. 1376 might want to use larger values.
799 1377
1378 IO::AIO::idle_timeout $seconds
1379 Sets the minimum idle timeout (default 10) after which worker
1380 threads are allowed to exit. SEe "IO::AIO::max_idle".
1381
800 $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 1382 IO::AIO::max_outstanding $maxreqs
801 This is a very bad function to use in interactive programs
802 because it blocks, and a bad way to reduce concurrency because
803 it is inexact: Better use an "aio_group" together with a feed
804 callback.
805
806 Sets the maximum number of outstanding requests to $nreqs. If 1383 Sets the maximum number of outstanding requests to $nreqs. If you do
807 you to queue up more than this number of requests, the next call 1384 queue up more than this number of requests, the next call to
808 to the "poll_cb" (and "poll_some" and other functions calling 1385 "IO::AIO::poll_cb" (and other functions calling "poll_cb", such as
809 "poll_cb") function will block until the limit is no longer 1386 "IO::AIO::flush" or "IO::AIO::poll") will block until the limit is
810 exceeded. 1387 no longer exceeded.
811 1388
812 The default value is very large, so there is no practical limit 1389 In other words, this setting does not enforce a queue limit, but can
1390 be used to make poll functions block if the limit is exceeded.
1391
1392 This is a very bad function to use in interactive programs because
1393 it blocks, and a bad way to reduce concurrency because it is
1394 inexact: Better use an "aio_group" together with a feed callback.
1395
1396 It's main use is in scripts without an event loop - when you want to
1397 stat a lot of files, you can write somehting like this:
1398
1399 IO::AIO::max_outstanding 32;
1400
1401 for my $path (...) {
1402 aio_stat $path , ...;
1403 IO::AIO::poll_cb;
1404 }
1405
1406 IO::AIO::flush;
1407
1408 The call to "poll_cb" inside the loop will normally return
1409 instantly, but as soon as more thna 32 reqeusts are in-flight, it
1410 will block until some requests have been handled. This keeps the
1411 loop from pushing a large number of "aio_stat" requests onto the
1412 queue.
1413
1414 The default value for "max_outstanding" is very large, so there is
813 on the number of outstanding requests. 1415 no practical limit on the number of outstanding requests.
814
815 You can still queue as many requests as you want. Therefore,
816 "max_oustsanding" is mainly useful in simple scripts (with low
817 values) or as a stop gap to shield against fatal memory overflow
818 (with large values).
819 1416
820 STATISTICAL INFORMATION 1417 STATISTICAL INFORMATION
821 IO::AIO::nreqs 1418 IO::AIO::nreqs
822 Returns the number of requests currently in the ready, execute 1419 Returns the number of requests currently in the ready, execute or
823 or pending states (i.e. for which their callback has not been 1420 pending states (i.e. for which their callback has not been invoked
824 invoked yet). 1421 yet).
825 1422
826 Example: wait till there are no outstanding requests anymore: 1423 Example: wait till there are no outstanding requests anymore:
827 1424
828 IO::AIO::poll_wait, IO::AIO::poll_cb 1425 IO::AIO::poll_wait, IO::AIO::poll_cb
829 while IO::AIO::nreqs; 1426 while IO::AIO::nreqs;
830 1427
831 IO::AIO::nready 1428 IO::AIO::nready
832 Returns the number of requests currently in the ready state (not 1429 Returns the number of requests currently in the ready state (not yet
833 yet executed). 1430 executed).
834 1431
835 IO::AIO::npending 1432 IO::AIO::npending
836 Returns the number of requests currently in the pending state 1433 Returns the number of requests currently in the pending state
837 (executed, but not yet processed by poll_cb). 1434 (executed, but not yet processed by poll_cb).
1435
1436 MISCELLANEOUS FUNCTIONS
1437 IO::AIO implements some functions that might be useful, but are not
1438 asynchronous.
1439
1440 IO::AIO::sendfile $ofh, $ifh, $offset, $count
1441 Calls the "eio_sendfile_sync" function, which is like
1442 "aio_sendfile", but is blocking (this makes most sense if you know
1443 the input data is likely cached already and the output filehandle is
1444 set to non-blocking operations).
1445
1446 Returns the number of bytes copied, or -1 on error.
1447
1448 IO::AIO::fadvise $fh, $offset, $len, $advice
1449 Simply calls the "posix_fadvise" function (see its manpage for
1450 details). The following advice constants are available:
1451 "IO::AIO::FADV_NORMAL", "IO::AIO::FADV_SEQUENTIAL",
1452 "IO::AIO::FADV_RANDOM", "IO::AIO::FADV_NOREUSE",
1453 "IO::AIO::FADV_WILLNEED", "IO::AIO::FADV_DONTNEED".
1454
1455 On systems that do not implement "posix_fadvise", this function
1456 returns ENOSYS, otherwise the return value of "posix_fadvise".
1457
1458 IO::AIO::madvise $scalar, $offset, $len, $advice
1459 Simply calls the "posix_madvise" function (see its manpage for
1460 details). The following advice constants are available:
1461 "IO::AIO::MADV_NORMAL", "IO::AIO::MADV_SEQUENTIAL",
1462 "IO::AIO::MADV_RANDOM", "IO::AIO::MADV_WILLNEED",
1463 "IO::AIO::MADV_DONTNEED".
1464
1465 On systems that do not implement "posix_madvise", this function
1466 returns ENOSYS, otherwise the return value of "posix_madvise".
1467
1468 IO::AIO::mprotect $scalar, $offset, $len, $protect
1469 Simply calls the "mprotect" function on the preferably AIO::mmap'ed
1470 $scalar (see its manpage for details). The following protect
1471 constants are available: "IO::AIO::PROT_NONE", "IO::AIO::PROT_READ",
1472 "IO::AIO::PROT_WRITE", "IO::AIO::PROT_EXEC".
1473
1474 On systems that do not implement "mprotect", this function returns
1475 ENOSYS, otherwise the return value of "mprotect".
1476
1477 IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset]
1478 Memory-maps a file (or anonymous memory range) and attaches it to
1479 the given $scalar, which will act like a string scalar.
1480
1481 The only operations allowed on the scalar are "substr"/"vec" that
1482 don't change the string length, and most read-only operations such
1483 as copying it or searching it with regexes and so on.
1484
1485 Anything else is unsafe and will, at best, result in memory leaks.
1486
1487 The memory map associated with the $scalar is automatically removed
1488 when the $scalar is destroyed, or when the "IO::AIO::mmap" or
1489 "IO::AIO::munmap" functions are called.
1490
1491 This calls the "mmap"(2) function internally. See your system's
1492 manual page for details on the $length, $prot and $flags parameters.
1493
1494 The $length must be larger than zero and smaller than the actual
1495 filesize.
1496
1497 $prot is a combination of "IO::AIO::PROT_NONE",
1498 "IO::AIO::PROT_EXEC", "IO::AIO::PROT_READ" and/or
1499 "IO::AIO::PROT_WRITE",
1500
1501 $flags can be a combination of "IO::AIO::MAP_SHARED" or
1502 "IO::AIO::MAP_PRIVATE", or a number of system-specific flags (when
1503 not available, the are defined as 0): "IO::AIO::MAP_ANONYMOUS"
1504 (which is set to "MAP_ANON" if your system only provides this
1505 constant), "IO::AIO::MAP_HUGETLB", "IO::AIO::MAP_LOCKED",
1506 "IO::AIO::MAP_NORESERVE", "IO::AIO::MAP_POPULATE" or
1507 "IO::AIO::MAP_NONBLOCK"
1508
1509 If $fh is "undef", then a file descriptor of -1 is passed.
1510
1511 $offset is the offset from the start of the file - it generally must
1512 be a multiple of "IO::AIO::PAGESIZE" and defaults to 0.
1513
1514 Example:
1515
1516 use Digest::MD5;
1517 use IO::AIO;
1518
1519 open my $fh, "<verybigfile"
1520 or die "$!";
1521
1522 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh
1523 or die "verybigfile: $!";
1524
1525 my $fast_md5 = md5 $data;
1526
1527 IO::AIO::munmap $scalar
1528 Removes a previous mmap and undefines the $scalar.
1529
1530 IO::AIO::munlock $scalar, $offset = 0, $length = undef
1531 Calls the "munlock" function, undoing the effects of a previous
1532 "aio_mlock" call (see its description for details).
1533
1534 IO::AIO::munlockall
1535 Calls the "munlockall" function.
1536
1537 On systems that do not implement "munlockall", this function returns
1538 ENOSYS, otherwise the return value of "munlockall".
1539
1540 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags
1541 Calls the GNU/Linux splice(2) syscall, if available. If $r_off or
1542 $w_off are "undef", then "NULL" is passed for these, otherwise they
1543 should be the file offset.
1544
1545 The following symbol flag values are available:
1546 "IO::AIO::SPLICE_F_MOVE", "IO::AIO::SPLICE_F_NONBLOCK",
1547 "IO::AIO::SPLICE_F_MORE" and "IO::AIO::SPLICE_F_GIFT".
1548
1549 See the splice(2) manpage for details.
1550
1551 IO::AIO::tee $r_fh, $w_fh, $length, $flags
1552 Calls the GNU/Linux tee(2) syscall, see it's manpage and the
1553 description for "IO::AIO::splice" above for details.
1554
1555EVENT LOOP INTEGRATION
1556 It is recommended to use AnyEvent::AIO to integrate IO::AIO
1557 automatically into many event loops:
1558
1559 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...)
1560 use AnyEvent::AIO;
1561
1562 You can also integrate IO::AIO manually into many event loops, here are
1563 some examples of how to do this:
1564
1565 # EV integration
1566 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb;
1567
1568 # Event integration
1569 Event->io (fd => IO::AIO::poll_fileno,
1570 poll => 'r',
1571 cb => \&IO::AIO::poll_cb);
1572
1573 # Glib/Gtk2 integration
1574 add_watch Glib::IO IO::AIO::poll_fileno,
1575 in => sub { IO::AIO::poll_cb; 1 };
1576
1577 # Tk integration
1578 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "",
1579 readable => \&IO::AIO::poll_cb);
1580
1581 # Danga::Socket integration
1582 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno =>
1583 \&IO::AIO::poll_cb);
838 1584
839 FORK BEHAVIOUR 1585 FORK BEHAVIOUR
840 This module should do "the right thing" when the process using it 1586 Usage of pthreads in a program changes the semantics of fork
841 forks: 1587 considerably. Specifically, only async-safe functions can be called
1588 after fork. Perl doesn't know about this, so in general, you cannot call
1589 fork with defined behaviour in perl if pthreads are involved. IO::AIO
1590 uses pthreads, so this applies, but many other extensions and (for
1591 inexplicable reasons) perl itself often is linked against pthreads, so
1592 this limitation applies to quite a lot of perls.
842 1593
843 Before the fork, IO::AIO enters a quiescent state where no requests 1594 This module no longer tries to fight your OS, or POSIX. That means
844 can be added in other threads and no results will be processed. 1595 IO::AIO only works in the process that loaded it. Forking is fully
845 After the fork the parent simply leaves the quiescent state and 1596 supported, but using IO::AIO in the child is not.
846 continues request/result processing, while the child frees the
847 request/result queue (so that the requests started before the fork
848 will only be handled in the parent). Threads will be started on
849 demand until the limit set in the parent process has been reached
850 again.
851 1597
852 In short: the parent will, after a short pause, continue as if fork 1598 You might get around by not *using* IO::AIO before (or after) forking.
853 had not been called, while the child will act as if IO::AIO has not 1599 You could also try to call the IO::AIO::reinit function in the child:
854 been used yet. 1600
1601 IO::AIO::reinit
1602 Abandons all current requests and I/O threads and simply
1603 reinitialises all data structures. This is not an operation
1604 supported by any standards, but happens to work on GNU/Linux and
1605 some newer BSD systems.
1606
1607 The only reasonable use for this function is to call it after
1608 forking, if "IO::AIO" was used in the parent. Calling it while
1609 IO::AIO is active in the process will result in undefined behaviour.
1610 Calling it at any time will also result in any undefined (by POSIX)
1611 behaviour.
855 1612
856 MEMORY USAGE 1613 MEMORY USAGE
857 Per-request usage: 1614 Per-request usage:
858 1615
859 Each aio request uses - depending on your architecture - around 1616 Each aio request uses - depending on your architecture - around 100-200
860 100-200 bytes of memory. In addition, stat requests need a stat 1617 bytes of memory. In addition, stat requests need a stat buffer (possibly
861 buffer (possibly a few hundred bytes), readdir requires a result 1618 a few hundred bytes), readdir requires a result buffer and so on. Perl
862 buffer and so on. Perl scalars and other data passed into aio 1619 scalars and other data passed into aio requests will also be locked and
863 requests will also be locked and will consume memory till the 1620 will consume memory till the request has entered the done state.
864 request has entered the done state.
865 1621
866 This is now awfully much, so queuing lots of requests is not usually 1622 This is not awfully much, so queuing lots of requests is not usually a
867 a problem. 1623 problem.
868 1624
869 Per-thread usage: 1625 Per-thread usage:
870 1626
871 In the execution phase, some aio requests require more memory for 1627 In the execution phase, some aio requests require more memory for
872 temporary buffers, and each thread requires a stack and other data 1628 temporary buffers, and each thread requires a stack and other data
873 structures (usually around 16k-128k, depending on the OS). 1629 structures (usually around 16k-128k, depending on the OS).
874 1630
875KNOWN BUGS 1631KNOWN BUGS
876 Known bugs will be fixed in the next release. 1632 Known bugs will be fixed in the next release.
877 1633
878SEE ALSO 1634SEE ALSO
879 Coro::AIO. 1635 AnyEvent::AIO for easy integration into event loops, Coro::AIO for a
1636 more natural syntax.
880 1637
881AUTHOR 1638AUTHOR
882 Marc Lehmann <schmorp@schmorp.de> 1639 Marc Lehmann <schmorp@schmorp.de>
883 http://home.schmorp.de/ 1640 http://home.schmorp.de/
884 1641

Diff Legend

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