… | |
… | |
3 | |
3 | |
4 | SYNOPSIS |
4 | SYNOPSIS |
5 | use IO::AIO; |
5 | use IO::AIO; |
6 | |
6 | |
7 | aio_open "/etc/passwd", O_RDONLY, 0, sub { |
7 | aio_open "/etc/passwd", 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 | |
14 | aio_read $fh, 30000, 1024, $buffer, 0, sub { |
15 | aio_read $fh, 30000, 1024, $buffer, 0, sub { |
15 | $_[0] > 0 or die "read error: $!"; |
16 | $_[0] > 0 or die "read error: $!"; |
16 | }; |
17 | }; |
17 | |
18 | |
18 | # AnyEvent |
19 | # version 2+ has request and group objects |
|
|
20 | use IO::AIO 2; |
|
|
21 | |
|
|
22 | aioreq_pri 4; # give next request a very high priority |
|
|
23 | my $req = aio_unlink "/tmp/file", sub { }; |
|
|
24 | $req->cancel; # cancel request if still in queue |
|
|
25 | |
|
|
26 | my $grp = aio_group sub { print "all stats done\n" }; |
|
|
27 | add $grp aio_stat "..." for ...; |
|
|
28 | |
|
|
29 | # AnyEvent integration |
19 | open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!"; |
30 | open my $fh, "<&=" . IO::AIO::poll_fileno or die "$!"; |
20 | my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb }); |
31 | my $w = AnyEvent->io (fh => $fh, poll => 'r', cb => sub { IO::AIO::poll_cb }); |
21 | |
32 | |
22 | # Event |
33 | # Event integration |
23 | Event->io (fd => IO::AIO::poll_fileno, |
34 | Event->io (fd => IO::AIO::poll_fileno, |
24 | poll => 'r', |
35 | poll => 'r', |
25 | cb => \&IO::AIO::poll_cb); |
36 | cb => \&IO::AIO::poll_cb); |
26 | |
37 | |
27 | # Glib/Gtk2 |
38 | # Glib/Gtk2 integration |
28 | add_watch Glib::IO IO::AIO::poll_fileno, |
39 | add_watch Glib::IO IO::AIO::poll_fileno, |
29 | in => sub { IO::AIO::poll_cb; 1 }; |
40 | in => sub { IO::AIO::poll_cb; 1 }; |
30 | |
41 | |
31 | # Tk |
42 | # Tk integration |
32 | Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "", |
43 | Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "", |
33 | readable => \&IO::AIO::poll_cb); |
44 | readable => \&IO::AIO::poll_cb); |
34 | |
45 | |
35 | # Danga::Socket |
46 | # Danga::Socket integration |
36 | Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => |
47 | Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => |
37 | \&IO::AIO::poll_cb); |
48 | \&IO::AIO::poll_cb); |
38 | |
49 | |
39 | DESCRIPTION |
50 | DESCRIPTION |
40 | This module implements asynchronous I/O using whatever means your |
51 | This module implements asynchronous I/O using whatever means your |
41 | operating system supports. |
52 | operating system supports. |
42 | |
53 | |
|
|
54 | Asynchronous means that operations that can normally block your program |
|
|
55 | (e.g. reading from disk) will be done asynchronously: the operation will |
|
|
56 | still block, but you can do something else in the meantime. This is |
|
|
57 | extremely useful for programs that need to stay interactive even when |
|
|
58 | doing heavy I/O (GUI programs, high performance network servers etc.), |
|
|
59 | but can also be used to easily do operations in parallel that are |
|
|
60 | normally done sequentially, e.g. stat'ing many files, which is much |
|
|
61 | faster on a RAID volume or over NFS when you do a number of stat |
|
|
62 | operations concurrently. |
|
|
63 | |
|
|
64 | While most of this works on all types of file descriptors (for example |
|
|
65 | sockets), using these functions on file descriptors that support |
|
|
66 | nonblocking operation (again, sockets, pipes etc.) is very inefficient |
|
|
67 | or might not work (aio_read fails on sockets/pipes/fifos). Use an event |
|
|
68 | loop for that (such as the Event module): IO::AIO will naturally fit |
|
|
69 | into such an event loop itself. |
|
|
70 | |
43 | Currently, a number of threads are started that execute your read/writes |
71 | In this version, a number of threads are started that execute your |
44 | and signal their completion. You don't need thread support in your libc |
72 | requests and signal their completion. You don't need thread support in |
45 | or perl, and the threads created by this module will not be visible to |
73 | perl, and the threads created by this module will not be visible to |
46 | the pthreads library. In the future, this module might make use of the |
74 | perl. In the future, this module might make use of the native aio |
47 | native aio functions available on many operating systems. However, they |
75 | functions available on many operating systems. However, they are often |
48 | are often not well-supported (Linux doesn't allow them on normal files |
76 | not well-supported or restricted (GNU/Linux doesn't allow them on normal |
49 | currently, for example), and they would only support aio_read and |
77 | files currently, for example), and they would only support aio_read and |
50 | aio_write, so the remaining functionality would have to be implemented |
78 | aio_write, so the remaining functionality would have to be implemented |
51 | using threads anyway. |
79 | using threads anyway. |
52 | |
80 | |
53 | Although the module will work with in the presence of other threads, it |
81 | Although the module will work with in the presence of other (Perl-) |
54 | is currently not reentrant, so use appropriate locking yourself, always |
82 | threads, it is currently not reentrant in any way, so use appropriate |
55 | call "poll_cb" from within the same thread, or never call "poll_cb" (or |
83 | locking yourself, always call "poll_cb" from within the same thread, or |
56 | other "aio_" functions) recursively. |
84 | never call "poll_cb" (or other "aio_" functions) recursively. |
|
|
85 | |
|
|
86 | EXAMPLE |
|
|
87 | This is a simple example that uses the Event module and loads |
|
|
88 | /etc/passwd asynchronously: |
|
|
89 | |
|
|
90 | use Fcntl; |
|
|
91 | use Event; |
|
|
92 | use IO::AIO; |
|
|
93 | |
|
|
94 | # register the IO::AIO callback with Event |
|
|
95 | Event->io (fd => IO::AIO::poll_fileno, |
|
|
96 | poll => 'r', |
|
|
97 | cb => \&IO::AIO::poll_cb); |
|
|
98 | |
|
|
99 | # queue the request to open /etc/passwd |
|
|
100 | aio_open "/etc/passwd", O_RDONLY, 0, sub { |
|
|
101 | my $fh = shift |
|
|
102 | or die "error while opening: $!"; |
|
|
103 | |
|
|
104 | # stat'ing filehandles is generally non-blocking |
|
|
105 | my $size = -s $fh; |
|
|
106 | |
|
|
107 | # queue a request to read the file |
|
|
108 | my $contents; |
|
|
109 | aio_read $fh, 0, $size, $contents, 0, sub { |
|
|
110 | $_[0] == $size |
|
|
111 | or die "short read: $!"; |
|
|
112 | |
|
|
113 | close $fh; |
|
|
114 | |
|
|
115 | # file contents now in $contents |
|
|
116 | print $contents; |
|
|
117 | |
|
|
118 | # exit event loop and program |
|
|
119 | Event::unloop; |
|
|
120 | }; |
|
|
121 | }; |
|
|
122 | |
|
|
123 | # possibly queue up other requests, or open GUI windows, |
|
|
124 | # check for sockets etc. etc. |
|
|
125 | |
|
|
126 | # process events as long as there are some: |
|
|
127 | Event::loop; |
|
|
128 | |
|
|
129 | REQUEST ANATOMY AND LIFETIME |
|
|
130 | Every "aio_*" function creates a request. which is a C data structure |
|
|
131 | not directly visible to Perl. |
|
|
132 | |
|
|
133 | If called in non-void context, every request function returns a Perl |
|
|
134 | object representing the request. In void context, nothing is returned, |
|
|
135 | which saves a bit of memory. |
|
|
136 | |
|
|
137 | The perl object is a fairly standard ref-to-hash object. The hash |
|
|
138 | contents are not used by IO::AIO so you are free to store anything you |
|
|
139 | like in it. |
|
|
140 | |
|
|
141 | During their existance, aio requests travel through the following |
|
|
142 | states, in order: |
|
|
143 | |
|
|
144 | ready |
|
|
145 | Immediately after a request is created it is put into the ready |
|
|
146 | state, waiting for a thread to execute it. |
|
|
147 | |
|
|
148 | execute |
|
|
149 | A thread has accepted the request for processing and is currently |
|
|
150 | executing it (e.g. blocking in read). |
|
|
151 | |
|
|
152 | pending |
|
|
153 | The request has been executed and is waiting for result processing. |
|
|
154 | |
|
|
155 | While request submission and execution is fully asynchronous, result |
|
|
156 | processing is not and relies on the perl interpreter calling |
|
|
157 | "poll_cb" (or another function with the same effect). |
|
|
158 | |
|
|
159 | result |
|
|
160 | The request results are processed synchronously by "poll_cb". |
|
|
161 | |
|
|
162 | The "poll_cb" function will process all outstanding aio requests by |
|
|
163 | calling their callbacks, freeing memory associated with them and |
|
|
164 | managing any groups they are contained in. |
|
|
165 | |
|
|
166 | done |
|
|
167 | Request has reached the end of its lifetime and holds no resources |
|
|
168 | anymore (except possibly for the Perl object, but its connection to |
|
|
169 | the actual aio request is severed and calling its methods will |
|
|
170 | either do nothing or result in a runtime error). |
57 | |
171 | |
58 | FUNCTIONS |
172 | FUNCTIONS |
59 | AIO FUNCTIONS |
173 | AIO REQUEST FUNCTIONS |
60 | All the "aio_*" calls are more or less thin wrappers around the syscall |
174 | All the "aio_*" calls are more or less thin wrappers around the syscall |
61 | with the same name (sans "aio_"). The arguments are similar or |
175 | with the same name (sans "aio_"). The arguments are similar or |
62 | identical, and they all accept an additional (and optional) $callback |
176 | identical, and they all accept an additional (and optional) $callback |
63 | argument which must be a code reference. This code reference will get |
177 | argument which must be a code reference. This code reference will get |
64 | called with the syscall return code (e.g. most syscalls return -1 on |
178 | called with the syscall return code (e.g. most syscalls return -1 on |
… | |
… | |
66 | argument when the given syscall has been executed asynchronously. |
180 | argument when the given syscall has been executed asynchronously. |
67 | |
181 | |
68 | All functions expecting a filehandle keep a copy of the filehandle |
182 | All functions expecting a filehandle keep a copy of the filehandle |
69 | internally until the request has finished. |
183 | internally until the request has finished. |
70 | |
184 | |
|
|
185 | All functions return request objects of type IO::AIO::REQ that allow |
|
|
186 | further manipulation of those requests while they are in-flight. |
|
|
187 | |
71 | The pathnames you pass to these routines *must* be absolute and encoded |
188 | The pathnames you pass to these routines *must* be absolute and encoded |
72 | in byte form. The reason for the former is that at the time the request |
189 | as octets. The reason for the former is that at the time the request is |
73 | is being executed, the current working directory could have changed. |
190 | being executed, the current working directory could have changed. |
74 | Alternatively, you can make sure that you never change the current |
191 | Alternatively, you can make sure that you never change the current |
75 | working directory. |
192 | working directory anywhere in the program and then use relative paths. |
76 | |
193 | |
77 | To encode pathnames to byte form, either make sure you either: a) always |
194 | To encode pathnames as octets, either make sure you either: a) always |
78 | pass in filenames you got from outside (command line, readdir etc.), b) |
195 | pass in filenames you got from outside (command line, readdir etc.) |
79 | are ASCII or ISO 8859-1, c) use the Encode module and encode your |
196 | without tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module |
80 | pathnames to the locale (or other) encoding in effect in the user |
197 | and encode your pathnames to the locale (or other) encoding in effect in |
81 | environment, d) use Glib::filename_from_unicode on unicode filenames or |
198 | the user environment, d) use Glib::filename_from_unicode on unicode |
82 | e) use something else. |
199 | filenames or e) use something else to ensure your scalar has the correct |
|
|
200 | contents. |
|
|
201 | |
|
|
202 | This works, btw. independent of the internal UTF-8 bit, which IO::AIO |
|
|
203 | handles correctly wether it is set or not. |
|
|
204 | |
|
|
205 | $prev_pri = aioreq_pri [$pri] |
|
|
206 | Returns the priority value that would be used for the next request |
|
|
207 | and, if $pri is given, sets the priority for the next aio request. |
|
|
208 | |
|
|
209 | The default priority is 0, the minimum and maximum priorities are -4 |
|
|
210 | and 4, respectively. Requests with higher priority will be serviced |
|
|
211 | first. |
|
|
212 | |
|
|
213 | The priority will be reset to 0 after each call to one of the |
|
|
214 | "aio_*" functions. |
|
|
215 | |
|
|
216 | Example: open a file with low priority, then read something from it |
|
|
217 | with higher priority so the read request is serviced before other |
|
|
218 | low priority open requests (potentially spamming the cache): |
|
|
219 | |
|
|
220 | aioreq_pri -3; |
|
|
221 | aio_open ..., sub { |
|
|
222 | return unless $_[0]; |
|
|
223 | |
|
|
224 | aioreq_pri -2; |
|
|
225 | aio_read $_[0], ..., sub { |
|
|
226 | ... |
|
|
227 | }; |
|
|
228 | }; |
|
|
229 | |
|
|
230 | aioreq_nice $pri_adjust |
|
|
231 | Similar to "aioreq_pri", but subtracts the given value from the |
|
|
232 | current priority, so the effect is cumulative. |
83 | |
233 | |
84 | aio_open $pathname, $flags, $mode, $callback->($fh) |
234 | aio_open $pathname, $flags, $mode, $callback->($fh) |
85 | Asynchronously open or create a file and call the callback with a |
235 | Asynchronously open or create a file and call the callback with a |
86 | newly created filehandle for the file. |
236 | newly created filehandle for the file. |
87 | |
237 | |
… | |
… | |
92 | They are the same as used by "sysopen". |
242 | They are the same as used by "sysopen". |
93 | |
243 | |
94 | Likewise, $mode specifies the mode of the newly created file, if it |
244 | Likewise, $mode specifies the mode of the newly created file, if it |
95 | didn't exist and "O_CREAT" has been given, just like perl's |
245 | didn't exist and "O_CREAT" has been given, just like perl's |
96 | "sysopen", except that it is mandatory (i.e. use 0 if you don't |
246 | "sysopen", except that it is mandatory (i.e. use 0 if you don't |
97 | create new files, and 0666 or 0777 if you do). |
247 | create new files, and 0666 or 0777 if you do). Note that the $mode |
|
|
248 | will be modified by the umask in effect then the request is being |
|
|
249 | executed, so better never change the umask. |
98 | |
250 | |
99 | Example: |
251 | Example: |
100 | |
252 | |
101 | aio_open "/etc/passwd", O_RDONLY, 0, sub { |
253 | aio_open "/etc/passwd", O_RDONLY, 0, sub { |
102 | if ($_[0]) { |
254 | if ($_[0]) { |
… | |
… | |
196 | |
348 | |
197 | aio_unlink $pathname, $callback->($status) |
349 | aio_unlink $pathname, $callback->($status) |
198 | Asynchronously unlink (delete) a file and call the callback with the |
350 | Asynchronously unlink (delete) a file and call the callback with the |
199 | result code. |
351 | result code. |
200 | |
352 | |
|
|
353 | aio_mknod $path, $mode, $dev, $callback->($status) |
|
|
354 | [EXPERIMENTAL] |
|
|
355 | |
|
|
356 | Asynchronously create a device node (or fifo). See mknod(2). |
|
|
357 | |
|
|
358 | The only (POSIX-) portable way of calling this function is: |
|
|
359 | |
|
|
360 | aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... |
|
|
361 | |
|
|
362 | aio_link $srcpath, $dstpath, $callback->($status) |
|
|
363 | Asynchronously create a new link to the existing object at $srcpath |
|
|
364 | at the path $dstpath and call the callback with the result code. |
|
|
365 | |
|
|
366 | aio_symlink $srcpath, $dstpath, $callback->($status) |
|
|
367 | Asynchronously create a new symbolic link to the existing object at |
|
|
368 | $srcpath at the path $dstpath and call the callback with the result |
|
|
369 | code. |
|
|
370 | |
|
|
371 | aio_readlink $path, $callback->($link) |
|
|
372 | Asynchronously read the symlink specified by $path and pass it to |
|
|
373 | the callback. If an error occurs, nothing or undef gets passed to |
|
|
374 | the callback. |
|
|
375 | |
|
|
376 | aio_rename $srcpath, $dstpath, $callback->($status) |
|
|
377 | Asynchronously rename the object at $srcpath to $dstpath, just as |
|
|
378 | rename(2) and call the callback with the result code. |
|
|
379 | |
|
|
380 | aio_mkdir $pathname, $mode, $callback->($status) |
|
|
381 | Asynchronously mkdir (create) a directory and call the callback with |
|
|
382 | the result code. $mode will be modified by the umask at the time the |
|
|
383 | request is executed, so do not change your umask. |
|
|
384 | |
201 | aio_rmdir $pathname, $callback->($status) |
385 | aio_rmdir $pathname, $callback->($status) |
202 | Asynchronously rmdir (delete) a directory and call the callback with |
386 | Asynchronously rmdir (delete) a directory and call the callback with |
203 | the result code. |
387 | the result code. |
204 | |
388 | |
205 | aio_readdir $pathname, $callback->($entries) |
389 | aio_readdir $pathname, $callback->($entries) |
… | |
… | |
208 | will not be sorted, and will NOT include the "." and ".." entries. |
392 | will not be sorted, and will NOT include the "." and ".." entries. |
209 | |
393 | |
210 | The callback a single argument which is either "undef" or an |
394 | The callback a single argument which is either "undef" or an |
211 | array-ref with the filenames. |
395 | array-ref with the filenames. |
212 | |
396 | |
|
|
397 | aio_load $path, $data, $callback->($status) |
|
|
398 | This is a composite request that tries to fully load the given file |
|
|
399 | into memory. Status is the same as with aio_read. |
|
|
400 | |
|
|
401 | aio_copy $srcpath, $dstpath, $callback->($status) |
|
|
402 | Try to copy the *file* (directories not supported as either source |
|
|
403 | or destination) from $srcpath to $dstpath and call the callback with |
|
|
404 | the 0 (error) or -1 ok. |
|
|
405 | |
|
|
406 | This is a composite request that it creates the destination file |
|
|
407 | with mode 0200 and copies the contents of the source file into it |
|
|
408 | using "aio_sendfile", followed by restoring atime, mtime, access |
|
|
409 | mode and uid/gid, in that order. |
|
|
410 | |
|
|
411 | If an error occurs, the partial destination file will be unlinked, |
|
|
412 | if possible, except when setting atime, mtime, access mode and |
|
|
413 | uid/gid, where errors are being ignored. |
|
|
414 | |
|
|
415 | aio_move $srcpath, $dstpath, $callback->($status) |
|
|
416 | Try to move the *file* (directories not supported as either source |
|
|
417 | or destination) from $srcpath to $dstpath and call the callback with |
|
|
418 | the 0 (error) or -1 ok. |
|
|
419 | |
|
|
420 | This is a composite request that tries to rename(2) the file first. |
|
|
421 | If rename files with "EXDEV", it copies the file with "aio_copy" |
|
|
422 | and, if that is successful, unlinking the $srcpath. |
|
|
423 | |
213 | aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) |
424 | aio_scandir $path, $maxreq, $callback->($dirs, $nondirs) |
214 | Scans a directory (similar to "aio_readdir") and tries to separate |
425 | Scans a directory (similar to "aio_readdir") but additionally tries |
215 | the entries of directory $path into two sets of names, ones you can |
426 | to efficiently separate the entries of directory $path into two sets |
216 | recurse into (directories), and ones you cannot recurse into |
427 | of names, directories you can recurse into (directories), and ones |
217 | (everything else). |
428 | you cannot recurse into (everything else, including symlinks to |
|
|
429 | directories). |
218 | |
430 | |
219 | "aio_scandir" is a composite request that consists of many |
431 | "aio_scandir" is a composite request that creates of many sub |
220 | aio-primitives. $maxreq specifies the maximum number of outstanding |
432 | requests_ $maxreq specifies the maximum number of outstanding aio |
221 | aio requests that this function generates. If it is "<= 0", then a |
433 | requests that this function generates. If it is "<= 0", then a |
222 | suitable default will be chosen (currently 8). |
434 | suitable default will be chosen (currently 4). |
223 | |
435 | |
224 | On error, the callback is called without arguments, otherwise it |
436 | On error, the callback is called without arguments, otherwise it |
225 | receives two array-refs with path-relative entry names. |
437 | receives two array-refs with path-relative entry names. |
226 | |
438 | |
227 | Example: |
439 | Example: |
… | |
… | |
236 | |
448 | |
237 | The "aio_readdir" cannot be avoided, but "stat()"'ing every entry |
449 | The "aio_readdir" cannot be avoided, but "stat()"'ing every entry |
238 | can. |
450 | can. |
239 | |
451 | |
240 | After reading the directory, the modification time, size etc. of the |
452 | After reading the directory, the modification time, size etc. of the |
241 | directory before and after the readdir is checked, and if they |
453 | directory before and after the readdir is checked, and if they match |
242 | match, the link count will be used to decide how many entries are |
454 | (and isn't the current time), the link count will be used to decide |
243 | directories (if >= 2). Otherwise, no knowledge of the number of |
455 | how many entries are directories (if >= 2). Otherwise, no knowledge |
244 | subdirectories will be assumed. |
456 | of the number of subdirectories will be assumed. |
245 | |
457 | |
246 | Then entires will be sorted into likely directories (everything |
458 | Then entries will be sorted into likely directories (everything |
247 | without a non-initial dot) and likely non-directories (everything |
459 | without a non-initial dot currently) and likely non-directories |
248 | else). Then every entry + "/." will be "stat"'ed, likely directories |
460 | (everything else). Then every entry plus an appended "/." will be |
249 | first. This is often faster because filesystems might detect the |
461 | "stat"'ed, likely directories first. If that succeeds, it assumes |
250 | type of the entry without reading the inode data (e.g. ext2fs |
|
|
251 | filetype feature). If that succeeds, it assumes that the entry is a |
|
|
252 | directory or a symlink to directory (which will be checked |
462 | that the entry is a directory or a symlink to directory (which will |
253 | seperately). |
463 | be checked seperately). This is often faster than stat'ing the entry |
|
|
464 | itself because filesystems might detect the type of the entry |
|
|
465 | without reading the inode data (e.g. ext2fs filetype feature). |
254 | |
466 | |
255 | If the known number of directories has been reached, the rest of the |
467 | If the known number of directories (link count - 2) has been |
256 | entries is assumed to be non-directories. |
468 | reached, the rest of the entries is assumed to be non-directories. |
|
|
469 | |
|
|
470 | This only works with certainty on POSIX (= UNIX) filesystems, which |
|
|
471 | fortunately are the vast majority of filesystems around. |
|
|
472 | |
|
|
473 | It will also likely work on non-POSIX filesystems with reduced |
|
|
474 | efficiency as those tend to return 0 or 1 as link counts, which |
|
|
475 | disables the directory counting heuristic. |
|
|
476 | |
|
|
477 | aio_rmtree $path, $callback->($status) |
|
|
478 | Delete a directory tree starting (and including) $path, return the |
|
|
479 | status of the final "rmdir" only. This is a composite request that |
|
|
480 | uses "aio_scandir" to recurse into and rmdir directories, and unlink |
|
|
481 | everything else. |
257 | |
482 | |
258 | aio_fsync $fh, $callback->($status) |
483 | aio_fsync $fh, $callback->($status) |
259 | Asynchronously call fsync on the given filehandle and call the |
484 | Asynchronously call fsync on the given filehandle and call the |
260 | callback with the fsync result code. |
485 | callback with the fsync result code. |
261 | |
486 | |
… | |
… | |
264 | callback with the fdatasync result code. |
489 | callback with the fdatasync result code. |
265 | |
490 | |
266 | If this call isn't available because your OS lacks it or it couldn't |
491 | If this call isn't available because your OS lacks it or it couldn't |
267 | be detected, it will be emulated by calling "fsync" instead. |
492 | be detected, it will be emulated by calling "fsync" instead. |
268 | |
493 | |
|
|
494 | aio_group $callback->(...) |
|
|
495 | This is a very special aio request: Instead of doing something, it |
|
|
496 | is a container for other aio requests, which is useful if you want |
|
|
497 | to bundle many requests into a single, composite, request with a |
|
|
498 | definite callback and the ability to cancel the whole request with |
|
|
499 | its subrequests. |
|
|
500 | |
|
|
501 | Returns an object of class IO::AIO::GRP. See its documentation below |
|
|
502 | for more info. |
|
|
503 | |
|
|
504 | Example: |
|
|
505 | |
|
|
506 | my $grp = aio_group sub { |
|
|
507 | print "all stats done\n"; |
|
|
508 | }; |
|
|
509 | |
|
|
510 | add $grp |
|
|
511 | (aio_stat ...), |
|
|
512 | (aio_stat ...), |
|
|
513 | ...; |
|
|
514 | |
|
|
515 | aio_nop $callback->() |
|
|
516 | This is a special request - it does nothing in itself and is only |
|
|
517 | used for side effects, such as when you want to add a dummy request |
|
|
518 | to a group so that finishing the requests in the group depends on |
|
|
519 | executing the given code. |
|
|
520 | |
|
|
521 | While this request does nothing, it still goes through the execution |
|
|
522 | phase and still requires a worker thread. Thus, the callback will |
|
|
523 | not be executed immediately but only after other requests in the |
|
|
524 | queue have entered their execution phase. This can be used to |
|
|
525 | measure request latency. |
|
|
526 | |
|
|
527 | IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED* |
|
|
528 | Mainly used for debugging and benchmarking, this aio request puts |
|
|
529 | one of the request workers to sleep for the given time. |
|
|
530 | |
|
|
531 | While it is theoretically handy to have simple I/O scheduling |
|
|
532 | requests like sleep and file handle readable/writable, the overhead |
|
|
533 | this creates is immense (it blocks a thread for a long time) so do |
|
|
534 | not use this function except to put your application under |
|
|
535 | artificial I/O pressure. |
|
|
536 | |
|
|
537 | IO::AIO::REQ CLASS |
|
|
538 | All non-aggregate "aio_*" functions return an object of this class when |
|
|
539 | called in non-void context. |
|
|
540 | |
|
|
541 | cancel $req |
|
|
542 | Cancels the request, if possible. Has the effect of skipping |
|
|
543 | execution when entering the execute state and skipping calling the |
|
|
544 | callback when entering the the result state, but will leave the |
|
|
545 | request otherwise untouched. That means that requests that currently |
|
|
546 | execute will not be stopped and resources held by the request will |
|
|
547 | not be freed prematurely. |
|
|
548 | |
|
|
549 | cb $req $callback->(...) |
|
|
550 | Replace (or simply set) the callback registered to the request. |
|
|
551 | |
|
|
552 | IO::AIO::GRP CLASS |
|
|
553 | This class is a subclass of IO::AIO::REQ, so all its methods apply to |
|
|
554 | objects of this class, too. |
|
|
555 | |
|
|
556 | A IO::AIO::GRP object is a special request that can contain multiple |
|
|
557 | other aio requests. |
|
|
558 | |
|
|
559 | You create one by calling the "aio_group" constructing function with a |
|
|
560 | callback that will be called when all contained requests have entered |
|
|
561 | the "done" state: |
|
|
562 | |
|
|
563 | my $grp = aio_group sub { |
|
|
564 | print "all requests are done\n"; |
|
|
565 | }; |
|
|
566 | |
|
|
567 | You add requests by calling the "add" method with one or more |
|
|
568 | "IO::AIO::REQ" objects: |
|
|
569 | |
|
|
570 | $grp->add (aio_unlink "..."); |
|
|
571 | |
|
|
572 | add $grp aio_stat "...", sub { |
|
|
573 | $_[0] or return $grp->result ("error"); |
|
|
574 | |
|
|
575 | # add another request dynamically, if first succeeded |
|
|
576 | add $grp aio_open "...", sub { |
|
|
577 | $grp->result ("ok"); |
|
|
578 | }; |
|
|
579 | }; |
|
|
580 | |
|
|
581 | This makes it very easy to create composite requests (see the source of |
|
|
582 | "aio_move" for an application) that work and feel like simple requests. |
|
|
583 | |
|
|
584 | * The IO::AIO::GRP objects will be cleaned up during calls to |
|
|
585 | "IO::AIO::poll_cb", just like any other request. |
|
|
586 | * They can be canceled like any other request. Canceling will cancel not |
|
|
587 | only the request itself, but also all requests it contains. |
|
|
588 | * They can also can also be added to other IO::AIO::GRP objects. |
|
|
589 | * You must not add requests to a group from within the group callback |
|
|
590 | (or any later time). |
|
|
591 | |
|
|
592 | Their lifetime, simplified, looks like this: when they are empty, they |
|
|
593 | will finish very quickly. If they contain only requests that are in the |
|
|
594 | "done" state, they will also finish. Otherwise they will continue to |
|
|
595 | exist. |
|
|
596 | |
|
|
597 | That means after creating a group you have some time to add requests. |
|
|
598 | And in the callbacks of those requests, you can add further requests to |
|
|
599 | the group. And only when all those requests have finished will the the |
|
|
600 | group itself finish. |
|
|
601 | |
|
|
602 | add $grp ... |
|
|
603 | $grp->add (...) |
|
|
604 | Add one or more requests to the group. Any type of IO::AIO::REQ can |
|
|
605 | be added, including other groups, as long as you do not create |
|
|
606 | circular dependencies. |
|
|
607 | |
|
|
608 | Returns all its arguments. |
|
|
609 | |
|
|
610 | $grp->cancel_subs |
|
|
611 | Cancel all subrequests and clears any feeder, but not the group |
|
|
612 | request itself. Useful when you queued a lot of events but got a |
|
|
613 | result early. |
|
|
614 | |
|
|
615 | $grp->result (...) |
|
|
616 | Set the result value(s) that will be passed to the group callback |
|
|
617 | when all subrequests have finished and set thre groups errno to the |
|
|
618 | current value of errno (just like calling "errno" without an error |
|
|
619 | number). By default, no argument will be passed and errno is zero. |
|
|
620 | |
|
|
621 | $grp->errno ([$errno]) |
|
|
622 | Sets the group errno value to $errno, or the current value of errno |
|
|
623 | when the argument is missing. |
|
|
624 | |
|
|
625 | Every aio request has an associated errno value that is restored |
|
|
626 | when the callback is invoked. This method lets you change this value |
|
|
627 | from its default (0). |
|
|
628 | |
|
|
629 | Calling "result" will also set errno, so make sure you either set $! |
|
|
630 | before the call to "result", or call c<errno> after it. |
|
|
631 | |
|
|
632 | feed $grp $callback->($grp) |
|
|
633 | Sets a feeder/generator on this group: every group can have an |
|
|
634 | attached generator that generates requests if idle. The idea behind |
|
|
635 | this is that, although you could just queue as many requests as you |
|
|
636 | want in a group, this might starve other requests for a potentially |
|
|
637 | long time. For example, "aio_scandir" might generate hundreds of |
|
|
638 | thousands "aio_stat" requests, delaying any later requests for a |
|
|
639 | long time. |
|
|
640 | |
|
|
641 | To avoid this, and allow incremental generation of requests, you can |
|
|
642 | instead a group and set a feeder on it that generates those |
|
|
643 | requests. The feed callback will be called whenever there are few |
|
|
644 | enough (see "limit", below) requests active in the group itself and |
|
|
645 | is expected to queue more requests. |
|
|
646 | |
|
|
647 | The feed callback can queue as many requests as it likes (i.e. "add" |
|
|
648 | does not impose any limits). |
|
|
649 | |
|
|
650 | If the feed does not queue more requests when called, it will be |
|
|
651 | automatically removed from the group. |
|
|
652 | |
|
|
653 | If the feed limit is 0, it will be set to 2 automatically. |
|
|
654 | |
|
|
655 | Example: |
|
|
656 | |
|
|
657 | # stat all files in @files, but only ever use four aio requests concurrently: |
|
|
658 | |
|
|
659 | my $grp = aio_group sub { print "finished\n" }; |
|
|
660 | limit $grp 4; |
|
|
661 | feed $grp sub { |
|
|
662 | my $file = pop @files |
|
|
663 | or return; |
|
|
664 | |
|
|
665 | add $grp aio_stat $file, sub { ... }; |
|
|
666 | }; |
|
|
667 | |
|
|
668 | limit $grp $num |
|
|
669 | Sets the feeder limit for the group: The feeder will be called |
|
|
670 | whenever the group contains less than this many requests. |
|
|
671 | |
|
|
672 | Setting the limit to 0 will pause the feeding process. |
|
|
673 | |
269 | SUPPORT FUNCTIONS |
674 | SUPPORT FUNCTIONS |
|
|
675 | EVENT PROCESSING AND EVENT LOOP INTEGRATION |
270 | $fileno = IO::AIO::poll_fileno |
676 | $fileno = IO::AIO::poll_fileno |
271 | Return the *request result pipe file descriptor*. This filehandle |
677 | Return the *request result pipe file descriptor*. This filehandle |
272 | must be polled for reading by some mechanism outside this module |
678 | must be polled for reading by some mechanism outside this module |
273 | (e.g. Event or select, see below or the SYNOPSIS). If the pipe |
679 | (e.g. Event or select, see below or the SYNOPSIS). If the pipe |
274 | becomes readable you have to call "poll_cb" to check the results. |
680 | becomes readable you have to call "poll_cb" to check the results. |
275 | |
681 | |
276 | See "poll_cb" for an example. |
682 | See "poll_cb" for an example. |
277 | |
683 | |
278 | IO::AIO::poll_cb |
684 | IO::AIO::poll_cb |
279 | Process all outstanding events on the result pipe. You have to call |
685 | Process some outstanding events on the result pipe. You have to call |
280 | this regularly. Returns the number of events processed. Returns |
686 | this regularly. Returns the number of events processed. Returns |
281 | immediately when no events are outstanding. |
687 | immediately when no events are outstanding. The amount of events |
|
|
688 | processed depends on the settings of "IO::AIO::max_poll_req" and |
|
|
689 | "IO::AIO::max_poll_time". |
|
|
690 | |
|
|
691 | If not all requests were processed for whatever reason, the |
|
|
692 | filehandle will still be ready when "poll_cb" returns. |
282 | |
693 | |
283 | Example: Install an Event watcher that automatically calls |
694 | Example: Install an Event watcher that automatically calls |
284 | IO::AIO::poll_cb with high priority: |
695 | IO::AIO::poll_cb with high priority: |
285 | |
696 | |
286 | Event->io (fd => IO::AIO::poll_fileno, |
697 | Event->io (fd => IO::AIO::poll_fileno, |
287 | poll => 'r', async => 1, |
698 | poll => 'r', async => 1, |
288 | cb => \&IO::AIO::poll_cb); |
699 | cb => \&IO::AIO::poll_cb); |
289 | |
700 | |
|
|
701 | IO::AIO::max_poll_reqs $nreqs |
|
|
702 | IO::AIO::max_poll_time $seconds |
|
|
703 | These set the maximum number of requests (default 0, meaning |
|
|
704 | infinity) that are being processed by "IO::AIO::poll_cb" in one |
|
|
705 | call, respectively the maximum amount of time (default 0, meaning |
|
|
706 | infinity) spent in "IO::AIO::poll_cb" to process requests (more |
|
|
707 | correctly the mininum amount of time "poll_cb" is allowed to use). |
|
|
708 | |
|
|
709 | Setting "max_poll_time" to a non-zero value creates an overhead of |
|
|
710 | one syscall per request processed, which is not normally a problem |
|
|
711 | unless your callbacks are really really fast or your OS is really |
|
|
712 | really slow (I am not mentioning Solaris here). Using |
|
|
713 | "max_poll_reqs" incurs no overhead. |
|
|
714 | |
|
|
715 | Setting these is useful if you want to ensure some level of |
|
|
716 | interactiveness when perl is not fast enough to process all requests |
|
|
717 | in time. |
|
|
718 | |
|
|
719 | For interactive programs, values such as 0.01 to 0.1 should be fine. |
|
|
720 | |
|
|
721 | Example: Install an Event watcher that automatically calls |
|
|
722 | IO::AIO::poll_cb with low priority, to ensure that other parts of |
|
|
723 | the program get the CPU sometimes even under high AIO load. |
|
|
724 | |
|
|
725 | # try not to spend much more than 0.1s in poll_cb |
|
|
726 | IO::AIO::max_poll_time 0.1; |
|
|
727 | |
|
|
728 | # use a low priority so other tasks have priority |
|
|
729 | Event->io (fd => IO::AIO::poll_fileno, |
|
|
730 | poll => 'r', nice => 1, |
|
|
731 | cb => &IO::AIO::poll_cb); |
|
|
732 | |
290 | IO::AIO::poll_wait |
733 | IO::AIO::poll_wait |
|
|
734 | If there are any outstanding requests and none of them in the result |
291 | Wait till the result filehandle becomes ready for reading (simply |
735 | phase, wait till the result filehandle becomes ready for reading |
292 | does a "select" on the filehandle. This is useful if you want to |
736 | (simply does a "select" on the filehandle. This is useful if you |
293 | synchronously wait for some requests to finish). |
737 | want to synchronously wait for some requests to finish). |
294 | |
738 | |
295 | See "nreqs" for an example. |
739 | See "nreqs" for an example. |
296 | |
740 | |
|
|
741 | IO::AIO::poll |
|
|
742 | Waits until some requests have been handled. |
|
|
743 | |
|
|
744 | Returns the number of requests processed, but is otherwise strictly |
|
|
745 | equivalent to: |
|
|
746 | |
|
|
747 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
748 | |
297 | IO::AIO::nreqs |
749 | IO::AIO::flush |
298 | Returns the number of requests currently outstanding (i.e. for which |
750 | Wait till all outstanding AIO requests have been handled. |
299 | their callback has not been invoked yet). |
|
|
300 | |
751 | |
301 | Example: wait till there are no outstanding requests anymore: |
752 | Strictly equivalent to: |
302 | |
753 | |
303 | IO::AIO::poll_wait, IO::AIO::poll_cb |
754 | IO::AIO::poll_wait, IO::AIO::poll_cb |
304 | while IO::AIO::nreqs; |
755 | while IO::AIO::nreqs; |
305 | |
756 | |
306 | IO::AIO::flush |
757 | CONTROLLING THE NUMBER OF THREADS |
307 | Wait till all outstanding AIO requests have been handled. |
|
|
308 | |
|
|
309 | Strictly equivalent to: |
|
|
310 | |
|
|
311 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
312 | while IO::AIO::nreqs; |
|
|
313 | |
|
|
314 | IO::AIO::poll |
|
|
315 | Waits until some requests have been handled. |
|
|
316 | |
|
|
317 | Strictly equivalent to: |
|
|
318 | |
|
|
319 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
320 | if IO::AIO::nreqs; |
|
|
321 | |
|
|
322 | IO::AIO::min_parallel $nthreads |
758 | IO::AIO::min_parallel $nthreads |
323 | Set the minimum number of AIO threads to $nthreads. The current |
759 | Set the minimum number of AIO threads to $nthreads. The current |
324 | default is 4, which means four asynchronous operations can be done |
760 | default is 8, which means eight asynchronous operations can execute |
325 | at one time (the number of outstanding operations, however, is |
761 | concurrently at any one time (the number of outstanding requests, |
326 | unlimited). |
762 | however, is unlimited). |
327 | |
763 | |
328 | IO::AIO starts threads only on demand, when an AIO request is queued |
764 | IO::AIO starts threads only on demand, when an AIO request is queued |
329 | and no free thread exists. |
765 | and no free thread exists. Please note that queueing up a hundred |
|
|
766 | requests can create demand for a hundred threads, even if it turns |
|
|
767 | out that everything is in the cache and could have been processed |
|
|
768 | faster by a single thread. |
330 | |
769 | |
331 | It is recommended to keep the number of threads low, as some Linux |
770 | It is recommended to keep the number of threads relatively low, as |
332 | kernel versions will scale negatively with the number of threads |
771 | some Linux kernel versions will scale negatively with the number of |
333 | (higher parallelity => MUCH higher latency). With current Linux 2.6 |
772 | threads (higher parallelity => MUCH higher latency). With current |
334 | versions, 4-32 threads should be fine. |
773 | Linux 2.6 versions, 4-32 threads should be fine. |
335 | |
774 | |
336 | Under most circumstances you don't need to call this function, as |
775 | Under most circumstances you don't need to call this function, as |
337 | the module selects a default that is suitable for low to moderate |
776 | the module selects a default that is suitable for low to moderate |
338 | load. |
777 | load. |
339 | |
778 | |
… | |
… | |
349 | ensure that all threads are killed and that there are no outstanding |
788 | ensure that all threads are killed and that there are no outstanding |
350 | requests. |
789 | requests. |
351 | |
790 | |
352 | Under normal circumstances you don't need to call this function. |
791 | Under normal circumstances you don't need to call this function. |
353 | |
792 | |
|
|
793 | IO::AIO::max_idle $nthreads |
|
|
794 | Limit the number of threads (default: 4) that are allowed to idle |
|
|
795 | (i.e., threads that did not get a request to process within 10 |
|
|
796 | seconds). That means if a thread becomes idle while $nthreads other |
|
|
797 | threads are also idle, it will free its resources and exit. |
|
|
798 | |
|
|
799 | This is useful when you allow a large number of threads (e.g. 100 or |
|
|
800 | 1000) to allow for extremely high load situations, but want to free |
|
|
801 | resources under normal circumstances (1000 threads can easily |
|
|
802 | consume 30MB of RAM). |
|
|
803 | |
|
|
804 | The default is probably ok in most situations, especially if thread |
|
|
805 | creation is fast. If thread creation is very slow on your system you |
|
|
806 | might want to use larger values. |
|
|
807 | |
354 | $oldnreqs = IO::AIO::max_outstanding $nreqs |
808 | $oldmaxreqs = IO::AIO::max_outstanding $maxreqs |
|
|
809 | This is a very bad function to use in interactive programs because |
|
|
810 | it blocks, and a bad way to reduce concurrency because it is |
|
|
811 | inexact: Better use an "aio_group" together with a feed callback. |
|
|
812 | |
355 | Sets the maximum number of outstanding requests to $nreqs. If you |
813 | Sets the maximum number of outstanding requests to $nreqs. If you to |
356 | try to queue up more than this number of requests, the caller will |
814 | queue up more than this number of requests, the next call to the |
357 | block until some requests have been handled. |
815 | "poll_cb" (and "poll_some" and other functions calling "poll_cb") |
|
|
816 | function will block until the limit is no longer exceeded. |
358 | |
817 | |
359 | The default is very large, so normally there is no practical limit. |
818 | The default value is very large, so there is no practical limit on |
360 | If you queue up many requests in a loop it often improves speed if |
819 | the number of outstanding requests. |
361 | you set this to a relatively low number, such as 100. |
|
|
362 | |
820 | |
363 | Under normal circumstances you don't need to call this function. |
821 | You can still queue as many requests as you want. Therefore, |
|
|
822 | "max_oustsanding" is mainly useful in simple scripts (with low |
|
|
823 | values) or as a stop gap to shield against fatal memory overflow |
|
|
824 | (with large values). |
|
|
825 | |
|
|
826 | STATISTICAL INFORMATION |
|
|
827 | IO::AIO::nreqs |
|
|
828 | Returns the number of requests currently in the ready, execute or |
|
|
829 | pending states (i.e. for which their callback has not been invoked |
|
|
830 | yet). |
|
|
831 | |
|
|
832 | Example: wait till there are no outstanding requests anymore: |
|
|
833 | |
|
|
834 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
835 | while IO::AIO::nreqs; |
|
|
836 | |
|
|
837 | IO::AIO::nready |
|
|
838 | Returns the number of requests currently in the ready state (not yet |
|
|
839 | executed). |
|
|
840 | |
|
|
841 | IO::AIO::npending |
|
|
842 | Returns the number of requests currently in the pending state |
|
|
843 | (executed, but not yet processed by poll_cb). |
364 | |
844 | |
365 | FORK BEHAVIOUR |
845 | FORK BEHAVIOUR |
|
|
846 | This module should do "the right thing" when the process using it forks: |
|
|
847 | |
366 | Before the fork, IO::AIO enters a quiescent state where no requests can |
848 | Before the fork, IO::AIO enters a quiescent state where no requests can |
367 | be added in other threads and no results will be processed. After the |
849 | be added in other threads and no results will be processed. After the |
368 | fork the parent simply leaves the quiescent state and continues |
850 | fork the parent simply leaves the quiescent state and continues |
369 | request/result processing, while the child clears the request/result |
851 | request/result processing, while the child frees the request/result |
370 | queue (so the requests started before the fork will only be handled in |
852 | queue (so that the requests started before the fork will only be handled |
371 | the parent). Threats will be started on demand until the limit ste in |
853 | in the parent). Threads will be started on demand until the limit set in |
372 | the parent process has been reached again. |
854 | the parent process has been reached again. |
373 | |
855 | |
|
|
856 | In short: the parent will, after a short pause, continue as if fork had |
|
|
857 | not been called, while the child will act as if IO::AIO has not been |
|
|
858 | used yet. |
|
|
859 | |
|
|
860 | MEMORY USAGE |
|
|
861 | Per-request usage: |
|
|
862 | |
|
|
863 | Each aio request uses - depending on your architecture - around 100-200 |
|
|
864 | bytes of memory. In addition, stat requests need a stat buffer (possibly |
|
|
865 | a few hundred bytes), readdir requires a result buffer and so on. Perl |
|
|
866 | scalars and other data passed into aio requests will also be locked and |
|
|
867 | will consume memory till the request has entered the done state. |
|
|
868 | |
|
|
869 | This is now awfully much, so queuing lots of requests is not usually a |
|
|
870 | problem. |
|
|
871 | |
|
|
872 | Per-thread usage: |
|
|
873 | |
|
|
874 | In the execution phase, some aio requests require more memory for |
|
|
875 | temporary buffers, and each thread requires a stack and other data |
|
|
876 | structures (usually around 16k-128k, depending on the OS). |
|
|
877 | |
|
|
878 | KNOWN BUGS |
|
|
879 | Known bugs will be fixed in the next release. |
|
|
880 | |
374 | SEE ALSO |
881 | SEE ALSO |
375 | Coro, Linux::AIO. |
882 | Coro::AIO. |
376 | |
883 | |
377 | AUTHOR |
884 | AUTHOR |
378 | Marc Lehmann <schmorp@schmorp.de> |
885 | Marc Lehmann <schmorp@schmorp.de> |
379 | http://home.schmorp.de/ |
886 | http://home.schmorp.de/ |
380 | |
887 | |