… | |
… | |
3 | IO::AIO - Asynchronous Input/Output |
3 | IO::AIO - Asynchronous Input/Output |
4 | |
4 | |
5 | =head1 SYNOPSIS |
5 | =head1 SYNOPSIS |
6 | |
6 | |
7 | use IO::AIO; |
7 | use IO::AIO; |
|
|
8 | |
|
|
9 | aio_open "/etc/passwd", O_RDONLY, 0, sub { |
|
|
10 | my ($fh) = @_; |
|
|
11 | ... |
|
|
12 | }; |
|
|
13 | |
|
|
14 | aio_unlink "/tmp/file", sub { }; |
|
|
15 | |
|
|
16 | aio_read $fh, 30000, 1024, $buffer, 0, sub { |
|
|
17 | $_[0] > 0 or die "read error: $!"; |
|
|
18 | }; |
|
|
19 | |
|
|
20 | # Event |
|
|
21 | Event->io (fd => IO::AIO::poll_fileno, |
|
|
22 | poll => 'r', |
|
|
23 | cb => \&IO::AIO::poll_cb); |
|
|
24 | |
|
|
25 | # Glib/Gtk2 |
|
|
26 | add_watch Glib::IO IO::AIO::poll_fileno, |
|
|
27 | in => sub { IO::AIO::poll_cb; 1 }; |
|
|
28 | |
|
|
29 | # Tk |
|
|
30 | Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "", |
|
|
31 | readable => \&IO::AIO::poll_cb); |
|
|
32 | |
|
|
33 | # Danga::Socket |
|
|
34 | Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => |
|
|
35 | \&IO::AIO::poll_cb); |
|
|
36 | |
8 | |
37 | |
9 | =head1 DESCRIPTION |
38 | =head1 DESCRIPTION |
10 | |
39 | |
11 | This module implements asynchronous I/O using whatever means your |
40 | This module implements asynchronous I/O using whatever means your |
12 | operating system supports. |
41 | operating system supports. |
… | |
… | |
19 | not well-supported (Linux doesn't allow them on normal files currently, |
48 | not well-supported (Linux doesn't allow them on normal files currently, |
20 | for example), and they would only support aio_read and aio_write, so the |
49 | for example), and they would only support aio_read and aio_write, so the |
21 | remaining functionality would have to be implemented using threads anyway. |
50 | remaining functionality would have to be implemented using threads anyway. |
22 | |
51 | |
23 | Although the module will work with in the presence of other threads, it is |
52 | Although the module will work with in the presence of other threads, it is |
24 | currently not reentrant, so use appropriate locking yourself. |
53 | currently not reentrant, so use appropriate locking yourself, always call |
25 | |
54 | C<poll_cb> from within the same thread, or never call C<poll_cb> (or other |
26 | =head2 API NOTES |
55 | C<aio_> functions) recursively. |
27 | |
|
|
28 | All the C<aio_*> calls are more or less thin wrappers around the syscall |
|
|
29 | with the same name (sans C<aio_>). The arguments are similar or identical, |
|
|
30 | and they all accept an additional C<$callback> argument which must be |
|
|
31 | a code reference. This code reference will get called with the syscall |
|
|
32 | return code (e.g. most syscalls return C<-1> on error, unlike perl, which |
|
|
33 | usually delivers "false") as it's sole argument when the given syscall has |
|
|
34 | been executed asynchronously. |
|
|
35 | |
|
|
36 | All functions that expect a filehandle will also accept a file descriptor. |
|
|
37 | |
|
|
38 | The filenames you pass to these routines I<must> be absolute. The reason |
|
|
39 | is that at the time the request is being executed, the current working |
|
|
40 | directory could have changed. Alternatively, you can make sure that you |
|
|
41 | never change the current working directory. |
|
|
42 | |
|
|
43 | =over 4 |
|
|
44 | |
56 | |
45 | =cut |
57 | =cut |
46 | |
58 | |
47 | package IO::AIO; |
59 | package IO::AIO; |
48 | |
60 | |
|
|
61 | no warnings; |
|
|
62 | |
49 | use base 'Exporter'; |
63 | use base 'Exporter'; |
50 | |
64 | |
51 | use Fcntl (); |
65 | use Fcntl (); |
52 | |
66 | |
53 | BEGIN { |
67 | BEGIN { |
54 | $VERSION = 0.1; |
68 | $VERSION = 1.2; |
55 | |
69 | |
56 | @EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink |
70 | @EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink |
57 | aio_fsync aio_fdatasync aio_readahead); |
71 | aio_rmdir aio_symlink aio_fsync aio_fdatasync aio_readahead); |
58 | @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs); |
72 | @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs); |
59 | |
73 | |
60 | require XSLoader; |
74 | require XSLoader; |
61 | XSLoader::load IO::AIO, $VERSION; |
75 | XSLoader::load IO::AIO, $VERSION; |
62 | } |
76 | } |
63 | |
77 | |
64 | =item IO::AIO::min_parallel $nthreads |
78 | =head1 FUNCTIONS |
65 | |
79 | |
66 | Set the minimum number of AIO threads to C<$nthreads>. The default is |
80 | =head2 AIO FUNCTIONS |
67 | C<1>, which means a single asynchronous operation can be done at one time |
|
|
68 | (the number of outstanding operations, however, is unlimited). |
|
|
69 | |
81 | |
70 | It is recommended to keep the number of threads low, as some linux |
82 | All the C<aio_*> calls are more or less thin wrappers around the syscall |
71 | kernel versions will scale negatively with the number of threads (higher |
83 | with the same name (sans C<aio_>). The arguments are similar or identical, |
72 | parallelity => MUCH higher latency). |
84 | and they all accept an additional (and optional) C<$callback> argument |
|
|
85 | which must be a code reference. This code reference will get called with |
|
|
86 | the syscall return code (e.g. most syscalls return C<-1> on error, unlike |
|
|
87 | perl, which usually delivers "false") as it's sole argument when the given |
|
|
88 | syscall has been executed asynchronously. |
73 | |
89 | |
74 | Under normal circumstances you don't need to call this function, as this |
90 | All functions expecting a filehandle keep a copy of the filehandle |
75 | module automatically starts a single async thread. |
91 | internally until the request has finished. |
76 | |
92 | |
77 | =item IO::AIO::max_parallel $nthreads |
93 | The pathnames you pass to these routines I<must> be absolute and |
|
|
94 | encoded in byte form. The reason for the former is that at the time the |
|
|
95 | request is being executed, the current working directory could have |
|
|
96 | changed. Alternatively, you can make sure that you never change the |
|
|
97 | current working directory. |
78 | |
98 | |
79 | Sets the maximum number of AIO threads to C<$nthreads>. If more than |
99 | To encode pathnames to byte form, either make sure you either: a) |
80 | the specified number of threads are currently running, kill them. This |
100 | always pass in filenames you got from outside (command line, readdir |
81 | function blocks until the limit is reached. |
101 | etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode |
|
|
102 | your pathnames to the locale (or other) encoding in effect in the user |
|
|
103 | environment, d) use Glib::filename_from_unicode on unicode filenames or e) |
|
|
104 | use something else. |
82 | |
105 | |
83 | This module automatically runs C<max_parallel 0> at program end, to ensure |
106 | =over 4 |
84 | that all threads are killed and that there are no outstanding requests. |
|
|
85 | |
|
|
86 | Under normal circumstances you don't need to call this function. |
|
|
87 | |
|
|
88 | =item $fileno = IO::AIO::poll_fileno |
|
|
89 | |
|
|
90 | Return the I<request result pipe filehandle>. This filehandle must be |
|
|
91 | polled for reading by some mechanism outside this module (e.g. Event |
|
|
92 | or select, see below). If the pipe becomes readable you have to call |
|
|
93 | C<poll_cb> to check the results. |
|
|
94 | |
|
|
95 | See C<poll_cb> for an example. |
|
|
96 | |
|
|
97 | =item IO::AIO::poll_cb |
|
|
98 | |
|
|
99 | Process all outstanding events on the result pipe. You have to call this |
|
|
100 | regularly. Returns the number of events processed. Returns immediately |
|
|
101 | when no events are outstanding. |
|
|
102 | |
|
|
103 | You can use Event to multiplex, e.g.: |
|
|
104 | |
|
|
105 | Event->io (fd => IO::AIO::poll_fileno, |
|
|
106 | poll => 'r', async => 1, |
|
|
107 | cb => \&IO::AIO::poll_cb); |
|
|
108 | |
|
|
109 | =item IO::AIO::poll_wait |
|
|
110 | |
|
|
111 | Wait till the result filehandle becomes ready for reading (simply does a |
|
|
112 | select on the filehandle. This is useful if you want to synchronously wait |
|
|
113 | for some requests to finish). |
|
|
114 | |
|
|
115 | See C<nreqs> for an example. |
|
|
116 | |
|
|
117 | =item IO::AIO::nreqs |
|
|
118 | |
|
|
119 | Returns the number of requests currently outstanding. |
|
|
120 | |
|
|
121 | Example: wait till there are no outstanding requests anymore: |
|
|
122 | |
|
|
123 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
124 | while IO::AIO::nreqs; |
|
|
125 | |
107 | |
126 | =item aio_open $pathname, $flags, $mode, $callback |
108 | =item aio_open $pathname, $flags, $mode, $callback |
127 | |
109 | |
128 | Asynchronously open or create a file and call the callback with a newly |
110 | Asynchronously open or create a file and call the callback with a newly |
129 | created filehandle for the file. |
111 | created filehandle for the file. |
130 | |
112 | |
131 | The pathname passed to C<aio_open> must be absolute. See API NOTES, above, |
113 | The pathname passed to C<aio_open> must be absolute. See API NOTES, above, |
132 | for an explanation. |
114 | for an explanation. |
133 | |
115 | |
134 | The C<$mode> argument is a bitmask. See the C<Fcntl> module for a |
116 | The C<$flags> argument is a bitmask. See the C<Fcntl> module for a |
135 | list. They are the same as used in C<sysopen>. |
117 | list. They are the same as used by C<sysopen>. |
|
|
118 | |
|
|
119 | Likewise, C<$mode> specifies the mode of the newly created file, if it |
|
|
120 | didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>, |
|
|
121 | except that it is mandatory (i.e. use C<0> if you don't create new files, |
|
|
122 | and C<0666> or C<0777> if you do). |
136 | |
123 | |
137 | Example: |
124 | Example: |
138 | |
125 | |
139 | aio_open "/etc/passwd", O_RDONLY, 0, sub { |
126 | aio_open "/etc/passwd", O_RDONLY, 0, sub { |
140 | if ($_[0]) { |
127 | if ($_[0]) { |
… | |
… | |
147 | |
134 | |
148 | =item aio_close $fh, $callback |
135 | =item aio_close $fh, $callback |
149 | |
136 | |
150 | Asynchronously close a file and call the callback with the result |
137 | Asynchronously close a file and call the callback with the result |
151 | code. I<WARNING:> although accepted, you should not pass in a perl |
138 | code. I<WARNING:> although accepted, you should not pass in a perl |
152 | filehandle here, as perl will likely close the file descriptor itself when |
139 | filehandle here, as perl will likely close the file descriptor another |
153 | the filehandle is destroyed. Normally, you can safely call perls C<close> |
140 | time when the filehandle is destroyed. Normally, you can safely call perls |
154 | or just let filehandles go out of scope. |
141 | C<close> or just let filehandles go out of scope. |
|
|
142 | |
|
|
143 | This is supposed to be a bug in the API, so that might change. It's |
|
|
144 | therefore best to avoid this function. |
155 | |
145 | |
156 | =item aio_read $fh,$offset,$length, $data,$dataoffset,$callback |
146 | =item aio_read $fh,$offset,$length, $data,$dataoffset,$callback |
157 | |
147 | |
158 | =item aio_write $fh,$offset,$length, $data,$dataoffset,$callback |
148 | =item aio_write $fh,$offset,$length, $data,$dataoffset,$callback |
159 | |
149 | |
160 | Reads or writes C<length> bytes from the specified C<fh> and C<offset> |
150 | Reads or writes C<length> bytes from the specified C<fh> and C<offset> |
161 | into the scalar given by C<data> and offset C<dataoffset> and calls the |
151 | into the scalar given by C<data> and offset C<dataoffset> and calls the |
162 | callback without the actual number of bytes read (or -1 on error, just |
152 | callback without the actual number of bytes read (or -1 on error, just |
163 | like the syscall). |
153 | like the syscall). |
164 | |
154 | |
165 | Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at |
155 | Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at |
166 | offset C<0> within the scalar: |
156 | offset C<0> within the scalar: |
167 | |
157 | |
168 | aio_read $fh, 7, 15, $buffer, 0, sub { |
158 | aio_read $fh, 7, 15, $buffer, 0, sub { |
169 | $_[0] >= 0 or die "read error: $!"; |
159 | $_[0] > 0 or die "read error: $!"; |
170 | print "read <$buffer>\n"; |
160 | print "read $_[0] bytes: <$buffer>\n"; |
171 | }; |
161 | }; |
172 | |
162 | |
173 | =item aio_readahead $fh,$offset,$length, $callback |
163 | =item aio_readahead $fh,$offset,$length, $callback |
174 | |
164 | |
175 | Asynchronously reads the specified byte range into the page cache, using |
|
|
176 | the C<readahead> syscall. If that syscall doesn't exist the status will be |
|
|
177 | C<-1> and C<$!> is set to ENOSYS. |
|
|
178 | |
|
|
179 | readahead() populates the page cache with data from a file so that |
165 | C<aio_readahead> populates the page cache with data from a file so that |
180 | subsequent reads from that file will not block on disk I/O. The C<$offset> |
166 | subsequent reads from that file will not block on disk I/O. The C<$offset> |
181 | argument specifies the starting point from which data is to be read and |
167 | argument specifies the starting point from which data is to be read and |
182 | C<$length> specifies the number of bytes to be read. I/O is performed in |
168 | C<$length> specifies the number of bytes to be read. I/O is performed in |
183 | whole pages, so that offset is effectively rounded down to a page boundary |
169 | whole pages, so that offset is effectively rounded down to a page boundary |
184 | and bytes are read up to the next page boundary greater than or equal to |
170 | and bytes are read up to the next page boundary greater than or equal to |
185 | (off-set+length). aio_readahead() does not read beyond the end of the |
171 | (off-set+length). C<aio_readahead> does not read beyond the end of the |
186 | file. The current file offset of the file is left unchanged. |
172 | file. The current file offset of the file is left unchanged. |
|
|
173 | |
|
|
174 | If that syscall doesn't exist (likely if your OS isn't Linux) it will be |
|
|
175 | emulated by simply reading the data, which would have a similar effect. |
187 | |
176 | |
188 | =item aio_stat $fh_or_path, $callback |
177 | =item aio_stat $fh_or_path, $callback |
189 | |
178 | |
190 | =item aio_lstat $fh, $callback |
179 | =item aio_lstat $fh, $callback |
191 | |
180 | |
… | |
… | |
210 | =item aio_unlink $pathname, $callback |
199 | =item aio_unlink $pathname, $callback |
211 | |
200 | |
212 | Asynchronously unlink (delete) a file and call the callback with the |
201 | Asynchronously unlink (delete) a file and call the callback with the |
213 | result code. |
202 | result code. |
214 | |
203 | |
|
|
204 | =item aio_rmdir $pathname, $callback |
|
|
205 | |
|
|
206 | Asynchronously rmdir (delete) a directory and call the callback with the |
|
|
207 | result code. |
|
|
208 | |
215 | =item aio_fsync $fh, $callback |
209 | =item aio_fsync $fh, $callback |
216 | |
210 | |
217 | Asynchronously call fsync on the given filehandle and call the callback |
211 | Asynchronously call fsync on the given filehandle and call the callback |
218 | with the fsync result code. |
212 | with the fsync result code. |
219 | |
213 | |
220 | =item aio_fdatasync $fh, $callback |
214 | =item aio_fdatasync $fh, $callback |
221 | |
215 | |
222 | Asynchronously call fdatasync on the given filehandle and call the |
216 | Asynchronously call fdatasync on the given filehandle and call the |
223 | callback with the fdatasync result code. |
217 | callback with the fdatasync result code. |
|
|
218 | |
|
|
219 | If this call isn't available because your OS lacks it or it couldn't be |
|
|
220 | detected, it will be emulated by calling C<fsync> instead. |
|
|
221 | |
|
|
222 | =back |
|
|
223 | |
|
|
224 | =head2 SUPPORT FUNCTIONS |
|
|
225 | |
|
|
226 | =over 4 |
|
|
227 | |
|
|
228 | =item $fileno = IO::AIO::poll_fileno |
|
|
229 | |
|
|
230 | Return the I<request result pipe file descriptor>. This filehandle must be |
|
|
231 | polled for reading by some mechanism outside this module (e.g. Event or |
|
|
232 | select, see below or the SYNOPSIS). If the pipe becomes readable you have |
|
|
233 | to call C<poll_cb> to check the results. |
|
|
234 | |
|
|
235 | See C<poll_cb> for an example. |
|
|
236 | |
|
|
237 | =item IO::AIO::poll_cb |
|
|
238 | |
|
|
239 | Process all outstanding events on the result pipe. You have to call this |
|
|
240 | regularly. Returns the number of events processed. Returns immediately |
|
|
241 | when no events are outstanding. |
|
|
242 | |
|
|
243 | Example: Install an Event watcher that automatically calls |
|
|
244 | IO::AIO::poll_cb with high priority: |
|
|
245 | |
|
|
246 | Event->io (fd => IO::AIO::poll_fileno, |
|
|
247 | poll => 'r', async => 1, |
|
|
248 | cb => \&IO::AIO::poll_cb); |
|
|
249 | |
|
|
250 | =item IO::AIO::poll_wait |
|
|
251 | |
|
|
252 | Wait till the result filehandle becomes ready for reading (simply does a |
|
|
253 | C<select> on the filehandle. This is useful if you want to synchronously wait |
|
|
254 | for some requests to finish). |
|
|
255 | |
|
|
256 | See C<nreqs> for an example. |
|
|
257 | |
|
|
258 | =item IO::AIO::nreqs |
|
|
259 | |
|
|
260 | Returns the number of requests currently outstanding (i.e. for which their |
|
|
261 | callback has not been invoked yet). |
|
|
262 | |
|
|
263 | Example: wait till there are no outstanding requests anymore: |
|
|
264 | |
|
|
265 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
266 | while IO::AIO::nreqs; |
|
|
267 | |
|
|
268 | =item IO::AIO::flush |
|
|
269 | |
|
|
270 | Wait till all outstanding AIO requests have been handled. |
|
|
271 | |
|
|
272 | Strictly equivalent to: |
|
|
273 | |
|
|
274 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
275 | while IO::AIO::nreqs; |
|
|
276 | |
|
|
277 | =item IO::AIO::poll |
|
|
278 | |
|
|
279 | Waits until some requests have been handled. |
|
|
280 | |
|
|
281 | Strictly equivalent to: |
|
|
282 | |
|
|
283 | IO::AIO::poll_wait, IO::AIO::poll_cb |
|
|
284 | if IO::AIO::nreqs; |
|
|
285 | |
|
|
286 | =item IO::AIO::min_parallel $nthreads |
|
|
287 | |
|
|
288 | Set the minimum number of AIO threads to C<$nthreads>. The default is |
|
|
289 | C<1>, which means a single asynchronous operation can be done at one time |
|
|
290 | (the number of outstanding operations, however, is unlimited). |
|
|
291 | |
|
|
292 | It is recommended to keep the number of threads low, as some Linux |
|
|
293 | kernel versions will scale negatively with the number of threads (higher |
|
|
294 | parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32 |
|
|
295 | threads should be fine. |
|
|
296 | |
|
|
297 | Under normal circumstances you don't need to call this function, as this |
|
|
298 | module automatically starts some threads (the exact number might change, |
|
|
299 | and is currently 4). |
|
|
300 | |
|
|
301 | =item IO::AIO::max_parallel $nthreads |
|
|
302 | |
|
|
303 | Sets the maximum number of AIO threads to C<$nthreads>. If more than |
|
|
304 | the specified number of threads are currently running, kill them. This |
|
|
305 | function blocks until the limit is reached. |
|
|
306 | |
|
|
307 | This module automatically runs C<max_parallel 0> at program end, to ensure |
|
|
308 | that all threads are killed and that there are no outstanding requests. |
|
|
309 | |
|
|
310 | Under normal circumstances you don't need to call this function. |
|
|
311 | |
|
|
312 | =item $oldnreqs = IO::AIO::max_outstanding $nreqs |
|
|
313 | |
|
|
314 | Sets the maximum number of outstanding requests to C<$nreqs>. If you |
|
|
315 | try to queue up more than this number of requests, the caller will block until |
|
|
316 | some requests have been handled. |
|
|
317 | |
|
|
318 | The default is very large, so normally there is no practical limit. If you |
|
|
319 | queue up many requests in a loop it it often improves speed if you set |
|
|
320 | this to a relatively low number, such as C<100>. |
|
|
321 | |
|
|
322 | Under normal circumstances you don't need to call this function. |
|
|
323 | |
|
|
324 | =back |
224 | |
325 | |
225 | =cut |
326 | =cut |
226 | |
327 | |
227 | # support function to convert a fd into a perl filehandle |
328 | # support function to convert a fd into a perl filehandle |
228 | sub _fd2fh { |
329 | sub _fd2fh { |
229 | return undef if $_[0] < 0; |
330 | return undef if $_[0] < 0; |
230 | |
331 | |
231 | # try to be perl5.6-compatible |
332 | # try to generate nice filehandles |
232 | local *AIO_FH; |
333 | my $sym = "IO::AIO::fd#$_[0]"; |
233 | open AIO_FH, "+<&=$_[0]" |
334 | local *$sym; |
|
|
335 | |
|
|
336 | open *$sym, "+<&=$_[0]" # usually works under any unix |
|
|
337 | or open *$sym, "<&=$_[0]" # cygwin needs this |
|
|
338 | or open *$sym, ">&=$_[0]" # or this |
234 | or return undef; |
339 | or return undef; |
235 | |
340 | |
236 | *AIO_FH |
341 | *$sym |
237 | } |
342 | } |
238 | |
343 | |
239 | min_parallel 4; |
344 | min_parallel 4; |
240 | |
345 | |
241 | END { |
346 | END { |
242 | max_parallel 0; |
347 | max_parallel 0; |
243 | } |
348 | } |
244 | |
349 | |
245 | 1; |
350 | 1; |
246 | |
351 | |
247 | =back |
352 | =head2 FORK BEHAVIOUR |
248 | |
353 | |
249 | =head1 BUGS |
354 | Before the fork IO::AIO first handles all outstanding requests - if other |
250 | |
355 | threads add requests during this period, this time is prolonged. It then |
251 | - could be optimized to use more semaphores instead of filehandles. |
356 | enters a quiescent state where no requests can be added in other threads |
|
|
357 | and no results will be processed. After the fork the parent simply leaves |
|
|
358 | the quiescent state and continues request processing, while the child will |
|
|
359 | free the request and result queue and start the same number of threads as |
|
|
360 | were in use by the parent. |
252 | |
361 | |
253 | =head1 SEE ALSO |
362 | =head1 SEE ALSO |
254 | |
363 | |
255 | L<Coro>, L<Linux::AIO>. |
364 | L<Coro>, L<Linux::AIO>. |
256 | |
365 | |