ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Linux-AIO/AIO.pm
(Generate patch)

Comparing Linux-AIO/AIO.pm (file contents):
Revision 1.12 by root, Sat May 18 21:48:36 2002 UTC vs.
Revision 1.30 by root, Sat Jul 9 23:30:00 2005 UTC

6 6
7 use Linux::AIO; 7 use Linux::AIO;
8 8
9=head1 DESCRIPTION 9=head1 DESCRIPTION
10 10
11This module implements asynchroneous i/o using the means available to 11This module implements asynchronous I/O using the means available to Linux
12linux - clone. It does not hook into the POSIX aio_* functions because 12- clone. It does not hook into the POSIX aio_* functions because Linux
13linux does not yet support these in the kernel. Instead, a number of 13does not yet support these in the kernel (even as of 2.6.12, only O_DIRECT
14threads are started that execute your read/writes and signal their 14files are supported) and even if, it would only allow aio_read and write,
15completion. 15not open, stat and so on.
16
17Instead, in this module a number of (non-posix) threads are started that
18execute your read/writes and signal their completion. You don't need
19thread support in your libc or perl, and the threads created by this
20module will not be visible to the pthreads library.
21
22NOTICE: the threads created by this module will automatically be killed
23when the thread calling min_parallel exits. Make sure you only ever call
24min_parallel from the same thread that loaded this module.
25
26Although the module will work with in the presence of other threads, it is
27not reentrant, so use appropriate locking yourself.
28
29=head2 API NOTES
30
31All the C<aio_*> calls are more or less thin wrappers around the syscall
32with the same name (sans C<aio_>). The arguments are similar or identical,
33and they all accept an additional C<$callback> argument which must be
34a code reference. This code reference will get called with the syscall
35return code (e.g. most syscalls return C<-1> on error, unlike perl, which
36usually delivers "false") as it's sole argument when the given syscall has
37been executed asynchronously.
38
39All functions that expect a filehandle will also accept a file descriptor.
16 40
17=over 4 41=over 4
18 42
19=cut 43=cut
20 44
21package Linux::AIO; 45package Linux::AIO;
22 46
23use base 'Exporter'; 47use base 'Exporter';
24 48
25BEGIN { 49BEGIN {
26 $VERSION = 0.111; 50 $VERSION = 1.7;
27 51
28 @EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat); 52 @EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink
53 aio_fsync aio_fdatasync aio_readahead);
29 @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs); 54 @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs);
30 55
31 require XSLoader; 56 require XSLoader;
32 XSLoader::load Linux::AIO, $VERSION; 57 XSLoader::load Linux::AIO, $VERSION;
33} 58}
34 59
35=item Linux::AIO::min_parallel($nthreads) 60=item Linux::AIO::min_parallel $nthreads
36 61
37Set the minimum number of AIO threads to $nthreads. You I<have> to call 62Set the minimum number of AIO threads to C<$nthreads>. The default is
38this function with a positive number at leats once, otherwise no threads 63C<1>, which means a single asynchronous operation can be done at one time
39will be started and you aio-operations will seem to hang. 64(the number of outstanding operations, however, is unlimited).
65
66It is recommended to keep the number of threads low, as some linux
67kernel versions will scale negatively with the number of threads (higher
68parallelity => MUCH higher latency).
69
70Under normal circumstances you don't need to call this function, as this
71module automatically starts a single async thread.
72
73=item Linux::AIO::max_parallel $nthreads
74
75Sets the maximum number of AIO threads to C<$nthreads>. If more than
76the specified number of threads are currently running, kill them. This
77function blocks until the limit is reached.
78
79This module automatically runs C<max_parallel 0> at program end, to ensure
80that all threads are killed and that there are no outstanding requests.
81
82Under normal circumstances you don't need to call this function.
40 83
41=item $fileno = Linux::AIO::poll_fileno 84=item $fileno = Linux::AIO::poll_fileno
42 85
43Return the request result pipe filehandle. This filehandle must be polled 86Return the I<request result pipe filehandle>. This filehandle must be
44for reading. If the pipe becomes readable you have to call C<poll_cb>. 87polled for reading by some mechanism outside this module (e.g. Event
88or select, see below). If the pipe becomes readable you have to call
89C<poll_cb> to check the results.
90
91See C<poll_cb> for an example.
45 92
46=item Linux::AIO::poll_cb 93=item Linux::AIO::poll_cb
47 94
48Process all outstanding events on the result pipe. You have to call this 95Process all outstanding events on the result pipe. You have to call this
49regularly. Returns the number of events processed. 96regularly. Returns the number of events processed. Returns immediately
97when no events are outstanding.
50 98
51You can use Event to multiplex, e.g.: 99You can use Event to multiplex, e.g.:
52 100
53 Event->io(fd => Linux::AIO::poll_fileno, 101 Event->io (fd => Linux::AIO::poll_fileno,
54 poll => 'r', async => 1, 102 poll => 'r', async => 1,
55 cb => \&Linux::AIO::poll_cb ); 103 cb => \&Linux::AIO::poll_cb);
56 104
105=item Linux::AIO::poll_wait
106
107Wait till the result filehandle becomes ready for reading (simply does a
108select on the filehandle. This is useful if you want to synchronously wait
109for some requests to finish).
110
111See C<nreqs> for an example.
57 112
58=item Linux::AIO::nreqs 113=item Linux::AIO::nreqs
59 114
60Returns the number of requests currently outstanding. 115Returns the number of requests currently outstanding.
61 116
117Example: wait till there are no outstanding requests anymore:
118
119 Linux::AIO::poll_wait while Linux::AIO::nreqs;
120
62=item aio_open($pathname, $flags, $mode, $callback) 121=item aio_open $pathname, $flags, $mode, $callback
63 122
64Asynchronously open or create a file and call the callback with the 123Asynchronously open or create a file and call the callback with the
65filedescriptor. 124filedescriptor (NOT a perl filehandle, sorry for that, but watch out, this
125might change in the future).
66 126
127The C<$mode> argument is a bitmask. See the C<Fcntl> module for a
128list. They are the same as used in C<sysopen>.
129
130Example:
131
132 aio_open "/etc/passwd", O_RDONLY, 0, sub {
133 if ($_[0] >= 0) {
134 open my $fh, "<&$_[0]"; # create a copy for perl
135 aio_close $_[0], sub { }; # close the aio handle
136 print "open successful, fh is $fh\n";
137 ...
138 } else {
139 die "open failed: $!\n";
140 }
141 };
142
67=item aio_close($fh, $callback) 143=item aio_close $fh, $callback
68 144
69Asynchronously close a file and call the callback with the result code. 145Asynchronously close a file and call the callback with the result code.
70 146
71=item aio_read($fh,$offset,$length, $data,$dataoffset,$callback) 147=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback
72 148
73=item aio_write($fh,$offset,$length, $data,$dataoffset,$callback) 149=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback
74 150
75Reads or writes C<length> bytes from the specified C<fh> and C<offset> 151Reads or writes C<length> bytes from the specified C<fh> and C<offset>
76into the scalar given by C<data> and offset C<dataoffset> and calls the 152into the scalar given by C<data> and offset C<dataoffset> and calls the
77callback without the actual number of bytes read (or undef on error). 153callback without the actual number of bytes read (or -1 on error, just
154like the syscall).
78 155
156Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at
157offset C<0> within the scalar:
158
159 aio_read $fh, 7, 15, $buffer, 0, sub {
160 $_[0] >= 0 or die "read error: $!";
161 print "read <$buffer>\n";
162 };
163
164=item aio_readahead $fh,$offset,$length, $callback
165
166Asynchronously reads the specified byte range into the page cache, using
167the C<readahead> syscall.
168
169readahead() populates the page cache with data from a file so that
170subsequent reads from that file will not block on disk I/O. The C<$offset>
171argument specifies the starting point from which data is to be read and
172C<$length> specifies the number of bytes to be read. I/O is performed in
173whole pages, so that offset is effectively rounded down to a page boundary
174and bytes are read up to the next page boundary greater than or equal to
175(off-set+length). aio_readahead() does not read beyond the end of the
176file. The current file offset of the file is left unchanged.
177
79=item aio_stat($fh_or_path,$callback) 178=item aio_stat $fh_or_path, $callback
80 179
81=item aio_lstat($fh,$callback) 180=item aio_lstat $fh, $callback
82 181
83Works like perl's C<stat> or C<lstat> in void context, i.e. the callback 182Works like perl's C<stat> or C<lstat> in void context. The callback will
84will be called after the stat and the results will be available using 183be called after the stat and the results will be available using C<stat _>
85C<stat _> or C<-s _> etc... 184or C<-s _> etc...
86 185
87Currently, the stats are always 64-bit-stats, i.e. instead of returning an 186Currently, the stats are always 64-bit-stats, i.e. instead of returning an
88error when stat'ing a large file, the results will be silently truncated 187error when stat'ing a large file, the results will be silently truncated
89unless perl itself is compiled with large file support. 188unless perl itself is compiled with large file support.
90 189
190Example: Print the length of F</etc/passwd>:
191
192 aio_stat "/etc/passwd", sub {
193 $_[0] and die "stat failed: $!";
194 print "size is ", -s _, "\n";
195 };
196
197=item aio_unlink $pathname, $callback
198
199Asynchronously unlink (delete) a file and call the callback with the
200result code.
201
202=item aio_fsync $fh, $callback
203
204Asynchronously call fsync on the given filehandle and call the callback
205with the fsync result code.
206
207=item aio_fdatasync $fh, $callback
208
209Asynchronously call fdatasync on the given filehandle and call the
210callback with the fdatasync result code.
211
91=cut 212=cut
213
214min_parallel 1;
92 215
93END { 216END {
94 max_parallel 0; 217 max_parallel 0;
95} 218}
96 219
98 221
99=back 222=back
100 223
101=head1 BUGS 224=head1 BUGS
102 225
103This module has not yet been extensively tested. Watch out! 226This module has been extensively tested in a large and very busy webserver
227for many years now.
104 228
105 - perl-threads/fork interaction poorly tested.
106 - aio_open gives a fd, but all other functions expect a filehandle. 229 - aio_open gives a fd, but all other functions expect a perl filehandle.
107 230
108=head1 SEE ALSO 231=head1 SEE ALSO
109 232
110L<Coro>. 233L<Coro>.
111 234
112=head1 AUTHOR 235=head1 AUTHOR
113 236
114 Marc Lehmann <pcg@goof.com> 237 Marc Lehmann <schmorp@schmorp.de>
115 http://www.goof.com/pcg/marc/ 238 http://home.schmorp.de/
116 239
117=cut 240=cut
118 241

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines