Linux-AIO/AIO.pm

=head1 NAME

Linux::AIO - linux-specific aio implemented using clone

=head1 SYNOPSIS

 use Linux::AIO;

=head1 DESCRIPTION

This module implements asynchronous i/o using the means available to linux
- clone. It does not hook into the POSIX aio_* functions because linux
does not yet support these in the kernel (and even if, it would only allow
aio_read and write, not open and stat).

Instead, in this module a number of (non-posix) threads are started that
execute your read/writes and signal their completion. You don't need
thread support in your libc or perl, and the threads created by this
module will not be visible to the pthreads library.

NOTICE: the threads created by this module will automatically be killed
when the thread calling min_parallel exits. Make sure you only ever call
min_parallel from the same thread that loaded this module.

Although the module will work with threads, it is not reentrant, so use
appropriate locking yourself.

=over 4

=cut

package Linux::AIO;

use base 'Exporter';

BEGIN {
   $VERSION = 1.3;

   @EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink);
   @EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs);

   require XSLoader;
   XSLoader::load Linux::AIO, $VERSION;
}

=item Linux::AIO::min_parallel $nthreads

Set the minimum number of AIO threads to C<$nthreads>. The default is
C<1>, which means a single asynchronous operation can be done at one time
(the number of outstanding operations, however, is unlimited).

It is recommended to keep the number of threads low, as many linux
kernel versions will scale negatively with the number of threads (higher
parallelity => MUCH higher latency).

=item $fileno = Linux::AIO::poll_fileno

Return the I<request result pipe filehandle>. This filehandle must be
polled for reading by some mechanism outside this module (e.g. Event
or select, see below). If the pipe becomes readable you have to call
C<poll_cb> to check the results.

=item Linux::AIO::poll_cb

Process all outstanding events on the result pipe. You have to call this
regularly. Returns the number of events processed. Returns immediately
when no events are outstanding.

You can use Event to multiplex, e.g.:

   Event->io (fd => Linux::AIO::poll_fileno,
              poll => 'r', async => 1,
              cb => \&Linux::AIO::poll_cb );


=item Linux::AIO::nreqs

Returns the number of requests currently outstanding.

=item aio_open  $pathname, $flags, $mode, $callback

Asynchronously open or create a file and call the callback with the
filedescriptor (NOT a perl filehandle, sorry for that, but watch out, this
might change in the future).

=item aio_close $fh, $callback

Asynchronously close a file and call the callback with the result code.

=item aio_read  $fh,$offset,$length, $data,$dataoffset,$callback

=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback

Reads or writes C<length> bytes from the specified C<fh> and C<offset>
into the scalar given by C<data> and offset C<dataoffset> and calls the
callback without the actual number of bytes read (or C<undef> on error).

=item aio_stat  $fh_or_path, $callback

=item aio_lstat $fh, $callback

Works like perl's C<stat> or C<lstat> in void context. The callback will
be called after the stat and the results will be available using C<stat _>
or C<-s _> etc...

Currently, the stats are always 64-bit-stats, i.e. instead of returning an
error when stat'ing a large file, the results will be silently truncated
unless perl itself is compiled with large file support.

=item aio_unlink  $pathname, $callback

Asynchronously unlink a file.

=cut

min_parallel 1;

END {
   max_parallel 0;
}

1;

=back

=head1 BUGS

This module has been extensively tested in a large and very busy webserver
for many years now.

   - aio_open gives a fd, but all other functions expect a perl filehandle.

=head1 SEE ALSO

L<Coro>.

=head1 AUTHOR

 Marc Lehmann <pcg@goof.com>
 http://home.schmorp.de/

=cut

Revision:	1.20
Committed:	Wed Jan 12 20:37:11 2005 UTC (19 years, 4 months ago) by root
Branch:	MAIN
Changes since 1.19:	+1 -1 lines
Log Message:	* empty log message *
#	User	Rev	Content
1	root	1.1	=head1 NAME
2
3			Linux::AIO - linux-specific aio implemented using clone
4
5			=head1 SYNOPSIS
6
7			use Linux::AIO;
8
9			=head1 DESCRIPTION
10
11	root	1.14	This module implements asynchronous i/o using the means available to linux
12			- clone. It does not hook into the POSIX aio_* functions because linux
13			does not yet support these in the kernel (and even if, it would only allow
14			aio_read and write, not open and stat).
15
16			Instead, in this module a number of (non-posix) threads are started that
17			execute your read/writes and signal their completion. You don't need
18			thread support in your libc or perl, and the threads created by this
19			module will not be visible to the pthreads library.
20	root	1.3
21	root	1.17	NOTICE: the threads created by this module will automatically be killed
22			when the thread calling min_parallel exits. Make sure you only ever call
23			min_parallel from the same thread that loaded this module.
24
25	root	1.16	Although the module will work with threads, it is not reentrant, so use
26			appropriate locking yourself.
27
28	root	1.1	=over 4
29
30			=cut
31
32			package Linux::AIO;
33
34	root	1.2	use base 'Exporter';
35
36	root	1.1	BEGIN {
37	root	1.19	$VERSION = 1.3;
38	root	1.1
39	root	1.17	@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink);
40	root	1.2	@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs);
41
42	root	1.1	require XSLoader;
43			XSLoader::load Linux::AIO, $VERSION;
44			}
45
46	root	1.14	=item Linux::AIO::min_parallel $nthreads
47	root	1.1
48	root	1.16	Set the minimum number of AIO threads to C<$nthreads>. The default is
49			C<1>, which means a single asynchronous operation can be done at one time
50			(the number of outstanding operations, however, is unlimited).
51	root	1.1
52	root	1.14	It is recommended to keep the number of threads low, as many linux
53			kernel versions will scale negatively with the number of threads (higher
54			parallelity => MUCH higher latency).
55
56	root	1.2	=item $fileno = Linux::AIO::poll_fileno
57
58	root	1.14	Return the I<request result pipe filehandle>. This filehandle must be
59			polled for reading by some mechanism outside this module (e.g. Event
60			or select, see below). If the pipe becomes readable you have to call
61			C<poll_cb> to check the results.
62	root	1.2
63			=item Linux::AIO::poll_cb
64
65			Process all outstanding events on the result pipe. You have to call this
66	root	1.14	regularly. Returns the number of events processed. Returns immediately
67			when no events are outstanding.
68	root	1.2
69	root	1.3	You can use Event to multiplex, e.g.:
70
71	root	1.14	Event->io (fd => Linux::AIO::poll_fileno,
72			poll => 'r', async => 1,
73			cb => \&Linux::AIO::poll_cb );
74	root	1.3
75
76	root	1.2	=item Linux::AIO::nreqs
77
78			Returns the number of requests currently outstanding.
79	root	1.4
80	root	1.14	=item aio_open $pathname, $flags, $mode, $callback
81	root	1.10
82			Asynchronously open or create a file and call the callback with the
83	root	1.14	filedescriptor (NOT a perl filehandle, sorry for that, but watch out, this
84			might change in the future).
85	root	1.10
86	root	1.14	=item aio_close $fh, $callback
87	root	1.10
88			Asynchronously close a file and call the callback with the result code.
89
90	root	1.14	=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback
91	root	1.9
92	root	1.14	=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback
93	root	1.4
94			Reads or writes C<length> bytes from the specified C<fh> and C<offset>
95			into the scalar given by C<data> and offset C<dataoffset> and calls the
96	root	1.14	callback without the actual number of bytes read (or C<undef> on error).
97	root	1.4
98	root	1.14	=item aio_stat $fh_or_path, $callback
99	root	1.4
100	root	1.14	=item aio_lstat $fh, $callback
101	root	1.5
102	root	1.14	Works like perl's C<stat> or C<lstat> in void context. The callback will
103			be called after the stat and the results will be available using C<stat _>
104			or C<-s _> etc...
105	root	1.5
106	root	1.10	Currently, the stats are always 64-bit-stats, i.e. instead of returning an
107			error when stat'ing a large file, the results will be silently truncated
108			unless perl itself is compiled with large file support.
109	root	1.17
110			=item aio_unlink $pathname, $callback
111
112	root	1.18	Asynchronously unlink a file.
113	root	1.2
114	root	1.1	=cut
115	root	1.16
116			min_parallel 1;
117	root	1.2
118			END {
119			max_parallel 0;
120			}
121	root	1.1
122			1;
123
124			=back
125
126			=head1 BUGS
127
128	root	1.14	This module has been extensively tested in a large and very busy webserver
129			for many years now.
130	root	1.1
131	root	1.14	- aio_open gives a fd, but all other functions expect a perl filehandle.
132	root	1.6
133	root	1.1	=head1 SEE ALSO
134
135			L<Coro>.
136
137			=head1 AUTHOR
138
139			Marc Lehmann <pcg@goof.com>
140	root	1.20	http://home.schmorp.de/
141	root	1.1
142			=cut
143