IO-AIO/AIO.pm

=head1 NAME

IO::AIO - Asynchronous Input/Output

=head1 SYNOPSIS

 use IO::AIO;

=head1 DESCRIPTION

This module implements asynchronous I/O using whatever means your
operating system supports.

Currently, a number of 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. In the future, this module might make use of the native
aio functions available on many operating systems. However, they are often
not well-supported (Linux doesn't allow them on normal files currently,
for example), and they would only support aio_read and aio_write, so the
remaining functionality would have to be implemented using threads anyway.

Although the module will work with in the presence of other threads, it is
currently not reentrant, so use appropriate locking yourself.

=head2 API NOTES

All the C<aio_*> calls are more or less thin wrappers around the syscall
with the same name (sans C<aio_>). The arguments are similar or identical,
and they all accept an additional C<$callback> argument which must be
a code reference. This code reference will get called with the syscall
return code (e.g. most syscalls return C<-1> on error, unlike perl, which
usually delivers "false") as it's sole argument when the given syscall has
been executed asynchronously.

All functions that expect a filehandle will also accept a file descriptor.

The filenames you pass to these routines I<must> be absolute. The reason
is that at the time the request is being executed, the current working
directory could have changed. Alternatively, you can make sure that you
never change the current working directory.

=over 4

=cut

package IO::AIO;

use base 'Exporter';

use Fcntl ();

BEGIN {
   $VERSION = 0.2;

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

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

=item IO::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 some Linux
kernel versions will scale negatively with the number of threads (higher
parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32
threads should be fine.

Under normal circumstances you don't need to call this function, as this
module automatically starts some threads (the exact number might change,
and is currently 4).

=item IO::AIO::max_parallel $nthreads

Sets the maximum number of AIO threads to C<$nthreads>. If more than
the specified number of threads are currently running, kill them. This
function blocks until the limit is reached.

This module automatically runs C<max_parallel 0> at program end, to ensure
that all threads are killed and that there are no outstanding requests.

Under normal circumstances you don't need to call this function.

=item $oldnreqs = IO::AIO::max_outstanding $nreqs

Sets the maximum number of outstanding requests to C<$nreqs>. If you
try to queue up more than this number of requests, the caller will block until
some requests have been handled.

The default is very large, so normally there is no practical limit. If you
queue up many requests in a loop it it often improves speed if you set
this to a relatively low number, such as C<100>.

Under normal circumstances you don't need to call this function.

=item $fileno = IO::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.

See C<poll_cb> for an example.

=item IO::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 => IO::AIO::poll_fileno,
              poll => 'r', async => 1,
              cb => \&IO::AIO::poll_cb);

=item IO::AIO::poll_wait

Wait till the result filehandle becomes ready for reading (simply does a
select on the filehandle. This is useful if you want to synchronously wait
for some requests to finish).

See C<nreqs> for an example.

=item IO::AIO::nreqs

Returns the number of requests currently outstanding.

Example: wait till there are no outstanding requests anymore:

   IO::AIO::poll_wait, IO::AIO::poll_cb
      while IO::AIO::nreqs;

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

Asynchronously open or create a file and call the callback with a newly
created filehandle for the file.

The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
for an explanation.

The C<$mode> argument is a bitmask. See the C<Fcntl> module for a
list. They are the same as used in C<sysopen>.

Example:

   aio_open "/etc/passwd", O_RDONLY, 0, sub {
      if ($_[0]) {
         print "open successful, fh is $_[0]\n";
         ...
      } else {
         die "open failed: $!\n";
      }
   };

=item aio_close $fh, $callback

Asynchronously close a file and call the callback with the result
code. I<WARNING:> although accepted, you should not pass in a perl
filehandle here, as perl will likely close the file descriptor itself when
the filehandle is destroyed. Normally, you can safely call perls C<close>
or just let filehandles go out of scope.

=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 -1 on error, just
like the syscall).

Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at
offset C<0> within the scalar:

   aio_read $fh, 7, 15, $buffer, 0, sub {
      $_[0] >= 0 or die "read error: $!";
      print "read <$buffer>\n";
   };

=item aio_readahead $fh,$offset,$length, $callback

Asynchronously reads the specified byte range into the page cache, using
the C<readahead> syscall. If that syscall doesn't exist the status will be
C<-1> and C<$!> is set to ENOSYS.

readahead() populates the page cache with data from a file so that
subsequent reads from that file will not block on disk I/O. The C<$offset>
argument specifies the starting point from which data is to be read and
C<$length> specifies the number of bytes to be read. I/O is performed in
whole pages, so that offset is effectively rounded down to a page boundary
and bytes are read up to the next page boundary greater than or equal to
(off-set+length). aio_readahead() does not read beyond the end of the
file. The current file offset of the file is left unchanged.

=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...

The pathname passed to C<aio_stat> must be absolute. See API NOTES, above,
for an explanation.

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.

Example: Print the length of F</etc/passwd>:

   aio_stat "/etc/passwd", sub {
      $_[0] and die "stat failed: $!";
      print "size is ", -s _, "\n";
   };

=item aio_unlink $pathname, $callback

Asynchronously unlink (delete) a file and call the callback with the
result code.

=item aio_fsync $fh, $callback

Asynchronously call fsync on the given filehandle and call the callback
with the fsync result code.

=item aio_fdatasync $fh, $callback

Asynchronously call fdatasync on the given filehandle and call the
callback with the fdatasync result code.

=cut

# support function to convert a fd into a perl filehandle
sub _fd2fh {
   return undef if $_[0] < 0;

   # try to be perl5.6-compatible
   local *AIO_FH;
   open AIO_FH, "+<&=$_[0]"
      or return undef;

   *AIO_FH
}

min_parallel 4;

END {
   max_parallel 0;
}

1;

=back

=head1 BUGS

   - could be optimized to use more semaphores instead of filehandles.

=head1 SEE ALSO

L<Coro>, L<Linux::AIO>.

=head1 AUTHOR

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

=cut

Revision:	1.4
Committed:	Sun Jul 10 20:57:00 2005 UTC (18 years, 10 months ago) by root
Branch:	MAIN
CVS Tags:	rel-0_2
Changes since 1.3:	+18 -4 lines
Log Message:	* empty log message *
#	User	Rev	Content
1	root	1.1	=head1 NAME
2
3			IO::AIO - Asynchronous Input/Output
4
5			=head1 SYNOPSIS
6
7			use IO::AIO;
8
9			=head1 DESCRIPTION
10
11			This module implements asynchronous I/O using whatever means your
12	root	1.2	operating system supports.
13	root	1.1
14	root	1.2	Currently, a number of threads are started that execute your read/writes
15			and signal their completion. You don't need thread support in your libc or
16			perl, and the threads created by this module will not be visible to the
17			pthreads library. In the future, this module might make use of the native
18			aio functions available on many operating systems. However, they are often
19			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
21			remaining functionality would have to be implemented using threads anyway.
22	root	1.1
23			Although the module will work with in the presence of other threads, it is
24	root	1.2	currently not reentrant, so use appropriate locking yourself.
25	root	1.1
26			=head2 API NOTES
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
45			=cut
46
47			package IO::AIO;
48
49			use base 'Exporter';
50
51	root	1.2	use Fcntl ();
52
53	root	1.1	BEGIN {
54	root	1.3	$VERSION = 0.2;
55	root	1.1
56			@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink
57			aio_fsync aio_fdatasync aio_readahead);
58	root	1.4	@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel max_outstanding nreqs);
59	root	1.1
60			require XSLoader;
61			XSLoader::load IO::AIO, $VERSION;
62			}
63
64			=item IO::AIO::min_parallel $nthreads
65
66			Set the minimum number of AIO threads to C<$nthreads>. The default is
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
70	root	1.4	It is recommended to keep the number of threads low, as some Linux
71	root	1.1	kernel versions will scale negatively with the number of threads (higher
72	root	1.4	parallelity => MUCH higher latency). With current Linux 2.6 versions, 4-32
73			threads should be fine.
74	root	1.1
75			Under normal circumstances you don't need to call this function, as this
76	root	1.4	module automatically starts some threads (the exact number might change,
77			and is currently 4).
78	root	1.1
79			=item IO::AIO::max_parallel $nthreads
80
81			Sets the maximum number of AIO threads to C<$nthreads>. If more than
82			the specified number of threads are currently running, kill them. This
83			function blocks until the limit is reached.
84
85			This module automatically runs C<max_parallel 0> at program end, to ensure
86			that all threads are killed and that there are no outstanding requests.
87
88			Under normal circumstances you don't need to call this function.
89
90	root	1.4	=item $oldnreqs = IO::AIO::max_outstanding $nreqs
91
92			Sets the maximum number of outstanding requests to C<$nreqs>. If you
93			try to queue up more than this number of requests, the caller will block until
94			some requests have been handled.
95
96			The default is very large, so normally there is no practical limit. If you
97			queue up many requests in a loop it it often improves speed if you set
98			this to a relatively low number, such as C<100>.
99
100			Under normal circumstances you don't need to call this function.
101
102	root	1.1	=item $fileno = IO::AIO::poll_fileno
103
104			Return the I<request result pipe filehandle>. This filehandle must be
105			polled for reading by some mechanism outside this module (e.g. Event
106			or select, see below). If the pipe becomes readable you have to call
107			C<poll_cb> to check the results.
108
109			See C<poll_cb> for an example.
110
111			=item IO::AIO::poll_cb
112
113			Process all outstanding events on the result pipe. You have to call this
114			regularly. Returns the number of events processed. Returns immediately
115			when no events are outstanding.
116
117			You can use Event to multiplex, e.g.:
118
119			Event->io (fd => IO::AIO::poll_fileno,
120			poll => 'r', async => 1,
121			cb => \&IO::AIO::poll_cb);
122
123			=item IO::AIO::poll_wait
124
125			Wait till the result filehandle becomes ready for reading (simply does a
126			select on the filehandle. This is useful if you want to synchronously wait
127			for some requests to finish).
128
129			See C<nreqs> for an example.
130
131			=item IO::AIO::nreqs
132
133			Returns the number of requests currently outstanding.
134
135			Example: wait till there are no outstanding requests anymore:
136
137			IO::AIO::poll_wait, IO::AIO::poll_cb
138			while IO::AIO::nreqs;
139
140			=item aio_open $pathname, $flags, $mode, $callback
141
142	root	1.2	Asynchronously open or create a file and call the callback with a newly
143			created filehandle for the file.
144	root	1.1
145			The pathname passed to C<aio_open> must be absolute. See API NOTES, above,
146			for an explanation.
147
148			The C<$mode> argument is a bitmask. See the C<Fcntl> module for a
149			list. They are the same as used in C<sysopen>.
150
151			Example:
152
153			aio_open "/etc/passwd", O_RDONLY, 0, sub {
154	root	1.2	if ($_[0]) {
155			print "open successful, fh is $_[0]\n";
156	root	1.1	...
157			} else {
158			die "open failed: $!\n";
159			}
160			};
161
162			=item aio_close $fh, $callback
163
164	root	1.2	Asynchronously close a file and call the callback with the result
165			code. I<WARNING:> although accepted, you should not pass in a perl
166			filehandle here, as perl will likely close the file descriptor itself when
167			the filehandle is destroyed. Normally, you can safely call perls C<close>
168			or just let filehandles go out of scope.
169	root	1.1
170			=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback
171
172			=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback
173
174			Reads or writes C<length> bytes from the specified C<fh> and C<offset>
175			into the scalar given by C<data> and offset C<dataoffset> and calls the
176			callback without the actual number of bytes read (or -1 on error, just
177			like the syscall).
178
179			Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at
180			offset C<0> within the scalar:
181
182			aio_read $fh, 7, 15, $buffer, 0, sub {
183			$_[0] >= 0 or die "read error: $!";
184			print "read <$buffer>\n";
185			};
186
187			=item aio_readahead $fh,$offset,$length, $callback
188
189			Asynchronously reads the specified byte range into the page cache, using
190			the C<readahead> syscall. If that syscall doesn't exist the status will be
191			C<-1> and C<$!> is set to ENOSYS.
192
193			readahead() populates the page cache with data from a file so that
194			subsequent reads from that file will not block on disk I/O. The C<$offset>
195			argument specifies the starting point from which data is to be read and
196			C<$length> specifies the number of bytes to be read. I/O is performed in
197			whole pages, so that offset is effectively rounded down to a page boundary
198			and bytes are read up to the next page boundary greater than or equal to
199			(off-set+length). aio_readahead() does not read beyond the end of the
200			file. The current file offset of the file is left unchanged.
201
202			=item aio_stat $fh_or_path, $callback
203
204			=item aio_lstat $fh, $callback
205
206			Works like perl's C<stat> or C<lstat> in void context. The callback will
207			be called after the stat and the results will be available using C<stat _>
208			or C<-s _> etc...
209
210			The pathname passed to C<aio_stat> must be absolute. See API NOTES, above,
211			for an explanation.
212
213			Currently, the stats are always 64-bit-stats, i.e. instead of returning an
214			error when stat'ing a large file, the results will be silently truncated
215			unless perl itself is compiled with large file support.
216
217			Example: Print the length of F</etc/passwd>:
218
219			aio_stat "/etc/passwd", sub {
220			$_[0] and die "stat failed: $!";
221			print "size is ", -s _, "\n";
222			};
223
224			=item aio_unlink $pathname, $callback
225
226			Asynchronously unlink (delete) a file and call the callback with the
227			result code.
228
229			=item aio_fsync $fh, $callback
230
231			Asynchronously call fsync on the given filehandle and call the callback
232			with the fsync result code.
233
234			=item aio_fdatasync $fh, $callback
235
236			Asynchronously call fdatasync on the given filehandle and call the
237			callback with the fdatasync result code.
238
239			=cut
240
241	root	1.2	# support function to convert a fd into a perl filehandle
242			sub _fd2fh {
243			return undef if $_[0] < 0;
244
245			# try to be perl5.6-compatible
246			local *AIO_FH;
247			open AIO_FH, "+<&=$_[0]"
248			or return undef;
249
250			*AIO_FH
251			}
252
253	root	1.1	min_parallel 4;
254
255			END {
256			max_parallel 0;
257			}
258
259			1;
260
261			=back
262
263			=head1 BUGS
264
265	root	1.2	- could be optimized to use more semaphores instead of filehandles.
266	root	1.1
267			=head1 SEE ALSO
268
269			L<Coro>, L<Linux::AIO>.
270
271			=head1 AUTHOR
272
273			Marc Lehmann <schmorp@schmorp.de>
274			http://home.schmorp.de/
275
276			=cut
277