| 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 |
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 (even as of 2.6.12, only O_DIRECT |
| 14 |
files are supported) and even if, it would only allow aio_read and write, |
| 15 |
not open, stat and so on. |
| 16 |
|
| 17 |
Instead, in this module a number of (non-posix) threads are started that |
| 18 |
execute your read/writes and signal their completion. You don't need |
| 19 |
thread support in your libc or perl, and the threads created by this |
| 20 |
module will not be visible to the pthreads library. |
| 21 |
|
| 22 |
NOTICE: the threads created by this module will automatically be killed |
| 23 |
when the thread calling min_parallel exits. Make sure you only ever call |
| 24 |
min_parallel from the same thread that loaded this module. |
| 25 |
|
| 26 |
Although the module will work with in the presence of other threads, it is |
| 27 |
not reentrant, so use appropriate locking yourself. |
| 28 |
|
| 29 |
=over 4 |
| 30 |
|
| 31 |
=cut |
| 32 |
|
| 33 |
package Linux::AIO; |
| 34 |
|
| 35 |
use base 'Exporter'; |
| 36 |
|
| 37 |
BEGIN { |
| 38 |
$VERSION = 1.6; |
| 39 |
|
| 40 |
@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink); |
| 41 |
@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs); |
| 42 |
|
| 43 |
require XSLoader; |
| 44 |
XSLoader::load Linux::AIO, $VERSION; |
| 45 |
} |
| 46 |
|
| 47 |
=item Linux::AIO::min_parallel $nthreads |
| 48 |
|
| 49 |
Set the minimum number of AIO threads to C<$nthreads>. The default is |
| 50 |
C<1>, which means a single asynchronous operation can be done at one time |
| 51 |
(the number of outstanding operations, however, is unlimited). |
| 52 |
|
| 53 |
It is recommended to keep the number of threads low, as some linux |
| 54 |
kernel versions will scale negatively with the number of threads (higher |
| 55 |
parallelity => MUCH higher latency). |
| 56 |
|
| 57 |
=item Linux::AIO::max_parallel $nthreads |
| 58 |
|
| 59 |
Sets the maximum number of AIO threads to C<$nthreads>. If more than |
| 60 |
the specified number of threads are currently running, kill them. This |
| 61 |
function blocks until the limit is reached. |
| 62 |
|
| 63 |
This module automatically runs C<max_parallel 0> at program end, to ensure |
| 64 |
that all threads are killed and that there are no outstanding requests. |
| 65 |
|
| 66 |
=item $fileno = Linux::AIO::poll_fileno |
| 67 |
|
| 68 |
Return the I<request result pipe filehandle>. This filehandle must be |
| 69 |
polled for reading by some mechanism outside this module (e.g. Event |
| 70 |
or select, see below). If the pipe becomes readable you have to call |
| 71 |
C<poll_cb> to check the results. |
| 72 |
|
| 73 |
=item Linux::AIO::poll_cb |
| 74 |
|
| 75 |
Process all outstanding events on the result pipe. You have to call this |
| 76 |
regularly. Returns the number of events processed. Returns immediately |
| 77 |
when no events are outstanding. |
| 78 |
|
| 79 |
You can use Event to multiplex, e.g.: |
| 80 |
|
| 81 |
Event->io (fd => Linux::AIO::poll_fileno, |
| 82 |
poll => 'r', async => 1, |
| 83 |
cb => \&Linux::AIO::poll_cb ); |
| 84 |
|
| 85 |
=item Linux::AIO::poll_wait |
| 86 |
|
| 87 |
Wait till the result filehandle becomes ready for reading (simply does a |
| 88 |
select on the filehandle. This is useful if you want to synchronously wait |
| 89 |
for some requests to finish). |
| 90 |
|
| 91 |
=item Linux::AIO::nreqs |
| 92 |
|
| 93 |
Returns the number of requests currently outstanding. |
| 94 |
|
| 95 |
=item aio_open $pathname, $flags, $mode, $callback |
| 96 |
|
| 97 |
Asynchronously open or create a file and call the callback with the |
| 98 |
filedescriptor (NOT a perl filehandle, sorry for that, but watch out, this |
| 99 |
might change in the future). |
| 100 |
|
| 101 |
=item aio_close $fh, $callback |
| 102 |
|
| 103 |
Asynchronously close a file and call the callback with the result code. |
| 104 |
|
| 105 |
=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback |
| 106 |
|
| 107 |
=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback |
| 108 |
|
| 109 |
Reads or writes C<length> bytes from the specified C<fh> and C<offset> |
| 110 |
into the scalar given by C<data> and offset C<dataoffset> and calls the |
| 111 |
callback without the actual number of bytes read (or C<undef> on error). |
| 112 |
|
| 113 |
=item aio_stat $fh_or_path, $callback |
| 114 |
|
| 115 |
=item aio_lstat $fh, $callback |
| 116 |
|
| 117 |
Works like perl's C<stat> or C<lstat> in void context. The callback will |
| 118 |
be called after the stat and the results will be available using C<stat _> |
| 119 |
or C<-s _> etc... |
| 120 |
|
| 121 |
Currently, the stats are always 64-bit-stats, i.e. instead of returning an |
| 122 |
error when stat'ing a large file, the results will be silently truncated |
| 123 |
unless perl itself is compiled with large file support. |
| 124 |
|
| 125 |
=item aio_unlink $pathname, $callback |
| 126 |
|
| 127 |
Asynchronously unlink a file. |
| 128 |
|
| 129 |
=cut |
| 130 |
|
| 131 |
min_parallel 1; |
| 132 |
|
| 133 |
END { |
| 134 |
max_parallel 0; |
| 135 |
} |
| 136 |
|
| 137 |
1; |
| 138 |
|
| 139 |
=back |
| 140 |
|
| 141 |
=head1 BUGS |
| 142 |
|
| 143 |
This module has been extensively tested in a large and very busy webserver |
| 144 |
for many years now. |
| 145 |
|
| 146 |
- aio_open gives a fd, but all other functions expect a perl filehandle. |
| 147 |
|
| 148 |
=head1 SEE ALSO |
| 149 |
|
| 150 |
L<Coro>. |
| 151 |
|
| 152 |
=head1 AUTHOR |
| 153 |
|
| 154 |
Marc Lehmann <schmorp@schmorp.de> |
| 155 |
http://home.schmorp.de/ |
| 156 |
|
| 157 |
=cut |
| 158 |
|
