| 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 |
=head2 API NOTES |
| 30 |
|
| 31 |
All the C<aio_*> calls are more or less thin wrappers around the syscall |
| 32 |
with the same name (sans C<aio_>). The arguments are similar or identical, |
| 33 |
and they all accept an additional C<$callback> argument which must be |
| 34 |
a code reference. This code reference will get called with the syscall |
| 35 |
return code (e.g. most syscalls return C<-1> on error, unlike perl, which |
| 36 |
usually delivers "false") as it's sole argument when the given syscall has |
| 37 |
been executed asynchronously. |
| 38 |
|
| 39 |
All functions that expect a filehandle will also accept a file descriptor. |
| 40 |
|
| 41 |
The filenames you pass to these routines I<must> be absolute. The reason |
| 42 |
is that at the time the request is being executed, the current working |
| 43 |
directory could have changed. Alternatively, you can make sure that you |
| 44 |
never change the current workign directory. |
| 45 |
|
| 46 |
=over 4 |
| 47 |
|
| 48 |
=cut |
| 49 |
|
| 50 |
package Linux::AIO; |
| 51 |
|
| 52 |
use base 'Exporter'; |
| 53 |
|
| 54 |
BEGIN { |
| 55 |
$VERSION = 1.71; |
| 56 |
|
| 57 |
@EXPORT = qw(aio_read aio_write aio_open aio_close aio_stat aio_lstat aio_unlink |
| 58 |
aio_fsync aio_fdatasync aio_readahead); |
| 59 |
@EXPORT_OK = qw(poll_fileno poll_cb min_parallel max_parallel nreqs); |
| 60 |
|
| 61 |
require XSLoader; |
| 62 |
XSLoader::load Linux::AIO, $VERSION; |
| 63 |
} |
| 64 |
|
| 65 |
=item Linux::AIO::min_parallel $nthreads |
| 66 |
|
| 67 |
Set the minimum number of AIO threads to C<$nthreads>. The default is |
| 68 |
C<1>, which means a single asynchronous operation can be done at one time |
| 69 |
(the number of outstanding operations, however, is unlimited). |
| 70 |
|
| 71 |
It is recommended to keep the number of threads low, as some linux |
| 72 |
kernel versions will scale negatively with the number of threads (higher |
| 73 |
parallelity => MUCH higher latency). |
| 74 |
|
| 75 |
Under normal circumstances you don't need to call this function, as this |
| 76 |
module automatically starts a single async thread. |
| 77 |
|
| 78 |
=item Linux::AIO::max_parallel $nthreads |
| 79 |
|
| 80 |
Sets the maximum number of AIO threads to C<$nthreads>. If more than |
| 81 |
the specified number of threads are currently running, kill them. This |
| 82 |
function blocks until the limit is reached. |
| 83 |
|
| 84 |
This module automatically runs C<max_parallel 0> at program end, to ensure |
| 85 |
that all threads are killed and that there are no outstanding requests. |
| 86 |
|
| 87 |
Under normal circumstances you don't need to call this function. |
| 88 |
|
| 89 |
=item $fileno = Linux::AIO::poll_fileno |
| 90 |
|
| 91 |
Return the I<request result pipe filehandle>. This filehandle must be |
| 92 |
polled for reading by some mechanism outside this module (e.g. Event |
| 93 |
or select, see below). If the pipe becomes readable you have to call |
| 94 |
C<poll_cb> to check the results. |
| 95 |
|
| 96 |
See C<poll_cb> for an example. |
| 97 |
|
| 98 |
=item Linux::AIO::poll_cb |
| 99 |
|
| 100 |
Process all outstanding events on the result pipe. You have to call this |
| 101 |
regularly. Returns the number of events processed. Returns immediately |
| 102 |
when no events are outstanding. |
| 103 |
|
| 104 |
You can use Event to multiplex, e.g.: |
| 105 |
|
| 106 |
Event->io (fd => Linux::AIO::poll_fileno, |
| 107 |
poll => 'r', async => 1, |
| 108 |
cb => \&Linux::AIO::poll_cb); |
| 109 |
|
| 110 |
=item Linux::AIO::poll_wait |
| 111 |
|
| 112 |
Wait till the result filehandle becomes ready for reading (simply does a |
| 113 |
select on the filehandle. This is useful if you want to synchronously wait |
| 114 |
for some requests to finish). |
| 115 |
|
| 116 |
See C<nreqs> for an example. |
| 117 |
|
| 118 |
=item Linux::AIO::nreqs |
| 119 |
|
| 120 |
Returns the number of requests currently outstanding. |
| 121 |
|
| 122 |
Example: wait till there are no outstanding requests anymore: |
| 123 |
|
| 124 |
Linux::AIO::poll_wait, Linux::AIO::poll_cb |
| 125 |
while Linux::AIO::nreqs; |
| 126 |
|
| 127 |
=item aio_open $pathname, $flags, $mode, $callback |
| 128 |
|
| 129 |
Asynchronously open or create a file and call the callback with the |
| 130 |
filedescriptor (NOT a perl filehandle, sorry for that, but watch out, this |
| 131 |
might change in the future). |
| 132 |
|
| 133 |
The C<$mode> argument is a bitmask. See the C<Fcntl> module for a |
| 134 |
list. They are the same as used in C<sysopen>. |
| 135 |
|
| 136 |
Example: |
| 137 |
|
| 138 |
aio_open "/etc/passwd", O_RDONLY, 0, sub { |
| 139 |
if ($_[0] >= 0) { |
| 140 |
open my $fh, "<&$_[0]"; # create a copy for perl |
| 141 |
aio_close $_[0], sub { }; # close the aio handle |
| 142 |
print "open successful, fh is $fh\n"; |
| 143 |
... |
| 144 |
} else { |
| 145 |
die "open failed: $!\n"; |
| 146 |
} |
| 147 |
}; |
| 148 |
|
| 149 |
=item aio_close $fh, $callback |
| 150 |
|
| 151 |
Asynchronously close a file and call the callback with the result code. |
| 152 |
|
| 153 |
=item aio_read $fh,$offset,$length, $data,$dataoffset,$callback |
| 154 |
|
| 155 |
=item aio_write $fh,$offset,$length, $data,$dataoffset,$callback |
| 156 |
|
| 157 |
Reads or writes C<length> bytes from the specified C<fh> and C<offset> |
| 158 |
into the scalar given by C<data> and offset C<dataoffset> and calls the |
| 159 |
callback without the actual number of bytes read (or -1 on error, just |
| 160 |
like the syscall). |
| 161 |
|
| 162 |
Example: Read 15 bytes at offset 7 into scalar C<$buffer>, strating at |
| 163 |
offset C<0> within the scalar: |
| 164 |
|
| 165 |
aio_read $fh, 7, 15, $buffer, 0, sub { |
| 166 |
$_[0] >= 0 or die "read error: $!"; |
| 167 |
print "read <$buffer>\n"; |
| 168 |
}; |
| 169 |
|
| 170 |
=item aio_readahead $fh,$offset,$length, $callback |
| 171 |
|
| 172 |
Asynchronously reads the specified byte range into the page cache, using |
| 173 |
the C<readahead> syscall. |
| 174 |
|
| 175 |
readahead() populates the page cache with data from a file so that |
| 176 |
subsequent reads from that file will not block on disk I/O. The C<$offset> |
| 177 |
argument specifies the starting point from which data is to be read and |
| 178 |
C<$length> specifies the number of bytes to be read. I/O is performed in |
| 179 |
whole pages, so that offset is effectively rounded down to a page boundary |
| 180 |
and bytes are read up to the next page boundary greater than or equal to |
| 181 |
(off-set+length). aio_readahead() does not read beyond the end of the |
| 182 |
file. The current file offset of the file is left unchanged. |
| 183 |
|
| 184 |
=item aio_stat $fh_or_path, $callback |
| 185 |
|
| 186 |
=item aio_lstat $fh, $callback |
| 187 |
|
| 188 |
Works like perl's C<stat> or C<lstat> in void context. The callback will |
| 189 |
be called after the stat and the results will be available using C<stat _> |
| 190 |
or C<-s _> etc... |
| 191 |
|
| 192 |
The pathname passed to C<aio_stat> must be absolute. See API NOTES, above, |
| 193 |
for an explanation. |
| 194 |
|
| 195 |
Currently, the stats are always 64-bit-stats, i.e. instead of returning an |
| 196 |
error when stat'ing a large file, the results will be silently truncated |
| 197 |
unless perl itself is compiled with large file support. |
| 198 |
|
| 199 |
Example: Print the length of F</etc/passwd>: |
| 200 |
|
| 201 |
aio_stat "/etc/passwd", sub { |
| 202 |
$_[0] and die "stat failed: $!"; |
| 203 |
print "size is ", -s _, "\n"; |
| 204 |
}; |
| 205 |
|
| 206 |
=item aio_unlink $pathname, $callback |
| 207 |
|
| 208 |
Asynchronously unlink (delete) a file and call the callback with the |
| 209 |
result code. |
| 210 |
|
| 211 |
=item aio_fsync $fh, $callback |
| 212 |
|
| 213 |
Asynchronously call fsync on the given filehandle and call the callback |
| 214 |
with the fsync result code. |
| 215 |
|
| 216 |
=item aio_fdatasync $fh, $callback |
| 217 |
|
| 218 |
Asynchronously call fdatasync on the given filehandle and call the |
| 219 |
callback with the fdatasync result code. |
| 220 |
|
| 221 |
=cut |
| 222 |
|
| 223 |
min_parallel 1; |
| 224 |
|
| 225 |
END { |
| 226 |
max_parallel 0; |
| 227 |
} |
| 228 |
|
| 229 |
1; |
| 230 |
|
| 231 |
=back |
| 232 |
|
| 233 |
=head1 BUGS |
| 234 |
|
| 235 |
This module has been extensively tested in a large and very busy webserver |
| 236 |
for many years now. |
| 237 |
|
| 238 |
- aio_open gives a fd, but all other functions expect a perl filehandle. |
| 239 |
|
| 240 |
=head1 SEE ALSO |
| 241 |
|
| 242 |
L<Coro>. |
| 243 |
|
| 244 |
=head1 AUTHOR |
| 245 |
|
| 246 |
Marc Lehmann <schmorp@schmorp.de> |
| 247 |
http://home.schmorp.de/ |
| 248 |
|
| 249 |
=cut |
| 250 |
|
