| 1 |
root |
1.1 |
NAME |
| 2 |
|
|
Linux::AIO - linux-specific aio implemented using clone |
| 3 |
|
|
|
| 4 |
|
|
SYNOPSIS |
| 5 |
|
|
use Linux::AIO; |
| 6 |
|
|
|
| 7 |
|
|
DESCRIPTION |
| 8 |
|
|
This module implements asynchronous i/o using the means available to |
| 9 |
|
|
linux - clone. It does not hook into the POSIX aio_* functions because |
| 10 |
|
|
linux does not yet support these in the kernel (and even if, it would |
| 11 |
|
|
only allow aio_read and write, not open and stat). |
| 12 |
|
|
|
| 13 |
|
|
Instead, in this module a number of (non-posix) threads are started that |
| 14 |
|
|
execute your read/writes and signal their completion. You don't need |
| 15 |
|
|
thread support in your libc or perl, and the threads created by this |
| 16 |
|
|
module will not be visible to the pthreads library. |
| 17 |
|
|
|
| 18 |
|
|
NOTICE: the threads created by this module will automatically be killed |
| 19 |
|
|
when the thread calling min_parallel exits. Make sure you only ever call |
| 20 |
|
|
min_parallel from the same thread that loaded this module. |
| 21 |
|
|
|
| 22 |
|
|
Although the module will work with threads, it is not reentrant, so use |
| 23 |
|
|
appropriate locking yourself. |
| 24 |
|
|
|
| 25 |
|
|
Linux::AIO::min_parallel $nthreads |
| 26 |
|
|
Set the minimum number of AIO threads to $nthreads. The default is |
| 27 |
|
|
1, which means a single asynchronous operation can be done at one |
| 28 |
|
|
time (the number of outstanding operations, however, is unlimited). |
| 29 |
|
|
|
| 30 |
root |
1.2 |
It is recommended to keep the number of threads low, as some linux |
| 31 |
root |
1.1 |
kernel versions will scale negatively with the number of threads |
| 32 |
|
|
(higher parallelity => MUCH higher latency). |
| 33 |
|
|
|
| 34 |
root |
1.2 |
Linux::AIO::max_parallel $nthreads |
| 35 |
|
|
Sets the maximum number of AIO threads to $nthreads. If more than |
| 36 |
|
|
the specified number of threads are currently running, kill them. |
| 37 |
|
|
This function blocks until the limit is reached. |
| 38 |
|
|
|
| 39 |
|
|
This module automatically runs "max_parallel 0" at program end, to |
| 40 |
|
|
ensure that all threads are killed and that there are no outstanding |
| 41 |
|
|
requests. |
| 42 |
|
|
|
| 43 |
root |
1.1 |
$fileno = Linux::AIO::poll_fileno |
| 44 |
|
|
Return the *request result pipe filehandle*. This filehandle must be |
| 45 |
|
|
polled for reading by some mechanism outside this module (e.g. Event |
| 46 |
|
|
or select, see below). If the pipe becomes readable you have to call |
| 47 |
|
|
"poll_cb" to check the results. |
| 48 |
|
|
|
| 49 |
|
|
Linux::AIO::poll_cb |
| 50 |
|
|
Process all outstanding events on the result pipe. You have to call |
| 51 |
|
|
this regularly. Returns the number of events processed. Returns |
| 52 |
|
|
immediately when no events are outstanding. |
| 53 |
|
|
|
| 54 |
|
|
You can use Event to multiplex, e.g.: |
| 55 |
|
|
|
| 56 |
|
|
Event->io (fd => Linux::AIO::poll_fileno, |
| 57 |
|
|
poll => 'r', async => 1, |
| 58 |
|
|
cb => \&Linux::AIO::poll_cb ); |
| 59 |
|
|
|
| 60 |
root |
1.2 |
Linux::AIO::poll_wait |
| 61 |
|
|
Wait till the result filehandle becomes ready for reading (simply |
| 62 |
|
|
does a select on the filehandle. This is useful if you want to |
| 63 |
|
|
synchronously wait for some requests to finish). |
| 64 |
|
|
|
| 65 |
root |
1.1 |
Linux::AIO::nreqs |
| 66 |
|
|
Returns the number of requests currently outstanding. |
| 67 |
|
|
|
| 68 |
|
|
aio_open $pathname, $flags, $mode, $callback |
| 69 |
|
|
Asynchronously open or create a file and call the callback with the |
| 70 |
|
|
filedescriptor (NOT a perl filehandle, sorry for that, but watch |
| 71 |
|
|
out, this might change in the future). |
| 72 |
|
|
|
| 73 |
|
|
aio_close $fh, $callback |
| 74 |
|
|
Asynchronously close a file and call the callback with the result |
| 75 |
|
|
code. |
| 76 |
|
|
|
| 77 |
|
|
aio_read $fh,$offset,$length, $data,$dataoffset,$callback |
| 78 |
|
|
aio_write $fh,$offset,$length, $data,$dataoffset,$callback |
| 79 |
|
|
Reads or writes "length" bytes from the specified "fh" and "offset" |
| 80 |
|
|
into the scalar given by "data" and offset "dataoffset" and calls |
| 81 |
|
|
the callback without the actual number of bytes read (or "undef" on |
| 82 |
|
|
error). |
| 83 |
|
|
|
| 84 |
|
|
aio_stat $fh_or_path, $callback |
| 85 |
|
|
aio_lstat $fh, $callback |
| 86 |
|
|
Works like perl's "stat" or "lstat" in void context. The callback |
| 87 |
|
|
will be called after the stat and the results will be available |
| 88 |
|
|
using "stat _" or "-s _" etc... |
| 89 |
|
|
|
| 90 |
|
|
Currently, the stats are always 64-bit-stats, i.e. instead of |
| 91 |
|
|
returning an error when stat'ing a large file, the results will be |
| 92 |
|
|
silently truncated unless perl itself is compiled with large file |
| 93 |
|
|
support. |
| 94 |
|
|
|
| 95 |
|
|
aio_unlink $pathname, $callback |
| 96 |
|
|
Asynchronously unlink a file. |
| 97 |
|
|
|
| 98 |
|
|
BUGS |
| 99 |
|
|
This module has been extensively tested in a large and very busy |
| 100 |
|
|
webserver for many years now. |
| 101 |
|
|
|
| 102 |
|
|
- aio_open gives a fd, but all other functions expect a perl filehandle. |
| 103 |
|
|
|
| 104 |
|
|
SEE ALSO |
| 105 |
|
|
Coro. |
| 106 |
|
|
|
| 107 |
|
|
AUTHOR |
| 108 |
|
|
Marc Lehmann <schmorp@schmorp.de> |
| 109 |
|
|
http://home.schmorp.de/ |
| 110 |
|
|
|
