… | |
… | |
168 | use common::sense; |
168 | use common::sense; |
169 | |
169 | |
170 | use base 'Exporter'; |
170 | use base 'Exporter'; |
171 | |
171 | |
172 | BEGIN { |
172 | BEGIN { |
173 | our $VERSION = '3.9'; |
173 | our $VERSION = '3.93'; |
174 | |
174 | |
175 | our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close |
175 | our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close |
176 | aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx |
176 | aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx |
177 | aio_scandir aio_symlink aio_readlink aio_sync aio_fsync |
177 | aio_scandir aio_symlink aio_readlink aio_realpath aio_sync aio_fsync |
178 | aio_fdatasync aio_sync_file_range aio_pathsync aio_readahead |
178 | aio_fdatasync aio_sync_file_range aio_fallocate |
|
|
179 | aio_pathsync aio_readahead |
179 | aio_rename aio_link aio_move aio_copy aio_group |
180 | aio_rename aio_link aio_move aio_copy aio_group |
180 | aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown |
181 | aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown |
181 | aio_chmod aio_utime aio_truncate |
182 | aio_chmod aio_utime aio_truncate |
182 | aio_msync aio_mtouch aio_mlock aio_mlockall |
183 | aio_msync aio_mtouch aio_mlock aio_mlockall |
183 | aio_statvfs); |
184 | aio_statvfs); |
… | |
… | |
222 | aio_unlink $pathname, $callback->($status) |
223 | aio_unlink $pathname, $callback->($status) |
223 | aio_mknod $path, $mode, $dev, $callback->($status) |
224 | aio_mknod $path, $mode, $dev, $callback->($status) |
224 | aio_link $srcpath, $dstpath, $callback->($status) |
225 | aio_link $srcpath, $dstpath, $callback->($status) |
225 | aio_symlink $srcpath, $dstpath, $callback->($status) |
226 | aio_symlink $srcpath, $dstpath, $callback->($status) |
226 | aio_readlink $path, $callback->($link) |
227 | aio_readlink $path, $callback->($link) |
|
|
228 | aio_realpath $path, $callback->($link) |
227 | aio_rename $srcpath, $dstpath, $callback->($status) |
229 | aio_rename $srcpath, $dstpath, $callback->($status) |
228 | aio_mkdir $pathname, $mode, $callback->($status) |
230 | aio_mkdir $pathname, $mode, $callback->($status) |
229 | aio_rmdir $pathname, $callback->($status) |
231 | aio_rmdir $pathname, $callback->($status) |
230 | aio_readdir $pathname, $callback->($entries) |
232 | aio_readdir $pathname, $callback->($entries) |
231 | aio_readdirx $pathname, $flags, $callback->($entries, $flags) |
233 | aio_readdirx $pathname, $flags, $callback->($entries, $flags) |
… | |
… | |
436 | |
438 | |
437 | Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts |
439 | Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts |
438 | reading at byte offset C<$in_offset>, and starts writing at the current |
440 | reading at byte offset C<$in_offset>, and starts writing at the current |
439 | file offset of C<$out_fh>. Because of that, it is not safe to issue more |
441 | file offset of C<$out_fh>. Because of that, it is not safe to issue more |
440 | than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each |
442 | than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each |
441 | other. |
443 | other. The same C<$in_fh> works fine though, as this function does not |
|
|
444 | move or use the file offset of C<$in_fh>. |
442 | |
445 | |
443 | Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than |
446 | Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than |
444 | are written, and there is no way to find out how many bytes have been read |
447 | are written, and there is no way to find out how many more bytes have been |
445 | from C<aio_sendfile> alone, as C<aio_sendfile> only provides the number of |
448 | read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the |
446 | bytes written to C<$out_fh>. Only if the result value equals C<$length> |
449 | number of bytes written to C<$out_fh>. Only if the result value equals |
447 | one can assume that C<$length> bytes have been read. |
450 | C<$length> one can assume that C<$length> bytes have been read. |
448 | |
451 | |
449 | Unlike with other C<aio_> functions, it makes a lot of sense to use |
452 | Unlike with other C<aio_> functions, it makes a lot of sense to use |
450 | C<aio_sendfile> on non-blocking sockets, as long as one end (typically |
453 | C<aio_sendfile> on non-blocking sockets, as long as one end (typically |
451 | the C<$in_fh>) is a file - the file I/O will then be asynchronous, while |
454 | the C<$in_fh>) is a file - the file I/O will then be asynchronous, while |
452 | the socket I/O will be non-blocking. Note, however, that you can run into |
455 | the socket I/O will be non-blocking. Note, however, that you can run |
453 | a trap where C<aio_sendfile> reads some data with readahead, then fails |
456 | into a trap where C<aio_sendfile> reads some data with readahead, then |
454 | to write all data, and when the socket is ready the next time, the data |
457 | fails to write all data, and when the socket is ready the next time, the |
455 | in the cache is already lost, forcing C<aio_sendfile> to again hit the |
458 | data in the cache is already lost, forcing C<aio_sendfile> to again hit |
456 | disk. Explicit C<aio_read> + C<aio_write> let's you control resource usage |
459 | the disk. Explicit C<aio_read> + C<aio_write> let's you better control |
457 | much better. |
460 | resource usage. |
458 | |
461 | |
459 | This call tries to make use of a native C<sendfile> syscall to provide |
462 | This call tries to make use of a native C<sendfile>-like syscall to |
460 | zero-copy operation. For this to work, C<$out_fh> should refer to a |
463 | provide zero-copy operation. For this to work, C<$out_fh> should refer to |
461 | socket, and C<$in_fh> should refer to an mmap'able file. |
464 | a socket, and C<$in_fh> should refer to an mmap'able file. |
462 | |
465 | |
463 | If a native sendfile cannot be found or it fails with C<ENOSYS>, |
466 | If a native sendfile cannot be found or it fails with C<ENOSYS>, |
464 | C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or C<ENOTSOCK>, |
467 | C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or |
465 | it will be emulated, so you can call C<aio_sendfile> on any type of |
468 | C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any |
466 | filehandle regardless of the limitations of the operating system. |
469 | type of filehandle regardless of the limitations of the operating system. |
|
|
470 | |
|
|
471 | As native sendfile syscalls (as practically any non-POSIX interface hacked |
|
|
472 | together in a hurry to improve benchmark numbers) tend to be rather buggy |
|
|
473 | on many systems, this implementation tries to work around some known bugs |
|
|
474 | in Linux and FreeBSD kernels (probably others, too), but that might fail, |
|
|
475 | so you really really should check the return value of C<aio_sendfile> - |
|
|
476 | fewre bytes than expected might have been transferred. |
467 | |
477 | |
468 | |
478 | |
469 | =item aio_readahead $fh,$offset,$length, $callback->($retval) |
479 | =item aio_readahead $fh,$offset,$length, $callback->($retval) |
470 | |
480 | |
471 | C<aio_readahead> populates the page cache with data from a file so that |
481 | C<aio_readahead> populates the page cache with data from a file so that |
… | |
… | |
635 | Asynchronously read the symlink specified by C<$path> and pass it to |
645 | Asynchronously read the symlink specified by C<$path> and pass it to |
636 | the callback. If an error occurs, nothing or undef gets passed to the |
646 | the callback. If an error occurs, nothing or undef gets passed to the |
637 | callback. |
647 | callback. |
638 | |
648 | |
639 | |
649 | |
|
|
650 | =item aio_realpath $path, $callback->($path) |
|
|
651 | |
|
|
652 | Asynchronously make the path absolute and resolve any symlinks in |
|
|
653 | C<$path>. The resulting path only consists of directories (Same as |
|
|
654 | L<Cwd::realpath>). |
|
|
655 | |
|
|
656 | This request can be used to get the absolute path of the current working |
|
|
657 | directory by passing it a path of F<.> (a single dot). |
|
|
658 | |
|
|
659 | |
640 | =item aio_rename $srcpath, $dstpath, $callback->($status) |
660 | =item aio_rename $srcpath, $dstpath, $callback->($status) |
641 | |
661 | |
642 | Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as |
662 | Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as |
643 | rename(2) and call the callback with the result code. |
663 | rename(2) and call the callback with the result code. |
644 | |
664 | |
… | |
… | |
862 | if ($_[0] && $! == EXDEV) { |
882 | if ($_[0] && $! == EXDEV) { |
863 | aioreq_pri $pri; |
883 | aioreq_pri $pri; |
864 | add $grp aio_copy $src, $dst, sub { |
884 | add $grp aio_copy $src, $dst, sub { |
865 | $grp->result ($_[0]); |
885 | $grp->result ($_[0]); |
866 | |
886 | |
867 | if (!$_[0]) { |
887 | unless ($_[0]) { |
868 | aioreq_pri $pri; |
888 | aioreq_pri $pri; |
869 | add $grp aio_unlink $src; |
889 | add $grp aio_unlink $src; |
870 | } |
890 | } |
871 | }; |
891 | }; |
872 | } else { |
892 | } else { |
… | |
… | |
1564 | Sets the minimum idle timeout (default 10) after which worker threads are |
1584 | Sets the minimum idle timeout (default 10) after which worker threads are |
1565 | allowed to exit. SEe C<IO::AIO::max_idle>. |
1585 | allowed to exit. SEe C<IO::AIO::max_idle>. |
1566 | |
1586 | |
1567 | =item IO::AIO::max_outstanding $maxreqs |
1587 | =item IO::AIO::max_outstanding $maxreqs |
1568 | |
1588 | |
|
|
1589 | Sets the maximum number of outstanding requests to C<$nreqs>. If |
|
|
1590 | you do queue up more than this number of requests, the next call to |
|
|
1591 | C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as |
|
|
1592 | C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no |
|
|
1593 | longer exceeded. |
|
|
1594 | |
|
|
1595 | In other words, this setting does not enforce a queue limit, but can be |
|
|
1596 | used to make poll functions block if the limit is exceeded. |
|
|
1597 | |
1569 | This is a very bad function to use in interactive programs because it |
1598 | This is a very bad function to use in interactive programs because it |
1570 | blocks, and a bad way to reduce concurrency because it is inexact: Better |
1599 | blocks, and a bad way to reduce concurrency because it is inexact: Better |
1571 | use an C<aio_group> together with a feed callback. |
1600 | use an C<aio_group> together with a feed callback. |
1572 | |
1601 | |
1573 | Sets the maximum number of outstanding requests to C<$nreqs>. If you |
1602 | It's main use is in scripts without an event loop - when you want to stat |
1574 | do queue up more than this number of requests, the next call to the |
1603 | a lot of files, you can write somehting like this: |
1575 | C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>) |
|
|
1576 | function will block until the limit is no longer exceeded. |
|
|
1577 | |
1604 | |
1578 | The default value is very large, so there is no practical limit on the |
1605 | IO::AIO::max_outstanding 32; |
1579 | number of outstanding requests. |
|
|
1580 | |
1606 | |
1581 | You can still queue as many requests as you want. Therefore, |
1607 | for my $path (...) { |
1582 | C<max_outstanding> is mainly useful in simple scripts (with low values) or |
1608 | aio_stat $path , ...; |
1583 | as a stop gap to shield against fatal memory overflow (with large values). |
1609 | IO::AIO::poll_cb; |
|
|
1610 | } |
|
|
1611 | |
|
|
1612 | IO::AIO::flush; |
|
|
1613 | |
|
|
1614 | The call to C<poll_cb> inside the loop will normally return instantly, but |
|
|
1615 | as soon as more thna C<32> reqeusts are in-flight, it will block until |
|
|
1616 | some requests have been handled. This keeps the loop from pushing a large |
|
|
1617 | number of C<aio_stat> requests onto the queue. |
|
|
1618 | |
|
|
1619 | The default value for C<max_outstanding> is very large, so there is no |
|
|
1620 | practical limit on the number of outstanding requests. |
1584 | |
1621 | |
1585 | =back |
1622 | =back |
1586 | |
1623 | |
1587 | =head3 STATISTICAL INFORMATION |
1624 | =head3 STATISTICAL INFORMATION |
1588 | |
1625 | |
… | |
… | |
1764 | Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => |
1801 | Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => |
1765 | \&IO::AIO::poll_cb); |
1802 | \&IO::AIO::poll_cb); |
1766 | |
1803 | |
1767 | =head2 FORK BEHAVIOUR |
1804 | =head2 FORK BEHAVIOUR |
1768 | |
1805 | |
1769 | This module should do "the right thing" when the process using it forks: |
1806 | Usage of pthreads in a program changes the semantics of fork |
|
|
1807 | considerably. Specifically, only async-safe functions can be called after |
|
|
1808 | fork. Perl doesn't know about this, so in general, you cannot call fork |
|
|
1809 | with defined behaviour in perl if pthreads are involved. IO::AIO uses |
|
|
1810 | pthreads, so this applies, but many other extensions and (for inexplicable |
|
|
1811 | reasons) perl itself often is linked against pthreads, so this limitation |
|
|
1812 | applies to quite a lot of perls. |
1770 | |
1813 | |
1771 | Before the fork, IO::AIO enters a quiescent state where no requests |
1814 | This module no longer tries to fight your OS, or POSIX. That means IO::AIO |
1772 | can be added in other threads and no results will be processed. After |
1815 | only works in the process that loaded it. Forking is fully supported, but |
1773 | the fork the parent simply leaves the quiescent state and continues |
1816 | using IO::AIO in the child is not. |
1774 | request/result processing, while the child frees the request/result queue |
|
|
1775 | (so that the requests started before the fork will only be handled in the |
|
|
1776 | parent). Threads will be started on demand until the limit set in the |
|
|
1777 | parent process has been reached again. |
|
|
1778 | |
1817 | |
1779 | In short: the parent will, after a short pause, continue as if fork had |
1818 | You might get around by not I<using> IO::AIO before (or after) |
1780 | not been called, while the child will act as if IO::AIO has not been used |
1819 | forking. You could also try to call the L<IO::AIO::reinit> function in the |
1781 | yet. |
1820 | child: |
|
|
1821 | |
|
|
1822 | =over 4 |
|
|
1823 | |
|
|
1824 | =item IO::AIO::reinit |
|
|
1825 | |
|
|
1826 | Abondons all current requests and I/O threads and simply reinitialises all |
|
|
1827 | data structures. This is not an operation suppported by any standards, but |
|
|
1828 | happens to work on GNU/Linux and some newer BSD systems. |
|
|
1829 | |
|
|
1830 | The only reasonable use for this function is to call it after forking, if |
|
|
1831 | C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in |
|
|
1832 | the process will result in undefined behaviour. Calling it at any time |
|
|
1833 | will also result in any undefined (by POSIX) behaviour. |
|
|
1834 | |
|
|
1835 | =back |
1782 | |
1836 | |
1783 | =head2 MEMORY USAGE |
1837 | =head2 MEMORY USAGE |
1784 | |
1838 | |
1785 | Per-request usage: |
1839 | Per-request usage: |
1786 | |
1840 | |