ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/IO-AIO/AIO.pm
(Generate patch)

Comparing IO-AIO/AIO.pm (file contents):
Revision 1.85 by root, Sat Oct 28 01:40:30 2006 UTC vs.
Revision 1.92 by root, Wed Nov 8 01:57:42 2006 UTC

… …
61etc.), but can also be used to easily do operations in parallel that are 61etc.), but can also be used to easily do operations in parallel that are
62normally done sequentially, e.g. stat'ing many files, which is much faster 62normally done sequentially, e.g. stat'ing many files, which is much faster
63on a RAID volume or over NFS when you do a number of stat operations 63on a RAID volume or over NFS when you do a number of stat operations
64concurrently. 64concurrently.
65 65
66While this works on all types of file descriptors (for example sockets), 66While most of this works on all types of file descriptors (for example
67using these functions on file descriptors that support nonblocking 67sockets), using these functions on file descriptors that support
68operation (again, sockets, pipes etc.) is very inefficient. Use an event 68nonblocking operation (again, sockets, pipes etc.) is very inefficient or
69might not work (aio_read fails on sockets/pipes/fifos). Use an event loop
69loop for that (such as the L<Event|Event> module): IO::AIO will naturally 70for that (such as the L<Event|Event> module): IO::AIO will naturally fit
70fit into such an event loop itself. 71into such an event loop itself.
71 72
72In this version, a number of threads are started that execute your 73In this version, a number of threads are started that execute your
73requests and signal their completion. You don't need thread support 74requests and signal their completion. You don't need thread support
74in perl, and the threads created by this module will not be visible 75in perl, and the threads created by this module will not be visible
75to perl. In the future, this module might make use of the native aio 76to perl. In the future, this module might make use of the native aio
… …
82Although the module will work with in the presence of other (Perl-) 83Although the module will work with in the presence of other (Perl-)
83threads, it is currently not reentrant in any way, so use appropriate 84threads, it is currently not reentrant in any way, so use appropriate
84locking yourself, always call C<poll_cb> from within the same thread, or 85locking yourself, always call C<poll_cb> from within the same thread, or
85never call C<poll_cb> (or other C<aio_> functions) recursively. 86never call C<poll_cb> (or other C<aio_> functions) recursively.
86 87
88=head2 EXAMPLE
89
90This is a simple example that uses the Event module and loads
91F</etc/passwd> asynchronously:
92
93 use Fcntl;
94 use Event;
95 use IO::AIO;
96
97 # register the IO::AIO callback with Event
98 Event->io (fd => IO::AIO::poll_fileno,
99 poll => 'r',
100 cb => \&IO::AIO::poll_cb);
101
102 # queue the request to open /etc/passwd
103 aio_open "/etc/passwd", O_RDONLY, 0, sub {
104 my $fh = $_[0]
105 or die "error while opening: $!";
106
107 # stat'ing filehandles is generally non-blocking
108 my $size = -s $fh;
109
110 # queue a request to read the file
111 my $contents;
112 aio_read $fh, 0, $size, $contents, 0, sub {
113 $_[0] == $size
114 or die "short read: $!";
115
116 close $fh;
117
118 # file contents now in $contents
119 print $contents;
120
121 # exit event loop and program
122 Event::unloop;
123 };
124 };
125
126 # possibly queue up other requests, or open GUI windows,
127 # check for sockets etc. etc.
128
129 # process events as long as there are some:
130 Event::loop;
131
87=head1 REQUEST ANATOMY AND LIFETIME 132=head1 REQUEST ANATOMY AND LIFETIME
88 133
89Every C<aio_*> function creates a request. which is a C data structure not 134Every C<aio_*> function creates a request. which is a C data structure not
90directly visible to Perl. 135directly visible to Perl.
91 136
… …
132Request has reached the end of its lifetime and holds no resources anymore 177Request has reached the end of its lifetime and holds no resources anymore
133(except possibly for the Perl object, but its connection to the actual 178(except possibly for the Perl object, but its connection to the actual
134aio request is severed and calling its methods will either do nothing or 179aio request is severed and calling its methods will either do nothing or
135result in a runtime error). 180result in a runtime error).
136 181
182=back
183
137=cut 184=cut
138 185
139package IO::AIO; 186package IO::AIO;
140 187
141no warnings; 188no warnings;
142use strict 'vars'; 189use strict 'vars';
143 190
144use base 'Exporter'; 191use base 'Exporter';
145 192
146BEGIN { 193BEGIN {
147 our $VERSION = '2.0'; 194 our $VERSION = '2.2';
148 195
149 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat 196 our @AIO_REQ = qw(aio_sendfile aio_read aio_write aio_open aio_close aio_stat
150 aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink 197 aio_lstat aio_unlink aio_rmdir aio_readdir aio_scandir aio_symlink
151 aio_fsync aio_fdatasync aio_readahead aio_rename aio_link aio_move 198 aio_readlink aio_fsync aio_fdatasync aio_readahead aio_rename aio_link
152 aio_copy aio_group aio_nop aio_mknod); 199 aio_move aio_copy aio_group aio_nop aio_mknod);
153 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); 200 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice));
154 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 201 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush
155 min_parallel max_parallel nreqs nready npending); 202 min_parallel max_parallel max_idle
203 nreqs nready npending nthreads
204 max_poll_time max_poll_reqs);
156 205
157 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 206 @IO::AIO::GRP::ISA = 'IO::AIO::REQ';
158 207
159 require XSLoader; 208 require XSLoader;
160 XSLoader::load ("IO::AIO", $VERSION); 209 XSLoader::load ("IO::AIO", $VERSION);
161} 210}
162 211
163=head1 FUNCTIONS 212=head1 FUNCTIONS
164 213
165=head2 AIO FUNCTIONS 214=head2 AIO REQUEST FUNCTIONS
166 215
167All the C<aio_*> calls are more or less thin wrappers around the syscall 216All the C<aio_*> calls are more or less thin wrappers around the syscall
168with the same name (sans C<aio_>). The arguments are similar or identical, 217with the same name (sans C<aio_>). The arguments are similar or identical,
169and they all accept an additional (and optional) C<$callback> argument 218and they all accept an additional (and optional) C<$callback> argument
170which must be a code reference. This code reference will get called with 219which must be a code reference. This code reference will get called with
… …
173syscall has been executed asynchronously. 222syscall has been executed asynchronously.
174 223
175All functions expecting a filehandle keep a copy of the filehandle 224All functions expecting a filehandle keep a copy of the filehandle
176internally until the request has finished. 225internally until the request has finished.
177 226
178All requests return objects of type L<IO::AIO::REQ> that allow further 227All functions return request objects of type L<IO::AIO::REQ> that allow
179manipulation of those requests while they are in-flight. 228further manipulation of those requests while they are in-flight.
180 229
181The pathnames you pass to these routines I<must> be absolute and 230The pathnames you pass to these routines I<must> be absolute and
182encoded in byte form. The reason for the former is that at the time the 231encoded as octets. The reason for the former is that at the time the
183request is being executed, the current working directory could have 232request is being executed, the current working directory could have
184changed. Alternatively, you can make sure that you never change the 233changed. Alternatively, you can make sure that you never change the
185current working directory. 234current working directory anywhere in the program and then use relative
235paths.
186 236
187To encode pathnames to byte form, either make sure you either: a) 237To encode pathnames as octets, either make sure you either: a) always pass
188always pass in filenames you got from outside (command line, readdir 238in filenames you got from outside (command line, readdir etc.) without
189etc.), b) are ASCII or ISO 8859-1, c) use the Encode module and encode 239tinkering, b) are ASCII or ISO 8859-1, c) use the Encode module and encode
190your pathnames to the locale (or other) encoding in effect in the user 240your pathnames to the locale (or other) encoding in effect in the user
191environment, d) use Glib::filename_from_unicode on unicode filenames or e) 241environment, d) use Glib::filename_from_unicode on unicode filenames or e)
192use something else. 242use something else to ensure your scalar has the correct contents.
243
244This works, btw. independent of the internal UTF-8 bit, which IO::AIO
245handles correctly wether it is set or not.
193 246
194=over 4 247=over 4
195 248
196=item $prev_pri = aioreq_pri [$pri] 249=item $prev_pri = aioreq_pri [$pri]
197 250
… …
220 }; 273 };
221 274
222=item aioreq_nice $pri_adjust 275=item aioreq_nice $pri_adjust
223 276
224Similar to C<aioreq_pri>, but subtracts the given value from the current 277Similar to C<aioreq_pri>, but subtracts the given value from the current
225priority, so effects are cumulative. 278priority, so the effect is cumulative.
226 279
227=item aio_open $pathname, $flags, $mode, $callback->($fh) 280=item aio_open $pathname, $flags, $mode, $callback->($fh)
228 281
229Asynchronously open or create a file and call the callback with a newly 282Asynchronously open or create a file and call the callback with a newly
230created filehandle for the file. 283created filehandle for the file.
… …
347Asynchronously unlink (delete) a file and call the callback with the 400Asynchronously unlink (delete) a file and call the callback with the
348result code. 401result code.
349 402
350=item aio_mknod $path, $mode, $dev, $callback->($status) 403=item aio_mknod $path, $mode, $dev, $callback->($status)
351 404
405[EXPERIMENTAL]
406
352Asynchronously create a device node (or fifo). See mknod(2). 407Asynchronously create a device node (or fifo). See mknod(2).
353 408
354The only portable (POSIX) way of calling this function is: 409The only (POSIX-) portable way of calling this function is:
355 410
356 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ... 411 aio_mknod $path, IO::AIO::S_IFIFO | $mode, 0, sub { ...
357 412
358=item aio_link $srcpath, $dstpath, $callback->($status) 413=item aio_link $srcpath, $dstpath, $callback->($status)
359 414
… …
362 417
363=item aio_symlink $srcpath, $dstpath, $callback->($status) 418=item aio_symlink $srcpath, $dstpath, $callback->($status)
364 419
365Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 420Asynchronously create a new symbolic link to the existing object at C<$srcpath> at
366the path C<$dstpath> and call the callback with the result code. 421the path C<$dstpath> and call the callback with the result code.
422
423=item aio_readlink $path, $callback->($link)
424
425Asynchronously read the symlink specified by C<$path> and pass it to
426the callback. If an error occurs, nothing or undef gets passed to the
427callback.
367 428
368=item aio_rename $srcpath, $dstpath, $callback->($status) 429=item aio_rename $srcpath, $dstpath, $callback->($status)
369 430
370Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 431Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as
371rename(2) and call the callback with the result code. 432rename(2) and call the callback with the result code.
… …
845 906
846=back 907=back
847 908
848=head2 SUPPORT FUNCTIONS 909=head2 SUPPORT FUNCTIONS
849 910
911=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
912
850=over 4 913=over 4
851 914
852=item $fileno = IO::AIO::poll_fileno 915=item $fileno = IO::AIO::poll_fileno
853 916
854Return the I<request result pipe file descriptor>. This filehandle must be 917Return the I<request result pipe file descriptor>. This filehandle must be
… …
858 921
859See C<poll_cb> for an example. 922See C<poll_cb> for an example.
860 923
861=item IO::AIO::poll_cb 924=item IO::AIO::poll_cb
862 925
863Process all outstanding events on the result pipe. You have to call this 926Process some outstanding events on the result pipe. You have to call this
864regularly. Returns the number of events processed. Returns immediately 927regularly. Returns the number of events processed. Returns immediately
865when no events are outstanding. 928when no events are outstanding. The amount of events processed depends on
929the settings of C<IO::AIO::max_poll_req> and C<IO::AIO::max_poll_time>.
866 930
867If not all requests were processed for whatever reason, the filehandle 931If not all requests were processed for whatever reason, the filehandle
868will still be ready when C<poll_cb> returns. 932will still be ready when C<poll_cb> returns.
869 933
870Example: Install an Event watcher that automatically calls 934Example: Install an Event watcher that automatically calls
… …
872 936
873 Event->io (fd => IO::AIO::poll_fileno, 937 Event->io (fd => IO::AIO::poll_fileno,
874 poll => 'r', async => 1, 938 poll => 'r', async => 1,
875 cb => \&IO::AIO::poll_cb); 939 cb => \&IO::AIO::poll_cb);
876 940
877=item IO::AIO::poll_some $max_requests 941=item IO::AIO::max_poll_reqs $nreqs
878 942
879Similar to C<poll_cb>, but only processes up to C<$max_requests> requests 943=item IO::AIO::max_poll_time $seconds
880at a time.
881 944
882Useful if you want to ensure some level of interactiveness when perl is 945These set the maximum number of requests (default C<0>, meaning infinity)
883not fast enough to process all requests in time. 946that are being processed by C<IO::AIO::poll_cb> in one call, respectively
947the maximum amount of time (default C<0>, meaning infinity) spent in
948C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount
949of time C<poll_cb> is allowed to use).
950
951Setting C<max_poll_time> to a non-zero value creates an overhead of one
952syscall per request processed, which is not normally a problem unless your
953callbacks are really really fast or your OS is really really slow (I am
954not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead.
955
956Setting these is useful if you want to ensure some level of
957interactiveness when perl is not fast enough to process all requests in
958time.
959
960For interactive programs, values such as C<0.01> to C<0.1> should be fine.
884 961
885Example: Install an Event watcher that automatically calls 962Example: Install an Event watcher that automatically calls
886IO::AIO::poll_some with low priority, to ensure that other parts of the 963IO::AIO::poll_cb with low priority, to ensure that other parts of the
887program get the CPU sometimes even under high AIO load. 964program get the CPU sometimes even under high AIO load.
888 965
966 # try not to spend much more than 0.1s in poll_cb
967 IO::AIO::max_poll_time 0.1;
968
969 # use a low priority so other tasks have priority
889 Event->io (fd => IO::AIO::poll_fileno, 970 Event->io (fd => IO::AIO::poll_fileno,
890 poll => 'r', nice => 1, 971 poll => 'r', nice => 1,
891 cb => sub { IO::AIO::poll_some 256 }); 972 cb => &IO::AIO::poll_cb);
892 973
893=item IO::AIO::poll_wait 974=item IO::AIO::poll_wait
894 975
895Wait till the result filehandle becomes ready for reading (simply does a 976If there are any outstanding requests, wait till the result filehandle
896C<select> on the filehandle. This is useful if you want to synchronously wait 977becomes ready for reading (simply does a C<select> on the filehandle. This
897for some requests to finish). 978is useful if you want to synchronously wait for some requests to finish).
898 979
899See C<nreqs> for an example. 980See C<nreqs> for an example.
900 981
982=item IO::AIO::poll
983
984Waits until some requests have been handled.
985
986Returns the number of requests processed, but is otherwise strictly
987equivalent to:
988
989 IO::AIO::poll_wait, IO::AIO::poll_cb
990 if IO::AIO::nreqs;
991
901=item IO::AIO::nreqs 992=item IO::AIO::flush
902 993
903Returns the number of requests currently in the ready, execute or pending 994Wait till all outstanding AIO requests have been handled.
904states (i.e. for which their callback has not been invoked yet).
905 995
906Example: wait till there are no outstanding requests anymore: 996Strictly equivalent to:
907 997
908 IO::AIO::poll_wait, IO::AIO::poll_cb 998 IO::AIO::poll_wait, IO::AIO::poll_cb
909 while IO::AIO::nreqs; 999 while IO::AIO::nreqs;
910 1000
911=item IO::AIO::nready 1001=head3 CONTROLLING THE NUMBER OF THREADS
912
913Returns the number of requests currently in the ready state (not yet
914executed).
915
916=item IO::AIO::npending
917
918Returns the number of requests currently in the pending state (executed,
919but not yet processed by poll_cb).
920
921=item IO::AIO::flush
922
923Wait till all outstanding AIO requests have been handled.
924
925Strictly equivalent to:
926
927 IO::AIO::poll_wait, IO::AIO::poll_cb
928 while IO::AIO::nreqs;
929
930=item IO::AIO::poll
931
932Waits until some requests have been handled.
933
934Strictly equivalent to:
935
936 IO::AIO::poll_wait, IO::AIO::poll_cb
937 if IO::AIO::nreqs;
938 1002
939=item IO::AIO::min_parallel $nthreads 1003=item IO::AIO::min_parallel $nthreads
940 1004
941Set the minimum number of AIO threads to C<$nthreads>. The current 1005Set the minimum number of AIO threads to C<$nthreads>. The current
942default is C<8>, which means eight asynchronous operations can execute 1006default is C<8>, which means eight asynchronous operations can execute
943concurrently at any one time (the number of outstanding requests, 1007concurrently at any one time (the number of outstanding requests,
944however, is unlimited). 1008however, is unlimited).
945 1009
946IO::AIO starts threads only on demand, when an AIO request is queued and 1010IO::AIO starts threads only on demand, when an AIO request is queued and
947no free thread exists. 1011no free thread exists. Please note that queueing up a hundred requests can
1012create demand for a hundred threads, even if it turns out that everything
1013is in the cache and could have been processed faster by a single thread.
948 1014
949It is recommended to keep the number of threads relatively low, as some 1015It is recommended to keep the number of threads relatively low, as some
950Linux kernel versions will scale negatively with the number of threads 1016Linux kernel versions will scale negatively with the number of threads
951(higher parallelity => MUCH higher latency). With current Linux 2.6 1017(higher parallelity => MUCH higher latency). With current Linux 2.6
952versions, 4-32 threads should be fine. 1018versions, 4-32 threads should be fine.
… …
966This module automatically runs C<max_parallel 0> at program end, to ensure 1032This module automatically runs C<max_parallel 0> at program end, to ensure
967that all threads are killed and that there are no outstanding requests. 1033that all threads are killed and that there are no outstanding requests.
968 1034
969Under normal circumstances you don't need to call this function. 1035Under normal circumstances you don't need to call this function.
970 1036
1037=item IO::AIO::max_idle $nthreads
1038
1039Limit the number of threads (default: 4) that are allowed to idle (i.e.,
1040threads that did not get a request to process within 10 seconds). That
1041means if a thread becomes idle while C<$nthreads> other threads are also
1042idle, it will free its resources and exit.
1043
1044This is useful when you allow a large number of threads (e.g. 100 or 1000)
1045to allow for extremely high load situations, but want to free resources
1046under normal circumstances (1000 threads can easily consume 30MB of RAM).
1047
1048The default is probably ok in most situations, especially if thread
1049creation is fast. If thread creation is very slow on your system you might
1050want to use larger values.
1051
971=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs 1052=item $oldmaxreqs = IO::AIO::max_outstanding $maxreqs
972 1053
973This is a very bad function to use in interactive programs because it 1054This is a very bad function to use in interactive programs because it
974blocks, and a bad way to reduce concurrency because it is inexact: Better 1055blocks, and a bad way to reduce concurrency because it is inexact: Better
975use an C<aio_group> together with a feed callback. 1056use an C<aio_group> together with a feed callback.
… …
983number of outstanding requests. 1064number of outstanding requests.
984 1065
985You can still queue as many requests as you want. Therefore, 1066You can still queue as many requests as you want. Therefore,
986C<max_oustsanding> is mainly useful in simple scripts (with low values) or 1067C<max_oustsanding> is mainly useful in simple scripts (with low values) or
987as a stop gap to shield against fatal memory overflow (with large values). 1068as a stop gap to shield against fatal memory overflow (with large values).
1069
1070=head3 STATISTICAL INFORMATION
1071
1072=item IO::AIO::nreqs
1073
1074Returns the number of requests currently in the ready, execute or pending
1075states (i.e. for which their callback has not been invoked yet).
1076
1077Example: wait till there are no outstanding requests anymore:
1078
1079 IO::AIO::poll_wait, IO::AIO::poll_cb
1080 while IO::AIO::nreqs;
1081
1082=item IO::AIO::nready
1083
1084Returns the number of requests currently in the ready state (not yet
1085executed).
1086
1087=item IO::AIO::npending
1088
1089Returns the number of requests currently in the pending state (executed,
1090but not yet processed by poll_cb).
988 1091
989=back 1092=back
990 1093
991=cut 1094=cut
992 1095

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines