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

Comparing OpenCL/OpenCL.pm (file contents):
Revision 1.52 by root, Tue Apr 24 14:24:42 2012 UTC vs.
Revision 1.63 by root, Sun Apr 29 18:24:35 2012 UTC

43 43
44OpenCL::Event objects are used to signal when something is complete. 44OpenCL::Event objects are used to signal when something is complete.
45 45
46=head2 HELPFUL RESOURCES 46=head2 HELPFUL RESOURCES
47 47
48The OpenCL spec used to develop this module (1.2 spec was available, but 48The OpenCL specs used to develop this module:
49no implementation was available to me :).
50 49
51 http://www.khronos.org/registry/cl/specs/opencl-1.1.pdf 50 http://www.khronos.org/registry/cl/specs/opencl-1.1.pdf
51 http://www.khronos.org/registry/cl/specs/opencl-1.2.pdf
52 http://www.khronos.org/registry/cl/specs/opencl-1.2-extensions.pdf
52 53
53OpenCL manpages: 54OpenCL manpages:
54 55
55 http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/ 56 http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/
57 http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/
56 58
57If you are into UML class diagrams, the following diagram might help - if 59If you are into UML class diagrams, the following diagram might help - if
58not, it will be mildly cobfusing: 60not, it will be mildly confusing (also, the class hierarchy of this module
61is much more fine-grained):
59 62
60 http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/classDiagram.html 63 http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/classDiagram.html
61 64
62Here's a tutorial from AMD (very AMD-centric, too), not sure how useful it 65Here's a tutorial from AMD (very AMD-centric, too), not sure how useful it
63is, but at least it's free of charge: 66is, but at least it's free of charge:
64 67
65 http://developer.amd.com/zones/OpenCLZone/courses/Documents/Introduction_to_OpenCL_Programming%20Training_Guide%20%28201005%29.pdf 68 http://developer.amd.com/zones/OpenCLZone/courses/Documents/Introduction_to_OpenCL_Programming%20Training_Guide%20%28201005%29.pdf
138=head2 Create a buffer with some predefined data, read it back synchronously, 141=head2 Create a buffer with some predefined data, read it back synchronously,
139then asynchronously. 142then asynchronously.
140 143
141 my $buf = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, "helmut"); 144 my $buf = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, "helmut");
142 145
143 $queue->enqueue_read_buffer ($buf, 1, 1, 3, my $data); 146 $queue->read_buffer ($buf, 1, 1, 3, my $data);
144 print "$data\n"; 147 print "$data\n";
145 148
146 my $ev = $queue->enqueue_read_buffer ($buf, 0, 1, 3, my $data); 149 my $ev = $queue->read_buffer ($buf, 0, 1, 3, my $data);
147 $ev->wait; 150 $ev->wait;
148 print "$data\n"; # prints "elm" 151 print "$data\n"; # prints "elm"
149 152
150=head2 Create and build a program, then create a kernel out of one of its 153=head2 Create and build a program, then create a kernel out of one of its
151functions. 154functions.
171 # set buffer 174 # set buffer
172 $kernel->set_buffer (0, $input); 175 $kernel->set_buffer (0, $input);
173 $kernel->set_buffer (1, $output); 176 $kernel->set_buffer (1, $output);
174 177
175 # execute it for all 4 numbers 178 # execute it for all 4 numbers
176 $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 179 $queue->nd_range_kernel ($kernel, undef, [4], undef);
177 180
178 # enqueue a synchronous read 181 # enqueue a synchronous read
179 $queue->enqueue_read_buffer ($output, 1, 0, OpenCL::SIZEOF_FLOAT * 4, my $data); 182 $queue->read_buffer ($output, 1, 0, OpenCL::SIZEOF_FLOAT * 4, my $data);
180 183
181 # print the results: 184 # print the results:
182 printf "%s\n", join ", ", unpack "f*", $data; 185 printf "%s\n", join ", ", unpack "f*", $data;
183 186
184=head2 The same enqueue operations as before, but assuming an out-of-order queue, 187=head2 The same enqueue operations as before, but assuming an out-of-order queue,
185showing off barriers. 188showing off barriers.
186 189
187 # execute it for all 4 numbers 190 # execute it for all 4 numbers
188 $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 191 $queue->nd_range_kernel ($kernel, undef, [4], undef);
189 192
190 # enqueue a barrier to ensure in-order execution 193 # enqueue a barrier to ensure in-order execution
191 $queue->enqueue_barrier; 194 $queue->barrier;
192 195
193 # enqueue an async read 196 # enqueue an async read
194 $queue->enqueue_read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data); 197 $queue->read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data);
195 198
196 # wait for all requests to finish 199 # wait for all requests to finish
197 $queue->finish; 200 $queue->finish;
198 201
199=head2 The same enqueue operations as before, but assuming an out-of-order queue, 202=head2 The same enqueue operations as before, but assuming an out-of-order queue,
200showing off event objects and wait lists. 203showing off event objects and wait lists.
201 204
202 # execute it for all 4 numbers 205 # execute it for all 4 numbers
203 my $ev = $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 206 my $ev = $queue->nd_range_kernel ($kernel, undef, [4], undef);
204 207
205 # enqueue an async read 208 # enqueue an async read
206 $ev = $queue->enqueue_read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data, $ev); 209 $ev = $queue->read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data, $ev);
207 210
208 # wait for the last event to complete 211 # wait for the last event to complete
209 $ev->wait; 212 $ev->wait;
210 213
211=head2 Use the OpenGL module to share a texture between OpenCL and OpenGL and draw some julia 214=head2 Use the OpenGL module to share a texture between OpenCL and OpenGL and draw some julia
270 273
271 # program compiled, kernel ready, now draw and loop 274 # program compiled, kernel ready, now draw and loop
272 275
273 for (my $time; ; ++$time) { 276 for (my $time; ; ++$time) {
274 # acquire objects from opengl 277 # acquire objects from opengl
275 $queue->enqueue_acquire_gl_objects ([$tex]); 278 $queue->acquire_gl_objects ([$tex]);
276 279
277 # configure and run our kernel 280 # configure and run our kernel
278 $kernel->set_image2d (0, $tex); 281 $kernel->set_image2d (0, $tex);
279 $kernel->set_float (1, $time); 282 $kernel->set_float (1, $time);
280 $queue->enqueue_nd_range_kernel ($kernel, undef, [256, 256], undef); 283 $queue->nd_range_kernel ($kernel, undef, [256, 256], undef);
281 284
282 # release objects to opengl again 285 # release objects to opengl again
283 $queue->enqueue_release_gl_objects ([$tex]); 286 $queue->release_gl_objects ([$tex]);
284 287
285 # wait 288 # wait
286 $queue->finish; 289 $queue->finish;
287 290
288 # now draw the texture, the defaults should be all right 291 # now draw the texture, the defaults should be all right
376 379
377For this to work, the OpenGL library must be loaded, a GLX context must 380For this to work, the OpenGL library must be loaded, a GLX context must
378have been created and be made current, and C<dlsym> must be available and 381have been created and be made current, and C<dlsym> must be available and
379capable of finding the function via C<RTLD_DEFAULT>. 382capable of finding the function via C<RTLD_DEFAULT>.
380 383
384=head2 EVENT SYSTEM
385
386OpenCL can generate a number of (potentially) asynchronous events, for
387example, after compiling a program, to signal a context-related error or,
388perhaps most important, to signal completion of queued jobs (by setting
389callbacks on OpenCL::Event objects).
390
391To facilitate this, this module maintains an event queue - each
392time an asynchronous event happens, it is queued, and perl will be
393interrupted. This is implemented via the L<Async::Interrupt> module. In
394addition, this module has L<AnyEvent> support, so it can seamlessly
395integrate itself into many event loops.
396
397Since this module is a bit hard to understand, here are some case examples:
398
399=head3 Don't use callbacks.
400
401When your program never uses any callbacks, then there will never be any
402notifications you need to take care of, and therefore no need to worry
403about all this.
404
405You can achieve a great deal by explicitly waiting for events, or using
406barriers and flush calls. In many programs, there is no need at all to
407tinker with asynchronous events.
408
409=head3 Use AnyEvent
410
411This module automatically registers a watcher that invokes all outstanding
412event callbacks when AnyEvent is initialised (and block asynchronous
413interruptions). Using this mode of operations is the safest and most
414recommended one.
415
416To use this, simply use AnyEvent and this module normally, make sure you
417have an event loop running:
418
419 use Gtk2 -init;
420 use AnyEvent;
421
422 # initialise AnyEvent, by creating a watcher, or:
423 AnyEvent::detect;
424
425 my $e = $queue->marker;
426 $e->cb (sub {
427 warn "opencl is finished\n";
428 })
429
430 main Gtk2;
431
432Note that this module will not initialise AnyEvent for you. Before
433AnyEvent is initialised, the module will asynchronously interrupt perl
434instead. To avoid any surprises, it's best to explicitly initialise
435AnyEvent.
436
437You can temporarily enable asynchronous interruptions (see next paragraph)
438by calling C<$OpenCL::INTERRUPT->unblock> and disable them again by
439calling C<$OpenCL::INTERRUPT->block>.
440
441=head3 Let yourself be interrupted at any time
442
443This mode is the default unless AnyEvent is loaded and initialised. In
444this mode, OpenCL asynchronously interrupts a running perl program. The
445emphasis is on both I<asynchronously> and I<running> here.
446
447Asynchronously means that perl might execute your callbacks at any
448time. For example, in the following code (I<THAT YOU SHOULD NOT COPY>),
449the C<until> loop following the marker call will be interrupted by the
450callback:
451
452 my $e = $queue->marker;
453 my $flag;
454 $e->cb (sub { $flag = 1 });
455 1 until $flag;
456 # $flag is now 1
457
458The reason why you shouldn't blindly copy the above code is that
459busy waiting is a really really bad thing, and really really bad for
460performance.
461
462While at first this asynchronous business might look exciting, it can be
463really hard, because you need to be prepared for the callback code to be
464executed at any time, which limits the amount of things the callback code
465can do safely.
466
467This can be mitigated somewhat by using C<<
468$OpenCL::INTERRUPT->scope_block >> (see the L<Async::Interrupt>
469documentation for details).
470
471The other problem is that your program must be actively I<running> to be
472interrupted. When you calculate stuff, your program is running. When you
473hang in some C functions or other block execution (by calling C<sleep>,
474C<select>, running an event loop and so on), your program is waiting, not
475running.
476
477One way around that would be to attach a read watcher to your event loop,
478listening for events on C<< $OpenCL::INTERRUPT->pipe_fileno >>, using a
479dummy callback (C<sub { }>) to temporarily execute some perl code.
480
481That is then awfully close to using the built-in AnyEvent support above,
482though, so consider that one instead.
483
484=head3 Be creative
485
486OpenCL exports the L<Async::Interrupt> object it uses in the global
487variable C<$OpenCL::INTERRUPT>. You can configure it in any way you like.
488
489So if you want to feel like a real pro, err, wait, if you feel no risk
490menas no fun, you can experiment by implementing your own mode of
491operations.
492
381=cut 493=cut
382 494
383package OpenCL; 495package OpenCL;
384 496
385use common::sense; 497use common::sense;
498use Carp ();
499use Async::Interrupt ();
500
501our $POLL_FUNC; # set by XS
386 502
387BEGIN { 503BEGIN {
388 our $VERSION = '0.96'; 504 our $VERSION = '0.98';
389 505
390 require XSLoader; 506 require XSLoader;
391 XSLoader::load (__PACKAGE__, $VERSION); 507 XSLoader::load (__PACKAGE__, $VERSION);
392 508
393 @OpenCL::Platform::ISA = 509 @OpenCL::Platform::ISA =
422=item $int = OpenCL::errno 538=item $int = OpenCL::errno
423 539
424The last error returned by a function - it's only valid after an error occured 540The last error returned by a function - it's only valid after an error occured
425and before calling another OpenCL function. 541and before calling another OpenCL function.
426 542
427=item $str = OpenCL::err2str $errval 543=item $str = OpenCL::err2str [$errval]
428 544
429Comverts an error value into a human readable string. 545Converts an error value into a human readable string. IF no error value is
546given, then the last error will be used (as returned by OpenCL::errno).
430 547
431=item $str = OpenCL::enum2str $enum 548=item $str = OpenCL::enum2str $enum
432 549
433Converts most enum values (of parameter names, image format constants, 550Converts most enum values (of parameter names, image format constants,
434object types, addressing and filter modes, command types etc.) into a 551object types, addressing and filter modes, command types etc.) into a
440 557
441Returns all available OpenCL::Platform objects. 558Returns all available OpenCL::Platform objects.
442 559
443L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html> 560L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html>
444 561
445=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $notify = undef 562=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr
446 563
447Tries to create a context from a default device and platform - never worked for me. 564Tries to create a context from a default device and platform type - never worked for me.
448 565
449L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 566L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
450 567
568=item $ctx = OpenCL::context $properties, \@devices, $callback->($err, $pvt) = $print_stderr)
569
570Create a new OpenCL::Context object using the given device object(s). This
571function isn't implemented yet, use C<< $platform->context >> instead.
572
573L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
574
451=item OpenCL::wait_for_events $wait_events... 575=item OpenCL::wait_for_events $wait_events...
452 576
453Waits for all events to complete. 577Waits for all events to complete.
454 578
455L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 579L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
580
581=item OpenCL::poll
582
583Checks if there are any outstanding events (see L<EVENT SYSTEM>) and
584invokes their callbacks.
585
586=item $OpenCL::INTERRUPT
587
588The L<Async::Interrupt> object used to signal asynchronous events (see
589L<EVENT SYSTEM>).
590
591=cut
592
593our $INTERRUPT = new Async::Interrupt c_cb => [$POLL_FUNC, 0];
594
595&_eq_initialise ($INTERRUPT->signal_func);
596
597=item $OpenCL::WATCHER
598
599The L<AnyEvent> watcher object used to watch for asynchronous events (see
600L<EVENT SYSTEM>). This variable is C<undef> until L<AnyEvent> has been
601loaded I<and> initialised (e.g. by calling C<AnyEvent::detect>).
602
603=cut
604
605our $WATCHER;
606
607sub _init_anyevent {
608 $INTERRUPT->block;
609 $WATCHER = AE::io ($INTERRUPT->pipe_fileno, 0, sub { $INTERRUPT->handle });
610}
611
612if (defined $AnyEvent::MODEL) {
613 _init_anyevent;
614} else {
615 push @AnyEvent::post_detect, \&_init_anyevent;
616}
456 617
457=back 618=back
458 619
459=head2 THE OpenCL::Object CLASS 620=head2 THE OpenCL::Object CLASS
460 621
472C<IV> type in your code and cast that to the correct type. 633C<IV> type in your code and cast that to the correct type.
473 634
474=cut 635=cut
475 636
476sub OpenCL::Object::id { 637sub OpenCL::Object::id {
638 ref $_[0] eq "SCALAR"
477 ${$_[0]} 639 ? ${ $_[0] }
640 : $_[0][0]
478} 641}
479 642
480=back 643=back
481 644
482=head2 THE OpenCL::Platform CLASS 645=head2 THE OpenCL::Platform CLASS
485 648
486=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL) 649=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL)
487 650
488Returns a list of matching OpenCL::Device objects. 651Returns a list of matching OpenCL::Device objects.
489 652
490=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $notify = undef) 653=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr)
491 654
492Tries to create a context. Never worked for me, and you need devices explicitly anyway. 655Tries to create a context. Never worked for me, and you need devices explicitly anyway.
493 656
494L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 657L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
495 658
496=item $ctx = $platform->context ($properties = undef, @$devices, $notify = undef) 659=item $ctx = $platform->context ($properties, \@devices, $callback->($err, $pvt) = $print_stderr)
497 660
498Create a new OpenCL::Context object using the given device object(s)- a 661Create a new OpenCL::Context object using the given device object(s)- a
499CL_CONTEXT_PLATFORM property is supplied automatically. 662CL_CONTEXT_PLATFORM property is supplied automatically.
500 663
501L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html> 664L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
844 my ($self, $prog, $options) = @_; 1007 my ($self, $prog, $options) = @_;
845 1008
846 $prog = $self->program_with_source ($prog) 1009 $prog = $self->program_with_source ($prog)
847 unless ref $prog; 1010 unless ref $prog;
848 1011
1012 eval { $prog->build (undef, $options); 1 }
1013 or errno == BUILD_PROGRAM_FAILURE
1014 or errno == INVALID_BINARY # workaround nvidia bug
1015 or Carp::croak "OpenCL::Context->build_program: " . err2str;
1016
1017 # we check status for all devices
849 for my $dev ($self->devices) { 1018 for my $dev ($self->devices) {
850 eval { $prog->build ($dev, $options); 1 } 1019 $prog->build_status ($dev) == BUILD_SUCCESS
851 or Carp::croak "Building OpenCL program for device '" . $dev->name . "' failed:\n" 1020 or Carp::croak "Building OpenCL program for device '" . $dev->name . "' failed:\n"
852 . $prog->build_log ($dev); 1021 . $prog->build_log ($dev);
853 } 1022 }
854 1023
855 $prog 1024 $prog
856} 1025}
857 1026
881=item $buf = $ctx->buffer_sv ($flags, $data) 1050=item $buf = $ctx->buffer_sv ($flags, $data)
882 1051
883Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and 1052Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and
884initialise it with the given data values. 1053initialise it with the given data values.
885 1054
886=item $img = $ctx->image ($self, $flags, $channel_order, $channel_type, $type, $width, $height, $depth, $array_size = 0, $row_pitch = 0, $slice_pitch = 0, $num_mip_level = 0, $num_samples = 0, $*data = &PL_sv_undef) 1055=item $img = $ctx->image ($self, $flags, $channel_order, $channel_type, $type, $width, $height, $depth = 0, $array_size = 0, $row_pitch = 0, $slice_pitch = 0, $num_mip_level = 0, $num_samples = 0, $*data = &PL_sv_undef)
887 1056
888Creates a new OpenCL::Image object and optionally initialises it with 1057Creates a new OpenCL::Image object and optionally initialises it with
889the given data values. 1058the given data values.
890 1059
891L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateImage.html> 1060L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateImage.html>
987=back 1156=back
988 1157
989=head2 THE OpenCL::Queue CLASS 1158=head2 THE OpenCL::Queue CLASS
990 1159
991An OpenCL::Queue represents an execution queue for OpenCL. You execute 1160An OpenCL::Queue represents an execution queue for OpenCL. You execute
992requests by calling their respective C<enqueue_xxx> method and waitinf for 1161requests by calling their respective method and waiting for it to complete
993it to complete in some way. 1162in some way.
994 1163
995All the enqueue methods return an event object that can be used to wait 1164Most methods that enqueue some request return an event object that can
996for completion, unless the method is called in void context, in which case 1165be used to wait for completion (optionally using a callback), unless
997no event object is created. 1166the method is called in void context, in which case no event object is
1167created.
998 1168
999They also allow you to specify any number of other event objects that this 1169They also allow you to specify any number of other event objects that this
1000request has to wait for before it starts executing, by simply passing the 1170request has to wait for before it starts executing, by simply passing the
1001event objects as extra parameters to the enqueue methods. To simplify 1171event objects as extra parameters to the enqueue methods. To simplify
1002program design, this module ignores any C<undef> values in the list of 1172program design, this module ignores any C<undef> values in the list of
1003events. This makes it possible to code operations such as this, without 1173events. This makes it possible to code operations such as this, without
1004having to put a valid event object into C<$event> first: 1174having to put a valid event object into C<$event> first:
1005 1175
1006 $event = $queue->enqueue_xxx (..., $event); 1176 $event = $queue->xxx (..., $event);
1007 1177
1008Queues execute in-order by default, without any parallelism, so in most 1178Queues execute in-order by default, without any parallelism, so in most
1009cases (i.e. you use only one queue) it's not necessary to wait for or 1179cases (i.e. you use only one queue) it's not necessary to wait for or
1010create event objects, althoguh an our of order queue is often a bit 1180create event objects, althoguh an our of order queue is often a bit
1011faster. 1181faster.
1012 1182
1013=over 4 1183=over 4
1014 1184
1015=item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...) 1185=item $ev = $queue->read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...)
1016 1186
1017Reads data from buffer into the given string. 1187Reads data from buffer into the given string.
1018 1188
1019L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBuffer.html> 1189L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBuffer.html>
1020 1190
1021=item $ev = $queue->enqueue_write_buffer ($buffer, $blocking, $offset, $data, $wait_events...) 1191=item $ev = $queue->write_buffer ($buffer, $blocking, $offset, $data, $wait_events...)
1022 1192
1023Writes data to buffer from the given string. 1193Writes data to buffer from the given string.
1024 1194
1025L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBuffer.html> 1195L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBuffer.html>
1026 1196
1027=item $ev = $queue->enqueue_copy_buffer ($src, $dst, $src_offset, $dst_offset, $len, $wait_events...) 1197=item $ev = $queue->copy_buffer ($src, $dst, $src_offset, $dst_offset, $len, $wait_events...)
1028 1198
1029L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBuffer.html> 1199L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBuffer.html>
1030 1200
1031=item $ev = $queue->enqueue_read_buffer_rect (OpenCL::Memory buf, cl_bool blocking, $buf_x, $buf_y, $buf_z, $host_x, $host_y, $host_z, $width, $height, $depth, $buf_row_pitch, $buf_slice_pitch, $host_row_pitch, $host_slice_pitch, $data, $wait_events...) 1201=item $ev = $queue->read_buffer_rect (OpenCL::Memory buf, cl_bool blocking, $buf_x, $buf_y, $buf_z, $host_x, $host_y, $host_z, $width, $height, $depth, $buf_row_pitch, $buf_slice_pitch, $host_row_pitch, $host_slice_pitch, $data, $wait_events...)
1032 1202
1033http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBufferRect.html 1203http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBufferRect.html
1034 1204
1035=item $ev = $queue->enqueue_write_buffer_rect (OpenCL::Memory buf, cl_bool blocking, $buf_x, $buf_y, $buf_z, $host_x, $host_y, $host_z, $width, $height, $depth, $buf_row_pitch, $buf_slice_pitch, $host_row_pitch, $host_slice_pitch, $data, $wait_events...) 1205=item $ev = $queue->write_buffer_rect (OpenCL::Memory buf, cl_bool blocking, $buf_x, $buf_y, $buf_z, $host_x, $host_y, $host_z, $width, $height, $depth, $buf_row_pitch, $buf_slice_pitch, $host_row_pitch, $host_slice_pitch, $data, $wait_events...)
1036 1206
1037http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBufferRect.html 1207http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBufferRect.html
1038 1208
1039=item $ev = $queue->enqueue_read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) 1209=item $ev = $queue->read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...)
1040 1210
1041L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferRect.html> 1211L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferRect.html>
1042 1212
1043=item $ev = $queue->enqueue_copy_buffer_to_image ($src_buffer, $dst_image, $src_offset, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...) 1213=item $ev = $queue->copy_buffer_to_image ($src_buffer, $dst_image, $src_offset, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...)
1044 1214
1045L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html> 1215L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html>
1046 1216
1047=item $ev = $queue->enqueue_write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) 1217=item $ev = $queue->write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...)
1048 1218
1049L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html> 1219L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html>
1050 1220
1051=item $ev = $queue->enqueue_copy_image ($src_image, $dst_image, $src_x, $src_y, $src_z, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...) 1221=item $ev = $queue->copy_image ($src_image, $dst_image, $src_x, $src_y, $src_z, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...)
1052 1222
1053L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html> 1223L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html>
1054 1224
1055=item $ev = $queue->enqueue_copy_image_to_buffer ($src_image, $dst_image, $src_x, $src_y, $src_z, $width, $height, $depth, $dst_offset, $wait_events...) 1225=item $ev = $queue->copy_image_to_buffer ($src_image, $dst_image, $src_x, $src_y, $src_z, $width, $height, $depth, $dst_offset, $wait_events...)
1056 1226
1057L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImageToBuffer.html> 1227L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImageToBuffer.html>
1058 1228
1059=item $ev = $queue->enqueue_copy_buffer_rect ($src, $dst, $src_x, $src_y, $src_z, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $src_row_pitch, $src_slice_pitch, $dst_row_pitch, $dst_slice_pitch, $wait_event...) 1229=item $ev = $queue->copy_buffer_rect ($src, $dst, $src_x, $src_y, $src_z, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $src_row_pitch, $src_slice_pitch, $dst_row_pitch, $dst_slice_pitch, $wait_event...)
1060 1230
1061Yeah. 1231Yeah.
1062 1232
1063L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>. 1233L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>.
1064 1234
1065=item $ev = $queue->enqueue_fill_buffer ($mem, $pattern, $offset, $size, ...) 1235=item $ev = $queue->fill_buffer ($mem, $pattern, $offset, $size, ...)
1066 1236
1067Fills the given buffer object with repeated applications of C<$pattern>, 1237Fills the given buffer object with repeated applications of C<$pattern>,
1068starting at C<$offset> for C<$size> octets. 1238starting at C<$offset> for C<$size> octets.
1069 1239
1070L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillBuffer.html> 1240L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillBuffer.html>
1071 1241
1072=item $ev = $queue->enqueue_fill_image ($img, $r, $g, $b, $a, $x, $y, $z, $width, $height, $depth, ...) 1242=item $ev = $queue->fill_image ($img, $r, $g, $b, $a, $x, $y, $z, $width, $height, $depth, ...)
1073 1243
1074Fills the given image area with the given rgba colour components. The 1244Fills the given image area with the given rgba colour components. The
1075components are normally floating point values between C<0> and C<1>, 1245components are normally floating point values between C<0> and C<1>,
1076except when the image channel data type is a signe dor unsigned 1246except when the image channel data type is a signe dor unsigned
1077unnormalised format, in which case the range is determined by the format. 1247unnormalised format, in which case the range is determined by the format.
1078 1248
1079L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillImage.html> 1249L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillImage.html>
1080 1250
1081=item $ev = $queue->enqueue_task ($kernel, $wait_events...) 1251=item $ev = $queue->task ($kernel, $wait_events...)
1082 1252
1083L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html> 1253L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html>
1084 1254
1085=item $ev = $queue->enqueue_nd_range_kernel ($kernel, @$global_work_offset, @$global_work_size, @$local_work_size, $wait_events...) 1255=item $ev = $queue->nd_range_kernel ($kernel, \@global_work_offset, \@global_work_size, \@local_work_size, $wait_events...)
1086 1256
1087Enqueues a kernel execution. 1257Enqueues a kernel execution.
1088 1258
1089@$global_work_size must be specified as a reference to an array of 1259\@global_work_size must be specified as a reference to an array of
1090integers specifying the work sizes (element counts). 1260integers specifying the work sizes (element counts).
1091 1261
1092@$global_work_offset must be either C<undef> (in which case all offsets 1262\@global_work_offset must be either C<undef> (in which case all offsets
1093are C<0>), or a reference to an array of work offsets, with the same number 1263are C<0>), or a reference to an array of work offsets, with the same number
1094of elements as @$global_work_size. 1264of elements as \@global_work_size.
1095 1265
1096@$local_work_size must be either C<undef> (in which case the 1266\@local_work_size must be either C<undef> (in which case the
1097implementation is supposed to choose good local work sizes), or a 1267implementation is supposed to choose good local work sizes), or a
1098reference to an array of local work sizes, with the same number of 1268reference to an array of local work sizes, with the same number of
1099elements as @$global_work_size. 1269elements as \@global_work_size.
1100 1270
1101L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html> 1271L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html>
1102 1272
1103=item $ev = $queue->enqueue_acquire_gl_objects ([object, ...], $wait_events...) 1273=item $ev = $queue->acquire_gl_objects ([object, ...], $wait_events...)
1104 1274
1105Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired 1275Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired
1106for subsequent OpenCL usage. 1276for subsequent OpenCL usage.
1107 1277
1108L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueAcquireGLObjects.html> 1278L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueAcquireGLObjects.html>
1109 1279
1110=item $ev = $queue->enqueue_release_gl_objects ([object, ...], $wait_events...) 1280=item $ev = $queue->release_gl_objects ([object, ...], $wait_events...)
1111 1281
1112Enqueues a list (an array-ref of OpenCL::Memory objects) to be released 1282Enqueues a list (an array-ref of OpenCL::Memory objects) to be released
1113for subsequent OpenGL usage. 1283for subsequent OpenGL usage.
1114 1284
1115L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReleaseGLObjects.html> 1285L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReleaseGLObjects.html>
1116 1286
1117=item $ev = $queue->enqueue_wait_for_events ($wait_events...) 1287=item $ev = $queue->wait_for_events ($wait_events...)
1118 1288
1119L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html> 1289L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html>
1120 1290
1121=item $ev = $queue->enqueue_marker ($wait_events...) 1291=item $ev = $queue->marker ($wait_events...)
1122 1292
1123L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueMarkerWithWaitList.html> 1293L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueMarkerWithWaitList.html>
1124 1294
1125=item $ev = $queue->enqueue_barrier ($wait_events...) 1295=item $ev = $queue->barrier ($wait_events...)
1126 1296
1127L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueBarrierWithWaitList.html> 1297L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueBarrierWithWaitList.html>
1128 1298
1129=item $queue->flush 1299=item $queue->flush
1130 1300
1253OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D, 1423OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D,
1254OpenCL::Image2DArray and OpenCL::Image3D. 1424OpenCL::Image2DArray and OpenCL::Image3D.
1255 1425
1256=over 4 1426=over 4
1257 1427
1258=item $packed_value = $ev->image_info ($name) 1428=item $packed_value = $image->image_info ($name)
1259 1429
1260See C<< $platform->info >> for details. 1430See C<< $platform->info >> for details.
1261 1431
1262The reason this method is not called C<info> is that there already is an 1432The reason this method is not called C<info> is that there already is an
1263C<< ->info >> method inherited from C<OpenCL::Memory>. 1433C<< ->info >> method inherited from C<OpenCL::Memory>.
1264 1434
1265L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html> 1435L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html>
1266 1436
1437=item ($channel_order, $channel_data_type) = $image->format
1438
1439Returns the channel order and type used to create the image by calling
1440C<clGetImageInfo> with C<CL_IMAGE_FORMAT>.
1441
1267=for gengetinfo begin image 1442=for gengetinfo begin image
1268 1443
1269=item $int = $image->element_size 1444=item $int = $image->element_size
1270 1445
1271Calls C<clGetImageInfo> with C<CL_IMAGE_ELEMENT_SIZE> and returns the result. 1446Calls C<clGetImageInfo> with C<CL_IMAGE_ELEMENT_SIZE> and returns the result.
1344 1519
1345=head2 THE OpenCL::Program CLASS 1520=head2 THE OpenCL::Program CLASS
1346 1521
1347=over 4 1522=over 4
1348 1523
1349=item $program->build ($device, $options = "") 1524=item $program->build (\@devices = undef, $options = "", $cb->($program) = undef)
1350 1525
1351Tries to build the program with the given options. See also the 1526Tries to build the program with the given options. See also the
1352C<$ctx->build> convenience function. 1527C<$ctx->build> convenience function.
1353 1528
1529If a callback is specified, then it will be called when compilation is
1530finished. Note that many OpenCL implementations block your program while
1531compiling whether you use a callback or not. See C<build_async> if you
1532want to make sure the build is done in the background.
1533
1534Note that some OpenCL implementations act up badly, and don't call the
1535callback in some error cases (but call it in others). This implementation
1536assumes the callback will always be called, and leaks memory if this is
1537not so. So best make sure you don't pass in invalid values.
1538
1539Some implementations fail with C<OpenCL::INVALID_BINARY> when the
1540compilation state is successful but some later stage fails.
1541
1354L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html> 1542L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html>
1543
1544=item $program->build_async (\@devices = undef, $options = "", $cb->($program) = undef)
1545
1546Similar to C<< ->build >>, except it starts a thread, and never fails (you
1547need to check the compilation status form the callback, or by polling).
1355 1548
1356=item $packed_value = $program->build_info ($device, $name) 1549=item $packed_value = $program->build_info ($device, $name)
1357 1550
1358Similar to C<< $platform->info >>, but returns build info for a previous 1551Similar to C<< $platform->info >>, but returns build info for a previous
1359build attempt for the given device. 1552build attempt for the given device.
1501 1694
1502Calls C<clGetKernelWorkGroupInfo> with C<CL_KERNEL_PRIVATE_MEM_SIZE> and returns the result. 1695Calls C<clGetKernelWorkGroupInfo> with C<CL_KERNEL_PRIVATE_MEM_SIZE> and returns the result.
1503 1696
1504=for gengetinfo end kernel_work_group 1697=for gengetinfo end kernel_work_group
1505 1698
1699=item $kernel->setf ($format, ...)
1700
1701Sets the arguments of a kernel. Since OpenCL 1.1 doesn't have a generic
1702way to set arguments (and with OpenCL 1.2 it might be rather slow), you
1703need to specify a format argument, much as with C<printf>, to tell OpenCL
1704what type of argument it is.
1705
1706The format arguments are single letters:
1707
1708 c char
1709 C unsigned char
1710 s short
1711 S unsigned short
1712 i int
1713 I unsigned int
1714 l long
1715 L unsigned long
1716
1717 h half float (0..65535)
1718 f float
1719 d double
1720
1721 z local (octet size)
1722
1723 m memory object (buffer or image)
1724 a sampler
1725 e event
1726
1727Space characters in the format string are ignored.
1728
1729Example: set the arguments for a kernel that expects an int, two floats, a buffer and an image.
1730
1731 $kernel->setf ("i ff mm", 5, 0.5, 3, $buffer, $image);
1732
1506=item $kernel->set_TYPE ($index, $value) 1733=item $kernel->set_TYPE ($index, $value)
1507 1734
1735=item $kernel->set_char ($index, $value)
1736
1737=item $kernel->set_uchar ($index, $value)
1738
1739=item $kernel->set_short ($index, $value)
1740
1741=item $kernel->set_ushort ($index, $value)
1742
1743=item $kernel->set_int ($index, $value)
1744
1745=item $kernel->set_uint ($index, $value)
1746
1747=item $kernel->set_long ($index, $value)
1748
1749=item $kernel->set_ulong ($index, $value)
1750
1751=item $kernel->set_half ($index, $value)
1752
1753=item $kernel->set_float ($index, $value)
1754
1755=item $kernel->set_double ($index, $value)
1756
1757=item $kernel->set_memory ($index, $value)
1758
1759=item $kernel->set_buffer ($index, $value)
1760
1761=item $kernel->set_image ($index, $value)
1762
1763=item $kernel->set_sampler ($index, $value)
1764
1765=item $kernel->set_local ($index, $value)
1766
1767=item $kernel->set_event ($index, $value)
1768
1508This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>. 1769This is a family of methods to set the kernel argument with the number
1509 1770C<$index> to the give C<$value>.
1510TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>,
1511C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>,
1512C<image2d>, C<image3d>, C<sampler>, C<local> or C<event>.
1513 1771
1514Chars and integers (including the half type) are specified as integers, 1772Chars and integers (including the half type) are specified as integers,
1515float and double as floating point values, memory/buffer/image2d/image3d 1773float and double as floating point values, memory/buffer/image must be
1516must be an object of that type or C<undef>, local-memory arguments are 1774an object of that type or C<undef>, local-memory arguments are set by
1517set by specifying the size, and sampler and event must be objects of that 1775specifying the size, and sampler and event must be objects of that type.
1518type. 1776
1777Note that C<set_memory> works for all memory objects (all types of buffers
1778and images) - the main purpose of the more specific C<set_TYPE> functions
1779is type checking.
1519 1780
1520Setting an argument for a kernel does NOT keep a reference to the object - 1781Setting an argument for a kernel does NOT keep a reference to the object -
1521for example, if you set an argument to some image object, free the image, 1782for example, if you set an argument to some image object, free the image,
1522and call the kernel, you will run into undefined behaviour. 1783and call the kernel, you will run into undefined behaviour.
1523 1784
1536 1797
1537Waits for the event to complete. 1798Waits for the event to complete.
1538 1799
1539L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 1800L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
1540 1801
1802=item $ev->cb ($exec_callback_type, $callback->($event, $event_command_exec_status))
1803
1804Adds a callback to the callback stack for the given event type. There is
1805no way to remove a callback again.
1806
1807L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetEventCallback.html>
1808
1541=item $packed_value = $ev->info ($name) 1809=item $packed_value = $ev->info ($name)
1542 1810
1543See C<< $platform->info >> for details. 1811See C<< $platform->info >> for details.
1544 1812
1545L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html> 1813L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html>
1605 1873
1606=over 4 1874=over 4
1607 1875
1608=item $ev->set_status ($execution_status) 1876=item $ev->set_status ($execution_status)
1609 1877
1878Sets the execution status of the user event. Can only be called once,
1879either with OpenCL::COMPLETE or a negative number as status.
1880
1610L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html> 1881L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html>
1611 1882
1612=back 1883=back
1613 1884
1614=cut 1885=cut

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines