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

Comparing OpenCL/OpenCL.pm (file contents):
Revision 1.38 by root, Thu Apr 19 19:36:24 2012 UTC vs.
Revision 1.56 by root, Tue Apr 24 23:53:12 2012 UTC

157 $id = get_global_id (0); 157 $id = get_global_id (0);
158 output [id] = input [id] * input [id]; 158 output [id] = input [id] * input [id];
159 } 159 }
160 '; 160 ';
161 161
162 my $prog = $ctx->program_with_source ($src); 162 my $prog = $ctx->build_program ($src);
163
164 # build croaks on compile errors, so catch it and print the compile errors
165 eval { $prog->build ($dev); 1 }
166 or die $prog->build_log;
167
168 my $kernel = $prog->kernel ("squareit"); 163 my $kernel = $prog->kernel ("squareit");
169 164
170=head2 Create some input and output float buffers, then call the 165=head2 Create some input and output float buffers, then call the
171'squareit' kernel on them. 166'squareit' kernel on them.
172 167
260 m.x = fabs (fmod (m.x + time * 0.05f, 4.f)) - 2.f; 255 m.x = fabs (fmod (m.x + time * 0.05f, 4.f)) - 2.f;
261 256
262 float2 z = m; 257 float2 z = m;
263 float2 c = (float2)(sin (time * 0.05005), cos (time * 0.06001)); 258 float2 c = (float2)(sin (time * 0.05005), cos (time * 0.06001));
264 259
265 for (int i = 0; i < 100 && dot (z, z) < 4.f; ++i) 260 for (int i = 0; i < 25 && dot (z, z) < 4.f; ++i)
266 z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c; 261 z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c;
267 262
268 float3 colour = (float3)(z.x, z.y, z.x * z.y); 263 float3 colour = (float3)(z.x, z.y, z.x * z.y);
269 write_imagef (img, (int2)(get_global_id (0), get_global_id (1)), (float4)(colour * p.x * p.x, 1.)); 264 write_imagef (img, (int2)(get_global_id (0), get_global_id (1)), (float4)(colour * p.x * p.x, 1.));
270 } 265 }
271 EOF 266 EOF
267
272 my $prog = $ctx->program_with_source ($src); 268 my $prog = $ctx->build_program ($src);
273 eval { $prog->build ($dev); 1 }
274 or die $prog->build_log ($dev);
275
276 my $kernel = $prog->kernel ("juliatunnel"); 269 my $kernel = $prog->kernel ("juliatunnel");
277 270
278 # program compiled, kernel ready, now draw and loop 271 # program compiled, kernel ready, now draw and loop
279 272
280 for (my $time; ; ++$time) { 273 for (my $time; ; ++$time) {
288 281
289 # release objects to opengl again 282 # release objects to opengl again
290 $queue->enqueue_release_gl_objects ([$tex]); 283 $queue->enqueue_release_gl_objects ([$tex]);
291 284
292 # wait 285 # wait
293 $queue->flush; 286 $queue->finish;
294 287
295 # now draw the texture, the defaults should be all right 288 # now draw the texture, the defaults should be all right
296 glTexParameterf GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST; 289 glTexParameterf GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST;
297 290
298 glEnable GL_TEXTURE_2D; 291 glEnable GL_TEXTURE_2D;
336=item * Structures are often specified by flattening out their components 329=item * Structures are often specified by flattening out their components
337as with short vectors, and returned as arrayrefs. 330as with short vectors, and returned as arrayrefs.
338 331
339=item * When enqueuing commands, the wait list is specified by adding 332=item * When enqueuing commands, the wait list is specified by adding
340extra arguments to the function - anywhere a C<$wait_events...> argument 333extra arguments to the function - anywhere a C<$wait_events...> argument
341is documented this can be any number of event objects. 334is documented this can be any number of event objects. As an extsnion
335implemented by this module, C<undef> values will be ignored in the event
336list.
342 337
343=item * When enqueuing commands, if the enqueue method is called in void 338=item * When enqueuing commands, if the enqueue method is called in void
344context, no event is created. In all other contexts an event is returned 339context, no event is created. In all other contexts an event is returned
345by the method. 340by the method.
346 341
381 376
382For this to work, the OpenGL library must be loaded, a GLX context must 377For this to work, the OpenGL library must be loaded, a GLX context must
383have been created and be made current, and C<dlsym> must be available and 378have been created and be made current, and C<dlsym> must be available and
384capable of finding the function via C<RTLD_DEFAULT>. 379capable of finding the function via C<RTLD_DEFAULT>.
385 380
381=head2 EVENT SYSTEM
382
383OpenCL can generate a number of (potentially) asynchronous events, for
384example, after compiling a program, to signal a context-related error or,
385perhaps most important, to signal completion of queued jobs (by setting
386callbacks on OpenCL::Event objects).
387
388To facilitate this, this module maintains an event queue - each
389time an asynchronous event happens, it is queued, and perl will be
390interrupted. This is implemented via the L<Async::Interrupt> module. In
391addition, this module has L<AnyEvent> support, so it can seamlessly
392integrate itself into many event loops.
393
394Since this module is a bit hard to understand, here are some case examples:
395
396=head3 Don't use callbacks.
397
398When your program never uses any callbacks, then there will never be any
399notifications you need to take care of, and therefore no need to worry
400about all this.
401
402You can achieve a great deal by explicitly waiting for events, or using
403barriers and flush calls. In many programs, there is no need at all to
404tinker with asynchronous events.
405
406=head3 Use AnyEvent
407
408This module automatically registers a watcher that invokes all outstanding
409event callbacks when AnyEvent is initialised (and block asynchronous
410interruptions). Using this mode of operations is the safest and most
411recommended one.
412
413To use this, simply use AnyEvent and this module normally, make sure you
414have an event loop running:
415
416 use Gtk2 -init;
417 use AnyEvent;
418
419 # initialise AnyEvent, by creating a watcher, or:
420 AnyEvent::detect;
421
422 my $e = $queue->enqueue_marker;
423 $e->cb (sub {
424 warn "opencl is finished\n";
425 })
426
427 main Gtk2;
428
429Note that this module will not initialise AnyEvent for you. Before
430AnyEvent is initialised, the module will asynchronously interrupt perl
431instead. To avoid any surprises, it's best to explicitly initialise
432AnyEvent.
433
434You can temporarily enable asynchronous interruptions (see next paragraph)
435by calling C<$OpenCL::INTERRUPT->unblock> and disable them again by
436calling C<$OpenCL::INTERRUPT->block>.
437
438=head3 Let yourself be interrupted at any time
439
440This mode is the default unless AnyEvent is loaded and initialised. In
441this mode, OpenCL asynchronously interrupts a running perl program. The
442emphasis is on both I<asynchronously> and I<running> here.
443
444Asynchronously means that perl might execute your callbacks at any
445time. For example, in the following code (I<THAT YOU SHOULD NOT COPY>),
446the C<until> loop following the marker call will be interrupted by the
447callback:
448
449 my $e = $queue->enqueue_marker;
450 my $flag;
451 $e->cb (sub { $flag = 1 });
452 1 until $flag;
453 # $flag is now 1
454
455The reason why you shouldn't blindly copy the above code is that
456busy waiting is a really really bad thing, and really really bad for
457performance.
458
459While at first this asynchronous business might look exciting, it can be
460really hard, because you need to be prepared for the callback code to be
461executed at any time, which limits the amount of things the callback code
462can do safely.
463
464This can be mitigated somewhat by using C<<
465$OpenCL::INTERRUPT->scope_block >> (see the L<Async::Interrupt>
466documentation for details).
467
468The other problem is that your program must be actively I<running> to be
469interrupted. When you calculate stuff, your program is running. When you
470hang in some C functions or other block execution (by calling C<sleep>,
471C<select>, running an event loop and so on), your program is waiting, not
472running.
473
474One way around that would be to attach a read watcher to your event loop,
475listening for events on C<< $OpenCL::INTERRUPT->pipe_fileno >>, using a
476dummy callback (C<sub { }>) to temporarily execute some perl code.
477
478That is then awfully close to using the built-in AnyEvent support above,
479though, so consider that one instead.
480
481=head3 Be creative
482
483OpenCL exports the L<Async::Interrupt> object it uses in the global
484variable C<$OpenCL::INTERRUPT>. You can configure it in any way you like.
485
486So if you want to feel like a real pro, err, wait, if you feel no risk
487menas no fun, you can experiment by implementing your own mode of
488operations.
489
490=cut
491
492package OpenCL;
493
494use common::sense;
495use Async::Interrupt ();
496
497our $POLL_FUNC; # set by XS
498
499BEGIN {
500 our $VERSION = '0.97';
501
502 require XSLoader;
503 XSLoader::load (__PACKAGE__, $VERSION);
504
505 @OpenCL::Platform::ISA =
506 @OpenCL::Device::ISA =
507 @OpenCL::Context::ISA =
508 @OpenCL::Queue::ISA =
509 @OpenCL::Memory::ISA =
510 @OpenCL::Sampler::ISA =
511 @OpenCL::Program::ISA =
512 @OpenCL::Kernel::ISA =
513 @OpenCL::Event::ISA = OpenCL::Object::;
514
515 @OpenCL::Buffer::ISA =
516 @OpenCL::Image::ISA = OpenCL::Memory::;
517
518 @OpenCL::BufferObj::ISA = OpenCL::Buffer::;
519
520 @OpenCL::Image2D::ISA =
521 @OpenCL::Image3D::ISA =
522 @OpenCL::Image2DArray::ISA =
523 @OpenCL::Image1D::ISA =
524 @OpenCL::Image1DArray::ISA =
525 @OpenCL::Image1DBuffer::ISA = OpenCL::Image::;
526
527 @OpenCL::UserEvent::ISA = OpenCL::Event::;
528}
529
386=head2 THE OpenCL PACKAGE 530=head2 THE OpenCL PACKAGE
387 531
388=over 4 532=over 4
389 533
390=item $int = OpenCL::errno 534=item $int = OpenCL::errno
408 552
409Returns all available OpenCL::Platform objects. 553Returns all available OpenCL::Platform objects.
410 554
411L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html> 555L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html>
412 556
413=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $notify = undef 557=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr
414 558
415Tries to create a context from a default device and platform - never worked for me. 559Tries to create a context from a default device and platform type - never worked for me.
416 560
417L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 561L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
418 562
563=item $ctx = OpenCL::context $properties, \@devices, $callback->($err, $pvt) = $print_stderr)
564
565Create a new OpenCL::Context object using the given device object(s). This
566function isn't implemented yet, use C<< $platform->context >> instead.
567
568L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
569
419=item OpenCL::wait_for_events $wait_events... 570=item OpenCL::wait_for_events $wait_events...
420 571
421Waits for all events to complete. 572Waits for all events to complete.
422 573
423L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 574L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
424 575
576=item OpenCL::poll
577
578Checks if there are any outstanding events (see L<EVENT SYSTEM>) and
579invokes their callbacks.
580
581=item $OpenCL::INTERRUPT
582
583The L<Async::Interrupt> object used to signal asynchronous events (see
584L<EVENT SYSTEM>).
585
586=cut
587
588our $INTERRUPT = new Async::Interrupt c_cb => [$POLL_FUNC, 0];
589
590&_eq_initialise ($INTERRUPT->signal_func);
591
592=item $OpenCL::WATCHER
593
594The L<AnyEvent> watcher object used to watch for asynchronous events (see
595L<EVENT SYSTEM>). This variable is C<undef> until L<AnyEvent> has been
596loaded I<and> initialised (e.g. by calling C<AnyEvent::detect>).
597
598=cut
599
600our $WATCHER;
601
602sub _init_anyevent {
603 $INTERRUPT->block;
604 $WATCHER = AE::io ($INTERRUPT->pipe_fileno, 0, sub { $INTERRUPT->handle });
605}
606
607if (defined $AnyEvent::MODEL) {
608 _init_anyevent;
609} else {
610 push @AnyEvent::post_detect, \&_init_anyevent;
611}
612
425=back 613=back
426 614
615=head2 THE OpenCL::Object CLASS
616
617This is the base class for all objects in the OpenCL module. The only
618method it implements is the C<id> method, which is only useful if you want
619to interface to OpenCL on the C level.
620
621=over 4
622
623=item $iv = $obj->id
624
625OpenCL objects are represented by pointers or integers on the C level. If
626you want to interface to an OpenCL object directly on the C level, then
627you need this value, which is returned by this method. You should use an
628C<IV> type in your code and cast that to the correct type.
629
630=cut
631
632sub OpenCL::Object::id {
633 ref $_[0] eq "SCALAR"
634 ? ${ $_[0] }
635 : $_[0][0]
636}
637
638=back
639
427=head2 THE OpenCL::Platform CLASS 640=head2 THE OpenCL::Platform CLASS
428 641
429=over 4 642=over 4
430 643
431=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL) 644=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL)
432 645
433Returns a list of matching OpenCL::Device objects. 646Returns a list of matching OpenCL::Device objects.
434 647
435=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $notify = undef) 648=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr)
436 649
437Tries to create a context. Never worked for me, and you need devices explicitly anyway. 650Tries to create a context. Never worked for me, and you need devices explicitly anyway.
438 651
439L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 652L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
440 653
441=item $ctx = $platform->context ($properties = undef, @$devices, $notify = undef) 654=item $ctx = $platform->context ($properties, \@devices, $callback->($err, $pvt) = $print_stderr)
442 655
443Create a new OpenCL::Context object using the given device object(s)- a 656Create a new OpenCL::Context object using the given device object(s)- a
444CL_CONTEXT_PLATFORM property is supplied automatically. 657CL_CONTEXT_PLATFORM property is supplied automatically.
445 658
446L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html> 659L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
454It's best to avoid this method and use one of the following convenience 667It's best to avoid this method and use one of the following convenience
455wrappers. 668wrappers.
456 669
457L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html> 670L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html>
458 671
672=item $platform->unload_compiler
673
674Attempts to unload the compiler for this platform, for endless
675profit. Does nothing on OpenCL 1.1.
676
677L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clUnloadPlatformCompiler.html>
678
459=for gengetinfo begin platform 679=for gengetinfo begin platform
460 680
461=item $string = $platform->profile 681=item $string = $platform->profile
462 682
463Calls C<clGetPlatformInfo> with C<CL_PLATFORM_PROFILE> and returns the result. 683Calls C<clGetPlatformInfo> with C<CL_PLATFORM_PROFILE> and returns the result.
748 968
749=item @device_partition_property_exts = $device->affinity_domains_ext 969=item @device_partition_property_exts = $device->affinity_domains_ext
750 970
751Calls C<clGetDeviceInfo> with C<CL_DEVICE_AFFINITY_DOMAINS_EXT> and returns the result. 971Calls C<clGetDeviceInfo> with C<CL_DEVICE_AFFINITY_DOMAINS_EXT> and returns the result.
752 972
753=item $uint = $device->reference_count_ext 973=item $uint = $device->reference_count_ext
754 974
755Calls C<clGetDeviceInfo> with C<CL_DEVICE_REFERENCE_COUNT_EXT > and returns the result. 975Calls C<clGetDeviceInfo> with C<CL_DEVICE_REFERENCE_COUNT_EXT> and returns the result.
756 976
757=item @device_partition_property_exts = $device->partition_style_ext 977=item @device_partition_property_exts = $device->partition_style_ext
758 978
759Calls C<clGetDeviceInfo> with C<CL_DEVICE_PARTITION_STYLE_EXT> and returns the result. 979Calls C<clGetDeviceInfo> with C<CL_DEVICE_PARTITION_STYLE_EXT> and returns the result.
760 980
764 984
765=head2 THE OpenCL::Context CLASS 985=head2 THE OpenCL::Context CLASS
766 986
767=over 4 987=over 4
768 988
989=item $prog = $ctx->build_program ($program, $options = "")
990
991This convenience function tries to build the program on all devices in
992the context. If the build fails, then the function will C<croak> with the
993build log. Otherwise ti returns the program object.
994
995The C<$program> can either be a C<OpenCL::Program> object or a string
996containing the program. In the latter case, a program objetc will be
997created automatically.
998
999=cut
1000
1001sub OpenCL::Context::build_program {
1002 my ($self, $prog, $options) = @_;
1003
1004 require Carp;
1005
1006 $prog = $self->program_with_source ($prog)
1007 unless ref $prog;
1008
1009 # we build separately per device so we instantly know which one failed
1010 for my $dev ($self->devices) {
1011 eval { $prog->build ([$dev], $options); 1 }
1012 or Carp::croak ("Building OpenCL program for device '" . $dev->name . "' failed:\n"
1013 . $prog->build_log ($dev));
1014 }
1015
1016 $prog
1017}
1018
769=item $queue = $ctx->queue ($device, $properties) 1019=item $queue = $ctx->queue ($device, $properties)
770 1020
771Create a new OpenCL::Queue object from the context and the given device. 1021Create a new OpenCL::Queue object from the context and the given device.
772 1022
773L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html> 1023L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html>
1024
1025Example: create an out-of-order queue.
1026
1027 $queue = $ctx->queue ($device, OpenCL::QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE);
774 1028
775=item $ev = $ctx->user_event 1029=item $ev = $ctx->user_event
776 1030
777Creates a new OpenCL::UserEvent object. 1031Creates a new OpenCL::UserEvent object.
778 1032
788=item $buf = $ctx->buffer_sv ($flags, $data) 1042=item $buf = $ctx->buffer_sv ($flags, $data)
789 1043
790Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and 1044Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and
791initialise it with the given data values. 1045initialise it with the given data values.
792 1046
1047=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)
1048
1049Creates a new OpenCL::Image object and optionally initialises it with
1050the given data values.
1051
1052L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateImage.html>
1053
793=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $row_pitch = 0, $data = undef) 1054=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $row_pitch = 0, $data = undef)
794 1055
795Creates a new OpenCL::Image2D object and optionally initialises it with 1056Creates a new OpenCL::Image2D object and optionally initialises it with
796the given data values. 1057the given data values.
797 1058
809Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object that refers to the given 1070Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object that refers to the given
810OpenGL buffer object. 1071OpenGL buffer object.
811 1072
812http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLBuffer.html 1073http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLBuffer.html
813 1074
1075=item $img = $ctx->gl_texture ($flags, $target, $miplevel, $texture)
1076
1077Creates a new OpenCL::Image object that refers to the given OpenGL
1078texture object or buffer.
1079
1080http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateFromGLTexture.html
1081
814=item $ctx->gl_texture2d ($flags, $target, $miplevel, $texture) 1082=item $img = $ctx->gl_texture2d ($flags, $target, $miplevel, $texture)
815 1083
816Creates a new OpenCL::Image2D object that refers to the given OpenGL 1084Creates a new OpenCL::Image2D object that refers to the given OpenGL
8172D texture object. 10852D texture object.
818 1086
819http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture2D.html 1087http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture2D.html
820 1088
821=item $ctx->gl_texture3d ($flags, $target, $miplevel, $texture) 1089=item $img = $ctx->gl_texture3d ($flags, $target, $miplevel, $texture)
822 1090
823Creates a new OpenCL::Image3D object that refers to the given OpenGL 1091Creates a new OpenCL::Image3D object that refers to the given OpenGL
8243D texture object. 10923D texture object.
825 1093
826http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture3D.html 1094http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture3D.html
889for completion, unless the method is called in void context, in which case 1157for completion, unless the method is called in void context, in which case
890no event object is created. 1158no event object is created.
891 1159
892They also allow you to specify any number of other event objects that this 1160They also allow you to specify any number of other event objects that this
893request has to wait for before it starts executing, by simply passing the 1161request has to wait for before it starts executing, by simply passing the
894event objects as extra parameters to the enqueue methods. 1162event objects as extra parameters to the enqueue methods. To simplify
1163program design, this module ignores any C<undef> values in the list of
1164events. This makes it possible to code operations such as this, without
1165having to put a valid event object into C<$event> first:
1166
1167 $event = $queue->enqueue_xxx (..., $event);
895 1168
896Queues execute in-order by default, without any parallelism, so in most 1169Queues execute in-order by default, without any parallelism, so in most
897cases (i.e. you use only one queue) it's not necessary to wait for or 1170cases (i.e. you use only one queue) it's not necessary to wait for or
898create event objects. 1171create event objects, althoguh an our of order queue is often a bit
1172faster.
899 1173
900=over 4 1174=over 4
901 1175
902=item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...) 1176=item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...)
903 1177
946=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...) 1220=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...)
947 1221
948Yeah. 1222Yeah.
949 1223
950L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>. 1224L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>.
1225
1226=item $ev = $queue->enqueue_fill_buffer ($mem, $pattern, $offset, $size, ...)
1227
1228Fills the given buffer object with repeated applications of C<$pattern>,
1229starting at C<$offset> for C<$size> octets.
1230
1231L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillBuffer.html>
1232
1233=item $ev = $queue->enqueue_fill_image ($img, $r, $g, $b, $a, $x, $y, $z, $width, $height, $depth, ...)
1234
1235Fills the given image area with the given rgba colour components. The
1236components are normally floating point values between C<0> and C<1>,
1237except when the image channel data type is a signe dor unsigned
1238unnormalised format, in which case the range is determined by the format.
1239
1240L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillImage.html>
951 1241
952=item $ev = $queue->enqueue_task ($kernel, $wait_events...) 1242=item $ev = $queue->enqueue_task ($kernel, $wait_events...)
953 1243
954L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html> 1244L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html>
955 1245
969reference to an array of local work sizes, with the same number of 1259reference to an array of local work sizes, with the same number of
970elements as @$global_work_size. 1260elements as @$global_work_size.
971 1261
972L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html> 1262L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html>
973 1263
974=item $ev = $queue->enqueue_marker ($wait_events...)
975
976L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMarker.html>
977
978=item $ev = $queue->enqueue_acquire_gl_objects ([object, ...], $wait_events...) 1264=item $ev = $queue->enqueue_acquire_gl_objects ([object, ...], $wait_events...)
979 1265
980Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired 1266Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired
981for subsequent OpenCL usage. 1267for subsequent OpenCL usage.
982 1268
991 1277
992=item $ev = $queue->enqueue_wait_for_events ($wait_events...) 1278=item $ev = $queue->enqueue_wait_for_events ($wait_events...)
993 1279
994L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html> 1280L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html>
995 1281
996=item $queue->enqueue_barrier 1282=item $ev = $queue->enqueue_marker ($wait_events...)
997 1283
1284L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueMarkerWithWaitList.html>
1285
1286=item $ev = $queue->enqueue_barrier ($wait_events...)
1287
998L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueBarrier.html> 1288L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueBarrierWithWaitList.html>
999 1289
1000=item $queue->flush 1290=item $queue->flush
1001 1291
1002L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html> 1292L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html>
1003 1293
1118 1408
1119=back 1409=back
1120 1410
1121=head2 THE OpenCL::Image CLASS 1411=head2 THE OpenCL::Image CLASS
1122 1412
1123This is the superclass of all image objects - OpenCL::Image2D and OpenCL::Image3D. 1413This is the superclass of all image objects - OpenCL::Image1D,
1414OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D,
1415OpenCL::Image2DArray and OpenCL::Image3D.
1124 1416
1125=over 4 1417=over 4
1126 1418
1127=item $packed_value = $ev->image_info ($name) 1419=item $packed_value = $image->image_info ($name)
1128 1420
1129See C<< $platform->info >> for details. 1421See C<< $platform->info >> for details.
1130 1422
1131The reason this method is not called C<info> is that there already is an 1423The reason this method is not called C<info> is that there already is an
1132C<< ->info >> method inherited from C<OpenCL::Memory>. 1424C<< ->info >> method inherited from C<OpenCL::Memory>.
1133 1425
1134L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html> 1426L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html>
1135 1427
1428=item ($channel_order, $channel_data_type) = $image->format
1429
1430Returns the channel order and type used to create the image by calling
1431C<clGetImageInfo> with C<CL_IMAGE_FORMAT>.
1432
1136=for gengetinfo begin image 1433=for gengetinfo begin image
1137 1434
1138=item $int = $image->element_size 1435=item $int = $image->element_size
1139 1436
1140Calls C<clGetImageInfo> with C<CL_IMAGE_ELEMENT_SIZE> and returns the result. 1437Calls C<clGetImageInfo> with C<CL_IMAGE_ELEMENT_SIZE> and returns the result.
1213 1510
1214=head2 THE OpenCL::Program CLASS 1511=head2 THE OpenCL::Program CLASS
1215 1512
1216=over 4 1513=over 4
1217 1514
1218=item $program->build ($device, $options = "") 1515=item $program->build (\@devices = undef, $options = "", $cb->($program) = undef)
1219 1516
1220Tries to build the program with the givne options. 1517Tries to build the program with the given options. See also the
1518C<$ctx->build> convenience function.
1519
1520If a callback is specified, then it will be called when compilation is
1521finished. Note that many OpenCL implementations block your program while
1522compiling whether you use a callback or not. See C<build_async> if you
1523want to make sure the build is done in the background.
1524
1525Note that some OpenCL implementations atc up badly, and don't call the
1526callback in some error cases (but call it in others). This implementation
1527assumes the callback will always be called, and leaks memory if this is
1528not so. So best make sure you don't pass in invalid values.
1221 1529
1222L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html> 1530L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html>
1531
1532=item $program->build_async (\@devices = undef, $options = "", $cb->($program) = undef)
1533
1534Similar to C<< ->build >>, except it starts a thread, and never fails (you
1535need to check the compilation status form the callback, or by polling).
1223 1536
1224=item $packed_value = $program->build_info ($device, $name) 1537=item $packed_value = $program->build_info ($device, $name)
1225 1538
1226Similar to C<< $platform->info >>, but returns build info for a previous 1539Similar to C<< $platform->info >>, but returns build info for a previous
1227build attempt for the given device. 1540build attempt for the given device.
1232 1545
1233Creates an OpenCL::Kernel object out of the named C<__kernel> function in 1546Creates an OpenCL::Kernel object out of the named C<__kernel> function in
1234the program. 1547the program.
1235 1548
1236L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html> 1549L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html>
1550
1551=item @kernels = $program->kernels_in_program
1552
1553Returns all kernels successfully compiled for all devices in program.
1554
1555http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernelsInProgram.html
1237 1556
1238=for gengetinfo begin program_build 1557=for gengetinfo begin program_build
1239 1558
1240=item $build_status = $program->build_status ($device) 1559=item $build_status = $program->build_status ($device)
1241 1560
1369 1688
1370This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>. 1689This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>.
1371 1690
1372TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>, 1691TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>,
1373C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>, 1692C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>,
1374C<image2d>, C<image3d>, C<sampler> or C<event>. 1693C<image2d>, C<image3d>, C<sampler>, C<local> or C<event>.
1375 1694
1376Chars and integers (including the half type) are specified as integers, 1695Chars and integers (including the half type) are specified as integers,
1377float and double as floating point values, memory/buffer/image2d/image3d 1696float and double as floating point values, memory/buffer/image2d/image3d
1378must be an object of that type or C<undef>, and sampler and event must be 1697must be an object of that type or C<undef>, local-memory arguments are
1379objects of that type. 1698set by specifying the size, and sampler and event must be objects of that
1699type.
1700
1701Setting an argument for a kernel does NOT keep a reference to the object -
1702for example, if you set an argument to some image object, free the image,
1703and call the kernel, you will run into undefined behaviour.
1380 1704
1381L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html> 1705L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html>
1382 1706
1383=back 1707=back
1384 1708
1393 1717
1394Waits for the event to complete. 1718Waits for the event to complete.
1395 1719
1396L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 1720L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
1397 1721
1722=item $ev->cb ($exec_callback_type, $callback->($event, $event_command_exec_status))
1723
1724Adds a callback to the callback stack for the given event type. There is
1725no way to remove a callback again.
1726
1727L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetEventCallback.html>
1728
1398=item $packed_value = $ev->info ($name) 1729=item $packed_value = $ev->info ($name)
1399 1730
1400See C<< $platform->info >> for details. 1731See C<< $platform->info >> for details.
1401 1732
1402L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html> 1733L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html>
1462 1793
1463=over 4 1794=over 4
1464 1795
1465=item $ev->set_status ($execution_status) 1796=item $ev->set_status ($execution_status)
1466 1797
1798Sets the execution status of the user event. Can only be called once,
1799either with OpenCL::COMPLETE or a negative number as status.
1800
1467L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html> 1801L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html>
1468 1802
1469=back 1803=back
1470 1804
1471=cut 1805=cut
1472
1473package OpenCL;
1474
1475use common::sense;
1476
1477BEGIN {
1478 our $VERSION = '0.95';
1479
1480 require XSLoader;
1481 XSLoader::load (__PACKAGE__, $VERSION);
1482
1483 @OpenCL::Buffer::ISA =
1484 @OpenCL::Image::ISA = OpenCL::Memory::;
1485
1486 @OpenCL::BufferObj::ISA = OpenCL::Buffer::;
1487
1488 @OpenCL::Image2D::ISA =
1489 @OpenCL::Image3D::ISA = OpenCL::Image::;
1490
1491 @OpenCL::UserEvent::ISA = OpenCL::Event::;
1492}
1493 1806
14941; 18071;
1495 1808
1496=head1 AUTHOR 1809=head1 AUTHOR
1497 1810

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines