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

Comparing OpenCL/OpenCL.pm (file contents):
Revision 1.8 by root, Wed Nov 16 06:22:20 2011 UTC vs.
Revision 1.64 by root, Mon Apr 30 23:12:43 2012 UTC

8 8
9=head1 DESCRIPTION 9=head1 DESCRIPTION
10 10
11This is an early release which might be useful, but hasn't seen much testing. 11This is an early release which might be useful, but hasn't seen much testing.
12 12
13=head2 OpenCL FROM 10000 FEET HEIGHT
14
15Here is a high level overview of OpenCL:
16
17First you need to find one or more OpenCL::Platforms (kind of like
18vendors) - usually there is only one.
19
20Each platform gives you access to a number of OpenCL::Device objects, e.g.
21your graphics card.
22
23From a platform and some device(s), you create an OpenCL::Context, which is
24a very central object in OpenCL: Once you have a context you can create
25most other objects:
26
27OpenCL::Program objects, which store source code and, after building for a
28specific device ("compiling and linking"), also binary programs. For each
29kernel function in a program you can then create an OpenCL::Kernel object
30which represents basically a function call with argument values.
31
32OpenCL::Memory objects of various flavours: OpenCL::Buffer objects (flat
33memory areas, think arrays or structs) and OpenCL::Image objects (think 2d
34or 3d array) for bulk data and input and output for kernels.
35
36OpenCL::Sampler objects, which are kind of like texture filter modes in
37OpenGL.
38
39OpenCL::Queue objects - command queues, which allow you to submit memory
40reads, writes and copies, as well as kernel calls to your devices. They
41also offer a variety of methods to synchronise request execution, for
42example with barriers or OpenCL::Event objects.
43
44OpenCL::Event objects are used to signal when something is complete.
45
13=head1 HELPFUL RESOURCES 46=head2 HELPFUL RESOURCES
14 47
15The OpenCL spec used to develop this module (1.2 spec was available, but 48The OpenCL specs used to develop this module:
16no implementation was available to me :).
17 49
18 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
19 53
20OpenCL manpages: 54OpenCL manpages:
21 55
22 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/
58
59If you are into UML class diagrams, the following diagram might help - if
60not, it will be mildly confusing (also, the class hierarchy of this module
61is much more fine-grained):
62
63 http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/classDiagram.html
64
65Here's a tutorial from AMD (very AMD-centric, too), not sure how useful it
66is, but at least it's free of charge:
67
68 http://developer.amd.com/zones/OpenCLZone/courses/Documents/Introduction_to_OpenCL_Programming%20Training_Guide%20%28201005%29.pdf
69
70And here's NVIDIA's OpenCL Best Practises Guide:
71
72 http://developer.download.nvidia.com/compute/cuda/3_2/toolkit/docs/OpenCL_Best_Practices_Guide.pdf
73
74=head1 BASIC WORKFLOW
75
76To get something done, you basically have to do this once (refer to the
77examples below for actual code, this is just a high-level description):
78
79Find some platform (e.g. the first one) and some device(s) (e.g. the first
80device of the platform), and create a context from those.
81
82Create program objects from your OpenCL source code, then build (compile)
83the programs for each device you want to run them on.
84
85Create kernel objects for all kernels you want to use (surprisingly, these
86are not device-specific).
87
88Then, to execute stuff, you repeat these steps, possibly resuing or
89sharing some buffers:
90
91Create some input and output buffers from your context. Set these as
92arguments to your kernel.
93
94Enqueue buffer writes to initialise your input buffers (when not
95initialised at creation time).
96
97Enqueue the kernel execution.
98
99Enqueue buffer reads for your output buffer to read results.
23 100
24=head1 EXAMPLES 101=head1 EXAMPLES
25 102
26=head2 Enumerate all devices and get contexts for them. 103=head2 Enumerate all devices and get contexts for them.
27 104
105Best run this once to get a feel for the platforms and devices in your
106system.
107
28 for my $platform (OpenCL::platforms) { 108 for my $platform (OpenCL::platforms) {
29 warn $platform->info (OpenCL::PLATFORM_NAME); 109 printf "platform: %s\n", $platform->name;
30 warn $platform->info (OpenCL::PLATFORM_EXTENSIONS); 110 printf "extensions: %s\n", $platform->extensions;
31 for my $device ($platform->devices) { 111 for my $device ($platform->devices) {
32 warn $device->info (OpenCL::DEVICE_NAME); 112 printf "+ device: %s\n", $device->name;
33 my $ctx = $device->context_simple; 113 my $ctx = $platform->context (undef, [$device]);
34 # do stuff 114 # do stuff
35 } 115 }
36 } 116 }
37 117
38=head2 Get a useful context and a command queue. 118=head2 Get a useful context and a command queue.
39 119
40 my $dev = ((OpenCL::platforms)[0]->devices)[0]; 120This is a useful boilerplate for any OpenCL program that only wants to use
41 my $ctx = $dev->context_simple; 121one device,
42 my $queue = $ctx->command_queue_simple ($dev); 122
123 my ($platform) = OpenCL::platforms; # find first platform
124 my ($dev) = $platform->devices; # find first device of platform
125 my $ctx = $platform->context (undef, [$dev]); # create context out of those
126 my $queue = $ctx->queue ($dev); # create a command queue for the device
43 127
44=head2 Print all supported image formats of a context. 128=head2 Print all supported image formats of a context.
45 129
130Best run this once for your context, to see whats available and how to
131gather information.
132
46 for my $type (OpenCL::MEM_OBJECT_IMAGE2D, OpenCL::MEM_OBJECT_IMAGE3D) { 133 for my $type (OpenCL::MEM_OBJECT_IMAGE2D, OpenCL::MEM_OBJECT_IMAGE3D) {
47 say "supported image formats for ", OpenCL::enum2str $type; 134 print "supported image formats for ", OpenCL::enum2str $type, "\n";
48 135
49 for my $f ($ctx->supported_image_formats (0, $type)) { 136 for my $f ($ctx->supported_image_formats (0, $type)) {
50 printf " %-10s %-20s\n", OpenCL::enum2str $f->[0], OpenCL::enum2str $f->[1]; 137 printf " %-10s %-20s\n", OpenCL::enum2str $f->[0], OpenCL::enum2str $f->[1];
51 } 138 }
52 } 139 }
54=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,
55then asynchronously. 142then asynchronously.
56 143
57 my $buf = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, "helmut"); 144 my $buf = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, "helmut");
58 145
59 $queue->enqueue_read_buffer ($buf, 1, 1, 3, my $data); 146 $queue->read_buffer ($buf, 1, 1, 3, my $data);
60 warn $data; 147 print "$data\n";
61 148
62 my $ev = $queue->enqueue_read_buffer ($buf, 0, 1, 3, my $data); 149 my $ev = $queue->read_buffer ($buf, 0, 1, 3, my $data);
63 $ev->wait; 150 $ev->wait;
64 warn $data; 151 print "$data\n"; # prints "elm"
65 152
66=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
67functions. 154functions.
68 155
69 my $src = ' 156 my $src = '
70 __kernel void 157 kernel void
71 squareit (__global float *input, __global float *output) 158 squareit (global float *input, global float *output)
72 { 159 {
73 size_t id = get_global_id (0); 160 $id = get_global_id (0);
74 output [id] = input [id] * input [id]; 161 output [id] = input [id] * input [id];
75 } 162 }
76 '; 163 ';
77 164
78 my $prog = $ctx->program_with_source ($src); 165 my $prog = $ctx->build_program ($src);
79
80 eval { $prog->build ($dev); 1 }
81 or die $prog->build_info ($dev, OpenCL::PROGRAM_BUILD_LOG);
82
83 my $kernel = $prog->kernel ("squareit"); 166 my $kernel = $prog->kernel ("squareit");
84 167
85=head2 Create some input and output float buffers, then call squareit on them. 168=head2 Create some input and output float buffers, then call the
169'squareit' kernel on them.
86 170
87 my $input = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, pack "f*", 1, 2, 3, 4.5); 171 my $input = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, pack "f*", 1, 2, 3, 4.5);
88 my $output = $ctx->buffer (0, OpenCL::SIZEOF_FLOAT * 5); 172 my $output = $ctx->buffer (0, OpenCL::SIZEOF_FLOAT * 5);
89 173
90 # set buffer 174 # set buffer
91 $kernel->set_buffer (0, $input); 175 $kernel->set_buffer (0, $input);
92 $kernel->set_buffer (1, $output); 176 $kernel->set_buffer (1, $output);
93 177
94 # execute it for all 4 numbers 178 # execute it for all 4 numbers
95 $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 179 $queue->nd_range_kernel ($kernel, undef, [4], undef);
96 180
97 # enqueue a synchronous read 181 # enqueue a synchronous read
98 $queue->enqueue_read_buffer ($output, 1, 0, OpenCL::SIZEOF_FLOAT * 4, my $data); 182 $queue->read_buffer ($output, 1, 0, OpenCL::SIZEOF_FLOAT * 4, my $data);
99 183
100 # print the results: 184 # print the results:
101 say join ", ", unpack "f*", $data; 185 printf "%s\n", join ", ", unpack "f*", $data;
102 186
103=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,
104showing off barriers. 188showing off barriers.
105 189
106 # execute it for all 4 numbers 190 # execute it for all 4 numbers
107 $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 191 $queue->nd_range_kernel ($kernel, undef, [4], undef);
108 192
109 # enqueue a barrier to ensure in-order execution 193 # enqueue a barrier to ensure in-order execution
110 $queue->enqueue_barrier; 194 $queue->barrier;
111 195
112 # enqueue an async read 196 # enqueue an async read
113 $queue->enqueue_read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data); 197 $queue->read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data);
114 198
115 # wait for all requests to finish 199 # wait for all requests to finish
116 $queue->finish; 200 $queue->finish;
117 201
118=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,
119showing off event objects and wait lists. 203showing off event objects and wait lists.
120 204
121 # execute it for all 4 numbers 205 # execute it for all 4 numbers
122 my $ev = $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 206 my $ev = $queue->nd_range_kernel ($kernel, undef, [4], undef);
123 207
124 # enqueue an async read 208 # enqueue an async read
125 $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);
126 210
127 # wait for the last event to complete 211 # wait for the last event to complete
128 $ev->wait; 212 $ev->wait;
129 213
214=head2 Use the OpenGL module to share a texture between OpenCL and OpenGL and draw some julia
215set tunnel effect.
216
217This is quite a long example to get you going - you can download it from
218L<http://cvs.schmorp.de/OpenCL/examples/juliaflight>.
219
220 use OpenGL ":all";
221 use OpenCL;
222
223 my $S = $ARGV[0] || 256; # window/texture size, smaller is faster
224
225 # open a window and create a gl texture
226 OpenGL::glpOpenWindow width => $S, height => $S;
227 my $texid = glGenTextures_p 1;
228 glBindTexture GL_TEXTURE_2D, $texid;
229 glTexImage2D_c GL_TEXTURE_2D, 0, GL_RGBA8, $S, $S, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0;
230
231 # find and use the first opencl device that let's us get a shared opengl context
232 my $platform;
233 my $dev;
234 my $ctx;
235
236 for (OpenCL::platforms) {
237 $platform = $_;
238 for ($platform->devices) {
239 $dev = $_;
240 $ctx = $platform->context ([OpenCL::GLX_DISPLAY_KHR, undef, OpenCL::GL_CONTEXT_KHR, undef], [$dev])
241 and last;
242 }
243 }
244
245 $ctx
246 or die "cannot find suitable OpenCL device\n";
247
248 my $queue = $ctx->queue ($dev);
249
250 # now attach an opencl image2d object to the opengl texture
251 my $tex = $ctx->gl_texture2d (OpenCL::MEM_WRITE_ONLY, GL_TEXTURE_2D, 0, $texid);
252
253 # now the boring opencl code
254 my $src = <<EOF;
255 kernel void
256 juliatunnel (write_only image2d_t img, float time)
257 {
258 int2 xy = (int2)(get_global_id (0), get_global_id (1));
259 float2 p = convert_float2 (xy) / $S.f * 2.f - 1.f;
260
261 float2 m = (float2)(1.f, p.y) / fabs (p.x); // tunnel
262 m.x = fabs (fmod (m.x + time * 0.05f, 4.f) - 2.f);
263
264 float2 z = m;
265 float2 c = (float2)(sin (time * 0.01133f), cos (time * 0.02521f));
266
267 for (int i = 0; i < 25 && dot (z, z) < 4.f; ++i) // standard julia
268 z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c;
269
270 float3 colour = (float3)(z.x, z.y, atan2 (z.y, z.x));
271 write_imagef (img, xy, (float4)(colour * p.x * p.x, 1.));
272 }
273 EOF
274
275 my $prog = $ctx->build_program ($src);
276 my $kernel = $prog->kernel ("juliatunnel");
277
278 # program compiled, kernel ready, now draw and loop
279
280 for (my $time; ; ++$time) {
281 # acquire objects from opengl
282 $queue->acquire_gl_objects ([$tex]);
283
284 # configure and run our kernel
285 $kernel->setf ("mf", $tex, $time*2); # mf = memory object, float
286 $queue->nd_range_kernel ($kernel, undef, [$S, $S], undef);
287
288 # release objects to opengl again
289 $queue->release_gl_objects ([$tex]);
290
291 # wait
292 $queue->finish;
293
294 # now draw the texture, the defaults should be all right
295 glTexParameterf GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST;
296
297 glEnable GL_TEXTURE_2D;
298 glBegin GL_QUADS;
299 glTexCoord2f 0, 1; glVertex3i -1, -1, -1;
300 glTexCoord2f 0, 0; glVertex3i 1, -1, -1;
301 glTexCoord2f 1, 0; glVertex3i 1, 1, -1;
302 glTexCoord2f 1, 1; glVertex3i -1, 1, -1;
303 glEnd;
304
305 glXSwapBuffers;
306
307 select undef, undef, undef, 1/60;
308 }
309
130=head1 DOCUMENTATION 310=head1 DOCUMENTATION
131 311
132=head2 BASIC CONVENTIONS 312=head2 BASIC CONVENTIONS
133 313
134This is not a 1:1 C-style translation of OpenCL to Perl - instead I 314This is not a one-to-one C-style translation of OpenCL to Perl - instead
135attempted to make the interface as type-safe as possible and introducing 315I attempted to make the interface as type-safe as possible by introducing
136object syntax where it makes sense. There are a number of important 316object syntax where it makes sense. There are a number of important
137differences between the OpenCL C API and this module: 317differences between the OpenCL C API and this module:
138 318
139=over 4 319=over 4
140 320
141=item * Object lifetime managament is automatic - there is no need 321=item * Object lifetime managament is automatic - there is no need
142to free objects explicitly (C<clReleaseXXX>), the release function 322to free objects explicitly (C<clReleaseXXX>), the release function
143is called automatically once all Perl references to it go away. 323is called automatically once all Perl references to it go away.
144 324
145=item * OpenCL uses CamelCase for function names (C<clGetPlatformInfo>), 325=item * OpenCL uses CamelCase for function names
326(e.g. C<clGetPlatformIDs>, C<clGetPlatformInfo>), while this module
146while this module uses underscores as word separator and often leaves out 327uses underscores as word separator and often leaves out prefixes
147prefixes (C<< $platform->info >>). 328(C<OpenCL::platforms>, C<< $platform->info >>).
148 329
149=item * OpenCL often specifies fixed vector function arguments as short 330=item * OpenCL often specifies fixed vector function arguments as short
150arrays (C<size_t origin[3]>), while this module explicitly expects the 331arrays (C<size_t origin[3]>), while this module explicitly expects the
151components as separate arguments- 332components as separate arguments (C<$orig_x, $orig_y, $orig_z>) in
333function calls.
152 334
153=item * Where possible, the row_pitch value is calculated from the perl 335=item * Structures are often specified by flattening out their components
154scalar length and need not be specified. 336as with short vectors, and returned as arrayrefs.
155 337
156=item * When enqueuing commands, the wait list is specified by adding 338=item * When enqueuing commands, the wait list is specified by adding
157extra arguments to the function - everywhere a C<$wait_events...> argument 339extra arguments to the function - anywhere a C<$wait_events...> argument
158is documented this can be any number of event objects. 340is documented this can be any number of event objects. As an extsnion
341implemented by this module, C<undef> values will be ignored in the event
342list.
159 343
160=item * When enqueuing commands, if the enqueue method is called in void 344=item * When enqueuing commands, if the enqueue method is called in void
161context, no event is created. In all other contexts an event is returned 345context, no event is created. In all other contexts an event is returned
162by the method. 346by the method.
163 347
183 ulong IV - Q 367 ulong IV - Q
184 float NV float f 368 float NV float f
185 half IV ushort S 369 half IV ushort S
186 double NV double d 370 double NV double d
187 371
372=head2 GLX SUPPORT
373
374Due to the sad state that OpenGL support is in in Perl (mostly the OpenGL
375module, which has little to no documentation and has little to no support
376for glX), this module, as a special extension, treats context creation
377properties C<OpenCL::GLX_DISPLAY_KHR> and C<OpenCL::GL_CONTEXT_KHR>
378specially: If either or both of these are C<undef>, then the OpenCL
379module tries to dynamically resolve C<glXGetCurrentDisplay> and
380C<glXGetCurrentContext>, call these functions and use their return values
381instead.
382
383For this to work, the OpenGL library must be loaded, a GLX context must
384have been created and be made current, and C<dlsym> must be available and
385capable of finding the function via C<RTLD_DEFAULT>.
386
387=head2 EVENT SYSTEM
388
389OpenCL can generate a number of (potentially) asynchronous events, for
390example, after compiling a program, to signal a context-related error or,
391perhaps most important, to signal completion of queued jobs (by setting
392callbacks on OpenCL::Event objects).
393
394To facilitate this, this module maintains an event queue - each
395time an asynchronous event happens, it is queued, and perl will be
396interrupted. This is implemented via the L<Async::Interrupt> module. In
397addition, this module has L<AnyEvent> support, so it can seamlessly
398integrate itself into many event loops.
399
400Since this module is a bit hard to understand, here are some case examples:
401
402=head3 Don't use callbacks.
403
404When your program never uses any callbacks, then there will never be any
405notifications you need to take care of, and therefore no need to worry
406about all this.
407
408You can achieve a great deal by explicitly waiting for events, or using
409barriers and flush calls. In many programs, there is no need at all to
410tinker with asynchronous events.
411
412=head3 Use AnyEvent
413
414This module automatically registers a watcher that invokes all outstanding
415event callbacks when AnyEvent is initialised (and block asynchronous
416interruptions). Using this mode of operations is the safest and most
417recommended one.
418
419To use this, simply use AnyEvent and this module normally, make sure you
420have an event loop running:
421
422 use Gtk2 -init;
423 use AnyEvent;
424
425 # initialise AnyEvent, by creating a watcher, or:
426 AnyEvent::detect;
427
428 my $e = $queue->marker;
429 $e->cb (sub {
430 warn "opencl is finished\n";
431 })
432
433 main Gtk2;
434
435Note that this module will not initialise AnyEvent for you. Before
436AnyEvent is initialised, the module will asynchronously interrupt perl
437instead. To avoid any surprises, it's best to explicitly initialise
438AnyEvent.
439
440You can temporarily enable asynchronous interruptions (see next paragraph)
441by calling C<$OpenCL::INTERRUPT->unblock> and disable them again by
442calling C<$OpenCL::INTERRUPT->block>.
443
444=head3 Let yourself be interrupted at any time
445
446This mode is the default unless AnyEvent is loaded and initialised. In
447this mode, OpenCL asynchronously interrupts a running perl program. The
448emphasis is on both I<asynchronously> and I<running> here.
449
450Asynchronously means that perl might execute your callbacks at any
451time. For example, in the following code (I<THAT YOU SHOULD NOT COPY>),
452the C<until> loop following the marker call will be interrupted by the
453callback:
454
455 my $e = $queue->marker;
456 my $flag;
457 $e->cb (sub { $flag = 1 });
458 1 until $flag;
459 # $flag is now 1
460
461The reason why you shouldn't blindly copy the above code is that
462busy waiting is a really really bad thing, and really really bad for
463performance.
464
465While at first this asynchronous business might look exciting, it can be
466really hard, because you need to be prepared for the callback code to be
467executed at any time, which limits the amount of things the callback code
468can do safely.
469
470This can be mitigated somewhat by using C<<
471$OpenCL::INTERRUPT->scope_block >> (see the L<Async::Interrupt>
472documentation for details).
473
474The other problem is that your program must be actively I<running> to be
475interrupted. When you calculate stuff, your program is running. When you
476hang in some C functions or other block execution (by calling C<sleep>,
477C<select>, running an event loop and so on), your program is waiting, not
478running.
479
480One way around that would be to attach a read watcher to your event loop,
481listening for events on C<< $OpenCL::INTERRUPT->pipe_fileno >>, using a
482dummy callback (C<sub { }>) to temporarily execute some perl code.
483
484That is then awfully close to using the built-in AnyEvent support above,
485though, so consider that one instead.
486
487=head3 Be creative
488
489OpenCL exports the L<Async::Interrupt> object it uses in the global
490variable C<$OpenCL::INTERRUPT>. You can configure it in any way you like.
491
492So if you want to feel like a real pro, err, wait, if you feel no risk
493menas no fun, you can experiment by implementing your own mode of
494operations.
495
496=cut
497
498package OpenCL;
499
500use common::sense;
501use Carp ();
502use Async::Interrupt ();
503
504our $POLL_FUNC; # set by XS
505
506BEGIN {
507 our $VERSION = '0.98';
508
509 require XSLoader;
510 XSLoader::load (__PACKAGE__, $VERSION);
511
512 @OpenCL::Platform::ISA =
513 @OpenCL::Device::ISA =
514 @OpenCL::Context::ISA =
515 @OpenCL::Queue::ISA =
516 @OpenCL::Memory::ISA =
517 @OpenCL::Sampler::ISA =
518 @OpenCL::Program::ISA =
519 @OpenCL::Kernel::ISA =
520 @OpenCL::Event::ISA = OpenCL::Object::;
521
522 @OpenCL::Buffer::ISA =
523 @OpenCL::Image::ISA = OpenCL::Memory::;
524
525 @OpenCL::BufferObj::ISA = OpenCL::Buffer::;
526
527 @OpenCL::Image2D::ISA =
528 @OpenCL::Image3D::ISA =
529 @OpenCL::Image2DArray::ISA =
530 @OpenCL::Image1D::ISA =
531 @OpenCL::Image1DArray::ISA =
532 @OpenCL::Image1DBuffer::ISA = OpenCL::Image::;
533
534 @OpenCL::UserEvent::ISA = OpenCL::Event::;
535}
536
188=head2 THE OpenCL PACKAGE 537=head2 THE OpenCL PACKAGE
189 538
190=over 4 539=over 4
191 540
192=item $int = OpenCL::errno 541=item $int = OpenCL::errno
193 542
194The last error returned by a function - it's only changed on errors. 543The last error returned by a function - it's only valid after an error occured
544and before calling another OpenCL function.
195 545
196=item $str = OpenCL::err2str $errval 546=item $str = OpenCL::err2str [$errval]
197 547
198Comverts an error value into a human readable string. 548Converts an error value into a human readable string. IF no error value is
549given, then the last error will be used (as returned by OpenCL::errno).
199 550
200=item $str = OpenCL::err2str $enum 551=item $str = OpenCL::enum2str $enum
201 552
202Converts most enum values (inof parameter names, image format constants, 553Converts most enum values (of parameter names, image format constants,
203object types, addressing and filter modes, command types etc.) into a 554object types, addressing and filter modes, command types etc.) into a
204human readbale string. When confronted with some random integer it can be 555human readable string. When confronted with some random integer it can be
205very helpful to pass it through this function to maybe get some readable 556very helpful to pass it through this function to maybe get some readable
206string out of it. 557string out of it.
207 558
208=item @platforms = OpenCL::platforms 559=item @platforms = OpenCL::platforms
209 560
210Returns all available OpenCL::Platform objects. 561Returns all available OpenCL::Platform objects.
211 562
212L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html> 563L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html>
213 564
214=item $ctx = OpenCL::context_from_type_simple $type = OpenCL::DEVICE_TYPE_DEFAULT 565=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr
215 566
216Tries to create a context from a default device and platform - never worked for me. 567Tries to create a context from a default device and platform type - never worked for me.
217 568
218L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 569L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
219 570
571=item $ctx = OpenCL::context $properties, \@devices, $callback->($err, $pvt) = $print_stderr)
572
573Create a new OpenCL::Context object using the given device object(s). This
574function isn't implemented yet, use C<< $platform->context >> instead.
575
576L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
577
220=item OpenCL::wait_for_events $wait_events... 578=item OpenCL::wait_for_events $wait_events...
221 579
222Waits for all events to complete. 580Waits for all events to complete.
223 581
224L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 582L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
225 583
584=item OpenCL::poll
585
586Checks if there are any outstanding events (see L<EVENT SYSTEM>) and
587invokes their callbacks.
588
589=item $OpenCL::INTERRUPT
590
591The L<Async::Interrupt> object used to signal asynchronous events (see
592L<EVENT SYSTEM>).
593
594=cut
595
596our $INTERRUPT = new Async::Interrupt c_cb => [$POLL_FUNC, 0];
597
598&_eq_initialise ($INTERRUPT->signal_func);
599
600=item $OpenCL::WATCHER
601
602The L<AnyEvent> watcher object used to watch for asynchronous events (see
603L<EVENT SYSTEM>). This variable is C<undef> until L<AnyEvent> has been
604loaded I<and> initialised (e.g. by calling C<AnyEvent::detect>).
605
606=cut
607
608our $WATCHER;
609
610sub _init_anyevent {
611 $INTERRUPT->block;
612 $WATCHER = AE::io ($INTERRUPT->pipe_fileno, 0, sub { $INTERRUPT->handle });
613}
614
615if (defined $AnyEvent::MODEL) {
616 _init_anyevent;
617} else {
618 push @AnyEvent::post_detect, \&_init_anyevent;
619}
620
226=back 621=back
227 622
623=head2 THE OpenCL::Object CLASS
624
625This is the base class for all objects in the OpenCL module. The only
626method it implements is the C<id> method, which is only useful if you want
627to interface to OpenCL on the C level.
628
629=over 4
630
631=item $iv = $obj->id
632
633OpenCL objects are represented by pointers or integers on the C level. If
634you want to interface to an OpenCL object directly on the C level, then
635you need this value, which is returned by this method. You should use an
636C<IV> type in your code and cast that to the correct type.
637
638=cut
639
640sub OpenCL::Object::id {
641 ref $_[0] eq "SCALAR"
642 ? ${ $_[0] }
643 : $_[0][0]
644}
645
646=back
647
228=head2 THE OpenCL::Platform CLASS 648=head2 THE OpenCL::Platform CLASS
229 649
230=over 4 650=over 4
231 651
652=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL)
653
654Returns a list of matching OpenCL::Device objects.
655
656=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr)
657
658Tries to create a context. Never worked for me, and you need devices explicitly anyway.
659
660L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
661
662=item $ctx = $platform->context ($properties, \@devices, $callback->($err, $pvt) = $print_stderr)
663
664Create a new OpenCL::Context object using the given device object(s)- a
665CL_CONTEXT_PLATFORM property is supplied automatically.
666
667L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
668
232=item $packed_value = $platform->info ($name) 669=item $packed_value = $platform->info ($name)
233 670
234Calls C<clGetPlatformInfo> and returns the packed, raw value - for 671Calls C<clGetPlatformInfo> and returns the packed, raw value - for
235strings, this will be the string, for other values you probably need to 672strings, this will be the string (possibly including terminating \0), for
236use the correct C<unpack>. This might get improved in the future. Hopefully. 673other values you probably need to use the correct C<unpack>.
674
675It's best to avoid this method and use one of the following convenience
676wrappers.
237 677
238L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html> 678L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html>
239 679
240=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL) 680=item $platform->unload_compiler
241 681
242Returns a list of matching OpenCL::Device objects. 682Attempts to unload the compiler for this platform, for endless
683profit. Does nothing on OpenCL 1.1.
243 684
244=item $ctx = $platform->context_from_type_simple ($type = OpenCL::DEVICE_TYPE_DEFAULT)
245
246Tries to create a context. Never worked for me.
247
248L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 685L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clUnloadPlatformCompiler.html>
686
687=for gengetinfo begin platform
688
689=item $string = $platform->profile
690
691Calls C<clGetPlatformInfo> with C<CL_PLATFORM_PROFILE> and returns the result.
692
693=item $string = $platform->version
694
695Calls C<clGetPlatformInfo> with C<CL_PLATFORM_VERSION> and returns the result.
696
697=item $string = $platform->name
698
699Calls C<clGetPlatformInfo> with C<CL_PLATFORM_NAME> and returns the result.
700
701=item $string = $platform->vendor
702
703Calls C<clGetPlatformInfo> with C<CL_PLATFORM_VENDOR> and returns the result.
704
705=item $string = $platform->extensions
706
707Calls C<clGetPlatformInfo> with C<CL_PLATFORM_EXTENSIONS> and returns the result.
708
709=for gengetinfo end platform
249 710
250=back 711=back
251 712
252=head2 THE OpenCL::Device CLASS 713=head2 THE OpenCL::Device CLASS
253 714
257 718
258See C<< $platform->info >> for details. 719See C<< $platform->info >> for details.
259 720
260L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetDeviceInfo.html> 721L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetDeviceInfo.html>
261 722
723=for gengetinfo begin device
724
725=item $device_type = $device->type
726
727Calls C<clGetDeviceInfo> with C<CL_DEVICE_TYPE> and returns the result.
728
729=item $uint = $device->vendor_id
730
731Calls C<clGetDeviceInfo> with C<CL_DEVICE_VENDOR_ID> and returns the result.
732
733=item $uint = $device->max_compute_units
734
735Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_COMPUTE_UNITS> and returns the result.
736
737=item $uint = $device->max_work_item_dimensions
738
739Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_WORK_ITEM_DIMENSIONS> and returns the result.
740
741=item $int = $device->max_work_group_size
742
743Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_WORK_GROUP_SIZE> and returns the result.
744
745=item @ints = $device->max_work_item_sizes
746
747Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_WORK_ITEM_SIZES> and returns the result.
748
749=item $uint = $device->preferred_vector_width_char
750
751Calls C<clGetDeviceInfo> with C<CL_DEVICE_PREFERRED_VECTOR_WIDTH_CHAR> and returns the result.
752
753=item $uint = $device->preferred_vector_width_short
754
755Calls C<clGetDeviceInfo> with C<CL_DEVICE_PREFERRED_VECTOR_WIDTH_SHORT> and returns the result.
756
757=item $uint = $device->preferred_vector_width_int
758
759Calls C<clGetDeviceInfo> with C<CL_DEVICE_PREFERRED_VECTOR_WIDTH_INT> and returns the result.
760
761=item $uint = $device->preferred_vector_width_long
762
763Calls C<clGetDeviceInfo> with C<CL_DEVICE_PREFERRED_VECTOR_WIDTH_LONG> and returns the result.
764
765=item $uint = $device->preferred_vector_width_float
766
767Calls C<clGetDeviceInfo> with C<CL_DEVICE_PREFERRED_VECTOR_WIDTH_FLOAT> and returns the result.
768
769=item $uint = $device->preferred_vector_width_double
770
771Calls C<clGetDeviceInfo> with C<CL_DEVICE_PREFERRED_VECTOR_WIDTH_DOUBLE> and returns the result.
772
773=item $uint = $device->max_clock_frequency
774
775Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_CLOCK_FREQUENCY> and returns the result.
776
777=item $bitfield = $device->address_bits
778
779Calls C<clGetDeviceInfo> with C<CL_DEVICE_ADDRESS_BITS> and returns the result.
780
781=item $uint = $device->max_read_image_args
782
783Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_READ_IMAGE_ARGS> and returns the result.
784
785=item $uint = $device->max_write_image_args
786
787Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_WRITE_IMAGE_ARGS> and returns the result.
788
789=item $ulong = $device->max_mem_alloc_size
790
791Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_MEM_ALLOC_SIZE> and returns the result.
792
793=item $int = $device->image2d_max_width
794
795Calls C<clGetDeviceInfo> with C<CL_DEVICE_IMAGE2D_MAX_WIDTH> and returns the result.
796
797=item $int = $device->image2d_max_height
798
799Calls C<clGetDeviceInfo> with C<CL_DEVICE_IMAGE2D_MAX_HEIGHT> and returns the result.
800
801=item $int = $device->image3d_max_width
802
803Calls C<clGetDeviceInfo> with C<CL_DEVICE_IMAGE3D_MAX_WIDTH> and returns the result.
804
805=item $int = $device->image3d_max_height
806
807Calls C<clGetDeviceInfo> with C<CL_DEVICE_IMAGE3D_MAX_HEIGHT> and returns the result.
808
809=item $int = $device->image3d_max_depth
810
811Calls C<clGetDeviceInfo> with C<CL_DEVICE_IMAGE3D_MAX_DEPTH> and returns the result.
812
813=item $uint = $device->image_support
814
815Calls C<clGetDeviceInfo> with C<CL_DEVICE_IMAGE_SUPPORT> and returns the result.
816
817=item $int = $device->max_parameter_size
818
819Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_PARAMETER_SIZE> and returns the result.
820
262=item $ctx = $device->context_simple 821=item $uint = $device->max_samplers
263 822
264Convenience function to create a new OpenCL::Context object. 823Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_SAMPLERS> and returns the result.
265 824
266L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html> 825=item $uint = $device->mem_base_addr_align
826
827Calls C<clGetDeviceInfo> with C<CL_DEVICE_MEM_BASE_ADDR_ALIGN> and returns the result.
828
829=item $uint = $device->min_data_type_align_size
830
831Calls C<clGetDeviceInfo> with C<CL_DEVICE_MIN_DATA_TYPE_ALIGN_SIZE> and returns the result.
832
833=item $device_fp_config = $device->single_fp_config
834
835Calls C<clGetDeviceInfo> with C<CL_DEVICE_SINGLE_FP_CONFIG> and returns the result.
836
837=item $device_mem_cache_type = $device->global_mem_cache_type
838
839Calls C<clGetDeviceInfo> with C<CL_DEVICE_GLOBAL_MEM_CACHE_TYPE> and returns the result.
840
841=item $uint = $device->global_mem_cacheline_size
842
843Calls C<clGetDeviceInfo> with C<CL_DEVICE_GLOBAL_MEM_CACHELINE_SIZE> and returns the result.
844
845=item $ulong = $device->global_mem_cache_size
846
847Calls C<clGetDeviceInfo> with C<CL_DEVICE_GLOBAL_MEM_CACHE_SIZE> and returns the result.
848
849=item $ulong = $device->global_mem_size
850
851Calls C<clGetDeviceInfo> with C<CL_DEVICE_GLOBAL_MEM_SIZE> and returns the result.
852
853=item $ulong = $device->max_constant_buffer_size
854
855Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_CONSTANT_BUFFER_SIZE> and returns the result.
856
857=item $uint = $device->max_constant_args
858
859Calls C<clGetDeviceInfo> with C<CL_DEVICE_MAX_CONSTANT_ARGS> and returns the result.
860
861=item $device_local_mem_type = $device->local_mem_type
862
863Calls C<clGetDeviceInfo> with C<CL_DEVICE_LOCAL_MEM_TYPE> and returns the result.
864
865=item $ulong = $device->local_mem_size
866
867Calls C<clGetDeviceInfo> with C<CL_DEVICE_LOCAL_MEM_SIZE> and returns the result.
868
869=item $boolean = $device->error_correction_support
870
871Calls C<clGetDeviceInfo> with C<CL_DEVICE_ERROR_CORRECTION_SUPPORT> and returns the result.
872
873=item $int = $device->profiling_timer_resolution
874
875Calls C<clGetDeviceInfo> with C<CL_DEVICE_PROFILING_TIMER_RESOLUTION> and returns the result.
876
877=item $boolean = $device->endian_little
878
879Calls C<clGetDeviceInfo> with C<CL_DEVICE_ENDIAN_LITTLE> and returns the result.
880
881=item $boolean = $device->available
882
883Calls C<clGetDeviceInfo> with C<CL_DEVICE_AVAILABLE> and returns the result.
884
885=item $boolean = $device->compiler_available
886
887Calls C<clGetDeviceInfo> with C<CL_DEVICE_COMPILER_AVAILABLE> and returns the result.
888
889=item $device_exec_capabilities = $device->execution_capabilities
890
891Calls C<clGetDeviceInfo> with C<CL_DEVICE_EXECUTION_CAPABILITIES> and returns the result.
892
893=item $command_queue_properties = $device->properties
894
895Calls C<clGetDeviceInfo> with C<CL_DEVICE_QUEUE_PROPERTIES> and returns the result.
896
897=item $ = $device->platform
898
899Calls C<clGetDeviceInfo> with C<CL_DEVICE_PLATFORM> and returns the result.
900
901=item $string = $device->name
902
903Calls C<clGetDeviceInfo> with C<CL_DEVICE_NAME> and returns the result.
904
905=item $string = $device->vendor
906
907Calls C<clGetDeviceInfo> with C<CL_DEVICE_VENDOR> and returns the result.
908
909=item $string = $device->driver_version
910
911Calls C<clGetDeviceInfo> with C<CL_DRIVER_VERSION> and returns the result.
912
913=item $string = $device->profile
914
915Calls C<clGetDeviceInfo> with C<CL_DEVICE_PROFILE> and returns the result.
916
917=item $string = $device->version
918
919Calls C<clGetDeviceInfo> with C<CL_DEVICE_VERSION> and returns the result.
920
921=item $string = $device->extensions
922
923Calls C<clGetDeviceInfo> with C<CL_DEVICE_EXTENSIONS> and returns the result.
924
925=item $uint = $device->preferred_vector_width_half
926
927Calls C<clGetDeviceInfo> with C<CL_DEVICE_PREFERRED_VECTOR_WIDTH_HALF> and returns the result.
928
929=item $uint = $device->native_vector_width_char
930
931Calls C<clGetDeviceInfo> with C<CL_DEVICE_NATIVE_VECTOR_WIDTH_CHAR> and returns the result.
932
933=item $uint = $device->native_vector_width_short
934
935Calls C<clGetDeviceInfo> with C<CL_DEVICE_NATIVE_VECTOR_WIDTH_SHORT> and returns the result.
936
937=item $uint = $device->native_vector_width_int
938
939Calls C<clGetDeviceInfo> with C<CL_DEVICE_NATIVE_VECTOR_WIDTH_INT> and returns the result.
940
941=item $uint = $device->native_vector_width_long
942
943Calls C<clGetDeviceInfo> with C<CL_DEVICE_NATIVE_VECTOR_WIDTH_LONG> and returns the result.
944
945=item $uint = $device->native_vector_width_float
946
947Calls C<clGetDeviceInfo> with C<CL_DEVICE_NATIVE_VECTOR_WIDTH_FLOAT> and returns the result.
948
949=item $uint = $device->native_vector_width_double
950
951Calls C<clGetDeviceInfo> with C<CL_DEVICE_NATIVE_VECTOR_WIDTH_DOUBLE> and returns the result.
952
953=item $uint = $device->native_vector_width_half
954
955Calls C<clGetDeviceInfo> with C<CL_DEVICE_NATIVE_VECTOR_WIDTH_HALF> and returns the result.
956
957=item $device_fp_config = $device->double_fp_config
958
959Calls C<clGetDeviceInfo> with C<CL_DEVICE_DOUBLE_FP_CONFIG> and returns the result.
960
961=item $device_fp_config = $device->half_fp_config
962
963Calls C<clGetDeviceInfo> with C<CL_DEVICE_HALF_FP_CONFIG> and returns the result.
964
965=item $boolean = $device->host_unified_memory
966
967Calls C<clGetDeviceInfo> with C<CL_DEVICE_HOST_UNIFIED_MEMORY> and returns the result.
968
969=item $device = $device->parent_device_ext
970
971Calls C<clGetDeviceInfo> with C<CL_DEVICE_PARENT_DEVICE_EXT> and returns the result.
972
973=item @device_partition_property_exts = $device->partition_types_ext
974
975Calls C<clGetDeviceInfo> with C<CL_DEVICE_PARTITION_TYPES_EXT> and returns the result.
976
977=item @device_partition_property_exts = $device->affinity_domains_ext
978
979Calls C<clGetDeviceInfo> with C<CL_DEVICE_AFFINITY_DOMAINS_EXT> and returns the result.
980
981=item $uint = $device->reference_count_ext
982
983Calls C<clGetDeviceInfo> with C<CL_DEVICE_REFERENCE_COUNT_EXT> and returns the result.
984
985=item @device_partition_property_exts = $device->partition_style_ext
986
987Calls C<clGetDeviceInfo> with C<CL_DEVICE_PARTITION_STYLE_EXT> and returns the result.
988
989=for gengetinfo end device
267 990
268=back 991=back
269 992
270=head2 THE OpenCL::Context CLASS 993=head2 THE OpenCL::Context CLASS
271 994
272=over 4 995=over 4
273 996
274=item $packed_value = $ctx->info ($name) 997=item $prog = $ctx->build_program ($program, $options = "")
275 998
276See C<< $platform->info >> for details. 999This convenience function tries to build the program on all devices in
1000the context. If the build fails, then the function will C<croak> with the
1001build log. Otherwise ti returns the program object.
277 1002
278L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetContextInfo.html> 1003The C<$program> can either be a C<OpenCL::Program> object or a string
1004containing the program. In the latter case, a program objetc will be
1005created automatically.
279 1006
280=item $queue = $ctx->command_queue_simple ($device) 1007=cut
281 1008
1009sub OpenCL::Context::build_program {
1010 my ($self, $prog, $options) = @_;
1011
1012 $prog = $self->program_with_source ($prog)
1013 unless ref $prog;
1014
1015 eval { $prog->build (undef, $options); 1 }
1016 or errno == BUILD_PROGRAM_FAILURE
1017 or errno == INVALID_BINARY # workaround nvidia bug
1018 or Carp::croak "OpenCL::Context->build_program: " . err2str;
1019
1020 # we check status for all devices
1021 for my $dev ($self->devices) {
1022 $prog->build_status ($dev) == BUILD_SUCCESS
1023 or Carp::croak "Building OpenCL program for device '" . $dev->name . "' failed:\n"
1024 . $prog->build_log ($dev);
1025 }
1026
1027 $prog
1028}
1029
1030=item $queue = $ctx->queue ($device, $properties)
1031
282Convenience function to create a new OpenCL::Queue object from the context and the given device. 1032Create a new OpenCL::Queue object from the context and the given device.
283 1033
284L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html> 1034L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html>
285 1035
1036Example: create an out-of-order queue.
1037
1038 $queue = $ctx->queue ($device, OpenCL::QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE);
1039
286=item $ev = $ctx->user_event 1040=item $ev = $ctx->user_event
287 1041
288Creates a new OpenCL::UserEvent object. 1042Creates a new OpenCL::UserEvent object.
289 1043
290L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateUserEvent.html> 1044L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateUserEvent.html>
291 1045
292=item $buf = $ctx->buffer ($flags, $len) 1046=item $buf = $ctx->buffer ($flags, $len)
293 1047
294Creates a new OpenCL::Buffer object with the given flags and octet-size. 1048Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object with the
1049given flags and octet-size.
295 1050
296L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateBuffer.html> 1051L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateBuffer.html>
297 1052
298=item $buf = $ctx->buffer_sv ($flags, $data) 1053=item $buf = $ctx->buffer_sv ($flags, $data)
299 1054
300Creates a new OpenCL::Buffer object and initialise it with the given data values. 1055Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and
1056initialise it with the given data values.
301 1057
1058=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)
1059
1060Creates a new OpenCL::Image object and optionally initialises it with
1061the given data values.
1062
1063L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateImage.html>
1064
302=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $data) 1065=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $row_pitch = 0, $data = undef)
303 1066
304Creates a new OpenCL::Image2D object and optionally initialises it with the given data values. 1067Creates a new OpenCL::Image2D object and optionally initialises it with
1068the given data values.
305 1069
306L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage2D.html> 1070L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage2D.html>
307 1071
308=item $img = $ctx->image3d ($flags, $channel_order, $channel_type, $width, $height, $depth, $slice_pitch, $data) 1072=item $img = $ctx->image3d ($flags, $channel_order, $channel_type, $width, $height, $depth, $row_pitch = 0, $slice_pitch = 0, $data = undef)
309 1073
310Creates a new OpenCL::Image3D object and optionally initialises it with the given data values. 1074Creates a new OpenCL::Image3D object and optionally initialises it with
1075the given data values.
311 1076
312L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage3D.html> 1077L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage3D.html>
1078
1079=item $buffer = $ctx->gl_buffer ($flags, $bufobj)
1080
1081Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object that refers to the given
1082OpenGL buffer object.
1083
1084http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLBuffer.html
1085
1086=item $img = $ctx->gl_texture ($flags, $target, $miplevel, $texture)
1087
1088Creates a new OpenCL::Image object that refers to the given OpenGL
1089texture object or buffer.
1090
1091http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateFromGLTexture.html
1092
1093=item $img = $ctx->gl_texture2d ($flags, $target, $miplevel, $texture)
1094
1095Creates a new OpenCL::Image2D object that refers to the given OpenGL
10962D texture object.
1097
1098http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture2D.html
1099
1100=item $img = $ctx->gl_texture3d ($flags, $target, $miplevel, $texture)
1101
1102Creates a new OpenCL::Image3D object that refers to the given OpenGL
11033D texture object.
1104
1105http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture3D.html
1106
1107=item $ctx->gl_renderbuffer ($flags, $renderbuffer)
1108
1109Creates a new OpenCL::Image2D object that refers to the given OpenGL
1110render buffer.
1111
1112http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLRenderbuffer.html
313 1113
314=item @formats = $ctx->supported_image_formats ($flags, $image_type) 1114=item @formats = $ctx->supported_image_formats ($flags, $image_type)
315 1115
316Returns a list of matching image formats - each format is an arrayref with 1116Returns a list of matching image formats - each format is an arrayref with
317two values, $channel_order and $channel_type, in it. 1117two values, $channel_order and $channel_type, in it.
328 1128
329Creates a new OpenCL::Program object from the given source code. 1129Creates a new OpenCL::Program object from the given source code.
330 1130
331L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateProgramWithSource.html> 1131L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateProgramWithSource.html>
332 1132
1133=item $packed_value = $ctx->info ($name)
1134
1135See C<< $platform->info >> for details.
1136
1137L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetContextInfo.html>
1138
1139=for gengetinfo begin context
1140
1141=item $uint = $context->reference_count
1142
1143Calls C<clGetContextInfo> with C<CL_CONTEXT_REFERENCE_COUNT> and returns the result.
1144
1145=item @devices = $context->devices
1146
1147Calls C<clGetContextInfo> with C<CL_CONTEXT_DEVICES> and returns the result.
1148
1149=item @property_ints = $context->properties
1150
1151Calls C<clGetContextInfo> with C<CL_CONTEXT_PROPERTIES> and returns the result.
1152
1153=item $uint = $context->num_devices
1154
1155Calls C<clGetContextInfo> with C<CL_CONTEXT_NUM_DEVICES> and returns the result.
1156
1157=for gengetinfo end context
1158
333=back 1159=back
334 1160
335=head2 THE OpenCL::Queue CLASS 1161=head2 THE OpenCL::Queue CLASS
336 1162
337An OpenCL::Queue represents an execution queue for OpenCL. You execute 1163An OpenCL::Queue represents an execution queue for OpenCL. You execute
338requests by calling their respective C<enqueue_xxx> method and waitinf for 1164requests by calling their respective method and waiting for it to complete
339it to complete in some way. 1165in some way.
340 1166
341All the enqueue methods return an event object that can be used to wait 1167Most methods that enqueue some request return an event object that can
342for completion, unless the method is called in void context, in which case 1168be used to wait for completion (optionally using a callback), unless
343no event object is created. 1169the method is called in void context, in which case no event object is
1170created.
344 1171
345They also allow you to specify any number of other event objects that this 1172They also allow you to specify any number of other event objects that this
346request has to wait for before it starts executing, by simply passing the 1173request has to wait for before it starts executing, by simply passing the
347event objects as extra parameters to the enqueue methods. 1174event objects as extra parameters to the enqueue methods. To simplify
1175program design, this module ignores any C<undef> values in the list of
1176events. This makes it possible to code operations such as this, without
1177having to put a valid event object into C<$event> first:
1178
1179 $event = $queue->xxx (..., $event);
348 1180
349Queues execute in-order by default, without any parallelism, so in most 1181Queues execute in-order by default, without any parallelism, so in most
350cases (i.e. you use only one queue) it's not necessary to wait for or 1182cases (i.e. you use only one queue) it's not necessary to wait for or
351create event objects. 1183create event objects, althoguh an our of order queue is often a bit
1184faster.
352 1185
353=over 4 1186=over 4
354 1187
355=item $packed_value = $ctx->info ($name)
356
357See C<< $platform->info >> for details.
358
359L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetCommandQueueInfo.html>
360
361=item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...) 1188=item $ev = $queue->read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...)
362 1189
363Reads data from buffer into the given string. 1190Reads data from buffer into the given string.
364 1191
365L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBuffer.html> 1192L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBuffer.html>
366 1193
367=item $ev = $queue->enqueue_write_buffer ($buffer, $blocking, $offset, $data, $wait_events...) 1194=item $ev = $queue->write_buffer ($buffer, $blocking, $offset, $data, $wait_events...)
368 1195
369Writes data to buffer from the given string. 1196Writes data to buffer from the given string.
370 1197
371L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBuffer.html> 1198L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBuffer.html>
372 1199
373=item $ev = $queue->enqueue_copy_buffer ($src, $dst, $src_offset, $dst_offset, $len, $wait_events...) 1200=item $ev = $queue->copy_buffer ($src, $dst, $src_offset, $dst_offset, $len, $wait_events...)
374 1201
375L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBuffer.html> 1202L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBuffer.html>
376 1203
1204=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...)
1205
1206http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBufferRect.html
1207
1208=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...)
1209
1210http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBufferRect.html
1211
377=item $ev = $queue->enqueue_read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) 1212=item $ev = $queue->read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...)
1213
1214L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferRect.html>
1215
1216=item $ev = $queue->copy_buffer_to_image ($src_buffer, $dst_image, $src_offset, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...)
378 1217
379L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html> 1218L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html>
380 1219
381=item $ev = $queue->enqueue_write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $data, $wait_events...) 1220=item $ev = $queue->write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...)
382 1221
383L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html> 1222L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html>
384 1223
1224=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...)
1225
1226L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html>
1227
1228=item $ev = $queue->copy_image_to_buffer ($src_image, $dst_image, $src_x, $src_y, $src_z, $width, $height, $depth, $dst_offset, $wait_events...)
1229
1230L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImageToBuffer.html>
1231
385=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, 4dst_row_pitch, $dst_slice_pitch, $ait_event...) 1232=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...)
386 1233
387Yeah. 1234Yeah.
388 1235
389L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferRect.html>
390
391=item $ev = $queue->enqueue_copy_buffer_to_image (OpenCL::Buffer src, OpenCL::Image dst, size_t src_offset, size_t dst_x, size_t dst_y, size_t dst_z, size_t width, size_t height, size_t depth, ...)
392
393L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>. 1236L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>.
394 1237
395=item $ev = $queue->enqueue_copy_image (OpenCL::Image src, OpenCL::Buffer dst, size_t src_x, size_t src_y, size_t src_z, size_t dst_x, size_t dst_y, size_t dst_z, size_t width, size_t height, size_t depth, ...) 1238=item $ev = $queue->fill_buffer ($mem, $pattern, $offset, $size, ...)
396 1239
1240Fills the given buffer object with repeated applications of C<$pattern>,
1241starting at C<$offset> for C<$size> octets.
1242
1243L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillBuffer.html>
1244
1245=item $ev = $queue->fill_image ($img, $r, $g, $b, $a, $x, $y, $z, $width, $height, $depth, ...)
1246
1247Fills the given image area with the given rgba colour components. The
1248components are normally floating point values between C<0> and C<1>,
1249except when the image channel data type is a signe dor unsigned
1250unnormalised format, in which case the range is determined by the format.
1251
397L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html> 1252L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillImage.html>
398 1253
399=item $ev = $queue->enqueue_copy_image_to_buffer (OpenCL::Image src, OpenCL::Buffer dst, size_t src_x, size_t src_y, size_t src_z, size_t width, size_t height, size_t depth, size_t dst_offset, ...)
400
401L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImageToBuffer.html>
402
403=item $ev = $queue->enqueue_task ($kernel, $wait_events...) 1254=item $ev = $queue->task ($kernel, $wait_events...)
404 1255
405L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html> 1256L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html>
406 1257
407=item $ev = $queue->enqueue_nd_range_kernel ($kernel, @$global_work_offset, @$global_work_size, @$local_work_size, $wait_events...) 1258=item $ev = $queue->nd_range_kernel ($kernel, \@global_work_offset, \@global_work_size, \@local_work_size, $wait_events...)
408 1259
409Enqueues a kernel execution. 1260Enqueues a kernel execution.
410 1261
411@$global_work_size must be specified as a reference to an array of 1262\@global_work_size must be specified as a reference to an array of
412integers specifying the work sizes (element counts). 1263integers specifying the work sizes (element counts).
413 1264
414@$global_work_offset must be either C<undef> (in which case all offsets 1265\@global_work_offset must be either C<undef> (in which case all offsets
415are C<0>), or a reference to an array of work offsets, with the same number 1266are C<0>), or a reference to an array of work offsets, with the same number
416of elements as @$global_work_size. 1267of elements as \@global_work_size.
417 1268
418@$local_work_size must be either C<undef> (in which case the 1269\@local_work_size must be either C<undef> (in which case the
419implementation is supposed to choose good local work sizes), or a 1270implementation is supposed to choose good local work sizes), or a
420reference to an array of local work sizes, with the same number of 1271reference to an array of local work sizes, with the same number of
421elements as @$global_work_size. 1272elements as \@global_work_size.
422 1273
423L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html> 1274L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html>
424 1275
425=item $ev = $queue->enqueue_marker 1276=item $ev = $queue->acquire_gl_objects ([object, ...], $wait_events...)
426 1277
1278Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired
1279for subsequent OpenCL usage.
1280
427L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMarker.html> 1281L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueAcquireGLObjects.html>
428 1282
1283=item $ev = $queue->release_gl_objects ([object, ...], $wait_events...)
1284
1285Enqueues a list (an array-ref of OpenCL::Memory objects) to be released
1286for subsequent OpenGL usage.
1287
1288L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReleaseGLObjects.html>
1289
429=item $ev = $queue->enqueue_wait_for_events ($wait_events...) 1290=item $ev = $queue->wait_for_events ($wait_events...)
430 1291
431L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html> 1292L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html>
432 1293
433=item $queue->enqueue_barrier 1294=item $ev = $queue->marker ($wait_events...)
434 1295
1296L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueMarkerWithWaitList.html>
1297
1298=item $ev = $queue->barrier ($wait_events...)
1299
435L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueBarrier.html> 1300L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueBarrierWithWaitList.html>
436 1301
437=item $queue->flush 1302=item $queue->flush
438 1303
439L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html> 1304L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html>
440 1305
441=item $queue->finish 1306=item $queue->finish
442 1307
443L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFinish.html> 1308L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFinish.html>
444 1309
1310=item $packed_value = $queue->info ($name)
1311
1312See C<< $platform->info >> for details.
1313
1314L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetCommandQueueInfo.html>
1315
1316=for gengetinfo begin command_queue
1317
1318=item $ctx = $command_queue->context
1319
1320Calls C<clGetCommandQueueInfo> with C<CL_QUEUE_CONTEXT> and returns the result.
1321
1322=item $device = $command_queue->device
1323
1324Calls C<clGetCommandQueueInfo> with C<CL_QUEUE_DEVICE> and returns the result.
1325
1326=item $uint = $command_queue->reference_count
1327
1328Calls C<clGetCommandQueueInfo> with C<CL_QUEUE_REFERENCE_COUNT> and returns the result.
1329
1330=item $command_queue_properties = $command_queue->properties
1331
1332Calls C<clGetCommandQueueInfo> with C<CL_QUEUE_PROPERTIES> and returns the result.
1333
1334=for gengetinfo end command_queue
1335
445=back 1336=back
446 1337
447=head2 THE OpenCL::Memory CLASS 1338=head2 THE OpenCL::Memory CLASS
448 1339
449This the superclass of all memory objects - OpenCL::Buffer, OpenCL::Image, 1340This the superclass of all memory objects - OpenCL::Buffer, OpenCL::Image,
450OpenCL::Image2D and OpenCL::Image3D. The subclasses of this class 1341OpenCL::Image2D and OpenCL::Image3D.
451currently only exist to allow type-checking.
452 1342
453=over 4 1343=over 4
454 1344
455=item $packed_value = $memory->info ($name) 1345=item $packed_value = $memory->info ($name)
456 1346
457See C<< $platform->info >> for details. 1347See C<< $platform->info >> for details.
458 1348
459L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetMemObjectInfo.html> 1349L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetMemObjectInfo.html>
460 1350
1351=for gengetinfo begin mem
1352
1353=item $mem_object_type = $mem->type
1354
1355Calls C<clGetMemObjectInfo> with C<CL_MEM_TYPE> and returns the result.
1356
1357=item $mem_flags = $mem->flags
1358
1359Calls C<clGetMemObjectInfo> with C<CL_MEM_FLAGS> and returns the result.
1360
1361=item $int = $mem->size
1362
1363Calls C<clGetMemObjectInfo> with C<CL_MEM_SIZE> and returns the result.
1364
1365=item $ptr_value = $mem->host_ptr
1366
1367Calls C<clGetMemObjectInfo> with C<CL_MEM_HOST_PTR> and returns the result.
1368
1369=item $uint = $mem->map_count
1370
1371Calls C<clGetMemObjectInfo> with C<CL_MEM_MAP_COUNT> and returns the result.
1372
1373=item $uint = $mem->reference_count
1374
1375Calls C<clGetMemObjectInfo> with C<CL_MEM_REFERENCE_COUNT> and returns the result.
1376
1377=item $ctx = $mem->context
1378
1379Calls C<clGetMemObjectInfo> with C<CL_MEM_CONTEXT> and returns the result.
1380
1381=item $mem = $mem->associated_memobject
1382
1383Calls C<clGetMemObjectInfo> with C<CL_MEM_ASSOCIATED_MEMOBJECT> and returns the result.
1384
1385=item $int = $mem->offset
1386
1387Calls C<clGetMemObjectInfo> with C<CL_MEM_OFFSET> and returns the result.
1388
1389=for gengetinfo end mem
1390
1391=item ($type, $name) = $mem->gl_object_info
1392
1393Returns the OpenGL object type (e.g. OpenCL::GL_OBJECT_TEXTURE2D) and the
1394object "name" (e.g. the texture name) used to create this memory object.
1395
1396L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetGLObjectInfo.html>
1397
461=back 1398=back
462 1399
1400=head2 THE OpenCL::Buffer CLASS
1401
1402This is a subclass of OpenCL::Memory, and the superclass of
1403OpenCL::BufferObj. Its purpose is simply to distinguish between buffers
1404and sub-buffers.
1405
1406=head2 THE OpenCL::BufferObj CLASS
1407
1408This is a subclass of OpenCL::Buffer and thus OpenCL::Memory. It exists
1409because one cna create sub buffers of OpenLC::BufferObj objects, but not
1410sub buffers from these sub buffers.
1411
1412=over 4
1413
1414=item $subbuf = $buf_obj->sub_buffer_region ($flags, $origin, $size)
1415
1416Creates an OpenCL::Buffer objects from this buffer and returns it. The
1417C<buffer_create_type> is assumed to be C<CL_BUFFER_CREATE_TYPE_REGION>.
1418
1419L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateSubBuffer.html>
1420
1421=back
1422
1423=head2 THE OpenCL::Image CLASS
1424
1425This is the superclass of all image objects - OpenCL::Image1D,
1426OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D,
1427OpenCL::Image2DArray and OpenCL::Image3D.
1428
1429=over 4
1430
1431=item $packed_value = $image->image_info ($name)
1432
1433See C<< $platform->info >> for details.
1434
1435The reason this method is not called C<info> is that there already is an
1436C<< ->info >> method inherited from C<OpenCL::Memory>.
1437
1438L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html>
1439
1440=item ($channel_order, $channel_data_type) = $image->format
1441
1442Returns the channel order and type used to create the image by calling
1443C<clGetImageInfo> with C<CL_IMAGE_FORMAT>.
1444
1445=for gengetinfo begin image
1446
1447=item $int = $image->element_size
1448
1449Calls C<clGetImageInfo> with C<CL_IMAGE_ELEMENT_SIZE> and returns the result.
1450
1451=item $int = $image->row_pitch
1452
1453Calls C<clGetImageInfo> with C<CL_IMAGE_ROW_PITCH> and returns the result.
1454
1455=item $int = $image->slice_pitch
1456
1457Calls C<clGetImageInfo> with C<CL_IMAGE_SLICE_PITCH> and returns the result.
1458
1459=item $int = $image->width
1460
1461Calls C<clGetImageInfo> with C<CL_IMAGE_WIDTH> and returns the result.
1462
1463=item $int = $image->height
1464
1465Calls C<clGetImageInfo> with C<CL_IMAGE_HEIGHT> and returns the result.
1466
1467=item $int = $image->depth
1468
1469Calls C<clGetImageInfo> with C<CL_IMAGE_DEPTH> and returns the result.
1470
1471=for gengetinfo end image
1472
1473=for gengetinfo begin gl_texture
1474
1475=item $GLenum = $gl_texture->target
1476
1477Calls C<clGetGLTextureInfo> with C<CL_GL_TEXTURE_TARGET> and returns the result.
1478
1479=item $GLint = $gl_texture->gl_mipmap_level
1480
1481Calls C<clGetGLTextureInfo> with C<CL_GL_MIPMAP_LEVEL> and returns the result.
1482
1483=for gengetinfo end gl_texture
1484
1485=back
1486
463=head2 THE OpenCL::Sampler CLASS 1487=head2 THE OpenCL::Sampler CLASS
464 1488
465=over 4 1489=over 4
466 1490
467=item $packed_value = $sampler->info ($name) 1491=item $packed_value = $sampler->info ($name)
468 1492
469See C<< $platform->info >> for details. 1493See C<< $platform->info >> for details.
470 1494
471L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetSamplerInfo.html> 1495L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetSamplerInfo.html>
472 1496
1497=for gengetinfo begin sampler
1498
1499=item $uint = $sampler->reference_count
1500
1501Calls C<clGetSamplerInfo> with C<CL_SAMPLER_REFERENCE_COUNT> and returns the result.
1502
1503=item $ctx = $sampler->context
1504
1505Calls C<clGetSamplerInfo> with C<CL_SAMPLER_CONTEXT> and returns the result.
1506
1507=item $addressing_mode = $sampler->normalized_coords
1508
1509Calls C<clGetSamplerInfo> with C<CL_SAMPLER_NORMALIZED_COORDS> and returns the result.
1510
1511=item $filter_mode = $sampler->addressing_mode
1512
1513Calls C<clGetSamplerInfo> with C<CL_SAMPLER_ADDRESSING_MODE> and returns the result.
1514
1515=item $boolean = $sampler->filter_mode
1516
1517Calls C<clGetSamplerInfo> with C<CL_SAMPLER_FILTER_MODE> and returns the result.
1518
1519=for gengetinfo end sampler
1520
473=back 1521=back
474 1522
475=head2 THE OpenCL::Program CLASS 1523=head2 THE OpenCL::Program CLASS
476 1524
477=over 4 1525=over 4
478 1526
479=item $packed_value = $program->info ($name) 1527=item $program->build (\@devices = undef, $options = "", $cb->($program) = undef)
480 1528
481See C<< $platform->info >> for details.
482
483L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProgramInfo.html>
484
485=item $program->build ($device, $options = "")
486
487Tries to build the program with the givne options. 1529Tries to build the program with the given options. See also the
1530C<$ctx->build> convenience function.
1531
1532If a callback is specified, then it will be called when compilation is
1533finished. Note that many OpenCL implementations block your program while
1534compiling whether you use a callback or not. See C<build_async> if you
1535want to make sure the build is done in the background.
1536
1537Note that some OpenCL implementations act up badly, and don't call the
1538callback in some error cases (but call it in others). This implementation
1539assumes the callback will always be called, and leaks memory if this is
1540not so. So best make sure you don't pass in invalid values.
1541
1542Some implementations fail with C<OpenCL::INVALID_BINARY> when the
1543compilation state is successful but some later stage fails.
488 1544
489L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html> 1545L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html>
1546
1547=item $program->build_async (\@devices = undef, $options = "", $cb->($program) = undef)
1548
1549Similar to C<< ->build >>, except it starts a thread, and never fails (you
1550need to check the compilation status form the callback, or by polling).
490 1551
491=item $packed_value = $program->build_info ($device, $name) 1552=item $packed_value = $program->build_info ($device, $name)
492 1553
493Similar to C<< $platform->info >>, but returns build info for a previous 1554Similar to C<< $platform->info >>, but returns build info for a previous
494build attempt for the given device. 1555build attempt for the given device.
500Creates an OpenCL::Kernel object out of the named C<__kernel> function in 1561Creates an OpenCL::Kernel object out of the named C<__kernel> function in
501the program. 1562the program.
502 1563
503L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html> 1564L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html>
504 1565
1566=item @kernels = $program->kernels_in_program
1567
1568Returns all kernels successfully compiled for all devices in program.
1569
1570http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernelsInProgram.html
1571
1572=for gengetinfo begin program_build
1573
1574=item $build_status = $program->build_status ($device)
1575
1576Calls C<clGetProgramBuildInfo> with C<CL_PROGRAM_BUILD_STATUS> and returns the result.
1577
1578=item $string = $program->build_options ($device)
1579
1580Calls C<clGetProgramBuildInfo> with C<CL_PROGRAM_BUILD_OPTIONS> and returns the result.
1581
1582=item $string = $program->build_log ($device)
1583
1584Calls C<clGetProgramBuildInfo> with C<CL_PROGRAM_BUILD_LOG> and returns the result.
1585
1586=for gengetinfo end program_build
1587
1588=item $packed_value = $program->info ($name)
1589
1590See C<< $platform->info >> for details.
1591
1592L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProgramInfo.html>
1593
1594=for gengetinfo begin program
1595
1596=item $uint = $program->reference_count
1597
1598Calls C<clGetProgramInfo> with C<CL_PROGRAM_REFERENCE_COUNT> and returns the result.
1599
1600=item $ctx = $program->context
1601
1602Calls C<clGetProgramInfo> with C<CL_PROGRAM_CONTEXT> and returns the result.
1603
1604=item $uint = $program->num_devices
1605
1606Calls C<clGetProgramInfo> with C<CL_PROGRAM_NUM_DEVICES> and returns the result.
1607
1608=item @devices = $program->devices
1609
1610Calls C<clGetProgramInfo> with C<CL_PROGRAM_DEVICES> and returns the result.
1611
1612=item $string = $program->source
1613
1614Calls C<clGetProgramInfo> with C<CL_PROGRAM_SOURCE> and returns the result.
1615
1616=item @ints = $program->binary_sizes
1617
1618Calls C<clGetProgramInfo> with C<CL_PROGRAM_BINARY_SIZES> and returns the result.
1619
1620=for gengetinfo end program
1621
1622=item @blobs = $program->binaries
1623
1624Returns a string for the compiled binary for every device associated with
1625the program, empty strings indicate missing programs, and an empty result
1626means no program binaries are available.
1627
1628These "binaries" are often, in fact, informative low-level assembly
1629sources.
1630
1631L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProgramInfo.html>
1632
505=back 1633=back
506 1634
507=head2 THE OpenCL::Kernel CLASS 1635=head2 THE OpenCL::Kernel CLASS
508 1636
509=over 4 1637=over 4
512 1640
513See C<< $platform->info >> for details. 1641See C<< $platform->info >> for details.
514 1642
515L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetKernelInfo.html> 1643L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetKernelInfo.html>
516 1644
1645=for gengetinfo begin kernel
1646
1647=item $string = $kernel->function_name
1648
1649Calls C<clGetKernelInfo> with C<CL_KERNEL_FUNCTION_NAME> and returns the result.
1650
1651=item $uint = $kernel->num_args
1652
1653Calls C<clGetKernelInfo> with C<CL_KERNEL_NUM_ARGS> and returns the result.
1654
1655=item $uint = $kernel->reference_count
1656
1657Calls C<clGetKernelInfo> with C<CL_KERNEL_REFERENCE_COUNT> and returns the result.
1658
1659=item $ctx = $kernel->context
1660
1661Calls C<clGetKernelInfo> with C<CL_KERNEL_CONTEXT> and returns the result.
1662
1663=item $program = $kernel->program
1664
1665Calls C<clGetKernelInfo> with C<CL_KERNEL_PROGRAM> and returns the result.
1666
1667=for gengetinfo end kernel
1668
1669=item $packed_value = $kernel->work_group_info ($device, $name)
1670
1671See C<< $platform->info >> for details.
1672
1673The reason this method is not called C<info> is that there already is an
1674C<< ->info >> method.
1675
1676L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetKernelWorkGroupInfo.html>
1677
1678=for gengetinfo begin kernel_work_group
1679
1680=item $int = $kernel->work_group_size ($device)
1681
1682Calls C<clGetKernelWorkGroupInfo> with C<CL_KERNEL_WORK_GROUP_SIZE> and returns the result.
1683
1684=item @ints = $kernel->compile_work_group_size ($device)
1685
1686Calls C<clGetKernelWorkGroupInfo> with C<CL_KERNEL_COMPILE_WORK_GROUP_SIZE> and returns the result.
1687
1688=item $ulong = $kernel->local_mem_size ($device)
1689
1690Calls C<clGetKernelWorkGroupInfo> with C<CL_KERNEL_LOCAL_MEM_SIZE> and returns the result.
1691
1692=item $int = $kernel->preferred_work_group_size_multiple ($device)
1693
1694Calls C<clGetKernelWorkGroupInfo> with C<CL_KERNEL_PREFERRED_WORK_GROUP_SIZE_MULTIPLE> and returns the result.
1695
1696=item $ulong = $kernel->private_mem_size ($device)
1697
1698Calls C<clGetKernelWorkGroupInfo> with C<CL_KERNEL_PRIVATE_MEM_SIZE> and returns the result.
1699
1700=for gengetinfo end kernel_work_group
1701
1702=item $kernel->setf ($format, ...)
1703
1704Sets the arguments of a kernel. Since OpenCL 1.1 doesn't have a generic
1705way to set arguments (and with OpenCL 1.2 it might be rather slow), you
1706need to specify a format argument, much as with C<printf>, to tell OpenCL
1707what type of argument it is.
1708
1709The format arguments are single letters:
1710
1711 c char
1712 C unsigned char
1713 s short
1714 S unsigned short
1715 i int
1716 I unsigned int
1717 l long
1718 L unsigned long
1719
1720 h half float (0..65535)
1721 f float
1722 d double
1723
1724 z local (octet size)
1725
1726 m memory object (buffer or image)
1727 a sampler
1728 e event
1729
1730Space characters in the format string are ignored.
1731
1732Example: set the arguments for a kernel that expects an int, two floats, a buffer and an image.
1733
1734 $kernel->setf ("i ff mm", 5, 0.5, 3, $buffer, $image);
1735
517=item $kernel->set_TYPE ($index, $value) 1736=item $kernel->set_TYPE ($index, $value)
518 1737
1738=item $kernel->set_char ($index, $value)
1739
1740=item $kernel->set_uchar ($index, $value)
1741
1742=item $kernel->set_short ($index, $value)
1743
1744=item $kernel->set_ushort ($index, $value)
1745
1746=item $kernel->set_int ($index, $value)
1747
1748=item $kernel->set_uint ($index, $value)
1749
1750=item $kernel->set_long ($index, $value)
1751
1752=item $kernel->set_ulong ($index, $value)
1753
1754=item $kernel->set_half ($index, $value)
1755
1756=item $kernel->set_float ($index, $value)
1757
1758=item $kernel->set_double ($index, $value)
1759
1760=item $kernel->set_memory ($index, $value)
1761
1762=item $kernel->set_buffer ($index, $value)
1763
1764=item $kernel->set_image ($index, $value)
1765
1766=item $kernel->set_sampler ($index, $value)
1767
1768=item $kernel->set_local ($index, $value)
1769
1770=item $kernel->set_event ($index, $value)
1771
519This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>. 1772This is a family of methods to set the kernel argument with the number
520 1773C<$index> to the give C<$value>.
521TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>,
522C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>,
523C<image2d>, C<image3d>, C<sampler> or C<event>.
524 1774
525Chars and integers (including the half type) are specified as integers, 1775Chars and integers (including the half type) are specified as integers,
526float and double as floating point values, memory/buffer/image2d/image3d 1776float and double as floating point values, memory/buffer/image must be
527must be an object of that type or C<undef>, and sampler and event must be 1777an object of that type or C<undef>, local-memory arguments are set by
528objects of that type. 1778specifying the size, and sampler and event must be objects of that type.
1779
1780Note that C<set_memory> works for all memory objects (all types of buffers
1781and images) - the main purpose of the more specific C<set_TYPE> functions
1782is type checking.
1783
1784Setting an argument for a kernel does NOT keep a reference to the object -
1785for example, if you set an argument to some image object, free the image,
1786and call the kernel, you will run into undefined behaviour.
529 1787
530L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html> 1788L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html>
531 1789
532=back 1790=back
533 1791
536This is the superclass for all event objects (including OpenCL::UserEvent 1794This is the superclass for all event objects (including OpenCL::UserEvent
537objects). 1795objects).
538 1796
539=over 4 1797=over 4
540 1798
1799=item $ev->wait
1800
1801Waits for the event to complete.
1802
1803L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
1804
1805=item $ev->cb ($exec_callback_type, $callback->($event, $event_command_exec_status))
1806
1807Adds a callback to the callback stack for the given event type. There is
1808no way to remove a callback again.
1809
1810L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetEventCallback.html>
1811
541=item $packed_value = $ev->info ($name) 1812=item $packed_value = $ev->info ($name)
542 1813
543See C<< $platform->info >> for details. 1814See C<< $platform->info >> for details.
544 1815
545L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html> 1816L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html>
546 1817
547=item $ev->wait 1818=for gengetinfo begin event
548 1819
549Waits for the event to complete. 1820=item $queue = $event->command_queue
550 1821
1822Calls C<clGetEventInfo> with C<CL_EVENT_COMMAND_QUEUE> and returns the result.
1823
1824=item $command_type = $event->command_type
1825
1826Calls C<clGetEventInfo> with C<CL_EVENT_COMMAND_TYPE> and returns the result.
1827
1828=item $uint = $event->reference_count
1829
1830Calls C<clGetEventInfo> with C<CL_EVENT_REFERENCE_COUNT> and returns the result.
1831
1832=item $uint = $event->command_execution_status
1833
1834Calls C<clGetEventInfo> with C<CL_EVENT_COMMAND_EXECUTION_STATUS> and returns the result.
1835
1836=item $ctx = $event->context
1837
1838Calls C<clGetEventInfo> with C<CL_EVENT_CONTEXT> and returns the result.
1839
1840=for gengetinfo end event
1841
1842=item $packed_value = $ev->profiling_info ($name)
1843
1844See C<< $platform->info >> for details.
1845
1846The reason this method is not called C<info> is that there already is an
1847C<< ->info >> method.
1848
551L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 1849L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProfilingInfo.html>
1850
1851=for gengetinfo begin profiling
1852
1853=item $ulong = $event->profiling_command_queued
1854
1855Calls C<clGetEventProfilingInfo> with C<CL_PROFILING_COMMAND_QUEUED> and returns the result.
1856
1857=item $ulong = $event->profiling_command_submit
1858
1859Calls C<clGetEventProfilingInfo> with C<CL_PROFILING_COMMAND_SUBMIT> and returns the result.
1860
1861=item $ulong = $event->profiling_command_start
1862
1863Calls C<clGetEventProfilingInfo> with C<CL_PROFILING_COMMAND_START> and returns the result.
1864
1865=item $ulong = $event->profiling_command_end
1866
1867Calls C<clGetEventProfilingInfo> with C<CL_PROFILING_COMMAND_END> and returns the result.
1868
1869=for gengetinfo end profiling
552 1870
553=back 1871=back
554 1872
555=head2 THE OpenCL::UserEvent CLASS 1873=head2 THE OpenCL::UserEvent CLASS
556 1874
558 1876
559=over 4 1877=over 4
560 1878
561=item $ev->set_status ($execution_status) 1879=item $ev->set_status ($execution_status)
562 1880
1881Sets the execution status of the user event. Can only be called once,
1882either with OpenCL::COMPLETE or a negative number as status.
1883
563L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html> 1884L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html>
564 1885
565=back 1886=back
566 1887
567=cut 1888=cut
568
569package OpenCL;
570
571use common::sense;
572
573BEGIN {
574 our $VERSION = '0.03';
575
576 require XSLoader;
577 XSLoader::load (__PACKAGE__, $VERSION);
578
579 @OpenCL::Buffer::ISA =
580 @OpenCL::Image::ISA = OpenCL::Memory::;
581
582 @OpenCL::Image2D::ISA =
583 @OpenCL::Image3D::ISA = OpenCL::Image::;
584
585 @OpenCL::UserEvent::ISA = OpenCL::Event::;
586}
587 1889
5881; 18901;
589 1891
590=head1 AUTHOR 1892=head1 AUTHOR
591 1893

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines