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

Comparing OpenCL/OpenCL.pm (file contents):
Revision 1.9 by root, Thu Nov 17 01:36:52 2011 UTC vs.
Revision 1.72 by root, Fri May 4 14:49:42 2012 UTC

18vendors) - usually there is only one. 18vendors) - usually there is only one.
19 19
20Each platform gives you access to a number of OpenCL::Device objects, e.g. 20Each platform gives you access to a number of OpenCL::Device objects, e.g.
21your graphics card. 21your graphics card.
22 22
23From a platform and some devices, you create an OpenCL::Context, which is 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 24a very central object in OpenCL: Once you have a context you can create
25most other objects: 25most other objects:
26 26
27OpenCL::Program objects, which store source code and, after building 27OpenCL::Program objects, which store source code and, after building for a
28("compiling and linking"), also binary programs. For each kernel function 28specific device ("compiling and linking"), also binary programs. For each
29in a program you can then create an OpenCL::Kernel object which represents 29kernel function in a program you can then create an OpenCL::Kernel object
30basically a function call with argument values. 30which represents basically a function call with argument values.
31 31
32OpenCL::Memory objects of various flavours: OpenCL::Buffers objects (flat 32OpenCL::Memory objects of various flavours: OpenCL::Buffer objects (flat
33memory areas, think array) and OpenCL::Image objects (think 2d or 3d 33memory areas, think arrays or structs) and OpenCL::Image objects (think 2d
34array) for bulk data and input and output for kernels. 34or 3d array) for bulk data and input and output for kernels.
35 35
36OpenCL::Sampler objects, which are kind of like texture filter modes in 36OpenCL::Sampler objects, which are kind of like texture filter modes in
37OpenGL. 37OpenGL.
38 38
39OpenCL::Queue objects - command queues, which allow you to submit memory 39OpenCL::Queue objects - command queues, which allow you to submit memory
43 43
44OpenCL::Event objects are used to signal when something is complete. 44OpenCL::Event objects are used to signal when something is complete.
45 45
46=head2 HELPFUL RESOURCES 46=head2 HELPFUL RESOURCES
47 47
48The OpenCL spec used to develop this module (1.2 spec was available, but 48The OpenCL specs used to develop this module - download these and keept
49no implementation was available to me :). 49hema round, they are required reference material:
50 50
51 http://www.khronos.org/registry/cl/specs/opencl-1.1.pdf 51 http://www.khronos.org/registry/cl/specs/opencl-1.1.pdf
52 http://www.khronos.org/registry/cl/specs/opencl-1.2.pdf
53 http://www.khronos.org/registry/cl/specs/opencl-1.2-extensions.pdf
52 54
53OpenCL manpages: 55OpenCL manpages:
54 56
55 http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/ 57 http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/
58 http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/
59
60If you are into UML class diagrams, the following diagram might help - if
61not, it will be mildly confusing (also, the class hierarchy of this module
62is much more fine-grained):
63
64 http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/classDiagram.html
65
66Here's a tutorial from AMD (very AMD-centric, too), not sure how useful it
67is, but at least it's free of charge:
68
69 http://developer.amd.com/zones/OpenCLZone/courses/Documents/Introduction_to_OpenCL_Programming%20Training_Guide%20%28201005%29.pdf
70
71And here's NVIDIA's OpenCL Best Practises Guide:
72
73 http://developer.download.nvidia.com/compute/cuda/3_2/toolkit/docs/OpenCL_Best_Practices_Guide.pdf
56 74
57=head1 BASIC WORKFLOW 75=head1 BASIC WORKFLOW
58 76
59To get something done, you basically have to do this once: 77To get something done, you basically have to do this once (refer to the
78examples below for actual code, this is just a high-level description):
60 79
61Find some platform (e.g. the first one) and some device (e.g. the first 80Find some platform (e.g. the first one) and some device(s) (e.g. the first
62device you can find), and create a context from those. 81device of the platform), and create a context from those.
63 82
64Create a command queue from your context, and program objects from your 83Create program objects from your OpenCL source code, then build (compile)
65OpenCL source code, build the programs. 84the programs for each device you want to run them on.
66 85
67Create kernel objects for all kernels you want to use. 86Create kernel objects for all kernels you want to use (surprisingly, these
87are not device-specific).
68 88
69Then, to execute stuff, you repeat this: 89Then, to execute stuff, you repeat these steps, possibly resuing or
90sharing some buffers:
70 91
71Create some input and output buffers from your context. Initialise the 92Create some input and output buffers from your context. Set these as
72input buffers with data. Set these as arguments to your kernel. 93arguments to your kernel.
94
95Enqueue buffer writes to initialise your input buffers (when not
96initialised at creation time).
73 97
74Enqueue the kernel execution. 98Enqueue the kernel execution.
75 99
76Enqueue buffer reads for your output buffer to read results. 100Enqueue buffer reads for your output buffer to read results.
77 101
78The next section shows how this can be done.
79
80=head1 EXAMPLES 102=head1 EXAMPLES
81 103
82=head2 Enumerate all devices and get contexts for them. 104=head2 Enumerate all devices and get contexts for them.
83 105
106Best run this once to get a feel for the platforms and devices in your
107system.
108
84 for my $platform (OpenCL::platforms) { 109 for my $platform (OpenCL::platforms) {
85 warn $platform->info (OpenCL::PLATFORM_NAME); 110 printf "platform: %s\n", $platform->name;
86 warn $platform->info (OpenCL::PLATFORM_EXTENSIONS); 111 printf "extensions: %s\n", $platform->extensions;
87 for my $device ($platform->devices) { 112 for my $device ($platform->devices) {
88 warn $device->info (OpenCL::DEVICE_NAME); 113 printf "+ device: %s\n", $device->name;
89 my $ctx = $device->context_simple; 114 my $ctx = $platform->context (undef, [$device]);
90 # do stuff 115 # do stuff
91 } 116 }
92 } 117 }
93 118
94=head2 Get a useful context and a command queue. 119=head2 Get a useful context and a command queue.
95 120
96 my $dev = ((OpenCL::platforms)[0]->devices)[0]; 121This is a useful boilerplate for any OpenCL program that only wants to use
97 my $ctx = $dev->context_simple; 122one device,
98 my $queue = $ctx->queue ($dev); 123
124 my ($platform) = OpenCL::platforms; # find first platform
125 my ($dev) = $platform->devices; # find first device of platform
126 my $ctx = $platform->context (undef, [$dev]); # create context out of those
127 my $queue = $ctx->queue ($dev); # create a command queue for the device
99 128
100=head2 Print all supported image formats of a context. 129=head2 Print all supported image formats of a context.
101 130
131Best run this once for your context, to see whats available and how to
132gather information.
133
102 for my $type (OpenCL::MEM_OBJECT_IMAGE2D, OpenCL::MEM_OBJECT_IMAGE3D) { 134 for my $type (OpenCL::MEM_OBJECT_IMAGE2D, OpenCL::MEM_OBJECT_IMAGE3D) {
103 say "supported image formats for ", OpenCL::enum2str $type; 135 print "supported image formats for ", OpenCL::enum2str $type, "\n";
104 136
105 for my $f ($ctx->supported_image_formats (0, $type)) { 137 for my $f ($ctx->supported_image_formats (0, $type)) {
106 printf " %-10s %-20s\n", OpenCL::enum2str $f->[0], OpenCL::enum2str $f->[1]; 138 printf " %-10s %-20s\n", OpenCL::enum2str $f->[0], OpenCL::enum2str $f->[1];
107 } 139 }
108 } 140 }
110=head2 Create a buffer with some predefined data, read it back synchronously, 142=head2 Create a buffer with some predefined data, read it back synchronously,
111then asynchronously. 143then asynchronously.
112 144
113 my $buf = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, "helmut"); 145 my $buf = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, "helmut");
114 146
115 $queue->enqueue_read_buffer ($buf, 1, 1, 3, my $data); 147 $queue->read_buffer ($buf, 1, 1, 3, my $data);
116 warn $data; 148 print "$data\n";
117 149
118 my $ev = $queue->enqueue_read_buffer ($buf, 0, 1, 3, my $data); 150 my $ev = $queue->read_buffer ($buf, 0, 1, 3, my $data);
119 $ev->wait; 151 $ev->wait;
120 warn $data; 152 print "$data\n"; # prints "elm"
121 153
122=head2 Create and build a program, then create a kernel out of one of its 154=head2 Create and build a program, then create a kernel out of one of its
123functions. 155functions.
124 156
125 my $src = ' 157 my $src = '
126 __kernel void 158 kernel void
127 squareit (__global float *input, __global float *output) 159 squareit (global float *input, global float *output)
128 { 160 {
129 size_t id = get_global_id (0); 161 $id = get_global_id (0);
130 output [id] = input [id] * input [id]; 162 output [id] = input [id] * input [id];
131 } 163 }
132 '; 164 ';
133 165
134 my $prog = $ctx->program_with_source ($src); 166 my $prog = $ctx->build_program ($src);
135
136 eval { $prog->build ($dev); 1 }
137 or die $prog->build_info ($dev, OpenCL::PROGRAM_BUILD_LOG);
138
139 my $kernel = $prog->kernel ("squareit"); 167 my $kernel = $prog->kernel ("squareit");
140 168
141=head2 Create some input and output float buffers, then call squareit on them. 169=head2 Create some input and output float buffers, then call the
170'squareit' kernel on them.
142 171
143 my $input = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, pack "f*", 1, 2, 3, 4.5); 172 my $input = $ctx->buffer_sv (OpenCL::MEM_COPY_HOST_PTR, pack "f*", 1, 2, 3, 4.5);
144 my $output = $ctx->buffer (0, OpenCL::SIZEOF_FLOAT * 5); 173 my $output = $ctx->buffer (0, OpenCL::SIZEOF_FLOAT * 5);
145 174
146 # set buffer 175 # set buffer
147 $kernel->set_buffer (0, $input); 176 $kernel->set_buffer (0, $input);
148 $kernel->set_buffer (1, $output); 177 $kernel->set_buffer (1, $output);
149 178
150 # execute it for all 4 numbers 179 # execute it for all 4 numbers
151 $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 180 $queue->nd_range_kernel ($kernel, undef, [4], undef);
152 181
153 # enqueue a synchronous read 182 # enqueue a synchronous read
154 $queue->enqueue_read_buffer ($output, 1, 0, OpenCL::SIZEOF_FLOAT * 4, my $data); 183 $queue->read_buffer ($output, 1, 0, OpenCL::SIZEOF_FLOAT * 4, my $data);
155 184
156 # print the results: 185 # print the results:
157 say join ", ", unpack "f*", $data; 186 printf "%s\n", join ", ", unpack "f*", $data;
158 187
159=head2 The same enqueue operations as before, but assuming an out-of-order queue, 188=head2 The same enqueue operations as before, but assuming an out-of-order queue,
160showing off barriers. 189showing off barriers.
161 190
162 # execute it for all 4 numbers 191 # execute it for all 4 numbers
163 $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 192 $queue->nd_range_kernel ($kernel, undef, [4], undef);
164 193
165 # enqueue a barrier to ensure in-order execution 194 # enqueue a barrier to ensure in-order execution
166 $queue->enqueue_barrier; 195 $queue->barrier;
167 196
168 # enqueue an async read 197 # enqueue an async read
169 $queue->enqueue_read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data); 198 $queue->read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data);
170 199
171 # wait for all requests to finish 200 # wait for all requests to finish
172 $queue->finish; 201 $queue->finish;
173 202
174=head2 The same enqueue operations as before, but assuming an out-of-order queue, 203=head2 The same enqueue operations as before, but assuming an out-of-order queue,
175showing off event objects and wait lists. 204showing off event objects and wait lists.
176 205
177 # execute it for all 4 numbers 206 # execute it for all 4 numbers
178 my $ev = $queue->enqueue_nd_range_kernel ($kernel, undef, [4], undef); 207 my $ev = $queue->nd_range_kernel ($kernel, undef, [4], undef);
179 208
180 # enqueue an async read 209 # enqueue an async read
181 $ev = $queue->enqueue_read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data, $ev); 210 $ev = $queue->read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data, $ev);
182 211
183 # wait for the last event to complete 212 # wait for the last event to complete
184 $ev->wait; 213 $ev->wait;
185 214
215=head2 Use the OpenGL module to share a texture between OpenCL and OpenGL and draw some julia
216set flight effect.
217
218This is quite a long example to get you going - you can download it from
219L<http://cvs.schmorp.de/OpenCL/examples/juliaflight>.
220
221 use OpenGL ":all";
222 use OpenCL;
223
224 my $S = $ARGV[0] || 256; # window/texture size, smaller is faster
225
226 # open a window and create a gl texture
227 OpenGL::glpOpenWindow width => $S, height => $S;
228 my $texid = glGenTextures_p 1;
229 glBindTexture GL_TEXTURE_2D, $texid;
230 glTexImage2D_c GL_TEXTURE_2D, 0, GL_RGBA8, $S, $S, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0;
231
232 # find and use the first opencl device that let's us get a shared opengl context
233 my $platform;
234 my $dev;
235 my $ctx;
236
237 for (OpenCL::platforms) {
238 $platform = $_;
239 for ($platform->devices) {
240 $dev = $_;
241 $ctx = $platform->context ([OpenCL::GLX_DISPLAY_KHR, undef, OpenCL::GL_CONTEXT_KHR, undef], [$dev])
242 and last;
243 }
244 }
245
246 $ctx
247 or die "cannot find suitable OpenCL device\n";
248
249 my $queue = $ctx->queue ($dev);
250
251 # now attach an opencl image2d object to the opengl texture
252 my $tex = $ctx->gl_texture2d (OpenCL::MEM_WRITE_ONLY, GL_TEXTURE_2D, 0, $texid);
253
254 # now the boring opencl code
255 my $src = <<EOF;
256 kernel void
257 juliatunnel (write_only image2d_t img, float time)
258 {
259 int2 xy = (int2)(get_global_id (0), get_global_id (1));
260 float2 p = convert_float2 (xy) / $S.f * 2.f - 1.f;
261
262 float2 m = (float2)(1.f, p.y) / fabs (p.x); // tunnel
263 m.x = fabs (fmod (m.x + time * 0.05f, 4.f) - 2.f);
264
265 float2 z = m;
266 float2 c = (float2)(sin (time * 0.01133f), cos (time * 0.02521f));
267
268 for (int i = 0; i < 25 && dot (z, z) < 4.f; ++i) // standard julia
269 z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c;
270
271 float3 colour = (float3)(z.x, z.y, atan2 (z.y, z.x));
272 write_imagef (img, xy, (float4)(colour * p.x * p.x, 1.));
273 }
274 EOF
275
276 my $prog = $ctx->build_program ($src);
277 my $kernel = $prog->kernel ("juliatunnel");
278
279 # program compiled, kernel ready, now draw and loop
280
281 for (my $time; ; ++$time) {
282 # acquire objects from opengl
283 $queue->acquire_gl_objects ([$tex]);
284
285 # configure and run our kernel
286 $kernel->setf ("mf", $tex, $time*2); # mf = memory object, float
287 $queue->nd_range_kernel ($kernel, undef, [$S, $S], undef);
288
289 # release objects to opengl again
290 $queue->release_gl_objects ([$tex]);
291
292 # wait
293 $queue->finish;
294
295 # now draw the texture, the defaults should be all right
296 glTexParameterf GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST;
297
298 glEnable GL_TEXTURE_2D;
299 glBegin GL_QUADS;
300 glTexCoord2f 0, 1; glVertex3i -1, -1, -1;
301 glTexCoord2f 0, 0; glVertex3i 1, -1, -1;
302 glTexCoord2f 1, 0; glVertex3i 1, 1, -1;
303 glTexCoord2f 1, 1; glVertex3i -1, 1, -1;
304 glEnd;
305
306 glXSwapBuffers;
307
308 select undef, undef, undef, 1/60;
309 }
310
311=head2 How to modify the previous example to not rely on GL sharing.
312
313For those poor souls with only a sucky CPU OpenCL implementation, you
314currently have to read the image into some perl scalar, and then modify a
315texture or use glDrawPixels or so).
316
317First, when you don't need gl sharing, you can create the context much simpler:
318
319 $ctx = $platform->context (undef, [$dev])
320
321To use a texture, you would modify the above example by creating an
322OpenCL::Image manually instead of deriving it from a texture:
323
324 my $tex = $ctx->image2d (OpenCL::MEM_WRITE_ONLY, OpenCL::RGBA, OpenCL::UNORM_INT8, $S, $S);
325
326And in the draw loop, intead of acquire_gl_objects/release_gl_objects, you
327would read the image2d after the kernel has written it:
328
329 $queue->read_image ($tex, 0, 0, 0, 0, $S, $S, 1, 0, 0, my $data);
330
331And then you would upload the pixel data to the texture (or use glDrawPixels):
332
333 glTexSubImage2D_s GL_TEXTURE_2D, 0, 0, 0, $S, $S, GL_RGBA, GL_UNSIGNED_BYTE, $data;
334
335The fully modified example can be found at
336L<http://cvs.schmorp.de/OpenCL/examples/juliaflight-nosharing>.
337
338=head2 Julia sets look soooo 80ies.
339
340Then colour them differently, e.g. using orbit traps! Replace the loop and
341colour calculation from the previous examples by this:
342
343 float2 dm = (float2)(1.f, 1.f);
344
345 for (int i = 0; i < 25; ++i)
346 {
347 z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c;
348 dm = fmin (dm, (float2)(fabs (dot (z, z) - 1.f), fabs (z.x - 1.f)));
349 }
350
351 float3 colour = (float3)(dm.x * dm.y, dm.x * dm.y, dm.x);
352
353Also try C<-10.f> instead of C<-1.f>.
354
186=head1 DOCUMENTATION 355=head1 DOCUMENTATION
187 356
188=head2 BASIC CONVENTIONS 357=head2 BASIC CONVENTIONS
189 358
190This is not a 1:1 C-style translation of OpenCL to Perl - instead I 359This is not a one-to-one C-style translation of OpenCL to Perl - instead
191attempted to make the interface as type-safe as possible and introducing 360I attempted to make the interface as type-safe as possible by introducing
192object syntax where it makes sense. There are a number of important 361object syntax where it makes sense. There are a number of important
193differences between the OpenCL C API and this module: 362differences between the OpenCL C API and this module:
194 363
195=over 4 364=over 4
196 365
197=item * Object lifetime managament is automatic - there is no need 366=item * Object lifetime managament is automatic - there is no need
198to free objects explicitly (C<clReleaseXXX>), the release function 367to free objects explicitly (C<clReleaseXXX>), the release function
199is called automatically once all Perl references to it go away. 368is called automatically once all Perl references to it go away.
200 369
201=item * OpenCL uses CamelCase for function names (C<clGetPlatformInfo>), 370=item * OpenCL uses CamelCase for function names
371(e.g. C<clGetPlatformIDs>, C<clGetPlatformInfo>), while this module
202while this module uses underscores as word separator and often leaves out 372uses underscores as word separator and often leaves out prefixes
203prefixes (C<< $platform->info >>). 373(C<OpenCL::platforms>, C<< $platform->info >>).
204 374
205=item * OpenCL often specifies fixed vector function arguments as short 375=item * OpenCL often specifies fixed vector function arguments as short
206arrays (C<size_t origin[3]>), while this module explicitly expects the 376arrays (C<size_t origin[3]>), while this module explicitly expects the
207components as separate arguments- 377components as separate arguments (C<$orig_x, $orig_y, $orig_z>) in
378function calls.
208 379
209=item * Where possible, one of the pitch values is calculated from the 380=item * Structures are often specified by flattening out their components
210perl scalar length and need not be specified. 381as with short vectors, and returned as arrayrefs.
211 382
212=item * When enqueuing commands, the wait list is specified by adding 383=item * When enqueuing commands, the wait list is specified by adding
213extra arguments to the function - anywhere a C<$wait_events...> argument 384extra arguments to the function - anywhere a C<$wait_events...> argument
214is documented this can be any number of event objects. 385is documented this can be any number of event objects. As an extsnion
386implemented by this module, C<undef> values will be ignored in the event
387list.
215 388
216=item * When enqueuing commands, if the enqueue method is called in void 389=item * When enqueuing commands, if the enqueue method is called in void
217context, no event is created. In all other contexts an event is returned 390context, no event is created. In all other contexts an event is returned
218by the method. 391by the method.
219 392
239 ulong IV - Q 412 ulong IV - Q
240 float NV float f 413 float NV float f
241 half IV ushort S 414 half IV ushort S
242 double NV double d 415 double NV double d
243 416
417=head2 GLX SUPPORT
418
419Due to the sad state that OpenGL support is in in Perl (mostly the OpenGL
420module, which has little to no documentation and has little to no support
421for glX), this module, as a special extension, treats context creation
422properties C<OpenCL::GLX_DISPLAY_KHR> and C<OpenCL::GL_CONTEXT_KHR>
423specially: If either or both of these are C<undef>, then the OpenCL
424module tries to dynamically resolve C<glXGetCurrentDisplay> and
425C<glXGetCurrentContext>, call these functions and use their return values
426instead.
427
428For this to work, the OpenGL library must be loaded, a GLX context must
429have been created and be made current, and C<dlsym> must be available and
430capable of finding the function via C<RTLD_DEFAULT>.
431
432=head2 EVENT SYSTEM
433
434OpenCL can generate a number of (potentially) asynchronous events, for
435example, after compiling a program, to signal a context-related error or,
436perhaps most important, to signal completion of queued jobs (by setting
437callbacks on OpenCL::Event objects).
438
439To facilitate this, this module maintains an event queue - each
440time an asynchronous event happens, it is queued, and perl will be
441interrupted. This is implemented via the L<Async::Interrupt> module. In
442addition, this module has L<AnyEvent> support, so it can seamlessly
443integrate itself into many event loops.
444
445Since this module is a bit hard to understand, here are some case examples:
446
447=head3 Don't use callbacks.
448
449When your program never uses any callbacks, then there will never be any
450notifications you need to take care of, and therefore no need to worry
451about all this.
452
453You can achieve a great deal by explicitly waiting for events, or using
454barriers and flush calls. In many programs, there is no need at all to
455tinker with asynchronous events.
456
457=head3 Use AnyEvent
458
459This module automatically registers a watcher that invokes all outstanding
460event callbacks when AnyEvent is initialised (and block asynchronous
461interruptions). Using this mode of operations is the safest and most
462recommended one.
463
464To use this, simply use AnyEvent and this module normally, make sure you
465have an event loop running:
466
467 use Gtk2 -init;
468 use AnyEvent;
469
470 # initialise AnyEvent, by creating a watcher, or:
471 AnyEvent::detect;
472
473 my $e = $queue->marker;
474 $e->cb (sub {
475 warn "opencl is finished\n";
476 })
477
478 main Gtk2;
479
480Note that this module will not initialise AnyEvent for you. Before
481AnyEvent is initialised, the module will asynchronously interrupt perl
482instead. To avoid any surprises, it's best to explicitly initialise
483AnyEvent.
484
485You can temporarily enable asynchronous interruptions (see next paragraph)
486by calling C<$OpenCL::INTERRUPT->unblock> and disable them again by
487calling C<$OpenCL::INTERRUPT->block>.
488
489=head3 Let yourself be interrupted at any time
490
491This mode is the default unless AnyEvent is loaded and initialised. In
492this mode, OpenCL asynchronously interrupts a running perl program. The
493emphasis is on both I<asynchronously> and I<running> here.
494
495Asynchronously means that perl might execute your callbacks at any
496time. For example, in the following code (I<THAT YOU SHOULD NOT COPY>),
497the C<until> loop following the marker call will be interrupted by the
498callback:
499
500 my $e = $queue->marker;
501 my $flag;
502 $e->cb (sub { $flag = 1 });
503 1 until $flag;
504 # $flag is now 1
505
506The reason why you shouldn't blindly copy the above code is that
507busy waiting is a really really bad thing, and really really bad for
508performance.
509
510While at first this asynchronous business might look exciting, it can be
511really hard, because you need to be prepared for the callback code to be
512executed at any time, which limits the amount of things the callback code
513can do safely.
514
515This can be mitigated somewhat by using C<<
516$OpenCL::INTERRUPT->scope_block >> (see the L<Async::Interrupt>
517documentation for details).
518
519The other problem is that your program must be actively I<running> to be
520interrupted. When you calculate stuff, your program is running. When you
521hang in some C functions or other block execution (by calling C<sleep>,
522C<select>, running an event loop and so on), your program is waiting, not
523running.
524
525One way around that would be to attach a read watcher to your event loop,
526listening for events on C<< $OpenCL::INTERRUPT->pipe_fileno >>, using a
527dummy callback (C<sub { }>) to temporarily execute some perl code.
528
529That is then awfully close to using the built-in AnyEvent support above,
530though, so consider that one instead.
531
532=head3 Be creative
533
534OpenCL exports the L<Async::Interrupt> object it uses in the global
535variable C<$OpenCL::INTERRUPT>. You can configure it in any way you like.
536
537So if you want to feel like a real pro, err, wait, if you feel no risk
538menas no fun, you can experiment by implementing your own mode of
539operations.
540
541=cut
542
543package OpenCL;
544
545use common::sense;
546use Carp ();
547use Async::Interrupt ();
548
549our $POLL_FUNC; # set by XS
550
551BEGIN {
552 our $VERSION = '0.99';
553
554 require XSLoader;
555 XSLoader::load (__PACKAGE__, $VERSION);
556
557 @OpenCL::Platform::ISA =
558 @OpenCL::Device::ISA =
559 @OpenCL::Context::ISA =
560 @OpenCL::Queue::ISA =
561 @OpenCL::Memory::ISA =
562 @OpenCL::Sampler::ISA =
563 @OpenCL::Program::ISA =
564 @OpenCL::Kernel::ISA =
565 @OpenCL::Event::ISA = OpenCL::Object::;
566
567 @OpenCL::SubDevice::ISA = OpenCL::Device::;
568
569 @OpenCL::Buffer::ISA =
570 @OpenCL::Image::ISA = OpenCL::Memory::;
571
572 @OpenCL::BufferObj::ISA = OpenCL::Buffer::;
573
574 @OpenCL::Image2D::ISA =
575 @OpenCL::Image3D::ISA =
576 @OpenCL::Image2DArray::ISA =
577 @OpenCL::Image1D::ISA =
578 @OpenCL::Image1DArray::ISA =
579 @OpenCL::Image1DBuffer::ISA = OpenCL::Image::;
580
581 @OpenCL::UserEvent::ISA = OpenCL::Event::;
582
583 @OpenCL::MappedBuffer::ISA =
584 @OpenCL::MappedImage::ISA = OpenCL::Mapped::;
585}
586
244=head2 THE OpenCL PACKAGE 587=head2 THE OpenCL PACKAGE
245 588
246=over 4 589=over 4
247 590
248=item $int = OpenCL::errno 591=item $int = OpenCL::errno
249 592
250The last error returned by a function - it's only changed on errors. 593The last error returned by a function - it's only valid after an error occured
594and before calling another OpenCL function.
251 595
252=item $str = OpenCL::err2str $errval 596=item $str = OpenCL::err2str [$errval]
253 597
254Comverts an error value into a human readable string. 598Converts an error value into a human readable string. IF no error value is
599given, then the last error will be used (as returned by OpenCL::errno).
255 600
256=item $str = OpenCL::enum2str $enum 601=item $str = OpenCL::enum2str $enum
257 602
258Converts most enum values (inof parameter names, image format constants, 603Converts most enum values (of parameter names, image format constants,
259object types, addressing and filter modes, command types etc.) into a 604object types, addressing and filter modes, command types etc.) into a
260human readbale string. When confronted with some random integer it can be 605human readable string. When confronted with some random integer it can be
261very helpful to pass it through this function to maybe get some readable 606very helpful to pass it through this function to maybe get some readable
262string out of it. 607string out of it.
263 608
264=item @platforms = OpenCL::platforms 609=item @platforms = OpenCL::platforms
265 610
266Returns all available OpenCL::Platform objects. 611Returns all available OpenCL::Platform objects.
267 612
268L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html> 613L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html>
269 614
270=item $ctx = OpenCL::context_from_type_simple $type = OpenCL::DEVICE_TYPE_DEFAULT 615=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr
271 616
272Tries to create a context from a default device and platform - never worked for me. 617Tries to create a context from a default device and platform type - never worked for me.
618
619type: OpenCL::DEVICE_TYPE_DEFAULT, OpenCL::DEVICE_TYPE_CPU, OpenCL::DEVICE_TYPE_GPU,
620OpenCL::DEVICE_TYPE_ACCELERATOR, OpenCL::DEVICE_TYPE_CUSTOM, OpenCL::DEVICE_TYPE_ALL.
273 621
274L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 622L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
275 623
624=item $ctx = OpenCL::context $properties, \@devices, $callback->($err, $pvt) = $print_stderr)
625
626Create a new OpenCL::Context object using the given device object(s). This
627function isn't implemented yet, use C<< $platform->context >> instead.
628
629L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
630
276=item OpenCL::wait_for_events $wait_events... 631=item OpenCL::wait_for_events $wait_events...
277 632
278Waits for all events to complete. 633Waits for all events to complete.
279 634
280L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 635L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
281 636
637=item OpenCL::poll
638
639Checks if there are any outstanding events (see L<EVENT SYSTEM>) and
640invokes their callbacks.
641
642=item $OpenCL::INTERRUPT
643
644The L<Async::Interrupt> object used to signal asynchronous events (see
645L<EVENT SYSTEM>).
646
647=cut
648
649our $INTERRUPT = new Async::Interrupt c_cb => [$POLL_FUNC, 0];
650
651&_eq_initialise ($INTERRUPT->signal_func);
652
653=item $OpenCL::WATCHER
654
655The L<AnyEvent> watcher object used to watch for asynchronous events (see
656L<EVENT SYSTEM>). This variable is C<undef> until L<AnyEvent> has been
657loaded I<and> initialised (e.g. by calling C<AnyEvent::detect>).
658
659=cut
660
661our $WATCHER;
662
663sub _init_anyevent {
664 $INTERRUPT->block;
665 $WATCHER = AE::io ($INTERRUPT->pipe_fileno, 0, sub { $INTERRUPT->handle });
666}
667
668if (defined $AnyEvent::MODEL) {
669 _init_anyevent;
670} else {
671 push @AnyEvent::post_detect, \&_init_anyevent;
672}
673
282=back 674=back
283 675
676=head2 THE OpenCL::Object CLASS
677
678This is the base class for all objects in the OpenCL module. The only
679method it implements is the C<id> method, which is only useful if you want
680to interface to OpenCL on the C level.
681
682=over 4
683
684=item $iv = $obj->id
685
686OpenCL objects are represented by pointers or integers on the C level. If
687you want to interface to an OpenCL object directly on the C level, then
688you need this value, which is returned by this method. You should use an
689C<IV> type in your code and cast that to the correct type.
690
691=cut
692
693sub OpenCL::Object::id {
694 ref $_[0] eq "SCALAR"
695 ? ${ $_[0] }
696 : $_[0][0]
697}
698
699=back
700
284=head2 THE OpenCL::Platform CLASS 701=head2 THE OpenCL::Platform CLASS
285 702
286=over 4 703=over 4
287 704
705=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL)
706
707Returns a list of matching OpenCL::Device objects.
708
709=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr)
710
711Tries to create a context. Never worked for me, and you need devices explicitly anyway.
712
713L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
714
715=item $ctx = $platform->context ($properties, \@devices, $callback->($err, $pvt) = $print_stderr)
716
717Create a new OpenCL::Context object using the given device object(s)- a
718CL_CONTEXT_PLATFORM property is supplied automatically.
719
720L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
721
288=item $packed_value = $platform->info ($name) 722=item $packed_value = $platform->info ($name)
289 723
290Calls C<clGetPlatformInfo> and returns the packed, raw value - for 724Calls C<clGetPlatformInfo> and returns the packed, raw value - for
291strings, this will be the string, for other values you probably need to 725strings, this will be the string (possibly including terminating \0), for
292use the correct C<unpack>. This might get improved in the future. Hopefully. 726other values you probably need to use the correct C<unpack>.
727
728It's best to avoid this method and use one of the following convenience
729wrappers.
293 730
294L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html> 731L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html>
295 732
296=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL) 733=item $platform->unload_compiler
297 734
298Returns a list of matching OpenCL::Device objects. 735Attempts to unload the compiler for this platform, for endless
736profit. Does nothing on OpenCL 1.1.
299 737
300=item $ctx = $platform->context_from_type_simple ($type = OpenCL::DEVICE_TYPE_DEFAULT)
301
302Tries to create a context. Never worked for me.
303
304L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html> 738L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clUnloadPlatformCompiler.html>
739
740=for gengetinfo begin platform
741
742=item $string = $platform->profile
743
744Calls C<clGetPlatformInfo> with C<OpenCL::PLATFORM_PROFILE> and returns the result.
745
746=item $string = $platform->version
747
748Calls C<clGetPlatformInfo> with C<OpenCL::PLATFORM_VERSION> and returns the result.
749
750=item $string = $platform->name
751
752Calls C<clGetPlatformInfo> with C<OpenCL::PLATFORM_NAME> and returns the result.
753
754=item $string = $platform->vendor
755
756Calls C<clGetPlatformInfo> with C<OpenCL::PLATFORM_VENDOR> and returns the result.
757
758=item $string = $platform->extensions
759
760Calls C<clGetPlatformInfo> with C<OpenCL::PLATFORM_EXTENSIONS> and returns the result.
761
762=for gengetinfo end platform
305 763
306=back 764=back
307 765
308=head2 THE OpenCL::Device CLASS 766=head2 THE OpenCL::Device CLASS
309 767
313 771
314See C<< $platform->info >> for details. 772See C<< $platform->info >> for details.
315 773
316L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetDeviceInfo.html> 774L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetDeviceInfo.html>
317 775
318=item $ctx = $device->context_simple 776=item @devices = $device->sub_devices (\@properties)
319 777
320Convenience function to create a new OpenCL::Context object. 778Creates OpencL::SubDevice objects by partitioning an existing device.
321 779
322L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html> 780L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateSubDevices.html>
781
782=for gengetinfo begin device
783
784=item $device_type = $device->type
785
786Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_TYPE> and returns the result.
787
788=item $uint = $device->vendor_id
789
790Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_VENDOR_ID> and returns the result.
791
792=item $uint = $device->max_compute_units
793
794Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_COMPUTE_UNITS> and returns the result.
795
796=item $uint = $device->max_work_item_dimensions
797
798Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_WORK_ITEM_DIMENSIONS> and returns the result.
799
800=item $int = $device->max_work_group_size
801
802Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_WORK_GROUP_SIZE> and returns the result.
803
804=item @ints = $device->max_work_item_sizes
805
806Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_WORK_ITEM_SIZES> and returns the result.
807
808=item $uint = $device->preferred_vector_width_char
809
810Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PREFERRED_VECTOR_WIDTH_CHAR> and returns the result.
811
812=item $uint = $device->preferred_vector_width_short
813
814Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PREFERRED_VECTOR_WIDTH_SHORT> and returns the result.
815
816=item $uint = $device->preferred_vector_width_int
817
818Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PREFERRED_VECTOR_WIDTH_INT> and returns the result.
819
820=item $uint = $device->preferred_vector_width_long
821
822Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PREFERRED_VECTOR_WIDTH_LONG> and returns the result.
823
824=item $uint = $device->preferred_vector_width_float
825
826Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PREFERRED_VECTOR_WIDTH_FLOAT> and returns the result.
827
828=item $uint = $device->preferred_vector_width_double
829
830Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PREFERRED_VECTOR_WIDTH_DOUBLE> and returns the result.
831
832=item $uint = $device->max_clock_frequency
833
834Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_CLOCK_FREQUENCY> and returns the result.
835
836=item $bitfield = $device->address_bits
837
838Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_ADDRESS_BITS> and returns the result.
839
840=item $uint = $device->max_read_image_args
841
842Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_READ_IMAGE_ARGS> and returns the result.
843
844=item $uint = $device->max_write_image_args
845
846Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_WRITE_IMAGE_ARGS> and returns the result.
847
848=item $ulong = $device->max_mem_alloc_size
849
850Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_MEM_ALLOC_SIZE> and returns the result.
851
852=item $int = $device->image2d_max_width
853
854Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_IMAGE2D_MAX_WIDTH> and returns the result.
855
856=item $int = $device->image2d_max_height
857
858Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_IMAGE2D_MAX_HEIGHT> and returns the result.
859
860=item $int = $device->image3d_max_width
861
862Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_IMAGE3D_MAX_WIDTH> and returns the result.
863
864=item $int = $device->image3d_max_height
865
866Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_IMAGE3D_MAX_HEIGHT> and returns the result.
867
868=item $int = $device->image3d_max_depth
869
870Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_IMAGE3D_MAX_DEPTH> and returns the result.
871
872=item $uint = $device->image_support
873
874Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_IMAGE_SUPPORT> and returns the result.
875
876=item $int = $device->max_parameter_size
877
878Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_PARAMETER_SIZE> and returns the result.
879
880=item $uint = $device->max_samplers
881
882Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_SAMPLERS> and returns the result.
883
884=item $uint = $device->mem_base_addr_align
885
886Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MEM_BASE_ADDR_ALIGN> and returns the result.
887
888=item $uint = $device->min_data_type_align_size
889
890Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MIN_DATA_TYPE_ALIGN_SIZE> and returns the result.
891
892=item $device_fp_config = $device->single_fp_config
893
894Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_SINGLE_FP_CONFIG> and returns the result.
895
896=item $device_mem_cache_type = $device->global_mem_cache_type
897
898Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_GLOBAL_MEM_CACHE_TYPE> and returns the result.
899
900=item $uint = $device->global_mem_cacheline_size
901
902Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_GLOBAL_MEM_CACHELINE_SIZE> and returns the result.
903
904=item $ulong = $device->global_mem_cache_size
905
906Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_GLOBAL_MEM_CACHE_SIZE> and returns the result.
907
908=item $ulong = $device->global_mem_size
909
910Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_GLOBAL_MEM_SIZE> and returns the result.
911
912=item $ulong = $device->max_constant_buffer_size
913
914Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_CONSTANT_BUFFER_SIZE> and returns the result.
915
916=item $uint = $device->max_constant_args
917
918Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_MAX_CONSTANT_ARGS> and returns the result.
919
920=item $device_local_mem_type = $device->local_mem_type
921
922Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_LOCAL_MEM_TYPE> and returns the result.
923
924=item $ulong = $device->local_mem_size
925
926Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_LOCAL_MEM_SIZE> and returns the result.
927
928=item $boolean = $device->error_correction_support
929
930Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_ERROR_CORRECTION_SUPPORT> and returns the result.
931
932=item $int = $device->profiling_timer_resolution
933
934Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PROFILING_TIMER_RESOLUTION> and returns the result.
935
936=item $boolean = $device->endian_little
937
938Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_ENDIAN_LITTLE> and returns the result.
939
940=item $boolean = $device->available
941
942Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_AVAILABLE> and returns the result.
943
944=item $boolean = $device->compiler_available
945
946Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_COMPILER_AVAILABLE> and returns the result.
947
948=item $device_exec_capabilities = $device->execution_capabilities
949
950Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_EXECUTION_CAPABILITIES> and returns the result.
951
952=item $command_queue_properties = $device->properties
953
954Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_QUEUE_PROPERTIES> and returns the result.
955
956=item $ = $device->platform
957
958Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PLATFORM> and returns the result.
959
960=item $string = $device->name
961
962Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NAME> and returns the result.
963
964=item $string = $device->vendor
965
966Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_VENDOR> and returns the result.
967
968=item $string = $device->driver_version
969
970Calls C<clGetDeviceInfo> with C<OpenCL::DRIVER_VERSION> and returns the result.
971
972=item $string = $device->profile
973
974Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PROFILE> and returns the result.
975
976=item $string = $device->version
977
978Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_VERSION> and returns the result.
979
980=item $string = $device->extensions
981
982Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_EXTENSIONS> and returns the result.
983
984=item $uint = $device->preferred_vector_width_half
985
986Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PREFERRED_VECTOR_WIDTH_HALF> and returns the result.
987
988=item $uint = $device->native_vector_width_char
989
990Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NATIVE_VECTOR_WIDTH_CHAR> and returns the result.
991
992=item $uint = $device->native_vector_width_short
993
994Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NATIVE_VECTOR_WIDTH_SHORT> and returns the result.
995
996=item $uint = $device->native_vector_width_int
997
998Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NATIVE_VECTOR_WIDTH_INT> and returns the result.
999
1000=item $uint = $device->native_vector_width_long
1001
1002Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NATIVE_VECTOR_WIDTH_LONG> and returns the result.
1003
1004=item $uint = $device->native_vector_width_float
1005
1006Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NATIVE_VECTOR_WIDTH_FLOAT> and returns the result.
1007
1008=item $uint = $device->native_vector_width_double
1009
1010Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NATIVE_VECTOR_WIDTH_DOUBLE> and returns the result.
1011
1012=item $uint = $device->native_vector_width_half
1013
1014Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_NATIVE_VECTOR_WIDTH_HALF> and returns the result.
1015
1016=item $device_fp_config = $device->double_fp_config
1017
1018Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_DOUBLE_FP_CONFIG> and returns the result.
1019
1020=item $device_fp_config = $device->half_fp_config
1021
1022Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_HALF_FP_CONFIG> and returns the result.
1023
1024=item $boolean = $device->host_unified_memory
1025
1026Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_HOST_UNIFIED_MEMORY> and returns the result.
1027
1028=item $device = $device->parent_device_ext
1029
1030Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PARENT_DEVICE_EXT> and returns the result.
1031
1032=item @device_partition_property_exts = $device->partition_types_ext
1033
1034Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PARTITION_TYPES_EXT> and returns the result.
1035
1036=item @device_partition_property_exts = $device->affinity_domains_ext
1037
1038Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_AFFINITY_DOMAINS_EXT> and returns the result.
1039
1040=item $uint = $device->reference_count_ext
1041
1042Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_REFERENCE_COUNT_EXT> and returns the result.
1043
1044=item @device_partition_property_exts = $device->partition_style_ext
1045
1046Calls C<clGetDeviceInfo> with C<OpenCL::DEVICE_PARTITION_STYLE_EXT> and returns the result.
1047
1048=for gengetinfo end device
323 1049
324=back 1050=back
325 1051
326=head2 THE OpenCL::Context CLASS 1052=head2 THE OpenCL::Context CLASS
327 1053
1054An OpenCL::Context is basically a container, or manager, for a number of
1055devices of a platform. It is used to create all sorts of secondary objects
1056such as buffers, queues, programs and so on.
1057
1058All context creation functions and methods take a list of properties
1059(type-value pairs). All property values can be specified as integers -
1060some additionally support other types:
1061
328=over 4 1062=over 4
329 1063
330=item $packed_value = $ctx->info ($name) 1064=item OpenCL::CONTEXT_PLATFORM
331 1065
332See C<< $platform->info >> for details. 1066Also accepts OpenCL::Platform objects.
333 1067
334L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetContextInfo.html> 1068=item OpenCL::GLX_DISPLAY_KHR
1069
1070Also accepts C<undef>, in which case a deep and troubling hack is engaged
1071to find the current glx display (see L<GLX SUPPORT>).
1072
1073=item OpenCL::GL_CONTEXT_KHR
1074
1075Also accepts C<undef>, in which case a deep and troubling hack is engaged
1076to find the current glx context (see L<GLX SUPPORT>).
1077
1078=back
1079
1080=over 4
1081
1082=item $prog = $ctx->build_program ($program, $options = "")
1083
1084This convenience function tries to build the program on all devices in
1085the context. If the build fails, then the function will C<croak> with the
1086build log. Otherwise ti returns the program object.
1087
1088The C<$program> can either be a C<OpenCL::Program> object or a string
1089containing the program. In the latter case, a program objetc will be
1090created automatically.
1091
1092=cut
1093
1094sub OpenCL::Context::build_program {
1095 my ($self, $prog, $options) = @_;
1096
1097 $prog = $self->program_with_source ($prog)
1098 unless ref $prog;
1099
1100 eval { $prog->build (undef, $options); 1 }
1101 or errno == BUILD_PROGRAM_FAILURE
1102 or errno == INVALID_BINARY # workaround nvidia bug
1103 or Carp::croak "OpenCL::Context->build_program: " . err2str;
1104
1105 # we check status for all devices
1106 for my $dev ($self->devices) {
1107 $prog->build_status ($dev) == BUILD_SUCCESS
1108 or Carp::croak "Building OpenCL program for device '" . $dev->name . "' failed:\n"
1109 . $prog->build_log ($dev);
1110 }
1111
1112 $prog
1113}
335 1114
336=item $queue = $ctx->queue ($device, $properties) 1115=item $queue = $ctx->queue ($device, $properties)
337 1116
338Create a new OpenCL::Queue object from the context and the given device. 1117Create a new OpenCL::Queue object from the context and the given device.
339 1118
340L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html> 1119L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html>
341 1120
1121Example: create an out-of-order queue.
1122
1123 $queue = $ctx->queue ($device, OpenCL::QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE);
1124
342=item $ev = $ctx->user_event 1125=item $ev = $ctx->user_event
343 1126
344Creates a new OpenCL::UserEvent object. 1127Creates a new OpenCL::UserEvent object.
345 1128
346L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateUserEvent.html> 1129L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateUserEvent.html>
347 1130
348=item $buf = $ctx->buffer ($flags, $len) 1131=item $buf = $ctx->buffer ($flags, $len)
349 1132
350Creates a new OpenCL::Buffer object with the given flags and octet-size. 1133Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object with the
1134given flags and octet-size.
1135
1136flags: OpenCL::MEM_READ_WRITE, OpenCL::MEM_WRITE_ONLY, OpenCL::MEM_READ_ONLY,
1137OpenCL::MEM_USE_HOST_PTR, OpenCL::MEM_ALLOC_HOST_PTR, OpenCL::MEM_COPY_HOST_PTR,
1138OpenCL::MEM_HOST_WRITE_ONLY, OpenCL::MEM_HOST_READ_ONLY, OpenCL::MEM_HOST_NO_ACCESS.
351 1139
352L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateBuffer.html> 1140L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateBuffer.html>
353 1141
354=item $buf = $ctx->buffer_sv ($flags, $data) 1142=item $buf = $ctx->buffer_sv ($flags, $data)
355 1143
356Creates a new OpenCL::Buffer object and initialise it with the given data values. 1144Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and
1145initialise it with the given data values.
357 1146
1147=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)
1148
1149Creates a new OpenCL::Image object and optionally initialises it with
1150the given data values.
1151
1152channel_order: OpenCL::R, OpenCL::A, OpenCL::RG, OpenCL::RA, OpenCL::RGB,
1153OpenCL::RGBA, OpenCL::BGRA, OpenCL::ARGB, OpenCL::INTENSITY, OpenCL::LUMINANCE,
1154OpenCL::Rx, OpenCL::RGx, OpenCL::RGBx.
1155
1156channel_type: OpenCL::SNORM_INT8, OpenCL::SNORM_INT16, OpenCL::UNORM_INT8,
1157OpenCL::UNORM_INT16, OpenCL::UNORM_SHORT_565, OpenCL::UNORM_SHORT_555,
1158OpenCL::UNORM_INT_101010, OpenCL::SIGNED_INT8, OpenCL::SIGNED_INT16,
1159OpenCL::SIGNED_INT32, OpenCL::UNSIGNED_INT8, OpenCL::UNSIGNED_INT16,
1160OpenCL::UNSIGNED_INT32, OpenCL::HALF_FLOAT, OpenCL::FLOAT.
1161
1162
1163type: OpenCL::MEM_OBJECT_BUFFER, OpenCL::MEM_OBJECT_IMAGE2D,
1164OpenCL::MEM_OBJECT_IMAGE3D, OpenCL::MEM_OBJECT_IMAGE2D_ARRAY,
1165OpenCL::MEM_OBJECT_IMAGE1D, OpenCL::MEM_OBJECT_IMAGE1D_ARRAY,
1166OpenCL::MEM_OBJECT_IMAGE1D_BUFFER.
1167
1168L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateImage.html>
1169
358=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $data) 1170=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $row_pitch = 0, $data = undef)
359 1171
360Creates a new OpenCL::Image2D object and optionally initialises it with the given data values. 1172Creates a new OpenCL::Image2D object and optionally initialises it with
1173the given data values.
361 1174
362L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage2D.html> 1175L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage2D.html>
363 1176
364=item $img = $ctx->image3d ($flags, $channel_order, $channel_type, $width, $height, $depth, $slice_pitch, $data) 1177=item $img = $ctx->image3d ($flags, $channel_order, $channel_type, $width, $height, $depth, $row_pitch = 0, $slice_pitch = 0, $data = undef)
365 1178
366Creates a new OpenCL::Image3D object and optionally initialises it with the given data values. 1179Creates a new OpenCL::Image3D object and optionally initialises it with
1180the given data values.
367 1181
368L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage3D.html> 1182L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateImage3D.html>
1183
1184=item $buffer = $ctx->gl_buffer ($flags, $bufobj)
1185
1186Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object that refers to the given
1187OpenGL buffer object.
1188
1189flags: OpenCL::MEM_READ_WRITE, OpenCL::MEM_READ_ONLY, OpenCL::MEM_WRITE_ONLY.
1190
1191http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLBuffer.html
1192
1193=item $img = $ctx->gl_texture ($flags, $target, $miplevel, $texture)
1194
1195Creates a new OpenCL::Image object that refers to the given OpenGL
1196texture object or buffer.
1197
1198target: GL_TEXTURE_1D, GL_TEXTURE_1D_ARRAY, GL_TEXTURE_BUFFER,
1199GL_TEXTURE_2D, GL_TEXTURE_2D_ARRAY, GL_TEXTURE_3D,
1200GL_TEXTURE_CUBE_MAP_POSITIVE_X, GL_TEXTURE_CUBE_MAP_POSITIVE_Y,
1201GL_TEXTURE_CUBE_MAP_POSITIVE_Z, GL_TEXTURE_CUBE_MAP_NEGATIVE_X,
1202GL_TEXTURE_CUBE_MAP_NEGATIVE_Y, GL_TEXTURE_CUBE_MAP_NEGATIVE_Z,
1203GL_TEXTURE_RECTANGLE/GL_TEXTURE_RECTANGLE_ARB.
1204
1205http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateFromGLTexture.html
1206
1207=item $img = $ctx->gl_texture2d ($flags, $target, $miplevel, $texture)
1208
1209Creates a new OpenCL::Image2D object that refers to the given OpenGL
12102D texture object.
1211
1212http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture2D.html
1213
1214=item $img = $ctx->gl_texture3d ($flags, $target, $miplevel, $texture)
1215
1216Creates a new OpenCL::Image3D object that refers to the given OpenGL
12173D texture object.
1218
1219http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture3D.html
1220
1221=item $ctx->gl_renderbuffer ($flags, $renderbuffer)
1222
1223Creates a new OpenCL::Image2D object that refers to the given OpenGL
1224render buffer.
1225
1226http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLRenderbuffer.html
369 1227
370=item @formats = $ctx->supported_image_formats ($flags, $image_type) 1228=item @formats = $ctx->supported_image_formats ($flags, $image_type)
371 1229
372Returns a list of matching image formats - each format is an arrayref with 1230Returns a list of matching image formats - each format is an arrayref with
373two values, $channel_order and $channel_type, in it. 1231two values, $channel_order and $channel_type, in it.
376 1234
377=item $sampler = $ctx->sampler ($normalized_coords, $addressing_mode, $filter_mode) 1235=item $sampler = $ctx->sampler ($normalized_coords, $addressing_mode, $filter_mode)
378 1236
379Creates a new OpenCL::Sampler object. 1237Creates a new OpenCL::Sampler object.
380 1238
1239addressing_mode: OpenCL::ADDRESS_NONE, OpenCL::ADDRESS_CLAMP_TO_EDGE,
1240OpenCL::ADDRESS_CLAMP, OpenCL::ADDRESS_REPEAT, OpenCL::ADDRESS_MIRRORED_REPEAT.
1241
1242filter_mode: OpenCL::FILTER_NEAREST, OpenCL::FILTER_LINEAR.
1243
381L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateSampler.html> 1244L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateSampler.html>
382 1245
383=item $program = $ctx->program_with_source ($string) 1246=item $program = $ctx->program_with_source ($string)
384 1247
385Creates a new OpenCL::Program object from the given source code. 1248Creates a new OpenCL::Program object from the given source code.
386 1249
387L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateProgramWithSource.html> 1250L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateProgramWithSource.html>
388 1251
1252=item ($program, \@status) = $ctx->program_with_binary (\@devices, \@binaries)
1253
1254Creates a new OpenCL::Program object from the given binaries.
1255
1256L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateProgramWithBinary.html>
1257
1258Example: clone an existing program object that contains a successfully
1259compiled program, no matter how useless this is.
1260
1261 my $clone = $ctx->program_with_binary ([$prog->devices], [$prog->binaries]);
1262
1263=item $program = $ctx->program_with_built_in_kernels (\@devices, $kernel_names)
1264
1265Creates a new OpenCL::Program object from the given built-in kernel names.
1266
1267L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateProgramWithBuiltInKernels.html>
1268
1269=item $packed_value = $ctx->info ($name)
1270
1271See C<< $platform->info >> for details.
1272
1273L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetContextInfo.html>
1274
1275=for gengetinfo begin context
1276
1277=item $uint = $context->reference_count
1278
1279Calls C<clGetContextInfo> with C<OpenCL::CONTEXT_REFERENCE_COUNT> and returns the result.
1280
1281=item @devices = $context->devices
1282
1283Calls C<clGetContextInfo> with C<OpenCL::CONTEXT_DEVICES> and returns the result.
1284
1285=item @property_ints = $context->properties
1286
1287Calls C<clGetContextInfo> with C<OpenCL::CONTEXT_PROPERTIES> and returns the result.
1288
1289=item $uint = $context->num_devices
1290
1291Calls C<clGetContextInfo> with C<OpenCL::CONTEXT_NUM_DEVICES> and returns the result.
1292
1293=for gengetinfo end context
1294
389=back 1295=back
390 1296
391=head2 THE OpenCL::Queue CLASS 1297=head2 THE OpenCL::Queue CLASS
392 1298
393An OpenCL::Queue represents an execution queue for OpenCL. You execute 1299An OpenCL::Queue represents an execution queue for OpenCL. You execute
394requests by calling their respective C<enqueue_xxx> method and waitinf for 1300requests by calling their respective method and waiting for it to complete
395it to complete in some way. 1301in some way.
396 1302
397All the enqueue methods return an event object that can be used to wait 1303Most methods that enqueue some request return an event object that can
398for completion, unless the method is called in void context, in which case 1304be used to wait for completion (optionally using a callback), unless
399no event object is created. 1305the method is called in void context, in which case no event object is
1306created.
400 1307
401They also allow you to specify any number of other event objects that this 1308They also allow you to specify any number of other event objects that this
402request has to wait for before it starts executing, by simply passing the 1309request has to wait for before it starts executing, by simply passing the
403event objects as extra parameters to the enqueue methods. 1310event objects as extra parameters to the enqueue methods. To simplify
1311program design, this module ignores any C<undef> values in the list of
1312events. This makes it possible to code operations such as this, without
1313having to put a valid event object into C<$event> first:
1314
1315 $event = $queue->xxx (..., $event);
404 1316
405Queues execute in-order by default, without any parallelism, so in most 1317Queues execute in-order by default, without any parallelism, so in most
406cases (i.e. you use only one queue) it's not necessary to wait for or 1318cases (i.e. you use only one queue) it's not necessary to wait for or
407create event objects. 1319create event objects, althoguh an our of order queue is often a bit
1320faster.
408 1321
409=over 4 1322=over 4
410 1323
411=item $packed_value = $ctx->info ($name)
412
413See C<< $platform->info >> for details.
414
415L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetCommandQueueInfo.html>
416
417=item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...) 1324=item $ev = $queue->read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...)
418 1325
419Reads data from buffer into the given string. 1326Reads data from buffer into the given string.
420 1327
421L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBuffer.html> 1328L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBuffer.html>
422 1329
423=item $ev = $queue->enqueue_write_buffer ($buffer, $blocking, $offset, $data, $wait_events...) 1330=item $ev = $queue->write_buffer ($buffer, $blocking, $offset, $data, $wait_events...)
424 1331
425Writes data to buffer from the given string. 1332Writes data to buffer from the given string.
426 1333
427L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBuffer.html> 1334L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBuffer.html>
428 1335
429=item $ev = $queue->enqueue_copy_buffer ($src, $dst, $src_offset, $dst_offset, $len, $wait_events...) 1336=item $ev = $queue->copy_buffer ($src, $dst, $src_offset, $dst_offset, $len, $wait_events...)
430 1337
431L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBuffer.html> 1338L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBuffer.html>
432 1339
1340=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...)
1341
1342http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadBufferRect.html
1343
1344=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...)
1345
1346http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBufferRect.html
1347
1348=item $ev = $queue->copy_buffer_to_image ($src_buffer, $dst_image, $src_offset, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...)
1349
1350L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>
1351
433=item $ev = $queue->enqueue_read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) 1352=item $ev = $queue->read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...)
1353
1354C<$row_pitch> (and C<$slice_pitch>) can be C<0>, in which case the OpenCL
1355module uses the image width (and height) to supply default values.
434 1356
435L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html> 1357L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html>
436 1358
437=item $ev = $queue->enqueue_write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $data, $wait_events...) 1359=item $ev = $queue->write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...)
438 1360
1361C<$row_pitch> (and C<$slice_pitch>) can be C<0>, in which case the OpenCL
1362module uses the image width (and height) to supply default values.
439L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html> 1363L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html>
440 1364
1365=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...)
1366
1367L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html>
1368
1369=item $ev = $queue->copy_image_to_buffer ($src_image, $dst_image, $src_x, $src_y, $src_z, $width, $height, $depth, $dst_offset, $wait_events...)
1370
1371L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImageToBuffer.html>
1372
441=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...) 1373=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...)
442 1374
443Yeah. 1375Yeah.
444 1376
445L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferRect.html>
446
447=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, ...)
448
449L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>. 1377L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>.
450 1378
451=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, ...) 1379=item $ev = $queue->fill_buffer ($mem, $pattern, $offset, $size, ...)
452 1380
1381Fills the given buffer object with repeated applications of C<$pattern>,
1382starting at C<$offset> for C<$size> octets.
1383
1384L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillBuffer.html>
1385
1386=item $ev = $queue->fill_image ($img, $r, $g, $b, $a, $x, $y, $z, $width, $height, $depth, ...)
1387
1388Fills the given image area with the given rgba colour components. The
1389components are normally floating point values between C<0> and C<1>,
1390except when the image channel data type is a signe dor unsigned
1391unnormalised format, in which case the range is determined by the format.
1392
453L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html> 1393L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillImage.html>
454 1394
455=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, ...)
456
457L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImageToBuffer.html>
458
459=item $ev = $queue->enqueue_task ($kernel, $wait_events...) 1395=item $ev = $queue->task ($kernel, $wait_events...)
460 1396
461L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html> 1397L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html>
462 1398
463=item $ev = $queue->enqueue_nd_range_kernel ($kernel, @$global_work_offset, @$global_work_size, @$local_work_size, $wait_events...) 1399=item $ev = $queue->nd_range_kernel ($kernel, \@global_work_offset, \@global_work_size, \@local_work_size, $wait_events...)
464 1400
465Enqueues a kernel execution. 1401Enqueues a kernel execution.
466 1402
467@$global_work_size must be specified as a reference to an array of 1403\@global_work_size must be specified as a reference to an array of
468integers specifying the work sizes (element counts). 1404integers specifying the work sizes (element counts).
469 1405
470@$global_work_offset must be either C<undef> (in which case all offsets 1406\@global_work_offset must be either C<undef> (in which case all offsets
471are C<0>), or a reference to an array of work offsets, with the same number 1407are C<0>), or a reference to an array of work offsets, with the same number
472of elements as @$global_work_size. 1408of elements as \@global_work_size.
473 1409
474@$local_work_size must be either C<undef> (in which case the 1410\@local_work_size must be either C<undef> (in which case the
475implementation is supposed to choose good local work sizes), or a 1411implementation is supposed to choose good local work sizes), or a
476reference to an array of local work sizes, with the same number of 1412reference to an array of local work sizes, with the same number of
477elements as @$global_work_size. 1413elements as \@global_work_size.
478 1414
479L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html> 1415L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html>
480 1416
481=item $ev = $queue->enqueue_marker 1417=item $ev = $queue->migrate_mem_objects (\@mem_objects, $flags, $wait_events...)
482 1418
1419Migrates a number of OpenCL::Memory objects to or from the device.
1420
1421flags: OpenCL::MIGRATE_MEM_OBJECT_HOST, OpenCL::MIGRATE_MEM_OBJECT_CONTENT_UNDEFINED
1422
1423L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueMigrateMemObjects.html>
1424
1425=item $ev = $queue->acquire_gl_objects ([object, ...], $wait_events...)
1426
1427Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired
1428for subsequent OpenCL usage.
1429
483L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMarker.html> 1430L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueAcquireGLObjects.html>
484 1431
1432=item $ev = $queue->release_gl_objects ([object, ...], $wait_events...)
1433
1434Enqueues a list (an array-ref of OpenCL::Memory objects) to be released
1435for subsequent OpenGL usage.
1436
1437L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReleaseGLObjects.html>
1438
485=item $ev = $queue->enqueue_wait_for_events ($wait_events...) 1439=item $ev = $queue->wait_for_events ($wait_events...)
486 1440
487L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html> 1441L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html>
488 1442
489=item $queue->enqueue_barrier 1443=item $ev = $queue->marker ($wait_events...)
490 1444
1445L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueMarkerWithWaitList.html>
1446
1447=item $ev = $queue->barrier ($wait_events...)
1448
491L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueBarrier.html> 1449L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueBarrierWithWaitList.html>
492 1450
493=item $queue->flush 1451=item $queue->flush
494 1452
495L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html> 1453L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html>
496 1454
497=item $queue->finish 1455=item $queue->finish
498 1456
499L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFinish.html> 1457L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFinish.html>
500 1458
1459=item $packed_value = $queue->info ($name)
1460
1461See C<< $platform->info >> for details.
1462
1463L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetCommandQueueInfo.html>
1464
1465=for gengetinfo begin command_queue
1466
1467=item $ctx = $command_queue->context
1468
1469Calls C<clGetCommandQueueInfo> with C<OpenCL::QUEUE_CONTEXT> and returns the result.
1470
1471=item $device = $command_queue->device
1472
1473Calls C<clGetCommandQueueInfo> with C<OpenCL::QUEUE_DEVICE> and returns the result.
1474
1475=item $uint = $command_queue->reference_count
1476
1477Calls C<clGetCommandQueueInfo> with C<OpenCL::QUEUE_REFERENCE_COUNT> and returns the result.
1478
1479=item $command_queue_properties = $command_queue->properties
1480
1481Calls C<clGetCommandQueueInfo> with C<OpenCL::QUEUE_PROPERTIES> and returns the result.
1482
1483=for gengetinfo end command_queue
1484
501=back 1485=back
502 1486
1487=head3 MEMORY MAPPED BUFFERS
1488
1489OpenCL allows you to map buffers and images to host memory (read: perl
1490scalars). This is done much like reading or copying a buffer, by enqueuing
1491a map or unmap operation on the command queue.
1492
1493The map operations return an C<OpenCL::Mapped> object - see L<THE
1494OpenCL::Mapped CLASS> section for details on what to do with these
1495objects.
1496
1497The object will be unmapped automatically when the mapped object is
1498destroyed (you can use a barrier to make sure the unmap has finished,
1499before using the buffer in a kernel), but you can also enqueue an unmap
1500operation manually.
1501
1502=over 4
1503
1504=item $mapped_buffer = $queue->map_buffer ($buf, $blocking=1, $map_flags=OpenCL::MAP_READ|OpenCL::MAP_WRITE, $offset=0, $size=undef, $wait_events...)
1505
1506Maps the given buffer into host memory and returns an
1507C<OpenCL::MappedBuffer> object. If C<$size> is specified as undef, then
1508the map will extend to the end of the buffer.
1509
1510map_flags: OpenCL::MAP_READ, OpenCL::MAP_WRITE, OpenCL::MAP_WRITE_INVALIDATE_REGION.
1511
1512L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMapBuffer.html>
1513
1514Example: map the buffer $buf fully and replace the first 4 bytes by "abcd", then unmap.
1515
1516 {
1517 my $mapped = $queue->map_buffer ($buf, 1, OpenCL::MAP_WRITE);
1518 substr $$mapped, 0, 4, "abcd";
1519 } # asynchronously unmap because $mapped is destroyed
1520
1521=item $mapped_image = $queue->map_image ($img, $blocking=1, $map_flags=OpenCL::MAP_READ|OpenCL::MAP_WRITE, $x=0, $y=0, $z=0, $width=undef, $height=undef, $depth=undef, $wait_events...)
1522
1523Maps the given image area into host memory and return an
1524C<OpenCL::MappedImage> object.
1525
1526If any of C<$width>, C<$height> and/or C<$depth> are C<undef> then they
1527will be replaced by the maximum possible value.
1528
1529L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMapImage.html>
1530
1531Example: map an image (with OpenCL::UNSIGNED_INT8 channel type) and set
1532the first channel of the leftmost column to 5, then explicitly unmap
1533it. You are not necessarily meant to do it this way, this example just
1534shows you the accessors to use :)
1535
1536 my $mapped = $queue->map_image ($image, 1, OpenCL::MAP_WRITE);
1537
1538 $mapped->set ($_ * $mapped->row_pitch, pack "C", 5)
1539 for 0..$image->height;
1540
1541 $mapped->unmap;.
1542 $mapped->wait; # only needed for out of order queues normally
1543
1544=item $ev = $queue->unmap ($mapped, $wait_events...)
1545
1546Unmaps the data from host memory. You must not call any methods that
1547modify the data, or modify the data scalar directly, after calling this
1548method.
1549
1550The mapped event object will always be passed as part of the
1551$wait_events. The mapped event object will be replaced by the new event
1552object that this request creates.
1553
1554=back
1555
503=head2 THE OpenCL::Memory CLASS 1556=head2 THE OpenCL::Memory CLASS
504 1557
505This the superclass of all memory objects - OpenCL::Buffer, OpenCL::Image, 1558This the superclass of all memory objects - OpenCL::Buffer, OpenCL::Image,
506OpenCL::Image2D and OpenCL::Image3D. The subclasses of this class 1559OpenCL::Image2D and OpenCL::Image3D.
507currently only exist to allow type-checking.
508 1560
509=over 4 1561=over 4
510 1562
511=item $packed_value = $memory->info ($name) 1563=item $packed_value = $memory->info ($name)
512 1564
513See C<< $platform->info >> for details. 1565See C<< $platform->info >> for details.
514 1566
515L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetMemObjectInfo.html> 1567L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetMemObjectInfo.html>
516 1568
1569=for gengetinfo begin mem
1570
1571=item $mem_object_type = $mem->type
1572
1573Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_TYPE> and returns the result.
1574
1575=item $mem_flags = $mem->flags
1576
1577Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_FLAGS> and returns the result.
1578
1579=item $int = $mem->size
1580
1581Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_SIZE> and returns the result.
1582
1583=item $ptr_value = $mem->host_ptr
1584
1585Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_HOST_PTR> and returns the result.
1586
1587=item $uint = $mem->map_count
1588
1589Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_MAP_COUNT> and returns the result.
1590
1591=item $uint = $mem->reference_count
1592
1593Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_REFERENCE_COUNT> and returns the result.
1594
1595=item $ctx = $mem->context
1596
1597Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_CONTEXT> and returns the result.
1598
1599=item $mem = $mem->associated_memobject
1600
1601Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_ASSOCIATED_MEMOBJECT> and returns the result.
1602
1603=item $int = $mem->offset
1604
1605Calls C<clGetMemObjectInfo> with C<OpenCL::MEM_OFFSET> and returns the result.
1606
1607=for gengetinfo end mem
1608
1609=item ($type, $name) = $mem->gl_object_info
1610
1611Returns the OpenGL object type (e.g. OpenCL::GL_OBJECT_TEXTURE2D) and the
1612object "name" (e.g. the texture name) used to create this memory object.
1613
1614L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetGLObjectInfo.html>
1615
517=back 1616=back
518 1617
1618=head2 THE OpenCL::Buffer CLASS
1619
1620This is a subclass of OpenCL::Memory, and the superclass of
1621OpenCL::BufferObj. Its purpose is simply to distinguish between buffers
1622and sub-buffers.
1623
1624=head2 THE OpenCL::BufferObj CLASS
1625
1626This is a subclass of OpenCL::Buffer and thus OpenCL::Memory. It exists
1627because one cna create sub buffers of OpenLC::BufferObj objects, but not
1628sub buffers from these sub buffers.
1629
1630=over 4
1631
1632=item $subbuf = $buf_obj->sub_buffer_region ($flags, $origin, $size)
1633
1634Creates an OpenCL::Buffer objects from this buffer and returns it. The
1635C<buffer_create_type> is assumed to be C<CL_BUFFER_CREATE_TYPE_REGION>.
1636
1637L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateSubBuffer.html>
1638
1639=back
1640
1641=head2 THE OpenCL::Image CLASS
1642
1643This is the superclass of all image objects - OpenCL::Image1D,
1644OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D,
1645OpenCL::Image2DArray and OpenCL::Image3D.
1646
1647=over 4
1648
1649=item $packed_value = $image->image_info ($name)
1650
1651See C<< $platform->info >> for details.
1652
1653The reason this method is not called C<info> is that there already is an
1654C<< ->info >> method inherited from C<OpenCL::Memory>.
1655
1656L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html>
1657
1658=item ($channel_order, $channel_data_type) = $image->format
1659
1660Returns the channel order and type used to create the image by calling
1661C<clGetImageInfo> with C<CL_IMAGE_FORMAT>.
1662
1663=for gengetinfo begin image
1664
1665=item $int = $image->element_size
1666
1667Calls C<clGetImageInfo> with C<OpenCL::IMAGE_ELEMENT_SIZE> and returns the result.
1668
1669=item $int = $image->row_pitch
1670
1671Calls C<clGetImageInfo> with C<OpenCL::IMAGE_ROW_PITCH> and returns the result.
1672
1673=item $int = $image->slice_pitch
1674
1675Calls C<clGetImageInfo> with C<OpenCL::IMAGE_SLICE_PITCH> and returns the result.
1676
1677=item $int = $image->width
1678
1679Calls C<clGetImageInfo> with C<OpenCL::IMAGE_WIDTH> and returns the result.
1680
1681=item $int = $image->height
1682
1683Calls C<clGetImageInfo> with C<OpenCL::IMAGE_HEIGHT> and returns the result.
1684
1685=item $int = $image->depth
1686
1687Calls C<clGetImageInfo> with C<OpenCL::IMAGE_DEPTH> and returns the result.
1688
1689=for gengetinfo end image
1690
1691=for gengetinfo begin gl_texture
1692
1693=item $GLenum = $gl_texture->target
1694
1695Calls C<clGetGLTextureInfo> with C<OpenCL::GL_TEXTURE_TARGET> and returns the result.
1696
1697=item $GLint = $gl_texture->gl_mipmap_level
1698
1699Calls C<clGetGLTextureInfo> with C<OpenCL::GL_MIPMAP_LEVEL> and returns the result.
1700
1701=for gengetinfo end gl_texture
1702
1703=back
1704
519=head2 THE OpenCL::Sampler CLASS 1705=head2 THE OpenCL::Sampler CLASS
520 1706
521=over 4 1707=over 4
522 1708
523=item $packed_value = $sampler->info ($name) 1709=item $packed_value = $sampler->info ($name)
524 1710
525See C<< $platform->info >> for details. 1711See C<< $platform->info >> for details.
526 1712
527L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetSamplerInfo.html> 1713L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetSamplerInfo.html>
528 1714
1715=for gengetinfo begin sampler
1716
1717=item $uint = $sampler->reference_count
1718
1719Calls C<clGetSamplerInfo> with C<OpenCL::SAMPLER_REFERENCE_COUNT> and returns the result.
1720
1721=item $ctx = $sampler->context
1722
1723Calls C<clGetSamplerInfo> with C<OpenCL::SAMPLER_CONTEXT> and returns the result.
1724
1725=item $addressing_mode = $sampler->normalized_coords
1726
1727Calls C<clGetSamplerInfo> with C<OpenCL::SAMPLER_NORMALIZED_COORDS> and returns the result.
1728
1729=item $filter_mode = $sampler->addressing_mode
1730
1731Calls C<clGetSamplerInfo> with C<OpenCL::SAMPLER_ADDRESSING_MODE> and returns the result.
1732
1733=item $boolean = $sampler->filter_mode
1734
1735Calls C<clGetSamplerInfo> with C<OpenCL::SAMPLER_FILTER_MODE> and returns the result.
1736
1737=for gengetinfo end sampler
1738
529=back 1739=back
530 1740
531=head2 THE OpenCL::Program CLASS 1741=head2 THE OpenCL::Program CLASS
532 1742
533=over 4 1743=over 4
534 1744
535=item $packed_value = $program->info ($name) 1745=item $program->build (\@devices = undef, $options = "", $cb->($program) = undef)
536 1746
537See C<< $platform->info >> for details.
538
539L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProgramInfo.html>
540
541=item $program->build ($device, $options = "")
542
543Tries to build the program with the givne options. 1747Tries to build the program with the given options. See also the
1748C<$ctx->build> convenience function.
1749
1750If a callback is specified, then it will be called when compilation is
1751finished. Note that many OpenCL implementations block your program while
1752compiling whether you use a callback or not. See C<build_async> if you
1753want to make sure the build is done in the background.
1754
1755Note that some OpenCL implementations act up badly, and don't call the
1756callback in some error cases (but call it in others). This implementation
1757assumes the callback will always be called, and leaks memory if this is
1758not so. So best make sure you don't pass in invalid values.
1759
1760Some implementations fail with C<OpenCL::INVALID_BINARY> when the
1761compilation state is successful but some later stage fails.
1762
1763options: C<-D name>, C<-D name=definition>, C<-I dir>,
1764C<-cl-single-precision-constant>, C<-cl-denorms-are-zero>,
1765C<-cl-fp32-correctly-rounded-divide-sqrt>, C<-cl-opt-disable>,
1766C<-cl-mad-enable>, C<-cl-no-signed-zeros>, C<-cl-unsafe-math-optimizations>,
1767C<-cl-finite-math-only>, C<-cl-fast-relaxed-math>,
1768C<-w>, C<-Werror>, C<-cl-std=CL1.1/CL1.2>, C<-cl-kernel-arg-info>,
1769C<-create-library>, C<-enable-link-options>.
544 1770
545L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html> 1771L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html>
1772
1773=item $program->build_async (\@devices = undef, $options = "", $cb->($program) = undef)
1774
1775Similar to C<< ->build >>, except it starts a thread, and never fails (you
1776need to check the compilation status form the callback, or by polling).
1777
1778build_status: OpenCL::BUILD_SUCCESS, OpenCL::BUILD_NONE,
1779OpenCL::BUILD_ERROR, OpenCL::BUILD_IN_PROGRESS.
546 1780
547=item $packed_value = $program->build_info ($device, $name) 1781=item $packed_value = $program->build_info ($device, $name)
548 1782
549Similar to C<< $platform->info >>, but returns build info for a previous 1783Similar to C<< $platform->info >>, but returns build info for a previous
550build attempt for the given device. 1784build attempt for the given device.
556Creates an OpenCL::Kernel object out of the named C<__kernel> function in 1790Creates an OpenCL::Kernel object out of the named C<__kernel> function in
557the program. 1791the program.
558 1792
559L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html> 1793L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html>
560 1794
1795=item @kernels = $program->kernels_in_program
1796
1797Returns all kernels successfully compiled for all devices in program.
1798
1799http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernelsInProgram.html
1800
1801=for gengetinfo begin program_build
1802
1803=item $build_status = $program->build_status ($device)
1804
1805Calls C<clGetProgramBuildInfo> with C<OpenCL::PROGRAM_BUILD_STATUS> and returns the result.
1806
1807=item $string = $program->build_options ($device)
1808
1809Calls C<clGetProgramBuildInfo> with C<OpenCL::PROGRAM_BUILD_OPTIONS> and returns the result.
1810
1811=item $string = $program->build_log ($device)
1812
1813Calls C<clGetProgramBuildInfo> with C<OpenCL::PROGRAM_BUILD_LOG> and returns the result.
1814
1815=for gengetinfo end program_build
1816
1817=item $packed_value = $program->info ($name)
1818
1819See C<< $platform->info >> for details.
1820
1821L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProgramInfo.html>
1822
1823=for gengetinfo begin program
1824
1825=item $uint = $program->reference_count
1826
1827Calls C<clGetProgramInfo> with C<OpenCL::PROGRAM_REFERENCE_COUNT> and returns the result.
1828
1829=item $ctx = $program->context
1830
1831Calls C<clGetProgramInfo> with C<OpenCL::PROGRAM_CONTEXT> and returns the result.
1832
1833=item $uint = $program->num_devices
1834
1835Calls C<clGetProgramInfo> with C<OpenCL::PROGRAM_NUM_DEVICES> and returns the result.
1836
1837=item @devices = $program->devices
1838
1839Calls C<clGetProgramInfo> with C<OpenCL::PROGRAM_DEVICES> and returns the result.
1840
1841=item $string = $program->source
1842
1843Calls C<clGetProgramInfo> with C<OpenCL::PROGRAM_SOURCE> and returns the result.
1844
1845=item @ints = $program->binary_sizes
1846
1847Calls C<clGetProgramInfo> with C<OpenCL::PROGRAM_BINARY_SIZES> and returns the result.
1848
1849=for gengetinfo end program
1850
1851=item @blobs = $program->binaries
1852
1853Returns a string for the compiled binary for every device associated with
1854the program, empty strings indicate missing programs, and an empty result
1855means no program binaries are available.
1856
1857These "binaries" are often, in fact, informative low-level assembly
1858sources.
1859
1860L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProgramInfo.html>
1861
561=back 1862=back
562 1863
563=head2 THE OpenCL::Kernel CLASS 1864=head2 THE OpenCL::Kernel CLASS
564 1865
565=over 4 1866=over 4
568 1869
569See C<< $platform->info >> for details. 1870See C<< $platform->info >> for details.
570 1871
571L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetKernelInfo.html> 1872L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetKernelInfo.html>
572 1873
1874=for gengetinfo begin kernel
1875
1876=item $string = $kernel->function_name
1877
1878Calls C<clGetKernelInfo> with C<OpenCL::KERNEL_FUNCTION_NAME> and returns the result.
1879
1880=item $uint = $kernel->num_args
1881
1882Calls C<clGetKernelInfo> with C<OpenCL::KERNEL_NUM_ARGS> and returns the result.
1883
1884=item $uint = $kernel->reference_count
1885
1886Calls C<clGetKernelInfo> with C<OpenCL::KERNEL_REFERENCE_COUNT> and returns the result.
1887
1888=item $ctx = $kernel->context
1889
1890Calls C<clGetKernelInfo> with C<OpenCL::KERNEL_CONTEXT> and returns the result.
1891
1892=item $program = $kernel->program
1893
1894Calls C<clGetKernelInfo> with C<OpenCL::KERNEL_PROGRAM> and returns the result.
1895
1896=for gengetinfo end kernel
1897
1898=item $packed_value = $kernel->work_group_info ($device, $name)
1899
1900See C<< $platform->info >> for details.
1901
1902The reason this method is not called C<info> is that there already is an
1903C<< ->info >> method.
1904
1905L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetKernelWorkGroupInfo.html>
1906
1907=for gengetinfo begin kernel_work_group
1908
1909=item $int = $kernel->work_group_size ($device)
1910
1911Calls C<clGetKernelWorkGroupInfo> with C<OpenCL::KERNEL_WORK_GROUP_SIZE> and returns the result.
1912
1913=item @ints = $kernel->compile_work_group_size ($device)
1914
1915Calls C<clGetKernelWorkGroupInfo> with C<OpenCL::KERNEL_COMPILE_WORK_GROUP_SIZE> and returns the result.
1916
1917=item $ulong = $kernel->local_mem_size ($device)
1918
1919Calls C<clGetKernelWorkGroupInfo> with C<OpenCL::KERNEL_LOCAL_MEM_SIZE> and returns the result.
1920
1921=item $int = $kernel->preferred_work_group_size_multiple ($device)
1922
1923Calls C<clGetKernelWorkGroupInfo> with C<OpenCL::KERNEL_PREFERRED_WORK_GROUP_SIZE_MULTIPLE> and returns the result.
1924
1925=item $ulong = $kernel->private_mem_size ($device)
1926
1927Calls C<clGetKernelWorkGroupInfo> with C<OpenCL::KERNEL_PRIVATE_MEM_SIZE> and returns the result.
1928
1929=for gengetinfo end kernel_work_group
1930
1931=for gengetinfo begin kernel_arg_info
1932
1933=for gengetinfo end kernel_arg_info
1934
1935=item $kernel->setf ($format, ...)
1936
1937Sets the arguments of a kernel. Since OpenCL 1.1 doesn't have a generic
1938way to set arguments (and with OpenCL 1.2 it might be rather slow), you
1939need to specify a format argument, much as with C<printf>, to tell OpenCL
1940what type of argument it is.
1941
1942The format arguments are single letters:
1943
1944 c char
1945 C unsigned char
1946 s short
1947 S unsigned short
1948 i int
1949 I unsigned int
1950 l long
1951 L unsigned long
1952
1953 h half float (0..65535)
1954 f float
1955 d double
1956
1957 z local (octet size)
1958
1959 m memory object (buffer or image)
1960 a sampler
1961 e event
1962
1963Space characters in the format string are ignored.
1964
1965Example: set the arguments for a kernel that expects an int, two floats, a buffer and an image.
1966
1967 $kernel->setf ("i ff mm", 5, 0.5, 3, $buffer, $image);
1968
573=item $kernel->set_TYPE ($index, $value) 1969=item $kernel->set_TYPE ($index, $value)
574 1970
1971=item $kernel->set_char ($index, $value)
1972
1973=item $kernel->set_uchar ($index, $value)
1974
1975=item $kernel->set_short ($index, $value)
1976
1977=item $kernel->set_ushort ($index, $value)
1978
1979=item $kernel->set_int ($index, $value)
1980
1981=item $kernel->set_uint ($index, $value)
1982
1983=item $kernel->set_long ($index, $value)
1984
1985=item $kernel->set_ulong ($index, $value)
1986
1987=item $kernel->set_half ($index, $value)
1988
1989=item $kernel->set_float ($index, $value)
1990
1991=item $kernel->set_double ($index, $value)
1992
1993=item $kernel->set_memory ($index, $value)
1994
1995=item $kernel->set_buffer ($index, $value)
1996
1997=item $kernel->set_image ($index, $value)
1998
1999=item $kernel->set_sampler ($index, $value)
2000
2001=item $kernel->set_local ($index, $value)
2002
2003=item $kernel->set_event ($index, $value)
2004
575This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>. 2005This is a family of methods to set the kernel argument with the number
576 2006C<$index> to the give C<$value>.
577TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>,
578C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>,
579C<image2d>, C<image3d>, C<sampler> or C<event>.
580 2007
581Chars and integers (including the half type) are specified as integers, 2008Chars and integers (including the half type) are specified as integers,
582float and double as floating point values, memory/buffer/image2d/image3d 2009float and double as floating point values, memory/buffer/image must be
583must be an object of that type or C<undef>, and sampler and event must be 2010an object of that type or C<undef>, local-memory arguments are set by
584objects of that type. 2011specifying the size, and sampler and event must be objects of that type.
2012
2013Note that C<set_memory> works for all memory objects (all types of buffers
2014and images) - the main purpose of the more specific C<set_TYPE> functions
2015is type checking.
2016
2017Setting an argument for a kernel does NOT keep a reference to the object -
2018for example, if you set an argument to some image object, free the image,
2019and call the kernel, you will run into undefined behaviour.
585 2020
586L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html> 2021L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html>
587 2022
588=back 2023=back
589 2024
592This is the superclass for all event objects (including OpenCL::UserEvent 2027This is the superclass for all event objects (including OpenCL::UserEvent
593objects). 2028objects).
594 2029
595=over 4 2030=over 4
596 2031
2032=item $ev->wait
2033
2034Waits for the event to complete.
2035
2036L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
2037
2038=item $ev->cb ($exec_callback_type, $callback->($event, $event_command_exec_status))
2039
2040Adds a callback to the callback stack for the given event type. There is
2041no way to remove a callback again.
2042
2043L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetEventCallback.html>
2044
597=item $packed_value = $ev->info ($name) 2045=item $packed_value = $ev->info ($name)
598 2046
599See C<< $platform->info >> for details. 2047See C<< $platform->info >> for details.
600 2048
601L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html> 2049L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html>
602 2050
603=item $ev->wait 2051=for gengetinfo begin event
604 2052
605Waits for the event to complete. 2053=item $queue = $event->command_queue
606 2054
2055Calls C<clGetEventInfo> with C<OpenCL::EVENT_COMMAND_QUEUE> and returns the result.
2056
2057=item $command_type = $event->command_type
2058
2059Calls C<clGetEventInfo> with C<OpenCL::EVENT_COMMAND_TYPE> and returns the result.
2060
2061=item $uint = $event->reference_count
2062
2063Calls C<clGetEventInfo> with C<OpenCL::EVENT_REFERENCE_COUNT> and returns the result.
2064
2065=item $uint = $event->command_execution_status
2066
2067Calls C<clGetEventInfo> with C<OpenCL::EVENT_COMMAND_EXECUTION_STATUS> and returns the result.
2068
2069=item $ctx = $event->context
2070
2071Calls C<clGetEventInfo> with C<OpenCL::EVENT_CONTEXT> and returns the result.
2072
2073=for gengetinfo end event
2074
2075=item $packed_value = $ev->profiling_info ($name)
2076
2077See C<< $platform->info >> for details.
2078
2079The reason this method is not called C<info> is that there already is an
2080C<< ->info >> method.
2081
607L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html> 2082L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetProfilingInfo.html>
2083
2084=for gengetinfo begin profiling
2085
2086=item $ulong = $event->profiling_command_queued
2087
2088Calls C<clGetEventProfilingInfo> with C<OpenCL::PROFILING_COMMAND_QUEUED> and returns the result.
2089
2090=item $ulong = $event->profiling_command_submit
2091
2092Calls C<clGetEventProfilingInfo> with C<OpenCL::PROFILING_COMMAND_SUBMIT> and returns the result.
2093
2094=item $ulong = $event->profiling_command_start
2095
2096Calls C<clGetEventProfilingInfo> with C<OpenCL::PROFILING_COMMAND_START> and returns the result.
2097
2098=item $ulong = $event->profiling_command_end
2099
2100Calls C<clGetEventProfilingInfo> with C<OpenCL::PROFILING_COMMAND_END> and returns the result.
2101
2102=for gengetinfo end profiling
608 2103
609=back 2104=back
610 2105
611=head2 THE OpenCL::UserEvent CLASS 2106=head2 THE OpenCL::UserEvent CLASS
612 2107
614 2109
615=over 4 2110=over 4
616 2111
617=item $ev->set_status ($execution_status) 2112=item $ev->set_status ($execution_status)
618 2113
2114Sets the execution status of the user event. Can only be called once,
2115either with OpenCL::COMPLETE or a negative number as status.
2116
2117execution_status: OpenCL::COMPLETE or a negative integer.
2118
619L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html> 2119L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html>
620 2120
621=back 2121=back
622 2122
2123=head2 THE OpenCL::Mapped CLASS
2124
2125This class represents objects mapped into host memory. They are
2126represented by a blessed string scalar. The string data is the mapped
2127memory area, that is, if you read or write it, then the mapped object is
2128accessed directly.
2129
2130You must only ever use operations that modify the string in-place - for
2131example, a C<substr> that doesn't change the length, or maybe a regex that
2132doesn't change the length. Any other operation might cause the data to be
2133copied.
2134
2135When the object is destroyed it will enqueue an implicit unmap operation
2136on the queue that was used to create it.
2137
2138Keep in mind that you I<need> to unmap (or destroy) mapped objects before
2139OpenCL sees the changes, even if some implementations don't need this
2140sometimes.
2141
2142Example, replace the first two floats in the mapped buffer by 1 and 2.
2143
2144 my $mapped = $queue->map_buffer ($buf, ...
2145 $mapped->event->wait; # make sure it's there
2146
2147 # now replace first 8 bytes by new data, which is exactly 8 bytes long
2148 # we blindly assume device endianness to equal host endianness
2149 # (and of course, we assume iee 754 single precision floats :)
2150 substr $$mapped, 0, 8, pack "f*", 1, 2;
2151
2152=over 4
2153
2154=item $ev = $mapped->unmap ($wait_events...)
2155
2156Unmaps the mapped memory object, using the queue originally used to create
2157it, quite similarly to C<< $queue->unmap ($mapped, ...) >>.
2158
2159=item $bool = $mapped->mapped
2160
2161Returns whether the object is still mapped - true before an C<unmap> is
2162enqueued, false afterwards.
2163
2164=item $ev = $mapped->event
2165
2166Return the event object associated with the mapped object. Initially, this
2167will be the event object created when mapping the object, and after an
2168unmap, this will be the event object that the unmap operation created.
2169
2170=item $mapped->wait
2171
2172Same as C<< $mapped->event->wait >> - makes sure no operations on this
2173mapped object are outstanding.
2174
2175=item $bytes = $mapped->size
2176
2177Returns the size of the mapped area, in bytes. Same as C<length $$mapped>.
2178
2179=item $ptr = $mapped->ptr
2180
2181Returns the raw memory address of the mapped area.
2182
2183=item $mapped->set ($offset, $data)
2184
2185Replaces the data at the given C<$offset> in the memory area by the new
2186C<$data>. This method is safer than direct manipulation of C<$mapped>
2187because it does bounds-checking, but also slower.
2188
2189=item $data = $mapped->get ($offset, $length)
2190
2191Returns (without copying) a scalar representing the data at the given
2192C<$offset> and C<$length> in the mapped memory area. This is the same as
2193the following substr, except much slower;
2194
2195 $data = substr $$mapped, $offset, $length
2196
623=cut 2197=cut
624 2198
625package OpenCL; 2199sub OpenCL::Mapped::get {
626 2200 substr ${$_[0]}, $_[1], $_[2]
627use common::sense;
628
629BEGIN {
630 our $VERSION = '0.03';
631
632 require XSLoader;
633 XSLoader::load (__PACKAGE__, $VERSION);
634
635 @OpenCL::Buffer::ISA =
636 @OpenCL::Image::ISA = OpenCL::Memory::;
637
638 @OpenCL::Image2D::ISA =
639 @OpenCL::Image3D::ISA = OpenCL::Image::;
640
641 @OpenCL::UserEvent::ISA = OpenCL::Event::;
642} 2201}
2202
2203=back
2204
2205=head2 THE OpenCL::MappedBuffer CLASS
2206
2207This is a subclass of OpenCL::Mapped, representing mapped buffers.
2208
2209=head2 THE OpenCL::MappedImage CLASS
2210
2211This is a subclass of OpenCL::Mapped, representing mapped images.
2212
2213=over 4
2214
2215=item $bytes = $mapped->row_pitch
2216
2217=item $bytes = $mapped->slice_pitch
2218
2219Return the row or slice pitch of the image that has been mapped.
2220
2221=back
2222
2223
2224=cut
643 2225
6441; 22261;
645 2227
646=head1 AUTHOR 2228=head1 AUTHOR
647 2229

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines