… | |
… | |
160 | '; |
160 | '; |
161 | |
161 | |
162 | my $prog = $ctx->program_with_source ($src); |
162 | my $prog = $ctx->program_with_source ($src); |
163 | |
163 | |
164 | # build croaks on compile errors, so catch it and print the compile errors |
164 | # build croaks on compile errors, so catch it and print the compile errors |
165 | eval { $prog->build ($dev); 1 } |
165 | eval { $prog->build ($dev, "-cl-fast-relaxed-math"); 1 } |
166 | or die $prog->build_log; |
166 | or die $prog->build_log; |
167 | |
167 | |
168 | my $kernel = $prog->kernel ("squareit"); |
168 | my $kernel = $prog->kernel ("squareit"); |
169 | |
169 | |
170 | =head2 Create some input and output float buffers, then call the |
170 | =head2 Create some input and output float buffers, then call the |
… | |
… | |
210 | # enqueue an async read |
210 | # enqueue an async read |
211 | $ev = $queue->enqueue_read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data, $ev); |
211 | $ev = $queue->enqueue_read_buffer ($output, 0, 0, OpenCL::SIZEOF_FLOAT * 4, my $data, $ev); |
212 | |
212 | |
213 | # wait for the last event to complete |
213 | # wait for the last event to complete |
214 | $ev->wait; |
214 | $ev->wait; |
|
|
215 | |
|
|
216 | =head2 Use the OpenGL module to share a texture between OpenCL and OpenGL and draw some julia |
|
|
217 | set tunnel effect. |
|
|
218 | |
|
|
219 | This is quite a long example to get you going. |
|
|
220 | |
|
|
221 | use OpenGL ":all"; |
|
|
222 | use OpenCL; |
|
|
223 | |
|
|
224 | # open a window and create a gl texture |
|
|
225 | OpenGL::glpOpenWindow width => 256, height => 256; |
|
|
226 | my $texid = glGenTextures_p 1; |
|
|
227 | glBindTexture GL_TEXTURE_2D, $texid; |
|
|
228 | glTexImage2D_c GL_TEXTURE_2D, 0, GL_RGBA8, 256, 256, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0; |
|
|
229 | |
|
|
230 | # find and use the first opencl device that let's us get a shared opengl context |
|
|
231 | my $platform; |
|
|
232 | my $dev; |
|
|
233 | my $ctx; |
|
|
234 | |
|
|
235 | for (OpenCL::platforms) { |
|
|
236 | $platform = $_; |
|
|
237 | for ($platform->devices) { |
|
|
238 | $dev = $_; |
|
|
239 | $ctx = $platform->context ([OpenCL::GLX_DISPLAY_KHR, undef, OpenCL::GL_CONTEXT_KHR, undef], [$dev]) |
|
|
240 | and last; |
|
|
241 | } |
|
|
242 | } |
|
|
243 | |
|
|
244 | $ctx |
|
|
245 | or die "cannot find suitable OpenCL device\n"; |
|
|
246 | |
|
|
247 | my $queue = $ctx->queue ($dev); |
|
|
248 | |
|
|
249 | # now attach an opencl image2d object to the opengl texture |
|
|
250 | my $tex = $ctx->gl_texture2d (OpenCL::MEM_WRITE_ONLY, GL_TEXTURE_2D, 0, $texid); |
|
|
251 | |
|
|
252 | # now the boring opencl code |
|
|
253 | my $src = <<EOF; |
|
|
254 | kernel void |
|
|
255 | juliatunnel (write_only image2d_t img, float time) |
|
|
256 | { |
|
|
257 | float2 p = (float2)(get_global_id (0), get_global_id (1)) / 256.f * 2.f - 1.f; |
|
|
258 | |
|
|
259 | float2 m = (float2)(1.f, p.y) / fabs (p.x); |
|
|
260 | m.x = fabs (fmod (m.x + time * 0.05f, 4.f)) - 2.f; |
|
|
261 | |
|
|
262 | float2 z = m; |
|
|
263 | float2 c = (float2)(sin (time * 0.05005), cos (time * 0.06001)); |
|
|
264 | |
|
|
265 | for (int i = 0; i < 25 && dot (z, z) < 4.f; ++i) |
|
|
266 | z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c; |
|
|
267 | |
|
|
268 | float3 colour = (float3)(z.x, z.y, z.x * z.y); |
|
|
269 | write_imagef (img, (int2)(get_global_id (0), get_global_id (1)), (float4)(colour * p.x * p.x, 1.)); |
|
|
270 | } |
|
|
271 | EOF |
|
|
272 | my $prog = $ctx->program_with_source ($src); |
|
|
273 | eval { $prog->build ($dev); 1 } |
|
|
274 | or die $prog->build_log ($dev); |
|
|
275 | |
|
|
276 | my $kernel = $prog->kernel ("juliatunnel"); |
|
|
277 | |
|
|
278 | # program compiled, kernel ready, now draw and loop |
|
|
279 | |
|
|
280 | for (my $time; ; ++$time) { |
|
|
281 | # acquire objects from opengl |
|
|
282 | $queue->enqueue_acquire_gl_objects ([$tex]); |
|
|
283 | |
|
|
284 | # configure and run our kernel |
|
|
285 | $kernel->set_image2d (0, $tex); |
|
|
286 | $kernel->set_float (1, $time); |
|
|
287 | $queue->enqueue_nd_range_kernel ($kernel, undef, [256, 256], undef); |
|
|
288 | |
|
|
289 | # release objects to opengl again |
|
|
290 | $queue->enqueue_release_gl_objects ([$tex]); |
|
|
291 | |
|
|
292 | # wait |
|
|
293 | $queue->finish; |
|
|
294 | |
|
|
295 | # now draw the texture, the defaults should be all right |
|
|
296 | glTexParameterf GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST; |
|
|
297 | |
|
|
298 | glEnable GL_TEXTURE_2D; |
|
|
299 | glBegin GL_QUADS; |
|
|
300 | glTexCoord2f 0, 1; glVertex3i -1, -1, -1; |
|
|
301 | glTexCoord2f 0, 0; glVertex3i 1, -1, -1; |
|
|
302 | glTexCoord2f 1, 0; glVertex3i 1, 1, -1; |
|
|
303 | glTexCoord2f 1, 1; glVertex3i -1, 1, -1; |
|
|
304 | glEnd; |
|
|
305 | |
|
|
306 | glXSwapBuffers; |
|
|
307 | |
|
|
308 | select undef, undef, undef, 1/60; |
|
|
309 | } |
215 | |
310 | |
216 | =head1 DOCUMENTATION |
311 | =head1 DOCUMENTATION |
217 | |
312 | |
218 | =head2 BASIC CONVENTIONS |
313 | =head2 BASIC CONVENTIONS |
219 | |
314 | |
… | |
… | |
241 | =item * Structures are often specified by flattening out their components |
336 | =item * Structures are often specified by flattening out their components |
242 | as with short vectors, and returned as arrayrefs. |
337 | as with short vectors, and returned as arrayrefs. |
243 | |
338 | |
244 | =item * When enqueuing commands, the wait list is specified by adding |
339 | =item * When enqueuing commands, the wait list is specified by adding |
245 | extra arguments to the function - anywhere a C<$wait_events...> argument |
340 | extra arguments to the function - anywhere a C<$wait_events...> argument |
246 | is documented this can be any number of event objects. |
341 | is documented this can be any number of event objects. As an extsnion |
|
|
342 | implemented by this module, C<undef> values will be ignored in the event |
|
|
343 | list. |
247 | |
344 | |
248 | =item * When enqueuing commands, if the enqueue method is called in void |
345 | =item * When enqueuing commands, if the enqueue method is called in void |
249 | context, no event is created. In all other contexts an event is returned |
346 | context, no event is created. In all other contexts an event is returned |
250 | by the method. |
347 | by the method. |
251 | |
348 | |
… | |
… | |
275 | |
372 | |
276 | =head2 GLX SUPPORT |
373 | =head2 GLX SUPPORT |
277 | |
374 | |
278 | Due to the sad state that OpenGL support is in in Perl (mostly the OpenGL |
375 | Due to the sad state that OpenGL support is in in Perl (mostly the OpenGL |
279 | module, which has little to no documentation and has little to no support |
376 | module, which has little to no documentation and has little to no support |
280 | for glx), this module, as a special extension, treats context creation |
377 | for glX), this module, as a special extension, treats context creation |
281 | properties C<OpenCL::GLX_DISPLAY_KHR> and C<OpenCL::GL_CONTEXT_KHR> |
378 | properties C<OpenCL::GLX_DISPLAY_KHR> and C<OpenCL::GL_CONTEXT_KHR> |
282 | specially: If either or both of these are C<undef>, then the OpenCL |
379 | specially: If either or both of these are C<undef>, then the OpenCL |
283 | module tries to dynamically resolve C<glxGetCurrentDisplay> and |
380 | module tries to dynamically resolve C<glXGetCurrentDisplay> and |
284 | C<glxGetCurrentContext>, call these functions and use their return values |
381 | C<glXGetCurrentContext>, call these functions and use their return values |
285 | instead. |
382 | instead. |
286 | |
383 | |
287 | For this to work, the OpenGL library must be loaded, a GLX context must |
384 | For this to work, the OpenGL library must be loaded, a GLX context must |
288 | have been created and be made current, and C<dlsym> must be available and |
385 | have been created and be made current, and C<dlsym> must be available and |
289 | capable of finding the function via C<RTLD_DEFAULT>. |
386 | capable of finding the function via C<RTLD_DEFAULT>. |
… | |
… | |
653 | |
750 | |
654 | =item @device_partition_property_exts = $device->affinity_domains_ext |
751 | =item @device_partition_property_exts = $device->affinity_domains_ext |
655 | |
752 | |
656 | Calls C<clGetDeviceInfo> with C<CL_DEVICE_AFFINITY_DOMAINS_EXT> and returns the result. |
753 | Calls C<clGetDeviceInfo> with C<CL_DEVICE_AFFINITY_DOMAINS_EXT> and returns the result. |
657 | |
754 | |
658 | =item $uint = $device->reference_count_ext |
755 | =item $uint = $device->reference_count_ext |
659 | |
756 | |
660 | Calls C<clGetDeviceInfo> with C<CL_DEVICE_REFERENCE_COUNT_EXT > and returns the result. |
757 | Calls C<clGetDeviceInfo> with C<CL_DEVICE_REFERENCE_COUNT_EXT > and returns the result. |
661 | |
758 | |
662 | =item @device_partition_property_exts = $device->partition_style_ext |
759 | =item @device_partition_property_exts = $device->partition_style_ext |
663 | |
760 | |
… | |
… | |
674 | =item $queue = $ctx->queue ($device, $properties) |
771 | =item $queue = $ctx->queue ($device, $properties) |
675 | |
772 | |
676 | Create a new OpenCL::Queue object from the context and the given device. |
773 | Create a new OpenCL::Queue object from the context and the given device. |
677 | |
774 | |
678 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html> |
775 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html> |
|
|
776 | |
|
|
777 | Example: create an out-of-order queue. |
|
|
778 | |
|
|
779 | $queue = $ctx->queue ($device, OpenCL::QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE); |
679 | |
780 | |
680 | =item $ev = $ctx->user_event |
781 | =item $ev = $ctx->user_event |
681 | |
782 | |
682 | Creates a new OpenCL::UserEvent object. |
783 | Creates a new OpenCL::UserEvent object. |
683 | |
784 | |
… | |
… | |
794 | for completion, unless the method is called in void context, in which case |
895 | for completion, unless the method is called in void context, in which case |
795 | no event object is created. |
896 | no event object is created. |
796 | |
897 | |
797 | They also allow you to specify any number of other event objects that this |
898 | They also allow you to specify any number of other event objects that this |
798 | request has to wait for before it starts executing, by simply passing the |
899 | request has to wait for before it starts executing, by simply passing the |
799 | event objects as extra parameters to the enqueue methods. |
900 | event objects as extra parameters to the enqueue methods. To simplify |
|
|
901 | program design, this module ignores any C<undef> values in the list of |
|
|
902 | events. This makes it possible to code operations such as this, without |
|
|
903 | having to put a valid event object into C<$event> first: |
|
|
904 | |
|
|
905 | $event = $queue->enqueue_xxx (..., $event); |
800 | |
906 | |
801 | Queues execute in-order by default, without any parallelism, so in most |
907 | Queues execute in-order by default, without any parallelism, so in most |
802 | cases (i.e. you use only one queue) it's not necessary to wait for or |
908 | cases (i.e. you use only one queue) it's not necessary to wait for or |
803 | create event objects. |
909 | create event objects, althoguh an our of order queue is often a bit |
|
|
910 | faster. |
804 | |
911 | |
805 | =over 4 |
912 | =over 4 |
806 | |
913 | |
807 | =item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...) |
914 | =item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...) |
808 | |
915 | |
… | |
… | |
1274 | |
1381 | |
1275 | This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>. |
1382 | This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>. |
1276 | |
1383 | |
1277 | TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>, |
1384 | TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>, |
1278 | C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>, |
1385 | C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>, |
1279 | C<image2d>, C<image3d>, C<sampler> or C<event>. |
1386 | C<image2d>, C<image3d>, C<sampler>, C<local> or C<event>. |
1280 | |
1387 | |
1281 | Chars and integers (including the half type) are specified as integers, |
1388 | Chars and integers (including the half type) are specified as integers, |
1282 | float and double as floating point values, memory/buffer/image2d/image3d |
1389 | float and double as floating point values, memory/buffer/image2d/image3d |
1283 | must be an object of that type or C<undef>, and sampler and event must be |
1390 | must be an object of that type or C<undef>, local-memory arguments are |
1284 | objects of that type. |
1391 | set by specifying the size, and sampler and event must be objects of that |
|
|
1392 | type. |
1285 | |
1393 | |
1286 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html> |
1394 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html> |
1287 | |
1395 | |
1288 | =back |
1396 | =back |
1289 | |
1397 | |
… | |
… | |
1378 | package OpenCL; |
1486 | package OpenCL; |
1379 | |
1487 | |
1380 | use common::sense; |
1488 | use common::sense; |
1381 | |
1489 | |
1382 | BEGIN { |
1490 | BEGIN { |
1383 | our $VERSION = '0.92'; |
1491 | our $VERSION = '0.96'; |
1384 | |
1492 | |
1385 | require XSLoader; |
1493 | require XSLoader; |
1386 | XSLoader::load (__PACKAGE__, $VERSION); |
1494 | XSLoader::load (__PACKAGE__, $VERSION); |
1387 | |
1495 | |
1388 | @OpenCL::Buffer::ISA = |
1496 | @OpenCL::Buffer::ISA = |