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

Comparing OpenCL/OpenCL.pm (file contents):
Revision 1.35 by root, Thu Apr 19 13:18:31 2012 UTC vs.
Revision 1.50 by root, Tue Apr 24 13:30:49 2012 UTC

… …
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
217set tunnel effect.
218
219This 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 $queue->finish;
294
295 # now draw the texture, the defaults should be all right
296 glTexParameterf GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST;
297
298 glEnable GL_TEXTURE_2D;
299 glBegin GL_QUADS;
300 glTexCoord2f 0, 1; glVertex3i -1, -1, -1;
301 glTexCoord2f 0, 0; glVertex3i 1, -1, -1;
302 glTexCoord2f 1, 0; glVertex3i 1, 1, -1;
303 glTexCoord2f 1, 1; glVertex3i -1, 1, -1;
304 glEnd;
305
306 glXSwapBuffers;
307
308 select undef, undef, undef, 1/60;
309 }
215 310
216=head1 DOCUMENTATION 311=head1 DOCUMENTATION
217 312
218=head2 BASIC CONVENTIONS 313=head2 BASIC CONVENTIONS
219 314
… …
241=item * Structures are often specified by flattening out their components 336=item * Structures are often specified by flattening out their components
242as with short vectors, and returned as arrayrefs. 337as with short vectors, and returned as arrayrefs.
243 338
244=item * When enqueuing commands, the wait list is specified by adding 339=item * When enqueuing commands, the wait list is specified by adding
245extra arguments to the function - anywhere a C<$wait_events...> argument 340extra arguments to the function - anywhere a C<$wait_events...> argument
246is documented this can be any number of event objects. 341is documented this can be any number of event objects. As an extsnion
342implemented by this module, C<undef> values will be ignored in the event
343list.
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
249context, no event is created. In all other contexts an event is returned 346context, no event is created. In all other contexts an event is returned
250by the method. 347by the method.
251 348
… …
271 ulong IV - Q 368 ulong IV - Q
272 float NV float f 369 float NV float f
273 half IV ushort S 370 half IV ushort S
274 double NV double d 371 double NV double d
275 372
373=head2 GLX SUPPORT
374
375Due to the sad state that OpenGL support is in in Perl (mostly the OpenGL
376module, which has little to no documentation and has little to no support
377for glX), this module, as a special extension, treats context creation
378properties C<OpenCL::GLX_DISPLAY_KHR> and C<OpenCL::GL_CONTEXT_KHR>
379specially: If either or both of these are C<undef>, then the OpenCL
380module tries to dynamically resolve C<glXGetCurrentDisplay> and
381C<glXGetCurrentContext>, call these functions and use their return values
382instead.
383
384For this to work, the OpenGL library must be loaded, a GLX context must
385have been created and be made current, and C<dlsym> must be available and
386capable of finding the function via C<RTLD_DEFAULT>.
387
276=head2 THE OpenCL PACKAGE 388=head2 THE OpenCL PACKAGE
277 389
278=over 4 390=over 4
279 391
280=item $int = OpenCL::errno 392=item $int = OpenCL::errno
… …
344It's best to avoid this method and use one of the following convenience 456It's best to avoid this method and use one of the following convenience
345wrappers. 457wrappers.
346 458
347L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html> 459L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformInfo.html>
348 460
461=item $platform->unload_compiler
462
463Attempts to unload the compiler for this platform, for endless
464profit. Does nothing on OpenCL 1.1.
465
466L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clUnloadPlatformCompiler.html>
467
349=for gengetinfo begin platform 468=for gengetinfo begin platform
350 469
351=item $string = $platform->profile 470=item $string = $platform->profile
352 471
353Calls C<clGetPlatformInfo> with C<CL_PLATFORM_PROFILE> and returns the result. 472Calls C<clGetPlatformInfo> with C<CL_PLATFORM_PROFILE> and returns the result.
… …
638 757
639=item @device_partition_property_exts = $device->affinity_domains_ext 758=item @device_partition_property_exts = $device->affinity_domains_ext
640 759
641Calls C<clGetDeviceInfo> with C<CL_DEVICE_AFFINITY_DOMAINS_EXT> and returns the result. 760Calls C<clGetDeviceInfo> with C<CL_DEVICE_AFFINITY_DOMAINS_EXT> and returns the result.
642 761
643=item $uint = $device->reference_count_ext 762=item $uint = $device->reference_count_ext
644 763
645Calls C<clGetDeviceInfo> with C<CL_DEVICE_REFERENCE_COUNT_EXT > and returns the result. 764Calls C<clGetDeviceInfo> with C<CL_DEVICE_REFERENCE_COUNT_EXT> and returns the result.
646 765
647=item @device_partition_property_exts = $device->partition_style_ext 766=item @device_partition_property_exts = $device->partition_style_ext
648 767
649Calls C<clGetDeviceInfo> with C<CL_DEVICE_PARTITION_STYLE_EXT> and returns the result. 768Calls C<clGetDeviceInfo> with C<CL_DEVICE_PARTITION_STYLE_EXT> and returns the result.
650 769
… …
659=item $queue = $ctx->queue ($device, $properties) 778=item $queue = $ctx->queue ($device, $properties)
660 779
661Create a new OpenCL::Queue object from the context and the given device. 780Create a new OpenCL::Queue object from the context and the given device.
662 781
663L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html> 782L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateCommandQueue.html>
783
784Example: create an out-of-order queue.
785
786 $queue = $ctx->queue ($device, OpenCL::QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE);
664 787
665=item $ev = $ctx->user_event 788=item $ev = $ctx->user_event
666 789
667Creates a new OpenCL::UserEvent object. 790Creates a new OpenCL::UserEvent object.
668 791
… …
678=item $buf = $ctx->buffer_sv ($flags, $data) 801=item $buf = $ctx->buffer_sv ($flags, $data)
679 802
680Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and 803Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object and
681initialise it with the given data values. 804initialise it with the given data values.
682 805
806=item $img = $ctx->image ($self, $flags, $channel_order, $channel_type, $type, $width, $height, $depth, $array_size = 0, $row_pitch = 0, $slice_pitch = 0, $num_mip_level = 0, $num_samples = 0, $*data = &PL_sv_undef)
807
808Creates a new OpenCL::Image object and optionally initialises it with
809the given data values.
810
811L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateImage.html>
812
683=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $row_pitch = 0, $data = undef) 813=item $img = $ctx->image2d ($flags, $channel_order, $channel_type, $width, $height, $row_pitch = 0, $data = undef)
684 814
685Creates a new OpenCL::Image2D object and optionally initialises it with 815Creates a new OpenCL::Image2D object and optionally initialises it with
686the given data values. 816the given data values.
687 817
… …
699Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object that refers to the given 829Creates a new OpenCL::Buffer (actually OpenCL::BufferObj) object that refers to the given
700OpenGL buffer object. 830OpenGL buffer object.
701 831
702http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLBuffer.html 832http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLBuffer.html
703 833
834=item $img = $ctx->gl_texture ($flags, $target, $miplevel, $texture)
835
836Creates a new OpenCL::Image object that refers to the given OpenGL
837texture object or buffer.
838
839http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clCreateFromGLTexture.html
840
704=item $ctx->gl_texture2d ($flags, $target, $miplevel, $texture) 841=item $img = $ctx->gl_texture2d ($flags, $target, $miplevel, $texture)
705 842
706Creates a new OpenCL::Image2D object that refers to the given OpenGL 843Creates a new OpenCL::Image2D object that refers to the given OpenGL
7072D texture object. 8442D texture object.
708 845
709http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture2D.html 846http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture2D.html
710 847
711=item $ctx->gl_texture3d ($flags, $target, $miplevel, $texture) 848=item $img = $ctx->gl_texture3d ($flags, $target, $miplevel, $texture)
712 849
713Creates a new OpenCL::Image3D object that refers to the given OpenGL 850Creates a new OpenCL::Image3D object that refers to the given OpenGL
7143D texture object. 8513D texture object.
715 852
716http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture3D.html 853http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateFromGLTexture3D.html
… …
779for completion, unless the method is called in void context, in which case 916for completion, unless the method is called in void context, in which case
780no event object is created. 917no event object is created.
781 918
782They also allow you to specify any number of other event objects that this 919They also allow you to specify any number of other event objects that this
783request has to wait for before it starts executing, by simply passing the 920request has to wait for before it starts executing, by simply passing the
784event objects as extra parameters to the enqueue methods. 921event objects as extra parameters to the enqueue methods. To simplify
922program design, this module ignores any C<undef> values in the list of
923events. This makes it possible to code operations such as this, without
924having to put a valid event object into C<$event> first:
925
926 $event = $queue->enqueue_xxx (..., $event);
785 927
786Queues execute in-order by default, without any parallelism, so in most 928Queues execute in-order by default, without any parallelism, so in most
787cases (i.e. you use only one queue) it's not necessary to wait for or 929cases (i.e. you use only one queue) it's not necessary to wait for or
788create event objects. 930create event objects, althoguh an our of order queue is often a bit
931faster.
789 932
790=over 4 933=over 4
791 934
792=item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...) 935=item $ev = $queue->enqueue_read_buffer ($buffer, $blocking, $offset, $len, $data, $wait_events...)
793 936
… …
859reference to an array of local work sizes, with the same number of 1002reference to an array of local work sizes, with the same number of
860elements as @$global_work_size. 1003elements as @$global_work_size.
861 1004
862L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html> 1005L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueNDRangeKernel.html>
863 1006
864=item $ev = $queue->enqueue_marker ($wait_events...)
865
866L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMarker.html>
867
868=item $ev = $queue->enqueue_acquire_gl_objects ([object, ...], $wait_events...) 1007=item $ev = $queue->enqueue_acquire_gl_objects ([object, ...], $wait_events...)
869 1008
870Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired 1009Enqueues a list (an array-ref of OpenCL::Memory objects) to be acquired
871for subsequent OpenCL usage. 1010for subsequent OpenCL usage.
872 1011
… …
881 1020
882=item $ev = $queue->enqueue_wait_for_events ($wait_events...) 1021=item $ev = $queue->enqueue_wait_for_events ($wait_events...)
883 1022
884L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html> 1023L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWaitForEvents.html>
885 1024
886=item $queue->enqueue_barrier 1025=item $ev = $queue->enqueue_marker ($wait_events...)
887 1026
1027L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueMarkerWithWaitList.html>
1028
1029=item $ev = $queue->enqueue_barrier ($wait_events...)
1030
888L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueBarrier.html> 1031L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueBarrierWithWaitList.html>
889 1032
890=item $queue->flush 1033=item $queue->flush
891 1034
892L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html> 1035L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clFlush.html>
893 1036
… …
1008 1151
1009=back 1152=back
1010 1153
1011=head2 THE OpenCL::Image CLASS 1154=head2 THE OpenCL::Image CLASS
1012 1155
1013This is the superclass of all image objects - OpenCL::Image2D and OpenCL::Image3D. 1156This is the superclass of all image objects - OpenCL::Image1D,
1157OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D,
1158OpenCL::Image2DArray and OpenCL::Image3D.
1014 1159
1015=over 4 1160=over 4
1016 1161
1017=item $packed_value = $ev->image_info ($name) 1162=item $packed_value = $ev->image_info ($name)
1018 1163
… …
1053 1198
1054=for gengetinfo begin gl_texture 1199=for gengetinfo begin gl_texture
1055 1200
1056=item $GLenum = $gl_texture->target 1201=item $GLenum = $gl_texture->target
1057 1202
1058Calls C<clGetGlTextureInfo> with C<CL_GL_TEXTURE_TARGET> and returns the result. 1203Calls C<clGetGLTextureInfo> with C<CL_GL_TEXTURE_TARGET> and returns the result.
1059 1204
1060=item $GLint = $gl_texture->gl_mipmap_level 1205=item $GLint = $gl_texture->gl_mipmap_level
1061 1206
1062Calls C<clGetGlTextureInfo> with C<CL_GL_MIPMAP_LEVEL> and returns the result. 1207Calls C<clGetGLTextureInfo> with C<CL_GL_MIPMAP_LEVEL> and returns the result.
1063 1208
1064=for gengetinfo end gl_texture 1209=for gengetinfo end gl_texture
1065 1210
1066=back 1211=back
1067 1212
… …
1122 1267
1123Creates an OpenCL::Kernel object out of the named C<__kernel> function in 1268Creates an OpenCL::Kernel object out of the named C<__kernel> function in
1124the program. 1269the program.
1125 1270
1126L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html> 1271L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernel.html>
1272
1273=item @kernels = $program->kernels_in_program
1274
1275Returns all kernels successfully compiled for all devices in program.
1276
1277http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateKernelsInProgram.html
1127 1278
1128=for gengetinfo begin program_build 1279=for gengetinfo begin program_build
1129 1280
1130=item $build_status = $program->build_status ($device) 1281=item $build_status = $program->build_status ($device)
1131 1282
… …
1259 1410
1260This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>. 1411This is a family of methods to set the kernel argument with the number C<$index> to the give C<$value>.
1261 1412
1262TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>, 1413TYPE is one of C<char>, C<uchar>, C<short>, C<ushort>, C<int>, C<uint>,
1263C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>, 1414C<long>, C<ulong>, C<half>, C<float>, C<double>, C<memory>, C<buffer>,
1264C<image2d>, C<image3d>, C<sampler> or C<event>. 1415C<image2d>, C<image3d>, C<sampler>, C<local> or C<event>.
1265 1416
1266Chars and integers (including the half type) are specified as integers, 1417Chars and integers (including the half type) are specified as integers,
1267float and double as floating point values, memory/buffer/image2d/image3d 1418float and double as floating point values, memory/buffer/image2d/image3d
1268must be an object of that type or C<undef>, and sampler and event must be 1419must be an object of that type or C<undef>, local-memory arguments are
1269objects of that type. 1420set by specifying the size, and sampler and event must be objects of that
1421type.
1422
1423Setting an argument for a kernel does NOT keep a reference to the object -
1424for example, if you set an argument to some image object, free the image,
1425and call the kernel, you will run into undefined behaviour.
1270 1426
1271L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html> 1427L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetKernelArg.html>
1272 1428
1273=back 1429=back
1274 1430
… …
1363package OpenCL; 1519package OpenCL;
1364 1520
1365use common::sense; 1521use common::sense;
1366 1522
1367BEGIN { 1523BEGIN {
1368 our $VERSION = '0.92'; 1524 our $VERSION = '0.96';
1369 1525
1370 require XSLoader; 1526 require XSLoader;
1371 XSLoader::load (__PACKAGE__, $VERSION); 1527 XSLoader::load (__PACKAGE__, $VERSION);
1372 1528
1373 @OpenCL::Buffer::ISA = 1529 @OpenCL::Buffer::ISA =
1374 @OpenCL::Image::ISA = OpenCL::Memory::; 1530 @OpenCL::Image::ISA = OpenCL::Memory::;
1375 1531
1376 @OpenCL::BufferObj::ISA = OpenCL::Buffer::; 1532 @OpenCL::BufferObj::ISA = OpenCL::Buffer::;
1377 1533
1378 @OpenCL::Image2D::ISA = 1534 @OpenCL::Image2D::ISA =
1535 @OpenCL::Image3D::ISA =
1536 @OpenCL::Image2DArray::ISA =
1537 @OpenCL::Image1D::ISA =
1538 @OpenCL::Image1DArray::ISA =
1379 @OpenCL::Image3D::ISA = OpenCL::Image::; 1539 @OpenCL::Image1DBuffer::ISA = OpenCL::Image::;
1380 1540
1381 @OpenCL::UserEvent::ISA = OpenCL::Event::; 1541 @OpenCL::UserEvent::ISA = OpenCL::Event::;
1382} 1542}
1383 1543
13841; 15441;
1385 1545
1386=head1 AUTHOR 1546=head1 AUTHOR

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines