… | |
… | |
212 | $ev->wait; |
212 | $ev->wait; |
213 | |
213 | |
214 | =head2 Use the OpenGL module to share a texture between OpenCL and OpenGL and draw some julia |
214 | =head2 Use the OpenGL module to share a texture between OpenCL and OpenGL and draw some julia |
215 | set tunnel effect. |
215 | set tunnel effect. |
216 | |
216 | |
217 | This is quite a long example to get you going. |
217 | This is quite a long example to get you going - you can download it from |
|
|
218 | L<http://cvs.schmorp.de/OpenCL/examples/juliaflight>. |
218 | |
219 | |
219 | use OpenGL ":all"; |
220 | use OpenGL ":all"; |
220 | use OpenCL; |
221 | use OpenCL; |
221 | |
222 | |
|
|
223 | my $S = $ARGV[0] || 256; # window/texture size, smaller is faster |
|
|
224 | |
222 | # open a window and create a gl texture |
225 | # open a window and create a gl texture |
223 | OpenGL::glpOpenWindow width => 256, height => 256; |
226 | OpenGL::glpOpenWindow width => $S, height => $S; |
224 | my $texid = glGenTextures_p 1; |
227 | my $texid = glGenTextures_p 1; |
225 | glBindTexture GL_TEXTURE_2D, $texid; |
228 | glBindTexture GL_TEXTURE_2D, $texid; |
226 | glTexImage2D_c GL_TEXTURE_2D, 0, GL_RGBA8, 256, 256, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0; |
229 | glTexImage2D_c GL_TEXTURE_2D, 0, GL_RGBA8, $S, $S, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0; |
227 | |
230 | |
228 | # find and use the first opencl device that let's us get a shared opengl context |
231 | # find and use the first opencl device that let's us get a shared opengl context |
229 | my $platform; |
232 | my $platform; |
230 | my $dev; |
233 | my $dev; |
231 | my $ctx; |
234 | my $ctx; |
… | |
… | |
250 | # now the boring opencl code |
253 | # now the boring opencl code |
251 | my $src = <<EOF; |
254 | my $src = <<EOF; |
252 | kernel void |
255 | kernel void |
253 | juliatunnel (write_only image2d_t img, float time) |
256 | juliatunnel (write_only image2d_t img, float time) |
254 | { |
257 | { |
255 | float2 p = (float2)(get_global_id (0), get_global_id (1)) / 256.f * 2.f - 1.f; |
258 | int2 xy = (int2)(get_global_id (0), get_global_id (1)); |
|
|
259 | float2 p = convert_float2 (xy) / $S.f * 2.f - 1.f; |
256 | |
260 | |
257 | float2 m = (float2)(1.f, p.y) / fabs (p.x); |
261 | float2 m = (float2)(1.f, p.y) / fabs (p.x); // tunnel |
258 | m.x = fabs (fmod (m.x + time * 0.05f, 4.f)) - 2.f; |
262 | m.x = fabs (fmod (m.x + time * 0.05f, 4.f) - 2.f); |
259 | |
263 | |
260 | float2 z = m; |
264 | float2 z = m; |
261 | float2 c = (float2)(sin (time * 0.05005), cos (time * 0.06001)); |
265 | float2 c = (float2)(sin (time * 0.01133f), cos (time * 0.02521f)); |
262 | |
266 | |
263 | for (int i = 0; i < 25 && dot (z, z) < 4.f; ++i) |
267 | for (int i = 0; i < 25 && dot (z, z) < 4.f; ++i) // standard julia |
264 | z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c; |
268 | z = (float2)(z.x * z.x - z.y * z.y, 2.f * z.x * z.y) + c; |
265 | |
269 | |
266 | float3 colour = (float3)(z.x, z.y, z.x * z.y); |
270 | float3 colour = (float3)(z.x, z.y, atan2 (z.y, z.x)); |
267 | write_imagef (img, (int2)(get_global_id (0), get_global_id (1)), (float4)(colour * p.x * p.x, 1.)); |
271 | write_imagef (img, xy, (float4)(colour * p.x * p.x, 1.)); |
268 | } |
272 | } |
269 | EOF |
273 | EOF |
270 | |
274 | |
271 | my $prog = $ctx->build_program ($src); |
275 | my $prog = $ctx->build_program ($src); |
272 | my $kernel = $prog->kernel ("juliatunnel"); |
276 | my $kernel = $prog->kernel ("juliatunnel"); |
… | |
… | |
276 | for (my $time; ; ++$time) { |
280 | for (my $time; ; ++$time) { |
277 | # acquire objects from opengl |
281 | # acquire objects from opengl |
278 | $queue->acquire_gl_objects ([$tex]); |
282 | $queue->acquire_gl_objects ([$tex]); |
279 | |
283 | |
280 | # configure and run our kernel |
284 | # configure and run our kernel |
281 | $kernel->set_image2d (0, $tex); |
285 | $kernel->setf ("mf", $tex, $time*2); # mf = memory object, float |
282 | $kernel->set_float (1, $time); |
|
|
283 | $queue->nd_range_kernel ($kernel, undef, [256, 256], undef); |
286 | $queue->nd_range_kernel ($kernel, undef, [$S, $S], undef); |
284 | |
287 | |
285 | # release objects to opengl again |
288 | # release objects to opengl again |
286 | $queue->release_gl_objects ([$tex]); |
289 | $queue->release_gl_objects ([$tex]); |
287 | |
290 | |
288 | # wait |
291 | # wait |
… | |
… | |
301 | |
304 | |
302 | glXSwapBuffers; |
305 | glXSwapBuffers; |
303 | |
306 | |
304 | select undef, undef, undef, 1/60; |
307 | select undef, undef, undef, 1/60; |
305 | } |
308 | } |
|
|
309 | |
|
|
310 | =head2 How to modify the previous example to not rely on GL sharing. |
|
|
311 | |
|
|
312 | For those poor souls with only a sucky CPU OpenCL implementation, you |
|
|
313 | currently have to read the image into some perl scalar, and then modify a |
|
|
314 | texture or use glDrawPixels or so). |
|
|
315 | |
|
|
316 | First, when you don't need gl sharing, you can create the context much simpler: |
|
|
317 | |
|
|
318 | $ctx = $platform->context (undef, [$dev]) |
|
|
319 | |
|
|
320 | To use a texture, you would modify the above example by creating an |
|
|
321 | OpenCL::Image manually instead of deriving it from a texture: |
|
|
322 | |
|
|
323 | my $tex = $ctx->image2d (OpenCL::MEM_WRITE_ONLY, OpenCL::RGBA, OpenCL::UNORM_INT8, $S, $S); |
|
|
324 | |
|
|
325 | And in the darw loop, intead of acquire_gl_objects/release_gl_objects, you |
|
|
326 | would read the image2d after the kernel has written it: |
|
|
327 | |
|
|
328 | $queue->read_image ($tex, 0, 0, 0, 0, $S, $S, 1, 0, 0, my $data); |
|
|
329 | |
|
|
330 | And then you would upload the pixel data to the texture (or use glDrawPixels): |
|
|
331 | |
|
|
332 | glTexSubImage2D_s GL_TEXTURE_2D, 0, 0, 0, $S, $S, GL_RGBA, GL_UNSIGNED_BYTE, $data; |
|
|
333 | |
|
|
334 | The fully modified example can be found at |
|
|
335 | L<http://cvs.schmorp.de/OpenCL/examples/juliaflight-nosharing>. |
306 | |
336 | |
307 | =head1 DOCUMENTATION |
337 | =head1 DOCUMENTATION |
308 | |
338 | |
309 | =head2 BASIC CONVENTIONS |
339 | =head2 BASIC CONVENTIONS |
310 | |
340 | |
… | |
… | |
527 | @OpenCL::Image1D::ISA = |
557 | @OpenCL::Image1D::ISA = |
528 | @OpenCL::Image1DArray::ISA = |
558 | @OpenCL::Image1DArray::ISA = |
529 | @OpenCL::Image1DBuffer::ISA = OpenCL::Image::; |
559 | @OpenCL::Image1DBuffer::ISA = OpenCL::Image::; |
530 | |
560 | |
531 | @OpenCL::UserEvent::ISA = OpenCL::Event::; |
561 | @OpenCL::UserEvent::ISA = OpenCL::Event::; |
|
|
562 | |
|
|
563 | @OpenCL::MappedBuffer = |
|
|
564 | @OpenCL::MappedImage = OpenCL::Mapped::; |
532 | } |
565 | } |
533 | |
566 | |
534 | =head2 THE OpenCL PACKAGE |
567 | =head2 THE OpenCL PACKAGE |
535 | |
568 | |
536 | =over 4 |
569 | =over 4 |
… | |
… | |
1204 | |
1237 | |
1205 | =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...) |
1238 | =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...) |
1206 | |
1239 | |
1207 | http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBufferRect.html |
1240 | http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteBufferRect.html |
1208 | |
1241 | |
|
|
1242 | =item $ev = $queue->copy_buffer_to_image ($src_buffer, $dst_image, $src_offset, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...) |
|
|
1243 | |
|
|
1244 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html> |
|
|
1245 | |
1209 | =item $ev = $queue->read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) |
1246 | =item $ev = $queue->read_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) |
1210 | |
1247 | |
1211 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferRect.html> |
1248 | C<$row_pitch> (and C<$slice_pitch>) can be C<0>, in which case the OpenCL |
1212 | |
1249 | module uses the image width (and height) to supply default values. |
1213 | =item $ev = $queue->copy_buffer_to_image ($src_buffer, $dst_image, $src_offset, $dst_x, $dst_y, $dst_z, $width, $height, $depth, $wait_events...) |
|
|
1214 | |
1250 | |
1215 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html> |
1251 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueReadImage.html> |
1216 | |
1252 | |
1217 | =item $ev = $queue->write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) |
1253 | =item $ev = $queue->write_image ($src, $blocking, $x, $y, $z, $width, $height, $depth, $row_pitch, $slice_pitch, $data, $wait_events...) |
1218 | |
1254 | |
|
|
1255 | C<$row_pitch> (and C<$slice_pitch>) can be C<0>, in which case the OpenCL |
|
|
1256 | module uses the image width (and height) to supply default values. |
1219 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html> |
1257 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueWriteImage.html> |
1220 | |
1258 | |
1221 | =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...) |
1259 | =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...) |
1222 | |
1260 | |
1223 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html> |
1261 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyImage.html> |
… | |
… | |
1327 | =item $command_queue_properties = $command_queue->properties |
1365 | =item $command_queue_properties = $command_queue->properties |
1328 | |
1366 | |
1329 | Calls C<clGetCommandQueueInfo> with C<CL_QUEUE_PROPERTIES> and returns the result. |
1367 | Calls C<clGetCommandQueueInfo> with C<CL_QUEUE_PROPERTIES> and returns the result. |
1330 | |
1368 | |
1331 | =for gengetinfo end command_queue |
1369 | =for gengetinfo end command_queue |
|
|
1370 | |
|
|
1371 | =back |
|
|
1372 | |
|
|
1373 | =head3 MEMORY MAPPED BUFFERS |
|
|
1374 | |
|
|
1375 | OpenCL allows you to map buffers and images to host memory (read: perl |
|
|
1376 | scalars). This is done much like reading or copying a buffer, by enqueuing |
|
|
1377 | a map or unmap operation on the command queue. |
|
|
1378 | |
|
|
1379 | The map operations return a C<OpenCL::Mapped> object - see L<THE |
|
|
1380 | OpenCL::Mapped CLASS> section for details on what to do with these |
|
|
1381 | objects. |
|
|
1382 | |
|
|
1383 | The object will be unmapped automatically when the mapped object is |
|
|
1384 | destroyed (you can use a barrier to make sure the unmap has finished, |
|
|
1385 | before using the buffer in a kernel), but you can also enqueue an unmap |
|
|
1386 | operation manually. |
|
|
1387 | |
|
|
1388 | =over 4 |
|
|
1389 | |
|
|
1390 | =item $mapped_buffer = $queue->map_buffer ($buf, $data, $blocking=1, $map_flags=OpenCL::MAP_READ|OpenCL::MAP_WRITE, $offset=0, $size=0, $wait_events...) |
|
|
1391 | |
|
|
1392 | Maps the given buffer into host memory and returns a C<OpenCL::MappedBuffer> object. |
|
|
1393 | |
|
|
1394 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMapBuffer.html> |
|
|
1395 | |
|
|
1396 | =item $mapped_image = $queue->map_image ($img, $data, $blocking=1, $map_flags=OpenCL::MAP_READ|OpenCL::MAP_WRITE, $x=0, $y=0, $z=0, $width=0, $height=0, $depth=0, $wait_events...) |
|
|
1397 | |
|
|
1398 | Maps the given image area into host memory and return a |
|
|
1399 | C<OpenCL::MappedImage> object. Although there are default values for most |
|
|
1400 | arguments, you currently have to specify all arguments, otherwise the call |
|
|
1401 | will fail. |
|
|
1402 | |
|
|
1403 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueMapImage.html> |
|
|
1404 | |
|
|
1405 | =item $ev = $queue->unmap ($mapped, $wait_events...) |
|
|
1406 | |
|
|
1407 | Unmaps the data from host memory. You must not call any methods that |
|
|
1408 | modify the data, or modify the data scalar directly, after calling this |
|
|
1409 | method. |
|
|
1410 | |
|
|
1411 | The mapped event object will always be passed as part of the |
|
|
1412 | $wait_events. The mapped event object will be replaced by the new event |
|
|
1413 | object that this request creates. |
1332 | |
1414 | |
1333 | =back |
1415 | =back |
1334 | |
1416 | |
1335 | =head2 THE OpenCL::Memory CLASS |
1417 | =head2 THE OpenCL::Memory CLASS |
1336 | |
1418 | |
… | |
… | |
1880 | |
1962 | |
1881 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html> |
1963 | L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html> |
1882 | |
1964 | |
1883 | =back |
1965 | =back |
1884 | |
1966 | |
|
|
1967 | =head2 THE OpenCL::Mapped CLASS |
|
|
1968 | |
|
|
1969 | This class represents objects mapped into host memory. They are |
|
|
1970 | represented by a blessed string scalar. The string data is the mapped |
|
|
1971 | memory area, that is, if you read or write it, then the mapped object is |
|
|
1972 | accessed directly. |
|
|
1973 | |
|
|
1974 | You must only ever use operations that modify the string in-place - for |
|
|
1975 | example, a C<substr> that doesn't change the length, or maybe a regex that |
|
|
1976 | doesn't change the length. Any other operation might cause the data to be |
|
|
1977 | copied. |
|
|
1978 | |
|
|
1979 | When the object is destroyed it will enqueue an implicit unmap operation |
|
|
1980 | on the queue that was used to create it. |
|
|
1981 | |
|
|
1982 | Example, replace the first two floats in the mapped buffer by 1 and 2. |
|
|
1983 | |
|
|
1984 | my $mapped = $queue->map_buffer ($buf, ... |
|
|
1985 | $mapped->event->wait; # make sure it's there |
|
|
1986 | |
|
|
1987 | # now replace first 8 bytes by new data, which is exactly 8 bytes long |
|
|
1988 | # we blindly assume device endianness to equal host endianness |
|
|
1989 | # (and of course, we assume iee 754 single precision floats :) |
|
|
1990 | substr $$mapped, 0, 8, pack "f*", 1, 2; |
|
|
1991 | |
|
|
1992 | =over 4 |
|
|
1993 | |
|
|
1994 | =item $bool = $mapped->mapped |
|
|
1995 | |
|
|
1996 | Returns whether the object is still mapped - true before an C<unmap> is |
|
|
1997 | enqueued, false afterwards. |
|
|
1998 | |
|
|
1999 | =item $ev = $mapped->event |
|
|
2000 | |
|
|
2001 | Return the event object associated with the mapped object. Initially, this |
|
|
2002 | will be the event object created when mapping the object, and after an |
|
|
2003 | unmap, this will be the event object that the unmap operation created. |
|
|
2004 | |
|
|
2005 | =item $mapped->wait |
|
|
2006 | |
|
|
2007 | Same as C<< $mapped->event->wait >> - makes sure no operations on this |
|
|
2008 | mapped object are outstanding. |
|
|
2009 | |
|
|
2010 | =item $bytes = $mapped->size |
|
|
2011 | |
|
|
2012 | Returns the size of the mapped area, in bytes. Same as C<length $$mapped>. |
|
|
2013 | |
|
|
2014 | =item $ptr = $mapped->ptr |
|
|
2015 | |
|
|
2016 | Returns the raw memory address of the mapped area - same as C<$mapped+0>. |
|
|
2017 | |
|
|
2018 | =back |
|
|
2019 | |
|
|
2020 | =head2 THE OpenCL::MappedBuffer CLASS |
|
|
2021 | |
|
|
2022 | This is a subclass of OpenCL::Mapped, representing mapped buffers. |
|
|
2023 | |
|
|
2024 | =over 4 |
|
|
2025 | |
|
|
2026 | =back |
|
|
2027 | |
|
|
2028 | =head2 THE OpenCL::MappedImage CLASS |
|
|
2029 | |
|
|
2030 | This is a subclass of OpenCL::Mapped, representing mapped images. |
|
|
2031 | |
|
|
2032 | =over 4 |
|
|
2033 | |
|
|
2034 | =back |
|
|
2035 | |
|
|
2036 | |
1885 | =cut |
2037 | =cut |
1886 | |
2038 | |
1887 | 1; |
2039 | 1; |
1888 | |
2040 | |
1889 | =head1 AUTHOR |
2041 | =head1 AUTHOR |