May 15, 2015, 08:22 (GMT) |
Cycles: Disable bake OpenCL kernel for NVidia devices prior to sm_30 Driver fails to compile kernel in reasonable time for those devices here, so for easier testing of the OpenCL split kernel work disabling bake kernel for now. |
May 15, 2015, 08:22 (GMT) |
Cycles: Pass device compute capabilities to kernel via build options This way it's possible to do device-selective feature disabling/enabling. Currently only supported for NVidia devices via OpenCL extension. |
May 15, 2015, 08:22 (GMT) |
OpenCL wrangler: Add some of the extension defines |
May 15, 2015, 08:22 (GMT) |
Cycles: Cleanup, move build options string calculation into the device class This way it's easier to access platform name, device ID and other stuff which might be needed to define build options. |
May 15, 2015, 07:14 (GMT) |
Fix T42510: Limit by seams fails in edge/vert mode |
May 15, 2015, 07:07 (GMT) |
BMesh: add loop-shell walker |
May 15, 2015, 01:29 (GMT) |
Correct recent cleanup Removed call which was still needed |
May 15, 2015, 00:55 (GMT) |
BMesh: rename loop walker -> edgeloop |
May 14, 2015, 23:25 (GMT) |
Add material reorder buttons for Cycles as well |
May 14, 2015, 23:08 (GMT) |
Fix some issues found by Coverity Scan Some of them are just brain dead code, some are potential bugs. |
May 14, 2015, 19:48 (GMT) |
Cycles: Enable object motion blur on Intel OpenCL platform This required allocating some memory related on object transform needed by ShaderData and currently it is done for all the platforms. Since we're targeting full feature-complete platforms this is rather acceptable at this point and in the future we'll do selective NO_HAIR/NO_SSS/NO_BLUR kernels. This is experimental still and in fact there're some major issues on NVidia platform and it's not really clear if it's a bug in compiler, some uninitizlied variable or other kind of issue. |
May 14, 2015, 19:42 (GMT) |
Cycles: Various fixes Some stupid fixes like spaces around operator and missing semicolon, plus fix for wrong detecting of ShaderData SOA size. Thar was harmless since there's only one closure array, but still better to fix this. |
May 14, 2015, 18:44 (GMT) |
Cycles: Remove meaningless ifdef checks for features in device_opencl This file was actually checking for features enabled on CPU and surely all of them were enabled, so removing them does not cause any difference. ideally we'll need to do runtime feature detection and just pass some stuff as NULL to the kernel, or maybe also have variadic kernel entry points which is also possible quite easily. |
May 14, 2015, 18:35 (GMT) |
Cycles: Enable camera motion blur in split kernel for Intel/NVidia It's good for testing and seems to work quite reliably here. This probably not totally cheap in terms of performance, but this we could solve quite easily by selective kernel compilation once other things are tested/proved to be reliable. |
May 14, 2015, 17:45 (GMT) |
Cycles: add some quick logging. |
May 14, 2015, 17:42 (GMT) |
Fix missing object drawing update when done exporting. |
May 14, 2015, 17:17 (GMT) |
cleanup: GPUBufferCopyFunc defined twice |
Revision 6b8ca6a by Julian Eisel (temp_widgets_c++_experiment, temp_widgets_files_refactor, temp_widgets_update_tagging, UI-experiments, wiggly-widgets) May 14, 2015, 16:52 (GMT) |
Merge branch 'master' into wiggly-widgets |
May 14, 2015, 16:51 (GMT) |
Merge branch 'master' into UI-experiments |
May 14, 2015, 15:09 (GMT) |
Fix fluid transformation matrix. There's a little bit of float precision issue, but I'll call it a day for now... |
|
|
|


Master Commits
MiikaHweb | 2003-2021