Revision 8d96ea8 by Campbell Barton June 14, 2014, 08:12 (GMT) |
BLI_rand: add BLI_rng_get_float_unit_v2 |
Revision b54793e by Campbell Barton June 14, 2014, 07:55 (GMT) |
Editmesh: remove redundant normal calculation call for subdivide |
Revision 577e4f8 by Campbell Barton June 14, 2014, 07:40 (GMT) |
UI: remove OpenGL calls from ED_region_init, now handled elsewhere |
Revision 8cd9d78 by Campbell Barton June 14, 2014, 07:34 (GMT) |
Replace sqrt with hypot for wipe-effect & transform code |
Revision 94d4b31 by Campbell Barton June 14, 2014, 07:10 (GMT) |
Math Lib: mat3_to_eulo2 & mat3_to_eul2 mixed float/double differently replace sqrt with hypotf to avoid precision loss instead |
Revision 019a8bc by Campbell Barton June 14, 2014, 07:00 (GMT) |
Quiet warning in Cycles |
Revision 29e8c46 by Campbell Barton June 14, 2014, 06:27 (GMT) |
Rename 'Extend Vertex' to be clear it operates on many vertices |
Revision 788f485 by Campbell Barton June 14, 2014, 06:27 (GMT) |
Comment unused macro |
Revision 746f0ad by Campbell Barton June 13, 2014, 22:27 (GMT) |
Polyfill2d: use kd-tree Simple search for intersections became slow for larger concave ngons (100+) Tested to work with ngons up to 75k sides, performance is approx ~6x faster then scanfill. This is a 2D version of BLI_kdtree with modifications: - nodes can be removed - an index -> node map is stored (especially for tessellation) |
Revision 19b1da2 by Campbell Barton June 13, 2014, 22:21 (GMT) |
Polyfill2d: avoid calculating polygon winding (its known in all cases) |
Revision 7529e36 by Campbell Barton June 13, 2014, 22:21 (GMT) |
Polyfill2d: Switch directions on concave triangles Better topology and minor speedup |
Revision f0f45ee by Campbell Barton June 13, 2014, 22:21 (GMT) |
Polyfill2d: replace array with linklist, faster resizing approx 4.0x speedup |
Revision 0ce3a75 by Thomas Dinges June 13, 2014, 21:40 (GMT) |
Cycles: Add support for uchar4 attributes. * Added support for uchar4 attributes to Cycles' attribute system. * This is used for Vertex Colors now, which saves some memory (4 unsigned characters, instead of 4 floats). * GPU Texture Limit on sm_20 and sm_21 decreased from 95 to 94, because we need a new texture for the uchar4 attributes. This is no problem for sm_30 or newer. Part of my GSoC 2014. |
Revision 7e20583 by Sergey Sharybin June 13, 2014, 21:35 (GMT) |
Attempt to fix sign conversion error happening on buildbot |
Revision 6cd5954 by Thomas Dinges June 13, 2014, 21:23 (GMT) |
Implement GLSL code for XYZ nodes. |
Revision 2c105dd by Sergey Sharybin June 13, 2014, 21:10 (GMT) |
Follow up for previous commit: need to tweak args for scons as well |
Revision 1d982ea by jens verwiebe June 13, 2014, 21:08 (GMT) |
OSX: update codesigning rules to 2.71 |
Revision b48d0a6 by Sergey Sharybin June 13, 2014, 21:02 (GMT) |
Cycles: Fix compilation error of AVX2 kernel with GCC/Clang |
Revision be182d9 by Thomas Dinges June 13, 2014, 20:26 (GMT) |
Code cleanup. |
Revision 866c7fb by Thomas Dinges June 13, 2014, 20:26 (GMT) |
Cycles: Add an AVX2 CPU kernel. This kernel is compiled with AVX2, FMA3, and BMI compiler flags. At the moment only Intel Haswell benefits from this, but future AMD CPUs will have these instructions as well. Makes rendering on Haswell CPUs a few percent faster, only benchmarked with clang on OS X though. Part of my GSoC 2014. |
|
|
|


Master Commits
MiikaHweb | 2003-2021