diff options
author | Sergey Sharybin <sergey.vfx@gmail.com> | 2015-04-27 22:13:03 +0300 |
---|---|---|
committer | Sergey Sharybin <sergey.vfx@gmail.com> | 2015-04-27 22:38:32 +0300 |
commit | ae7d84dbc1936ef7ddd00f9c22d074389f97f04f (patch) | |
tree | 7103596bc78785ed769e90c1570d654cc14a839d /intern/cycles/render/buffers.cpp | |
parent | 278ff15c7f3c4a91ba049b7a13ecf8a4183ebf67 (diff) |
Cycles: Use native saturate function for CUDA
This more a workaround for CUDA optimizer which can't optimize clamp(x, 0, 1)
into a single instruction and uses 4 instructions instead.
Original patch by @lockal with own modification:
Don't make changes outside of the kernel. They don't make any difference
anyway and term saturate() has a bit different meaning outside of kernel.
This gives around 2% of speedup in Barcelona file, but in more complex shader
setups with lots of math nodes with clamping speedup could be much nicer.
Subscribers: dingto
Projects: #cycles
Differential Revision: https://developer.blender.org/D1224
Diffstat (limited to 'intern/cycles/render/buffers.cpp')
-rw-r--r-- | intern/cycles/render/buffers.cpp | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/intern/cycles/render/buffers.cpp b/intern/cycles/render/buffers.cpp index 5202bf5862c..dc00da1bf88 100644 --- a/intern/cycles/render/buffers.cpp +++ b/intern/cycles/render/buffers.cpp @@ -187,7 +187,7 @@ bool RenderBuffers::get_pass_rect(PassType type, float exposure, int sample, int else if(type == PASS_MIST) { for(int i = 0; i < size; i++, in += pass_stride, pixels++) { float f = *in; - pixels[0] = clamp(f*scale_exposure, 0.0f, 1.0f); + pixels[0] = saturate(f*scale_exposure); } } #ifdef WITH_CYCLES_DEBUG @@ -298,7 +298,7 @@ bool RenderBuffers::get_pass_rect(PassType type, float exposure, int sample, int pixels[2] = f.z*scale_exposure; /* clamp since alpha might be > 1.0 due to russian roulette */ - pixels[3] = clamp(f.w*scale, 0.0f, 1.0f); + pixels[3] = saturate(f.w*scale); } } } |