diff options
| author | Sergey Sharybin <sergey.vfx@gmail.com> | 2016-08-01 16:40:46 +0300 |
|---|---|---|
| committer | Sergey Sharybin <sergey.vfx@gmail.com> | 2016-08-01 16:54:29 +0300 |
| commit | 6353ecb996898b4ce2fe8065130ed1f5ea3b6989 (patch) | |
| tree | b6d620152e4ff7920465d8396fe443dc9b3ffc56 /intern/cycles/kernel/kernel_path_surface.h | |
| parent | 7065022f7aa23ba13d2999e1e40162a8f480af0e (diff) | |
Cycles: Tweaks to support CUDA 8 toolkit
All the changes are mainly giving explicit tips on inlining functions,
so they match how inlining worked with previous toolkit.
This make kernel compiled by CUDA 8 render in average with same speed
as previous kernels. Some scenes are somewhat faster, some of them are
somewhat slower. But slowdown is within 1% so far.
On a positive side it allows us to enable newer generation cards on
buildbots (so GTX 10x0 will be officially supported soon).
Diffstat (limited to 'intern/cycles/kernel/kernel_path_surface.h')
| -rw-r--r-- | intern/cycles/kernel/kernel_path_surface.h | 9 |
1 files changed, 7 insertions, 2 deletions
diff --git a/intern/cycles/kernel/kernel_path_surface.h b/intern/cycles/kernel/kernel_path_surface.h index 74b1ae0ca32..250b8e92a45 100644 --- a/intern/cycles/kernel/kernel_path_surface.h +++ b/intern/cycles/kernel/kernel_path_surface.h @@ -222,8 +222,13 @@ ccl_device_inline void kernel_path_surface_connect_light(KernelGlobals *kg, ccl_ #endif /* path tracing: bounce off or through surface to with new direction stored in ray */ -ccl_device_inline bool kernel_path_surface_bounce(KernelGlobals *kg, ccl_addr_space RNG *rng, - ShaderData *sd, ccl_addr_space float3 *throughput, ccl_addr_space PathState *state, PathRadiance *L, ccl_addr_space Ray *ray) +ccl_device bool kernel_path_surface_bounce(KernelGlobals *kg, + ccl_addr_space RNG *rng, + ShaderData *sd, + ccl_addr_space float3 *throughput, + ccl_addr_space PathState *state, + PathRadiance *L, + ccl_addr_space Ray *ray) { /* no BSDF? we can stop here */ if(ccl_fetch(sd, flag) & SD_BSDF) { |