diff options
| author | Sergey Sharybin <sergey.vfx@gmail.com> | 2016-01-14 12:53:05 +0300 |
|---|---|---|
| committer | Sergey Sharybin <sergey.vfx@gmail.com> | 2016-01-14 12:53:05 +0300 |
| commit | 1f273cec00feddb1065847e3c8163cdcf8a6d89a (patch) | |
| tree | 5d24030df25f02e3122340eabcc591767ef05576 /intern/cycles/kernel/svm/svm_tex_coord.h | |
| parent | 5af103fe008aac0cb20631871bbee16f319835ed (diff) | |
Cycles: Tweak inline policy for some functions
The goal is to make Experimental kernel closer in performance to the
official kernel, avoiding spills and such.
There should not be big impact on official kernel, own tests showed
few percent performance drop on laptop's GPU. CPU was always the
same speed on AVX, AVX2 and SSE4.1 CPUs i've been testing here.
This seems to be the last essential step before we can get rid of
Experimental kernel and enable SSS officially on GPU without causing
some major performance issues.
Surely some more tweaks are possibly required, but that we can do
for until cows go home anyway.
Diffstat (limited to 'intern/cycles/kernel/svm/svm_tex_coord.h')
| -rw-r--r-- | intern/cycles/kernel/svm/svm_tex_coord.h | 24 |
1 files changed, 12 insertions, 12 deletions
diff --git a/intern/cycles/kernel/svm/svm_tex_coord.h b/intern/cycles/kernel/svm/svm_tex_coord.h index eebd9bee420..27fed89fdf7 100644 --- a/intern/cycles/kernel/svm/svm_tex_coord.h +++ b/intern/cycles/kernel/svm/svm_tex_coord.h @@ -99,12 +99,12 @@ ccl_device void svm_node_tex_coord(KernelGlobals *kg, stack_store_float3(stack, out_offset, data); } -ccl_device void svm_node_tex_coord_bump_dx(KernelGlobals *kg, - ShaderData *sd, - int path_flag, - float *stack, - uint4 node, - int *offset) +ccl_device_inline void svm_node_tex_coord_bump_dx(KernelGlobals *kg, + ShaderData *sd, + int path_flag, + float *stack, + uint4 node, + int *offset) { #ifdef __RAY_DIFFERENTIALS__ float3 data; @@ -184,12 +184,12 @@ ccl_device void svm_node_tex_coord_bump_dx(KernelGlobals *kg, #endif } -ccl_device void svm_node_tex_coord_bump_dy(KernelGlobals *kg, - ShaderData *sd, - int path_flag, - float *stack, - uint4 node, - int *offset) +ccl_device_inline void svm_node_tex_coord_bump_dy(KernelGlobals *kg, + ShaderData *sd, + int path_flag, + float *stack, + uint4 node, + int *offset) { #ifdef __RAY_DIFFERENTIALS__ float3 data; |