Program options for executing the kernel programs. More...
#include <irendering_kernel_programs.h>
Public Attributes | |
mi::Sint32 | max_registers |
Define the maximum number of registers to use for the program. More... | |
mi::math::Vector_struct< mi::Uint32, 2 > | block_size |
Define the execution block size of the rendering kernel using the program. More... | |
mi::Uint32 | debug_flags |
Internal debug options. More... | |
Program options for executing the kernel programs.
mi::math::Vector_struct<mi::Uint32, 2> nv::index::IRendering_kernel_program::Program_options::block_size |
Define the execution block size of the rendering kernel using the program.
Special values: (0, 0) - let NVIDIA IndeX decide the block size) (default value: (0, 0))
mi::Uint32 nv::index::IRendering_kernel_program::Program_options::debug_flags |
Internal debug options.
(default value: 0)
mi::Sint32 nv::index::IRendering_kernel_program::Program_options::max_registers |
Define the maximum number of registers to use for the program.
Special values: 0 - use NVIDIA IndeX internal limit -1 - no limit, let CUDA decide the register count (default value: 0)