Page MenuHome
Differential D14081 Diff 48207 intern/cycles/kernel/device/gpu/kernel.h

Changeset View

Standalone View

View Options

intern/cycles/kernel/device/gpu/kernel.h

Show First 20 Lines • Show All 289 Lines • ▼ Show 20 Lines ccl_gpu_kernel_signature(integrator_queued_paths_array,
ccl_global int *indices, ccl_global int *indices,
ccl_global int *num_indices, ccl_global int *num_indices,
int kernel_index) int kernel_index)
{ {
ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, path, queued_kernel) == kernel_index, ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, path, queued_kernel) == kernel_index,
int kernel_index); int kernel_index);
ccl_gpu_kernel_lambda_pass.kernel_index = kernel_index; ccl_gpu_kernel_lambda_pass.kernel_index = kernel_index;
gpu_parallel_active_index_array<GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE>( gpu_parallel_active_index_array(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE,
num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass); num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass);
} }
ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE) ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE)
ccl_gpu_kernel_signature(integrator_queued_shadow_paths_array, ccl_gpu_kernel_signature(integrator_queued_shadow_paths_array,
int num_states, int num_states,
ccl_global int *indices, ccl_global int *indices,
ccl_global int *num_indices, ccl_global int *num_indices,
int kernel_index) int kernel_index)
{ {
ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, shadow_path, queued_kernel) == kernel_index, ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, shadow_path, queued_kernel) == kernel_index,
int kernel_index); int kernel_index);
ccl_gpu_kernel_lambda_pass.kernel_index = kernel_index; ccl_gpu_kernel_lambda_pass.kernel_index = kernel_index;
gpu_parallel_active_index_array<GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE>( gpu_parallel_active_index_array(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE,
num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass); num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass);
} }
ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE) ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE)
ccl_gpu_kernel_signature(integrator_active_paths_array, ccl_gpu_kernel_signature(integrator_active_paths_array,
int num_states, int num_states,
ccl_global int *indices, ccl_global int *indices,
ccl_global int *num_indices) ccl_global int *num_indices)
{ {
ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, path, queued_kernel) != 0); ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, path, queued_kernel) != 0);
gpu_parallel_active_index_array<GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE>( gpu_parallel_active_index_array(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE,
num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass); num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass);
} }
ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE) ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE)
ccl_gpu_kernel_signature(integrator_terminated_paths_array, ccl_gpu_kernel_signature(integrator_terminated_paths_array,
int num_states, int num_states,
ccl_global int *indices, ccl_global int *indices,
ccl_global int *num_indices, ccl_global int *num_indices,
int indices_offset) int indices_offset)
{ {
ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, path, queued_kernel) == 0); ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, path, queued_kernel) == 0);
gpu_parallel_active_index_array<GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE>( gpu_parallel_active_index_array(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE,
num_states, indices + indices_offset, num_indices, ccl_gpu_kernel_lambda_pass); num_states, indices + indices_offset, num_indices, ccl_gpu_kernel_lambda_pass);
} }
ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE) ccl_gpu_kernel_threads(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE)
ccl_gpu_kernel_signature(integrator_terminated_shadow_paths_array, ccl_gpu_kernel_signature(integrator_terminated_shadow_paths_array,
int num_states, int num_states,
ccl_global int *indices, ccl_global int *indices,
ccl_global int *num_indices, ccl_global int *num_indices,
int indices_offset) int indices_offset)
{ {
ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, shadow_path, queued_kernel) == 0); ccl_gpu_kernel_lambda(INTEGRATOR_STATE(state, shadow_path, queued_kernel) == 0);
gpu_parallel_active_index_array<GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE>( gpu_parallel_active_index_array(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE,
num_states, indices + indices_offset, num_indices, ccl_gpu_kernel_lambda_pass); num_states, indices + indices_offset, num_indices, ccl_gpu_kernel_lambda_pass);
} }
ccl_gpu_kernel_threads(GPU_PARALLEL_SORTED_INDEX_DEFAULT_BLOCK_SIZE) ccl_gpu_kernel_threads(GPU_PARALLEL_SORTED_INDEX_DEFAULT_BLOCK_SIZE)
ccl_gpu_kernel_signature(integrator_sorted_paths_array, ccl_gpu_kernel_signature(integrator_sorted_paths_array,
int num_states, int num_states,
int num_states_limit, int num_states_limit,
ccl_global int *indices, ccl_global int *indices,
Show All 26 Lines ccl_gpu_kernel_signature(integrator_compact_paths_array,
ccl_global int *num_indices, ccl_global int *num_indices,
int num_active_paths) int num_active_paths)
{ {
ccl_gpu_kernel_lambda((state >= num_active_paths) && ccl_gpu_kernel_lambda((state >= num_active_paths) &&
(INTEGRATOR_STATE(state, path, queued_kernel) != 0), (INTEGRATOR_STATE(state, path, queued_kernel) != 0),
int num_active_paths); int num_active_paths);
ccl_gpu_kernel_lambda_pass.num_active_paths = num_active_paths; ccl_gpu_kernel_lambda_pass.num_active_paths = num_active_paths;
gpu_parallel_active_index_array<GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE>( gpu_parallel_active_index_array(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE,
num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass); num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass);
} }
ccl_gpu_kernel_threads(GPU_PARALLEL_SORTED_INDEX_DEFAULT_BLOCK_SIZE) ccl_gpu_kernel_threads(GPU_PARALLEL_SORTED_INDEX_DEFAULT_BLOCK_SIZE)
ccl_gpu_kernel_signature(integrator_compact_states, ccl_gpu_kernel_signature(integrator_compact_states,
ccl_global const int *active_terminated_states, ccl_global const int *active_terminated_states,
const int active_states_offset, const int active_states_offset,
const int terminated_states_offset, const int terminated_states_offset,
Show All 16 Lines ccl_gpu_kernel_signature(integrator_compact_shadow_paths_array,
ccl_global int *num_indices, ccl_global int *num_indices,
int num_active_paths) int num_active_paths)
{ {
ccl_gpu_kernel_lambda((state >= num_active_paths) && ccl_gpu_kernel_lambda((state >= num_active_paths) &&
(INTEGRATOR_STATE(state, shadow_path, queued_kernel) != 0), (INTEGRATOR_STATE(state, shadow_path, queued_kernel) != 0),
int num_active_paths); int num_active_paths);
ccl_gpu_kernel_lambda_pass.num_active_paths = num_active_paths; ccl_gpu_kernel_lambda_pass.num_active_paths = num_active_paths;
gpu_parallel_active_index_array<GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE>( gpu_parallel_active_index_array(GPU_PARALLEL_ACTIVE_INDEX_DEFAULT_BLOCK_SIZE,
num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass); num_states, indices, num_indices, ccl_gpu_kernel_lambda_pass);
} }
ccl_gpu_kernel_threads(GPU_PARALLEL_SORTED_INDEX_DEFAULT_BLOCK_SIZE) ccl_gpu_kernel_threads(GPU_PARALLEL_SORTED_INDEX_DEFAULT_BLOCK_SIZE)
ccl_gpu_kernel_signature(integrator_compact_shadow_states, ccl_gpu_kernel_signature(integrator_compact_shadow_states,
ccl_global const int *active_terminated_states, ccl_global const int *active_terminated_states,
const int active_states_offset, const int active_states_offset,
const int terminated_states_offset, const int terminated_states_offset,
▲ Show 20 Lines • Show All 520 Lines • Show Last 20 Lines