|
Blender V5.3
|
Go to the source code of this file.
Macros | |
| #define | gpu_parallel_sort_bucket_pass gpu_parallel_sort_bucket_pass_impl |
| #define | gpu_parallel_sort_write_pass gpu_parallel_sort_write_pass_impl |
Functions | |
| CCL_NAMESPACE_BEGIN ccl_device_inline void | gpu_parallel_sort_bucket_pass_impl (const uint num_states, const uint partition_size, const uint max_shaders, const uint queued_kernel, ccl_global ushort *d_queued_kernel, ccl_global uint *d_shader_sort_key, ccl_global int *partition_key_offsets) |
| ccl_device_inline void | gpu_parallel_sort_write_pass_impl (const uint num_states, const uint partition_size, const int num_states_limit, ccl_global int *indices, const uint max_shaders, const uint queued_kernel, ccl_global ushort *d_queued_kernel, ccl_global uint *d_shader_sort_key, ccl_global int *partition_key_offsets) |
| template<typename GetKeyOp> | |
| __device__ void | gpu_parallel_sorted_index_array (const uint state_index, const uint num_states, const int num_states_limit, ccl_global int *indices, ccl_global int *num_indices, ccl_global int *key_counter, ccl_global int *key_prefix_sum, GetKeyOp get_key_op) |
| #define gpu_parallel_sort_bucket_pass gpu_parallel_sort_bucket_pass_impl |
Definition at line 135 of file parallel_sorted_index.h.
| #define gpu_parallel_sort_write_pass gpu_parallel_sort_write_pass_impl |
Definition at line 256 of file parallel_sorted_index.h.
| CCL_NAMESPACE_BEGIN ccl_device_inline void gpu_parallel_sort_bucket_pass_impl | ( | const uint | num_states, |
| const uint | partition_size, | ||
| const uint | max_shaders, | ||
| const uint | queued_kernel, | ||
| ccl_global ushort * | d_queued_kernel, | ||
| ccl_global uint * | d_shader_sort_key, | ||
| ccl_global int * | partition_key_offsets ) |
Definition at line 18 of file parallel_sorted_index.h.
References __KERNEL_METAL__, __KERNEL_ONEAPI__, ccl_device_inline, ccl_global, ccl_gpu_block_dim_x, ccl_gpu_block_idx_x, ccl_gpu_local_syncthreads, ccl_gpu_shared, ccl_gpu_syncthreads, ccl_gpu_thread_idx_x, gpu_parallel_sort_bucket_pass_impl(), local_size, min, and num_states.
Referenced by gpu_parallel_sort_bucket_pass_impl().
| ccl_device_inline void gpu_parallel_sort_write_pass_impl | ( | const uint | num_states, |
| const uint | partition_size, | ||
| const int | num_states_limit, | ||
| ccl_global int * | indices, | ||
| const uint | max_shaders, | ||
| const uint | queued_kernel, | ||
| ccl_global ushort * | d_queued_kernel, | ||
| ccl_global uint * | d_shader_sort_key, | ||
| ccl_global int * | partition_key_offsets ) |
Definition at line 139 of file parallel_sorted_index.h.
References __KERNEL_METAL__, __KERNEL_ONEAPI__, ccl_device_inline, ccl_global, ccl_gpu_block_dim_x, ccl_gpu_block_idx_x, ccl_gpu_local_syncthreads, ccl_gpu_shared, ccl_gpu_syncthreads, ccl_gpu_thread_idx_x, gpu_parallel_sort_write_pass_impl(), index, local_size, min, and num_states.
Referenced by gpu_parallel_sort_write_pass_impl().
| __device__ void gpu_parallel_sorted_index_array | ( | const uint | state_index, |
| const uint | num_states, | ||
| const int | num_states_limit, | ||
| ccl_global int * | indices, | ||
| ccl_global int * | num_indices, | ||
| ccl_global int * | key_counter, | ||
| ccl_global int * | key_prefix_sum, | ||
| GetKeyOp | get_key_op ) |
Definition at line 261 of file parallel_sorted_index.h.
References __device__, atomic_fetch_and_add_uint32(), ccl_global, gpu_parallel_sorted_index_array(), GPU_PARALLEL_SORTED_INDEX_INACTIVE_KEY, index, and num_states.
Referenced by gpu_parallel_sorted_index_array().