libflame
revision_anchor
|
Data Structures | |
struct | FLA_Obj_gpu_struct |
struct | FLASH_Queue_variables |
struct | FLASH_Queue_variables |
Typedefs | |
typedef struct FLA_Obj_gpu_struct | FLA_Obj_gpu |
typedef struct FLASH_Queue_variables | FLASH_Queue_vars |
typedef struct FLA_Obj_gpu_struct FLA_Obj_gpu |
typedef struct FLASH_Queue_variables FLASH_Queue_vars |
References FLA_Obj_view::base, FLA_Obj_gpu_struct::clean, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_variables::gpu, FLASH_Queue_variables::gpu_lock, FLA_Obj_gpu_struct::obj, FLA_Obj_gpu_struct::request, and FLASH_Queue_variables::victim.
Referenced by FLASH_Queue_check_gpu().
FLA_Bool FLASH_Queue_check_gpu | ( | FLASH_Task * | t, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLA_Obj_col_stride(), FLA_Obj_elemtype(), FLA_Obj_length(), FLA_Obj_width(), FLASH_Queue_check_block_gpu(), FLASH_Queue_get_num_threads(), i, FLASH_Task_s::input_arg, FLASH_Task_s::n_input_args, FLASH_Task_s::n_output_args, FLA_Obj_gpu_struct::obj, FLASH_Task_s::output_arg, and FLASH_Task_s::thread.
Referenced by FLASH_Queue_exec_gpu().
void FLASH_Queue_create_gpu | ( | int | thread, |
void * | arg | ||
) |
References FLASH_Queue_variables::block_size, FLA_Obj_gpu_struct::buffer_gpu, FLASH_Queue_variables::datatype, FLASH_Queue_alloc_gpu(), FLASH_Queue_bind_gpu(), FLASH_Queue_get_enabled_gpu(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_variables::gpu, and i.
Referenced by FLASH_Queue_exec_parallel_function().
void FLASH_Queue_destroy_gpu | ( | int | thread, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLA_Obj_gpu_struct::buffer_gpu, FLA_Obj_gpu_struct::clean, FLASH_Queue_free_gpu(), FLASH_Queue_get_enabled_gpu(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_read_gpu(), FLASH_Queue_variables::gpu, i, and FLA_Obj_gpu_struct::obj.
Referenced by FLASH_Queue_exec_parallel_function().
void FLASH_Queue_exec | ( | void | ) |
References FLASH_Queue_variables::all_lock, FLA_Obj_view::base, FLA_Obj_gpu_struct::buffer_gpu, FLASH_Queue_variables::cac_lock, FLASH_Queue_variables::cache, FLA_Obj_gpu_struct::clean, FLASH_Queue_variables::dep_lock, FLA_Clock(), FLA_free(), FLA_is_owner(), FLA_Lock_destroy(), FLA_Lock_init(), FLA_malloc(), FLA_shfree(), FLA_shmalloc(), FLASH_Queue_exec_parallel(), FLASH_Queue_exec_parallel_function(), FLASH_Queue_exec_simulation(), FLASH_Queue_get_block_size(), FLASH_Queue_get_cache_size(), FLASH_Queue_get_caching(), FLASH_Queue_get_cores_per_cache(), FLASH_Queue_get_cores_per_queue(), FLASH_Queue_get_data_affinity(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_get_num_tasks(), FLASH_Queue_get_num_threads(), FLASH_Queue_get_verbose_output(), FLASH_Queue_get_work_stealing(), FLASH_Queue_init_tasks(), FLASH_Queue_reset(), FLASH_Queue_set_caching(), FLASH_Queue_set_data_affinity(), FLASH_Queue_set_parallel_time(), FLASH_Queue_set_work_stealing(), FLASH_Queue_verbose_output(), FLASH_Task_free(), for(), FLASH_Queue_variables::gpu, FLASH_Queue_variables::gpu_lock, FLASH_Queue_variables::gpu_log, FLASH_Queue_s::head, i, FLASH_Queue_variables::n_caches, FLASH_Queue_variables::n_queues, FLASH_Queue_variables::n_ready, FLASH_Queue_s::n_tasks, FLASH_Queue_variables::n_wait, FLA_Obj_gpu_struct::obj, FLASH_Queue_variables::pc, FLASH_Queue_variables::prefetch, RCCE_wtime(), FLA_Obj_gpu_struct::request, FLASH_Queue_variables::run_lock, FLASH_Queue_variables::size, Synch_all(), FLASH_Queue_s::tail, FLASH_Queue_variables::task_queue, FLASH_Queue_variables::victim, FLASH_Queue_variables::wait_queue, and FLASH_Queue_variables::war_lock.
FLA_Bool FLASH_Queue_exec_gpu | ( | FLASH_Task * | t, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLASH_Task_s::enabled_gpu, FLA_free(), FLA_Lock_acquire(), FLA_Lock_release(), FLA_malloc(), FLA_Obj_col_stride(), FLA_Obj_elemtype(), FLA_Obj_length(), FLA_Obj_width(), FLASH_Queue_check_gpu(), FLASH_Queue_exec_task(), FLASH_Queue_exec_task_gpu(), FLASH_Queue_flush_block_gpu(), FLASH_Queue_get_enabled_gpu(), FLASH_Queue_get_num_threads(), FLASH_Queue_invalidate_block_gpu(), FLASH_Queue_mark_gpu(), FLASH_Queue_update_gpu(), FLASH_Queue_wait_enqueue(), FLASH_Task_s::hit, i, FLASH_Task_s::input_arg, FLASH_Task_s::n_input_args, FLASH_Task_s::n_output_args, FLA_Obj_gpu_struct::obj, FLASH_Task_s::output_arg, FLASH_Task_s::queue, FLASH_Queue_variables::run_lock, and FLASH_Task_s::thread.
Referenced by FLASH_Queue_exec_parallel_function().
void FLASH_Queue_exec_parallel | ( | void * | arg | ) |
References FLA_Check_pthread_create_result(), FLA_Check_pthread_join_result(), FLA_free(), FLA_malloc(), FLASH_Queue_exec_parallel_function(), FLASH_Queue_get_num_threads(), and i.
Referenced by FLASH_Queue_exec().
void* FLASH_Queue_exec_parallel_function | ( | void * | arg | ) |
References FLASH_Thread_s::args, FLASH_Task_s::cache, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_create_gpu(), FLASH_Queue_destroy_gpu(), FLASH_Queue_exec_gpu(), FLASH_Queue_exec_task(), FLASH_Queue_flush_gpu(), FLASH_Queue_get_caching(), FLASH_Queue_get_cores_per_cache(), FLASH_Queue_get_enabled_gpu(), FLASH_Queue_get_num_tasks(), FLASH_Queue_get_num_threads(), FLASH_Queue_get_work_stealing(), FLASH_Queue_prefetch(), FLASH_Queue_update_cache(), FLASH_Queue_wait_dequeue(), FLASH_Queue_work_stealing(), FLASH_Task_free_parallel(), FLASH_Task_update_dependencies(), i, FLASH_Thread_s::id, FLASH_Queue_variables::n_queues, RCCE_acquire_lock(), RCCE_release_lock(), RCCE_ue(), and FLASH_Task_s::thread.
Referenced by FLASH_Queue_exec(), and FLASH_Queue_exec_parallel().
void FLASH_Queue_exec_simulation | ( | void * | arg | ) |
References FLASH_Task_s::cache, FLASH_Task_s::dep_arg_head, FLA_free(), FLA_malloc(), FLASH_Queue_exec_task(), FLASH_Queue_get_cores_per_cache(), FLASH_Queue_get_num_tasks(), FLASH_Queue_get_num_threads(), FLASH_Queue_get_verbose_output(), FLASH_Queue_prefetch(), FLASH_Queue_update_cache(), FLASH_Queue_wait_dequeue(), FLASH_Queue_wait_enqueue(), FLASH_Task_free(), i, FLASH_Task_s::n_dep_args, FLASH_Queue_variables::n_queues, FLASH_Task_s::n_ready, FLASH_Task_s::name, FLASH_Dep_s::next_dep, FLASH_Queue_variables::pc, RCCE_acquire_lock(), RCCE_release_lock(), RCCE_ue(), RCCE_wtime(), Synch_all(), FLASH_Dep_s::task, and FLASH_Task_s::thread.
Referenced by FLASH_Queue_exec().
void FLASH_Queue_flush_block_gpu | ( | FLA_Obj | obj, |
int | thread, | ||
void * | arg | ||
) |
References FLA_Obj_view::base, FLA_Obj_gpu_struct::buffer_gpu, FLA_Obj_gpu_struct::clean, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_read_gpu(), FLASH_Queue_variables::gpu, FLASH_Queue_variables::gpu_lock, FLA_Obj_gpu_struct::obj, and FLA_Obj_gpu_struct::request.
Referenced by FLASH_Queue_exec_gpu().
void FLASH_Queue_flush_gpu | ( | int | thread, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLA_Obj_gpu_struct::buffer_gpu, FLA_Obj_gpu_struct::clean, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_enabled_gpu(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_read_gpu(), FLASH_Queue_variables::gpu, FLASH_Queue_variables::gpu_lock, FLASH_Queue_variables::gpu_log, i, FLA_Obj_gpu_struct::obj, and FLA_Obj_gpu_struct::request.
Referenced by FLASH_Queue_exec_parallel_function().
void FLASH_Queue_init_tasks | ( | void * | arg | ) |
References FLA_Obj_view::base, FLASH_Queue_variables::block_size, FLASH_Queue_variables::datatype, FLASH_Task_s::dep_arg_head, FLA_is_owner(), FLA_Obj_col_stride(), FLA_Obj_datatype(), FLA_Obj_datatype_size(), FLA_Obj_elemtype(), FLA_Obj_free_buffer_task(), FLA_Obj_length(), FLA_Obj_width(), FLASH_Queue_get_data_affinity(), FLASH_Queue_get_head_task(), FLASH_Queue_get_num_tasks(), FLASH_Queue_get_tail_task(), FLASH_Queue_wait_enqueue(), FLASH_Task_s::func, FLASH_Task_s::height, i, FLASH_Task_s::input_arg, FLA_Obj_struct::m_index, FLASH_Task_s::n_dep_args, FLA_Obj_struct::n_index, FLASH_Task_s::n_input_args, FLASH_Task_s::n_macro_args, FLASH_Task_s::n_output_args, FLASH_Queue_variables::n_queues, FLASH_Task_s::n_ready, FLASH_Queue_variables::n_ready, FLASH_Task_s::n_war_args, FLA_Obj_struct::n_write_blocks, FLASH_Dep_s::next_dep, FLASH_Task_s::next_task, FLA_Obj_gpu_struct::obj, FLASH_Task_s::output_arg, FLASH_Queue_variables::prefetch, FLASH_Task_s::prev_task, FLASH_Task_s::queue, RCCE_acquire_lock(), RCCE_release_lock(), FLASH_Queue_variables::size, FLASH_Dep_s::task, and FLASH_Queue_variables::task_queue.
Referenced by FLASH_Queue_exec().
void FLASH_Queue_invalidate_block_gpu | ( | FLA_Obj | obj, |
int | thread, | ||
void * | arg | ||
) |
References FLA_Obj_view::base, FLA_Obj_gpu_struct::clean, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_variables::gpu, FLASH_Queue_variables::gpu_lock, FLA_Obj_gpu_struct::obj, and FLA_Obj_gpu_struct::request.
Referenced by FLASH_Queue_exec_gpu(), and FLASH_Queue_update_gpu().
void FLASH_Queue_mark_gpu | ( | FLASH_Task * | t, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLA_Obj_gpu_struct::clean, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_variables::gpu, FLASH_Queue_variables::gpu_lock, i, FLASH_Task_s::n_output_args, FLA_Obj_gpu_struct::obj, FLASH_Task_s::output_arg, FLA_Obj_gpu_struct::request, and FLASH_Task_s::thread.
Referenced by FLASH_Queue_exec_gpu().
void FLASH_Queue_prefetch | ( | int | cache, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLASH_Queue_variables::cache, FLASH_Queue_prefetch_block(), i, FLA_Obj_gpu_struct::obj, FLASH_Queue_variables::prefetch, and FLASH_Queue_variables::size.
Referenced by FLASH_Queue_exec_parallel_function(), and FLASH_Queue_exec_simulation().
void FLASH_Queue_prefetch_block | ( | FLA_Obj | obj | ) |
References FLA_Obj_datatype(), FLA_Obj_elem_size(), FLA_Obj_length(), FLA_Obj_width(), FLASH_Queue_get_cache_line_size(), i, scomplex::real, and dcomplex::real.
Referenced by FLASH_Queue_prefetch().
void FLASH_Queue_update_block_gpu | ( | FLA_Obj | obj, |
void ** | buffer_gpu, | ||
int | thread, | ||
void * | arg | ||
) |
References FLA_Obj_view::base, FLA_Obj_gpu_struct::buffer_gpu, FLA_Obj_gpu_struct::clean, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_read_gpu(), FLASH_Queue_write_gpu(), FLASH_Queue_variables::gpu, FLASH_Queue_variables::gpu_lock, FLA_Obj_gpu_struct::obj, FLA_Obj_gpu_struct::request, and FLASH_Queue_variables::victim.
Referenced by FLASH_Queue_update_gpu().
void FLASH_Queue_update_cache | ( | FLASH_Task * | t, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLASH_Task_s::cache, FLA_Obj_col_stride(), FLA_Obj_elemtype(), FLA_Obj_length(), FLA_Obj_width(), FLASH_Queue_update_cache_block(), i, FLASH_Task_s::input_arg, FLASH_Task_s::n_input_args, FLASH_Task_s::n_output_args, FLA_Obj_gpu_struct::obj, and FLASH_Task_s::output_arg.
Referenced by FLASH_Queue_exec_parallel_function(), and FLASH_Queue_exec_simulation().
References FLA_Obj_view::base, FLASH_Queue_variables::cac_lock, FLASH_Queue_variables::cache, FLA_Lock_acquire(), FLA_Lock_release(), i, FLASH_Queue_variables::n_caches, and FLASH_Queue_variables::size.
Referenced by FLASH_Queue_update_cache().
void FLASH_Queue_update_gpu | ( | FLASH_Task * | t, |
void ** | input_arg, | ||
void ** | output_arg, | ||
void * | arg | ||
) |
References FLA_Obj_view::base, FLASH_Queue_get_num_threads(), FLASH_Queue_invalidate_block_gpu(), FLASH_Queue_update_block_gpu(), i, FLASH_Task_s::input_arg, FLASH_Task_s::n_input_args, FLASH_Task_s::n_output_args, FLASH_Task_s::output_arg, and FLASH_Task_s::thread.
Referenced by FLASH_Queue_exec_gpu().
FLASH_Task* FLASH_Queue_wait_dequeue | ( | int | queue, |
int | cache, | ||
void * | arg | ||
) |
References FLASH_Queue_variables::cac_lock, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_caching(), FLASH_Queue_get_enabled_gpu(), FLASH_Queue_wait_dequeue_block(), FLASH_Queue_variables::gpu_lock, FLASH_Queue_s::head, FLASH_Queue_s::n_tasks, FLASH_Queue_variables::n_wait, FLASH_Task_s::next_wait, FLASH_Queue_variables::pc, FLASH_Task_s::prev_wait, FLASH_Queue_s::tail, FLASH_Queue_variables::task_queue, and FLASH_Queue_variables::wait_queue.
Referenced by FLASH_Queue_exec_parallel_function(), FLASH_Queue_exec_simulation(), and FLASH_Task_update_dependencies().
FLASH_Task* FLASH_Queue_wait_dequeue_block | ( | int | queue, |
int | cache, | ||
void * | arg | ||
) |
References FLA_Obj_view::base, FLASH_Queue_variables::cache, FLA_Obj_elemtype(), FLASH_Queue_get_enabled_gpu(), FLASH_Queue_get_gpu_num_blocks(), FLASH_Queue_variables::gpu, FLASH_Queue_s::head, FLASH_Task_s::hit, i, FLASH_Task_s::n_output_args, FLASH_Queue_s::n_tasks, FLASH_Task_s::next_wait, FLA_Obj_gpu_struct::obj, FLASH_Task_s::output_arg, FLASH_Queue_variables::size, and FLASH_Queue_variables::wait_queue.
Referenced by FLASH_Queue_wait_dequeue().
void FLASH_Queue_wait_enqueue | ( | FLASH_Task * | t, |
void * | arg | ||
) |
References FLASH_Queue_get_sorting(), FLASH_Queue_s::head, FLASH_Task_s::height, i, FLASH_Queue_s::n_tasks, FLASH_Queue_variables::n_wait, FLASH_Task_s::next_wait, FLASH_Task_s::order, FLASH_Queue_variables::pc, FLASH_Task_s::prev_wait, FLASH_Task_s::queue, FLASH_Queue_s::tail, FLASH_Queue_variables::task_queue, and FLASH_Queue_variables::wait_queue.
Referenced by FLASH_Queue_exec_gpu(), FLASH_Queue_exec_simulation(), FLASH_Queue_init_tasks(), FLASH_Task_update_binding(), and FLASH_Task_update_dependencies().
FLASH_Task* FLASH_Queue_work_stealing | ( | int | queue, |
void * | arg | ||
) |
References FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_s::head, FLASH_Queue_variables::n_queues, FLASH_Queue_s::n_tasks, FLASH_Task_s::next_wait, FLASH_Task_s::prev_wait, FLASH_Task_s::queue, FLASH_Queue_variables::run_lock, FLASH_Queue_s::tail, and FLASH_Queue_variables::wait_queue.
Referenced by FLASH_Queue_exec_parallel_function().
void FLASH_Task_free_parallel | ( | FLASH_Task * | t, |
void * | arg | ||
) |
References FLA_Obj_view::base, FLASH_Task_s::dep_arg_head, FLASH_Task_s::fla_arg, FLA_free(), FLA_Lock_acquire(), FLA_Lock_release(), FLA_Obj_col_stride(), FLA_Obj_elemtype(), FLA_Obj_length(), FLA_Obj_width(), FLASH_Queue_get_num_threads(), i, FLASH_Task_s::input_arg, FLASH_Task_s::int_arg, FLASH_Task_s::n_dep_args, FLASH_Task_s::n_input_args, FLASH_Task_s::n_output_args, FLA_Obj_struct::n_read_blocks, FLA_Obj_struct::n_read_tasks, FLASH_Dep_s::next_dep, FLA_Obj_gpu_struct::obj, FLASH_Task_s::output_arg, FLA_Obj_struct::read_task_head, FLA_Obj_struct::read_task_tail, FLASH_Queue_variables::war_lock, and FLA_Obj_struct::write_task.
Referenced by FLASH_Queue_exec_parallel_function().
FLASH_Task* FLASH_Task_update_binding | ( | FLASH_Task * | t, |
FLASH_Task * | r, | ||
void * | arg | ||
) |
References FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_sorting(), FLASH_Queue_wait_enqueue(), FLASH_Task_s::height, FLASH_Task_s::hit, FLASH_Task_s::queue, and FLASH_Queue_variables::run_lock.
Referenced by FLASH_Task_update_dependencies().
FLASH_Task* FLASH_Task_update_dependencies | ( | FLASH_Task * | t, |
void * | arg | ||
) |
References FLASH_Task_s::cache, FLASH_Task_s::dep_arg_head, FLASH_Queue_variables::dep_lock, FLA_Lock_acquire(), FLA_Lock_release(), FLASH_Queue_get_caching(), FLASH_Queue_get_num_threads(), FLASH_Queue_get_work_stealing(), FLASH_Queue_wait_dequeue(), FLASH_Queue_wait_enqueue(), FLASH_Task_update_binding(), i, FLASH_Task_s::n_dep_args, FLASH_Task_s::n_ready, FLASH_Queue_variables::n_ready, FLASH_Dep_s::next_dep, FLASH_Task_s::order, FLASH_Task_s::queue, RCCE_acquire_lock(), RCCE_release_lock(), FLASH_Queue_variables::run_lock, and FLASH_Dep_s::task.
Referenced by FLASH_Queue_exec_parallel_function().
int RCCE_acquire_lock | ( | int | ) |
int RCCE_release_lock | ( | int | ) |
int RCCE_ue | ( | void | ) |
Referenced by FLASH_Queue_exec_parallel_function(), and FLASH_Queue_exec_simulation().
double RCCE_wtime | ( | void | ) |
Referenced by FLASH_Queue_exec(), and FLASH_Queue_exec_simulation().
void Synch_all | ( | ) |
Referenced by FLASH_Queue_exec(), and FLASH_Queue_exec_simulation().