1
0
mirror of https://github.com/godotengine/godot.git synced 2025-11-07 12:30:27 +00:00

Remove ThreadWorkPool, replace by WorkerThreadPool

The former needs to be allocated once per usage. The later is shared for all threads, which is more efficient.
It can also be better debugged.
This commit is contained in:
Juan Linietsky
2022-07-23 19:12:41 +02:00
parent 3bd74cd67b
commit c7255388e1
31 changed files with 248 additions and 515 deletions

View File

@@ -2518,14 +2518,14 @@ void RendererSceneCull::render_camera(RID p_render_buffers, RID p_camera, RID p_
RENDER_TIMESTAMP("Update Occlusion Buffer")
// For now just cull on the first camera
RendererSceneOcclusionCull::get_singleton()->buffer_update(p_viewport, camera_data.main_transform, camera_data.main_projection, camera_data.is_orthogonal, RendererThreadPool::singleton->thread_work_pool);
RendererSceneOcclusionCull::get_singleton()->buffer_update(p_viewport, camera_data.main_transform, camera_data.main_projection, camera_data.is_orthogonal);
_render_scene(&camera_data, p_render_buffers, environment, camera->effects, camera->visible_layers, p_scenario, p_viewport, p_shadow_atlas, RID(), -1, p_screen_mesh_lod_threshold, true, r_render_info);
#endif
}
void RendererSceneCull::_visibility_cull_threaded(uint32_t p_thread, VisibilityCullData *cull_data) {
uint32_t total_threads = RendererThreadPool::singleton->thread_work_pool.get_thread_count();
uint32_t total_threads = WorkerThreadPool::get_singleton()->get_thread_count();
uint32_t bin_from = p_thread * cull_data->cull_count / total_threads;
uint32_t bin_to = (p_thread + 1 == total_threads) ? cull_data->cull_count : ((p_thread + 1) * cull_data->cull_count / total_threads);
@@ -2622,7 +2622,7 @@ bool RendererSceneCull::_visibility_parent_check(const CullData &p_cull_data, co
void RendererSceneCull::_scene_cull_threaded(uint32_t p_thread, CullData *cull_data) {
uint32_t cull_total = cull_data->scenario->instance_data.size();
uint32_t total_threads = RendererThreadPool::singleton->thread_work_pool.get_thread_count();
uint32_t total_threads = WorkerThreadPool::get_singleton()->get_thread_count();
uint32_t cull_from = p_thread * cull_total / total_threads;
uint32_t cull_to = (p_thread + 1 == total_threads) ? cull_total : ((p_thread + 1) * cull_total / total_threads);
@@ -2919,7 +2919,8 @@ void RendererSceneCull::_render_scene(const RendererSceneRender::CameraData *p_c
}
if (visibility_cull_data.cull_count > thread_cull_threshold) {
RendererThreadPool::singleton->thread_work_pool.do_work(RendererThreadPool::singleton->thread_work_pool.get_thread_count(), this, &RendererSceneCull::_visibility_cull_threaded, &visibility_cull_data);
WorkerThreadPool::GroupID group_task = WorkerThreadPool::get_singleton()->add_template_group_task(this, &RendererSceneCull::_visibility_cull_threaded, &visibility_cull_data, WorkerThreadPool::get_singleton()->get_thread_count(), -1, true, SNAME("VisibilityCullInstances"));
WorkerThreadPool::get_singleton()->wait_for_group_task_completion(group_task);
} else {
_visibility_cull(visibility_cull_data, visibility_cull_data.cull_offset, visibility_cull_data.cull_offset + visibility_cull_data.cull_count);
}
@@ -3024,7 +3025,8 @@ void RendererSceneCull::_render_scene(const RendererSceneRender::CameraData *p_c
scene_cull_result_threads[i].clear();
}
RendererThreadPool::singleton->thread_work_pool.do_work(scene_cull_result_threads.size(), this, &RendererSceneCull::_scene_cull_threaded, &cull_data);
WorkerThreadPool::GroupID group_task = WorkerThreadPool::get_singleton()->add_template_group_task(this, &RendererSceneCull::_scene_cull_threaded, &cull_data, scene_cull_result_threads.size(), -1, true, SNAME("RenderCullInstances"));
WorkerThreadPool::get_singleton()->wait_for_group_task_completion(group_task);
for (uint32_t i = 0; i < scene_cull_result_threads.size(); i++) {
scene_cull_result.append_from(scene_cull_result_threads[i]);
@@ -4030,14 +4032,14 @@ RendererSceneCull::RendererSceneCull() {
}
scene_cull_result.init(&rid_cull_page_pool, &geometry_instance_cull_page_pool, &instance_cull_page_pool);
scene_cull_result_threads.resize(RendererThreadPool::singleton->thread_work_pool.get_thread_count());
scene_cull_result_threads.resize(WorkerThreadPool::get_singleton()->get_thread_count());
for (uint32_t i = 0; i < scene_cull_result_threads.size(); i++) {
scene_cull_result_threads[i].init(&rid_cull_page_pool, &geometry_instance_cull_page_pool, &instance_cull_page_pool);
}
indexer_update_iterations = GLOBAL_GET("rendering/limits/spatial_indexer/update_iterations_per_frame");
thread_cull_threshold = GLOBAL_GET("rendering/limits/spatial_indexer/threaded_cull_minimum_instances");
thread_cull_threshold = MAX(thread_cull_threshold, (uint32_t)RendererThreadPool::singleton->thread_work_pool.get_thread_count()); //make sure there is at least one thread per CPU
thread_cull_threshold = MAX(thread_cull_threshold, (uint32_t)WorkerThreadPool::get_singleton()->get_thread_count()); //make sure there is at least one thread per CPU
taa_jitter_array.resize(TAA_JITTER_COUNT);
for (int i = 0; i < TAA_JITTER_COUNT; i++) {