Use multiple threads to import.

- For now everything imports multithreaded by default (should work I guess, let's test).
- Controllable per importer

Early test benchmark. 64 large textures (importing as lossless, _not_ as vram) on a mobile i7, 12 threads:
Importing goes down from 46 to 7 seconds.

For VRAM I will change the logic to use a compressing thread in a subsequent PR, as well as implementing Betsy.

1 files changed, 1 insertions, 1 deletions

diff --git a/scene/3d/gpu_particles_collision_3d.cpp b/scene/3d/gpu_particles_collision_3d.cpp
index 97241be60f..628b823f89 100644
--- a/scene/3d/gpu_particles_collision_3d.cpp
+++ b/scene/3d/gpu_particles_collision_3d.cpp
@@ -346,7 +346,7 @@ void GPUParticlesCollisionSDF::_compute_sdf(ComputeSDFParams *params) {
 	ThreadWorkPool work_pool;
 	work_pool.init();
 	work_pool.begin_work(params->size.z, this, &GPUParticlesCollisionSDF::_compute_sdf_z, params);
-	while (work_pool.get_work_index() < (uint32_t)params->size.z) {
+	while (!work_pool.is_done_dispatching()) {
 		OS::get_singleton()->delay_usec(10000);
 		bake_step_function(work_pool.get_work_index() * 100 / params->size.z, "Baking SDF");
 	}