Fix: Light and shadow optimizations #742

Merged

SakulFlee merged 15 commits from fix_light_optimization into main

2026-08-01 02:08:06 +02:00

SakulFlee commented

2026-07-31 23:05:21 +02:00

Owner

This introduces staggered light and shadow updates.
Instead of recalculating everything every frame, we keep a list of "dirty" items and update one per frame. Say we render at 60 FPS, that would be 60x lighting updates per second. It is near real-time while having an enormous performance save: From ~3500 commands before we are down to ~500 now!

This introduces staggered light and shadow updates. Instead of recalculating everything every frame, we keep a list of "dirty" items and update **one per frame**. Say we render at 60 FPS, that would be 60x lighting updates per second. It is near real-time while having an enormous performance save: From ~3500 commands before we are down to ~500 now!

SakulFlee self-assigned this

2026-07-31 23:05:21 +02:00

SakulFlee added 11 commits

2026-07-31 23:05:21 +02:00

Add LightSlotIndex, ShadowDirtyFlag, PrevPosition components 9d63ec8c40

Add LightSlotTracker, ShadowMapCache, StaggeredLightConfig resources with slot allocation and shadow caching 408817faa7

Rewrite realize_lights() for incremental per-slot GPU buffer writes with position change detection for shadow dirty flagging e27038255d

Add stagger system for per-frame light/shadow update budgeting with round-robin scheduling ef00c78e5b

Modify ShadowRenderer::render() to accept dirty set and skip clean lights without re-rendering 9f7c98bf1b

Wire stagger system into module_runtime and renderer to pass dirty shadow set per frame 59e97ffd5c

Fix compilation errors: borrow checker, mut annotations, and pattern matching 0927380a0e

Fix shadow rendering and performance bugs: skip clean 2D depth passes, fill matrix_bytes for clean slots, restore realize_lights early return, use LightSlotIndex in collect_shadow_lights 9e1746d697

Fix FPS: cap shader light loop with light_count uniform (144 FPS) 5a03b9e997

The pre-allocated light buffer (MAX_LIGHTS=256) caused the PBR shader to
loop 256 times per fragment via arrayLength(&light_store), running the
full BRDF+shadow path 256 times instead of ~6 actual lights. This made
the model render pass ~3x slower (6.8ms→22ms in RenderDoc).

Changes:
- Add binding 13 (light_count uniform) to world bind group layout
- Compute active_light_count from LightDescriptorEcs dense.len() each frame
- Write to a static 16-byte UNIFORM|COPY_DST buffer (LIGHT_COUNT_BUF)
- Shader: loop bound changed from arrayLength(&light_store) to light_count
- Shader: add zero-intensity early continue guard (light.color.w == 0.0)
- Also moved global_model_dirty check before realize_models (Fix B1) to
  catch newly-imported models (e.g. DamagedHelmet) whose ModelDirty flag
  is cleared by realize_models before the check runs

Fix shadow texture reallocation wiping cube depth data mid-render 0e7afbe507

On frame 1 (bootstrap), ensure_cubes(2) was called mid-loop when the
second point light was processed, reallocating the cube depth texture
from 1 to 2 cubes. This dropped the first point light's freshly-rendered
cube-0 depth data into the old (dropped) texture, leaving cube 0 as
uninitialized garbage in the new texture. The shader's textureSampleCompare
for that cube returned shadowed (0) for all fragments, making the Room 4
point light contribute nothing — dark room, no shadows.

Changes:
- Pre-compute needed_cubes (count of enabled point lights) and
  needed_layers (spot/directional cascade sum) BEFORE the render loop
- Call ensure_cubes/ensure_layers up-front to final capacity
- Track grew = cubes_grew || layers_grew; when true, force_all_dirty
  makes all lights re-render this frame so no previously-stored depth
  is silently lost after a texture growth realloc
- Remove redundant mid-loop ensure_cubes and post-loop ensure_layers

Fix round-robin starvation: reserve floor budget for proactive shadow refresh

ci/woodpecker/pr/linting Pipeline failed

Details

ci/woodpecker/pr/test unknown status

Details

99208c9a99

The dirty-first pass consumed the entire per-frame budget when lights
were constantly dirty (e.g. the temporary LightAnimator marking Room-1
spot dirty every frame), starving the proactive round-robin refresh.
This meant clean shadow maps were never re-rendered, preventing newly
imported models from ever appearing in shadow maps.

Changes:
- Reserve max(1, budget/2) for round-robin when clean casters exist
  (0 reserve when no clean casters — all budget goes to dirty)
- Dirty lights get up to budget - reserve slots (still priority)
- Round-robin uses the remainder, ensuring proactive refresh runs
  even when some lights are always dirty

SakulFlee added 1 commit

2026-07-31 23:05:32 +02:00

Merge branch 'main' into fix_light_optimization

ci/woodpecker/pr/linting Pipeline failed

Details

ci/woodpecker/pr/test unknown status

Details

24c6eeee95

Ghost added 1 commit

2026-07-31 23:09:33 +02:00