perf(gen): the landing work was per-voxel -- 342s to 2.6s, same world

The owner played 64f5d61 and reported the truth: "generation still takes ages... TunnelNetwork doesn't even finish generating, as big as the surface world" while SurfaceWorld completed in under 10s, and stopping PIE hung for 30+ seconds. e08b49a's 21x was real and irrelevant. It fixed per-CHUNK state rebuilds, which dominate the commandlet export path. The game is bound by per-VOXEL evaluation. Same symptom, different bottleneck, and nobody had measured the second one. WHY PIE HUNG bShuttingDown was never checked inside VoxelMarchingCubesMesher.cpp -- zero occurrences. A worker inside GenerateMesh could not abort, so 64f5d61's correct unconditional EndPlay wait was waiting on tiles physically unable to stop. The hang was not a teardown bug; it was evidence that one tile took tens of seconds. Added shutdown polling per slab/row in the mesher and through GenerateTileResult; aborted results carry Epoch + bAborted and ProcessQueue drops them. EndPlay still waits unconditionally -- the 3s mark only logs now. PIE stop 3.409s / 44.759s -> 4 ms teardown, 38 us worker drain WHERE THE PER-VOXEL COST WAS The owner's own comparison was the sharpest instrument available. Profiled both: GetDensityAt TunnelNetwork 12.64 us/sample SurfaceWorld 1.32 tunnel candidates evaluated 7.088 per sample 0 tunnel-core candidates evaluated 7.201 per sample 0 room candidates evaluated 1.000 per sample 0 largest buckets: RoomGraphSource 2.57, TunnelCorePosts 0.76, PassageCarveOp 0.22, PassageStructuralPosts 0.20 The three regressing commits map cleanly, and one is innocent: 7a20eee drainage tree -- cache-build work, NOT per-voxel 413c15d landings -- repeated full passage modification per sample 55bf807 landing rooms -- room SDF + tunnel/core evaluation per sample Fixes: FPassageCarveOp carves only, landing/support moved to the structural MC post; tunnel core support and SDF fused into one pass; room graph caches shared per worker on complete deterministic keys; tunnel density cache uses open addressing with 4096 slots. Room cache rebuilds fell to 9 per 42,875 samples, so what remains is genuine O(local cached room/tunnel count) work. TunnelNetwork 128^3, seed 1337 slot 1, same region: 342.2 s / 163.161 us/vox -> 2.623 s / 1.251 us/vox (130x) Geometry SHA-256 aaf95559..066b4, IDENTICAL to the pre-optimisation run at the same roughness. 153,692 triangles either way. Verified independently of the agent's own report. NOT FIXED, and the first one is the next real target: - L4 tiles still cost 16.93 s. Hit rate falls 99.9% (L0) -> 23.7% (L4), but 1024 -> 4096 slots only moved 18.39s -> 16.93s and the 1,927 misses ARE the distinct key count. That is low reuse, not eviction. A bigger cache will not save it. - TunnelNetwork is still ~10x SurfaceWorld (0.354s vs 0.028s on a matched 32^3). - The live PIE tile table is SurfaceWorld: this checkout's PIE map is configured that way. TunnelNetwork went through the same mesher and density path via the commandlet, but no TunnelNetwork tile has been timed in an actual game session. Export numbers are not presented as game numbers. Still outstanding, untouched here: the inter-strate passage has intersections it should not have, routes through open passage when it should not, and should descend from a low point of the current strate. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012aheo2crrZmHuYtPpdY2Cb
2026-09-08 23:47:53 +02:00
1 parent 64f5d61e08
commit 0576e800c4
@@ -8,6 +8,7 @@
 #include "VoxelTypes.h"  // For CHUNK_SIZE, VOXEL_SIZE, WorldToChunkCoord
 #include "VoxelCaveMorphology.h"  // For VoxelSDF and VoxelHash
 #include "VoxelDensityPrimitives.h"  // Shared passage carve polarity/strength
+#include "VoxelDensityProfile.h"  // Opt-in targeted per-voxel attribution
 #include "VoxelTerrainOpDefinition.h"  // For UVoxelTerrainOpDefinition::ApplyTo
 #include "VoxelBiomeDefinition.h"  // For UVoxelBiomeDefinition (biome context flatten)
 #include "UObject/UObjectGlobals.h"
@@ -1687,6 +1688,12 @@ float UVoxelStrateManager::EvaluateModifierSDF(float WorldX, float WorldY, float
    // The shortlist is rebuilt once for a (manager, version, chunk) and is shared by the tube,
    // landing, and floor paths. No source-fit stencil or topology search is allowed below it.
    const TArray<int32>& Nearby = VF_GetNearbyPassages(this, ChunkCoord);
+    if (VoxelDensityProfile::IsEnabled())
+    {
+        VoxelDensityProfile::AddCounter(
+            VoxelDensityProfile::ECounter::PassageCandidates,
+            static_cast<uint64>(Nearby.Num()));
+    }

    if (Nearby.Num() == 0) return FLT_MAX;   // no passage near this chunk → no carve

@@ -1707,6 +1714,11 @@ float UVoxelStrateManager::EvaluateModifierSDF(float WorldX, float WorldY, float
        // capsule chain unconditionally — the dominant lag source once passages became
        // 12-segment worms. Now far passages cost a single squared-distance compare.
        if (FVector::DistSquared(Pos, P.BoundCenter) > P.BoundRadiusSq) continue;
+        if (VoxelDensityProfile::IsEnabled())
+        {
+            VoxelDensityProfile::AddCounter(
+                VoxelDensityProfile::ECounter::PassageEvaluated);
+        }

        if (P.ControlPoints.Num() >= 2)
        {
@@ -1748,6 +1760,8 @@ void UVoxelStrateManager::ApplyPassageModifier(
    float& Density, float WorldX, float WorldY, float WorldZ,
    float BaseDensity, float SealThickness) const
 {
+    VoxelDensityProfile::FScopedTimer ProfileTimer(
+        VoxelDensityProfile::EBucket::PassageModifier);
    const float ModSDF = EvaluateModifierSDF(WorldX, WorldY, WorldZ);
    VF_ApplyPassageCarving(Density, ModSDF, BaseDensity, SealThickness);
    ApplyPassageLandingAir(Density, WorldX, WorldY, WorldZ, BaseDensity, SealThickness);
@@ -1781,10 +1795,22 @@ void UVoxelStrateManager::ApplyPassageModifier(
    ApplyPassageTunnelAir(Density, WorldX, WorldY, WorldZ, BaseDensity, SealThickness);
 }

+void UVoxelStrateManager::ApplyPassageCarvingOnly(
+    float& Density, float WorldX, float WorldY, float WorldZ,
+    float BaseDensity, float SealThickness) const
+{
+    VoxelDensityProfile::FScopedTimer ProfileTimer(
+        VoxelDensityProfile::EBucket::PassageModifier);
+    const float ModSDF = EvaluateModifierSDF(WorldX, WorldY, WorldZ);
+    VF_ApplyPassageCarving(Density, ModSDF, BaseDensity, SealThickness);
+}
+
 void UVoxelStrateManager::ApplyPassageLandingAir(
    float& Density, float WorldX, float WorldY, float WorldZ,
    float BaseDensity, float SealThickness) const
 {
+    VoxelDensityProfile::FScopedTimer ProfileTimer(
+        VoxelDensityProfile::EBucket::PassageLandingAir);
    const FIntVector ChunkCoord(
        FMath::FloorToInt(WorldX / (float)CHUNK_SIZE),
        FMath::FloorToInt(WorldY / (float)CHUNK_SIZE),
@@ -1846,6 +1872,8 @@ void UVoxelStrateManager::ApplyPassageTunnelAir(
    float& Density, float WorldX, float WorldY, float WorldZ,
    float BaseDensity, float SealThickness) const
 {
+    VoxelDensityProfile::FScopedTimer ProfileTimer(
+        VoxelDensityProfile::EBucket::PassageTunnelAir);
    const FVector Position(WorldX, WorldY, WorldZ);
    int32 PassageIndex = INDEX_NONE;
    float FloorZ = 0.0f;
@@ -1874,6 +1902,8 @@ void UVoxelStrateManager::ApplyPassageTunnelAirMC(
 void UVoxelStrateManager::ApplyPassageLandingFloorMC(
    float& Density, float WorldX, float WorldY, float WorldZ, float BaseDensity) const
 {
+    VoxelDensityProfile::FScopedTimer ProfileTimer(
+        VoxelDensityProfile::EBucket::PassageLandingFloor);
    const FIntVector ChunkCoord(
        FMath::FloorToInt(WorldX / (float)CHUNK_SIZE),
        FMath::FloorToInt(WorldY / (float)CHUNK_SIZE),
@@ -1914,10 +1944,102 @@ static bool VF_IsPassageRoomFloor(
            <= FMath::Max(Landing.HalfWidth - 1.0f, 0.0f);
 }

+void UVoxelStrateManager::ApplyPassageStructuralPostsMC(
+    float& Density, float WorldX, float WorldY, float WorldZ,
+    float BaseDensity, float SealThickness) const
+{
+    VoxelDensityProfile::FScopedTimer ProfileTimer(
+        VoxelDensityProfile::EBucket::PassageStructuralPosts);
+
+    const FIntVector ChunkCoord(
+        FMath::FloorToInt(WorldX / (float)CHUNK_SIZE),
+        FMath::FloorToInt(WorldY / (float)CHUNK_SIZE),
+        FMath::FloorToInt(WorldZ / (float)CHUNK_SIZE));
+    const TArray<int32>& Nearby = VF_GetNearbyPassages(this, ChunkCoord);
+    if (Nearby.Num() == 0) return;
+
+    const FVector Position(WorldX, WorldY, WorldZ);
+    const bool bSuppressFloor =
+        VoxelPassageGeometry::VerticalShaftConnectorAirMarker();
+    bool bAnyPassageFloor = false;
+    bool bAnyRoomFloor = false;
+    bool bWalkableAir = false;
+    float MinLandingSDF = FLT_MAX;
+
+    // The four old MC backstops all walked the same immutable per-chunk shortlist. Gather their
+    // predicates in one pass; the writes below retain the old order (landing air, landing floor,
+    // tunnel air, room floor), including the vertical-shaft floor suppression marker.
+    for (const int32 PassageIndex : Nearby)
+    {
+        if (!Passages.IsValidIndex(PassageIndex)) continue;
+        const FVoxelPassage& Passage = Passages[PassageIndex];
+
+        if (!bSuppressFloor
+            && (VF_IsPassageLandingFloor(Position, Passage.UpperLanding)
+                || VF_IsPassageLandingFloor(Position, Passage.LowerLanding)
+                || VF_IsWalkableTunnelFloor(Passage, Position)))
+        {
+            bAnyPassageFloor = true;
+        }
+
+        if (!bSuppressFloor
+            && (VF_IsPassageRoomFloor(Position, Passage.UpperLanding)
+                || VF_IsPassageRoomFloor(Position, Passage.LowerLanding)))
+        {
+            bAnyRoomFloor = true;
+        }
+
+        if (!bWalkableAir
+            && FVector::DistSquared(Position, Passage.BoundCenter) <= Passage.BoundRadiusSq
+            && VF_IsWalkableTunnelAir(Passage, Position))
+        {
+            bWalkableAir = true;
+        }
+
+        if (FVector::DistSquared(Position, Passage.BoundCenter) <= Passage.BoundRadiusSq)
+        {
+            const FVoxelPassageLanding* Landings[] = {
+                &Passage.UpperLanding, &Passage.LowerLanding };
+            for (const FVoxelPassageLanding* Landing : Landings)
+            {
+                MinLandingSDF = FMath::Min(
+                    MinLandingSDF,
+                    VF_EvaluatePassageLandingSDF(Position, *Landing));
+            }
+        }
+    }
+
+    if (!bAnyPassageFloor && MinLandingSDF < FLT_MAX)
+    {
+        float InternalDensity = -Density;
+        VF_ApplyPassageLandingCarving(
+            InternalDensity, MinLandingSDF, BaseDensity, SealThickness);
+        Density = -InternalDensity;
+    }
+
+    if (bAnyPassageFloor)
+    {
+        Density = FMath::Min(Density, -BaseDensity);
+    }
+
+    if (bWalkableAir)
+    {
+        Density = FMath::Max(
+            Density, BaseDensity * 2.0f + SealThickness + 4.0f);
+    }
+
+    if (bAnyRoomFloor)
+    {
+        Density = FMath::Min(Density, -BaseDensity);
+    }
+}
+
 void UVoxelStrateManager::ApplyPassageLandingRoomFloorMC(
    float& Density, float WorldX, float WorldY, float WorldZ,
    float BaseDensity) const
 {
+    VoxelDensityProfile::FScopedTimer ProfileTimer(
+        VoxelDensityProfile::EBucket::PassageLandingRoomFloor);
    const FIntVector ChunkCoord(
        FMath::FloorToInt(WorldX / (float)CHUNK_SIZE),
        FMath::FloorToInt(WorldY / (float)CHUNK_SIZE),