feat: implement CPU meshlet baking and update pipeline shaders

Introduce SimplifyOptions and SimplifyVertexOptions enums for mesh simplification, replacing magic numbers with type-safe flags. Update MeshOptApi with strongly-typed wrapper methods. Refactor MeshletUtility to use new enums and nullable delegates, and fix stride calculation for pointer arithmetic.

Rename NamingConventions.GetMethodName to GetName, update name removal logic to use "$TBare", and add ALL_CAPS style for constants. Update config files to match new naming conventions and add ALL_CAPS constant rule for meshopt. Refactor BindingParser and related classes to support constant member kind. Apply minor bug fixes and code style improvements throughout.

README_julian.md

@@ -16,4 +16,5 @@ Misaki has set up my environment to work from my own fork of "GhostEngine." Here
 
 We'll refine separately; I'll merge when...
 ...the functionality is complete and tested.
-**PR Description Tip:** Include concise changelog markdown.
+**PR Description Tip:** write everything in the description, do not include concise changelog markdown.
+Also, I should not include AGENT.md and README_julian.md in the PR.

src/Runtime/Ghost.Core/Ghost.Core.csproj

@@ -9,20 +9,19 @@
 
   <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|AnyCPU'">
     <IsAotCompatible>True</IsAotCompatible>
-    <DefineConstants>$(DefineConstants);PLATEFORME_WIN64</DefineConstants>
+    <DefineConstants>$(DefineConstants);ENABLE_DEBUG_LAYER</DefineConstants>
     <IsTrimmable>True</IsTrimmable>
   </PropertyGroup>
 
   <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|AnyCPU'">
     <IsAotCompatible>True</IsAotCompatible>
-    <DefineConstants>$(DefineConstants);PLATEFORME_WIN64</DefineConstants>
     <IsTrimmable>True</IsTrimmable>
   </PropertyGroup>
 
   <ItemGroup>
     <PackageReference Include="Misaki.HighPerformance" Version="1.0.4" />
-    <PackageReference Include="Misaki.HighPerformance.Jobs" Version="1.5.1" />
+    <PackageReference Include="Misaki.HighPerformance.Jobs" Version="1.5.3" />
-    <PackageReference Include="Misaki.HighPerformance.LowLevel" Version="1.4.4">
+    <PackageReference Include="Misaki.HighPerformance.LowLevel" Version="1.5.2">
       <PrivateAssets>all</PrivateAssets>
       <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
     </PackageReference>

src/Runtime/Ghost.Graphics.RHI/RootSignatureLayout.cs

@@ -72,4 +72,7 @@ public struct PerObjectData
     public uint vertexBuffer;
     public float3 worldBoundsMax;
     public uint indexBuffer;
+    public uint meshletBuffer;
+    public uint meshletVerticesBuffer;
+    public uint meshletTrianglesBuffer;
 };

src/Runtime/Ghost.Graphics/Core/Mesh.cs

@@ -6,6 +6,8 @@ using Misaki.HighPerformance.LowLevel.Collections;
 using Misaki.HighPerformance.LowLevel.Utilities;
 using Misaki.HighPerformance.Mathematics;
 using Misaki.HighPerformance.Mathematics.Geometry;
+using System.Runtime.InteropServices;
+using System.Runtime.CompilerServices;
 
 namespace Ghost.Graphics.Core;
 
@@ -68,6 +70,8 @@ public struct Mesh : IResourceReleasable
     private UnsafeList<uint> _indices;
     private MeshletMeshData _meshletData;
 
+    public MeshletMeshData MeshletData => _meshletData;
+
     internal bool IsMeshDataDirty
     {
         get; private set;
@@ -149,6 +153,22 @@ public struct Mesh : IResourceReleasable
         get; internal set;
     }
 
+    /// <summary>
+    /// Gets the handle to the meshlet vertices buffer on the GPU.
+    /// </summary>
+    public Handle<GraphicsBuffer> MeshletVerticesBuffer
+    {
+        get; internal set;
+    }
+
+    /// <summary>
+    /// Gets the handle to the meshlet triangles buffer on the GPU.
+    /// </summary>
+    public Handle<GraphicsBuffer> MeshletTrianglesBuffer
+    {
+        get; internal set;
+    }
+
     /// <summary>
     /// Gets the handle to the mesh data buffer on the GPU.
     /// </summary>
@@ -176,6 +196,92 @@ public struct Mesh : IResourceReleasable
         _meshletData.Dispose();
     }
 
+    public unsafe void CookMeshlets()
+    {
+        // 1. Prepare Configuration
+        var config = new ClodConfig
+        {
+            maxVertices = 64,
+            minTriangles = 32,
+            maxTriangles = 124,
+            partitionSize = 128,
+            clusterSpatial = true,
+            clusterFillWeight = 1.0f,
+            clusterSplitFactor = 1.0f,
+            simplifyRatio = 0.5f,
+            simplifyThreshold = 0.5f,
+            simplifyErrorMergePrevious = 0.5f,
+            simplifyErrorMergeAdditive = 0.5f,
+            simplifyErrorFactorSloppy = 1.0f,
+            simplifyErrorEdgeLimit = 1.0f,
+            optimizeBounds = true,
+            optimizeClusters = true
+        };
+
+        // 2. Map Mesh to ClodMesh
+        ClodMesh clodMesh = new ClodMesh
+        {
+            vertexPositions = (float*)_vertices.GetUnsafePtr(),
+            vertexCount = (nuint)_vertices.Count,
+            vertexPositionsStride = (nuint)sizeof(Vertex),
+            indices = (uint*)_indices.GetUnsafePtr(),
+            indexCount = (nuint)_indices.Count,
+            attributeProtectMask = 0
+        };
+
+        // 3. Build
+        MeshletUtility.Build(config, clodMesh, Unsafe.AsPointer(ref this), MeshletOutputCallback);
+    }
+
+    private static unsafe int MeshletOutputCallback(void* context, ClodGroup group, ClodCluster* clusters, nuint clusterCount)
+    {
+        Mesh* mesh = (Mesh*)context;
+        ref var data = ref mesh->_meshletData;
+
+        // Ensure lists are initialized
+        if (!data.groups.IsCreated) data.groups = new UnsafeList<MeshletGroup>(16, Allocator.Persistent);
+        if (!data.meshlets.IsCreated) data.meshlets = new UnsafeList<Meshlet>(64, Allocator.Persistent);
+        if (!data.meshletVertices.IsCreated) data.meshletVertices = new UnsafeList<uint>(128, Allocator.Persistent);
+        if (!data.meshletTriangles.IsCreated) data.meshletTriangles = new UnsafeList<byte>(128, Allocator.Persistent);
+
+        var meshletGroup = new MeshletGroup
+        {
+            meshletStartIndex = (uint)data.meshlets.Count,
+            meshletCount = (uint)clusterCount,
+            lodLevel = (uint)group.depth
+        };
+        data.groups.Add(meshletGroup);
+
+        for (nuint i = 0; i < clusterCount; i++)
+        {
+            var cluster = clusters[i];
+            
+            var meshlet = new Meshlet
+            {
+                vertexCount = (byte)cluster.vertexCount,
+                triangleCount = (byte)(cluster.indexCount / 3),
+                vertexOffset = (uint)data.meshletVertices.Count,
+                triangleOffset = (uint)data.meshletTriangles.Count,
+                groupIndex = (uint)data.groups.Count - 1
+            };
+            data.meshlets.Add(meshlet);
+
+            // Add indices
+            for (nuint j = 0; j < cluster.indexCount; j++)
+            {
+                data.meshletVertices.Add(cluster.indices[j]);
+            }
+            // Add triangles (packed indices or byte offsets)
+            // Assuming 8-bit local indices for meshlets as per standard convention
+            for (nuint j = 0; j < cluster.indexCount; j++)
+            {
+                data.meshletTriangles.Add((byte)j);
+            }
+        }
+
+        return 0;
+    }
+
     public readonly void ReleaseResource(IResourceDatabase database)
     {
         ReleaseCpuResources();
@@ -183,6 +289,8 @@ public struct Mesh : IResourceReleasable
         database.ReleaseResource(VertexBuffer.AsResource());
         database.ReleaseResource(IndexBuffer.AsResource());
         database.ReleaseResource(MeshLetBuffer.AsResource());
+        database.ReleaseResource(MeshletVerticesBuffer.AsResource());
+        database.ReleaseResource(MeshletTrianglesBuffer.AsResource());
         database.ReleaseResource(ObjectDataBuffer.AsResource());
     }
 }

src/Runtime/Ghost.Graphics/Core/RenderingContext.cs

@@ -157,6 +157,68 @@ public readonly unsafe ref struct RenderingContext
         }
     }
 
+    public void UploadMeshlets(Handle<Mesh> mesh)
+    {
+        var r = _resourceManager.GetMeshReference(mesh);
+        if (r.IsFailure) return;
+        
+        ref var meshRef = ref r.Value;
+        var meshletData = meshRef.MeshletData;
+
+        if (!meshletData.meshlets.IsCreated || meshletData.meshlets.Count == 0) return;
+
+        var meshletDesc = new BufferDesc
+        {
+            Size = (uint)(meshletData.meshlets.Count * sizeof(Meshlet)),
+            Stride = (uint)sizeof(Meshlet),
+            Usage = BufferUsage.Raw | BufferUsage.ShaderResource,
+            MemoryType = ResourceMemoryType.Default,
+        };
+        var verticesDesc = new BufferDesc
+        {
+            Size = (uint)(meshletData.meshletVertices.Count * sizeof(uint)),
+            Stride = sizeof(uint),
+            Usage = BufferUsage.Raw | BufferUsage.ShaderResource,
+            MemoryType = ResourceMemoryType.Default,
+        };
+        // Ensure size is multiple of 4 for Raw buffer
+        var trianglesSize = (uint)meshletData.meshletTriangles.Count;
+        trianglesSize = (trianglesSize + 3u) & ~3u;
+        var trianglesDesc = new BufferDesc
+        {
+            Size = trianglesSize,
+            Stride = sizeof(byte),
+            Usage = BufferUsage.Raw | BufferUsage.ShaderResource,
+            MemoryType = ResourceMemoryType.Default,
+        };
+
+        meshRef.MeshLetBuffer = _engine.ResourceAllocator.CreateBuffer(in meshletDesc, "Meshlets");
+        meshRef.MeshletVerticesBuffer = _engine.ResourceAllocator.CreateBuffer(in verticesDesc, "MeshletVertices");
+        meshRef.MeshletTrianglesBuffer = _engine.ResourceAllocator.CreateBuffer(in trianglesDesc, "MeshletTriangles");
+
+        TransitionBarrier(meshRef.MeshLetBuffer.AsResource(), false, BarrierLayout.Undefined, BarrierAccess.CopyDest, BarrierSync.Copy);
+        TransitionBarrier(meshRef.MeshletVerticesBuffer.AsResource(), false, BarrierLayout.Undefined, BarrierAccess.CopyDest, BarrierSync.Copy);
+        TransitionBarrier(meshRef.MeshletTrianglesBuffer.AsResource(), false, BarrierLayout.Undefined, BarrierAccess.CopyDest, BarrierSync.Copy);
+
+        _directCmd.UploadBuffer(meshRef.MeshLetBuffer, meshletData.meshlets.AsSpan());
+        _directCmd.UploadBuffer(meshRef.MeshletVerticesBuffer, meshletData.meshletVertices.AsSpan());
+        // Padding for triangle data if needed
+        if (trianglesSize > meshletData.meshletTriangles.Count)
+        {
+            var paddedData = new byte[trianglesSize];
+            meshletData.meshletTriangles.AsSpan().CopyTo(paddedData);
+            _directCmd.UploadBuffer(meshRef.MeshletTrianglesBuffer, paddedData.AsSpan());
+        }
+        else
+        {
+            _directCmd.UploadBuffer(meshRef.MeshletTrianglesBuffer, meshletData.meshletTriangles.AsSpan());
+        }
+
+        TransitionBarrier(meshRef.MeshLetBuffer.AsResource(), false, BarrierLayout.Undefined, BarrierAccess.ShaderResource, BarrierSync.NonPixelShading | BarrierSync.PixelShading);
+        TransitionBarrier(meshRef.MeshletVerticesBuffer.AsResource(), false, BarrierLayout.Undefined, BarrierAccess.ShaderResource, BarrierSync.NonPixelShading | BarrierSync.PixelShading);
+        TransitionBarrier(meshRef.MeshletTrianglesBuffer.AsResource(), false, BarrierLayout.Undefined, BarrierAccess.ShaderResource, BarrierSync.NonPixelShading | BarrierSync.PixelShading);
+    }
+
     public void UpdateObjectData(Handle<Mesh> mesh, float4x4 localToWorld)
     {
         var r = _resourceManager.GetMeshReference(mesh);
@@ -173,6 +235,9 @@ public readonly unsafe ref struct RenderingContext
             worldBoundsMax = meshData.BoundingBox.Max,
             vertexBuffer = _engine.ResourceDatabase.GetBindlessIndex(meshData.VertexBuffer.AsResource()),
             indexBuffer = _engine.ResourceDatabase.GetBindlessIndex(meshData.IndexBuffer.AsResource()),
+            meshletBuffer = meshData.MeshLetBuffer.IsInvalid ? 0 : _engine.ResourceDatabase.GetBindlessIndex(meshData.MeshLetBuffer.AsResource()),
+            meshletVerticesBuffer = meshData.MeshletVerticesBuffer.IsInvalid ? 0 : _engine.ResourceDatabase.GetBindlessIndex(meshData.MeshletVerticesBuffer.AsResource()),
+            meshletTrianglesBuffer = meshData.MeshletTrianglesBuffer.IsInvalid ? 0 : _engine.ResourceDatabase.GetBindlessIndex(meshData.MeshletTrianglesBuffer.AsResource()),
         };
 
         var bufferHandle = meshData.ObjectDataBuffer.AsResource();

src/Runtime/Ghost.Graphics/Meshlet/ClodBounds.cs

@@ -1,10 +0,0 @@
-using System.Numerics;
-
-namespace Ghost.Graphics.Meshlet;
-
-public struct ClodBounds
-{
-    public Vector3 center;
-    public float radius;
-    public float error;
-}

src/Runtime/Ghost.Graphics/Meshlet/ClodBoundsHelper.cs

@@ -1,48 +0,0 @@
-using System;
-using System.Numerics;
-using Ghost.MeshOptimizer;
-using Misaki.HighPerformance.LowLevel.Buffer;
-using Misaki.HighPerformance.LowLevel.Collections;
-
-namespace Ghost.Graphics.Meshlet;
-
-internal static class ClodBoundsHelper
-{
-    public static unsafe ClodBounds ComputeBounds(ClodMesh mesh, UnsafeList<uint> indices, float error)
-    {
-        var bounds = MeshOptApi.ComputeClusterBounds((uint*)indices.GetUnsafePtr(), (nuint)indices.Count, mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride);
-        return new ClodBounds
-        {
-            center = new Vector3(bounds.center[0], bounds.center[1], bounds.center[2]),
-            radius = bounds.radius,
-            error = error
-        };
-    }
-
-    public static unsafe ClodBounds MergeBounds(UnsafeList<Cluster> clusters, UnsafeList<int> group)
-    {
-        var boundsList = new UnsafeList<ClodBounds>(group.Count, Allocator.Temp);
-        for (int j = 0; j < group.Count; j++)
-            boundsList.Add(clusters[group[j]].bounds);
-
-        var merged = MeshOptApi.ComputeSphereBounds(
-            (float*)boundsList.GetUnsafePtr(),
-            (nuint)group.Count,
-            (nuint)sizeof(ClodBounds),
-            (float*)boundsList.GetUnsafePtr() + 3,
-            (nuint)sizeof(ClodBounds)
-        );
-
-        float maxError = 0.0f;
-        for (int j = 0; j < group.Count; j++)
-            maxError = Math.Max(maxError, clusters[group[j]].bounds.error);
-
-        boundsList.Dispose();
-        return new ClodBounds
-        {
-            center = new Vector3(merged.center[0], merged.center[1], merged.center[2]),
-            radius = merged.radius,
-            error = maxError
-        };
-    }
-}

src/Runtime/Ghost.Graphics/Meshlet/ClodBuilder.cs

@@ -1,176 +0,0 @@
-using System;
-using System.Diagnostics;
-using Ghost.MeshOptimizer;
-using Misaki.HighPerformance.LowLevel.Buffer;
-using Misaki.HighPerformance.LowLevel.Collections;
-
-namespace Ghost.Graphics.Meshlet;
-
-internal struct Cluster
-{
-    public nuint vertices;
-    public UnsafeList<uint> indices;
-    public int group;
-    public int refined;
-    public ClodBounds bounds;
-}
-
-public unsafe static class ClodBuilder
-{
-    public static nuint Build(ClodConfig config, ClodMesh mesh, void* outputContext, ClodOutputDelegate outputCallback)
-    {
-        Debug.Assert(mesh.vertexAttributesStride % (nuint)sizeof(float) == 0, "vertexAttributesStride must be a multiple of sizeof(float)");
-
-        var locks = new UnsafeList<byte>((int)mesh.vertexCount, Allocator.Temp);
-        locks.AsSpan().Fill(0);
-
-        var remap = new UnsafeList<uint>((int)mesh.vertexCount, Allocator.Temp);
-        remap.Resize((int)mesh.vertexCount);
-        MeshOptApi.GeneratePositionRemap((uint*)remap.GetUnsafePtr(), mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride);
-
-        if (mesh.attributeProtectMask != 0)
-        {
-            nuint maxAttributes = mesh.vertexAttributesStride / sizeof(float);
-            for (nuint i = 0; i < mesh.vertexCount; i++)
-            {
-                uint r = ((uint*)remap.GetUnsafePtr())[(int)i];
-                for (nuint j = 0; j < maxAttributes; j++)
-                {
-                    if ((r != i) && ((mesh.attributeProtectMask & (1u << (int)j)) != 0))
-                    {
-                        if (mesh.vertexAttributes[i * maxAttributes + j] != mesh.vertexAttributes[r * maxAttributes + j])
-                        {
-                            ((byte*)locks.GetUnsafePtr())[(int)i] |= (byte)(Api.meshopt_SimplifyVertex_Protect & 0xFF);
-                        }
-                    }
-                }
-            }
-        }
-
-        var clusters = ClodInternal.Clusterize(config, mesh, mesh.indices, mesh.indexCount, Allocator.Persistent);
-
-        for (int i = 0; i < clusters.Count; i++)
-        {
-            clusters[i].bounds = ClodBoundsHelper.ComputeBounds(mesh, clusters[i].indices, 0.0f);
-        }
-
-        var pending = new UnsafeList<int>(clusters.Count, Allocator.Temp);
-        for (int i = 0; i < clusters.Count; i++)
-            pending.Add(i);
-
-        int depth = 0;
-
-        while (pending.Count > 1)
-        {
-            var groups = ClodPartition.Partition(config, mesh, clusters, pending, remap, Allocator.Temp);
-            pending.Clear();
-
-            ClodBoundary.LockBoundary(locks, groups, clusters, remap, mesh.vertexLock);
-
-            for (int i = 0; i < groups.Count; i++)
-            {
-                var merged = new UnsafeList<uint>(groups[i].Count * (int)config.maxTriangles * 3, Allocator.Temp);
-                for (int j = 0; j < groups[i].Count; j++)
-                {
-                    var clusterIndices = clusters[groups[i][j]].indices;
-                    for (int k = 0; k < clusterIndices.Count; k++)
-                        merged.Add(clusterIndices[k]);
-                }
-
-                nuint targetSize = ((nuint)merged.Count / 3) * (nuint)config.simplifyRatio * 3;
-                var bounds = ClodBoundsHelper.MergeBounds(clusters, groups[i]);
-
-                float error = 0.0f;
-                var simplified = ClodSimplify.Simplify(config, mesh, merged, locks, targetSize, &error);
-
-                if ((nuint)simplified.Count > (nuint)(merged.Count * config.simplifyThreshold))
-                {
-                    bounds.error = float.MaxValue;
-                    OutputGroup(config, mesh, clusters, groups[i], bounds, depth, outputContext, outputCallback);
-                    merged.Dispose();
-                    continue;
-                }
-
-                bounds.error = Math.Max(bounds.error * config.simplifyErrorMergePrevious, error) + error * config.simplifyErrorMergeAdditive;
-
-                int refined = OutputGroup(config, mesh, clusters, groups[i], bounds, depth, outputContext, outputCallback);
-
-                for (int j = 0; j < groups[i].Count; j++)
-                    clusters[groups[i][j]].indices.Dispose();
-
-                var split = ClodInternal.Clusterize(config, mesh, (uint*)simplified.GetUnsafePtr(), (nuint)simplified.Count, Allocator.Persistent);
-                for (int j = 0; j < split.Count; j++)
-                {
-                    split[j].refined = refined;
-                    split[j].bounds = bounds;
-                    clusters.Add(split[j]);
-                    pending.Add(clusters.Count - 1);
-                }
-
-                split.Dispose();
-                merged.Dispose();
-            }
-
-            for (int i = 0; i < groups.Count; i++)
-                groups[i].Dispose();
-            groups.Dispose();
-
-            depth++;
-        }
-
-        if (pending.Count > 0)
-        {
-            var bounds = clusters[pending[0]].bounds;
-            bounds.error = float.MaxValue;
-            OutputGroup(config, mesh, clusters, pending, bounds, depth, outputContext, outputCallback);
-        }
-
-        nuint finalClusterCount = (nuint)clusters.Count;
-
-        for (int i = 0; i < clusters.Count; i++)
-            clusters[i].indices.Dispose();
-        clusters.Dispose();
-        locks.Dispose();
-        remap.Dispose();
-        pending.Dispose();
-
-        return finalClusterCount;
-    }
-
-    private static int OutputGroup(
-        ClodConfig config,
-        ClodMesh mesh,
-        UnsafeList<Cluster> clusters,
-        UnsafeList<int> group,
-        ClodBounds simplified,
-        int depth,
-        void* outputContext,
-        ClodOutputDelegate outputCallback
-    )
-    {
-        var groupClusters = new UnsafeList<ClodCluster>(group.Count, Allocator.Temp);
-
-        for (int i = 0; i < group.Count; i++)
-        {
-            ref var srcCluster = ref clusters[group[i]];
-            groupClusters.Add(new ClodCluster
-            {
-                refined = srcCluster.refined,
-                bounds = (config.optimizeBounds && srcCluster.refined != -1)
-                    ? ClodBoundsHelper.ComputeBounds(mesh, srcCluster.indices, srcCluster.bounds.error)
-                    : srcCluster.bounds,
-                indices = (uint*)srcCluster.indices.GetUnsafePtr(),
-                indexCount = (nuint)srcCluster.indices.Count,
-                vertexCount = srcCluster.vertices
-            });
-        }
-
-        var clodGroup = new ClodGroup { depth = depth, simplified = simplified };
-        int result = outputCallback != null
-            ? outputCallback(outputContext, clodGroup, (ClodCluster*)groupClusters.GetUnsafePtr(), (nuint)groupClusters.Count)
-            : -1;
-
-        groupClusters.Dispose();
-        return result;
-    }
-}

src/Runtime/Ghost.Graphics/Meshlet/ClodConfig.cs

@@ -1,37 +0,0 @@
-namespace Ghost.Graphics.Meshlet;
-
-public struct ClodConfig
-{
-    public nuint maxVertices;
-    public nuint minTriangles;
-    public nuint maxTriangles;
-
-    public bool partitionSpatial;
-    public bool partitionSort;
-    public nuint partitionSize;
-
-    public bool clusterSpatial;
-    public float clusterFillWeight;
-    public float clusterSplitFactor;
-
-    public float simplifyRatio;
-    public float simplifyThreshold;
-
-    public float simplifyErrorMergePrevious;
-    public float simplifyErrorMergeAdditive;
-
-    public float simplifyErrorFactorSloppy;
-
-    public float simplifyErrorEdgeLimit;
-
-    public bool simplifyPermissive;
-
-    public bool simplifyFallbackPermissive;
-    public bool simplifyFallbackSloppy;
-
-    public bool simplifyRegularize;
-
-    public bool optimizeBounds;
-
-    public bool optimizeClusters;
-}

src/Runtime/Ghost.Graphics/Meshlet/ClodInternal.cs

@@ -1,83 +0,0 @@
-using System;
-using Ghost.MeshOptimizer;
-using Misaki.HighPerformance.LowLevel.Buffer;
-using Misaki.HighPerformance.LowLevel.Collections;
-
-namespace Ghost.Graphics.Meshlet;
-
-internal static class ClodInternal
-{
-    public static unsafe UnsafeList<Cluster> Clusterize(ClodConfig config, ClodMesh mesh, uint* indices, nuint indexCount, Allocator allocator)
-    {
-        nuint maxMeshlets = MeshOptApi.BuildMeshletsBound(indexCount, config.maxVertices, config.minTriangles);
-
-        var meshlets = new UnsafeList<meshopt_Meshlet>((int)maxMeshlets, Allocator.Temp);
-        meshlets.Resize((int)maxMeshlets);
-        var meshletVertices = new UnsafeList<uint>((int)indexCount, Allocator.Temp);
-        meshletVertices.Resize((int)indexCount);
-        var meshletTriangles = new UnsafeList<byte>((int)indexCount, Allocator.Temp);
-        meshletTriangles.Resize((int)indexCount);
-
-        meshopt_Meshlet* pMeshlets = (meshopt_Meshlet*)meshlets.GetUnsafePtr();
-        uint* pMeshletVertices = (uint*)meshletVertices.GetUnsafePtr();
-        byte* pMeshletTriangles = (byte*)meshletTriangles.GetUnsafePtr();
-
-        nuint meshletCount;
-        if (config.clusterSpatial)
-        {
-            meshletCount = pMeshlets[0].BuildsSpatial(
-                pMeshletVertices, pMeshletTriangles,
-                indices, indexCount,
-                mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride,
-                config.maxVertices, config.minTriangles, config.maxTriangles,
-                config.clusterFillWeight
-            );
-        }
-        else
-        {
-            meshletCount = pMeshlets[0].BuildsFlex(
-                pMeshletVertices, pMeshletTriangles,
-                indices, indexCount,
-                mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride,
-                config.maxVertices, config.minTriangles, config.maxTriangles,
-                0.0f, config.clusterSplitFactor
-            );
-        }
-
-        var clusters = new UnsafeList<Cluster>((int)meshletCount, allocator);
-
-        for (nuint i = 0; i < meshletCount; i++)
-        {
-            ref var meshlet = ref pMeshlets[i];
-
-            if (config.optimizeClusters)
-            {
-                MeshOptApi.OptimizeMeshlet(
-                    pMeshletVertices + meshlet.vertex_offset,
-                    pMeshletTriangles + meshlet.triangle_offset,
-                    meshlet.triangle_count,
-                    meshlet.vertex_count
-                );
-            }
-
-            var cluster = new Cluster
-            {
-                vertices = meshlet.vertex_count,
-                indices = new UnsafeList<uint>((int)(meshlet.triangle_count * 3), allocator),
-                group = -1,
-                refined = -1
-            };
-
-            for (nuint j = 0; j < meshlet.triangle_count * 3; j++)
-                cluster.indices.Add(pMeshletVertices[meshlet.vertex_offset + pMeshletTriangles[meshlet.triangle_offset + j]]);
-
-            clusters.Add(cluster);
-        }
-
-        meshlets.Dispose();
-        meshletVertices.Dispose();
-        meshletTriangles.Dispose();
-
-        return clusters;
-    }
-}

src/Runtime/Ghost.Graphics/Meshlet/ClodInternal_Boundary.cs

@@ -1,48 +0,0 @@
-using Ghost.MeshOptimizer;
-using Misaki.HighPerformance.LowLevel.Buffer;
-using Misaki.HighPerformance.LowLevel.Collections;
-
-namespace Ghost.Graphics.Meshlet;
-
-internal static class ClodBoundary
-{
-    public static unsafe void LockBoundary(UnsafeList<byte> locks, UnsafeList<UnsafeList<int>> groups, UnsafeList<Cluster> clusters, UnsafeList<uint> remap, byte* vertexLock)
-    {
-        byte* pLocks = (byte*)locks.GetUnsafePtr();
-        uint* pRemap = (uint*)remap.GetUnsafePtr();
-
-        for (int i = 0; i < locks.Count; i++)
-            pLocks[i] = unchecked((byte)(pLocks[i] & ~((1 << 0) | (1 << 7))));
-
-        for (int i = 0; i < groups.Count; i++)
-        {
-            for (int j = 0; j < groups[i].Count; j++)
-            {
-                var cluster = clusters[groups[i][j]];
-                for (int k = 0; k < cluster.indices.Count; k++)
-                {
-                    uint r = pRemap[(int)cluster.indices[k]];
-                    pLocks[r] |= (byte)(pLocks[r] >> 7);
-                }
-            }
-
-            for (int j = 0; j < groups[i].Count; j++)
-            {
-                var cluster = clusters[groups[i][j]];
-                for (int k = 0; k < cluster.indices.Count; k++)
-                {
-                    uint r = pRemap[(int)cluster.indices[k]];
-                    pLocks[r] |= (byte)(1 << 7);
-                }
-            }
-        }
-
-        for (int i = 0; i < locks.Count; i++)
-        {
-            uint r = pRemap[i];
-            pLocks[i] = (byte)((pLocks[r] & 1) | (pLocks[i] & (byte)(Api.meshopt_SimplifyVertex_Protect & 0xFF)));
-            if (vertexLock != null)
-                pLocks[i] |= vertexLock[i];
-        }
-    }
-}

src/Runtime/Ghost.Graphics/Meshlet/ClodInternal_Partition.cs

@@ -1,64 +0,0 @@
-using System;
-using Ghost.MeshOptimizer;
-using Misaki.HighPerformance.LowLevel.Buffer;
-using Misaki.HighPerformance.LowLevel.Collections;
-
-namespace Ghost.Graphics.Meshlet;
-
-internal static class ClodPartition
-{
-    public static unsafe UnsafeList<UnsafeList<int>> Partition(ClodConfig config, ClodMesh mesh, UnsafeList<Cluster> clusters, UnsafeList<int> pending, UnsafeList<uint> remap, Allocator allocator)
-    {
-        if (pending.Count <= (int)config.partitionSize)
-        {
-            var single = new UnsafeList<UnsafeList<int>>(1, allocator);
-            single.Add(pending);
-            return single;
-        }
-
-        nuint totalIndexCount = 0;
-        for (int i = 0; i < pending.Count; i++)
-            totalIndexCount += (nuint)clusters[pending[i]].indices.Count;
-
-        var clusterIndices = new UnsafeList<uint>((int)totalIndexCount, Allocator.Temp);
-        var clusterCounts = new UnsafeList<uint>(pending.Count, Allocator.Temp);
-
-        nuint offset = 0;
-        for (int i = 0; i < pending.Count; i++)
-        {
-            var cluster = clusters[pending[i]];
-            clusterCounts.Add((uint)cluster.indices.Count);
-            for (int j = 0; j < cluster.indices.Count; j++)
-                clusterIndices.Add(((uint*)remap.GetUnsafePtr())[(int)cluster.indices[j]]);
-            offset += (nuint)cluster.indices.Count;
-        }
-
-        var clusterPart = new UnsafeList<uint>(pending.Count, Allocator.Temp);
-        clusterPart.Resize(pending.Count);
-
-        nuint partitionCount = MeshOptApi.PartitionClusters(
-            (uint*)clusterPart.GetUnsafePtr(),
-            (uint*)clusterIndices.GetUnsafePtr(),
-            totalIndexCount,
-            (uint*)clusterCounts.GetUnsafePtr(),
-            (nuint)pending.Count,
-            config.partitionSpatial ? mesh.vertexPositions : null,
-            (nuint)remap.Count,
-            mesh.vertexPositionsStride,
-            config.partitionSize
-        );
-
-        var partitions = new UnsafeList<UnsafeList<int>>((int)partitionCount, allocator);
-        for (nuint i = 0; i < partitionCount; i++)
-            partitions.Add(new UnsafeList<int>((int)(config.partitionSize + config.partitionSize / 3), allocator));
-
-        for (int i = 0; i < pending.Count; i++)
-            partitions[(int)((uint*)clusterPart.GetUnsafePtr())[i]].Add(pending[i]);
-
-        clusterIndices.Dispose();
-        clusterCounts.Dispose();
-        clusterPart.Dispose();
-
-        return partitions;
-    }
-}

src/Runtime/Ghost.Graphics/Meshlet/ClodMesh.cs

@@ -1,43 +0,0 @@
-using System;
-using Ghost.MeshOptimizer;
-using Misaki.HighPerformance.LowLevel.Collections;
-using Misaki.HighPerformance.LowLevel;
-using Misaki.HighPerformance.LowLevel.Buffer;
-
-namespace Ghost.Graphics.Meshlet;
-
-public unsafe struct ClodMesh
-{
-    public float* vertexPositions;
-    public nuint vertexCount;
-    public nuint vertexPositionsStride;
-
-    public float* vertexAttributes;
-    public nuint vertexAttributesStride;
-    public float* attributeWeights;
-    public nuint attributeCount;
-
-    public uint* indices;
-    public nuint indexCount;
-
-    public byte* vertexLock;
-    public uint attributeProtectMask;
-}
-
-public struct ClodGroup
-{
-    public int depth;
-    public ClodBounds simplified;
-}
-
-public unsafe struct ClodCluster
-{
-    public int refined;
-    public ClodBounds bounds;
-
-    public uint* indices;
-    public nuint indexCount;
-    public nuint vertexCount;
-}
-
-public unsafe delegate int ClodOutputDelegate(void* context, ClodGroup group, ClodCluster* clusters, nuint clusterCount);

src/Runtime/Ghost.Graphics/Meshlet/ClodSimplify.cs

@@ -1,109 +0,0 @@
-using System;
-using Ghost.MeshOptimizer;
-using Misaki.HighPerformance.LowLevel.Buffer;
-using Misaki.HighPerformance.LowLevel.Collections;
-
-namespace Ghost.Graphics.Meshlet;
-
-internal static class ClodSimplify
-{
-    public static unsafe UnsafeList<uint> Simplify(
-        ClodConfig config,
-        ClodMesh mesh,
-        UnsafeList<uint> indices,
-        UnsafeList<byte> locks,
-        nuint targetCount,
-        float* error
-    )
-    {
-        if (targetCount >= (nuint)indices.Count)
-            return indices;
-
-        var lod = new UnsafeList<uint>(indices.Count, Allocator.Temp);
-        lod.Resize(indices.Count);
-
-        uint options = (uint)(Api.meshopt_SimplifySparse | Api.meshopt_SimplifyErrorAbsolute);
-        if (config.simplifyPermissive)
-            options |= (uint)Api.meshopt_SimplifyPermissive;
-        if (config.simplifyRegularize)
-            options |= (uint)Api.meshopt_SimplifyRegularize;
-
-        nuint resultSize = MeshOptApi.SimplifyWithAttributes(
-            (uint*)lod.GetUnsafePtr(),
-            (uint*)indices.GetUnsafePtr(),
-            (nuint)indices.Count,
-            mesh.vertexPositions,
-            mesh.vertexCount,
-            mesh.vertexPositionsStride,
-            mesh.vertexAttributes,
-            mesh.vertexAttributesStride,
-            mesh.attributeWeights,
-            mesh.attributeCount,
-            (byte*)locks.GetUnsafePtr(),
-            targetCount,
-            float.MaxValue,
-            options,
-            error
-        );
-        lod.Resize((int)resultSize);
-
-        if ((nuint)lod.Count > targetCount && config.simplifyFallbackPermissive && !config.simplifyPermissive)
-        {
-            options |= (uint)Api.meshopt_SimplifyPermissive;
-            resultSize = MeshOptApi.SimplifyWithAttributes(
-                (uint*)lod.GetUnsafePtr(),
-                (uint*)indices.GetUnsafePtr(),
-                (nuint)indices.Count,
-                mesh.vertexPositions,
-                mesh.vertexCount,
-                mesh.vertexPositionsStride,
-                mesh.vertexAttributes,
-                mesh.vertexAttributesStride,
-                mesh.attributeWeights,
-                mesh.attributeCount,
-                (byte*)locks.GetUnsafePtr(),
-                targetCount,
-                float.MaxValue,
-                options,
-                error
-            );
-            lod.Resize((int)resultSize);
-        }
-
-        if ((nuint)lod.Count > targetCount && config.simplifyFallbackSloppy)
-        {
-            *error *= config.simplifyErrorFactorSloppy;
-        }
-
-        if (config.simplifyErrorEdgeLimit > 0)
-        {
-            float maxEdgeSq = 0;
-            uint* pIdx = (uint*)indices.GetUnsafePtr();
-            int posStride = (int)(mesh.vertexPositionsStride / sizeof(float));
-
-            for (int i = 0; i < indices.Count; i += 3)
-            {
-                uint a = pIdx[i], b = pIdx[i + 1], c = pIdx[i + 2];
-                float* va = mesh.vertexPositions + (a * (uint)posStride);
-                float* vb = mesh.vertexPositions + (b * (uint)posStride);
-                float* vc = mesh.vertexPositions + (c * (uint)posStride);
-
-                float dx, dy, dz;
-                dx = va[0] - vb[0]; dy = va[1] - vb[1]; dz = va[2] - vb[2];
-                float eab = dx * dx + dy * dy + dz * dz;
-                dx = va[0] - vc[0]; dy = va[1] - vc[1]; dz = va[2] - vc[2];
-                float eac = dx * dx + dy * dy + dz * dz;
-                dx = vb[0] - vc[0]; dy = vb[1] - vc[1]; dz = vb[2] - vc[2];
-                float ebc = dx * dx + dy * dy + dz * dz;
-
-                float emax = Math.Max(Math.Max(eab, eac), ebc);
-                float emin = Math.Min(Math.Min(eab, eac), ebc);
-                maxEdgeSq = Math.Max(maxEdgeSq, Math.Max(emin, emax / 4));
-            }
-
-            *error = Math.Min(*error, (float)Math.Sqrt(maxEdgeSq) * config.simplifyErrorEdgeLimit);
-        }
-
-        return lod;
-    }
-}

src/Runtime/Ghost.Graphics/RenderGraphModule/RenderGraphResourcePool.cs

@@ -129,8 +129,10 @@ internal sealed class RenderGraphResourceRegistry
             for (var i = 0; i < _resources.Count; i++)
             {
                 if (_resources[i].type == RenderGraphResourceType.Texture)
+                {
                     count++;
                 }
+            }
             return count;
         }
     }
@@ -142,8 +144,10 @@ internal sealed class RenderGraphResourceRegistry
             for (var i = 0; i < _resources.Count; i++)
             {
                 if (_resources[i].type == RenderGraphResourceType.Buffer)
+                {
                     count++;
                 }
+            }
             return count;
         }
     }
@@ -290,7 +294,9 @@ internal sealed class RenderGraphResourceRegistry
         {
             var res = _resources[i];
             if (res.type != RenderGraphResourceType.Texture || res.isImported)
+            {
                 continue;
+            }
 
             var desc = res.rgTextureDesc;
             if (desc.sizeMode == RGTextureSizeMode.Absolute)

src/Runtime/Ghost.Graphics/RenderPasses/MeshRenderPass.cs

@@ -194,6 +194,16 @@ internal class MeshRenderPass : IRenderPass
         MeshBuilder.CreateCube(0.75f, default, Misaki.HighPerformance.LowLevel.Buffer.Allocator.Persistent, out var vertices, out var indices);
 
         _mesh = ctx.CreateMesh(vertices, indices, true);
+        
+        // Cook meshlets for the mesh
+        var meshRef = ctx.ResourceManager.GetMeshReference(_mesh);
+        if (meshRef.IsSuccess)
+        {
+            meshRef.Value.CookMeshlets();
+        }
+
+        ctx.UploadMeshlets(_mesh);
+
         ctx.UpdateObjectData(_mesh, float4x4.identity);
 
         _textures = new Handle<Texture>[_textureFiles.Length];

src/Runtime/Ghost.Graphics/RenderPasses/ShaderCode.hlsl

@@ -8,30 +8,80 @@ struct PixelInput
     float4 uv : TEXCOORD0;
 };
 
-[numthreads(3, 1, 1)] // 3 threads per triangle
+struct Meshlet
+{
+    float4 boundingSphere;
+    float3 boundingBoxMin;
+    float3 boundingBoxMax;
+    uint vertexOffset;
+    uint triangleOffset;
+    uint groupIndex;
+    float parentError;
+    uint packedCounts; // byte vertexCount, byte triangleCount, byte localMaterialIndex, byte lodLevel
+};
+
+[numthreads(64, 1, 1)] // 64 threads for max 64 vertices and up to 124 triangles
 [OUTPUT_TRIANGLE_TOPOLOGY]
 void MSMain(
     uint3 groupThreadID : SV_GroupThreadID,
     uint groupID : SV_GroupID,
-    out vertices PixelInput outVerts[3],
+    out vertices PixelInput outVerts[64],
-    out indices uint3 outTris[1])
+    out indices uint3 outTris[124])
 {
-    uint vertexId = groupThreadID.x;
-
     PerObjectData perObjectData = LoadData<PerObjectData>(g_PushConstantData.perObjectBuffer, 0);
-    Vertex v = LoadVertexData(vertexId, groupID.x, perObjectData.vertexBuffer, perObjectData.indexBuffer);
 
-    SetMeshOutputCounts(3, 1);
+    ByteAddressBuffer meshletBuffer = GET_BUFFER(perObjectData.meshletBuffer);
+    Meshlet m = meshletBuffer.Load<Meshlet>(groupID.x * sizeof(Meshlet));
+
+    uint vertexCount = m.packedCounts & 0xFF;
+    uint triangleCount = (m.packedCounts >> 8) & 0xFF;
+
+    SetMeshOutputCounts(vertexCount, triangleCount);
+
+    ByteAddressBuffer meshletVerticesBuffer = GET_BUFFER(perObjectData.meshletVerticesBuffer);
+    ByteAddressBuffer meshletTrianglesBuffer = GET_BUFFER(perObjectData.meshletTrianglesBuffer);
 
     // Write vertex output
-    outVerts[vertexId].position = v.position;
+    if (groupThreadID.x < vertexCount)
-    outVerts[vertexId].color = v.color;
-    outVerts[vertexId].uv = v.uv;
-
-    // Thread 0 defines topology
-    if (vertexId == 0)
     {
-        outTris[0] = uint3(0, 1, 2);
+        uint vertexIndex = meshletVerticesBuffer.Load((m.vertexOffset + groupThreadID.x) * 4);
+        ByteAddressBuffer vertices = GET_BUFFER(perObjectData.vertexBuffer);
+        Vertex v = vertices.Load<Vertex>(vertexIndex * sizeof(Vertex));
+
+        // Basic MVP transform not needed if already in world space, but usually we need localToWorld and ViewProj
+        PerViewData perViewData = LoadData<PerViewData>(g_PushConstantData.perViewBuffer, 0);
+        float4 worldPos = mul(perObjectData.localToWorld, float4(v.position.xyz, 1.0f));
+        outVerts[groupThreadID.x].position = mul(perViewData.viewMatrix, worldPos);
+        outVerts[groupThreadID.x].position = mul(perViewData.projectionMatrix, outVerts[groupThreadID.x].position);
+
+        outVerts[groupThreadID.x].color = v.color;
+        outVerts[groupThreadID.x].uv = v.uv;
+    }
+
+    // Write triangle output (1 thread processes 1 triangle)
+    // We could pack 3 indices in a uint or just use byte offset
+    // In our CPU code, we packed it as individual bytes, so 3 bytes per triangle.
+    // For 124 triangles, we have 372 bytes.
+    if (groupThreadID.x < triangleCount)
+    {
+        uint triangleIndex = groupThreadID.x;
+        uint baseOffset = m.triangleOffset + triangleIndex * 3;
+
+        // Load 4 bytes to get the 3 index bytes
+        // Needs byte-aligned loading
+        uint wordOffset = baseOffset & ~3;
+        uint shift = (baseOffset & 3) * 8;
+        uint packedIndices1 = meshletTrianglesBuffer.Load(wordOffset);
+        uint packedIndices2 = meshletTrianglesBuffer.Load(wordOffset + 4);
+
+        uint64_t combined = ((uint64_t)packedIndices2 << 32) | packedIndices1;
+        uint packedIndices = (uint)(combined >> shift);
+
+        uint i0 = packedIndices & 0xFF;
+        uint i1 = (packedIndices >> 8) & 0xFF;
+        uint i2 = (packedIndices >> 16) & 0xFF;
+
+        outTris[triangleIndex] = uint3(i0, i1, i2);
     }
 }
 

src/Runtime/Ghost.Graphics/Shaders/Includes/Properties.hlsl

@@ -29,6 +29,9 @@ struct PerObjectData
     BYTE_ADDRESS_BUFFER vertexBuffer;
     float3 worldBoundsMax;
     BYTE_ADDRESS_BUFFER indexBuffer;
+    BYTE_ADDRESS_BUFFER meshletBuffer;
+    BYTE_ADDRESS_BUFFER meshletVerticesBuffer;
+    BYTE_ADDRESS_BUFFER meshletTrianglesBuffer;
 };
 
 PushConstantData g_PushConstantData : register(b0);

src/Runtime/Ghost.Graphics/Utilities/MeshletUtility.cs

@@ -0,0 +1,620 @@
+using Ghost.MeshOptimizer;
+using Misaki.HighPerformance.LowLevel.Buffer;
+using Misaki.HighPerformance.LowLevel.Collections;
+using Misaki.HighPerformance.LowLevel.Utilities;
+using Misaki.HighPerformance.Mathematics;
+using System.Diagnostics;
+
+namespace Ghost.Graphics.Utilities;
+
+internal struct Cluster : IDisposable
+{
+    public UnsafeList<uint> indices;
+    public ClodBounds bounds;
+    public nuint vertices;
+    public int group;
+    public int refined;
+
+    public void Dispose()
+    {
+        indices.Dispose();
+    }
+}
+
+/// <summary>
+/// Represents the bounding sphere and simplification error for a LOD cluster.
+/// </summary>
+public struct ClodBounds
+{
+    /// <summary> The center of the bounding sphere. </summary>
+    public float3 center;
+    /// <summary> The radius of the bounding sphere. </summary>
+    public float radius;
+    /// <summary> The simplification error associated with this LOD level. </summary>
+    public float error;
+}
+
+/// <summary>
+/// Configuration parameters for the cluster LOD generation pipeline.
+/// </summary>
+public struct ClodConfig
+{
+    /// <summary> The maximum number of vertices per meshlet. </summary>
+    public nuint maxVertices;
+    /// <summary> The minimum number of triangles per meshlet. </summary>
+    public nuint minTriangles;
+    /// <summary> The maximum number of triangles per meshlet. </summary>
+    public nuint maxTriangles;
+    /// <summary> Whether to use spatial partitioning during meshlet building. </summary>
+    public bool partitionSpatial;
+    /// <summary> Whether to sort clusters after partitioning. </summary>
+    public bool partitionSort;
+    /// <summary> The target size for partitions. </summary>
+    public nuint partitionSize;
+    /// <summary> Whether to cluster meshlets using spatial clustering. </summary>
+    public bool clusterSpatial;
+    /// <summary> Weight factor for cluster fill calculation. </summary>
+    public float clusterFillWeight;
+    /// <summary> Split factor for flexible clustering. </summary>
+    public float clusterSplitFactor;
+    /// <summary> The simplification ratio to achieve per LOD level. </summary>
+    public float simplifyRatio;
+    /// <summary> Threshold for stopping simplification. </summary>
+    public float simplifyThreshold;
+    /// <summary> Error factor used when merging previous LOD level errors. </summary>
+    public float simplifyErrorMergePrevious;
+    /// <summary> Additive error factor when merging LOD levels. </summary>
+    public float simplifyErrorMergeAdditive;
+    /// <summary> Error factor for sloppy simplification. </summary>
+    public float simplifyErrorFactorSloppy;
+    /// <summary> Edge length limit error factor. </summary>
+    public float simplifyErrorEdgeLimit;
+    /// <summary> Whether to allow permissive simplification. </summary>
+    public bool simplifyPermissive;
+    /// <summary> Whether to fallback to permissive simplification. </summary>
+    public bool simplifyFallbackPermissive;
+    /// <summary> Whether to fallback to sloppy simplification. </summary>
+    public bool simplifyFallbackSloppy;
+    /// <summary> Whether to regularize the mesh during simplification. </summary>
+    public bool simplifyRegularize;
+    /// <summary> Whether to optimize cluster bounds. </summary>
+    public bool optimizeBounds;
+    /// <summary> Whether to optimize clusters post-build. </summary>
+    public bool optimizeClusters;
+}
+
+/// <summary>
+/// Contains input data for the Cluster LOD generation pipeline.
+/// </summary>
+public unsafe struct ClodMesh
+{
+    /// <summary> Pointer to vertex position data (float array). </summary>
+    public float* vertexPositions;
+    /// <summary> Number of vertices in the mesh. </summary>
+    public nuint vertexCount;
+    /// <summary> Stride in bytes for vertex position data. </summary>
+    public nuint vertexPositionsStride;
+    /// <summary> Pointer to vertex attribute data (float array). </summary>
+    public float* vertexAttributes;
+    /// <summary> Stride in bytes for vertex attribute data. </summary>
+    public nuint vertexAttributesStride;
+    /// <summary> Pointer to attribute weights for simplification. </summary>
+    public float* attributeWeights;
+    /// <summary> Number of vertex attributes. </summary>
+    public nuint attributeCount;
+    /// <summary> Pointer to index data. </summary>
+    public uint* indices;
+    /// <summary> Number of indices in the mesh. </summary>
+    public nuint indexCount;
+    /// <summary> Pointer to per-vertex lock flags (1 byte per vertex). </summary>
+    public byte* vertexLock;
+    /// <summary> Mask indicating which attributes are protected during simplification. </summary>
+    public uint attributeProtectMask;
+}
+
+/// <summary>
+/// Defines a group of clusters in the LOD hierarchy.
+/// </summary>
+public struct ClodGroup
+{
+    /// <summary> LOD hierarchy depth of this group. </summary>
+    public int depth;
+    /// <summary> Bounding information for the simplified group. </summary>
+    public ClodBounds simplified;
+}
+
+/// <summary>
+/// Represents a cluster of meshlets in the LOD hierarchy.
+/// </summary>
+public unsafe struct ClodCluster
+{
+    /// <summary> Refinement level of the cluster. </summary>
+    public int refined;
+    /// <summary> Bounding info for the cluster. </summary>
+    public ClodBounds bounds;
+    /// <summary> Pointer to indices for this cluster. </summary>
+    public uint* indices;
+    /// <summary> Number of indices. </summary>
+    public nuint indexCount;
+    /// <summary> Number of vertices in the cluster. </summary>
+    public nuint vertexCount;
+}
+
+/// <summary>
+/// Delegate type for processing generated LOD groups.
+/// </summary>
+public unsafe delegate int ClodOutputDelegate(void* context, ClodGroup group, ClodCluster* clusters, nuint clusterCount);
+
+// FIX: UnsafeList and UnsafeArray are not same as std::vector.
+
+public static unsafe class MeshletUtility
+{
+    private static ClodBounds ComputeBounds(ClodMesh mesh, UnsafeList<uint> indices, float error)
+    {
+        var bounds = MeshOptApi.ComputeClusterBounds((uint*)indices.GetUnsafePtr(), (nuint)indices.Count, mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride);
+        return new ClodBounds
+        {
+            center = new float3(bounds.center[0], bounds.center[1], bounds.center[2]),
+            radius = bounds.radius,
+            error = error
+        };
+    }
+
+    private static ClodBounds MergeBounds(UnsafeList<Cluster> clusters, UnsafeList<int> group)
+    {
+        using var boundsList = new UnsafeArray<ClodBounds>(group.Count, Allocator.FreeList);
+        for (var j = 0; j < group.Count; j++)
+        {
+            boundsList[j] = (clusters[group[j]].bounds);
+        }
+
+        var merged = MeshOptApi.ComputeSphereBounds(
+            (float*)boundsList.GetUnsafePtr(),
+            (nuint)group.Count,
+            (nuint)sizeof(ClodBounds),
+            (float*)boundsList.GetUnsafePtr() + 3,
+            (nuint)sizeof(ClodBounds)
+        );
+
+        var maxError = 0.0f;
+        for (var j = 0; j < group.Count; j++)
+        {
+            maxError = Math.Max(maxError, clusters[group[j]].bounds.error);
+        }
+
+        return new ClodBounds
+        {
+            center = new float3(merged.center[0], merged.center[1], merged.center[2]),
+            radius = merged.radius,
+            error = maxError
+        };
+    }
+
+    private static UnsafeList<Cluster> Clusterize(ClodConfig config, ClodMesh mesh, uint* indices, nuint indexCount, Allocator allocator)
+    {
+        var maxMeshlets = MeshOptApi.BuildMeshletsBound(indexCount, config.maxVertices, config.minTriangles);
+
+        using var meshlets = new UnsafeArray<meshopt_Meshlet>((int)maxMeshlets, Allocator.FreeList);
+        using var meshletVertices = new UnsafeArray<uint>((int)indexCount, Allocator.FreeList);
+        using var meshletTriangles = new UnsafeArray<byte>((int)indexCount, Allocator.FreeList);
+
+        var pMeshlets = (meshopt_Meshlet*)meshlets.GetUnsafePtr();
+        var pMeshletVertices = (uint*)meshletVertices.GetUnsafePtr();
+        var pMeshletTriangles = (byte*)meshletTriangles.GetUnsafePtr();
+
+        nuint meshletCount;
+        if (config.clusterSpatial)
+        {
+            meshletCount = pMeshlets[0].BuildsSpatial(
+                pMeshletVertices, pMeshletTriangles,
+                indices, indexCount,
+                mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride,
+                config.maxVertices, config.minTriangles, config.maxTriangles,
+                config.clusterFillWeight
+            );
+        }
+        else
+        {
+            meshletCount = pMeshlets[0].BuildsFlex(
+                pMeshletVertices, pMeshletTriangles,
+                indices, indexCount,
+                mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride,
+                config.maxVertices, config.minTriangles, config.maxTriangles,
+                0.0f, config.clusterSplitFactor
+            );
+        }
+
+        var clusters = new UnsafeList<Cluster>((int)meshletCount, allocator);
+
+        for (nuint i = 0; i < meshletCount; i++)
+        {
+            ref var meshlet = ref pMeshlets[i];
+
+            if (config.optimizeClusters)
+            {
+                MeshOptApi.OptimizeMeshlet(
+                    pMeshletVertices + meshlet.vertex_offset,
+                    pMeshletTriangles + meshlet.triangle_offset,
+                    meshlet.triangle_count,
+                    meshlet.vertex_count
+                );
+            }
+
+            var cluster = new Cluster
+            {
+                vertices = meshlet.vertex_count,
+                indices = new UnsafeList<uint>((int)(meshlet.triangle_count * 3), Allocator.Persistent),
+                group = -1,
+                refined = -1
+            };
+
+            for (nuint j = 0; j < meshlet.triangle_count * 3; j++)
+            {
+                cluster.indices.Add(pMeshletVertices[meshlet.vertex_offset + pMeshletTriangles[meshlet.triangle_offset + j]]);
+            }
+
+            clusters.Add(cluster);
+        }
+
+        return clusters;
+    }
+
+    internal static void LockBoundary(UnsafeArray<byte> locks, UnsafeList<UnsafeList<int>> groups, UnsafeList<Cluster> clusters, UnsafeArray<uint> remap, byte* vertexLock)
+    {
+        var pLocks = (byte*)locks.GetUnsafePtr();
+        var pRemap = (uint*)remap.GetUnsafePtr();
+
+        for (var i = 0; i < locks.Length; i++)
+        {
+            pLocks[i] = unchecked((byte)(pLocks[i] & ~((1 << 0) | (1 << 7))));
+        }
+
+        for (var i = 0; i < groups.Count; i++)
+        {
+            for (var j = 0; j < groups[i].Count; j++)
+            {
+                var cluster = clusters[groups[i][j]];
+                for (var k = 0; k < cluster.indices.Count; k++)
+                {
+                    var r = pRemap[(int)cluster.indices[k]];
+                    pLocks[r] |= (byte)(pLocks[r] >> 7);
+                }
+            }
+
+            for (var j = 0; j < groups[i].Count; j++)
+            {
+                var cluster = clusters[groups[i][j]];
+                for (var k = 0; k < cluster.indices.Count; k++)
+                {
+                    var r = pRemap[(int)cluster.indices[k]];
+                    pLocks[r] |= 1 << 7;
+                }
+            }
+        }
+
+        for (var i = 0; i < locks.Length; i++)
+        {
+            var r = pRemap[i];
+            pLocks[i] = (byte)((pLocks[r] & 1) | (pLocks[i] & (byte)SimplifyVertexOptions.Protect & 0xFF));
+            if (vertexLock != null)
+            {
+                pLocks[i] |= vertexLock[i];
+            }
+        }
+    }
+
+    private static UnsafeList<UnsafeList<int>> Partition(ClodConfig config, ClodMesh mesh, UnsafeList<Cluster> clusters, UnsafeList<int> pending, UnsafeArray<uint> remap, Allocator allocator)
+    {
+        if (pending.Count <= (int)config.partitionSize)
+        {
+            var single = new UnsafeList<UnsafeList<int>>(1, allocator);
+            single.Add(pending);
+            return single;
+        }
+
+        nuint totalIndexCount = 0;
+        for (var i = 0; i < pending.Count; i++)
+        {
+            totalIndexCount += (nuint)clusters[pending[i]].indices.Count;
+        }
+
+        using var clusterIndices = new UnsafeList<uint>((int)totalIndexCount, Allocator.FreeList);
+        using var clusterCounts = new UnsafeList<uint>(pending.Count, Allocator.FreeList);
+
+        nuint offset = 0;
+        for (var i = 0; i < pending.Count; i++)
+        {
+            var cluster = clusters[pending[i]];
+            clusterCounts.Add((uint)cluster.indices.Count);
+            for (var j = 0; j < cluster.indices.Count; j++)
+            {
+                clusterIndices.Add(((uint*)remap.GetUnsafePtr())[(int)cluster.indices[j]]);
+            }
+
+            offset += (nuint)cluster.indices.Count;
+        }
+
+        using var clusterPart = new UnsafeArray<uint>(pending.Count, Allocator.FreeList);
+
+        var partitionCount = MeshOptApi.PartitionClusters(
+            (uint*)clusterPart.GetUnsafePtr(),
+            (uint*)clusterIndices.GetUnsafePtr(),
+            totalIndexCount,
+            (uint*)clusterCounts.GetUnsafePtr(),
+            (nuint)pending.Count,
+            config.partitionSpatial ? mesh.vertexPositions : null,
+            (nuint)remap.Length,
+            mesh.vertexPositionsStride,
+            config.partitionSize
+        );
+
+        var partitions = new UnsafeList<UnsafeList<int>>((int)partitionCount, allocator);
+        for (nuint i = 0; i < partitionCount; i++)
+        {
+            partitions.Add(new UnsafeList<int>((int)(config.partitionSize + config.partitionSize / 3), allocator));
+        }
+
+        for (var i = 0; i < pending.Count; i++)
+        {
+            partitions[(int)((uint*)clusterPart.GetUnsafePtr())[i]].Add(pending[i]);
+        }
+
+        return partitions;
+    }
+
+    private static int OutputGroup(ClodConfig config, ClodMesh mesh, UnsafeList<Cluster> clusters, UnsafeList<int> group, ClodBounds simplified, int depth, void* outputContext, ClodOutputDelegate? outputCallback)
+    {
+        using var groupClusters = new UnsafeList<ClodCluster>(group.Count, Allocator.FreeList);
+
+        for (var i = 0; i < group.Count; i++)
+        {
+            ref var srcCluster = ref clusters[group[i]];
+            groupClusters.Add(new ClodCluster
+            {
+                refined = srcCluster.refined,
+                bounds = (config.optimizeBounds && srcCluster.refined != -1)
+                    ? ComputeBounds(mesh, srcCluster.indices, srcCluster.bounds.error)
+                    : srcCluster.bounds,
+                indices = (uint*)srcCluster.indices.GetUnsafePtr(),
+                indexCount = (nuint)srcCluster.indices.Count,
+                vertexCount = srcCluster.vertices
+            });
+        }
+
+        var clodGroup = new ClodGroup { depth = depth, simplified = simplified };
+        var result = outputCallback != null
+            ? outputCallback(outputContext, clodGroup, (ClodCluster*)groupClusters.GetUnsafePtr(), (nuint)groupClusters.Count)
+            : -1;
+
+        return result;
+    }
+
+    public static UnsafeArray<uint> Simplify(ClodConfig config, ClodMesh mesh, ReadOnlyUnsafeCollection<uint> indices, ReadOnlyUnsafeCollection<byte> locks, nuint targetCount, float* error, Allocator allocator)
+    {
+        var lod = new UnsafeArray<uint>(indices.Count, allocator);
+
+        if (targetCount >= (nuint)indices.Count)
+        {
+            lod.CopyFrom(indices.AsSpan());
+            return lod;
+        }
+
+        var options = SimplifyOptions.Sparse | SimplifyOptions.ErrorAbsolute;
+        if (config.simplifyPermissive)
+        {
+            options |= SimplifyOptions.Permissive;
+        }
+
+        if (config.simplifyRegularize)
+        {
+            options |= SimplifyOptions.Regularize;
+        }
+
+        var resultSize = MeshOptApi.SimplifyWithAttributes(
+            (uint*)lod.GetUnsafePtr(),
+            (uint*)indices.GetUnsafePtr(),
+            (nuint)indices.Count,
+            mesh.vertexPositions,
+            mesh.vertexCount,
+            mesh.vertexPositionsStride,
+            mesh.vertexAttributes,
+            mesh.vertexAttributesStride,
+            mesh.attributeWeights,
+            mesh.attributeCount,
+            (byte*)locks.GetUnsafePtr(),
+            targetCount,
+            float.MaxValue,
+            options,
+            error
+        );
+
+        lod.Resize((int)resultSize);
+
+        if ((nuint)lod.Length > targetCount && config.simplifyFallbackPermissive && !config.simplifyPermissive)
+        {
+            options |= SimplifyOptions.Permissive;
+            resultSize = MeshOptApi.SimplifyWithAttributes(
+                (uint*)lod.GetUnsafePtr(),
+                (uint*)indices.GetUnsafePtr(),
+                (nuint)indices.Count,
+                mesh.vertexPositions,
+                mesh.vertexCount,
+                mesh.vertexPositionsStride,
+                mesh.vertexAttributes,
+                mesh.vertexAttributesStride,
+                mesh.attributeWeights,
+                mesh.attributeCount,
+                (byte*)locks.GetUnsafePtr(),
+                targetCount,
+                float.MaxValue,
+                options,
+                error
+            );
+
+            lod.Resize((int)resultSize);
+        }
+
+        if ((nuint)lod.Length > targetCount && config.simplifyFallbackSloppy)
+        {
+            *error *= config.simplifyErrorFactorSloppy;
+        }
+
+        if (config.simplifyErrorEdgeLimit > 0)
+        {
+            float maxEdgeSq = 0;
+            var pIdx = (uint*)indices.GetUnsafePtr();
+            var posStride = mesh.vertexPositionsStride / (nuint)sizeof(float);
+
+            for (var i = 0; i < indices.Count; i += 3)
+            {
+                uint a = pIdx[i], b = pIdx[i + 1], c = pIdx[i + 2];
+                var va = mesh.vertexPositions + (a * posStride);
+                var vb = mesh.vertexPositions + (b * posStride);
+                var vc = mesh.vertexPositions + (c * posStride);
+
+                float dx, dy, dz;
+                dx = va[0] - vb[0]; dy = va[1] - vb[1]; dz = va[2] - vb[2];
+                var eab = dx * dx + dy * dy + dz * dz;
+                dx = va[0] - vc[0]; dy = va[1] - vc[1]; dz = va[2] - vc[2];
+                var eac = dx * dx + dy * dy + dz * dz;
+                dx = vb[0] - vc[0]; dy = vb[1] - vc[1]; dz = vb[2] - vc[2];
+                var ebc = dx * dx + dy * dy + dz * dz;
+
+                var emax = Math.Max(Math.Max(eab, eac), ebc);
+                var emin = Math.Min(Math.Min(eab, eac), ebc);
+                maxEdgeSq = Math.Max(maxEdgeSq, Math.Max(emin, emax / 4));
+            }
+
+            *error = Math.Min(*error, (float)Math.Sqrt(maxEdgeSq) * config.simplifyErrorEdgeLimit);
+        }
+
+        return lod;
+    }
+
+    /// <summary>
+    /// Builds a cluster LOD hierarchy from the input mesh.
+    /// </summary>
+    /// <param name="config">The configuration parameters for the LOD building process.</param>
+    /// <param name="mesh">The input mesh data.</param>
+    /// <param name="outputContext">Optional context pointer passed to the output callback.</param>
+    /// <param name="outputCallback">Delegate invoked for each generated LOD group.</param>
+    /// <returns>The total count of generated clusters.</returns>
+    public static nuint Build(ClodConfig config, ClodMesh mesh, void* outputContext, ClodOutputDelegate? outputCallback)
+    {
+        Debug.Assert(mesh.vertexAttributesStride % sizeof(float) == 0, "vertexAttributesStride must be a multiple of sizeof(float)");
+
+        using var locks = new UnsafeArray<byte>((int)mesh.vertexCount, Allocator.FreeList, AllocationOption.Clear);
+        using var remap = new UnsafeArray<uint>((int)mesh.vertexCount, Allocator.FreeList);
+
+        MeshOptApi.GeneratePositionRemap((uint*)remap.GetUnsafePtr(), mesh.vertexPositions, mesh.vertexCount, mesh.vertexPositionsStride);
+
+        if (mesh.attributeProtectMask != 0)
+        {
+            var maxAttributes = mesh.vertexAttributesStride / sizeof(float);
+            for (nuint i = 0; i < mesh.vertexCount; i++)
+            {
+                var r = ((uint*)remap.GetUnsafePtr())[(int)i];
+                for (nuint j = 0; j < maxAttributes; j++)
+                {
+                    if ((r != i) && ((mesh.attributeProtectMask & (1u << (int)j)) != 0))
+                    {
+                        if (mesh.vertexAttributes[i * maxAttributes + j] != mesh.vertexAttributes[r * maxAttributes + j])
+                        {
+                            ((byte*)locks.GetUnsafePtr())[i] |= (byte)SimplifyVertexOptions.Protect & 0xFF;
+                        }
+                    }
+                }
+            }
+        }
+
+        using var clusters = Clusterize(config, mesh, mesh.indices, mesh.indexCount, Allocator.FreeList);
+
+        for (var i = 0; i < clusters.Count; i++)
+        {
+            clusters[i].bounds = ComputeBounds(mesh, clusters[i].indices, 0.0f);
+        }
+
+        using var pending = new UnsafeList<int>(clusters.Count, Allocator.FreeList);
+        for (var i = 0; i < clusters.Count; i++)
+        {
+            pending.Add(i);
+        }
+
+        var depth = 0;
+
+        while (pending.Count > 1)
+        {
+            using var groups = Partition(config, mesh, clusters, pending, remap, Allocator.FreeList);
+            pending.Clear();
+
+            LockBoundary(locks, groups, clusters, remap, mesh.vertexLock);
+
+            for (var i = 0; i < groups.Count; i++)
+            {
+                using var merged = new UnsafeList<uint>(groups[i].Count * (int)config.maxTriangles * 3, Allocator.FreeList);
+                for (var j = 0; j < groups[i].Count; j++)
+                {
+                    var clusterIndices = clusters[groups[i][j]].indices;
+                    for (var k = 0; k < clusterIndices.Count; k++)
+                    {
+                        merged.Add(clusterIndices[k]);
+                    }
+                }
+
+                var targetSize = ((nuint)merged.Count / 3) * (nuint)config.simplifyRatio * 3;
+                var bounds = MergeBounds(clusters, groups[i]);
+
+                var error = 0.0f;
+                using var simplified = Simplify(config, mesh, merged.AsReadOnly(), locks.AsReadOnly(), targetSize, &error, Allocator.FreeList);
+
+                if ((nuint)simplified.Length > (nuint)(merged.Count * config.simplifyThreshold))
+                {
+                    bounds.error = float.MaxValue;
+                    OutputGroup(config, mesh, clusters, groups[i], bounds, depth, outputContext, outputCallback);
+                    continue;
+                }
+
+                bounds.error = Math.Max(bounds.error * config.simplifyErrorMergePrevious, error) + error * config.simplifyErrorMergeAdditive;
+
+                var refined = OutputGroup(config, mesh, clusters, groups[i], bounds, depth, outputContext, outputCallback);
+
+                for (var j = 0; j < groups[i].Count; j++)
+                {
+                    clusters[groups[i][j]].Dispose();
+                }
+
+                using var split = Clusterize(config, mesh, (uint*)simplified.GetUnsafePtr(), (nuint)simplified.Length, Allocator.FreeList);
+                for (var j = 0; j < split.Count; j++)
+                {
+                    split[j].refined = refined;
+                    split[j].bounds = bounds;
+                    clusters.Add(split[j]);
+                    pending.Add(clusters.Count - 1);
+                }
+            }
+
+            for (var i = 0; i < groups.Count; i++)
+            {
+                groups[i].Dispose();
+            }
+
+            depth++;
+        }
+
+        if (pending.Count > 0)
+        {
+            var bounds = clusters[pending[0]].bounds;
+            bounds.error = float.MaxValue;
+            OutputGroup(config, mesh, clusters, pending, bounds, depth, outputContext, outputCallback);
+        }
+
+        var finalClusterCount = (nuint)clusters.Count;
+
+        for (var i = 0; i < clusters.Count; i++)
+        {
+            clusters[i].Dispose();
+        }
+
+        return finalClusterCount;
+    }
+}

src/ThridParty/Ghost.MeshOptimizer/MeshOptApi.cs

@@ -0,0 +1,90 @@
+namespace Ghost.MeshOptimizer;
+
+[Flags]
+public enum SimplifyOptions : uint
+{
+    LockBorder = 1 << 0,
+    Sparse = 1 << 1,
+    ErrorAbsolute = 1 << 2,
+    Prune = 1 << 3,
+    Regularize = 1 << 4,
+    Permissive = 1 << 5
+}
+
+[Flags]
+public enum SimplifyVertexOptions : byte
+{
+    Lock = 1 << 0,
+    Protect = 1 << 1
+}
+
+public unsafe partial struct MeshOptApi
+{
+    public const int VERSION = Api.MESHOPTIMIZER_VERSION;
+
+    /// <summary>
+    /// From: <see cref="Api.meshopt_simplify(uint*, uint*, nuint, float*, nuint, nuint, nuint, float, uint, float*)" />
+    /// </summary>
+    [System.Runtime.CompilerServices.MethodImpl(System.Runtime.CompilerServices.MethodImplOptions.AggressiveInlining)]
+    public static nuint Simplify(uint* destination, uint* indices, nuint index_count, float* vertex_positions, nuint vertex_count, nuint vertex_positions_stride, nuint target_index_count, float target_error, SimplifyOptions options, float* result_error)
+    {
+        return Api.meshopt_simplify(
+            destination,
+            indices,
+            index_count,
+            vertex_positions,
+            vertex_count,
+            vertex_positions_stride,
+            target_index_count,
+            target_error,
+            (uint)options,
+            result_error);
+    }
+
+    /// <summary>
+    /// From: <see cref="Api.meshopt_simplifyWithAttributes(uint*, uint*, nuint, float*, nuint, nuint, float*, nuint, float*, nuint, byte*, nuint, float, uint, float*)" />
+    /// </summary>
+    [System.Runtime.CompilerServices.MethodImpl(System.Runtime.CompilerServices.MethodImplOptions.AggressiveInlining)]
+    public static nuint SimplifyWithAttributes(uint* destination, uint* indices, nuint index_count, float* vertex_positions, nuint vertex_count, nuint vertex_positions_stride, float* vertex_attributes, nuint vertex_attributes_stride, float* attribute_weights, nuint attribute_count, byte* vertex_lock, nuint target_index_count, float target_error, SimplifyOptions options, float* result_error)
+    {
+        return Api.meshopt_simplifyWithAttributes(
+            destination,
+            indices,
+            index_count,
+            vertex_positions,
+            vertex_count,
+            vertex_positions_stride,
+            vertex_attributes,
+            vertex_attributes_stride,
+            attribute_weights,
+            attribute_count,
+            vertex_lock,
+            target_index_count,
+            target_error,
+            (uint)options,
+            result_error);
+    }
+
+    /// <summary>
+    /// From: <see cref="Api.meshopt_simplifyWithUpdate(uint*, nuint, float*, nuint, nuint, float*, nuint, float*, nuint, byte*, nuint, float, uint, float*)" />
+    /// </summary>
+    [System.Runtime.CompilerServices.MethodImpl(System.Runtime.CompilerServices.MethodImplOptions.AggressiveInlining)]
+    public static nuint SimplifyWithUpdate(uint* indices, nuint index_count, float* vertex_positions, nuint vertex_count, nuint vertex_positions_stride, float* vertex_attributes, nuint vertex_attributes_stride, float* attribute_weights, nuint attribute_count, byte* vertex_lock, nuint target_index_count, float target_error, SimplifyOptions options, float* result_error)
+    {
+        return Api.meshopt_simplifyWithUpdate(
+            indices,
+            index_count,
+            vertex_positions,
+            vertex_count,
+            vertex_positions_stride,
+            vertex_attributes,
+            vertex_attributes_stride,
+            attribute_weights,
+            attribute_count,
+            vertex_lock,
+            target_index_count,
+            target_error,
+            (uint)options,
+            result_error);
+    }
+}

src/Tools/Ghost.NativeWrapperGen/Emit/WrapperGeneratorEmitter.cs

@@ -276,7 +276,7 @@ public sealed class WrapperGeneratorEmitter
     {
         var func = routed.Function;
         var nameOpts = routed.Apply.Opts?.name;
-        var methodName = naming.GetMethodName(func.Name, nameOpts, routed.TargetStructName);
+        var methodName = naming.GetName(func.Name, nameOpts, routed.TargetStructName);
 
         // Build the parameter plan: for each native parameter, determine the public type
         // and how to pass it to the Api call (applying remaps).

src/Tools/Ghost.NativeWrapperGen/Model/NativeLibrary.cs

@@ -18,7 +18,6 @@ public sealed class NativeStruct
     public required bool IsList { get; init; }
     public required bool IsPointerList { get; init; }
     public string? ListElementType { get; init; }
-    public required bool IsElementLike { get; init; }
 }
 
 public sealed class NativeEnum
@@ -52,4 +51,5 @@ public enum NativeMemberKind
 {
     Field,
     Property,
+    Constant,
 }

src/Tools/Ghost.NativeWrapperGen/Parsing/BindingParser.cs

@@ -10,6 +10,7 @@ public sealed class BindingParser
 {
     public NativeLibrary Parse(string inputDirectory, WrapperConfig config)
     {
+        var members = new List<NativeMember>();
         var structs = new List<NativeStruct>();
         var enums = new List<NativeEnum>();
         var functions = new List<NativeFunction>();
@@ -33,18 +34,17 @@ public sealed class BindingParser
                     continue;
                 }
 
-                var members = ParseMembers(@struct);
+                var structMembers = ParseMembers(@struct);
-                var listInfo = TryMatchList(members);
+                var listInfo = TryMatchList(structMembers);
 
                 structs.Add(new NativeStruct
                 {
                     Name = @struct.Identifier.ValueText,
                     Namespace = namespaceName,
-                    Members = members,
+                    Members = structMembers,
                     IsList = listInfo.IsList,
                     IsPointerList = listInfo.IsPointerList,
                     ListElementType = listInfo.ListElementType,
-                    IsElementLike = members.Any(static m => m.Name == "element" && m.TypeName == "ufbx_element"),
                 });
             }
 

src/Tools/Ghost.NativeWrapperGen/Transform/NamingConventions.cs

@@ -17,13 +17,13 @@ public sealed class NamingConventions
     ///
     /// Supported remove tokens:
     ///   "PREFIX"            — strip the config's NativeTypePrefix from the start (e.g. "nvtt", "ufbx_")
-    ///   "NO_PREFIX($TSelf)" — strip the target struct name minus its type prefix from the start,
+    ///   "$TBare" — strip the target struct name minus its type prefix from the start,
     ///                         case-insensitively (e.g. NvttSurface → "Surface" stripped from "SurfaceWidth")
     ///
     /// nameOpts is the dynamic opts.name object from JSON (may be null).
     /// If no nameOpts are provided, the name is returned with only the library prefix stripped.
     /// </summary>
-    public string GetMethodName(string nativeFunctionName, dynamic? nameOpts, string targetStructName)
+    public string GetName(string nativeFunctionName, dynamic? nameOpts, string targetStructName)
     {
         var name = nativeFunctionName;
 
@@ -47,13 +47,8 @@ public sealed class NamingConventions
             {
                 name = StripPrefixIgnoreCase(name, _config.NativeTypePrefix);
             }
-            else if (token.StartsWith("NO_PREFIX(", StringComparison.Ordinal) && token.EndsWith(')'))
+            else if (string.Equals(token, "$TBare", StringComparison.Ordinal))
             {
-                // Extract $TSelf — it's the literal token "NO_PREFIX($TSelf)", so the struct name
-                // is resolved from the targetStructName argument passed in.
-                // Strip the config prefix from the struct name to get the "bare" part.
-                // Try prefix first, then suffix (handles both nvtt "SurfaceWidth"→"Width"
-                // and ufbx "free_scene"→"free_" styles).
                 var bareStructName = StripPrefixIgnoreCase(targetStructName, _config.NativeTypePrefix);
 
                 // Remove directly, the name maybe nvttSetOutputOptionsOutputHeader, if we only remove prefix and suffix, OutputOptions in the middle will be ignored, so we remove the bare struct name directly, case-insensitively.
@@ -66,7 +61,7 @@ public sealed class NamingConventions
         var style = nameOpts.style as string;
         if (!string.IsNullOrEmpty(style))
         {
-            if (string.Equals(style, "PascalCase", StringComparison.OrdinalIgnoreCase))
+            if (string.Equals(style, "PascalCase", StringComparison.Ordinal))
             {
                 int counter = 0;
                 Span<char> nameSpan = stackalloc char[name.Length];
@@ -83,10 +78,10 @@ public sealed class NamingConventions
 
                     if (name[i] == '_')
                     {
-                        while (name[i] == '_' && i < name.Length)
+                        do
                         {
                             i++;
-                        }
+                        } while (i < name.Length && name[i] == '_');
 
                         nameSpan[counter] = char.ToUpperInvariant(name[i]);
                         counter++;
@@ -98,6 +93,54 @@ public sealed class NamingConventions
                     counter++;
                 }
 
+                name = nameSpan[..counter].ToString();
+            }
+            else if (string.Equals(style, "ALL_CAPS", StringComparison.Ordinal))
+            {
+                int counter = 0;
+                Span<char> nameSpan = stackalloc char[name.Length * 2]; // Worst case, every character is uppercase and followed by an underscore.
+
+                for (int i = 0; i < name.Length; i++)
+                {
+                    // ___ to _
+                    if (name[i] == '_')
+                    {
+                        while (i + 1 < name.Length && name[i + 1] == '_')
+                        {
+                            i++;
+                        }
+
+                        nameSpan[counter] = '_';
+                        counter++;
+
+                        continue;
+                    }
+
+                    // AbC to AB_C
+                    if (i > 0 && char.IsUpper(name[i]) && char.IsLower(name[i - 1]))
+                    {
+                        nameSpan[counter] = '_';
+                        counter++;
+                    }
+
+                    // ABC to ABC
+                    while (i < name.Length && char.IsUpper(name[i]))
+                    {
+                        nameSpan[counter] = name[i];
+
+                        counter++;
+                        i++;
+                    }
+
+                    if (i == name.Length)
+                    {
+                        break;
+                    }
+
+                    nameSpan[counter] = char.ToUpperInvariant(name[i]);
+                    counter++;
+                }
+
                 name = nameSpan[..counter].ToString();
             }
         }

src/Tools/Ghost.NativeWrapperGen/configs/meshopt.json

@@ -24,7 +24,7 @@
           "name": {
             "remove": [
               "PREFIX",
-              "NO_PREFIX($TSelf)" // NO_PREFIX(NvttSurface) will change "NvttSurface" to "Surface", the prefix is determined by the "nativeTypePrefix" field at the top level of this config
+              "$TBare" // NO_PREFIX(NvttSurface) will change "NvttSurface" to "Surface", the prefix is determined by the "nativeTypePrefix" field at the top level of this config
             ],
             "style": "PascalCase"
           }
@@ -43,7 +43,7 @@
           "name": {
             "remove": [
               "PREFIX",
-              "NO_PREFIX($TSelf)"
+              "$TBare"
             ],
             "style": "PascalCase"
           }
@@ -64,6 +64,21 @@
           }
         }
       }
+    },
+    {
+      "filter": "CONST",
+      "targetType": "MeshOptApi",
+      "apply": {
+        "type": "CONST",
+        "opts": {
+          "name": {
+            "remove": [
+              "PREFIX"
+            ],
+            "style": "ALL_CAPS"
+          }
+        }
+      }
     }
   ]
 }

src/Tools/Ghost.NativeWrapperGen/configs/nvtt.json

@@ -68,7 +68,7 @@
           "name": {
             "remove": [
               "PREFIX",
-              "NO_PREFIX($TSelf)" // NO_PREFIX(NvttSurface) will change "NvttSurface" to "Surface", the prefix is determined by the "nativeTypePrefix" field at the top level of this config
+              "$TBare" // NO_PREFIX(NvttSurface) will change "NvttSurface" to "Surface", the prefix is determined by the "nativeTypePrefix" field at the top level of this config
             ]
           }
         }
@@ -86,21 +86,7 @@
           "name": {
             "remove": [
               "PREFIX",
-              "NO_PREFIX($TSelf)"
+              "$TBare"
-            ]
-          }
-        }
-      }
-    },
-    {
-      "filter": "EXTERN_API",
-      "targetType": "NvttApi",
-      "apply": {
-        "type": "STATIC_METHOD",
-        "opts": {
-          "name": {
-            "remove": [
-              "PREFIX"
             ]
           }
         }

src/Tools/Ghost.NativeWrapperGen/configs/ufbx.json

@@ -68,7 +68,7 @@
           "name": {
             "remove": [
               "PREFIX",
-              "NO_PREFIX($TSelf)" // NO_PREFIX(ufbx_scene) will output "scene" change "free_scene" to "free", the prefix is determined by the "nativeTypePrefix" field at the top level of this config
+              "$TBare" // NO_PREFIX(ufbx_scene) will output "scene" change "free_scene" to "free", the prefix is determined by the "nativeTypePrefix" field at the top level of this config
             ],
             "style": "PascalCase"
           }
@@ -87,7 +87,7 @@
           "name": {
             "remove": [
               "PREFIX",
-              "NO_PREFIX($TSelf)"
+              "$TBare"
             ],
             "style": "PascalCase"
           }