clusterizer: Separate sparse and dense adjacency builds

In sparse mode we still had a memset(counts) that cleared the entire
array; this is not necessary in sparse mode and can be meaningfully
slower especially if the region of memory is fresh and the pages need to
be committed.

We can instead clear every used index in a separate pass. This results
in a small further performance improvement. This also results in sparse
builds using almost completely separate code, at which point it's
cleaner to split it into a separate function, which is what this commit
also does.

src/clusterizer.cpp

@@ -30,10 +30,6 @@ static void buildTriangleAdjacency(TriangleAdjacency2& adjacency, const unsigned
 {
 	size_t face_count = index_count / 3;
 
-	// sparse mode can build adjacency more quickly by ignoring unused vertices
-	const unsigned int sparse_seen = 1u << 31;
-	bool sparse = vertex_count > index_count && index_count < sparse_seen;
-
 	// allocate arrays
 	adjacency.counts = allocator.allocate<unsigned int>(vertex_count);
 	adjacency.offsets = allocator.allocate<unsigned int>(vertex_count);
@@ -52,27 +48,68 @@ static void buildTriangleAdjacency(TriangleAdjacency2& adjacency, const unsigned
 	// fill offset table
 	unsigned int offset = 0;
 
-	if (sparse)
+	for (size_t i = 0; i < vertex_count; ++i)
 	{
-		// when using sparse mode this pass uses sparse_seen bit to tag visited vertices
-		for (size_t i = 0; i < index_count; ++i)
-		{
-			unsigned int v = indices[i];
+		adjacency.offsets[i] = offset;
+		offset += adjacency.counts[i];
+	}
 
-			if ((adjacency.counts[v] & sparse_seen) == 0)
-			{
-				adjacency.offsets[v] = offset;
-				offset += adjacency.counts[v];
-				adjacency.counts[v] |= sparse_seen;
-			}
-		}
+	assert(offset == index_count);
+
+	// fill triangle data
+	for (size_t i = 0; i < face_count; ++i)
+	{
+		unsigned int a = indices[i * 3 + 0], b = indices[i * 3 + 1], c = indices[i * 3 + 2];
+
+		adjacency.data[adjacency.offsets[a]++] = unsigned(i);
+		adjacency.data[adjacency.offsets[b]++] = unsigned(i);
+		adjacency.data[adjacency.offsets[c]++] = unsigned(i);
 	}
-	else
+
+	// fix offsets that have been disturbed by the previous pass
+	for (size_t i = 0; i < vertex_count; ++i)
 	{
-		for (size_t i = 0; i < vertex_count; ++i)
+		assert(adjacency.offsets[i] >= adjacency.counts[i]);
+		adjacency.offsets[i] -= adjacency.counts[i];
+	}
+}
+
+static void buildTriangleAdjacencySparse(TriangleAdjacency2& adjacency, const unsigned int* indices, size_t index_count, size_t vertex_count, meshopt_Allocator& allocator)
+{
+	size_t face_count = index_count / 3;
+
+	// sparse mode can build adjacency more quickly by ignoring unused vertices, using a bit to mark visited vertices
+	const unsigned int sparse_seen = 1u << 31;
+	assert(index_count < sparse_seen);
+
+	// allocate arrays
+	adjacency.counts = allocator.allocate<unsigned int>(vertex_count);
+	adjacency.offsets = allocator.allocate<unsigned int>(vertex_count);
+	adjacency.data = allocator.allocate<unsigned int>(index_count);
+
+	// fill triangle counts
+	for (size_t i = 0; i < index_count; ++i)
+		assert(indices[i] < vertex_count);
+
+	for (size_t i = 0; i < index_count; ++i)
+		adjacency.counts[indices[i]] = 0;
+
+	for (size_t i = 0; i < index_count; ++i)
+		adjacency.counts[indices[i]]++;
+
+	// fill offset table
+	unsigned int offset = 0;
+
+	// when using sparse mode this pass uses sparse_seen bit to tag visited vertices
+	for (size_t i = 0; i < index_count; ++i)
+	{
+		unsigned int v = indices[i];
+
+		if ((adjacency.counts[v] & sparse_seen) == 0)
 		{
-			adjacency.offsets[i] = offset;
-			offset += adjacency.counts[i];
+			adjacency.offsets[v] = offset;
+			offset += adjacency.counts[v];
+			adjacency.counts[v] |= sparse_seen;
 		}
 	}
 
@@ -89,28 +126,17 @@ static void buildTriangleAdjacency(TriangleAdjacency2& adjacency, const unsigned
 	}
 
 	// fix offsets that have been disturbed by the previous pass
-	if (sparse)
+	// when using sparse mode this pass also fixes counts (that were marked with sparse_seen)
+	for (size_t i = 0; i < index_count; ++i)
 	{
-		// when using sparse mode this pass also fixes counts (that were marked with sparse_seen)
-		for (size_t i = 0; i < index_count; ++i)
-		{
-			unsigned int v = indices[i];
-
-			if (adjacency.counts[v] & sparse_seen)
-			{
-				adjacency.counts[v] &= ~sparse_seen;
+		unsigned int v = indices[i];
 
-				assert(adjacency.offsets[v] >= adjacency.counts[v]);
-				adjacency.offsets[v] -= adjacency.counts[v];
-			}
-		}
-	}
-	else
-	{
-		for (size_t i = 0; i < vertex_count; ++i)
+		if (adjacency.counts[v] & sparse_seen)
 		{
-			assert(adjacency.offsets[i] >= adjacency.counts[i]);
-			adjacency.offsets[i] -= adjacency.counts[i];
+			adjacency.counts[v] &= ~sparse_seen;
+
+			assert(adjacency.offsets[v] >= adjacency.counts[v]);
+			adjacency.offsets[v] -= adjacency.counts[v];
 		}
 	}
 }
@@ -592,7 +618,10 @@ size_t meshopt_buildMeshlets(meshopt_Meshlet* meshlets, unsigned int* meshlet_ve
 	meshopt_Allocator allocator;
 
 	TriangleAdjacency2 adjacency = {};
-	buildTriangleAdjacency(adjacency, indices, index_count, vertex_count, allocator);
+	if (vertex_count > index_count && index_count < (1u << 31))
+		buildTriangleAdjacencySparse(adjacency, indices, index_count, vertex_count, allocator);
+	else
+		buildTriangleAdjacency(adjacency, indices, index_count, vertex_count, allocator);
 
 	// live triangle counts; note, we alias adjacency.counts as we remove triangles after emitting them so the counts always match
 	unsigned int* live_triangles = adjacency.counts;