OpenSubDiv: add support for an OpenGL evaluator

This evaluator is used in order to evaluate subdivision at render time, allowing for
faster renders of meshes with a subdivision surface modifier placed at the last
position in the modifier list.

When evaluating the subsurf modifier, we detect whether we can delegate evaluation
to the draw code. If so, the subdivision is first evaluated on the GPU using our own
custom evaluator (only the coarse data needs to be initially sent to the GPU), then,
buffers for the final `MeshBufferCache` are filled on the GPU using a set of
compute shaders. However, some buffers are still filled on the CPU side, if doing so
on the GPU is impractical (e.g. the line adjacency buffer used for x-ray, whose
logic is hardly GPU compatible).

This is done at the mesh buffer extraction level so that the result can be readily used
in the various OpenGL engines, without having to write custom geometry or tesselation
shaders.

We use our own subdivision evaluation shaders, instead of OpenSubDiv's vanilla one, in
order to control the data layout, and interpolation. For example, we store vertex colors
as compressed 16-bit integers, while OpenSubDiv's default evaluator only work for float
types.

In order to still access the modified geometry on the CPU side, for use in modifiers
or transform operators, a dedicated wrapper type is added `MESH_WRAPPER_TYPE_SUBD`.
Subdivision will be lazily evaluated via `BKE_object_get_evaluated_mesh` which will
create such a wrapper if possible. If the final subdivision surface is not needed on
the CPU side, `BKE_object_get_evaluated_mesh_no_subsurf` should be used.

Enabling or disabling GPU subdivision can be done through the user preferences (under
Viewport -> Subdivision).

See patch description for benchmarks.

Reviewed By: campbellbarton, jbakker, fclem, brecht, #eevee_viewport

Differential Revision: https://developer.blender.org/D12406

1 files changed, 67 insertions, 5 deletions

diff --git a/source/blender/draw/intern/mesh_extractors/extract_mesh_ibo_lines_adjacency.cc b/source/blender/draw/intern/mesh_extractors/extract_mesh_ibo_lines_adjacency.cc
index e7dabfa9ee2..6855feb51ed 100644
--- a/source/blender/draw/intern/mesh_extractors/extract_mesh_ibo_lines_adjacency.cc
+++ b/source/blender/draw/intern/mesh_extractors/extract_mesh_ibo_lines_adjacency.cc
@@ -26,6 +26,7 @@
 
 #include "MEM_guardedalloc.h"
 
+#include "draw_subdivision.h"
 #include "extract_mesh.h"
 
 namespace blender::draw {
@@ -44,6 +45,18 @@ struct MeshExtract_LineAdjacency_Data {
   uint *vert_to_loop;
 };
 
+static void line_adjacency_data_init(MeshExtract_LineAdjacency_Data *data,
+                                     uint vert_len,
+                                     uint loop_len,
+                                     uint tess_edge_len)
+{
+  data->vert_to_loop = static_cast<uint *>(MEM_callocN(sizeof(uint) * vert_len, __func__));
+
+  GPU_indexbuf_init(&data->elb, GPU_PRIM_LINES_ADJ, tess_edge_len, loop_len);
+  data->eh = BLI_edgehash_new_ex(__func__, tess_edge_len);
+  data->is_manifold = true;
+}
+
 static void extract_lines_adjacency_init(const MeshRenderData *mr,
                                          struct MeshBatchCache *UNUSED(cache),
                                          void *UNUSED(buf),
@@ -55,11 +68,7 @@ static void extract_lines_adjacency_init(const MeshRenderData *mr,
   uint tess_edge_len = mr->loop_len + mr->tri_len - mr->poly_len;
 
   MeshExtract_LineAdjacency_Data *data = static_cast<MeshExtract_LineAdjacency_Data *>(tls_data);
-  data->vert_to_loop = static_cast<uint *>(MEM_callocN(sizeof(uint) * mr->vert_len, __func__));
-
-  GPU_indexbuf_init(&data->elb, GPU_PRIM_LINES_ADJ, tess_edge_len, mr->loop_len);
-  data->eh = BLI_edgehash_new_ex(__func__, tess_edge_len);
-  data->is_manifold = true;
+  line_adjacency_data_init(data, mr->vert_len, mr->loop_len, tess_edge_len);
 }
 
 BLI_INLINE void lines_adjacency_triangle(
@@ -171,6 +180,56 @@ static void extract_lines_adjacency_finish(const MeshRenderData *UNUSED(mr),
   MEM_freeN(data->vert_to_loop);
 }
 
+static void extract_lines_adjacency_init_subdiv(const DRWSubdivCache *subdiv_cache,
+                                                const MeshRenderData *UNUSED(mr),
+                                                struct MeshBatchCache *UNUSED(cache),
+                                                void *UNUSED(buf),
+                                                void *_data)
+{
+  MeshExtract_LineAdjacency_Data *data = static_cast<MeshExtract_LineAdjacency_Data *>(_data);
+
+  /* For each polygon there is (loop + triangle - 1) edges. Since we only have quads, and a quad
+   * is split into 2 triangles, we have (loop + 2 - 1) = (loop + 1) edges for each quad, or in
+   * total: (number_of_loops + number_of_quads). */
+  const uint tess_len = subdiv_cache->num_subdiv_loops + subdiv_cache->num_subdiv_quads;
+  line_adjacency_data_init(
+      data, tess_len, subdiv_cache->num_subdiv_verts, subdiv_cache->num_subdiv_loops);
+}
+
+static void extract_lines_adjacency_iter_subdiv(const DRWSubdivCache *subdiv_cache,
+                                                const MeshRenderData *UNUSED(mr),
+                                                void *_data)
+{
+  MeshExtract_LineAdjacency_Data *data = static_cast<MeshExtract_LineAdjacency_Data *>(_data);
+
+  for (uint i = 0; i < subdiv_cache->num_subdiv_quads; i++) {
+    const uint loop_index = i * 4;
+    const uint l0 = loop_index + 0;
+    const uint l1 = loop_index + 1;
+    const uint l2 = loop_index + 2;
+    const uint l3 = loop_index + 3;
+
+    const uint v0 = subdiv_cache->subdiv_loop_subdiv_vert_index[l0];
+    const uint v1 = subdiv_cache->subdiv_loop_subdiv_vert_index[l1];
+    const uint v2 = subdiv_cache->subdiv_loop_subdiv_vert_index[l2];
+    const uint v3 = subdiv_cache->subdiv_loop_subdiv_vert_index[l3];
+
+    lines_adjacency_triangle(v0, v1, v2, l0, l1, l2, data);
+    lines_adjacency_triangle(v0, v2, v3, l0, l2, l3, data);
+  }
+}
+
+static void extract_lines_adjacency_finish_subdiv(const DRWSubdivCache *UNUSED(subdiv_cache),
+                                                  void *buf,
+                                                  void *_data)
+{
+  GPUIndexBuf *ibo = static_cast<GPUIndexBuf *>(buf);
+  MeshExtract_LineAdjacency_Data *data = static_cast<MeshExtract_LineAdjacency_Data *>(_data);
+  GPU_indexbuf_build_in_place(&data->elb, ibo);
+  BLI_edgehash_free(data->eh, nullptr);
+  MEM_freeN(data->vert_to_loop);
+}
+
 #undef NO_EDGE
 
 constexpr MeshExtract create_extractor_lines_adjacency()
@@ -180,6 +239,9 @@ constexpr MeshExtract create_extractor_lines_adjacency()
   extractor.iter_looptri_bm = extract_lines_adjacency_iter_looptri_bm;
   extractor.iter_looptri_mesh = extract_lines_adjacency_iter_looptri_mesh;
   extractor.finish = extract_lines_adjacency_finish;
+  extractor.init_subdiv = extract_lines_adjacency_init_subdiv;
+  extractor.iter_subdiv = extract_lines_adjacency_iter_subdiv;
+  extractor.finish_subdiv = extract_lines_adjacency_finish_subdiv;
   extractor.data_type = MR_DATA_NONE;
   extractor.data_size = sizeof(MeshExtract_LineAdjacency_Data);
   extractor.use_threading = false;