DrawManager: Use Compute Shader to Update Hair.

This patch will use compute shaders to create the VBO for hair.
The previous implementation uses transform feedback.

Timings before: between 0.000069s and 0.000362s.
Timings after:  between 0.000032s and 0.000092s.

Speedup isn't noticeable by end-users. The patch is used to test
the new compute shader pipeline and integrate it with the draw
manager. Allowing EEVEE, Workbench and other draw engines to
use compute shaders with the introduction of `DRW_shgroup_call_compute`
and `DRW_shgroup_vertex_buffer`.

Future improvements are possible by generating the index buffer
of hair directly on the GPU.

NOTE: that compute shaders aren't supported by Apple and still use
the transform feedback workaround.

Reviewed By: fclem

Differential Revision: https://developer.blender.org/D11057

1 files changed, 36 insertions, 0 deletions

diff --git a/source/blender/draw/intern/draw_manager_data.c b/source/blender/draw/intern/draw_manager_data.c
index 6bdc5305fed..3b852e7f8c8 100644
--- a/source/blender/draw/intern/draw_manager_data.c
+++ b/source/blender/draw/intern/draw_manager_data.c
@@ -47,6 +47,7 @@
 #endif
 
 #include "GPU_buffers.h"
+#include "GPU_capabilities.h"
 #include "GPU_material.h"
 #include "GPU_uniform_buffer.h"
 
@@ -446,6 +447,19 @@ void DRW_shgroup_uniform_vec4_array_copy(DRWShadingGroup *shgroup,
   }
 }
 
+void DRW_shgroup_vertex_buffer(DRWShadingGroup *shgroup,
+                               const char *name,
+                               GPUVertBuf *vertex_buffer)
+{
+  int location = GPU_shader_get_ssbo(shgroup->shader, name);
+  if (location == -1) {
+    BLI_assert(false && "Unable to locate binding of shader storage buffer objects.");
+    return;
+  }
+  drw_shgroup_uniform_create_ex(
+      shgroup, location, DRW_UNIFORM_VERTEX_BUFFER_AS_STORAGE, vertex_buffer, 0, 0, 1);
+}
+
 /** \} */
 
 /* -------------------------------------------------------------------- */
@@ -700,6 +714,17 @@ static void drw_command_draw_intance_range(
   cmd->inst_count = count;
 }
 
+static void drw_command_compute(DRWShadingGroup *shgroup,
+                                int groups_x_len,
+                                int groups_y_len,
+                                int groups_z_len)
+{
+  DRWCommandCompute *cmd = drw_command_create(shgroup, DRW_CMD_COMPUTE);
+  cmd->groups_x_len = groups_x_len;
+  cmd->groups_y_len = groups_y_len;
+  cmd->groups_z_len = groups_z_len;
+}
+
 static void drw_command_draw_procedural(DRWShadingGroup *shgroup,
                                         GPUBatch *batch,
                                         DRWResourceHandle handle,
@@ -815,6 +840,17 @@ void DRW_shgroup_call_instance_range(
   drw_command_draw_intance_range(shgroup, geom, handle, i_sta, i_ct);
 }
 
+void DRW_shgroup_call_compute(DRWShadingGroup *shgroup,
+                              int groups_x_len,
+                              int groups_y_len,
+                              int groups_z_len)
+{
+  BLI_assert(groups_x_len > 0 && groups_y_len > 0 && groups_z_len > 0);
+  BLI_assert(GPU_compute_shader_support());
+
+  drw_command_compute(shgroup, groups_x_len, groups_y_len, groups_z_len);
+}
+
 static void drw_shgroup_call_procedural_add_ex(DRWShadingGroup *shgroup,
                                                GPUBatch *geom,
                                                Object *ob,