Cycles: decouple shadow paths from main path on GPU

The motivation for this is twofold. It improves performance (5-10% on most
benchmark scenes), and will help  to bring back transparency support for the
ambient occlusion pass.

* Duplicate some members from the main path state in the shadow path state.
* Add shadow paths incrementally to the array similar to what we do for
  the shadow catchers.
* For the scheduling, allow running shade surface and shade volume kernels
  as long as there is enough space in the shadow paths array. If not, execute
  shadow kernels until it is empty.

* Add IntegratorShadowState and ConstIntegratorShadowState typedefs that
  can be different between CPU and GPU. For GPU both main and shadow paths
  juse have an integer for SoA access. Bt with CPU it's a different pointer
  type so we get type safety checks in code shared between CPU and GPU.
* For CPU, add a separate IntegratorShadowStateCPU struct embedded in
  IntegratorShadowState.
* Update various functions to take the shadow state, and make SVM take either
  type of state using templates.

Differential Revision: https://developer.blender.org/D12889

1 files changed, 25 insertions, 11 deletions

diff --git a/intern/cycles/kernel/device/cpu/kernel_arch_impl.h b/intern/cycles/kernel/device/cpu/kernel_arch_impl.h
index bf8667ac045..2b0eea4fb61 100644
--- a/intern/cycles/kernel/device/cpu/kernel_arch_impl.h
+++ b/intern/cycles/kernel/device/cpu/kernel_arch_impl.h
@@ -69,6 +69,18 @@ CCL_NAMESPACE_BEGIN
 #  define KERNEL_INVOKE(name, ...) integrator_##name(__VA_ARGS__)
 #endif
 
+/* TODO: Either use something like get_work_pixel(), or simplify tile which is passed here, so
+ * that it does not contain unused fields. */
+#define DEFINE_INTEGRATOR_INIT_KERNEL(name) \
+  bool KERNEL_FUNCTION_FULL_NAME(integrator_##name)(const KernelGlobalsCPU *kg, \
+                                                    IntegratorStateCPU *state, \
+                                                    KernelWorkTile *tile, \
+                                                    ccl_global float *render_buffer) \
+  { \
+    return KERNEL_INVOKE( \
+        name, kg, state, tile, render_buffer, tile->x, tile->y, tile->start_sample); \
+  }
+
 #define DEFINE_INTEGRATOR_KERNEL(name) \
   void KERNEL_FUNCTION_FULL_NAME(integrator_##name)(const KernelGlobalsCPU *kg, \
                                                     IntegratorStateCPU *state) \
@@ -83,30 +95,32 @@ CCL_NAMESPACE_BEGIN
     KERNEL_INVOKE(name, kg, state, render_buffer); \
   }
 
-/* TODO: Either use something like get_work_pixel(), or simplify tile which is passed here, so
- * that it does not contain unused fields. */
-#define DEFINE_INTEGRATOR_INIT_KERNEL(name) \
-  bool KERNEL_FUNCTION_FULL_NAME(integrator_##name)(const KernelGlobalsCPU *kg, \
-                                                    IntegratorStateCPU *state, \
-                                                    KernelWorkTile *tile, \
-                                                    ccl_global float *render_buffer) \
+#define DEFINE_INTEGRATOR_SHADOW_KERNEL(name) \
+  void KERNEL_FUNCTION_FULL_NAME(integrator_##name)(const KernelGlobalsCPU *kg, \
+                                                    IntegratorStateCPU *state) \
   { \
-    return KERNEL_INVOKE( \
-        name, kg, state, tile, render_buffer, tile->x, tile->y, tile->start_sample); \
+    KERNEL_INVOKE(name, kg, &state->shadow); \
+  }
+
+#define DEFINE_INTEGRATOR_SHADOW_SHADE_KERNEL(name) \
+  void KERNEL_FUNCTION_FULL_NAME(integrator_##name)( \
+      const KernelGlobalsCPU *kg, IntegratorStateCPU *state, ccl_global float *render_buffer) \
+  { \
+    KERNEL_INVOKE(name, kg, &state->shadow, render_buffer); \
   }
 
 DEFINE_INTEGRATOR_INIT_KERNEL(init_from_camera)
 DEFINE_INTEGRATOR_INIT_KERNEL(init_from_bake)
 DEFINE_INTEGRATOR_KERNEL(intersect_closest)
-DEFINE_INTEGRATOR_KERNEL(intersect_shadow)
 DEFINE_INTEGRATOR_KERNEL(intersect_subsurface)
 DEFINE_INTEGRATOR_KERNEL(intersect_volume_stack)
 DEFINE_INTEGRATOR_SHADE_KERNEL(shade_background)
 DEFINE_INTEGRATOR_SHADE_KERNEL(shade_light)
-DEFINE_INTEGRATOR_SHADE_KERNEL(shade_shadow)
 DEFINE_INTEGRATOR_SHADE_KERNEL(shade_surface)
 DEFINE_INTEGRATOR_SHADE_KERNEL(shade_volume)
 DEFINE_INTEGRATOR_SHADE_KERNEL(megakernel)
+DEFINE_INTEGRATOR_SHADOW_KERNEL(intersect_shadow)
+DEFINE_INTEGRATOR_SHADOW_SHADE_KERNEL(shade_shadow)
 
 /* --------------------------------------------------------------------
  * Shader evaluation.