Cycles: Record all possible volume intersections for SSS and camera checks

This replaces sequential ray moving followed with scene intersection with
single BVH traversal, which gives us all possible intersections.

Only implemented for CPU, due to qsort and a bigger memory usage on GPU
which we rather avoid. GPU still uses the regular bvh volume intersection code, while CPU now uses the new code.

This improves render performance for scenes with:
a) Camera inside volume mesh
b) SSS mesh intersecting a volume mesh/domain

In simple volume files (not much geometry) performance is roughly the same
(slightly faster). In files with a lot of geometry, the performance
increase is larger. bmps.blend with a volume shader and camera inside the
mesh, it renders ~10% faster here.

Patch by Sergey and myself.

Differential Revision: https://developer.blender.org/D1264

1 files changed, 1 insertions, 14 deletions

diff --git a/intern/cycles/kernel/kernel_shadow.h b/intern/cycles/kernel/kernel_shadow.h
index 8923fcebee5..d7c4fa02bcf 100644
--- a/intern/cycles/kernel/kernel_shadow.h
+++ b/intern/cycles/kernel/kernel_shadow.h
@@ -39,19 +39,6 @@ CCL_NAMESPACE_BEGIN
  * This is CPU only because of qsort, and malloc or high stack space usage to
  * record all these intersections. */
 
-ccl_device_noinline int shadow_intersections_compare(const void *a, const void *b)
-{
-	const Intersection *isect_a = (const Intersection*)a;
-	const Intersection *isect_b = (const Intersection*)b;
-
-	if(isect_a->t < isect_b->t)
-		return -1;
-	else if(isect_a->t > isect_b->t)
-		return 1;
-	else
-		return 0;
-}
-
 #define STACK_MAX_HITS 64
 
 ccl_device_inline bool shadow_blocked(KernelGlobals *kg, PathState *state, Ray *ray, float3 *shadow)
@@ -95,7 +82,7 @@ ccl_device_inline bool shadow_blocked(KernelGlobals *kg, PathState *state, Ray *
 			PathState ps = *state;
 #endif
 
-			qsort(hits, num_hits, sizeof(Intersection), shadow_intersections_compare);
+			qsort(hits, num_hits, sizeof(Intersection), intersections_compare);
 
 			for(int hit = 0; hit < num_hits; hit++, isect++) {
 				/* adjust intersection distance for moving ray forward */