Split (1)

train · ~2.3M rows (showing the first 1.36M)

name string	code string	asm string	file string
unsigned long embree::avx::createMortonCodeArray<embree::Instance>(embree::Instance*, embree::vector_t<embree::avx::BVHBuilderMorton::BuildPrim, embree::aligned_monitored_allocator<embree::avx::BVHBuilderMorton::BuildPrim, 8ul>>&, embree::BuildProgressMonitor&)::'lambda'(embree::range<unsigned long> const&, unsigned long)::operator()(embree::range<unsigned long> const&, unsigned long) const (.cold.1)	__forceinline QuaternionT( const T& r, const T& i, const T& j, const T& k ) : r(r), i(i), j(j), k(k) {}	vmovss 0x3c(%rdi), %xmm3 vmovss 0xc(%rdi), %xmm6 vmovss 0x1c(%rdi), %xmm4 vmovss 0x2c(%rdi), %xmm2 vmovaps (%rdi), %xmm1 vxorps %xmm0, %xmm0, %xmm0 vshufps $0xe9, %xmm0, %xmm1, %xmm1 # xmm1 = xmm1[1,2],xmm0[2,3] vblendps $0x4, 0x10(%rdi), %xmm1, %xmm1 # xmm1 = xmm1[0,1],mem[2],xmm1[3] vmulss %xmm6, %xmm6, %xmm8 vmulss %xmm3, %xmm3, %xmm9 vaddss %xmm8, %xmm9, %xmm5 vbroadcastss 0x1d1fa73(%rip), %xmm10 # 0x1f20ec0 vxorps %xmm4, %xmm10, %xmm7 vmulss %xmm4, %xmm7, %xmm7 vaddss %xmm5, %xmm7, %xmm5 vxorps %xmm2, %xmm10, %xmm10 vmulss %xmm2, %xmm10, %xmm10 vaddss %xmm5, %xmm10, %xmm11 vmulss %xmm2, %xmm3, %xmm5 vmulss %xmm4, %xmm6, %xmm12 vaddss %xmm5, %xmm12, %xmm13 vsubss %xmm5, %xmm12, %xmm5 vmulss %xmm2, %xmm6, %xmm12 vsubss %xmm8, %xmm9, %xmm9 vmulss %xmm4, %xmm4, %xmm8 vaddss %xmm9, %xmm8, %xmm8 vaddss %xmm8, %xmm10, %xmm8 vmulss %xmm4, %xmm3, %xmm10 vmulss %xmm6, %xmm3, %xmm3 vsubss %xmm10, %xmm12, %xmm6 vmulss %xmm2, %xmm4, %xmm4 vaddss %xmm10, %xmm12, %xmm10 vaddss %xmm3, %xmm4, %xmm12 vsubss %xmm3, %xmm4, %xmm3 vaddss %xmm13, %xmm13, %xmm4 vaddss %xmm6, %xmm6, %xmm6 vaddss %xmm7, %xmm9, %xmm7 vmulss %xmm2, %xmm2, %xmm2 vaddss %xmm7, %xmm2, %xmm7 vshufps $0x0, %xmm11, %xmm11, %xmm2 # xmm2 = xmm11[0,0,0,0] vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vshufps $0x0, %xmm6, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmovaps 0x1ceb229(%rip), %xmm9 # 0x1eec700 vmulps %xmm6, %xmm9, %xmm6 vmovsd 0x1ceb20d(%rip), %xmm11 # 0x1eec6f0 vmulps %xmm4, %xmm11, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vmovss 0x1ceb221(%rip), %xmm6 # 0x1eec714 vmulps %xmm6, %xmm2, %xmm2 vaddps %xmm4, %xmm2, %xmm2 vaddss %xmm5, %xmm5, %xmm4 vaddss %xmm12, %xmm12, %xmm5 vaddss %xmm10, %xmm10, %xmm10 vaddss %xmm3, %xmm3, %xmm12 vshufps $0x0, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[0,0,0,0] vshufps $0x0, %xmm8, %xmm8, %xmm4 # xmm4 = xmm8[0,0,0,0] vshufps $0x0, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vmulps %xmm5, %xmm9, %xmm5 vmulps %xmm4, %xmm11, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vshufps $0x0, %xmm10, %xmm10, %xmm4 # xmm4 = xmm10[0,0,0,0] vshufps $0x0, %xmm12, %xmm12, %xmm5 # xmm5 = xmm12[0,0,0,0] vshufps $0x0, %xmm7, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmulps %xmm7, %xmm9, %xmm7 vmulps %xmm5, %xmm11, %xmm5 vaddps %xmm7, %xmm5, %xmm5 vmulps %xmm6, %xmm4, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vaddps %xmm0, %xmm1, %xmm5 vbroadcastss (%rdi), %xmm1 vmulps %xmm0, %xmm4, %xmm6 vmulps %xmm0, %xmm3, %xmm0 vaddps %xmm6, %xmm0, %xmm0 vmulps %xmm2, %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x10(%rdi), %xmm1 vbroadcastss 0x14(%rdi), %xmm7 vmulps %xmm3, %xmm7, %xmm7 vaddps %xmm6, %xmm7, %xmm6 vmulps %xmm2, %xmm1, %xmm1 vaddps %xmm6, %xmm1, %xmm1 vbroadcastss 0x20(%rdi), %xmm6 vbroadcastss 0x24(%rdi), %xmm7 vbroadcastss 0x28(%rdi), %xmm8 vmulps %xmm4, %xmm8, %xmm8 vmulps %xmm3, %xmm7, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vmulps %xmm2, %xmm6, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0x30(%rdi), %xmm7 vbroadcastss 0x34(%rdi), %xmm8 vbroadcastss 0x38(%rdi), %xmm9 vmulps %xmm4, %xmm9, %xmm4 vmulps %xmm3, %xmm8, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vmulps %xmm2, %xmm7, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vaddps %xmm2, %xmm5, %xmm2 movq 0x58(%rsi), %rax vmovaps 0x10(%rax), %xmm3 vmovaps 0x20(%rax), %xmm4 vminps 0x30(%rax), %xmm3, %xmm3 vmaxps 0x40(%rax), %xmm4, %xmm4 vshufps $0x0, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[0,0,0,0] vshufps $0x55, %xmm3, %xmm3, %xmm7 # xmm7 = xmm3[1,1,1,1] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm2, %xmm3, %xmm3 vmulps %xmm1, %xmm7, %xmm7 vaddps %xmm3, %xmm7, %xmm8 vmulps %xmm0, %xmm5, %xmm5 vaddps %xmm5, %xmm8, %xmm9 vbroadcastss 0x1cea3f9(%rip), %xmm10 # 0x1eeba20 vminps %xmm9, %xmm10, %xmm10 vbroadcastss 0x1ceb54f(%rip), %xmm11 # 0x1eecb84 vmaxps %xmm9, %xmm11, %xmm9 vshufps $0xaa, %xmm4, %xmm4, %xmm11 # xmm11 = xmm4[2,2,2,2] vmulps %xmm6, %xmm11, %xmm6 vaddps %xmm6, %xmm2, %xmm2 vaddps %xmm2, %xmm7, %xmm6 vaddps %xmm6, %xmm5, %xmm7 vminps %xmm7, %xmm10, %xmm10 vmaxps %xmm7, %xmm9, %xmm7 vshufps $0x55, %xmm4, %xmm4, %xmm9 # xmm9 = xmm4[1,1,1,1] vmulps %xmm1, %xmm9, %xmm1 vaddps %xmm1, %xmm3, %xmm3 vaddps %xmm3, %xmm5, %xmm9 vminps %xmm9, %xmm10, %xmm10 vmaxps %xmm9, %xmm7, %xmm7 vaddps %xmm2, %xmm1, %xmm1 vaddps %xmm1, %xmm5, %xmm2 vminps %xmm2, %xmm10, %xmm5 vmaxps %xmm2, %xmm7, %xmm2 vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmulps %xmm4, %xmm0, %xmm0 vaddps %xmm0, %xmm8, %xmm4 vminps %xmm4, %xmm5, %xmm5 vmaxps %xmm4, %xmm2, %xmm2 vaddps %xmm6, %xmm0, %xmm4 vminps %xmm4, %xmm5, %xmm5 vmaxps %xmm4, %xmm2, %xmm2 vaddps %xmm3, %xmm0, %xmm3 vminps %xmm3, %xmm5, %xmm4 vmaxps %xmm3, %xmm2, %xmm2 vaddps %xmm1, %xmm0, %xmm0 vminps %xmm0, %xmm4, %xmm1 vmovaps %xmm1, (%rdx) vmaxps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, (%rcx) retq	/embree[P]embree/kernels/builders/../common/../../common/math/quaternion.h
unsigned long embree::avx::createMortonCodeArray<embree::Instance>(embree::Instance*, embree::vector_t<embree::avx::BVHBuilderMorton::BuildPrim, embree::aligned_monitored_allocator<embree::avx::BVHBuilderMorton::BuildPrim, 8ul>>&, embree::BuildProgressMonitor&)::'lambda0'(embree::range<unsigned long> const&, unsigned long)::operator()(embree::range<unsigned long> const&, unsigned long) const (.cold.1)	__forceinline QuaternionT( const T& r, const T& i, const T& j, const T& k ) : r(r), i(i), j(j), k(k) {}	vmovss 0x3c(%rdi), %xmm3 vmovss 0xc(%rdi), %xmm6 vmovss 0x1c(%rdi), %xmm4 vmovss 0x2c(%rdi), %xmm2 vmovaps (%rdi), %xmm1 vxorps %xmm0, %xmm0, %xmm0 vshufps $0xe9, %xmm0, %xmm1, %xmm1 # xmm1 = xmm1[1,2],xmm0[2,3] vblendps $0x4, 0x10(%rdi), %xmm1, %xmm1 # xmm1 = xmm1[0,1],mem[2],xmm1[3] vmulss %xmm6, %xmm6, %xmm8 vmulss %xmm3, %xmm3, %xmm9 vaddss %xmm8, %xmm9, %xmm5 vbroadcastss 0x1d1f7be(%rip), %xmm10 # 0x1f20ec0 vxorps %xmm4, %xmm10, %xmm7 vmulss %xmm4, %xmm7, %xmm7 vaddss %xmm5, %xmm7, %xmm5 vxorps %xmm2, %xmm10, %xmm10 vmulss %xmm2, %xmm10, %xmm10 vaddss %xmm5, %xmm10, %xmm11 vmulss %xmm2, %xmm3, %xmm5 vmulss %xmm4, %xmm6, %xmm12 vaddss %xmm5, %xmm12, %xmm13 vsubss %xmm5, %xmm12, %xmm5 vmulss %xmm2, %xmm6, %xmm12 vsubss %xmm8, %xmm9, %xmm9 vmulss %xmm4, %xmm4, %xmm8 vaddss %xmm9, %xmm8, %xmm8 vaddss %xmm8, %xmm10, %xmm8 vmulss %xmm4, %xmm3, %xmm10 vmulss %xmm6, %xmm3, %xmm3 vsubss %xmm10, %xmm12, %xmm6 vmulss %xmm2, %xmm4, %xmm4 vaddss %xmm10, %xmm12, %xmm10 vaddss %xmm3, %xmm4, %xmm12 vsubss %xmm3, %xmm4, %xmm3 vaddss %xmm13, %xmm13, %xmm4 vaddss %xmm6, %xmm6, %xmm6 vaddss %xmm7, %xmm9, %xmm7 vmulss %xmm2, %xmm2, %xmm2 vaddss %xmm7, %xmm2, %xmm7 vshufps $0x0, %xmm11, %xmm11, %xmm2 # xmm2 = xmm11[0,0,0,0] vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vshufps $0x0, %xmm6, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmovaps 0x1ceaf74(%rip), %xmm9 # 0x1eec700 vmulps %xmm6, %xmm9, %xmm6 vmovsd 0x1ceaf58(%rip), %xmm11 # 0x1eec6f0 vmulps %xmm4, %xmm11, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vmovss 0x1ceaf6c(%rip), %xmm6 # 0x1eec714 vmulps %xmm6, %xmm2, %xmm2 vaddps %xmm4, %xmm2, %xmm2 vaddss %xmm5, %xmm5, %xmm4 vaddss %xmm12, %xmm12, %xmm5 vaddss %xmm10, %xmm10, %xmm10 vaddss %xmm3, %xmm3, %xmm12 vshufps $0x0, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[0,0,0,0] vshufps $0x0, %xmm8, %xmm8, %xmm4 # xmm4 = xmm8[0,0,0,0] vshufps $0x0, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vmulps %xmm5, %xmm9, %xmm5 vmulps %xmm4, %xmm11, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vshufps $0x0, %xmm10, %xmm10, %xmm4 # xmm4 = xmm10[0,0,0,0] vshufps $0x0, %xmm12, %xmm12, %xmm5 # xmm5 = xmm12[0,0,0,0] vshufps $0x0, %xmm7, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmulps %xmm7, %xmm9, %xmm7 vmulps %xmm5, %xmm11, %xmm5 vaddps %xmm7, %xmm5, %xmm5 vmulps %xmm6, %xmm4, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vaddps %xmm0, %xmm1, %xmm5 vbroadcastss (%rdi), %xmm1 vmulps %xmm0, %xmm4, %xmm6 vmulps %xmm0, %xmm3, %xmm0 vaddps %xmm6, %xmm0, %xmm0 vmulps %xmm2, %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x10(%rdi), %xmm1 vbroadcastss 0x14(%rdi), %xmm7 vmulps %xmm3, %xmm7, %xmm7 vaddps %xmm6, %xmm7, %xmm6 vmulps %xmm2, %xmm1, %xmm1 vaddps %xmm6, %xmm1, %xmm1 vbroadcastss 0x20(%rdi), %xmm6 vbroadcastss 0x24(%rdi), %xmm7 vbroadcastss 0x28(%rdi), %xmm8 vmulps %xmm4, %xmm8, %xmm8 vmulps %xmm3, %xmm7, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vmulps %xmm2, %xmm6, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0x30(%rdi), %xmm7 vbroadcastss 0x34(%rdi), %xmm8 vbroadcastss 0x38(%rdi), %xmm9 vmulps %xmm4, %xmm9, %xmm4 vmulps %xmm3, %xmm8, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vmulps %xmm2, %xmm7, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vaddps %xmm2, %xmm5, %xmm2 movq 0x58(%rsi), %rax vmovaps 0x10(%rax), %xmm3 vmovaps 0x20(%rax), %xmm4 vminps 0x30(%rax), %xmm3, %xmm3 vmaxps 0x40(%rax), %xmm4, %xmm4 vshufps $0x0, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[0,0,0,0] vshufps $0x55, %xmm3, %xmm3, %xmm7 # xmm7 = xmm3[1,1,1,1] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm2, %xmm3, %xmm3 vmulps %xmm1, %xmm7, %xmm7 vaddps %xmm3, %xmm7, %xmm8 vmulps %xmm0, %xmm5, %xmm5 vaddps %xmm5, %xmm8, %xmm9 vbroadcastss 0x1cea144(%rip), %xmm10 # 0x1eeba20 vminps %xmm9, %xmm10, %xmm10 vbroadcastss 0x1ceb29a(%rip), %xmm11 # 0x1eecb84 vmaxps %xmm9, %xmm11, %xmm9 vshufps $0xaa, %xmm4, %xmm4, %xmm11 # xmm11 = xmm4[2,2,2,2] vmulps %xmm6, %xmm11, %xmm6 vaddps %xmm6, %xmm2, %xmm2 vaddps %xmm2, %xmm7, %xmm6 vaddps %xmm6, %xmm5, %xmm7 vminps %xmm7, %xmm10, %xmm10 vmaxps %xmm7, %xmm9, %xmm7 vshufps $0x55, %xmm4, %xmm4, %xmm9 # xmm9 = xmm4[1,1,1,1] vmulps %xmm1, %xmm9, %xmm1 vaddps %xmm1, %xmm3, %xmm3 vaddps %xmm3, %xmm5, %xmm9 vminps %xmm9, %xmm10, %xmm10 vmaxps %xmm9, %xmm7, %xmm7 vaddps %xmm2, %xmm1, %xmm1 vaddps %xmm1, %xmm5, %xmm2 vminps %xmm2, %xmm10, %xmm5 vmaxps %xmm2, %xmm7, %xmm2 vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmulps %xmm4, %xmm0, %xmm0 vaddps %xmm0, %xmm8, %xmm4 vminps %xmm4, %xmm5, %xmm5 vmaxps %xmm4, %xmm2, %xmm2 vaddps %xmm6, %xmm0, %xmm4 vminps %xmm4, %xmm5, %xmm5 vmaxps %xmm4, %xmm2, %xmm2 vaddps %xmm3, %xmm0, %xmm3 vminps %xmm3, %xmm5, %xmm4 vmaxps %xmm3, %xmm2, %xmm2 vaddps %xmm1, %xmm0, %xmm0 vminps %xmm0, %xmm4, %xmm1 vmovaps %xmm1, (%rdx) vmaxps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, (%rcx) retq nop	/embree[P]embree/kernels/builders/../common/../../common/math/quaternion.h
embree::avx::BVHNIntersector1<8, 257, false, embree::avx::VirtualCurveIntersector1>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1388, %rsp # imm = 0x1388 movq %rdx, 0x18(%rsp) movq %rdi, 0x10(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x20210c vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x20210c vmovaps 0x10(%rsi), %xmm2 vdpps $0x7f, %xmm2, %xmm2, %xmm5 vrsqrtss %xmm5, %xmm5, %xmm6 vmovss 0x1cead4b(%rip), %xmm3 # 0x1eec718 vmulss %xmm3, %xmm6, %xmm7 vmovss 0x1ceb1a7(%rip), %xmm4 # 0x1eecb80 vmulss %xmm4, %xmm5, %xmm5 vmulss %xmm6, %xmm5, %xmm5 vmulss %xmm6, %xmm6, %xmm6 vmulss %xmm6, %xmm5, %xmm5 vsubss %xmm5, %xmm7, %xmm8 vbroadcastss 0x10(%rsi), %ymm5 vmovups %ymm5, 0x180(%rsp) vshufps $0x0, %xmm8, %xmm8, %xmm5 # xmm5 = xmm8[0,0,0,0] vmulps %xmm5, %xmm2, %xmm6 vshufpd $0x1, %xmm6, %xmm6, %xmm9 # xmm9 = xmm6[1,0] vmovshdup %xmm6, %xmm7 # xmm7 = xmm6[1,1,3,3] vbroadcastss 0x1d1f4a8(%rip), %xmm10 # 0x1f20ec0 vxorps %xmm7, %xmm10, %xmm11 vxorps %xmm7, %xmm7, %xmm7 vunpckhps %xmm7, %xmm6, %xmm12 # xmm12 = xmm6[2],xmm7[2],xmm6[3],xmm7[3] vmovss %xmm11, %xmm7, %xmm11 # xmm11 = xmm11[0],xmm7[1,2,3] vshufps $0x41, %xmm11, %xmm12, %xmm11 # xmm11 = xmm12[1,0],xmm11[0,1] vxorpd %xmm10, %xmm9, %xmm9 vinsertps $0x2a, %xmm6, %xmm9, %xmm9 # xmm9 = xmm9[0],zero,xmm6[0],zero vdpps $0x7f, %xmm11, %xmm11, %xmm10 vdpps $0x7f, %xmm9, %xmm9, %xmm12 vcmpltps %xmm10, %xmm12, %xmm10 vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vblendvps %xmm10, %xmm11, %xmm9, %xmm9 vdpps $0x7f, %xmm9, %xmm9, %xmm10 vbroadcastss 0x14(%rsi), %ymm11 vmovups %ymm11, 0x160(%rsp) vrsqrtss %xmm10, %xmm10, %xmm11 vmulss %xmm4, %xmm10, %xmm10 vmulss %xmm11, %xmm10, %xmm10 vmulss %xmm11, %xmm11, %xmm12 vmulss %xmm12, %xmm10, %xmm10 vbroadcastss 0x18(%rsi), %ymm12 vmovups %ymm12, 0x140(%rsp) vmulss %xmm3, %xmm11, %xmm11 vsubss %xmm10, %xmm11, %xmm10 vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vmulps %xmm10, %xmm9, %xmm9 vshufps $0xc9, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,2,0,3] vshufps $0xc9, %xmm6, %xmm6, %xmm11 # xmm11 = xmm6[1,2,0,3] vmulps %xmm9, %xmm11, %xmm11 vmulps %xmm6, %xmm10, %xmm10 vsubps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[1,2,0,3] vdpps $0x7f, %xmm10, %xmm10, %xmm11 leaq 0x1a0(%rsp), %rcx vmovss %xmm8, (%rcx) vrsqrtss %xmm11, %xmm11, %xmm8 vmulss %xmm3, %xmm8, %xmm3 vmulss %xmm4, %xmm11, %xmm4 vmulss %xmm4, %xmm8, %xmm4 vmulss %xmm8, %xmm8, %xmm8 vmulss %xmm4, %xmm8, %xmm4 vsubss %xmm4, %xmm3, %xmm3 vshufps $0x0, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmulps %xmm3, %xmm10, %xmm3 vmulps %xmm6, %xmm5, %xmm4 vunpcklps %xmm4, %xmm9, %xmm5 # xmm5 = xmm9[0],xmm4[0],xmm9[1],xmm4[1] vunpckhps %xmm4, %xmm9, %xmm4 # xmm4 = xmm9[2],xmm4[2],xmm9[3],xmm4[3] vunpcklps %xmm7, %xmm3, %xmm6 # xmm6 = xmm3[0],xmm7[0],xmm3[1],xmm7[1] vunpckhps %xmm7, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm7[2],xmm3[3],xmm7[3] vunpcklps %xmm3, %xmm4, %xmm3 # xmm3 = xmm4[0],xmm3[0],xmm4[1],xmm3[1] vunpcklps %xmm6, %xmm5, %xmm4 # xmm4 = xmm5[0],xmm6[0],xmm5[1],xmm6[1] vunpckhps %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm6[2],xmm5[3],xmm6[3] vmovaps %xmm4, 0x10(%rcx) vmovaps %xmm5, 0x20(%rcx) vmovaps %xmm3, 0x30(%rcx) leaq 0x1e8(%rsp), %r9 movq 0x70(%rax), %rax vmaxss 0xc(%rsi), %xmm1, %xmm3 vbroadcastss 0x1d1f37b(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm2, %xmm4 vbroadcastss 0x1cef492(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm2, %xmm2 movq %rax, -0x8(%r9) vrcpps %xmm2, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1ceab9e(%rip), %xmm5 # 0x1eec714 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vbroadcastss (%rsi), %ymm15 vbroadcastss 0x4(%rsi), %ymm6 vbroadcastss 0x8(%rsi), %ymm14 vaddps %xmm2, %xmm4, %xmm2 xorl %r10d, %r10d vucomiss %xmm1, %xmm2 setb %r10b vshufps $0x0, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm7 vmovshdup %xmm2, %xmm4 # xmm4 = xmm2[1,1,3,3] vshufps $0x55, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm8 vshufpd $0x1, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[1,0] vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] shll $0x5, %r10d xorl %r14d, %r14d vucomiss %xmm1, %xmm4 vinsertf128 $0x1, %xmm2, %ymm2, %ymm4 setb %r14b shll $0x5, %r14d orq $0x40, %r14 xorl %r13d, %r13d vucomiss %xmm1, %xmm5 setb %r13b shll $0x5, %r13d orq $0x80, %r13 movq %r10, %rbp xorq $0x20, %rbp movq %r14, %r12 xorq $0x20, %r12 movq %r13, %rbx xorq $0x20, %rbx vshufps $0x0, %xmm3, %xmm3, %xmm1 # xmm1 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm5 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm9 leaq 0x1e0(%rsp), %r11 vmovups %ymm15, 0x120(%rsp) vmovups %ymm6, 0x20(%rsp) vmovups %ymm14, 0x100(%rsp) vmovups %ymm7, 0xe0(%rsp) vmovups %ymm8, 0xc0(%rsp) vmovups %ymm4, 0xa0(%rsp) vmovups %ymm5, 0x80(%rsp) vmovups %ymm9, 0x60(%rsp) movq %rsi, 0x8(%rsp) movq %r10, (%rsp) cmpq %r11, %r9 je 0x20210c movq -0x8(%r9), %rcx addq $-0x8, %r9 movq %rcx, %rax andq $0xf, %rax jne 0x201d56 vmovaps 0x40(%rcx,%r10), %ymm0 vsubps %ymm15, %ymm0, %ymm0 vmulps %ymm0, %ymm7, %ymm0 vmovaps 0x40(%rcx,%r14), %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm8, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rcx,%r13), %ymm1 vsubps %ymm14, %ymm1, %ymm1 vmulps %ymm1, %ymm4, %ymm1 vmovaps 0x40(%rcx,%rbp), %ymm2 vsubps %ymm15, %ymm2, %ymm2 vmulps %ymm2, %ymm7, %ymm2 vmovaps 0x40(%rcx,%r12), %ymm3 vsubps %ymm6, %ymm3, %ymm3 vmulps %ymm3, %ymm8, %ymm3 vminps %ymm3, %ymm2, %ymm2 vmovaps 0x40(%rcx,%rbx), %ymm3 vsubps %ymm14, %ymm3, %ymm3 vmulps %ymm3, %ymm4, %ymm3 vmaxps %ymm5, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vminps %ymm9, %ymm3, %ymm1 vminps %ymm1, %ymm2, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r15d movb $0x1, %al testb %al, %al je 0x201d5f testq %r15, %r15 je 0x201d63 andq $-0x10, %rcx bsfq %r15, %rax leaq -0x1(%r15), %rdi xorl %edx, %edx movq (%rcx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r15, %rdi jne 0x201d68 movq %rax, %rcx testl %edx, %edx je 0x201c84 jmp 0x202065 cmpl $0x2, %eax je 0x201daf xorl %eax, %eax jmp 0x201d11 pushq $0x6 jmp 0x201d65 pushq $0x4 popq %rdx jmp 0x201d49 movq %rax, (%r9) addq $0x8, %r9 bsfq %rdi, %r8 leaq -0x1(%rdi), %rax movq (%rcx,%r8,8), %r8 prefetcht0 (%r8) prefetcht0 0x40(%r8) prefetcht0 0x80(%r8) prefetcht0 0xc0(%r8) andq %rdi, %rax je 0x201daa movq %r8, (%r9) addq $0x8, %r9 bsfq %rax, %r8 leaq -0x1(%rax), %rdi jmp 0x201d77 movq %r8, %rcx jmp 0x201d49 movq %rcx, %rax andq $-0x10, %rax vmovaps 0x100(%rax), %ymm5 vmovups 0x140(%rsp), %ymm0 vmulps %ymm5, %ymm0, %ymm7 vmovaps 0x120(%rax), %ymm4 vmulps %ymm4, %ymm0, %ymm8 vmovaps 0x140(%rax), %ymm3 vmulps %ymm3, %ymm0, %ymm9 vmovaps 0x40(%rax), %ymm2 vmovaps 0x60(%rax), %ymm1 vmovaps 0x80(%rax), %ymm0 vmovups %ymm0, 0x40(%rsp) vmovaps 0xa0(%rax), %ymm6 vmovups 0x160(%rsp), %ymm12 vmulps %ymm6, %ymm12, %ymm10 vaddps %ymm7, %ymm10, %ymm10 vmovaps 0xc0(%rax), %ymm7 vmulps %ymm7, %ymm12, %ymm11 vaddps %ymm11, %ymm8, %ymm11 vmovaps 0xe0(%rax), %ymm8 vmulps %ymm8, %ymm12, %ymm12 vaddps %ymm12, %ymm9, %ymm9 vmovups 0x180(%rsp), %ymm13 vmulps %ymm2, %ymm13, %ymm12 vaddps %ymm12, %ymm10, %ymm10 vmulps %ymm1, %ymm13, %ymm12 vaddps %ymm12, %ymm11, %ymm11 vmulps %ymm0, %ymm13, %ymm12 vaddps %ymm12, %ymm9, %ymm9 vbroadcastss 0x1d1f060(%rip), %ymm13 # 0x1f20ec4 vandps %ymm13, %ymm10, %ymm12 vbroadcastss 0x1cef176(%rip), %ymm0 # 0x1ef0fe8 vcmpltps %ymm0, %ymm12, %ymm12 vblendvps %ymm12, %ymm0, %ymm10, %ymm10 vandps %ymm13, %ymm11, %ymm12 vcmpltps %ymm0, %ymm12, %ymm12 vblendvps %ymm12, %ymm0, %ymm11, %ymm11 vandps %ymm13, %ymm9, %ymm12 vcmpltps %ymm0, %ymm12, %ymm12 vblendvps %ymm12, %ymm0, %ymm9, %ymm12 vrcpps %ymm10, %ymm9 vmulps %ymm10, %ymm9, %ymm10 vbroadcastss 0x1cea864(%rip), %ymm0 # 0x1eec714 vsubps %ymm10, %ymm0, %ymm10 vrcpps %ymm11, %ymm13 vmulps %ymm10, %ymm9, %ymm10 vaddps %ymm10, %ymm9, %ymm9 vmulps %ymm11, %ymm13, %ymm10 vsubps %ymm10, %ymm0, %ymm10 vmulps %ymm10, %ymm13, %ymm10 vrcpps %ymm12, %ymm11 vaddps %ymm10, %ymm13, %ymm10 vmulps %ymm12, %ymm11, %ymm12 vsubps %ymm12, %ymm0, %ymm12 vmulps %ymm12, %ymm11, %ymm12 vaddps %ymm12, %ymm11, %ymm11 vmulps %ymm5, %ymm14, %ymm5 vaddps 0x160(%rax), %ymm5, %ymm5 vbroadcastss 0x1d1efba(%rip), %ymm0 # 0x1f20ec0 vxorps %ymm0, %ymm9, %ymm12 vmulps %ymm4, %ymm14, %ymm4 vaddps 0x180(%rax), %ymm4, %ymm4 vxorps %ymm0, %ymm10, %ymm13 vmulps %ymm3, %ymm14, %ymm3 vaddps 0x1a0(%rax), %ymm3, %ymm3 vxorps %ymm0, %ymm11, %ymm14 vmulps 0x20(%rsp), %ymm6, %ymm6 vaddps %ymm5, %ymm6, %ymm5 vmulps 0x20(%rsp), %ymm7, %ymm6 vaddps %ymm4, %ymm6, %ymm4 vmulps 0x20(%rsp), %ymm8, %ymm6 vaddps %ymm3, %ymm6, %ymm3 vmulps %ymm2, %ymm15, %ymm2 vaddps %ymm5, %ymm2, %ymm2 vmulps %ymm2, %ymm12, %ymm2 vmulps %ymm1, %ymm15, %ymm1 vaddps %ymm4, %ymm1, %ymm1 vmulps %ymm1, %ymm13, %ymm1 vmulps 0x40(%rsp), %ymm15, %ymm0 vaddps %ymm3, %ymm0, %ymm0 vmulps %ymm0, %ymm14, %ymm0 vaddps %ymm2, %ymm9, %ymm4 vaddps %ymm1, %ymm10, %ymm5 vaddps %ymm0, %ymm11, %ymm3 vextractf128 $0x1, %ymm4, %xmm6 vextractf128 $0x1, %ymm2, %xmm7 vpminsd %xmm6, %xmm7, %xmm8 vpminsd %xmm4, %xmm2, %xmm9 vinsertf128 $0x1, %xmm8, %ymm9, %ymm8 vextractf128 $0x1, %ymm5, %xmm9 vextractf128 $0x1, %ymm1, %xmm10 vpminsd %xmm9, %xmm10, %xmm11 vpminsd %xmm5, %xmm1, %xmm12 vinsertf128 $0x1, %xmm11, %ymm12, %ymm11 vextractf128 $0x1, %ymm3, %xmm12 vextractf128 $0x1, %ymm0, %xmm13 vpminsd %xmm12, %xmm13, %xmm14 vpminsd %xmm3, %xmm0, %xmm15 vinsertf128 $0x1, %xmm14, %ymm15, %ymm14 vmovups 0x120(%rsp), %ymm15 vmaxps %ymm14, %ymm11, %ymm11 vmovups 0x100(%rsp), %ymm14 vpmaxsd %xmm6, %xmm7, %xmm6 vmovups 0xe0(%rsp), %ymm7 vpmaxsd %xmm4, %xmm2, %xmm2 vinsertf128 $0x1, %xmm6, %ymm2, %ymm2 vmovups 0x20(%rsp), %ymm6 vpmaxsd %xmm9, %xmm10, %xmm4 vmovups 0x60(%rsp), %ymm9 vpmaxsd %xmm5, %xmm1, %xmm1 vmovups 0x80(%rsp), %ymm5 vinsertf128 $0x1, %xmm4, %ymm1, %ymm1 vpmaxsd %xmm12, %xmm13, %xmm4 vpmaxsd %xmm3, %xmm0, %xmm0 vinsertf128 $0x1, %xmm4, %ymm0, %ymm0 vmovups 0xa0(%rsp), %ymm4 vminps %ymm0, %ymm1, %ymm0 vmaxps %ymm8, %ymm5, %ymm1 vmovups 0xc0(%rsp), %ymm8 vmaxps %ymm11, %ymm1, %ymm1 vminps %ymm2, %ymm9, %ymm2 vminps %ymm0, %ymm2, %ymm0 vcmpleps %ymm0, %ymm1, %ymm0 jmp 0x201d0b cmpl $0x6, %edx jne 0x202103 andq $-0x10, %rcx movzbl (%rcx), %eax movq 0x10(%rsp), %rdx movq 0x8(%rdx), %r8 shll $0x6, %eax leaq 0x1a0(%rsp), %rdi movq 0x18(%rsp), %rdx movq %r9, 0x40(%rsp) vzeroupper callq *0x8(%r8,%rax) leaq 0x1e0(%rsp), %r11 vmovups 0x60(%rsp), %ymm9 vmovups 0x80(%rsp), %ymm5 vmovups 0xa0(%rsp), %ymm4 vmovups 0xc0(%rsp), %ymm8 vmovups 0xe0(%rsp), %ymm7 movq (%rsp), %r10 vmovups 0x100(%rsp), %ymm14 vmovups 0x20(%rsp), %ymm6 vmovups 0x120(%rsp), %ymm15 movq 0x40(%rsp), %r9 movq 0x8(%rsp), %rsi xorl %edx, %edx testb %al, %al je 0x202103 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rdx cmpl $0x3, %edx jne 0x201c73 addq $0x1388, %rsp # imm = 0x1388 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16781328, false, embree::avx::VirtualCurveIntersector1>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1408, %rsp # imm = 0x1408 movq %rdx, 0x18(%rsp) movq %rdi, 0x10(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x202a31 vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x202a31 vxorps %xmm12, %xmm12, %xmm12 vmovaps 0x10(%rsi), %xmm1 vdpps $0x7f, %xmm1, %xmm1, %xmm4 vrsqrtss %xmm4, %xmm4, %xmm5 vmovss 0x1cea59e(%rip), %xmm2 # 0x1eec718 vmulss %xmm2, %xmm5, %xmm6 vmovss 0x1cea9fa(%rip), %xmm3 # 0x1eecb80 vmulss %xmm3, %xmm4, %xmm4 vmulss %xmm5, %xmm4, %xmm4 vmulss %xmm5, %xmm5, %xmm5 vmulss %xmm5, %xmm4, %xmm4 vsubss %xmm4, %xmm6, %xmm7 vbroadcastss 0x10(%rsi), %ymm4 vmovups %ymm4, 0x200(%rsp) vbroadcastss 0x14(%rsi), %ymm4 vmovups %ymm4, 0x1e0(%rsp) vbroadcastss 0x18(%rsi), %ymm4 vmovups %ymm4, 0x1c0(%rsp) vshufps $0x0, %xmm7, %xmm7, %xmm4 # xmm4 = xmm7[0,0,0,0] vmulps %xmm4, %xmm1, %xmm5 vshufpd $0x1, %xmm5, %xmm5, %xmm8 # xmm8 = xmm5[1,0] vmovshdup %xmm5, %xmm6 # xmm6 = xmm5[1,1,3,3] vbroadcastss 0x1d1ecde(%rip), %xmm9 # 0x1f20ec0 vxorps %xmm6, %xmm9, %xmm10 vxorps %xmm6, %xmm6, %xmm6 vunpckhps %xmm6, %xmm5, %xmm11 # xmm11 = xmm5[2],xmm6[2],xmm5[3],xmm6[3] vmovss %xmm10, %xmm6, %xmm10 # xmm10 = xmm10[0],xmm6[1,2,3] vshufps $0x41, %xmm10, %xmm11, %xmm10 # xmm10 = xmm11[1,0],xmm10[0,1] vxorpd %xmm9, %xmm8, %xmm8 vinsertps $0x2a, %xmm5, %xmm8, %xmm8 # xmm8 = xmm8[0],zero,xmm5[0],zero vdpps $0x7f, %xmm10, %xmm10, %xmm9 vdpps $0x7f, %xmm8, %xmm8, %xmm11 vcmpltps %xmm9, %xmm11, %xmm9 vshufps $0x0, %xmm9, %xmm9, %xmm9 # xmm9 = xmm9[0,0,0,0] vblendvps %xmm9, %xmm10, %xmm8, %xmm8 vdpps $0x7f, %xmm8, %xmm8, %xmm9 vrsqrtss %xmm9, %xmm9, %xmm10 vmulss %xmm2, %xmm10, %xmm11 vmulss %xmm3, %xmm9, %xmm9 vmulss %xmm10, %xmm9, %xmm9 vmulss %xmm10, %xmm10, %xmm10 vmulss %xmm10, %xmm9, %xmm9 vsubss %xmm9, %xmm11, %xmm9 vshufps $0x0, %xmm9, %xmm9, %xmm9 # xmm9 = xmm9[0,0,0,0] vmulps %xmm9, %xmm8, %xmm8 vshufps $0xc9, %xmm8, %xmm8, %xmm9 # xmm9 = xmm8[1,2,0,3] vshufps $0xc9, %xmm5, %xmm5, %xmm10 # xmm10 = xmm5[1,2,0,3] vmulps %xmm8, %xmm10, %xmm10 vmulps %xmm5, %xmm9, %xmm9 vsubps %xmm10, %xmm9, %xmm9 vshufps $0xc9, %xmm9, %xmm9, %xmm9 # xmm9 = xmm9[1,2,0,3] vdpps $0x7f, %xmm9, %xmm9, %xmm10 leaq 0x220(%rsp), %rcx vmovss %xmm7, (%rcx) vrsqrtss %xmm10, %xmm10, %xmm7 vmulss %xmm2, %xmm7, %xmm2 vmulss %xmm3, %xmm10, %xmm3 vmulss %xmm7, %xmm3, %xmm3 vmulss %xmm7, %xmm7, %xmm7 vmulss %xmm7, %xmm3, %xmm3 vsubss %xmm3, %xmm2, %xmm2 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmulps %xmm2, %xmm9, %xmm2 vmulps %xmm5, %xmm4, %xmm3 vunpcklps %xmm3, %xmm8, %xmm4 # xmm4 = xmm8[0],xmm3[0],xmm8[1],xmm3[1] vunpckhps %xmm3, %xmm8, %xmm3 # xmm3 = xmm8[2],xmm3[2],xmm8[3],xmm3[3] vunpcklps %xmm6, %xmm2, %xmm5 # xmm5 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] vunpckhps %xmm6, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm6[2],xmm2[3],xmm6[3] vunpcklps %xmm2, %xmm3, %xmm2 # xmm2 = xmm3[0],xmm2[0],xmm3[1],xmm2[1] vunpcklps %xmm5, %xmm4, %xmm3 # xmm3 = xmm4[0],xmm5[0],xmm4[1],xmm5[1] vunpckhps %xmm5, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm5[2],xmm4[3],xmm5[3] vmovaps %xmm3, 0x10(%rcx) vmovaps %xmm4, 0x20(%rcx) vmovaps %xmm2, 0x30(%rcx) leaq 0x268(%rsp), %r9 vmaxss 0xc(%rsi), %xmm12, %xmm2 vbroadcastss 0x1d1ebd4(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm1, %xmm3 vbroadcastss 0x1ceeceb(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vblendvps %xmm3, %xmm4, %xmm1, %xmm1 movq 0x70(%rax), %rax vrcpps %xmm1, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vbroadcastss 0x1cea3f7(%rip), %xmm4 # 0x1eec714 vsubps %xmm1, %xmm4, %xmm1 vmulps %xmm1, %xmm3, %xmm1 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vbroadcastss 0x8(%rsi), %ymm8 vaddps %xmm1, %xmm3, %xmm1 xorl %r10d, %r10d vucomiss %xmm12, %xmm1 movq %rax, -0x8(%r9) setb %r10b vshufps $0x0, %xmm1, %xmm1, %xmm3 # xmm3 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm9 vmovshdup %xmm1, %xmm3 # xmm3 = xmm1[1,1,3,3] vshufps $0x55, %xmm1, %xmm1, %xmm4 # xmm4 = xmm1[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm10 vshufpd $0x1, %xmm1, %xmm1, %xmm4 # xmm4 = xmm1[1,0] vshufps $0xaa, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vinsertf128 $0x1, %xmm1, %ymm1, %ymm11 shll $0x5, %r10d xorl %r13d, %r13d vucomiss %xmm12, %xmm3 setb %r13b shll $0x5, %r13d orq $0x40, %r13 xorl %r14d, %r14d vucomiss %xmm12, %xmm4 setb %r14b shll $0x5, %r14d orq $0x80, %r14 movq %r10, %rbp xorq $0x20, %rbp movq %r13, %r12 xorq $0x20, %r12 movq %r14, %rbx xorq $0x20, %rbx vshufps $0x0, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm12 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm13 leaq 0x260(%rsp), %r11 vmovups %ymm6, 0x60(%rsp) vmovups %ymm7, 0x40(%rsp) vmovups %ymm8, 0x20(%rsp) vmovups %ymm9, 0x120(%rsp) vmovups %ymm10, 0x100(%rsp) vmovups %ymm11, 0xe0(%rsp) vmovups %ymm12, 0xc0(%rsp) vmovups %ymm13, 0xa0(%rsp) movq %rsi, 0x8(%rsp) movq %r10, (%rsp) cmpq %r11, %r9 je 0x202a31 movq -0x8(%r9), %rcx addq $-0x8, %r9 testb $0x8, %cl jne 0x202533 vmovss 0x1c(%rsi), %xmm1 movl %ecx, %edx andl $0x7, %edx movq %rcx, %rax andq $-0x10, %rax cmpq $0x3, %rdx je 0x202588 vshufps $0x0, %xmm1, %xmm1, %xmm0 # xmm0 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmulps 0x100(%rax,%r10), %ymm0, %ymm1 vaddps 0x40(%rax,%r10), %ymm1, %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmulps 0x100(%rax,%r13), %ymm0, %ymm2 vaddps 0x40(%rax,%r13), %ymm2, %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm10, %ymm2 vmulps 0x100(%rax,%r14), %ymm0, %ymm3 vaddps 0x40(%rax,%r14), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm11, %ymm3 vmulps 0x100(%rax,%rbp), %ymm0, %ymm4 vmaxps %ymm3, %ymm2, %ymm2 vaddps 0x40(%rax,%rbp), %ymm4, %ymm3 vsubps %ymm6, %ymm3, %ymm3 vmulps 0x100(%rax,%r12), %ymm0, %ymm4 vaddps 0x40(%rax,%r12), %ymm4, %ymm4 vmulps %ymm3, %ymm9, %ymm3 vsubps %ymm7, %ymm4, %ymm4 vmulps 0x100(%rax,%rbx), %ymm0, %ymm5 vmulps %ymm4, %ymm10, %ymm4 vaddps 0x40(%rax,%rbx), %ymm5, %ymm5 vsubps %ymm8, %ymm5, %ymm5 vmulps %ymm5, %ymm11, %ymm5 vminps %ymm5, %ymm4, %ymm4 vmaxps %ymm1, %ymm12, %ymm1 vmaxps %ymm2, %ymm1, %ymm1 vminps %ymm3, %ymm13, %ymm2 vminps %ymm4, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 cmpl $0x6, %edx je 0x202910 vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r15d testb $0x8, %cl jne 0x202581 testq %r15, %r15 je 0x202908 andq $-0x10, %rcx bsfq %r15, %rax leaq -0x1(%r15), %rdi xorl %edx, %edx movq (%rcx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r15, %rdi jne 0x20293d movq %rax, %rcx testl %edx, %edx je 0x20242f jmp 0x202987 pushq $0x6 jmp 0x20290a vmovaps 0x40(%rax), %ymm2 vmovups %ymm2, 0x160(%rsp) vmovaps 0xa0(%rax), %ymm6 vmovaps 0xc0(%rax), %ymm14 vmovaps 0xe0(%rax), %ymm3 vmovups %ymm3, 0x140(%rsp) vmovaps 0x100(%rax), %ymm9 vmovaps 0x120(%rax), %ymm5 vmovaps 0x140(%rax), %ymm7 vshufps $0x0, %xmm1, %xmm1, %xmm4 # xmm4 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm12 vmovss 0x1cea132(%rip), %xmm0 # 0x1eec714 vsubss %xmm1, %xmm0, %xmm13 vmulps 0x1c0(%rax), %ymm12, %ymm1 vmulps 0x1e0(%rax), %ymm12, %ymm4 vmulps 0x200(%rax), %ymm12, %ymm8 vmulss 0x1ce941e(%rip), %xmm13, %xmm10 # 0x1eeba24 vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vinsertf128 $0x1, %xmm10, %ymm10, %ymm10 vaddps %ymm1, %ymm10, %ymm0 vmovups %ymm0, 0x80(%rsp) vaddps %ymm4, %ymm10, %ymm0 vmovups %ymm0, 0x1a0(%rsp) vaddps %ymm8, %ymm10, %ymm0 vmovups %ymm0, 0x180(%rsp) vmovups 0x1c0(%rsp), %ymm0 vmulps %ymm0, %ymm9, %ymm1 vmulps %ymm5, %ymm0, %ymm4 vmulps %ymm7, %ymm0, %ymm10 vmovups 0x1e0(%rsp), %ymm0 vmulps %ymm6, %ymm0, %ymm11 vaddps %ymm1, %ymm11, %ymm1 vmulps %ymm0, %ymm14, %ymm11 vmovaps %ymm14, %ymm8 vaddps %ymm4, %ymm11, %ymm4 vmulps %ymm3, %ymm0, %ymm11 vaddps %ymm10, %ymm11, %ymm14 vmovups 0x200(%rsp), %ymm0 vmulps %ymm2, %ymm0, %ymm10 vaddps %ymm1, %ymm10, %ymm1 vmovaps 0x60(%rax), %ymm10 vmulps %ymm0, %ymm10, %ymm11 vaddps %ymm4, %ymm11, %ymm4 vmovaps 0x80(%rax), %ymm11 vmulps %ymm0, %ymm11, %ymm15 vaddps %ymm14, %ymm15, %ymm14 vbroadcastss 0x1d1e816(%rip), %ymm0 # 0x1f20ec4 vandps %ymm0, %ymm1, %ymm15 vbroadcastss 0x1cee92d(%rip), %ymm2 # 0x1ef0fe8 vcmpltps %ymm2, %ymm15, %ymm15 vblendvps %ymm15, %ymm2, %ymm1, %ymm1 vandps %ymm0, %ymm4, %ymm15 vcmpltps %ymm2, %ymm15, %ymm15 vblendvps %ymm15, %ymm2, %ymm4, %ymm4 vandps %ymm0, %ymm14, %ymm15 vcmpltps %ymm2, %ymm15, %ymm15 vblendvps %ymm15, %ymm2, %ymm14, %ymm3 vshufps $0x0, %xmm13, %xmm13, %xmm13 # xmm13 = xmm13[0,0,0,0] vmulps 0x220(%rax), %ymm12, %ymm14 vmulps 0x240(%rax), %ymm12, %ymm15 vinsertf128 $0x1, %xmm13, %ymm13, %ymm0 vmulps 0x260(%rax), %ymm12, %ymm2 vaddps %ymm0, %ymm14, %ymm12 vaddps %ymm0, %ymm15, %ymm13 vrcpps %ymm1, %ymm15 vaddps %ymm2, %ymm0, %ymm14 vmulps %ymm1, %ymm15, %ymm0 vbroadcastss 0x1ce9fef(%rip), %ymm2 # 0x1eec714 vsubps %ymm0, %ymm2, %ymm0 vmulps %ymm0, %ymm15, %ymm0 vaddps %ymm0, %ymm15, %ymm15 vrcpps %ymm4, %ymm0 vmulps %ymm0, %ymm4, %ymm1 vsubps %ymm1, %ymm2, %ymm1 vmulps %ymm1, %ymm0, %ymm1 vaddps %ymm1, %ymm0, %ymm4 vrcpps %ymm3, %ymm0 vmulps %ymm0, %ymm3, %ymm1 vsubps %ymm1, %ymm2, %ymm1 vmulps %ymm1, %ymm0, %ymm1 vaddps %ymm1, %ymm0, %ymm1 vmulps 0x20(%rsp), %ymm9, %ymm0 vaddps 0x160(%rax), %ymm0, %ymm0 vmulps 0x40(%rsp), %ymm6, %ymm2 vaddps %ymm0, %ymm2, %ymm0 vmulps 0x20(%rsp), %ymm5, %ymm2 vaddps 0x180(%rax), %ymm2, %ymm2 vmulps 0x40(%rsp), %ymm8, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vmulps 0x20(%rsp), %ymm7, %ymm3 vaddps 0x1a0(%rax), %ymm3, %ymm3 vmovups 0x140(%rsp), %ymm5 vmulps 0x40(%rsp), %ymm5, %ymm5 vaddps %ymm3, %ymm5, %ymm3 vmovups 0x160(%rsp), %ymm5 vmulps 0x60(%rsp), %ymm5, %ymm5 vaddps %ymm0, %ymm5, %ymm5 vmulps 0x60(%rsp), %ymm10, %ymm0 vaddps %ymm2, %ymm0, %ymm6 vmulps 0x60(%rsp), %ymm11, %ymm0 vaddps %ymm3, %ymm0, %ymm7 vmovups 0x80(%rsp), %ymm0 vsubps %ymm5, %ymm0, %ymm0 vmovups 0x1a0(%rsp), %ymm2 vsubps %ymm6, %ymm2, %ymm2 vmovups 0x180(%rsp), %ymm3 vsubps %ymm7, %ymm3, %ymm8 vmulps %ymm0, %ymm15, %ymm3 vmulps %ymm4, %ymm2, %ymm2 vmulps %ymm1, %ymm8, %ymm0 vsubps %ymm5, %ymm12, %ymm5 vsubps %ymm6, %ymm13, %ymm6 vsubps %ymm7, %ymm14, %ymm7 vmulps %ymm5, %ymm15, %ymm5 vmulps %ymm4, %ymm6, %ymm4 vmulps %ymm1, %ymm7, %ymm1 vextractf128 $0x1, %ymm5, %xmm6 vextractf128 $0x1, %ymm3, %xmm7 vpminsd %xmm6, %xmm7, %xmm8 vpminsd %xmm5, %xmm3, %xmm9 vinsertf128 $0x1, %xmm8, %ymm9, %ymm8 vextractf128 $0x1, %ymm4, %xmm9 vextractf128 $0x1, %ymm2, %xmm10 vpminsd %xmm9, %xmm10, %xmm11 vpminsd %xmm4, %xmm2, %xmm12 vinsertf128 $0x1, %xmm11, %ymm12, %ymm11 vextractf128 $0x1, %ymm1, %xmm12 vextractf128 $0x1, %ymm0, %xmm13 vpminsd %xmm12, %xmm13, %xmm14 vpminsd %xmm1, %xmm0, %xmm15 vinsertf128 $0x1, %xmm14, %ymm15, %ymm14 vmaxps %ymm14, %ymm11, %ymm11 vpmaxsd %xmm6, %xmm7, %xmm6 vmovups 0x40(%rsp), %ymm7 vpmaxsd %xmm5, %xmm3, %xmm3 vinsertf128 $0x1, %xmm6, %ymm3, %ymm3 vmovups 0x60(%rsp), %ymm6 vpmaxsd %xmm9, %xmm10, %xmm5 vmovups 0x100(%rsp), %ymm10 vmovups 0x120(%rsp), %ymm9 vpmaxsd %xmm4, %xmm2, %xmm2 vinsertf128 $0x1, %xmm5, %ymm2, %ymm2 vpmaxsd %xmm12, %xmm13, %xmm4 vmovups 0xa0(%rsp), %ymm13 vmovups 0xc0(%rsp), %ymm12 vpmaxsd %xmm1, %xmm0, %xmm0 vinsertf128 $0x1, %xmm4, %ymm0, %ymm0 vminps %ymm0, %ymm2, %ymm0 vmaxps %ymm8, %ymm12, %ymm1 vmovups 0x20(%rsp), %ymm8 vmaxps %ymm11, %ymm1, %ymm1 vmovups 0xe0(%rsp), %ymm11 vminps %ymm3, %ymm13, %ymm2 vminps %ymm0, %ymm2, %ymm0 vcmpleps %ymm0, %ymm1, %ymm0 vmovmskps %ymm0, %r15d jmp 0x202533 pushq $0x4 popq %rdx jmp 0x202574 vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x202522 movq %rax, (%r9) addq $0x8, %r9 bsfq %rdi, %r8 leaq -0x1(%rdi), %rax movq (%rcx,%r8,8), %r8 prefetcht0 (%r8) prefetcht0 0x40(%r8) prefetcht0 0x80(%r8) prefetcht0 0xc0(%r8) andq %rdi, %rax je 0x20297f movq %r8, (%r9) addq $0x8, %r9 bsfq %rax, %r8 leaq -0x1(%rax), %rdi jmp 0x20294c movq %r8, %rcx jmp 0x202574 cmpl $0x6, %edx jne 0x202a28 andq $-0x10, %rcx movzbl (%rcx), %eax movq 0x10(%rsp), %rdx movq 0x8(%rdx), %r8 shll $0x6, %eax leaq 0x220(%rsp), %rdi movq 0x18(%rsp), %rdx movq %r9, 0x80(%rsp) vzeroupper callq *0x8(%r8,%rax) leaq 0x260(%rsp), %r11 vmovups 0xa0(%rsp), %ymm13 vmovups 0xc0(%rsp), %ymm12 vmovups 0xe0(%rsp), %ymm11 vmovups 0x100(%rsp), %ymm10 vmovups 0x120(%rsp), %ymm9 movq (%rsp), %r10 vmovups 0x20(%rsp), %ymm8 vmovups 0x40(%rsp), %ymm7 vmovups 0x60(%rsp), %ymm6 movq 0x80(%rsp), %r9 movq 0x8(%rsp), %rsi xorl %edx, %edx testb %al, %al je 0x202a28 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rdx cmpl $0x3, %edx jne 0x20241e addq $0x1408, %rsp # imm = 0x1408 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 257, true, embree::avx::VirtualCurveIntersector1>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x13e8, %rsp # imm = 0x13E8 movq %rdx, 0x18(%rsp) movq %rdi, 0x10(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x20327b vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x20327b vmovaps 0x10(%rsi), %xmm2 vdpps $0x7f, %xmm2, %xmm2, %xmm5 vrsqrtss %xmm5, %xmm5, %xmm6 vmovss 0x1ce9c7f(%rip), %xmm3 # 0x1eec718 vmulss %xmm3, %xmm6, %xmm7 vmovss 0x1cea0db(%rip), %xmm4 # 0x1eecb80 vmulss %xmm4, %xmm5, %xmm5 vmulss %xmm6, %xmm5, %xmm5 vmulss %xmm6, %xmm6, %xmm6 vmulss %xmm6, %xmm5, %xmm5 vsubss %xmm5, %xmm7, %xmm8 vbroadcastss 0x10(%rsi), %ymm5 vmovups %ymm5, 0x1e0(%rsp) vbroadcastss 0x14(%rsi), %ymm5 vmovups %ymm5, 0x1c0(%rsp) vbroadcastss 0x18(%rsi), %ymm5 vmovups %ymm5, 0x1a0(%rsp) vshufps $0x0, %xmm8, %xmm8, %xmm5 # xmm5 = xmm8[0,0,0,0] vmulps %xmm5, %xmm2, %xmm6 vshufpd $0x1, %xmm6, %xmm6, %xmm9 # xmm9 = xmm6[1,0] vmovshdup %xmm6, %xmm7 # xmm7 = xmm6[1,1,3,3] vbroadcastss 0x1d1e3be(%rip), %xmm10 # 0x1f20ec0 vxorps %xmm7, %xmm10, %xmm11 vxorps %xmm7, %xmm7, %xmm7 vunpckhps %xmm7, %xmm6, %xmm12 # xmm12 = xmm6[2],xmm7[2],xmm6[3],xmm7[3] vmovss %xmm11, %xmm7, %xmm11 # xmm11 = xmm11[0],xmm7[1,2,3] vshufps $0x41, %xmm11, %xmm12, %xmm11 # xmm11 = xmm12[1,0],xmm11[0,1] vxorpd %xmm10, %xmm9, %xmm9 vinsertps $0x2a, %xmm6, %xmm9, %xmm9 # xmm9 = xmm9[0],zero,xmm6[0],zero vdpps $0x7f, %xmm11, %xmm11, %xmm10 vdpps $0x7f, %xmm9, %xmm9, %xmm12 vcmpltps %xmm10, %xmm12, %xmm10 vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vblendvps %xmm10, %xmm11, %xmm9, %xmm9 vdpps $0x7f, %xmm9, %xmm9, %xmm10 vrsqrtss %xmm10, %xmm10, %xmm11 vmulss %xmm3, %xmm11, %xmm12 vmulss %xmm4, %xmm10, %xmm10 vmulss %xmm11, %xmm10, %xmm10 vmulss %xmm11, %xmm11, %xmm11 vmulss %xmm11, %xmm10, %xmm10 vsubss %xmm10, %xmm12, %xmm10 vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vmulps %xmm10, %xmm9, %xmm9 vshufps $0xc9, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,2,0,3] vshufps $0xc9, %xmm6, %xmm6, %xmm11 # xmm11 = xmm6[1,2,0,3] vmulps %xmm9, %xmm11, %xmm11 vmulps %xmm6, %xmm10, %xmm10 vsubps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[1,2,0,3] vdpps $0x7f, %xmm10, %xmm10, %xmm11 leaq 0x200(%rsp), %rcx vmovss %xmm8, (%rcx) vrsqrtss %xmm11, %xmm11, %xmm8 vmulss %xmm3, %xmm8, %xmm3 vmulss %xmm4, %xmm11, %xmm4 vmulss %xmm4, %xmm8, %xmm4 vmulss %xmm8, %xmm8, %xmm8 vmulss %xmm4, %xmm8, %xmm4 vsubss %xmm4, %xmm3, %xmm3 vshufps $0x0, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmulps %xmm3, %xmm10, %xmm3 vmulps %xmm6, %xmm5, %xmm4 vunpcklps %xmm4, %xmm9, %xmm5 # xmm5 = xmm9[0],xmm4[0],xmm9[1],xmm4[1] vunpckhps %xmm4, %xmm9, %xmm4 # xmm4 = xmm9[2],xmm4[2],xmm9[3],xmm4[3] vunpcklps %xmm7, %xmm3, %xmm6 # xmm6 = xmm3[0],xmm7[0],xmm3[1],xmm7[1] vunpckhps %xmm7, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm7[2],xmm3[3],xmm7[3] vunpcklps %xmm3, %xmm4, %xmm3 # xmm3 = xmm4[0],xmm3[0],xmm4[1],xmm3[1] vunpcklps %xmm6, %xmm5, %xmm4 # xmm4 = xmm5[0],xmm6[0],xmm5[1],xmm6[1] vunpckhps %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm6[2],xmm5[3],xmm6[3] vmovaps %xmm4, 0x10(%rcx) vmovaps %xmm5, 0x20(%rcx) vmovaps %xmm3, 0x30(%rcx) leaq 0x248(%rsp), %r9 movq 0x70(%rax), %rax vmaxss 0xc(%rsi), %xmm1, %xmm3 vbroadcastss 0x1d1e2af(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm2, %xmm4 vbroadcastss 0x1cee3c6(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vbroadcastss 0x1ce9ae4(%rip), %xmm5 # 0x1eec714 vdivps %xmm2, %xmm5, %xmm2 vbroadcastss 0x1d1e323(%rip), %xmm5 # 0x1f20f60 vblendvps %xmm4, %xmm5, %xmm2, %xmm2 movq %rax, -0x8(%r9) vbroadcastss 0x1d1d2c0(%rip), %xmm4 # 0x1f1ff10 vmulps %xmm4, %xmm2, %xmm5 vbroadcastss 0x1d1d2b7(%rip), %xmm4 # 0x1f1ff14 vbroadcastss (%rsi), %ymm15 vbroadcastss 0x4(%rsi), %ymm7 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss 0x8(%rsi), %ymm14 xorl %r10d, %r10d vucomiss %xmm1, %xmm5 setb %r10b vshufps $0x0, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm8 vmovshdup %xmm5, %xmm6 # xmm6 = xmm5[1,1,3,3] vshufps $0x55, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vshufpd $0x1, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[1,0] vshufps $0xaa, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[2,2,2,2] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufps $0x0, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[0,0,0,0] shll $0x5, %r10d xorl %r14d, %r14d vucomiss %xmm1, %xmm6 vinsertf128 $0x1, %xmm5, %ymm5, %ymm6 vshufps $0x55, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm5 vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vinsertf128 $0x1, %xmm2, %ymm2, %ymm11 setb %r14b shll $0x5, %r14d orq $0x40, %r14 xorl %r13d, %r13d vucomiss %xmm1, %xmm4 setb %r13b shll $0x5, %r13d orq $0x80, %r13 movq %r10, %rbp xorq $0x20, %rbp movq %r14, %r12 xorq $0x20, %r12 movq %r13, %rbx xorq $0x20, %rbx vshufps $0x0, %xmm3, %xmm3, %xmm1 # xmm1 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm4 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm12 leaq 0x240(%rsp), %r11 vmovups %ymm15, 0x180(%rsp) vmovups %ymm7, 0x20(%rsp) vmovups %ymm14, 0x40(%rsp) vmovups %ymm8, 0x160(%rsp) vmovups %ymm9, 0x140(%rsp) vmovups %ymm10, 0x120(%rsp) vmovups %ymm6, 0x100(%rsp) vmovups %ymm5, 0xe0(%rsp) vmovups %ymm11, 0xc0(%rsp) vmovups %ymm4, 0xa0(%rsp) vmovups %ymm12, 0x80(%rsp) movq %rsi, 0x8(%rsp) movq %r10, (%rsp) cmpq %r11, %r9 je 0x20327b movq -0x8(%r9), %rcx addq $-0x8, %r9 movq %rcx, %rax andq $0xf, %rax jne 0x202e71 vmovaps 0x40(%rcx,%r10), %ymm0 vsubps %ymm15, %ymm0, %ymm0 vmulps %ymm0, %ymm8, %ymm0 vmovaps 0x40(%rcx,%r14), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rcx,%r13), %ymm1 vsubps %ymm14, %ymm1, %ymm1 vmulps %ymm1, %ymm10, %ymm1 vmovaps 0x40(%rcx,%rbp), %ymm2 vsubps %ymm15, %ymm2, %ymm2 vmulps %ymm2, %ymm6, %ymm2 vmovaps 0x40(%rcx,%r12), %ymm3 vsubps %ymm7, %ymm3, %ymm3 vmulps %ymm3, %ymm5, %ymm3 vminps %ymm3, %ymm2, %ymm2 vmovaps 0x40(%rcx,%rbx), %ymm3 vsubps %ymm14, %ymm3, %ymm3 vmulps %ymm3, %ymm11, %ymm3 vmaxps %ymm4, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vminps %ymm12, %ymm3, %ymm1 vminps %ymm1, %ymm2, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r15d movb $0x1, %al testb %al, %al je 0x202e7a testq %r15, %r15 je 0x202e7e andq $-0x10, %rcx bsfq %r15, %rax leaq -0x1(%r15), %rdi xorl %edx, %edx movq (%rcx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r15, %rdi jne 0x202e83 movq %rax, %rcx testl %edx, %edx je 0x202d9f jmp 0x2031b9 cmpl $0x2, %eax je 0x202eca xorl %eax, %eax jmp 0x202e2c pushq $0x6 jmp 0x202e80 pushq $0x4 popq %rdx jmp 0x202e64 movq %rax, (%r9) addq $0x8, %r9 bsfq %rdi, %r8 leaq -0x1(%rdi), %rax movq (%rcx,%r8,8), %r8 prefetcht0 (%r8) prefetcht0 0x40(%r8) prefetcht0 0x80(%r8) prefetcht0 0xc0(%r8) andq %rdi, %rax je 0x202ec5 movq %r8, (%r9) addq $0x8, %r9 bsfq %rax, %r8 leaq -0x1(%rax), %rdi jmp 0x202e92 movq %r8, %rcx jmp 0x202e64 movq %rcx, %rax andq $-0x10, %rax vmovaps 0x100(%rax), %ymm5 vmovups 0x1a0(%rsp), %ymm0 vmulps %ymm5, %ymm0, %ymm7 vmovaps 0x120(%rax), %ymm3 vmulps %ymm3, %ymm0, %ymm8 vmovaps 0x140(%rax), %ymm4 vmulps %ymm4, %ymm0, %ymm9 vmovaps 0x40(%rax), %ymm2 vmovaps 0x60(%rax), %ymm1 vmovaps 0x80(%rax), %ymm0 vmovups %ymm0, 0x60(%rsp) vmovaps 0xa0(%rax), %ymm6 vmovups 0x1c0(%rsp), %ymm12 vmulps %ymm6, %ymm12, %ymm10 vaddps %ymm7, %ymm10, %ymm10 vmovaps 0xc0(%rax), %ymm7 vmulps %ymm7, %ymm12, %ymm11 vaddps %ymm11, %ymm8, %ymm11 vmovaps 0xe0(%rax), %ymm8 vmulps %ymm8, %ymm12, %ymm12 vaddps %ymm12, %ymm9, %ymm9 vmovups 0x1e0(%rsp), %ymm13 vmulps %ymm2, %ymm13, %ymm12 vaddps %ymm12, %ymm10, %ymm10 vmulps %ymm1, %ymm13, %ymm12 vaddps %ymm12, %ymm11, %ymm11 vmulps %ymm0, %ymm13, %ymm12 vaddps %ymm12, %ymm9, %ymm9 vbroadcastss 0x1d1df45(%rip), %ymm13 # 0x1f20ec4 vandps %ymm13, %ymm10, %ymm12 vbroadcastss 0x1cee05b(%rip), %ymm0 # 0x1ef0fe8 vcmpltps %ymm0, %ymm12, %ymm12 vblendvps %ymm12, %ymm0, %ymm10, %ymm10 vandps %ymm13, %ymm11, %ymm12 vcmpltps %ymm0, %ymm12, %ymm12 vblendvps %ymm12, %ymm0, %ymm11, %ymm11 vandps %ymm13, %ymm9, %ymm12 vcmpltps %ymm0, %ymm12, %ymm12 vblendvps %ymm12, %ymm0, %ymm9, %ymm12 vrcpps %ymm10, %ymm9 vmulps %ymm10, %ymm9, %ymm10 vbroadcastss 0x1ce9749(%rip), %ymm0 # 0x1eec714 vsubps %ymm10, %ymm0, %ymm10 vmulps %ymm10, %ymm9, %ymm10 vaddps %ymm10, %ymm9, %ymm9 vrcpps %ymm11, %ymm10 vmulps %ymm11, %ymm10, %ymm11 vsubps %ymm11, %ymm0, %ymm11 vmulps %ymm11, %ymm10, %ymm11 vaddps %ymm11, %ymm10, %ymm10 vrcpps %ymm12, %ymm11 vmulps %ymm12, %ymm11, %ymm12 vsubps %ymm12, %ymm0, %ymm12 vmulps %ymm12, %ymm11, %ymm12 vaddps %ymm12, %ymm11, %ymm11 vbroadcastss 0x1d1deab(%rip), %ymm0 # 0x1f20ec0 vxorps %ymm0, %ymm9, %ymm12 vxorps %ymm0, %ymm10, %ymm13 vmulps %ymm5, %ymm14, %ymm5 vaddps 0x160(%rax), %ymm5, %ymm5 vxorps %ymm0, %ymm11, %ymm14 vmulps 0x40(%rsp), %ymm3, %ymm3 vaddps 0x180(%rax), %ymm3, %ymm3 vmulps 0x40(%rsp), %ymm4, %ymm4 vaddps 0x1a0(%rax), %ymm4, %ymm4 vmulps 0x20(%rsp), %ymm6, %ymm6 vaddps %ymm5, %ymm6, %ymm5 vmulps 0x20(%rsp), %ymm7, %ymm6 vaddps %ymm3, %ymm6, %ymm3 vmulps 0x20(%rsp), %ymm8, %ymm6 vaddps %ymm4, %ymm6, %ymm4 vmulps %ymm2, %ymm15, %ymm2 vaddps %ymm5, %ymm2, %ymm2 vmulps %ymm2, %ymm12, %ymm2 vmulps %ymm1, %ymm15, %ymm1 vaddps %ymm3, %ymm1, %ymm1 vmulps %ymm1, %ymm13, %ymm1 vmulps 0x60(%rsp), %ymm15, %ymm0 vaddps %ymm4, %ymm0, %ymm0 vmulps %ymm0, %ymm14, %ymm0 vaddps %ymm2, %ymm9, %ymm4 vaddps %ymm1, %ymm10, %ymm5 vaddps %ymm0, %ymm11, %ymm3 vextractf128 $0x1, %ymm4, %xmm6 vextractf128 $0x1, %ymm2, %xmm7 vpminsd %xmm6, %xmm7, %xmm8 vpminsd %xmm4, %xmm2, %xmm9 vinsertf128 $0x1, %xmm8, %ymm9, %ymm8 vextractf128 $0x1, %ymm5, %xmm9 vextractf128 $0x1, %ymm1, %xmm10 vpminsd %xmm9, %xmm10, %xmm11 vpminsd %xmm5, %xmm1, %xmm12 vinsertf128 $0x1, %xmm11, %ymm12, %ymm11 vextractf128 $0x1, %ymm3, %xmm12 vextractf128 $0x1, %ymm0, %xmm13 vpminsd %xmm12, %xmm13, %xmm14 vpminsd %xmm3, %xmm0, %xmm15 vinsertf128 $0x1, %xmm14, %ymm15, %ymm14 vmovups 0x180(%rsp), %ymm15 vmaxps %ymm14, %ymm11, %ymm11 vmovups 0x40(%rsp), %ymm14 vpmaxsd %xmm6, %xmm7, %xmm6 vmovups 0x20(%rsp), %ymm7 vpmaxsd %xmm4, %xmm2, %xmm2 vinsertf128 $0x1, %xmm6, %ymm2, %ymm2 vmovups 0x100(%rsp), %ymm6 vpmaxsd %xmm9, %xmm10, %xmm4 vmovups 0x120(%rsp), %ymm10 vmovups 0x140(%rsp), %ymm9 vpmaxsd %xmm5, %xmm1, %xmm1 vmovups 0xe0(%rsp), %ymm5 vinsertf128 $0x1, %xmm4, %ymm1, %ymm1 vpmaxsd %xmm12, %xmm13, %xmm4 vmovups 0x80(%rsp), %ymm12 vpmaxsd %xmm3, %xmm0, %xmm0 vinsertf128 $0x1, %xmm4, %ymm0, %ymm0 vmovups 0xa0(%rsp), %ymm4 vminps %ymm0, %ymm1, %ymm0 vmaxps %ymm8, %ymm4, %ymm1 vmovups 0x160(%rsp), %ymm8 vmaxps %ymm11, %ymm1, %ymm1 vmovups 0xc0(%rsp), %ymm11 vminps %ymm2, %ymm12, %ymm2 vminps %ymm0, %ymm2, %ymm0 vbroadcastss 0x1d1cd72(%rip), %ymm2 # 0x1f1ff10 vmulps %ymm2, %ymm1, %ymm1 vbroadcastss 0x1d1cd69(%rip), %ymm2 # 0x1f1ff14 vmulps %ymm2, %ymm0, %ymm0 vcmpleps %ymm0, %ymm1, %ymm0 jmp 0x202e26 cmpl $0x6, %edx jne 0x203272 andq $-0x10, %rcx movzbl (%rcx), %eax movq 0x10(%rsp), %rdx movq 0x8(%rdx), %r8 shll $0x6, %eax leaq 0x200(%rsp), %rdi movq 0x18(%rsp), %rdx movq %r9, 0x60(%rsp) vzeroupper callq *0x8(%r8,%rax) leaq 0x240(%rsp), %r11 vmovups 0x80(%rsp), %ymm12 vmovups 0xa0(%rsp), %ymm4 vmovups 0xc0(%rsp), %ymm11 vmovups 0xe0(%rsp), %ymm5 vmovups 0x100(%rsp), %ymm6 vmovups 0x120(%rsp), %ymm10 vmovups 0x140(%rsp), %ymm9 vmovups 0x160(%rsp), %ymm8 movq (%rsp), %r10 vmovups 0x40(%rsp), %ymm14 vmovups 0x20(%rsp), %ymm7 vmovups 0x180(%rsp), %ymm15 movq 0x60(%rsp), %r9 movq 0x8(%rsp), %rsi xorl %edx, %edx testb %al, %al je 0x203272 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rdx cmpl $0x3, %edx jne 0x202d8e addq $0x13e8, %rsp # imm = 0x13E8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16781328, true, embree::avx::VirtualCurveIntersector1>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1468, %rsp # imm = 0x1468 movq %rdx, 0x18(%rsp) movq %rdi, 0x10(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x203c3a vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x203c3a vxorps %xmm12, %xmm12, %xmm12 vmovaps 0x10(%rsi), %xmm1 vdpps $0x7f, %xmm1, %xmm1, %xmm4 vrsqrtss %xmm4, %xmm4, %xmm5 vmovss 0x1ce9430(%rip), %xmm2 # 0x1eec718 vmulss %xmm2, %xmm5, %xmm6 vmovss 0x1ce988c(%rip), %xmm3 # 0x1eecb80 vmulss %xmm3, %xmm4, %xmm4 vmulss %xmm5, %xmm4, %xmm4 vmulss %xmm5, %xmm5, %xmm5 vmulss %xmm5, %xmm4, %xmm4 vbroadcastss 0x10(%rsi), %ymm5 vmovups %ymm5, 0x260(%rsp) vbroadcastss 0x14(%rsi), %ymm5 vmovups %ymm5, 0x240(%rsp) vsubss %xmm4, %xmm6, %xmm7 vbroadcastss 0x18(%rsi), %ymm4 vmovups %ymm4, 0x220(%rsp) leaq 0x280(%rsp), %rcx vshufps $0x0, %xmm7, %xmm7, %xmm4 # xmm4 = xmm7[0,0,0,0] vmulps %xmm4, %xmm1, %xmm5 vshufpd $0x1, %xmm5, %xmm5, %xmm8 # xmm8 = xmm5[1,0] vmovshdup %xmm5, %xmm6 # xmm6 = xmm5[1,1,3,3] vbroadcastss 0x1d1db68(%rip), %xmm9 # 0x1f20ec0 vxorps %xmm6, %xmm9, %xmm10 vxorps %xmm6, %xmm6, %xmm6 vunpckhps %xmm6, %xmm5, %xmm11 # xmm11 = xmm5[2],xmm6[2],xmm5[3],xmm6[3] vmovss %xmm10, %xmm6, %xmm10 # xmm10 = xmm10[0],xmm6[1,2,3] vshufps $0x41, %xmm10, %xmm11, %xmm10 # xmm10 = xmm11[1,0],xmm10[0,1] vxorpd %xmm9, %xmm8, %xmm8 vinsertps $0x2a, %xmm5, %xmm8, %xmm8 # xmm8 = xmm8[0],zero,xmm5[0],zero vdpps $0x7f, %xmm10, %xmm10, %xmm9 vdpps $0x7f, %xmm8, %xmm8, %xmm11 vcmpltps %xmm9, %xmm11, %xmm9 vshufps $0x0, %xmm9, %xmm9, %xmm9 # xmm9 = xmm9[0,0,0,0] vblendvps %xmm9, %xmm10, %xmm8, %xmm8 vdpps $0x7f, %xmm8, %xmm8, %xmm9 vmovss %xmm7, (%rcx) vrsqrtss %xmm9, %xmm9, %xmm7 vmulss %xmm2, %xmm7, %xmm10 vmulss %xmm3, %xmm9, %xmm9 vmulss %xmm7, %xmm9, %xmm9 vmulss %xmm7, %xmm7, %xmm7 vmulss %xmm7, %xmm9, %xmm7 vsubss %xmm7, %xmm10, %xmm7 vshufps $0x0, %xmm7, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmulps %xmm7, %xmm8, %xmm7 vshufps $0xc9, %xmm7, %xmm7, %xmm8 # xmm8 = xmm7[1,2,0,3] vshufps $0xc9, %xmm5, %xmm5, %xmm9 # xmm9 = xmm5[1,2,0,3] vmulps %xmm7, %xmm9, %xmm9 vmulps %xmm5, %xmm8, %xmm8 vsubps %xmm9, %xmm8, %xmm8 vshufps $0xc9, %xmm8, %xmm8, %xmm8 # xmm8 = xmm8[1,2,0,3] vdpps $0x7f, %xmm8, %xmm8, %xmm9 vrsqrtss %xmm9, %xmm9, %xmm10 vmulss %xmm2, %xmm10, %xmm2 vmulss %xmm3, %xmm9, %xmm3 vmulss %xmm3, %xmm10, %xmm3 vmulss %xmm10, %xmm10, %xmm9 vmulss %xmm3, %xmm9, %xmm3 vsubss %xmm3, %xmm2, %xmm2 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmulps %xmm2, %xmm8, %xmm2 vmulps %xmm5, %xmm4, %xmm3 vunpcklps %xmm3, %xmm7, %xmm4 # xmm4 = xmm7[0],xmm3[0],xmm7[1],xmm3[1] vunpckhps %xmm3, %xmm7, %xmm3 # xmm3 = xmm7[2],xmm3[2],xmm7[3],xmm3[3] vunpcklps %xmm6, %xmm2, %xmm5 # xmm5 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] vunpckhps %xmm6, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm6[2],xmm2[3],xmm6[3] vunpcklps %xmm2, %xmm3, %xmm2 # xmm2 = xmm3[0],xmm2[0],xmm3[1],xmm2[1] vunpcklps %xmm5, %xmm4, %xmm3 # xmm3 = xmm4[0],xmm5[0],xmm4[1],xmm5[1] vunpckhps %xmm5, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm5[2],xmm4[3],xmm5[3] vmovaps %xmm3, 0x10(%rcx) vmovaps %xmm4, 0x20(%rcx) vmovaps %xmm2, 0x30(%rcx) leaq 0x2c8(%rsp), %r9 movq 0x70(%rax), %rax vmaxss 0xc(%rsi), %xmm12, %xmm2 movq %rax, -0x8(%r9) vbroadcastss 0x1d1da65(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm1, %xmm3 vbroadcastss 0x1cedb7c(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1ce929a(%rip), %xmm4 # 0x1eec714 vdivps %xmm1, %xmm4, %xmm1 vbroadcastss 0x1d1dad9(%rip), %xmm4 # 0x1f20f60 vblendvps %xmm3, %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d1ca7a(%rip), %xmm3 # 0x1f1ff10 vmulps %xmm3, %xmm1, %xmm3 vbroadcastss 0x1d1ca71(%rip), %xmm4 # 0x1f1ff14 vmulps %xmm4, %xmm1, %xmm1 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vbroadcastss 0x8(%rsi), %ymm8 xorl %r10d, %r10d vucomiss %xmm12, %xmm3 setb %r10b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm11 vshufps $0x0, %xmm1, %xmm1, %xmm3 # xmm3 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm13 vshufps $0x55, %xmm1, %xmm1, %xmm3 # xmm3 = xmm1[1,1,1,1] vinsertf128 $0x1, %xmm3, %ymm3, %ymm14 vshufps $0xaa, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vinsertf128 $0x1, %xmm1, %ymm1, %ymm15 shll $0x5, %r10d xorl %r13d, %r13d vucomiss %xmm12, %xmm4 setb %r13b shll $0x5, %r13d orq $0x40, %r13 xorl %r14d, %r14d vucomiss %xmm12, %xmm5 setb %r14b shll $0x5, %r14d orq $0x80, %r14 movq %r10, %rbp xorq $0x20, %rbp movq %r13, %r12 xorq $0x20, %r12 movq %r14, %rbx xorq $0x20, %rbx vshufps $0x0, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm5 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm12 leaq 0x2c0(%rsp), %r11 vmovups %ymm6, 0x60(%rsp) vmovups %ymm7, 0x40(%rsp) vmovups %ymm8, 0x20(%rsp) vmovups %ymm9, 0x180(%rsp) vmovups %ymm10, 0x160(%rsp) vmovups %ymm11, 0x140(%rsp) vmovups %ymm13, 0x120(%rsp) vmovups %ymm14, 0x100(%rsp) vmovups %ymm15, 0xe0(%rsp) vmovups %ymm5, 0xc0(%rsp) vmovups %ymm12, 0xa0(%rsp) movq %rsi, 0x8(%rsp) movq %r10, (%rsp) cmpq %r11, %r9 je 0x203c3a movq -0x8(%r9), %rcx addq $-0x8, %r9 testb $0x8, %cl jne 0x2036e9 vmovss 0x1c(%rsi), %xmm2 movl %ecx, %edx andl $0x7, %edx movq %rcx, %rax andq $-0x10, %rax cmpq $0x3, %rdx je 0x20373e vshufps $0x0, %xmm2, %xmm2, %xmm0 # xmm0 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmulps 0x100(%rax,%r10), %ymm0, %ymm1 vaddps 0x40(%rax,%r10), %ymm1, %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmulps 0x100(%rax,%r13), %ymm0, %ymm2 vaddps 0x40(%rax,%r13), %ymm2, %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm10, %ymm2 vmulps 0x100(%rax,%r14), %ymm0, %ymm3 vaddps 0x40(%rax,%r14), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm11, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmaxps %ymm1, %ymm5, %ymm1 vmulps 0x100(%rax,%rbp), %ymm0, %ymm3 vaddps 0x40(%rax,%rbp), %ymm3, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vsubps %ymm6, %ymm3, %ymm2 vmulps 0x100(%rax,%r12), %ymm0, %ymm3 vmulps %ymm2, %ymm13, %ymm2 vaddps 0x40(%rax,%r12), %ymm3, %ymm3 vsubps %ymm7, %ymm3, %ymm3 vmulps 0x100(%rax,%rbx), %ymm0, %ymm4 vaddps 0x40(%rax,%rbx), %ymm4, %ymm4 vmulps %ymm3, %ymm14, %ymm3 vsubps %ymm8, %ymm4, %ymm4 vmulps %ymm4, %ymm15, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm2, %ymm12, %ymm2 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 cmpl $0x6, %edx je 0x203afe vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r15d testb $0x8, %cl jne 0x203737 testq %r15, %r15 je 0x203af6 andq $-0x10, %rcx bsfq %r15, %rax leaq -0x1(%r15), %rdi xorl %edx, %edx movq (%rcx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r15, %rdi jne 0x203b2b movq %rax, %rcx testl %edx, %edx je 0x2035e5 jmp 0x203b75 pushq $0x6 jmp 0x203af8 vmovaps 0xa0(%rax), %ymm4 vmovaps 0xc0(%rax), %ymm15 vmovaps 0xe0(%rax), %ymm1 vmovaps 0x100(%rax), %ymm10 vmovaps 0x120(%rax), %ymm6 vmovaps 0x140(%rax), %ymm3 vshufps $0x0, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm5, %ymm5, %ymm7 vmovss 0x1ce8f93(%rip), %xmm0 # 0x1eec714 vsubss %xmm2, %xmm0, %xmm2 vshufps $0x0, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[0,0,0,0] vmulps 0x1c0(%rax), %ymm7, %ymm8 vmulps 0x1e0(%rax), %ymm7, %ymm9 vinsertf128 $0x1, %xmm5, %ymm5, %ymm12 vmulps 0x200(%rax), %ymm7, %ymm11 vmulss 0x1ce8274(%rip), %xmm2, %xmm2 # 0x1eeba24 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vaddps %ymm2, %ymm8, %ymm0 vmovups %ymm0, 0x80(%rsp) vaddps %ymm2, %ymm9, %ymm0 vmovups %ymm0, 0x200(%rsp) vmulps 0x220(%rax), %ymm7, %ymm8 vmulps 0x240(%rax), %ymm7, %ymm9 vaddps %ymm2, %ymm11, %ymm11 vmulps 0x260(%rax), %ymm7, %ymm2 vaddps %ymm8, %ymm12, %ymm0 vmovups %ymm0, 0x1e0(%rsp) vaddps %ymm9, %ymm12, %ymm0 vmovups %ymm0, 0x1c0(%rsp) vaddps %ymm2, %ymm12, %ymm0 vmovups %ymm0, 0x1a0(%rsp) vmovups 0x220(%rsp), %ymm0 vmulps %ymm0, %ymm10, %ymm2 vmulps %ymm6, %ymm0, %ymm12 vmulps %ymm3, %ymm0, %ymm13 vmovups 0x240(%rsp), %ymm0 vmulps %ymm4, %ymm0, %ymm14 vaddps %ymm2, %ymm14, %ymm2 vmulps %ymm0, %ymm15, %ymm14 vmovaps %ymm15, %ymm8 vaddps %ymm12, %ymm14, %ymm14 vmulps %ymm1, %ymm0, %ymm12 vmovaps %ymm1, %ymm9 vaddps %ymm13, %ymm12, %ymm15 vmovaps 0x40(%rax), %ymm12 vmovups 0x260(%rsp), %ymm0 vmulps %ymm0, %ymm12, %ymm13 vaddps %ymm2, %ymm13, %ymm2 vmovaps 0x60(%rax), %ymm13 vmulps %ymm0, %ymm13, %ymm5 vaddps %ymm5, %ymm14, %ymm5 vmovaps 0x80(%rax), %ymm14 vmulps %ymm0, %ymm14, %ymm7 vaddps %ymm7, %ymm15, %ymm7 vbroadcastss 0x1d1d62d(%rip), %ymm0 # 0x1f20ec4 vandps %ymm0, %ymm2, %ymm15 vbroadcastss 0x1ced744(%rip), %ymm1 # 0x1ef0fe8 vcmpltps %ymm1, %ymm15, %ymm15 vblendvps %ymm15, %ymm1, %ymm2, %ymm2 vandps %ymm0, %ymm5, %ymm15 vcmpltps %ymm1, %ymm15, %ymm15 vblendvps %ymm15, %ymm1, %ymm5, %ymm5 vandps %ymm0, %ymm7, %ymm15 vcmpltps %ymm1, %ymm15, %ymm15 vblendvps %ymm15, %ymm1, %ymm7, %ymm7 vrcpps %ymm2, %ymm15 vmulps %ymm2, %ymm15, %ymm2 vbroadcastss 0x1ce8e36(%rip), %ymm1 # 0x1eec714 vsubps %ymm2, %ymm1, %ymm2 vrcpps %ymm5, %ymm0 vmulps %ymm2, %ymm15, %ymm2 vaddps %ymm2, %ymm15, %ymm15 vmulps %ymm0, %ymm5, %ymm2 vsubps %ymm2, %ymm1, %ymm2 vmulps %ymm2, %ymm0, %ymm2 vrcpps %ymm7, %ymm5 vaddps %ymm2, %ymm0, %ymm2 vmulps %ymm5, %ymm7, %ymm0 vsubps %ymm0, %ymm1, %ymm0 vmulps %ymm0, %ymm5, %ymm0 vaddps %ymm0, %ymm5, %ymm5 vmulps 0x20(%rsp), %ymm10, %ymm0 vaddps 0x160(%rax), %ymm0, %ymm0 vmulps 0x20(%rsp), %ymm6, %ymm6 vaddps 0x180(%rax), %ymm6, %ymm6 vmulps 0x40(%rsp), %ymm4, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vmulps 0x40(%rsp), %ymm8, %ymm1 vaddps %ymm6, %ymm1, %ymm1 vmulps 0x20(%rsp), %ymm3, %ymm3 vaddps 0x1a0(%rax), %ymm3, %ymm3 vmulps 0x40(%rsp), %ymm9, %ymm4 vaddps %ymm3, %ymm4, %ymm3 vmulps 0x60(%rsp), %ymm12, %ymm4 vaddps %ymm0, %ymm4, %ymm4 vmulps 0x60(%rsp), %ymm13, %ymm0 vaddps %ymm1, %ymm0, %ymm6 vmulps 0x60(%rsp), %ymm14, %ymm0 vaddps %ymm3, %ymm0, %ymm7 vmovups 0x80(%rsp), %ymm0 vsubps %ymm4, %ymm0, %ymm0 vmovups 0x200(%rsp), %ymm1 vsubps %ymm6, %ymm1, %ymm1 vsubps %ymm7, %ymm11, %ymm10 vmulps %ymm0, %ymm15, %ymm3 vmulps %ymm2, %ymm1, %ymm1 vmulps %ymm5, %ymm10, %ymm0 vmovups 0x1e0(%rsp), %ymm8 vsubps %ymm4, %ymm8, %ymm4 vmovups 0x1c0(%rsp), %ymm8 vsubps %ymm6, %ymm8, %ymm6 vmovups 0x1a0(%rsp), %ymm8 vsubps %ymm7, %ymm8, %ymm7 vmulps %ymm4, %ymm15, %ymm4 vmulps %ymm2, %ymm6, %ymm6 vmulps %ymm5, %ymm7, %ymm2 vextractf128 $0x1, %ymm4, %xmm5 vextractf128 $0x1, %ymm3, %xmm7 vpminsd %xmm5, %xmm7, %xmm8 vpminsd %xmm4, %xmm3, %xmm9 vinsertf128 $0x1, %xmm8, %ymm9, %ymm8 vextractf128 $0x1, %ymm6, %xmm9 vextractf128 $0x1, %ymm1, %xmm10 vpminsd %xmm9, %xmm10, %xmm11 vpminsd %xmm6, %xmm1, %xmm12 vinsertf128 $0x1, %xmm11, %ymm12, %ymm11 vextractf128 $0x1, %ymm2, %xmm12 vextractf128 $0x1, %ymm0, %xmm13 vpminsd %xmm12, %xmm13, %xmm14 vpminsd %xmm2, %xmm0, %xmm15 vinsertf128 $0x1, %xmm14, %ymm15, %ymm14 vmovups 0xe0(%rsp), %ymm15 vmaxps %ymm14, %ymm11, %ymm11 vmovups 0x100(%rsp), %ymm14 vpmaxsd %xmm5, %xmm7, %xmm5 vmovups 0x40(%rsp), %ymm7 vpmaxsd %xmm4, %xmm3, %xmm3 vinsertf128 $0x1, %xmm5, %ymm3, %ymm3 vmovups 0xc0(%rsp), %ymm5 vpmaxsd %xmm9, %xmm10, %xmm4 vmovups 0x160(%rsp), %ymm10 vmovups 0x180(%rsp), %ymm9 vpmaxsd %xmm6, %xmm1, %xmm1 vmovups 0x60(%rsp), %ymm6 vinsertf128 $0x1, %xmm4, %ymm1, %ymm1 vpmaxsd %xmm12, %xmm13, %xmm4 vmovups 0xa0(%rsp), %ymm12 vmovups 0x120(%rsp), %ymm13 vpmaxsd %xmm2, %xmm0, %xmm0 vinsertf128 $0x1, %xmm4, %ymm0, %ymm0 vminps %ymm0, %ymm1, %ymm0 vmaxps %ymm8, %ymm5, %ymm1 vmovups 0x20(%rsp), %ymm8 vmaxps %ymm11, %ymm1, %ymm1 vmovups 0x140(%rsp), %ymm11 vminps %ymm3, %ymm12, %ymm2 vminps %ymm0, %ymm2, %ymm0 vbroadcastss 0x1d1c439(%rip), %ymm2 # 0x1f1ff10 vmulps %ymm2, %ymm1, %ymm1 vbroadcastss 0x1d1c430(%rip), %ymm2 # 0x1f1ff14 vmulps %ymm2, %ymm0, %ymm0 vcmpleps %ymm0, %ymm1, %ymm0 vmovmskps %ymm0, %r15d jmp 0x2036e9 pushq $0x4 popq %rdx jmp 0x20372a vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x2036d8 movq %rax, (%r9) addq $0x8, %r9 bsfq %rdi, %r8 leaq -0x1(%rdi), %rax movq (%rcx,%r8,8), %r8 prefetcht0 (%r8) prefetcht0 0x40(%r8) prefetcht0 0x80(%r8) prefetcht0 0xc0(%r8) andq %rdi, %rax je 0x203b6d movq %r8, (%r9) addq $0x8, %r9 bsfq %rax, %r8 leaq -0x1(%rax), %rdi jmp 0x203b3a movq %r8, %rcx jmp 0x20372a cmpl $0x6, %edx jne 0x203c31 andq $-0x10, %rcx movzbl (%rcx), %eax movq 0x10(%rsp), %rdx movq 0x8(%rdx), %r8 shll $0x6, %eax leaq 0x280(%rsp), %rdi movq 0x18(%rsp), %rdx movq %r9, 0x80(%rsp) vzeroupper callq *0x8(%r8,%rax) leaq 0x2c0(%rsp), %r11 vmovups 0xa0(%rsp), %ymm12 vmovups 0xc0(%rsp), %ymm5 vmovups 0xe0(%rsp), %ymm15 vmovups 0x100(%rsp), %ymm14 vmovups 0x120(%rsp), %ymm13 vmovups 0x140(%rsp), %ymm11 vmovups 0x160(%rsp), %ymm10 vmovups 0x180(%rsp), %ymm9 movq (%rsp), %r10 vmovups 0x20(%rsp), %ymm8 vmovups 0x40(%rsp), %ymm7 vmovups 0x60(%rsp), %ymm6 movq 0x80(%rsp), %r9 movq 0x8(%rsp), %rsi xorl %edx, %edx testb %al, %al je 0x203c31 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rdx cmpl $0x3, %edx jne 0x2035d4 addq $0x1468, %rsp # imm = 0x1468 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMIntersector1Moeller<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1498, %rsp # imm = 0x1498 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x203c80 addq $0x1498, %rsp # imm = 0x1498 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %rsi, %r14 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x203c6b leaq 0x2f8(%rsp), %rdi movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmaxss 0xc(%r14), %xmm2, %xmm1 vmovaps 0x10(%r14), %xmm3 vbroadcastss 0x1d1d20d(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ced324(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1ce8a34(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss (%r14), %ymm6 vbroadcastss 0x4(%r14), %ymm7 vaddps %xmm3, %xmm4, %xmm3 vbroadcastss 0x8(%r14), %ymm8 xorl %r8d, %r8d vucomiss %xmm2, %xmm3 setb %r8b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm2, %xmm4 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm2, %xmm5 vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %rcx xorq $0x20, %rcx movq %r10, %r13 xorq $0x20, %r13 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm4 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm5 leaq 0x1f4c1ef(%rip), %rax # 0x214ff80 vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x120(%rsp) leaq 0x2f0(%rsp), %rbp vmovups %ymm6, 0x2d0(%rsp) vmovups %ymm7, 0x2b0(%rsp) vmovups %ymm8, 0x290(%rsp) vmovups %ymm9, 0x270(%rsp) vmovups %ymm10, 0x250(%rsp) vmovups %ymm3, 0x230(%rsp) movq %r13, 0x58(%rsp) vmovups %ymm4, 0x210(%rsp) vmovups %ymm5, 0x1f0(%rsp) cmpq %rbp, %rdi je 0x203c6b movq -0x8(%rdi), %rbx addq $-0x8, %rdi testb $0x8, %bl jne 0x203e89 vmovaps 0x40(%rbx,%r8), %ymm0 vsubps %ymm6, %ymm0, %ymm0 vmulps %ymm0, %ymm9, %ymm0 vmovaps 0x40(%rbx,%r9), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm10, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%r10), %ymm1 vsubps %ymm8, %ymm1, %ymm1 vmulps %ymm1, %ymm3, %ymm1 vmaxps %ymm4, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%r11), %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmovaps 0x40(%rbx,%rcx), %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm10, %ymm2 vminps %ymm2, %ymm1, %ymm1 vmovaps 0x40(%rbx,%r13), %ymm2 vsubps %ymm8, %ymm2, %ymm2 vmulps %ymm2, %ymm3, %ymm2 vminps %ymm5, %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r12d testb $0x8, %bl jne 0x203ecf testq %r12, %r12 je 0x203ed3 andq $-0x10, %rbx bsfq %r12, %rax leaq -0x1(%r12), %rsi xorl %r15d, %r15d movq (%rbx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r12, %rsi jne 0x203ed9 movq %rax, %rbx testl %r15d, %r15d je 0x203e08 jmp 0x203f25 pushq $0x6 jmp 0x203ed5 pushq $0x4 popq %r15 jmp 0x203ec4 movq %rcx, %r15 movq %rax, (%rdi) addq $0x8, %rdi bsfq %rsi, %rcx leaq -0x1(%rsi), %rax movq (%rbx,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rsi, %rax je 0x203f1a movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rsi jmp 0x203eeb movq %rcx, %rbx movq %r15, %rcx xorl %r15d, %r15d jmp 0x203ec4 cmpl $0x6, %r15d jne 0x2044dd movl %ebx, %esi andl $0xf, %esi addq $-0x8, %rsi setne %r15b je 0x2044da andq $-0x10, %rbx xorl %ebp, %ebp imulq $0xb0, %rbp, %r13 vmovaps 0x80(%rbx,%r13), %xmm9 vmovaps 0x40(%rbx,%r13), %xmm6 vmulps %xmm6, %xmm9, %xmm0 vmovaps 0x70(%rbx,%r13), %xmm10 vmovaps 0x50(%rbx,%r13), %xmm7 vmulps %xmm7, %xmm10, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, (%rsp) vmovaps 0x60(%rbx,%r13), %xmm11 vmulps %xmm7, %xmm11, %xmm1 vmovaps (%rbx,%r13), %xmm3 vmovaps 0x10(%rbx,%r13), %xmm13 vmovaps 0x20(%rbx,%r13), %xmm0 vmovaps 0x30(%rbx,%r13), %xmm8 vmulps %xmm8, %xmm9, %xmm2 vsubps %xmm1, %xmm2, %xmm5 vmulps %xmm8, %xmm10, %xmm2 vmulps %xmm6, %xmm11, %xmm12 vsubps %xmm2, %xmm12, %xmm4 vbroadcastss (%r14), %xmm12 vsubps %xmm12, %xmm3, %xmm2 vbroadcastss 0x4(%r14), %xmm12 vsubps %xmm12, %xmm13, %xmm3 vbroadcastss 0x8(%r14), %xmm12 vsubps %xmm12, %xmm0, %xmm1 vbroadcastss 0x14(%r14), %xmm12 vbroadcastss 0x18(%r14), %xmm13 vmulps %xmm1, %xmm12, %xmm14 vmulps %xmm3, %xmm13, %xmm15 vsubps %xmm14, %xmm15, %xmm14 vbroadcastss 0x10(%r14), %xmm15 vmulps %xmm2, %xmm13, %xmm0 vmovaps %xmm1, 0x60(%rsp) vmulps %xmm1, %xmm15, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm3, 0x40(%rsp) vmulps %xmm3, %xmm15, %xmm1 vmovaps %xmm2, 0x70(%rsp) vmulps %xmm2, %xmm12, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm4, 0x20(%rsp) vmulps %xmm4, %xmm13, %xmm2 vmovaps (%rsp), %xmm13 vmovaps %xmm5, 0x30(%rsp) vmulps %xmm5, %xmm12, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm15, %xmm13, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm1, %xmm9, %xmm9 vmulps %xmm0, %xmm10, %xmm10 vaddps %xmm10, %xmm9, %xmm9 vmulps %xmm14, %xmm11, %xmm10 vaddps %xmm9, %xmm10, %xmm10 vmulps %xmm1, %xmm7, %xmm1 vmulps %xmm0, %xmm6, %xmm0 vaddps %xmm0, %xmm1, %xmm0 vmovddup 0x1d1cef3(%rip), %xmm1 # xmm1 = mem[0,0] vandps %xmm1, %xmm2, %xmm9 vxorps %xmm10, %xmm9, %xmm7 vmulps %xmm14, %xmm8, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm8 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %xmm10, %xmm7, %xmm0 vcmpnltps %xmm10, %xmm8, %xmm1 vandps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1d1ce1b(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm2, %xmm6 vcmpneqps %xmm2, %xmm10, %xmm1 vandps %xmm1, %xmm0, %xmm0 vaddps %xmm7, %xmm8, %xmm1 vcmpleps %xmm6, %xmm1, %xmm1 vandps %xmm1, %xmm0, %xmm10 vtestps 0x120(%rsp), %xmm10 jne 0x2040e4 incq %rbp cmpq %rsi, %rbp setb %r15b jne 0x203f48 jmp 0x204485 vandps 0x120(%rsp), %xmm10, %xmm10 vmovaps 0x60(%rsp), %xmm0 vmulps 0x20(%rsp), %xmm0, %xmm0 vmovaps 0x40(%rsp), %xmm1 vmulps 0x30(%rsp), %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x70(%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm3 vbroadcastss 0xc(%r14), %xmm0 vmulps %xmm0, %xmm6, %xmm0 vcmpltps %xmm3, %xmm0, %xmm0 vbroadcastss 0x20(%r14), %xmm1 vmulps %xmm1, %xmm6, %xmm1 vcmpleps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm4 vtestps %xmm10, %xmm4 je 0x2040cf addq %rbx, %r13 vandps %xmm4, %xmm10, %xmm0 vmovaps %xmm7, 0x130(%rsp) vmovaps %xmm8, 0x140(%rsp) vmovaps %xmm3, 0x150(%rsp) vmovaps %xmm6, 0x160(%rsp) vmovaps %xmm0, 0x180(%rsp) vmovaps %xmm13, 0x1c0(%rsp) vmovaps 0x30(%rsp), %xmm1 vmovaps %xmm1, 0x1d0(%rsp) vmovaps 0x20(%rsp), %xmm1 vmovaps %xmm1, 0x1e0(%rsp) movq (%rdx), %rax movq %rax, 0x70(%rsp) vrcpps %xmm6, %xmm1 vmulps %xmm1, %xmm6, %xmm2 vbroadcastss 0x1ce8560(%rip), %xmm3 # 0x1eec714 vsubps %xmm2, %xmm3, %xmm2 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps 0x150(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x1b0(%rsp) vmulps 0x130(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x190(%rsp) vmulps 0x140(%rsp), %xmm1, %xmm1 vmovaps %xmm1, 0x1a0(%rsp) vmovmskps %xmm0, %eax movq %rcx, 0x10(%rsp) movq %rax, (%rsp) bsfq %rax, %rcx movq %rcx, 0x20(%rsp) movl 0x90(%r13,%rcx,4), %eax movq 0x70(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq %rax, 0x40(%rsp) movq (%rcx,%rax,8), %rax movl 0x24(%r14), %ecx movq %rax, 0x30(%rsp) testl %ecx, 0x34(%rax) je 0x20425c movq 0x10(%rdx), %rax movq %rax, 0x60(%rsp) cmpq $0x0, 0x10(%rax) movq 0x10(%rsp), %rcx jne 0x20428e movq 0x30(%rsp), %rax cmpq $0x0, 0x48(%rax) jne 0x20428e xorl %eax, %eax jmp 0x204274 movq 0x20(%rsp), %rax movq (%rsp), %rcx btcq %rax, %rcx movq %rcx, (%rsp) movb $0x1, %al movq 0x10(%rsp), %rcx testb %al, %al je 0x2044ec movq (%rsp), %rax testq %rax, %rax jne 0x2041ff jmp 0x2040cf movq %rsi, 0x88(%rsp) movq %r11, 0x90(%rsp) movq %r10, 0x98(%rsp) movq %r9, 0xa0(%rsp) movq %r8, 0xa8(%rsp) movq %rdi, 0xb0(%rsp) movq 0x20(%rsp), %rsi vmovss 0x190(%rsp,%rsi,4), %xmm0 vmovss 0x1a0(%rsp,%rsi,4), %xmm1 movq %rdx, 0xb8(%rsp) movq 0x8(%rdx), %rcx movl 0xa0(%r13,%rsi,4), %edx vmovss 0x1c0(%rsp,%rsi,4), %xmm2 vmovss 0x1d0(%rsp,%rsi,4), %xmm3 vmovss 0x1e0(%rsp,%rsi,4), %xmm4 vmovss %xmm2, 0xc0(%rsp) vmovss %xmm3, 0xc4(%rsp) vmovss %xmm4, 0xc8(%rsp) vmovss %xmm0, 0xcc(%rsp) vmovss %xmm1, 0xd0(%rsp) movl %edx, 0xd4(%rsp) movq 0x40(%rsp), %rax movl %eax, 0xd8(%rsp) movl (%rcx), %eax movl %eax, 0xdc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xe0(%rsp) vmovss 0x20(%r14), %xmm0 vmovss %xmm0, 0x40(%rsp) vmovss 0x1b0(%rsp,%rsi,4), %xmm0 vmovss %xmm0, 0x20(%r14) orl $-0x1, 0x1c(%rsp) leaq 0x1c(%rsp), %rax movq %rax, 0xf0(%rsp) movq 0x30(%rsp), %rdx movq 0x18(%rdx), %rax movq %rax, 0xf8(%rsp) movq %rcx, 0x100(%rsp) movq %r14, 0x108(%rsp) leaq 0xc0(%rsp), %rax movq %rax, 0x110(%rsp) movl $0x1, 0x118(%rsp) movq 0x48(%rdx), %rax testq %rax, %rax je 0x2043e3 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x204424 movq 0x60(%rsp), %rax movq 0x10(%rax), %rax testq %rax, %rax je 0x204420 movq 0x60(%rsp), %rcx testb $0x2, (%rcx) jne 0x204406 movq 0x30(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x204413 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x204424 xorl %eax, %eax jmp 0x204443 vmovss 0x40(%rsp), %xmm0 vmovss %xmm0, 0x20(%r14) movq (%rsp), %rax movq 0x20(%rsp), %rcx btcq %rcx, %rax movq %rax, (%rsp) movb $0x1, %al movq 0xb8(%rsp), %rdx movq 0xb0(%rsp), %rdi movq 0xa8(%rsp), %r8 movq 0xa0(%rsp), %r9 movq 0x98(%rsp), %r10 movq 0x90(%rsp), %r11 movq 0x10(%rsp), %rcx movq 0x88(%rsp), %rsi jmp 0x204274 vmovups 0x2d0(%rsp), %ymm6 vmovups 0x2b0(%rsp), %ymm7 vmovups 0x290(%rsp), %ymm8 vmovups 0x270(%rsp), %ymm9 vmovups 0x250(%rsp), %ymm10 vmovups 0x230(%rsp), %ymm3 movq 0x58(%rsp), %r13 vmovups 0x210(%rsp), %ymm4 vmovups 0x1f0(%rsp), %ymm5 leaq 0x2f0(%rsp), %rbp xorl %r15d, %r15d cmpl $0x3, %r15d jne 0x203df7 jmp 0x203c6b testb $0x1, %r15b vmovups 0x2d0(%rsp), %ymm6 vmovups 0x2b0(%rsp), %ymm7 vmovups 0x290(%rsp), %ymm8 vmovups 0x270(%rsp), %ymm9 vmovups 0x250(%rsp), %ymm10 vmovups 0x230(%rsp), %ymm3 movq 0x58(%rsp), %r13 vmovups 0x210(%rsp), %ymm4 vmovups 0x1f0(%rsp), %ymm5 leaq 0x2f0(%rsp), %rbp movl $0x0, %r15d je 0x2044dd movl $0xff800000, 0x20(%r14) # imm = 0xFF800000 pushq $0x3 popq %r15 jmp 0x2044dd nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, true, embree::avx::ArrayIntersector1<embree::avx::TriangleMvIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1558, %rsp # imm = 0x1558 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x2059ea vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x2059ea leaq 0x3b8(%rsp), %r8 movq 0x70(%rax), %rax movq %rax, -0x8(%r8) vmovaps 0x10(%rsi), %xmm3 vmaxss 0xc(%rsi), %xmm2, %xmm1 vbroadcastss 0x1d1bf69(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1cec080(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vbroadcastss 0x1ce779e(%rip), %xmm5 # 0x1eec714 vdivps %xmm3, %xmm5, %xmm3 vbroadcastss 0x1d1bfdd(%rip), %xmm5 # 0x1f20f60 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vbroadcastss 0x1d1af7e(%rip), %xmm4 # 0x1f1ff10 vmulps %xmm4, %xmm3, %xmm4 vbroadcastss (%rsi), %ymm7 vbroadcastss 0x4(%rsi), %ymm5 vmovups %ymm5, 0x390(%rsp) vbroadcastss 0x8(%rsi), %ymm5 vmovups %ymm5, 0x370(%rsp) vbroadcastss 0x1d1af52(%rip), %xmm5 # 0x1f1ff14 xorl %r9d, %r9d vucomiss %xmm2, %xmm4 vmulps %xmm5, %xmm3, %xmm5 setb %r9b vshufps $0x0, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovups %ymm3, 0x350(%rsp) vmovshdup %xmm4, %xmm6 # xmm6 = xmm4[1,1,3,3] vshufps $0x55, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[1,1,1,1] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovups %ymm3, 0x330(%rsp) vshufpd $0x1, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[1,0] vshufps $0xaa, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x310(%rsp) vshufps $0x0, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x2f0(%rsp) shll $0x5, %r9d xorl %r10d, %r10d vucomiss %xmm2, %xmm6 vshufps $0x55, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x2d0(%rsp) vshufps $0xaa, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[2,2,2,2] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x2b0(%rsp) setb %r10b shll $0x5, %r10d orq $0x40, %r10 xorl %r11d, %r11d vucomiss %xmm2, %xmm3 setb %r11b shll $0x5, %r11d orq $0x80, %r11 movq %r9, %r14 xorq $0x20, %r14 movq %r10, %r15 xorq $0x20, %r15 movq %r11, %rax xorq $0x20, %rax movq %rax, 0xa8(%rsp) vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovups %ymm1, 0x290(%rsp) vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovups %ymm0, 0x270(%rsp) leaq 0x1f4aeb5(%rip), %rax # 0x214ff80 vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x110(%rsp) vmovups %ymm7, 0x120(%rsp) movq %r14, 0x88(%rsp) movq %r15, 0x80(%rsp) leaq 0x3b0(%rsp), %rax cmpq %rax, %r8 je 0x2059ea movq -0x8(%r8), %rbx addq $-0x8, %r8 testb $0x8, %bl jne 0x2051d3 vmovaps 0x40(%rbx,%r9), %ymm0 vsubps %ymm7, %ymm0, %ymm0 vmulps 0x350(%rsp), %ymm0, %ymm0 vmovaps 0x40(%rbx,%r10), %ymm1 vmovups 0x390(%rsp), %ymm3 vsubps %ymm3, %ymm1, %ymm1 vmulps 0x330(%rsp), %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%r11), %ymm1 vmovups 0x370(%rsp), %ymm4 vsubps %ymm4, %ymm1, %ymm1 vmulps 0x310(%rsp), %ymm1, %ymm1 vmaxps 0x290(%rsp), %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%r14), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps 0x2f0(%rsp), %ymm1, %ymm1 vmovaps 0x40(%rbx,%r15), %ymm2 vsubps %ymm3, %ymm2, %ymm2 vmulps 0x2d0(%rsp), %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 movq 0xa8(%rsp), %rax vmovaps 0x40(%rbx,%rax), %ymm2 vsubps %ymm4, %ymm2, %ymm2 vmulps 0x2b0(%rsp), %ymm2, %ymm2 vminps 0x270(%rsp), %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r12d testb $0x8, %bl jne 0x205217 testq %r12, %r12 je 0x20521b andq $-0x10, %rbx bsfq %r12, %rax leaq -0x1(%r12), %rdi xorl %ebp, %ebp movq (%rbx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r12, %rdi jne 0x205220 movq %rax, %rbx testl %ebp, %ebp je 0x20510e jmp 0x205263 pushq $0x6 jmp 0x20521d pushq $0x4 popq %rbp jmp 0x20520d movq %rax, (%r8) addq $0x8, %r8 bsfq %rdi, %rcx leaq -0x1(%rdi), %rax movq (%rbx,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdi, %rax je 0x20525e movq %rcx, (%r8) addq $0x8, %r8 bsfq %rax, %rcx leaq -0x1(%rax), %rdi jmp 0x20522f movq %rcx, %rbx jmp 0x20520d cmpl $0x6, %ebp jne 0x2059e1 movl %ebx, %edi andl $0xf, %edi xorl %ebp, %ebp addq $-0x8, %rdi setne %al je 0x2059e1 andq $-0x10, %rbx xorl %ecx, %ecx movb %al, 0xe(%rsp) movq %rcx, 0xa0(%rsp) imulq $0xb0, %rcx, %rcx leaq 0xf(%rsp), %rax movq %rax, 0x1f0(%rsp) vbroadcastss (%rsi), %xmm1 vbroadcastss 0x4(%rsi), %xmm2 vbroadcastss 0x8(%rsi), %xmm3 vmovaps (%rbx,%rcx), %xmm0 vmovaps 0x10(%rbx,%rcx), %xmm4 vmovaps 0x20(%rbx,%rcx), %xmm5 vmovaps 0x30(%rbx,%rcx), %xmm6 vsubps %xmm1, %xmm0, %xmm7 vsubps %xmm2, %xmm4, %xmm14 vsubps %xmm3, %xmm5, %xmm10 vsubps %xmm1, %xmm6, %xmm0 vmovaps %xmm0, 0x50(%rsp) vmovaps 0x40(%rbx,%rcx), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps %xmm0, 0x20(%rsp) vmovaps 0x50(%rbx,%rcx), %xmm4 vsubps %xmm3, %xmm4, %xmm6 vmovaps 0x60(%rbx,%rcx), %xmm4 vsubps %xmm1, %xmm4, %xmm5 vmovaps 0x70(%rbx,%rcx), %xmm1 vsubps %xmm2, %xmm1, %xmm2 vmovaps %xmm2, 0x30(%rsp) movq %rcx, 0x10(%rsp) vmovaps 0x80(%rbx,%rcx), %xmm1 vsubps %xmm3, %xmm1, %xmm3 vmovaps %xmm3, 0x40(%rsp) vsubps %xmm7, %xmm5, %xmm13 vsubps %xmm14, %xmm2, %xmm15 vsubps %xmm10, %xmm3, %xmm0 vaddps %xmm2, %xmm14, %xmm1 vaddps %xmm3, %xmm10, %xmm4 vmulps %xmm0, %xmm1, %xmm8 vmulps %xmm4, %xmm15, %xmm9 vsubps %xmm8, %xmm9, %xmm12 vaddps %xmm5, %xmm7, %xmm8 vmulps %xmm4, %xmm13, %xmm4 vmovaps %xmm0, 0x170(%rsp) vmulps %xmm0, %xmm8, %xmm9 vsubps %xmm4, %xmm9, %xmm4 vmovaps %xmm15, 0x180(%rsp) vmulps %xmm15, %xmm8, %xmm8 vmovaps %xmm13, 0x190(%rsp) vmulps %xmm1, %xmm13, %xmm1 vsubps %xmm8, %xmm1, %xmm1 vbroadcastss 0x18(%rsi), %xmm0 vmulps %xmm1, %xmm0, %xmm1 vbroadcastss 0x14(%rsi), %xmm13 vmulps %xmm4, %xmm13, %xmm4 vaddps %xmm4, %xmm1, %xmm1 vbroadcastss 0x10(%rsi), %xmm11 vmulps %xmm12, %xmm11, %xmm4 vaddps %xmm1, %xmm4, %xmm9 vmovaps 0x20(%rsp), %xmm3 vsubps %xmm3, %xmm14, %xmm15 vsubps %xmm6, %xmm10, %xmm1 vmovaps %xmm14, 0x1b0(%rsp) vaddps %xmm3, %xmm14, %xmm2 vmovaps %xmm10, 0x1a0(%rsp) vaddps %xmm6, %xmm10, %xmm4 vmovaps %xmm6, %xmm10 vmovaps %xmm0, %xmm6 vmulps %xmm1, %xmm2, %xmm12 vmulps %xmm4, %xmm15, %xmm14 vsubps %xmm12, %xmm14, %xmm12 vmovaps 0x50(%rsp), %xmm0 vsubps %xmm0, %xmm7, %xmm14 vmulps %xmm4, %xmm14, %xmm4 vmovaps %xmm7, 0x90(%rsp) vaddps %xmm0, %xmm7, %xmm7 vmovaps %xmm1, 0x150(%rsp) vmulps %xmm1, %xmm7, %xmm8 vsubps %xmm4, %xmm8, %xmm4 vbroadcastss 0x1d1ba9f(%rip), %xmm8 # 0x1f20ec4 vmovaps %xmm15, 0x160(%rsp) vmulps %xmm7, %xmm15, %xmm7 vmulps %xmm2, %xmm14, %xmm1 vsubps %xmm7, %xmm1, %xmm1 vmulps %xmm1, %xmm6, %xmm1 vmulps %xmm4, %xmm13, %xmm4 vaddps %xmm4, %xmm1, %xmm1 vmulps %xmm12, %xmm11, %xmm4 vaddps %xmm1, %xmm4, %xmm4 vsubps %xmm5, %xmm0, %xmm1 vaddps %xmm5, %xmm0, %xmm7 vmovaps 0x30(%rsp), %xmm0 vsubps %xmm0, %xmm3, %xmm12 vaddps %xmm0, %xmm3, %xmm0 vmovaps 0x40(%rsp), %xmm2 vsubps %xmm2, %xmm10, %xmm15 vaddps %xmm2, %xmm10, %xmm2 vmulps %xmm0, %xmm15, %xmm3 vmulps %xmm2, %xmm12, %xmm5 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm2, %xmm1, %xmm2 vmulps %xmm7, %xmm15, %xmm5 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm7, %xmm12, %xmm5 vmovups 0x120(%rsp), %ymm7 vmulps %xmm0, %xmm1, %xmm0 vsubps %xmm5, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm0 vmovaps %xmm13, 0x50(%rsp) vmulps %xmm2, %xmm13, %xmm2 vaddps %xmm2, %xmm0, %xmm0 vmovaps %xmm11, 0x40(%rsp) vmulps %xmm3, %xmm11, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vaddps %xmm4, %xmm9, %xmm2 vaddps %xmm2, %xmm0, %xmm13 vminps %xmm4, %xmm9, %xmm2 vminps %xmm0, %xmm2, %xmm2 vandps %xmm8, %xmm13, %xmm5 vbroadcastss 0x1d1b9ee(%rip), %xmm3 # 0x1f20ecc vmovaps %xmm5, 0x20(%rsp) vmulps %xmm3, %xmm5, %xmm3 vbroadcastss 0x1d1b9cf(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm3, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm9, 0x30(%rsp) vmaxps %xmm4, %xmm9, %xmm5 vmaxps %xmm0, %xmm5, %xmm0 vcmpleps %xmm3, %xmm0, %xmm0 vorps %xmm0, %xmm2, %xmm0 vtestps 0x110(%rsp), %xmm0 je 0x205997 vmovaps %xmm6, %xmm10 vmovaps 0x170(%rsp), %xmm7 vmovaps 0x160(%rsp), %xmm11 vmovaps %xmm0, 0x140(%rsp) vmulps %xmm7, %xmm11, %xmm0 vmovaps 0x180(%rsp), %xmm9 vmovaps 0x150(%rsp), %xmm6 vmulps %xmm6, %xmm9, %xmm2 vsubps %xmm0, %xmm2, %xmm2 vmulps %xmm6, %xmm12, %xmm3 vmulps %xmm15, %xmm11, %xmm5 vsubps %xmm3, %xmm5, %xmm5 vandps %xmm0, %xmm8, %xmm0 vandps %xmm3, %xmm8, %xmm3 vcmpltps %xmm3, %xmm0, %xmm0 vblendvps %xmm0, %xmm2, %xmm5, %xmm0 vmulps %xmm15, %xmm14, %xmm2 vmulps %xmm7, %xmm14, %xmm3 vmovaps 0x190(%rsp), %xmm7 vmulps %xmm7, %xmm6, %xmm5 vsubps %xmm5, %xmm3, %xmm3 vmulps %xmm1, %xmm6, %xmm6 vsubps %xmm2, %xmm6, %xmm6 vandps %xmm5, %xmm8, %xmm5 vandps %xmm2, %xmm8, %xmm2 vcmpltps %xmm2, %xmm5, %xmm2 vblendvps %xmm2, %xmm3, %xmm6, %xmm2 vmulps %xmm1, %xmm11, %xmm1 vmulps %xmm7, %xmm11, %xmm3 vmulps %xmm9, %xmm14, %xmm5 vmulps %xmm12, %xmm14, %xmm6 vsubps %xmm5, %xmm3, %xmm3 vsubps %xmm1, %xmm6, %xmm6 vandps %xmm5, %xmm8, %xmm5 vandps %xmm1, %xmm8, %xmm1 vcmpltps %xmm1, %xmm5, %xmm1 vblendvps %xmm1, %xmm3, %xmm6, %xmm1 vmulps %xmm1, %xmm10, %xmm3 vmulps 0x50(%rsp), %xmm2, %xmm5 vaddps %xmm5, %xmm3, %xmm3 vmulps 0x40(%rsp), %xmm0, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vaddps %xmm3, %xmm3, %xmm5 vmulps 0x1a0(%rsp), %xmm1, %xmm3 vmulps 0x1b0(%rsp), %xmm2, %xmm6 vaddps %xmm6, %xmm3, %xmm3 vmulps 0x90(%rsp), %xmm0, %xmm6 vaddps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm3, %xmm3 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm7 vbroadcastss 0x1ce70e0(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm6, %xmm7 vaddps %xmm7, %xmm6, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0xc(%rsi), %xmm6 vcmpleps %xmm3, %xmm6, %xmm6 vbroadcastss 0x20(%rsi), %xmm7 vcmpleps %xmm7, %xmm3, %xmm7 vandps %xmm7, %xmm6, %xmm6 vmovups 0x120(%rsp), %ymm7 vcmpneqps 0x1ce63a0(%rip), %xmm5, %xmm5 # 0x1eeba10 vandps %xmm6, %xmm5, %xmm6 vmovaps 0x140(%rsp), %xmm5 vandps 0x110(%rsp), %xmm5, %xmm5 vpslld $0x1f, %xmm6, %xmm6 vpsrad $0x1f, %xmm6, %xmm6 vtestps %xmm5, %xmm6 je 0x205997 addq %rbx, 0x10(%rsp) vandps %xmm5, %xmm6, %xmm5 vmovaps 0x30(%rsp), %xmm6 vmovaps %xmm6, 0x1c0(%rsp) vmovaps %xmm4, 0x1d0(%rsp) vmovaps %xmm13, 0x1e0(%rsp) movq %rax, 0x1f0(%rsp) vmovaps %xmm5, 0x200(%rsp) vmovaps %xmm3, 0x230(%rsp) vmovaps %xmm0, 0x240(%rsp) vmovaps %xmm2, 0x250(%rsp) vmovaps %xmm1, 0x260(%rsp) movq (%rdx), %rax movq %rax, 0x50(%rsp) vrcpps %xmm13, %xmm0 vmulps %xmm0, %xmm13, %xmm1 vbroadcastss 0x1ce7000(%rip), %xmm2 # 0x1eec714 vsubps %xmm1, %xmm2, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1ceb8bf(%rip), %xmm1 # 0x1ef0fe8 vmovaps 0x20(%rsp), %xmm3 vcmpnltps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm0 vmulps %xmm0, %xmm6, %xmm1 vminps %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x210(%rsp) vmulps %xmm0, %xmm4, %xmm0 vminps %xmm2, %xmm0, %xmm0 vmovaps %xmm0, 0x220(%rsp) vmovmskps %xmm5, %r15d bsfq %r15, %r13 movq 0x10(%rsp), %rax movl 0x90(%rax,%r13,4), %eax movq 0x50(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rbp movl 0x24(%rsi), %ecx testl %ecx, 0x34(%rbp) je 0x20579d movq 0x10(%rdx), %r14 cmpq $0x0, 0x10(%r14) jne 0x2057b7 cmpq $0x0, 0x48(%rbp) jne 0x2057b7 xorl %eax, %eax jmp 0x2057a3 btcq %r13, %r15 movb $0x1, %al xorl %ebp, %ebp testb %al, %al je 0x2059c0 testq %r15, %r15 jne 0x20575e jmp 0x205997 movq %rdi, 0x40(%rsp) vmovss 0x210(%rsp,%r13,4), %xmm0 vmovss 0x220(%rsp,%r13,4), %xmm1 movq 0x8(%rdx), %rcx movq 0x10(%rsp), %rdi movl 0xa0(%rdi,%r13,4), %edi vmovss 0x240(%rsp,%r13,4), %xmm2 vmovss 0x250(%rsp,%r13,4), %xmm3 vmovss 0x260(%rsp,%r13,4), %xmm4 vmovss %xmm2, 0xb0(%rsp) vmovss %xmm3, 0xb4(%rsp) vmovss %xmm4, 0xb8(%rsp) vmovss %xmm0, 0xbc(%rsp) vmovss %xmm1, 0xc0(%rsp) movl %edi, 0xc4(%rsp) movl %eax, 0xc8(%rsp) movl (%rcx), %eax movl %eax, 0xcc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xd0(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0x30(%rsp) vmovss 0x230(%rsp,%r13,4), %xmm0 vmovss %xmm0, 0x20(%rsi) orl $-0x1, 0x1c(%rsp) leaq 0x1c(%rsp), %rax movq %rax, 0xe0(%rsp) movq 0x18(%rbp), %rax movq %rax, 0xe8(%rsp) movq %rcx, 0xf0(%rsp) movq %rsi, 0xf8(%rsp) leaq 0xb0(%rsp), %rax movq %rax, 0x100(%rsp) movl $0x1, 0x108(%rsp) movq 0x48(%rbp), %rax testq %rax, %rax movq %r8, 0x78(%rsp) movq %r9, 0x70(%rsp) movq %r10, 0x68(%rsp) movq %r11, 0x60(%rsp) je 0x20591e leaq 0xe0(%rsp), %rdi movq %rdx, 0x20(%rsp) movq %rsi, 0x90(%rsp) vzeroupper callq %rax movq 0x60(%rsp), %r11 movq 0x68(%rsp), %r10 movq 0x70(%rsp), %r9 vmovups 0x120(%rsp), %ymm7 movq 0x78(%rsp), %r8 movq 0x90(%rsp), %rsi movq 0x20(%rsp), %rdx movq 0xe0(%rsp), %rax cmpl $0x0, (%rax) je 0x20597a movq 0x10(%r14), %rax testq %rax, %rax je 0x205976 testb $0x2, (%r14) jne 0x205933 testb $0x40, 0x3e(%rbp) je 0x205969 leaq 0xe0(%rsp), %rdi movq %rdx, %r14 movq %rsi, %rbp vzeroupper callq %rax movq 0x60(%rsp), %r11 movq 0x68(%rsp), %r10 movq 0x70(%rsp), %r9 vmovups 0x120(%rsp), %ymm7 movq 0x78(%rsp), %r8 movq %rbp, %rsi movq %r14, %rdx movq 0xe0(%rsp), %rax cmpl $0x0, (%rax) je 0x20597a xorl %eax, %eax jmp 0x20598b vmovss 0x30(%rsp), %xmm0 vmovss %xmm0, 0x20(%rsi) btcq %r13, %r15 movb $0x1, %al xorl %ebp, %ebp movq 0x40(%rsp), %rdi jmp 0x2057a5 movq 0xa0(%rsp), %rcx incq %rcx cmpq %rdi, %rcx setb %al jne 0x205286 movq 0x88(%rsp), %r14 movq 0x80(%rsp), %r15 jmp 0x2059e1 testb $0x1, 0xe(%rsp) movq 0x88(%rsp), %r14 movq 0x80(%rsp), %r15 je 0x2059e1 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rbp cmpl $0x3, %ebp jne 0x2050f5 addq $0x1558, %rsp # imm = 0x1558 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMvMBIntersector1Moeller<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1498, %rsp # imm = 0x1498 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x207040 addq $0x1498, %rsp # imm = 0x1498 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %rsi, %r14 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x20702b leaq 0x2f8(%rsp), %rdi movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmaxss 0xc(%r14), %xmm2, %xmm1 vmovaps 0x10(%r14), %xmm3 vbroadcastss 0x1d19e4d(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce9f64(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1ce5674(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss (%r14), %ymm6 vbroadcastss 0x4(%r14), %ymm7 vaddps %xmm3, %xmm4, %xmm3 vbroadcastss 0x8(%r14), %ymm8 xorl %r8d, %r8d vucomiss %xmm2, %xmm3 setb %r8b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm2, %xmm4 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm2, %xmm5 vinsertf128 $0x1, %xmm3, %ymm3, %ymm5 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %r15 xorq $0x20, %r15 movq %r10, %r13 xorq $0x20, %r13 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm11 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm12 leaq 0x1f48e2f(%rip), %rax # 0x214ff80 vmovdqa 0xf0(%rax), %xmm0 vmovdqa %xmm0, 0x120(%rsp) leaq 0x2f0(%rsp), %rbp vmovups %ymm6, 0x2d0(%rsp) vmovups %ymm7, 0x2b0(%rsp) vmovups %ymm8, 0x290(%rsp) vmovups %ymm9, 0x270(%rsp) vmovups %ymm10, 0x250(%rsp) vmovups %ymm5, 0x230(%rsp) movq %r15, 0x68(%rsp) movq %r13, 0x60(%rsp) vmovups %ymm11, 0x210(%rsp) vmovups %ymm12, 0x1f0(%rsp) cmpq %rbp, %rdi je 0x20702b movq -0x8(%rdi), %r12 addq $-0x8, %rdi testb $0x8, %r12b jne 0x2072be movq %r12, %rax andq $-0x10, %rax vbroadcastss 0x1c(%r14), %ymm0 vmulps 0x100(%rax,%r8), %ymm0, %ymm1 vaddps 0x40(%rax,%r8), %ymm1, %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmulps 0x100(%rax,%r9), %ymm0, %ymm2 vaddps 0x40(%rax,%r9), %ymm2, %ymm2 vmaxps %ymm1, %ymm11, %ymm1 vsubps %ymm7, %ymm2, %ymm2 vmulps 0x100(%rax,%r10), %ymm0, %ymm3 vmulps %ymm2, %ymm10, %ymm2 vaddps 0x40(%rax,%r10), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm5, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r11), %ymm3, %ymm2 vsubps %ymm6, %ymm2, %ymm2 vmulps %ymm2, %ymm9, %ymm2 vminps %ymm2, %ymm12, %ymm2 vmulps 0x100(%rax,%r15), %ymm0, %ymm3 vaddps 0x40(%rax,%r15), %ymm3, %ymm3 vsubps %ymm7, %ymm3, %ymm3 vmulps %ymm3, %ymm10, %ymm3 vmulps 0x100(%rax,%r13), %ymm0, %ymm4 vaddps 0x40(%rax,%r13), %ymm4, %ymm4 vsubps %ymm8, %ymm4, %ymm4 vmulps %ymm4, %ymm5, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r12d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x207306 vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %ebx testb $0x8, %r12b jne 0x207302 testq %rbx, %rbx je 0x207333 andq $-0x10, %r12 bsfq %rbx, %rax leaq -0x1(%rbx), %rsi xorl %ecx, %ecx movq (%r12,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %rbx, %rsi jne 0x207338 movq %rax, %r12 testl %ecx, %ecx je 0x2071cd jmp 0x207380 pushq $0x6 jmp 0x207335 vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x2072ae pushq $0x4 popq %rcx jmp 0x2072f8 movq %rax, (%rdi) addq $0x8, %rdi bsfq %rsi, %rcx leaq -0x1(%rsi), %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rsi, %rax je 0x207376 movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rsi jmp 0x207347 movq %rcx, %r12 xorl %ecx, %ecx jmp 0x2072f8 cmpl $0x6, %ecx jne 0x20799d movl %r12d, %esi andl $0xf, %esi xorl %ecx, %ecx addq $-0x8, %rsi setne %bpl je 0x207995 andq $-0x10, %r12 xorl %r15d, %r15d imulq $0x140, %r15, %r13 # imm = 0x140 vbroadcastss 0x1c(%r14), %xmm0 vmulps 0x90(%r12,%r13), %xmm0, %xmm1 vaddps (%r12,%r13), %xmm1, %xmm7 vmulps 0xa0(%r12,%r13), %xmm0, %xmm1 vaddps 0x10(%r12,%r13), %xmm1, %xmm5 vmulps 0xb0(%r12,%r13), %xmm0, %xmm1 vaddps 0x20(%r12,%r13), %xmm1, %xmm6 vmulps 0xc0(%r12,%r13), %xmm0, %xmm1 vaddps 0x30(%r12,%r13), %xmm1, %xmm1 vmulps 0xd0(%r12,%r13), %xmm0, %xmm2 vaddps 0x40(%r12,%r13), %xmm2, %xmm2 vmulps 0xe0(%r12,%r13), %xmm0, %xmm4 vaddps 0x50(%r12,%r13), %xmm4, %xmm8 vmulps 0xf0(%r12,%r13), %xmm0, %xmm4 vaddps 0x60(%r12,%r13), %xmm4, %xmm9 vmulps 0x100(%r12,%r13), %xmm0, %xmm4 vaddps 0x70(%r12,%r13), %xmm4, %xmm10 vmulps 0x110(%r12,%r13), %xmm0, %xmm0 vaddps 0x80(%r12,%r13), %xmm0, %xmm0 vsubps %xmm1, %xmm7, %xmm3 vmovaps %xmm3, (%rsp) vsubps %xmm2, %xmm5, %xmm12 vmovaps %xmm12, 0x40(%rsp) vsubps %xmm8, %xmm6, %xmm8 vsubps %xmm7, %xmm9, %xmm9 vsubps %xmm5, %xmm10, %xmm10 vsubps %xmm6, %xmm0, %xmm11 vmulps %xmm11, %xmm12, %xmm0 vmulps %xmm10, %xmm8, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x30(%rsp) vmulps %xmm9, %xmm8, %xmm1 vmulps %xmm3, %xmm11, %xmm2 vsubps %xmm1, %xmm2, %xmm4 vmulps %xmm3, %xmm10, %xmm2 vmulps %xmm9, %xmm12, %xmm12 vsubps %xmm2, %xmm12, %xmm3 vbroadcastss (%r14), %xmm12 vbroadcastss 0x4(%r14), %xmm13 vbroadcastss 0x8(%r14), %xmm14 vbroadcastss 0x14(%r14), %xmm15 vsubps %xmm12, %xmm7, %xmm2 vbroadcastss 0x18(%r14), %xmm12 vsubps %xmm13, %xmm5, %xmm5 vsubps %xmm14, %xmm6, %xmm6 vmulps %xmm6, %xmm15, %xmm13 vmulps %xmm5, %xmm12, %xmm14 vsubps %xmm13, %xmm14, %xmm13 vbroadcastss 0x10(%r14), %xmm14 vmulps %xmm2, %xmm12, %xmm0 vmulps %xmm6, %xmm14, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmulps %xmm5, %xmm14, %xmm1 vmovaps %xmm2, 0x20(%rsp) vmulps %xmm2, %xmm15, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm3, 0x70(%rsp) vmulps %xmm3, %xmm12, %xmm2 vmulps %xmm4, %xmm15, %xmm12 vmovaps %xmm4, %xmm15 vaddps %xmm2, %xmm12, %xmm2 vmovaps 0x30(%rsp), %xmm3 vmulps %xmm3, %xmm14, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm1, %xmm11, %xmm11 vmulps %xmm0, %xmm10, %xmm10 vaddps %xmm11, %xmm10, %xmm10 vmulps %xmm13, %xmm9, %xmm9 vaddps %xmm10, %xmm9, %xmm10 vmulps %xmm1, %xmm8, %xmm1 vmulps 0x40(%rsp), %xmm0, %xmm0 vaddps %xmm1, %xmm0, %xmm0 vmovddup 0x1d19a1a(%rip), %xmm1 # xmm1 = mem[0,0] vandps %xmm1, %xmm2, %xmm9 vxorps %xmm10, %xmm9, %xmm7 vmulps (%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm8 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %xmm10, %xmm7, %xmm0 vcmpnltps %xmm10, %xmm8, %xmm1 vandps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1d19942(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm2, %xmm4 vcmpneqps %xmm2, %xmm10, %xmm1 vandps %xmm1, %xmm0, %xmm0 vaddps %xmm7, %xmm8, %xmm1 vcmpleps %xmm4, %xmm1, %xmm1 vandps %xmm1, %xmm0, %xmm10 vtestps 0x120(%rsp), %xmm10 jne 0x2075bd incq %r15 cmpq %rsi, %r15 setb %bpl jne 0x2073a6 jmp 0x207943 vmovaps %xmm3, %xmm14 vandps 0x120(%rsp), %xmm10, %xmm10 vmulps 0x70(%rsp), %xmm6, %xmm0 vmulps %xmm5, %xmm15, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x20(%rsp), %xmm3, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm3 vbroadcastss 0xc(%r14), %xmm0 vmulps %xmm0, %xmm4, %xmm0 vcmpltps %xmm3, %xmm0, %xmm0 vbroadcastss 0x20(%r14), %xmm1 vmulps %xmm1, %xmm4, %xmm1 vcmpleps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm5 vtestps %xmm10, %xmm5 je 0x2075a8 addq %r12, %r13 vandps %xmm5, %xmm10, %xmm0 vmovaps %xmm7, 0x130(%rsp) vmovaps %xmm8, 0x140(%rsp) vmovaps %xmm3, 0x150(%rsp) vmovaps %xmm4, 0x160(%rsp) vmovaps %xmm0, 0x180(%rsp) vmovaps %xmm14, 0x1c0(%rsp) vmovaps %xmm15, 0x1d0(%rsp) vmovaps 0x70(%rsp), %xmm1 vmovaps %xmm1, 0x1e0(%rsp) movq (%rdx), %rax movq %rax, 0x70(%rsp) vrcpps %xmm4, %xmm1 vmulps %xmm1, %xmm4, %xmm2 vbroadcastss 0x1ce5097(%rip), %xmm3 # 0x1eec714 vsubps %xmm2, %xmm3, %xmm2 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps 0x150(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x1b0(%rsp) vmulps 0x130(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x190(%rsp) vmulps 0x140(%rsp), %xmm1, %xmm1 vmovaps %xmm1, 0x1a0(%rsp) vmovmskps %xmm0, %eax movq %rax, (%rsp) bsfq %rax, %rcx movq %rcx, 0x30(%rsp) movl 0x120(%r13,%rcx,4), %eax movq 0x70(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq %rax, 0x20(%rsp) movq (%rcx,%rax,8), %rax movl 0x24(%r14), %ecx movq %rax, 0x40(%rsp) testl %ecx, 0x34(%rax) je 0x207720 movq 0x10(%rdx), %rax movq %rax, 0x58(%rsp) cmpq $0x0, 0x10(%rax) movl $0x0, %ecx jne 0x20774f movq 0x40(%rsp), %rax cmpq $0x0, 0x48(%rax) jne 0x20774f xorl %eax, %eax jmp 0x207735 movq 0x30(%rsp), %rax movq (%rsp), %rcx btcq %rax, %rcx movq %rcx, (%rsp) movb $0x1, %al xorl %ecx, %ecx testb %al, %al je 0x2079ab movq (%rsp), %rax testq %rax, %rax jne 0x2076c3 jmp 0x2075a8 movq %rsi, 0x88(%rsp) movq %r11, 0x90(%rsp) movq %r10, 0x98(%rsp) movq %r9, 0xa0(%rsp) movq %r8, 0xa8(%rsp) movq %rdi, 0xb0(%rsp) movq 0x30(%rsp), %rsi vmovss 0x190(%rsp,%rsi,4), %xmm0 vmovss 0x1a0(%rsp,%rsi,4), %xmm1 movq %rdx, 0xb8(%rsp) movq 0x8(%rdx), %rcx movl 0x130(%r13,%rsi,4), %edx vmovss 0x1c0(%rsp,%rsi,4), %xmm2 vmovss 0x1d0(%rsp,%rsi,4), %xmm3 vmovss 0x1e0(%rsp,%rsi,4), %xmm4 vmovss %xmm2, 0xc0(%rsp) vmovss %xmm3, 0xc4(%rsp) vmovss %xmm4, 0xc8(%rsp) vmovss %xmm0, 0xcc(%rsp) vmovss %xmm1, 0xd0(%rsp) movl %edx, 0xd4(%rsp) movq 0x20(%rsp), %rax movl %eax, 0xd8(%rsp) movl (%rcx), %eax movl %eax, 0xdc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xe0(%rsp) vmovss 0x20(%r14), %xmm0 vmovss %xmm0, 0x20(%rsp) vmovss 0x1b0(%rsp,%rsi,4), %xmm0 vmovss %xmm0, 0x20(%r14) orl $-0x1, 0x1c(%rsp) leaq 0x1c(%rsp), %rax movq %rax, 0xf0(%rsp) movq 0x40(%rsp), %rdx movq 0x18(%rdx), %rax movq %rax, 0xf8(%rsp) movq %rcx, 0x100(%rsp) movq %r14, 0x108(%rsp) leaq 0xc0(%rsp), %rax movq %rax, 0x110(%rsp) movl $0x1, 0x118(%rsp) movq 0x48(%rdx), %rax testq %rax, %rax je 0x2078a4 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x2078e5 movq 0x58(%rsp), %rax movq 0x10(%rax), %rax testq %rax, %rax je 0x2078e1 movq 0x58(%rsp), %rcx testb $0x2, (%rcx) jne 0x2078c7 movq 0x40(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x2078d4 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x2078e5 xorl %eax, %eax jmp 0x207904 vmovss 0x20(%rsp), %xmm0 vmovss %xmm0, 0x20(%r14) movq (%rsp), %rax movq 0x30(%rsp), %rcx btcq %rcx, %rax movq %rax, (%rsp) movb $0x1, %al movq 0xb8(%rsp), %rdx movq 0xb0(%rsp), %rdi movq 0xa8(%rsp), %r8 movq 0xa0(%rsp), %r9 movq 0x98(%rsp), %r10 movq 0x90(%rsp), %r11 xorl %ecx, %ecx movq 0x88(%rsp), %rsi jmp 0x207735 vmovups 0x2d0(%rsp), %ymm6 vmovups 0x2b0(%rsp), %ymm7 vmovups 0x290(%rsp), %ymm8 vmovups 0x270(%rsp), %ymm9 vmovups 0x250(%rsp), %ymm10 vmovups 0x230(%rsp), %ymm5 movq 0x68(%rsp), %r15 movq 0x60(%rsp), %r13 vmovups 0x210(%rsp), %ymm11 vmovups 0x1f0(%rsp), %ymm12 leaq 0x2f0(%rsp), %rbp cmpl $0x3, %ecx jne 0x2071bc jmp 0x20702b testb $0x1, %bpl vmovups 0x2d0(%rsp), %ymm6 vmovups 0x2b0(%rsp), %ymm7 vmovups 0x290(%rsp), %ymm8 vmovups 0x270(%rsp), %ymm9 vmovups 0x250(%rsp), %ymm10 vmovups 0x230(%rsp), %ymm5 movq 0x68(%rsp), %r15 movq 0x60(%rsp), %r13 vmovups 0x210(%rsp), %ymm11 vmovups 0x1f0(%rsp), %ymm12 leaq 0x2f0(%rsp), %rbp je 0x20799d movl $0xff800000, 0x20(%r14) # imm = 0xFF800000 pushq $0x3 popq %rcx jmp 0x20799d	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMiMBIntersector1Moeller<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x14f8, %rsp # imm = 0x14F8 movq %rdx, 0x8(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x207a4d addq $0x14f8, %rsp # imm = 0x14F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq vmovss 0x20(%rsi), %xmm0 vxorps %xmm5, %xmm5, %xmm5 vucomiss %xmm0, %xmm5 ja 0x207a38 leaq 0x358(%rsp), %rdi movq 0x70(%rax), %rax vmovaps 0x10(%rsi), %xmm2 vmaxss 0xc(%rsi), %xmm5, %xmm1 vbroadcastss 0x1d19449(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1ce9560(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vblendvps %xmm3, %xmm4, %xmm2, %xmm2 movq %rax, -0x8(%rdi) vrcpps %xmm2, %xmm3 vmulps %xmm2, %xmm3, %xmm2 vbroadcastss 0x1ce4c6c(%rip), %xmm4 # 0x1eec714 vsubps %xmm2, %xmm4, %xmm2 vmulps %xmm2, %xmm3, %xmm2 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vbroadcastss 0x8(%rsi), %ymm8 vaddps %xmm2, %xmm3, %xmm2 xorl %r8d, %r8d vucomiss %xmm5, %xmm2 setb %r8b vshufps $0x0, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm9 vmovshdup %xmm2, %xmm3 # xmm3 = xmm2[1,1,3,3] vshufps $0x55, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm10 vshufpd $0x1, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[1,0] vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm5, %xmm3 vinsertf128 $0x1, %xmm2, %ymm2, %ymm11 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm5, %xmm4 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %rbx xorq $0x20, %rbx movq %r10, %r14 xorq $0x20, %r14 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm12 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm13 leaq 0x1f48427(%rip), %rax # 0x214ff80 vmovdqa 0xf0(%rax), %xmm0 vmovdqa %xmm0, 0x140(%rsp) leaq 0x350(%rsp), %r15 movq %rsi, 0x10(%rsp) vmovups %ymm6, 0x330(%rsp) vmovups %ymm7, 0x310(%rsp) vmovups %ymm8, 0x2f0(%rsp) movq %r8, 0x80(%rsp) vmovups %ymm9, 0x2d0(%rsp) vmovups %ymm10, 0x2b0(%rsp) movq %r9, 0x78(%rsp) vmovups %ymm11, 0x290(%rsp) movq %r10, 0x70(%rsp) movq %r11, 0x68(%rsp) movq %rbx, 0x60(%rsp) movq %r14, 0x58(%rsp) vmovups %ymm12, 0x270(%rsp) vmovups %ymm13, 0x250(%rsp) cmpq %r15, %rdi je 0x207a38 movq -0x8(%rdi), %r13 addq $-0x8, %rdi testb $0x8, %r13b jne 0x207ce1 movq %r13, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rsi), %ymm0 vmulps 0x100(%rax,%r8), %ymm0, %ymm1 vaddps 0x40(%rax,%r8), %ymm1, %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmulps 0x100(%rax,%r9), %ymm0, %ymm2 vaddps 0x40(%rax,%r9), %ymm2, %ymm2 vmaxps %ymm1, %ymm12, %ymm1 vsubps %ymm7, %ymm2, %ymm2 vmulps 0x100(%rax,%r10), %ymm0, %ymm3 vmulps %ymm2, %ymm10, %ymm2 vaddps 0x40(%rax,%r10), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm11, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r11), %ymm3, %ymm2 vsubps %ymm6, %ymm2, %ymm2 vmulps %ymm2, %ymm9, %ymm2 vminps %ymm2, %ymm13, %ymm2 vmulps 0x100(%rax,%rbx), %ymm0, %ymm3 vaddps 0x40(%rax,%rbx), %ymm3, %ymm3 vsubps %ymm7, %ymm3, %ymm3 vmulps %ymm3, %ymm10, %ymm3 vmulps 0x100(%rax,%r14), %ymm0, %ymm4 vaddps 0x40(%rax,%r14), %ymm4, %ymm4 vsubps %ymm8, %ymm4, %ymm4 vmulps %ymm4, %ymm11, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r13d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x207d2b vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r12d testb $0x8, %r13b jne 0x207d27 testq %r12, %r12 je 0x207d58 andq $-0x10, %r13 bsfq %r12, %rax leaq -0x1(%r12), %rdx xorl %ebp, %ebp movq (%r13,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r12, %rdx jne 0x207d5d movq %rax, %r13 testl %ebp, %ebp je 0x207bf1 jmp 0x207da4 pushq $0x6 jmp 0x207d5a vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x207cd0 pushq $0x4 popq %rbp jmp 0x207d1d movq %rax, (%rdi) addq $0x8, %rdi bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r13,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x207d9c movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x207d6c movq %rcx, %r13 jmp 0x207d1d cmpl $0x6, %ebp jne 0x20860c movl %r13d, %eax andl $0xf, %eax xorl %ebp, %ebp addq $-0x8, %rax movq %rax, 0xd0(%rsp) setne %cl je 0x20860c movq %r12, 0x88(%rsp) movq %rdi, 0x90(%rsp) andq $-0x10, %r13 movq 0x8(%rsp), %rax movq (%rax), %rax movq %rax, 0x98(%rsp) xorl %eax, %eax movb %cl, 0x7(%rsp) movq %rax, 0xd8(%rsp) imulq $0x50, %rax, %rax vmovss 0x1c(%rsi), %xmm0 movl 0x30(%r13,%rax), %ecx movq 0x98(%rsp), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rcx,8), %rcx vmovss 0x28(%rcx), %xmm1 vmovss 0x2c(%rcx), %xmm2 vmovss 0x30(%rcx), %xmm3 vsubss %xmm2, %xmm0, %xmm0 vsubss %xmm2, %xmm3, %xmm2 vdivss %xmm2, %xmm0, %xmm0 vmulss %xmm0, %xmm1, %xmm6 vroundss $0x9, %xmm6, %xmm6, %xmm0 vaddss 0x1ce8b82(%rip), %xmm1, %xmm1 # 0x1ef09cc vminss %xmm1, %xmm0, %xmm0 vmaxss %xmm0, %xmm5, %xmm2 vcvttss2si %xmm2, %edx movslq %edx, %rdx movq 0xe0(%rcx), %rcx imulq $0x38, %rdx, %rsi movl (%r13,%rax), %ebp movl 0x4(%r13,%rax), %edx movq (%rcx,%rsi), %r12 movq 0x38(%rcx,%rsi), %rcx vmovups (%r12,%rbp,4), %xmm4 movl 0x10(%r13,%rax), %r15d vmovups (%r12,%r15,4), %xmm5 movl 0x20(%r13,%rax), %esi movq %rsi, 0x40(%rsp) vmovups (%r12,%rsi,4), %xmm0 vmovaps %xmm0, 0x30(%rsp) vmovups (%r12,%rdx,4), %xmm8 movl 0x14(%r13,%rax), %edi vmovups (%r12,%rdi,4), %xmm7 movl 0x24(%r13,%rax), %esi vmovups (%r12,%rsi,4), %xmm0 vmovaps %xmm0, 0x20(%rsp) movl 0x8(%r13,%rax), %ebx vmovups (%r12,%rbx,4), %xmm10 movl 0x18(%r13,%rax), %r10d vmovups (%r12,%r10,4), %xmm9 movl 0x28(%r13,%rax), %r8d vmovups (%r12,%r8,4), %xmm11 movl 0xc(%r13,%rax), %r14d vmovups (%r12,%r14,4), %xmm1 movl 0x1c(%r13,%rax), %r11d vmovups (%r12,%r11,4), %xmm14 movl 0x2c(%r13,%rax), %r9d vmovups (%r12,%r9,4), %xmm12 vmovups (%rcx,%rbp,4), %xmm15 vmovups (%rcx,%r15,4), %xmm13 vsubss %xmm2, %xmm6, %xmm6 vunpcklps %xmm10, %xmm4, %xmm2 # xmm2 = xmm4[0],xmm10[0],xmm4[1],xmm10[1] vunpckhps %xmm10, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm10[2],xmm4[3],xmm10[3] vunpcklps %xmm1, %xmm8, %xmm0 # xmm0 = xmm8[0],xmm1[0],xmm8[1],xmm1[1] vunpckhps %xmm1, %xmm8, %xmm1 # xmm1 = xmm8[2],xmm1[2],xmm8[3],xmm1[3] vmovups (%rcx,%rdx,4), %xmm3 vunpcklps %xmm1, %xmm4, %xmm1 # xmm1 = xmm4[0],xmm1[0],xmm4[1],xmm1[1] vmovaps %xmm1, 0x160(%rsp) vunpcklps %xmm0, %xmm2, %xmm8 # xmm8 = xmm2[0],xmm0[0],xmm2[1],xmm0[1] vunpckhps %xmm0, %xmm2, %xmm0 # xmm0 = xmm2[2],xmm0[2],xmm2[3],xmm0[3] vmovaps %xmm0, 0xb0(%rsp) vunpcklps %xmm9, %xmm5, %xmm0 # xmm0 = xmm5[0],xmm9[0],xmm5[1],xmm9[1] vunpckhps %xmm9, %xmm5, %xmm1 # xmm1 = xmm5[2],xmm9[2],xmm5[3],xmm9[3] vunpcklps %xmm14, %xmm7, %xmm2 # xmm2 = xmm7[0],xmm14[0],xmm7[1],xmm14[1] vunpckhps %xmm14, %xmm7, %xmm5 # xmm5 = xmm7[2],xmm14[2],xmm7[3],xmm14[3] vmovups (%rcx,%rbx,4), %xmm14 vunpcklps %xmm5, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm5[0],xmm1[1],xmm5[1] vmovaps %xmm1, 0xa0(%rsp) vunpcklps %xmm2, %xmm0, %xmm9 # xmm9 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x150(%rsp) vmovaps 0x30(%rsp), %xmm1 vunpcklps %xmm11, %xmm1, %xmm0 # xmm0 = xmm1[0],xmm11[0],xmm1[1],xmm11[1] vunpckhps %xmm11, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm11[2],xmm1[3],xmm11[3] vmovaps 0x20(%rsp), %xmm2 vunpcklps %xmm12, %xmm2, %xmm5 # xmm5 = xmm2[0],xmm12[0],xmm2[1],xmm12[1] vunpckhps %xmm12, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm12[2],xmm2[3],xmm12[3] vmovups (%rcx,%r14,4), %xmm11 vunpcklps %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] vmovaps %xmm1, 0x30(%rsp) vunpcklps %xmm5, %xmm0, %xmm1 # xmm1 = xmm0[0],xmm5[0],xmm0[1],xmm5[1] vmovaps %xmm1, 0xc0(%rsp) vunpckhps %xmm5, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm5[2],xmm0[3],xmm5[3] vmovaps %xmm0, 0x20(%rsp) vunpcklps %xmm14, %xmm15, %xmm0 # xmm0 = xmm15[0],xmm14[0],xmm15[1],xmm14[1] vunpckhps %xmm14, %xmm15, %xmm1 # xmm1 = xmm15[2],xmm14[2],xmm15[3],xmm14[3] vunpcklps %xmm11, %xmm3, %xmm5 # xmm5 = xmm3[0],xmm11[0],xmm3[1],xmm11[1] vunpckhps %xmm11, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm11[2],xmm3[3],xmm11[3] vmovups (%rcx,%r10,4), %xmm11 vunpcklps %xmm3, %xmm1, %xmm4 # xmm4 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vunpcklps %xmm5, %xmm0, %xmm3 # xmm3 = xmm0[0],xmm5[0],xmm0[1],xmm5[1] vunpckhps %xmm5, %xmm0, %xmm2 # xmm2 = xmm0[2],xmm5[2],xmm0[3],xmm5[3] vunpcklps %xmm11, %xmm13, %xmm0 # xmm0 = xmm13[0],xmm11[0],xmm13[1],xmm11[1] vunpckhps %xmm11, %xmm13, %xmm5 # xmm5 = xmm13[2],xmm11[2],xmm13[3],xmm11[3] vmovups (%rcx,%rdi,4), %xmm11 vmovups (%rcx,%r11,4), %xmm12 vunpcklps %xmm12, %xmm11, %xmm13 # xmm13 = xmm11[0],xmm12[0],xmm11[1],xmm12[1] vunpckhps %xmm12, %xmm11, %xmm11 # xmm11 = xmm11[2],xmm12[2],xmm11[3],xmm12[3] vunpcklps %xmm11, %xmm5, %xmm11 # xmm11 = xmm5[0],xmm11[0],xmm5[1],xmm11[1] vunpcklps %xmm13, %xmm0, %xmm1 # xmm1 = xmm0[0],xmm13[0],xmm0[1],xmm13[1] vunpckhps %xmm13, %xmm0, %xmm7 # xmm7 = xmm0[2],xmm13[2],xmm0[3],xmm13[3] movq 0x40(%rsp), %rdx vmovups (%rcx,%rdx,4), %xmm0 vmovups (%rcx,%r8,4), %xmm5 vunpcklps %xmm5, %xmm0, %xmm13 # xmm13 = xmm0[0],xmm5[0],xmm0[1],xmm5[1] vunpckhps %xmm5, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm5[2],xmm0[3],xmm5[3] vmovups (%rcx,%rsi,4), %xmm5 movq 0x10(%rsp), %rsi vmovups (%rcx,%r9,4), %xmm12 vunpcklps %xmm12, %xmm5, %xmm15 # xmm15 = xmm5[0],xmm12[0],xmm5[1],xmm12[1] vunpckhps %xmm12, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm12[2],xmm5[3],xmm12[3] vunpcklps %xmm5, %xmm0, %xmm10 # xmm10 = xmm0[0],xmm5[0],xmm0[1],xmm5[1] vunpcklps %xmm15, %xmm13, %xmm14 # xmm14 = xmm13[0],xmm15[0],xmm13[1],xmm15[1] vunpckhps %xmm15, %xmm13, %xmm15 # xmm15 = xmm13[2],xmm15[2],xmm13[3],xmm15[3] vshufps $0x0, %xmm6, %xmm6, %xmm0 # xmm0 = xmm6[0,0,0,0] vmovss 0x1ce46af(%rip), %xmm5 # 0x1eec714 vsubss %xmm6, %xmm5, %xmm5 vshufps $0x0, %xmm5, %xmm5, %xmm13 # xmm13 = xmm5[0,0,0,0] vmulps %xmm3, %xmm0, %xmm3 vmulps %xmm8, %xmm13, %xmm5 vaddps %xmm3, %xmm5, %xmm12 vmulps %xmm2, %xmm0, %xmm2 vmulps 0xb0(%rsp), %xmm13, %xmm3 vaddps %xmm2, %xmm3, %xmm5 vmulps %xmm4, %xmm0, %xmm2 vmulps 0x160(%rsp), %xmm13, %xmm3 vaddps %xmm2, %xmm3, %xmm6 vmulps %xmm1, %xmm0, %xmm1 vmulps %xmm9, %xmm13, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vmulps %xmm7, %xmm0, %xmm2 vmulps 0x150(%rsp), %xmm13, %xmm3 vaddps %xmm2, %xmm3, %xmm2 vmovaps 0x30(%r13,%rax), %xmm3 vmovaps %xmm3, 0x180(%rsp) vmovaps 0x40(%r13,%rax), %xmm3 movq 0xd8(%rsp), %rax vmulps %xmm0, %xmm11, %xmm7 vmulps 0xa0(%rsp), %xmm13, %xmm8 vaddps %xmm7, %xmm8, %xmm9 vmulps %xmm0, %xmm14, %xmm7 vmulps %xmm0, %xmm15, %xmm8 vmulps %xmm0, %xmm10, %xmm0 vmulps 0xc0(%rsp), %xmm13, %xmm10 vaddps %xmm7, %xmm10, %xmm10 vmulps 0x20(%rsp), %xmm13, %xmm7 vaddps %xmm7, %xmm8, %xmm11 vmulps 0x30(%rsp), %xmm13, %xmm7 vaddps %xmm0, %xmm7, %xmm0 vmovaps %xmm3, 0x170(%rsp) vsubps %xmm1, %xmm12, %xmm7 vmovaps %xmm5, %xmm1 vmovaps %xmm5, 0xb0(%rsp) vsubps %xmm2, %xmm5, %xmm8 vsubps %xmm9, %xmm6, %xmm9 vsubps %xmm12, %xmm10, %xmm10 vmovaps %xmm12, %xmm5 vsubps %xmm1, %xmm11, %xmm11 vsubps %xmm6, %xmm0, %xmm12 vmulps %xmm12, %xmm8, %xmm0 vmulps %xmm11, %xmm9, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x40(%rsp) vmulps %xmm10, %xmm9, %xmm1 vmulps %xmm7, %xmm12, %xmm2 vsubps %xmm1, %xmm2, %xmm4 vmulps %xmm7, %xmm11, %xmm2 vmulps %xmm10, %xmm8, %xmm3 vsubps %xmm2, %xmm3, %xmm3 vbroadcastss (%rsi), %xmm0 vbroadcastss 0x4(%rsi), %xmm13 vbroadcastss 0x8(%rsi), %xmm14 vbroadcastss 0x14(%rsi), %xmm15 vsubps %xmm0, %xmm5, %xmm2 vbroadcastss 0x18(%rsi), %xmm0 vmovaps 0xb0(%rsp), %xmm1 vsubps %xmm13, %xmm1, %xmm5 vsubps %xmm14, %xmm6, %xmm1 vmulps %xmm1, %xmm15, %xmm6 vmulps %xmm0, %xmm5, %xmm13 vsubps %xmm6, %xmm13, %xmm6 vbroadcastss 0x10(%rsi), %xmm13 vmulps %xmm0, %xmm2, %xmm14 vmovaps %xmm1, 0xc0(%rsp) vmulps %xmm1, %xmm13, %xmm1 vsubps %xmm14, %xmm1, %xmm1 vmovaps %xmm5, 0x20(%rsp) vmulps %xmm5, %xmm13, %xmm14 vmovaps %xmm2, 0x30(%rsp) vmulps %xmm2, %xmm15, %xmm2 vsubps %xmm14, %xmm2, %xmm2 vmovaps %xmm3, 0xa0(%rsp) vmulps %xmm0, %xmm3, %xmm0 vmulps %xmm4, %xmm15, %xmm14 vaddps %xmm0, %xmm14, %xmm0 vmovaps 0x40(%rsp), %xmm14 vmulps %xmm13, %xmm14, %xmm13 vaddps %xmm0, %xmm13, %xmm0 vmulps %xmm2, %xmm12, %xmm12 vmulps %xmm1, %xmm11, %xmm11 vaddps %xmm12, %xmm11, %xmm11 vmulps %xmm6, %xmm10, %xmm10 vaddps %xmm11, %xmm10, %xmm10 vmulps %xmm2, %xmm9, %xmm2 vmulps %xmm1, %xmm8, %xmm1 vaddps %xmm2, %xmm1, %xmm1 vmovddup 0x1d18d2b(%rip), %xmm2 # xmm2 = mem[0,0] vandps %xmm2, %xmm0, %xmm9 vxorps %xmm10, %xmm9, %xmm8 vmulps %xmm6, %xmm7, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vxorps %xmm1, %xmm9, %xmm7 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %xmm10, %xmm8, %xmm1 vcmpnltps %xmm10, %xmm7, %xmm2 vandps %xmm2, %xmm1, %xmm1 vbroadcastss 0x1d18c54(%rip), %xmm2 # 0x1f20ec4 vandps %xmm2, %xmm0, %xmm6 vcmpneqps %xmm0, %xmm10, %xmm0 vandps %xmm0, %xmm1, %xmm0 vaddps %xmm7, %xmm8, %xmm1 vcmpleps %xmm6, %xmm1, %xmm1 vandps %xmm1, %xmm0, %xmm10 vtestps 0x140(%rsp), %xmm10 jne 0x2082b3 incq %rax cmpq 0xd0(%rsp), %rax setb %cl vxorps %xmm5, %xmm5, %xmm5 jne 0x207df0 jmp 0x208589 vmovaps %xmm4, %xmm15 vandps 0x140(%rsp), %xmm10, %xmm10 vmovaps 0xc0(%rsp), %xmm0 vmulps 0xa0(%rsp), %xmm0, %xmm0 vmulps 0x20(%rsp), %xmm4, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x30(%rsp), %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm3 vbroadcastss 0xc(%rsi), %xmm0 vmulps %xmm0, %xmm6, %xmm0 vcmpltps %xmm3, %xmm0, %xmm0 vbroadcastss 0x20(%rsi), %xmm1 vmulps %xmm1, %xmm6, %xmm1 vcmpleps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm4 vtestps %xmm10, %xmm4 je 0x208296 vandps %xmm4, %xmm10, %xmm0 vmovaps %xmm8, 0x190(%rsp) vmovaps %xmm7, 0x1a0(%rsp) vmovaps %xmm3, 0x1b0(%rsp) vmovaps %xmm6, 0x1c0(%rsp) vmovaps %xmm0, 0x1e0(%rsp) vmovaps %xmm14, 0x220(%rsp) vmovaps %xmm15, 0x230(%rsp) vmovaps 0xa0(%rsp), %xmm1 vmovaps %xmm1, 0x240(%rsp) vrcpps %xmm6, %xmm1 vmulps %xmm1, %xmm6, %xmm2 vbroadcastss 0x1ce439b(%rip), %xmm3 # 0x1eec714 vsubps %xmm2, %xmm3, %xmm2 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps 0x1b0(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x210(%rsp) vmulps 0x190(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x1f0(%rsp) vmulps 0x1a0(%rsp), %xmm1, %xmm1 vmovaps %xmm1, 0x200(%rsp) vmovmskps %xmm0, %r15d movq %rax, %rbx bsfq %r15, %r12 movl 0x180(%rsp,%r12,4), %eax movq 0x98(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rbp movl 0x24(%rsi), %ecx testl %ecx, 0x34(%rbp) je 0x208404 movq 0x8(%rsp), %rcx movq 0x10(%rcx), %r14 cmpq $0x0, 0x10(%r14) jne 0x20841f cmpq $0x0, 0x48(%rbp) jne 0x20841f xorl %eax, %eax jmp 0x20840a btcq %r12, %r15 movb $0x1, %al testb %al, %al je 0x20861a testq %r15, %r15 movq %rbx, %rax jne 0x2083bf jmp 0x208296 vmovss 0x1f0(%rsp,%r12,4), %xmm0 vmovss 0x200(%rsp,%r12,4), %xmm1 movq 0x8(%rsp), %rcx movq 0x8(%rcx), %rcx movl 0x170(%rsp,%r12,4), %edx vmovss 0x220(%rsp,%r12,4), %xmm2 vmovss 0x230(%rsp,%r12,4), %xmm3 vmovss 0x240(%rsp,%r12,4), %xmm4 vmovss %xmm2, 0xe0(%rsp) vmovss %xmm3, 0xe4(%rsp) vmovss %xmm4, 0xe8(%rsp) vmovss %xmm0, 0xec(%rsp) vmovss %xmm1, 0xf0(%rsp) movl %edx, 0xf4(%rsp) movl %eax, 0xf8(%rsp) movl (%rcx), %eax movl %eax, 0xfc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0x100(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0x40(%rsp) vmovss 0x210(%rsp,%r12,4), %xmm0 vmovss %xmm0, 0x20(%rsi) orl $-0x1, 0x1c(%rsp) leaq 0x1c(%rsp), %rax movq %rax, 0x110(%rsp) movq 0x18(%rbp), %rax movq %rax, 0x118(%rsp) movq %rcx, 0x120(%rsp) movq %rsi, 0x128(%rsp) leaq 0xe0(%rsp), %rax movq %rax, 0x130(%rsp) movl $0x1, 0x138(%rsp) movq 0x48(%rbp), %rax testq %rax, %rax je 0x208536 leaq 0x110(%rsp), %rdi vzeroupper callq %rax movq 0x110(%rsp), %rax cmpl $0x0, (%rax) je 0x208569 movq 0x10(%r14), %rax testq %rax, %rax je 0x208565 testb $0x2, (%r14) jne 0x20854b testb $0x40, 0x3e(%rbp) je 0x208558 leaq 0x110(%rsp), %rdi vzeroupper callq %rax movq 0x110(%rsp), %rax cmpl $0x0, (%rax) je 0x208569 xorl %eax, %eax jmp 0x20857f movq 0x10(%rsp), %rax vmovss 0x40(%rsp), %xmm0 vmovss %xmm0, 0x20(%rax) btcq %r12, %r15 movb $0x1, %al movq 0x10(%rsp), %rsi jmp 0x20840a movq 0x90(%rsp), %rdi vmovups 0x330(%rsp), %ymm6 vmovups 0x310(%rsp), %ymm7 vmovups 0x2f0(%rsp), %ymm8 movq 0x80(%rsp), %r8 vmovups 0x2d0(%rsp), %ymm9 vmovups 0x2b0(%rsp), %ymm10 movq 0x78(%rsp), %r9 vmovups 0x290(%rsp), %ymm11 movq 0x70(%rsp), %r10 movq 0x68(%rsp), %r11 movq 0x60(%rsp), %rbx movq 0x58(%rsp), %r14 vmovups 0x270(%rsp), %ymm12 vmovups 0x250(%rsp), %ymm13 leaq 0x350(%rsp), %r15 movq 0x88(%rsp), %r12 xorl %ebp, %ebp cmpl $0x3, %ebp jne 0x207be0 jmp 0x207a38 testb $0x1, 0x7(%rsp) vxorps %xmm5, %xmm5, %xmm5 movq 0x90(%rsp), %rdi vmovups 0x330(%rsp), %ymm6 vmovups 0x310(%rsp), %ymm7 vmovups 0x2f0(%rsp), %ymm8 movq 0x80(%rsp), %r8 vmovups 0x2d0(%rsp), %ymm9 vmovups 0x2b0(%rsp), %ymm10 movq 0x78(%rsp), %r9 vmovups 0x290(%rsp), %ymm11 movq 0x70(%rsp), %r10 movq 0x68(%rsp), %r11 movq 0x60(%rsp), %rbx movq 0x58(%rsp), %r14 vmovups 0x270(%rsp), %ymm12 vmovups 0x250(%rsp), %ymm13 leaq 0x350(%rsp), %r15 movq 0x88(%rsp), %r12 movl $0x0, %ebp je 0x20860c movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rbp jmp 0x20860c	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, true, embree::avx::ArrayIntersector1<embree::avx::TriangleMvMBIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1568, %rsp # imm = 0x1568 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x2092de vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x2092de leaq 0x3c8(%rsp), %r8 movq 0x70(%rax), %rax movq %rax, -0x8(%r8) vmovaps 0x10(%rsi), %xmm3 vmaxss 0xc(%rsi), %xmm2, %xmm1 vbroadcastss 0x1d187b1(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce88c8(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vbroadcastss 0x1ce3fe6(%rip), %xmm5 # 0x1eec714 vdivps %xmm3, %xmm5, %xmm3 vbroadcastss 0x1d18825(%rip), %xmm5 # 0x1f20f60 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vbroadcastss 0x1d177c6(%rip), %xmm4 # 0x1f1ff10 vmulps %xmm4, %xmm3, %xmm4 vbroadcastss (%rsi), %ymm7 vbroadcastss 0x4(%rsi), %ymm5 vmovups %ymm5, 0x3a0(%rsp) vbroadcastss 0x8(%rsi), %ymm5 vmovups %ymm5, 0x380(%rsp) vbroadcastss 0x1d1779a(%rip), %xmm5 # 0x1f1ff14 xorl %r9d, %r9d vucomiss %xmm2, %xmm4 vmulps %xmm5, %xmm3, %xmm5 setb %r9b vshufps $0x0, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovups %ymm3, 0x360(%rsp) vmovshdup %xmm4, %xmm6 # xmm6 = xmm4[1,1,3,3] vshufps $0x55, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[1,1,1,1] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovups %ymm3, 0x340(%rsp) vshufpd $0x1, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[1,0] vshufps $0xaa, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x320(%rsp) vshufps $0x0, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x300(%rsp) shll $0x5, %r9d xorl %r10d, %r10d vucomiss %xmm2, %xmm6 vshufps $0x55, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x2e0(%rsp) vshufps $0xaa, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[2,2,2,2] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x2c0(%rsp) setb %r10b shll $0x5, %r10d orq $0x40, %r10 xorl %r11d, %r11d vucomiss %xmm2, %xmm3 setb %r11b shll $0x5, %r11d orq $0x80, %r11 movq %r9, %r14 xorq $0x20, %r14 movq %r10, %r15 xorq $0x20, %r15 movq %r11, %r13 xorq $0x20, %r13 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovdqu %ymm1, 0x2a0(%rsp) vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovups %ymm0, 0x280(%rsp) leaq 0x1f47705(%rip), %rax # 0x214ff80 vmovdqa 0xf0(%rax), %xmm0 vmovdqa %xmm0, 0x110(%rsp) leaq 0x3c0(%rsp), %rbp vmovups %ymm7, 0x120(%rsp) movq %r14, 0x70(%rsp) movq %r15, 0x68(%rsp) movq %r13, 0x60(%rsp) cmpq %rbp, %r8 je 0x2092de movq -0x8(%r8), %r12 addq $-0x8, %r8 testb $0x8, %r12b jne 0x2089ee movq %r12, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rsi), %ymm0 vmulps 0x100(%rax,%r9), %ymm0, %ymm1 vaddps 0x40(%rax,%r9), %ymm1, %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps 0x360(%rsp), %ymm1, %ymm1 vmulps 0x100(%rax,%r10), %ymm0, %ymm2 vaddps 0x40(%rax,%r10), %ymm2, %ymm2 vmovups 0x2a0(%rsp), %ymm3 vmaxps %ymm1, %ymm3, %ymm1 vmovups 0x3a0(%rsp), %ymm4 vsubps %ymm4, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmulps 0x340(%rsp), %ymm2, %ymm2 vaddps 0x40(%rax,%r11), %ymm3, %ymm3 vmovups 0x380(%rsp), %ymm5 vsubps %ymm5, %ymm3, %ymm3 vmulps 0x320(%rsp), %ymm3, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%r14), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r14), %ymm3, %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps 0x300(%rsp), %ymm2, %ymm2 vmovups 0x280(%rsp), %ymm3 vminps %ymm2, %ymm3, %ymm2 vmulps 0x100(%rax,%r15), %ymm0, %ymm3 vaddps 0x40(%rax,%r15), %ymm3, %ymm3 vsubps %ymm4, %ymm3, %ymm3 vmulps 0x2e0(%rsp), %ymm3, %ymm3 vmulps 0x100(%rax,%r13), %ymm0, %ymm4 vaddps 0x40(%rax,%r13), %ymm4, %ymm4 vsubps %ymm5, %ymm4, %ymm4 vmulps 0x2c0(%rsp), %ymm4, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r12d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x208a36 vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %ebx testb $0x8, %r12b jne 0x208a32 testq %rbx, %rbx je 0x208a63 andq $-0x10, %r12 bsfq %rbx, %rax leaq -0x1(%rbx), %rdi xorl %ecx, %ecx movq (%r12,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %rbx, %rdi jne 0x208a68 movq %rax, %r12 testl %ecx, %ecx je 0x2088bd jmp 0x208ab0 pushq $0x6 jmp 0x208a65 vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x2089de pushq $0x4 popq %rcx jmp 0x208a28 movq %rax, (%r8) addq $0x8, %r8 bsfq %rdi, %rcx leaq -0x1(%rdi), %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdi, %rax je 0x208aa6 movq %rcx, (%r8) addq $0x8, %r8 bsfq %rax, %rcx leaq -0x1(%rax), %rdi jmp 0x208a77 movq %rcx, %r12 xorl %ecx, %ecx jmp 0x208a28 cmpl $0x6, %ecx jne 0x2092d5 movl %r12d, %edi andl $0xf, %edi xorl %ecx, %ecx addq $-0x8, %rdi setne %al je 0x2092d5 andq $-0x10, %r12 xorl %r14d, %r14d movb %al, 0xe(%rsp) imulq $0x140, %r14, %rax # imm = 0x140 vbroadcastss 0x1c(%rsi), %xmm0 vmulps 0x90(%r12,%rax), %xmm0, %xmm1 vaddps (%r12,%rax), %xmm1, %xmm1 vmulps 0xa0(%r12,%rax), %xmm0, %xmm2 vaddps 0x10(%r12,%rax), %xmm2, %xmm2 vmulps 0xb0(%r12,%rax), %xmm0, %xmm3 vaddps 0x20(%r12,%rax), %xmm3, %xmm11 vmulps 0xc0(%r12,%rax), %xmm0, %xmm4 vaddps 0x30(%r12,%rax), %xmm4, %xmm4 vmulps 0xd0(%r12,%rax), %xmm0, %xmm5 vaddps 0x40(%r12,%rax), %xmm5, %xmm5 vmulps 0xe0(%r12,%rax), %xmm0, %xmm6 vaddps 0x50(%r12,%rax), %xmm6, %xmm12 vmulps 0xf0(%r12,%rax), %xmm0, %xmm6 vaddps 0x60(%r12,%rax), %xmm6, %xmm8 vmulps 0x100(%r12,%rax), %xmm0, %xmm6 vaddps 0x70(%r12,%rax), %xmm6, %xmm9 vmulps 0x110(%r12,%rax), %xmm0, %xmm0 movq %rax, 0x10(%rsp) vaddps 0x80(%r12,%rax), %xmm0, %xmm0 vbroadcastss (%rsi), %xmm10 vbroadcastss 0x4(%rsi), %xmm13 vbroadcastss 0x8(%rsi), %xmm14 vsubps %xmm10, %xmm1, %xmm7 vsubps %xmm13, %xmm2, %xmm3 vsubps %xmm14, %xmm11, %xmm1 vsubps %xmm10, %xmm4, %xmm2 vmovaps %xmm2, 0x30(%rsp) vsubps %xmm13, %xmm5, %xmm11 vsubps %xmm14, %xmm12, %xmm5 vsubps %xmm10, %xmm8, %xmm12 vsubps %xmm13, %xmm9, %xmm6 vsubps %xmm14, %xmm0, %xmm4 vmovaps %xmm4, 0x20(%rsp) vsubps %xmm7, %xmm12, %xmm14 vsubps %xmm3, %xmm6, %xmm15 vsubps %xmm1, %xmm4, %xmm0 vaddps %xmm3, %xmm6, %xmm2 vaddps %xmm1, %xmm4, %xmm4 vmulps %xmm2, %xmm0, %xmm8 vmulps %xmm4, %xmm15, %xmm9 vsubps %xmm8, %xmm9, %xmm13 vaddps %xmm7, %xmm12, %xmm8 vmulps %xmm4, %xmm14, %xmm4 vmovaps %xmm0, 0x190(%rsp) vmulps %xmm0, %xmm8, %xmm9 vsubps %xmm4, %xmm9, %xmm4 vmovaps %xmm15, 0x1a0(%rsp) vmulps %xmm8, %xmm15, %xmm8 vmovaps %xmm14, 0x1b0(%rsp) vmulps %xmm2, %xmm14, %xmm0 vsubps %xmm8, %xmm0, %xmm0 vbroadcastss 0x18(%rsi), %xmm10 vmulps %xmm0, %xmm10, %xmm0 vbroadcastss 0x14(%rsi), %xmm15 vmulps %xmm4, %xmm15, %xmm4 vaddps %xmm0, %xmm4, %xmm0 vbroadcastss 0x10(%rsi), %xmm2 vmovaps %xmm2, 0x80(%rsp) vmulps %xmm2, %xmm13, %xmm4 vaddps %xmm0, %xmm4, %xmm9 vmovaps %xmm11, %xmm4 vsubps %xmm11, %xmm3, %xmm2 vmovaps %xmm5, %xmm8 vsubps %xmm5, %xmm1, %xmm11 vmovaps %xmm3, 0x90(%rsp) vaddps %xmm4, %xmm3, %xmm0 vmovaps %xmm4, %xmm5 vmovaps %xmm1, 0x1c0(%rsp) vaddps %xmm1, %xmm8, %xmm4 vmovaps %xmm8, %xmm1 vmulps %xmm0, %xmm11, %xmm14 vmulps %xmm4, %xmm2, %xmm3 vsubps %xmm14, %xmm3, %xmm3 vmovaps 0x30(%rsp), %xmm14 vsubps %xmm14, %xmm7, %xmm13 vmulps %xmm4, %xmm13, %xmm4 vmovaps %xmm7, 0xa0(%rsp) vaddps %xmm7, %xmm14, %xmm7 vmulps %xmm7, %xmm11, %xmm8 vsubps %xmm4, %xmm8, %xmm4 vmovaps %xmm2, 0x180(%rsp) vmulps %xmm7, %xmm2, %xmm7 vmovaps %xmm13, 0x170(%rsp) vmulps %xmm0, %xmm13, %xmm0 vsubps %xmm7, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vmulps %xmm4, %xmm15, %xmm4 vaddps %xmm0, %xmm4, %xmm0 vmovaps 0x80(%rsp), %xmm13 vmulps %xmm3, %xmm13, %xmm3 vaddps %xmm0, %xmm3, %xmm0 vsubps %xmm12, %xmm14, %xmm8 vaddps %xmm12, %xmm14, %xmm3 vsubps %xmm6, %xmm5, %xmm12 vaddps %xmm6, %xmm5, %xmm2 vmovaps 0x20(%rsp), %xmm5 vsubps %xmm5, %xmm1, %xmm4 vaddps %xmm5, %xmm1, %xmm1 vmulps %xmm2, %xmm4, %xmm5 vmulps %xmm1, %xmm12, %xmm7 vsubps %xmm5, %xmm7, %xmm5 vmulps %xmm1, %xmm8, %xmm1 vmulps %xmm3, %xmm4, %xmm7 vsubps %xmm1, %xmm7, %xmm1 vmovups 0x120(%rsp), %ymm7 vmulps %xmm3, %xmm12, %xmm3 vmulps %xmm2, %xmm8, %xmm2 vsubps %xmm3, %xmm2, %xmm2 vmovaps %xmm10, 0x20(%rsp) vmulps %xmm2, %xmm10, %xmm2 vmovaps %xmm15, 0x30(%rsp) vmulps %xmm1, %xmm15, %xmm1 vaddps %xmm2, %xmm1, %xmm1 vmulps %xmm5, %xmm13, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm0, %xmm9, %xmm2 vaddps %xmm2, %xmm1, %xmm13 vminps %xmm0, %xmm9, %xmm2 vminps %xmm1, %xmm2, %xmm2 vbroadcastss 0x1d18146(%rip), %xmm3 # 0x1f20ec4 vmovaps %xmm3, %xmm10 vandps %xmm3, %xmm13, %xmm5 vbroadcastss 0x1d1813d(%rip), %xmm3 # 0x1f20ecc vmovaps %xmm5, 0x140(%rsp) vmulps %xmm3, %xmm5, %xmm3 vbroadcastss 0x1d1811b(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm3, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm9, 0x160(%rsp) vmovaps %xmm0, 0x150(%rsp) vmaxps %xmm0, %xmm9, %xmm5 vmaxps %xmm1, %xmm5, %xmm1 vcmpleps %xmm3, %xmm1, %xmm1 vorps %xmm1, %xmm2, %xmm0 leaq 0xf(%rsp), %rax movq %rax, 0x200(%rsp) vtestps 0x110(%rsp), %xmm0 je 0x209285 vmovaps 0x190(%rsp), %xmm7 vmovaps 0x180(%rsp), %xmm15 vmulps %xmm7, %xmm15, %xmm1 vmovaps 0x1a0(%rsp), %xmm9 vmulps %xmm11, %xmm9, %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmulps %xmm12, %xmm11, %xmm3 vmulps %xmm4, %xmm15, %xmm5 vsubps %xmm3, %xmm5, %xmm5 vandps %xmm1, %xmm10, %xmm1 vandps %xmm3, %xmm10, %xmm3 vcmpltps %xmm3, %xmm1, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm5 vmovaps 0x170(%rsp), %xmm14 vmulps %xmm4, %xmm14, %xmm1 vmulps %xmm7, %xmm14, %xmm2 vmovaps 0x1b0(%rsp), %xmm7 vmulps %xmm7, %xmm11, %xmm3 vsubps %xmm3, %xmm2, %xmm2 vmulps %xmm8, %xmm11, %xmm6 vsubps %xmm1, %xmm6, %xmm6 vandps %xmm3, %xmm10, %xmm3 vandps %xmm1, %xmm10, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm6, %xmm1 vmovaps %xmm0, %xmm4 vmulps %xmm8, %xmm15, %xmm0 vmulps %xmm7, %xmm15, %xmm2 vmulps %xmm14, %xmm9, %xmm3 vmulps %xmm12, %xmm14, %xmm6 vsubps %xmm3, %xmm2, %xmm2 vsubps %xmm0, %xmm6, %xmm6 vandps %xmm3, %xmm10, %xmm3 vandps %xmm0, %xmm10, %xmm0 vcmpltps %xmm0, %xmm3, %xmm0 vblendvps %xmm0, %xmm2, %xmm6, %xmm0 vmulps 0x20(%rsp), %xmm0, %xmm2 vmulps 0x30(%rsp), %xmm1, %xmm3 vaddps %xmm2, %xmm3, %xmm2 vmulps 0x80(%rsp), %xmm5, %xmm3 vaddps %xmm2, %xmm3, %xmm2 vaddps %xmm2, %xmm2, %xmm3 vmulps 0x1c0(%rsp), %xmm0, %xmm2 vmulps 0x90(%rsp), %xmm1, %xmm6 vaddps %xmm2, %xmm6, %xmm2 vmulps 0xa0(%rsp), %xmm5, %xmm6 vaddps %xmm2, %xmm6, %xmm2 vaddps %xmm2, %xmm2, %xmm2 vrcpps %xmm3, %xmm6 vmulps %xmm6, %xmm3, %xmm7 vbroadcastss 0x1ce3815(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm6, %xmm7 vaddps %xmm7, %xmm6, %xmm6 vmulps %xmm6, %xmm2, %xmm2 vbroadcastss 0xc(%rsi), %xmm6 vcmpleps %xmm2, %xmm6, %xmm6 vbroadcastss 0x20(%rsi), %xmm7 vcmpleps %xmm7, %xmm2, %xmm7 vandps %xmm7, %xmm6, %xmm6 vmovups 0x120(%rsp), %ymm7 vcmpneqps 0x1ce2ad5(%rip), %xmm3, %xmm3 # 0x1eeba10 vandps %xmm6, %xmm3, %xmm6 vandps 0x110(%rsp), %xmm4, %xmm3 vpslld $0x1f, %xmm6, %xmm6 vpsrad $0x1f, %xmm6, %xmm6 vtestps %xmm3, %xmm6 je 0x209285 addq %r12, 0x10(%rsp) vandps %xmm3, %xmm6, %xmm3 vmovaps 0x160(%rsp), %xmm6 vmovaps %xmm6, 0x1d0(%rsp) vmovaps 0x150(%rsp), %xmm4 vmovaps %xmm4, 0x1e0(%rsp) vmovaps %xmm13, 0x1f0(%rsp) movq %rax, 0x200(%rsp) vmovaps %xmm3, 0x210(%rsp) vmovaps %xmm2, 0x240(%rsp) vmovaps %xmm5, 0x250(%rsp) vmovaps %xmm1, 0x260(%rsp) vmovaps %xmm0, 0x270(%rsp) movq (%rdx), %rax movq %rax, 0x80(%rsp) vrcpps %xmm13, %xmm0 vmulps %xmm0, %xmm13, %xmm1 vbroadcastss 0x1ce372f(%rip), %xmm2 # 0x1eec714 vsubps %xmm1, %xmm2, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1ce7fee(%rip), %xmm1 # 0x1ef0fe8 vmovaps 0x140(%rsp), %xmm5 vcmpnltps %xmm1, %xmm5, %xmm1 vandps %xmm0, %xmm1, %xmm0 vmulps %xmm0, %xmm6, %xmm1 vminps %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x220(%rsp) vmulps %xmm0, %xmm4, %xmm0 vminps %xmm2, %xmm0, %xmm0 vmovaps %xmm0, 0x230(%rsp) vmovmskps %xmm3, %ebp movq %r14, 0x78(%rsp) bsfq %rbp, %r15 movq 0x10(%rsp), %rax movl 0x120(%rax,%r15,4), %eax movq 0x80(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %r13 movl 0x24(%rsi), %ecx testl %ecx, 0x34(%r13) je 0x209084 movq 0x10(%rdx), %r14 cmpq $0x0, 0x10(%r14) movl $0x0, %ecx jne 0x20909e cmpq $0x0, 0x48(%r13) jne 0x20909e xorl %eax, %eax movq 0x78(%rsp), %r14 jmp 0x20908c btcq %r15, %rbp movb $0x1, %al xorl %ecx, %ecx testb %al, %al je 0x2092ad testq %rbp, %rbp jne 0x209037 jmp 0x209285 movq %rdi, 0x30(%rsp) vmovss 0x220(%rsp,%r15,4), %xmm0 vmovss 0x230(%rsp,%r15,4), %xmm1 movq 0x8(%rdx), %rcx movq 0x10(%rsp), %rdi movl 0x130(%rdi,%r15,4), %edi vmovss 0x250(%rsp,%r15,4), %xmm2 vmovss 0x260(%rsp,%r15,4), %xmm3 vmovss 0x270(%rsp,%r15,4), %xmm4 vmovss %xmm2, 0xb0(%rsp) vmovss %xmm3, 0xb4(%rsp) vmovss %xmm4, 0xb8(%rsp) vmovss %xmm0, 0xbc(%rsp) vmovss %xmm1, 0xc0(%rsp) movl %edi, 0xc4(%rsp) movl %eax, 0xc8(%rsp) movl (%rcx), %eax movl %eax, 0xcc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xd0(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0x20(%rsp) vmovss 0x240(%rsp,%r15,4), %xmm0 vmovss %xmm0, 0x20(%rsi) orl $-0x1, 0x1c(%rsp) leaq 0x1c(%rsp), %rax movq %rax, 0xe0(%rsp) movq 0x18(%r13), %rax movq %rax, 0xe8(%rsp) movq %rcx, 0xf0(%rsp) movq %rsi, 0xf8(%rsp) leaq 0xb0(%rsp), %rax movq %rax, 0x100(%rsp) movl $0x1, 0x108(%rsp) movq 0x48(%r13), %rax testq %rax, %rax movq %r8, 0x58(%rsp) movq %r9, 0x50(%rsp) movq %r10, 0x48(%rsp) movq %r11, 0x40(%rsp) je 0x20920b leaq 0xe0(%rsp), %rdi movq %rdx, 0xa0(%rsp) movq %rsi, 0x90(%rsp) vzeroupper callq %rax movq 0x40(%rsp), %r11 movq 0x48(%rsp), %r10 movq 0x50(%rsp), %r9 vmovups 0x120(%rsp), %ymm7 movq 0x58(%rsp), %r8 movq 0x90(%rsp), %rsi movq 0xa0(%rsp), %rdx movq 0xe0(%rsp), %rax cmpl $0x0, (%rax) je 0x209268 movq 0x10(%r14), %rax testq %rax, %rax je 0x209264 testb $0x2, (%r14) jne 0x209221 testb $0x40, 0x3e(%r13) je 0x209257 leaq 0xe0(%rsp), %rdi movq %rdx, %r14 movq %rsi, %r13 vzeroupper callq %rax movq 0x40(%rsp), %r11 movq 0x48(%rsp), %r10 movq 0x50(%rsp), %r9 vmovups 0x120(%rsp), %ymm7 movq 0x58(%rsp), %r8 movq %r13, %rsi movq %r14, %rdx movq 0xe0(%rsp), %rax cmpl $0x0, (%rax) je 0x209268 xorl %eax, %eax jmp 0x209279 vmovss 0x20(%rsp), %xmm0 vmovss %xmm0, 0x20(%rsi) btcq %r15, %rbp movb $0x1, %al xorl %ecx, %ecx movq 0x30(%rsp), %rdi jmp 0x20907d incq %r14 cmpq %rdi, %r14 setb %al leaq 0x3c0(%rsp), %rbp jne 0x208ad5 movq 0x70(%rsp), %r14 movq 0x68(%rsp), %r15 movq 0x60(%rsp), %r13 jmp 0x2092d5 testb $0x1, 0xe(%rsp) movq 0x70(%rsp), %r14 movq 0x68(%rsp), %r15 movq 0x60(%rsp), %r13 leaq 0x3c0(%rsp), %rbp je 0x2092d5 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rcx cmpl $0x3, %ecx jne 0x2088ac addq $0x1568, %rsp # imm = 0x1568 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, true, embree::avx::ArrayIntersector1<embree::avx::TriangleMiMBIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x15a8, %rsp # imm = 0x15A8 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x20a1a0 vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x20a1a0 leaq 0x408(%rsp), %r8 movq 0x70(%rax), %rax movq %rax, -0x8(%r8) vxorps %xmm6, %xmm6, %xmm6 vmaxss 0xc(%rsi), %xmm6, %xmm1 vmovaps 0x10(%rsi), %xmm2 vbroadcastss 0x1d17b77(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1ce7c8e(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1ce33ac(%rip), %xmm4 # 0x1eec714 vdivps %xmm2, %xmm4, %xmm2 vbroadcastss 0x1d17beb(%rip), %xmm4 # 0x1f20f60 vblendvps %xmm3, %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d16b8c(%rip), %xmm3 # 0x1f1ff10 vmulps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1d16b83(%rip), %xmm4 # 0x1f1ff14 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss (%rsi), %ymm7 vbroadcastss 0x4(%rsi), %ymm4 vmovups %ymm4, 0x3e0(%rsp) vbroadcastss 0x8(%rsi), %ymm4 vmovups %ymm4, 0x3c0(%rsp) xorl %r9d, %r9d vucomiss %xmm6, %xmm3 setb %r9b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x3a0(%rsp) vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm5 vmovups %ymm5, 0x380(%rsp) vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovups %ymm3, 0x360(%rsp) vshufps $0x0, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovups %ymm3, 0x340(%rsp) vshufps $0x55, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovups %ymm3, 0x320(%rsp) vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vmovups %ymm2, 0x300(%rsp) shll $0x5, %r9d xorl %r10d, %r10d vucomiss %xmm6, %xmm4 setb %r10b shll $0x5, %r10d orq $0x40, %r10 xorl %r11d, %r11d vucomiss %xmm6, %xmm5 setb %r11b shll $0x5, %r11d orq $0x80, %r11 movq %r9, %r14 xorq $0x20, %r14 movq %r10, %r15 xorq $0x20, %r15 movq %r11, %r13 xorq $0x20, %r13 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovdqu %ymm1, 0x2e0(%rsp) vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovups %ymm0, 0x2c0(%rsp) leaq 0x1f46acb(%rip), %rax # 0x214ff80 vmovdqa 0xf0(%rax), %xmm0 vmovdqa %xmm0, 0x150(%rsp) movq %rdx, 0x48(%rsp) movq %rsi, 0x40(%rsp) vmovups %ymm7, 0x160(%rsp) movq %r9, 0xb0(%rsp) movq %r10, 0xa8(%rsp) movq %r11, 0xa0(%rsp) movq %r14, 0x98(%rsp) movq %r15, 0x90(%rsp) movq %r13, 0x88(%rsp) leaq 0x400(%rsp), %rax cmpq %rax, %r8 je 0x20a1a0 movq -0x8(%r8), %r12 addq $-0x8, %r8 testb $0x8, %r12b jne 0x209653 movq %r12, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rsi), %ymm0 vmulps 0x100(%rax,%r9), %ymm0, %ymm1 vaddps 0x40(%rax,%r9), %ymm1, %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps 0x3a0(%rsp), %ymm1, %ymm1 vmulps 0x100(%rax,%r10), %ymm0, %ymm2 vaddps 0x40(%rax,%r10), %ymm2, %ymm2 vmovups 0x2e0(%rsp), %ymm3 vmaxps %ymm1, %ymm3, %ymm1 vmovups 0x3e0(%rsp), %ymm4 vsubps %ymm4, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmulps 0x380(%rsp), %ymm2, %ymm2 vaddps 0x40(%rax,%r11), %ymm3, %ymm3 vmovups 0x3c0(%rsp), %ymm5 vsubps %ymm5, %ymm3, %ymm3 vmulps 0x360(%rsp), %ymm3, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%r14), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r14), %ymm3, %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps 0x340(%rsp), %ymm2, %ymm2 vmovups 0x2c0(%rsp), %ymm3 vminps %ymm2, %ymm3, %ymm2 vmulps 0x100(%rax,%r15), %ymm0, %ymm3 vaddps 0x40(%rax,%r15), %ymm3, %ymm3 vsubps %ymm4, %ymm3, %ymm3 vmulps 0x320(%rsp), %ymm3, %ymm3 vmulps 0x100(%rax,%r13), %ymm0, %ymm4 vaddps 0x40(%rax,%r13), %ymm4, %ymm4 vsubps %ymm5, %ymm4, %ymm4 vmulps 0x300(%rsp), %ymm4, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r12d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x20969b vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %ebx testb $0x8, %r12b jne 0x209697 testq %rbx, %rbx je 0x2096c8 andq $-0x10, %r12 bsfq %rbx, %rax leaq -0x1(%rbx), %rdi xorl %ebp, %ebp movq (%r12,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %rbx, %rdi jne 0x2096cd movq %rax, %r12 testl %ebp, %ebp je 0x209522 jmp 0x209713 pushq $0x6 jmp 0x2096ca vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x209643 pushq $0x4 popq %rbp jmp 0x20968d movq %rax, (%r8) addq $0x8, %r8 bsfq %rdi, %rcx leaq -0x1(%rdi), %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdi, %rax je 0x20970b movq %rcx, (%r8) addq $0x8, %r8 bsfq %rax, %rcx leaq -0x1(%rax), %rdi jmp 0x2096dc movq %rcx, %r12 jmp 0x20968d cmpl $0x6, %ebp jne 0x20a197 movl %r12d, %eax andl $0xf, %eax xorl %ebp, %ebp addq $-0x8, %rax setne %cl je 0x20a197 movq %r8, 0xb8(%rsp) andq $-0x10, %r12 movq (%rdx), %r8 xorl %edi, %edi movq %rax, 0xe0(%rsp) movq %r8, 0x50(%rsp) movb %cl, 0xe(%rsp) movq %rdi, 0xe8(%rsp) imulq $0x50, %rdi, %rdi vmovss 0x1c(%rsi), %xmm0 movl 0x30(%r12,%rdi), %eax movq 0x1e8(%r8), %rcx movq (%rcx,%rax,8), %rax vmovss 0x28(%rax), %xmm1 vmovss 0x2c(%rax), %xmm2 vmovss 0x30(%rax), %xmm3 vsubss %xmm2, %xmm0, %xmm0 vsubss %xmm2, %xmm3, %xmm2 vdivss %xmm2, %xmm0, %xmm0 vmulss %xmm0, %xmm1, %xmm6 vroundss $0x9, %xmm6, %xmm6, %xmm0 vaddss 0x1ce722b(%rip), %xmm1, %xmm1 # 0x1ef09cc vminss %xmm1, %xmm0, %xmm0 vxorps %xmm1, %xmm1, %xmm1 vmaxss %xmm0, %xmm1, %xmm2 vcvttss2si %xmm2, %ecx movslq %ecx, %rcx movq 0xe0(%rax), %rdx imulq $0x38, %rcx, %rcx movl (%r12,%rdi), %r13d movl 0x4(%r12,%rdi), %r15d movq (%rdx,%rcx), %rax movq 0x38(%rdx,%rcx), %rcx vmovups (%rax,%r13,4), %xmm4 movl 0x10(%r12,%rdi), %esi vmovups (%rax,%rsi,4), %xmm5 movl 0x20(%r12,%rdi), %edx movq %rdx, 0x30(%rsp) vmovups (%rax,%rdx,4), %xmm0 vmovaps %xmm0, 0x20(%rsp) vmovups (%rax,%r15,4), %xmm7 movl 0x14(%r12,%rdi), %edx vmovups (%rax,%rdx,4), %xmm8 movl 0x24(%r12,%rdi), %r8d movq %r8, 0x70(%rsp) vmovups (%rax,%r8,4), %xmm0 vmovaps %xmm0, 0x10(%rsp) movl 0x8(%r12,%rdi), %r14d vmovups (%rax,%r14,4), %xmm9 movl 0x18(%r12,%rdi), %r10d vmovups (%rax,%r10,4), %xmm11 movl 0x28(%r12,%rdi), %r8d vmovups (%rax,%r8,4), %xmm10 movl 0xc(%r12,%rdi), %ebp vmovups (%rax,%rbp,4), %xmm1 movl 0x1c(%r12,%rdi), %r11d vmovups (%rax,%r11,4), %xmm14 movl 0x2c(%r12,%rdi), %r9d vmovups (%rax,%r9,4), %xmm12 vmovups (%rcx,%r13,4), %xmm15 vmovups (%rcx,%rsi,4), %xmm13 vsubss %xmm2, %xmm6, %xmm6 vunpcklps %xmm9, %xmm4, %xmm2 # xmm2 = xmm4[0],xmm9[0],xmm4[1],xmm9[1] vunpckhps %xmm9, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm9[2],xmm4[3],xmm9[3] vunpcklps %xmm1, %xmm7, %xmm0 # xmm0 = xmm7[0],xmm1[0],xmm7[1],xmm1[1] vunpckhps %xmm1, %xmm7, %xmm1 # xmm1 = xmm7[2],xmm1[2],xmm7[3],xmm1[3] vmovups (%rcx,%r15,4), %xmm3 vunpcklps %xmm1, %xmm4, %xmm1 # xmm1 = xmm4[0],xmm1[0],xmm4[1],xmm1[1] vmovaps %xmm1, 0xc0(%rsp) vunpcklps %xmm0, %xmm2, %xmm9 # xmm9 = xmm2[0],xmm0[0],xmm2[1],xmm0[1] vunpckhps %xmm0, %xmm2, %xmm7 # xmm7 = xmm2[2],xmm0[2],xmm2[3],xmm0[3] vunpcklps %xmm11, %xmm5, %xmm0 # xmm0 = xmm5[0],xmm11[0],xmm5[1],xmm11[1] vunpckhps %xmm11, %xmm5, %xmm1 # xmm1 = xmm5[2],xmm11[2],xmm5[3],xmm11[3] vunpcklps %xmm14, %xmm8, %xmm2 # xmm2 = xmm8[0],xmm14[0],xmm8[1],xmm14[1] vunpckhps %xmm14, %xmm8, %xmm5 # xmm5 = xmm8[2],xmm14[2],xmm8[3],xmm14[3] vmovups (%rcx,%r14,4), %xmm11 vunpcklps %xmm5, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm5[0],xmm1[1],xmm5[1] vmovaps %xmm1, 0x60(%rsp) vunpcklps %xmm2, %xmm0, %xmm8 # xmm8 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vunpckhps %xmm2, %xmm0, %xmm5 # xmm5 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps 0x20(%rsp), %xmm1 vunpcklps %xmm10, %xmm1, %xmm0 # xmm0 = xmm1[0],xmm10[0],xmm1[1],xmm10[1] vunpckhps %xmm10, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm10[2],xmm1[3],xmm10[3] vmovaps 0x10(%rsp), %xmm4 vunpcklps %xmm12, %xmm4, %xmm2 # xmm2 = xmm4[0],xmm12[0],xmm4[1],xmm12[1] vunpckhps %xmm12, %xmm4, %xmm10 # xmm10 = xmm4[2],xmm12[2],xmm4[3],xmm12[3] vmovups (%rcx,%rbp,4), %xmm12 vunpcklps %xmm10, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm10[0],xmm1[1],xmm10[1] vmovaps %xmm1, 0x20(%rsp) vunpcklps %xmm2, %xmm0, %xmm1 # xmm1 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vmovaps %xmm1, 0xd0(%rsp) vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x10(%rsp) vunpcklps %xmm11, %xmm15, %xmm10 # xmm10 = xmm15[0],xmm11[0],xmm15[1],xmm11[1] vunpckhps %xmm11, %xmm15, %xmm0 # xmm0 = xmm15[2],xmm11[2],xmm15[3],xmm11[3] vunpcklps %xmm12, %xmm3, %xmm11 # xmm11 = xmm3[0],xmm12[0],xmm3[1],xmm12[1] vunpckhps %xmm12, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm12[2],xmm3[3],xmm12[3] vmovups (%rcx,%r10,4), %xmm12 vunpcklps %xmm3, %xmm0, %xmm4 # xmm4 = xmm0[0],xmm3[0],xmm0[1],xmm3[1] vunpcklps %xmm11, %xmm10, %xmm3 # xmm3 = xmm10[0],xmm11[0],xmm10[1],xmm11[1] vunpckhps %xmm11, %xmm10, %xmm1 # xmm1 = xmm10[2],xmm11[2],xmm10[3],xmm11[3] vunpcklps %xmm12, %xmm13, %xmm11 # xmm11 = xmm13[0],xmm12[0],xmm13[1],xmm12[1] vunpckhps %xmm12, %xmm13, %xmm10 # xmm10 = xmm13[2],xmm12[2],xmm13[3],xmm12[3] vmovups (%rcx,%rdx,4), %xmm12 vmovups (%rcx,%r11,4), %xmm13 vunpcklps %xmm13, %xmm12, %xmm14 # xmm14 = xmm12[0],xmm13[0],xmm12[1],xmm13[1] vunpckhps %xmm13, %xmm12, %xmm12 # xmm12 = xmm12[2],xmm13[2],xmm12[3],xmm13[3] vunpcklps %xmm12, %xmm10, %xmm10 # xmm10 = xmm10[0],xmm12[0],xmm10[1],xmm12[1] vunpcklps %xmm14, %xmm11, %xmm0 # xmm0 = xmm11[0],xmm14[0],xmm11[1],xmm14[1] vunpckhps %xmm14, %xmm11, %xmm2 # xmm2 = xmm11[2],xmm14[2],xmm11[3],xmm14[3] movq 0x30(%rsp), %rax vmovups (%rcx,%rax,4), %xmm11 vmovups (%rcx,%r8,4), %xmm12 movq 0x50(%rsp), %r8 movq 0x40(%rsp), %rsi movq 0x48(%rsp), %rdx vunpcklps %xmm12, %xmm11, %xmm14 # xmm14 = xmm11[0],xmm12[0],xmm11[1],xmm12[1] vunpckhps %xmm12, %xmm11, %xmm11 # xmm11 = xmm11[2],xmm12[2],xmm11[3],xmm12[3] movq 0x70(%rsp), %rax vmovups (%rcx,%rax,4), %xmm12 vmovups (%rcx,%r9,4), %xmm13 vunpcklps %xmm13, %xmm12, %xmm15 # xmm15 = xmm12[0],xmm13[0],xmm12[1],xmm13[1] vunpckhps %xmm13, %xmm12, %xmm12 # xmm12 = xmm12[2],xmm13[2],xmm12[3],xmm13[3] vunpcklps %xmm12, %xmm11, %xmm11 # xmm11 = xmm11[0],xmm12[0],xmm11[1],xmm12[1] vunpcklps %xmm15, %xmm14, %xmm13 # xmm13 = xmm14[0],xmm15[0],xmm14[1],xmm15[1] vunpckhps %xmm15, %xmm14, %xmm14 # xmm14 = xmm14[2],xmm15[2],xmm14[3],xmm15[3] vshufps $0x0, %xmm6, %xmm6, %xmm15 # xmm15 = xmm6[0,0,0,0] vmovss 0x1ce2d53(%rip), %xmm12 # 0x1eec714 vsubss %xmm6, %xmm12, %xmm6 vshufps $0x0, %xmm6, %xmm6, %xmm12 # xmm12 = xmm6[0,0,0,0] vmulps %xmm3, %xmm15, %xmm3 vmulps %xmm9, %xmm12, %xmm6 vaddps %xmm3, %xmm6, %xmm6 vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm7, %xmm12, %xmm3 vaddps %xmm1, %xmm3, %xmm7 vmulps %xmm4, %xmm15, %xmm1 vmulps 0xc0(%rsp), %xmm12, %xmm3 vaddps %xmm1, %xmm3, %xmm4 vmulps %xmm0, %xmm15, %xmm0 vmulps %xmm8, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm8 vmulps %xmm2, %xmm15, %xmm0 vmulps %xmm5, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm5 vmovaps 0x30(%r12,%rdi), %xmm0 vmovaps %xmm0, 0x200(%rsp) vmovaps 0x40(%r12,%rdi), %xmm0 movb 0xe(%rsp), %dil vmulps %xmm10, %xmm15, %xmm1 vmulps 0x60(%rsp), %xmm12, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vmulps %xmm13, %xmm15, %xmm2 vmulps %xmm14, %xmm15, %xmm3 vmulps %xmm11, %xmm15, %xmm9 vmulps 0xd0(%rsp), %xmm12, %xmm10 vaddps %xmm2, %xmm10, %xmm2 vmulps 0x10(%rsp), %xmm12, %xmm10 vaddps %xmm3, %xmm10, %xmm15 vmulps 0x20(%rsp), %xmm12, %xmm3 vaddps %xmm3, %xmm9, %xmm9 vmovaps %xmm0, 0x1f0(%rsp) vbroadcastss (%rsi), %xmm12 vbroadcastss 0x4(%rsi), %xmm13 vbroadcastss 0x8(%rsi), %xmm14 vsubps %xmm12, %xmm6, %xmm10 vsubps %xmm13, %xmm7, %xmm3 vsubps %xmm14, %xmm4, %xmm7 vsubps %xmm12, %xmm8, %xmm0 vmovaps %xmm0, 0x30(%rsp) vsubps %xmm13, %xmm5, %xmm0 vmovaps %xmm0, 0x10(%rsp) vsubps %xmm14, %xmm1, %xmm0 vmovaps %xmm0, 0x20(%rsp) vsubps %xmm12, %xmm2, %xmm12 vsubps %xmm13, %xmm15, %xmm2 vmovaps %xmm2, 0x60(%rsp) vsubps %xmm14, %xmm9, %xmm1 vmovaps %xmm1, 0x70(%rsp) vsubps %xmm10, %xmm12, %xmm14 vsubps %xmm3, %xmm2, %xmm15 vsubps %xmm7, %xmm1, %xmm0 vaddps %xmm3, %xmm2, %xmm4 vaddps %xmm7, %xmm1, %xmm5 vmulps %xmm4, %xmm0, %xmm8 vmulps %xmm5, %xmm15, %xmm9 vsubps %xmm8, %xmm9, %xmm13 vaddps %xmm10, %xmm12, %xmm8 vmulps %xmm5, %xmm14, %xmm5 vmovaps %xmm0, 0x1b0(%rsp) vmulps %xmm0, %xmm8, %xmm9 vsubps %xmm5, %xmm9, %xmm5 vmovaps %xmm15, 0x1c0(%rsp) vmulps %xmm8, %xmm15, %xmm8 vmovaps %xmm14, 0x1d0(%rsp) vmulps %xmm4, %xmm14, %xmm4 vsubps %xmm8, %xmm4, %xmm4 vbroadcastss 0x18(%rsi), %xmm6 vmulps %xmm4, %xmm6, %xmm4 vbroadcastss 0x14(%rsi), %xmm15 vmulps %xmm5, %xmm15, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vbroadcastss 0x10(%rsi), %xmm11 vmulps %xmm13, %xmm11, %xmm5 vaddps %xmm4, %xmm5, %xmm9 vmovaps 0x10(%rsp), %xmm2 vsubps %xmm2, %xmm3, %xmm0 vmovaps 0x20(%rsp), %xmm13 vsubps %xmm13, %xmm7, %xmm14 vmovaps %xmm3, 0xc0(%rsp) vaddps %xmm2, %xmm3, %xmm4 vmovaps %xmm7, 0x1e0(%rsp) vaddps %xmm7, %xmm13, %xmm5 vmulps %xmm4, %xmm14, %xmm7 vmulps %xmm5, %xmm0, %xmm3 vsubps %xmm7, %xmm3, %xmm3 vmovaps 0x30(%rsp), %xmm8 vsubps %xmm8, %xmm10, %xmm1 vmulps %xmm5, %xmm1, %xmm5 vmovaps %xmm10, 0xd0(%rsp) vaddps %xmm8, %xmm10, %xmm7 vmovaps %xmm8, %xmm10 vmovaps %xmm14, 0x190(%rsp) vmulps %xmm7, %xmm14, %xmm8 vmovaps %xmm1, %xmm14 vsubps %xmm5, %xmm8, %xmm5 vmovaps %xmm0, 0x1a0(%rsp) vmulps %xmm7, %xmm0, %xmm7 vmulps %xmm4, %xmm1, %xmm4 vsubps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm6, %xmm4 vmulps %xmm5, %xmm15, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vmulps %xmm3, %xmm11, %xmm3 vaddps %xmm4, %xmm3, %xmm4 vsubps %xmm12, %xmm10, %xmm5 vaddps %xmm12, %xmm10, %xmm3 vmovaps 0x60(%rsp), %xmm0 vsubps %xmm0, %xmm2, %xmm12 vaddps %xmm0, %xmm2, %xmm2 vmovaps 0x70(%rsp), %xmm0 vsubps %xmm0, %xmm13, %xmm8 vaddps %xmm0, %xmm13, %xmm0 vmulps %xmm2, %xmm8, %xmm1 vmulps %xmm0, %xmm12, %xmm7 vsubps %xmm1, %xmm7, %xmm1 vmulps %xmm0, %xmm5, %xmm0 vmulps %xmm3, %xmm8, %xmm7 vsubps %xmm0, %xmm7, %xmm0 vmovups 0x160(%rsp), %ymm7 vmulps %xmm3, %xmm12, %xmm3 vmulps %xmm2, %xmm5, %xmm2 vsubps %xmm3, %xmm2, %xmm2 vmovaps %xmm6, 0x30(%rsp) vmulps %xmm2, %xmm6, %xmm2 vmovaps %xmm15, 0x70(%rsp) vmulps %xmm0, %xmm15, %xmm0 vaddps %xmm2, %xmm0, %xmm0 vmovaps %xmm11, 0x60(%rsp) vmulps %xmm1, %xmm11, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm4, %xmm9, %xmm1 vaddps %xmm1, %xmm0, %xmm11 vminps %xmm4, %xmm9, %xmm1 vminps %xmm0, %xmm1, %xmm1 vbroadcastss 0x1d17247(%rip), %xmm2 # 0x1f20ec4 vandps %xmm2, %xmm11, %xmm3 vbroadcastss 0x1d17242(%rip), %xmm2 # 0x1f20ecc vmovaps %xmm3, 0x10(%rsp) vmulps %xmm2, %xmm3, %xmm2 vbroadcastss 0x1d17223(%rip), %xmm3 # 0x1f20ec0 vxorps %xmm3, %xmm2, %xmm3 vcmpnltps %xmm3, %xmm1, %xmm1 vmovaps %xmm9, 0x20(%rsp) vmaxps %xmm4, %xmm9, %xmm3 vmaxps %xmm0, %xmm3, %xmm0 vcmpleps %xmm2, %xmm0, %xmm0 vorps %xmm0, %xmm1, %xmm0 movb $0x0, 0xf(%rsp) leaq 0xf(%rsp), %rax movq %rax, 0x240(%rsp) vtestps 0x150(%rsp), %xmm0 je 0x20a0f4 vmovaps 0x1b0(%rsp), %xmm9 vmovaps 0x1a0(%rsp), %xmm10 vmovaps %xmm0, 0x180(%rsp) vmulps %xmm10, %xmm9, %xmm0 vmovaps 0x1c0(%rsp), %xmm15 vmovaps 0x190(%rsp), %xmm6 vmulps %xmm6, %xmm15, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm6, %xmm12, %xmm2 vmulps %xmm8, %xmm10, %xmm3 vsubps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d17195(%rip), %xmm7 # 0x1f20ec4 vandps %xmm7, %xmm0, %xmm0 vandps %xmm7, %xmm2, %xmm2 vcmpltps %xmm2, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm3, %xmm0 vmulps %xmm8, %xmm14, %xmm1 vmulps %xmm14, %xmm9, %xmm2 vmovaps 0x1d0(%rsp), %xmm9 vmulps %xmm6, %xmm9, %xmm3 vsubps %xmm3, %xmm2, %xmm2 vmulps %xmm5, %xmm6, %xmm6 vsubps %xmm1, %xmm6, %xmm6 vandps %xmm7, %xmm3, %xmm3 vandps %xmm7, %xmm1, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm6, %xmm1 vmulps %xmm5, %xmm10, %xmm2 vmulps %xmm10, %xmm9, %xmm3 vmulps %xmm14, %xmm15, %xmm5 vmulps %xmm12, %xmm14, %xmm6 vsubps %xmm5, %xmm3, %xmm3 vsubps %xmm2, %xmm6, %xmm6 vandps %xmm7, %xmm5, %xmm5 vandps %xmm7, %xmm2, %xmm2 vcmpltps %xmm2, %xmm5, %xmm2 vblendvps %xmm2, %xmm3, %xmm6, %xmm2 vmulps 0x30(%rsp), %xmm2, %xmm3 vmulps 0x70(%rsp), %xmm1, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vmulps 0x60(%rsp), %xmm0, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vaddps %xmm3, %xmm3, %xmm5 vmulps 0x1e0(%rsp), %xmm2, %xmm3 vmulps 0xc0(%rsp), %xmm1, %xmm6 vaddps %xmm3, %xmm6, %xmm3 vmulps 0xd0(%rsp), %xmm0, %xmm6 vaddps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm3, %xmm3 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm7 vbroadcastss 0x1ce2918(%rip), %xmm9 # 0x1eec714 vsubps %xmm7, %xmm9, %xmm7 vmulps %xmm7, %xmm6, %xmm7 vaddps %xmm7, %xmm6, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0xc(%rsi), %xmm6 vcmpleps %xmm3, %xmm6, %xmm6 vbroadcastss 0x20(%rsi), %xmm7 vcmpleps %xmm7, %xmm3, %xmm7 vandps %xmm7, %xmm6, %xmm6 vmovups 0x160(%rsp), %ymm7 vcmpneqps 0x1ce1bd8(%rip), %xmm5, %xmm5 # 0x1eeba10 vandps %xmm6, %xmm5, %xmm6 vmovaps 0x180(%rsp), %xmm5 vandps 0x150(%rsp), %xmm5, %xmm5 vpslld $0x1f, %xmm6, %xmm6 vpsrad $0x1f, %xmm6, %xmm6 vtestps %xmm5, %xmm6 je 0x20a0f4 vandps %xmm5, %xmm6, %xmm5 vmovaps 0x20(%rsp), %xmm8 vmovaps %xmm8, 0x210(%rsp) vmovaps %xmm4, 0x220(%rsp) vmovaps %xmm11, 0x230(%rsp) movq %rax, 0x240(%rsp) vmovaps %xmm5, 0x250(%rsp) vmovaps %xmm3, 0x280(%rsp) vmovaps %xmm0, 0x290(%rsp) vmovaps %xmm1, 0x2a0(%rsp) vmovaps %xmm2, 0x2b0(%rsp) vrcpps %xmm11, %xmm0 vmulps %xmm0, %xmm11, %xmm1 vbroadcastss 0x1ce2845(%rip), %xmm2 # 0x1eec714 vsubps %xmm1, %xmm2, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1ce7104(%rip), %xmm1 # 0x1ef0fe8 vmovaps 0x10(%rsp), %xmm3 vcmpnltps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm0 vmulps %xmm0, %xmm8, %xmm1 vminps %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x260(%rsp) vmulps %xmm0, %xmm4, %xmm0 vminps %xmm2, %xmm0, %xmm0 vmovaps %xmm0, 0x270(%rsp) vmovmskps %xmm5, %r13d bsfq %r13, %rbp movl 0x200(%rsp,%rbp,4), %eax movq 0x1e8(%r8), %rcx movq (%rcx,%rax,8), %r15 movl 0x24(%rsi), %ecx testl %ecx, 0x34(%r15) je 0x209f4e movq 0x10(%rdx), %r14 cmpq $0x0, 0x10(%r14) jne 0x209f68 cmpq $0x0, 0x48(%r15) jne 0x209f68 xorl %eax, %eax jmp 0x209f54 btcq %rbp, %r13 movb $0x1, %al xorl %ebp, %ebp testb %al, %al je 0x20a14f testq %r13, %r13 jne 0x209f19 jmp 0x20a0f4 vmovss 0x260(%rsp,%rbp,4), %xmm0 vmovss 0x270(%rsp,%rbp,4), %xmm1 movq 0x8(%rdx), %rcx movl 0x1f0(%rsp,%rbp,4), %edi vmovss 0x290(%rsp,%rbp,4), %xmm2 vmovss 0x2a0(%rsp,%rbp,4), %xmm3 vmovss 0x2b0(%rsp,%rbp,4), %xmm4 vmovss %xmm2, 0xf0(%rsp) vmovss %xmm3, 0xf4(%rsp) vmovss %xmm4, 0xf8(%rsp) vmovss %xmm0, 0xfc(%rsp) vmovss %xmm1, 0x100(%rsp) movl %edi, 0x104(%rsp) movl %eax, 0x108(%rsp) movl (%rcx), %eax movl %eax, 0x10c(%rsp) movl 0x4(%rcx), %eax movl %eax, 0x110(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0x30(%rsp) vmovss 0x280(%rsp,%rbp,4), %xmm0 vmovss %xmm0, 0x20(%rsi) orl $-0x1, 0x5c(%rsp) leaq 0x5c(%rsp), %rax movq %rax, 0x120(%rsp) movq 0x18(%r15), %rax movq %rax, 0x128(%rsp) movq %rcx, 0x130(%rsp) movq %rsi, 0x138(%rsp) leaq 0xf0(%rsp), %rax movq %rax, 0x140(%rsp) movl $0x1, 0x148(%rsp) movq 0x48(%r15), %rax testq %rax, %rax je 0x20a08b leaq 0x120(%rsp), %rdi vzeroupper callq %rax movq 0x50(%rsp), %r8 vmovups 0x160(%rsp), %ymm7 movq 0x40(%rsp), %rsi movq 0x48(%rsp), %rdx movq 0x120(%rsp), %rax cmpl $0x0, (%rax) je 0x20a0d7 movq 0x10(%r14), %rax testq %rax, %rax je 0x20a0d3 testb $0x2, (%r14) jne 0x20a0a1 testb $0x40, 0x3e(%r15) je 0x20a0c6 leaq 0x120(%rsp), %rdi vzeroupper callq %rax movq 0x50(%rsp), %r8 vmovups 0x160(%rsp), %ymm7 movq 0x40(%rsp), %rsi movq 0x48(%rsp), %rdx movq 0x120(%rsp), %rax cmpl $0x0, (%rax) je 0x20a0d7 xorl %eax, %eax jmp 0x20a0e8 vmovss 0x30(%rsp), %xmm0 vmovss %xmm0, 0x20(%rsi) btcq %rbp, %r13 movb $0x1, %al xorl %ebp, %ebp movb 0xe(%rsp), %dil jmp 0x209f56 movq 0xe8(%rsp), %rdi incq %rdi movq 0xe0(%rsp), %rax cmpq %rax, %rdi setb %cl jne 0x20974f movq 0xb8(%rsp), %r8 movq 0xb0(%rsp), %r9 movq 0xa8(%rsp), %r10 movq 0xa0(%rsp), %r11 movq 0x98(%rsp), %r14 movq 0x90(%rsp), %r15 movq 0x88(%rsp), %r13 xorl %ebp, %ebp jmp 0x20a197 testb $0x1, %dil movq 0xb8(%rsp), %r8 movq 0xb0(%rsp), %r9 movq 0xa8(%rsp), %r10 movq 0xa0(%rsp), %r11 movq 0x98(%rsp), %r14 movq 0x90(%rsp), %r15 movq 0x88(%rsp), %r13 je 0x20a197 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rbp cmpl $0x3, %ebp jne 0x209509 addq $0x15a8, %rsp # imm = 0x15A8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, false, embree::avx::ArrayIntersector1<embree::avx::QuadMvIntersector1Moeller<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x1640, %rsp # imm = 0x1640 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x20a1ea leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %rsi, %r14 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x20a1d8 leaq 0x488(%rsp), %rdi vmovaps 0x10(%r14), %xmm3 vmaxss 0xc(%r14), %xmm2, %xmm1 vbroadcastss 0x1d16cab(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce6dc2(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 movq 0x70(%rax), %rax vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1ce24ce(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss (%r14), %ymm6 vbroadcastss 0x4(%r14), %ymm7 vbroadcastss 0x8(%r14), %ymm8 vaddps %xmm3, %xmm4, %xmm3 xorl %r8d, %r8d vucomiss %xmm2, %xmm3 movq %rax, -0x8(%rdi) setb %r8b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm2, %xmm4 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm2, %xmm5 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %rcx xorq $0x20, %rcx movq %r10, %r15 xorq $0x20, %r15 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm4 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm5 leaq 0x1f45c85(%rip), %rax # 0x214ff80 vbroadcastf128 0xf0(%rax), %ymm0 # ymm0 = mem[0,1,0,1] vmovaps %ymm0, 0x260(%rsp) vperm2f128 $0x2, (%rax), %ymm0, %ymm1 # ymm1 = mem[0,1],ymm0[0,1] leaq 0x480(%rsp), %rbx vbroadcastss 0x1ce66a8(%rip), %ymm0 # 0x1ef09cc vbroadcastss 0x1ce23e7(%rip), %ymm2 # 0x1eec714 vmovaps %ymm1, 0x2a0(%rsp) vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x280(%rsp) vmovaps %ymm6, 0x240(%rsp) vmovaps %ymm7, 0x220(%rsp) vmovaps %ymm8, 0x200(%rsp) vmovaps %ymm9, 0x1e0(%rsp) vmovaps %ymm10, 0x1c0(%rsp) vmovaps %ymm3, 0x1a0(%rsp) movq %r15, 0x38(%rsp) vmovaps %ymm4, 0x180(%rsp) vmovaps %ymm5, 0x160(%rsp) cmpq %rbx, %rdi je 0x20a1d8 movq -0x8(%rdi), %r13 addq $-0x8, %rdi testb $0x8, %r13b jne 0x20a426 vmovaps 0x40(%r13,%r8), %ymm0 vsubps %ymm6, %ymm0, %ymm0 vmulps %ymm0, %ymm9, %ymm0 vmovaps 0x40(%r13,%r9), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm10, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r13,%r10), %ymm1 vsubps %ymm8, %ymm1, %ymm1 vmulps %ymm1, %ymm3, %ymm1 vmaxps %ymm4, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r13,%r11), %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmovaps 0x40(%r13,%rcx), %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm10, %ymm2 vminps %ymm2, %ymm1, %ymm1 vmovaps 0x40(%r13,%r15), %ymm2 vsubps %ymm8, %ymm2, %ymm2 vmulps %ymm2, %ymm3, %ymm2 vminps %ymm5, %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r12d testb $0x8, %r13b jne 0x20a46c testq %r12, %r12 je 0x20a470 andq $-0x10, %r13 bsfq %r12, %rax leaq -0x1(%r12), %rsi movq (%r13,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r12, %rsi jne 0x20a475 movq %rax, %r13 xorl %esi, %esi testl %esi, %esi je 0x20a3a3 jmp 0x20a4d5 pushq $0x6 jmp 0x20a472 pushq $0x4 popq %rsi jmp 0x20a462 movq %rdx, 0xc0(%rsp) movq %rbx, %rdx movq %rcx, %rbx movq %rax, (%rdi) addq $0x8, %rdi bsfq %rsi, %rcx leaq -0x1(%rsi), %rax movq (%r13,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rsi, %rax je 0x20a4c2 movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rsi jmp 0x20a492 movq %rcx, %r13 movq %rbx, %rcx movq %rdx, %rbx movq 0xc0(%rsp), %rdx jmp 0x20a460 cmpl $0x6, %esi jne 0x20ab4a movl %r13d, %eax andl $0xf, %eax xorl %esi, %esi addq $-0x8, %rax movq %rax, 0x98(%rsp) setne %bl je 0x20ab42 andq $-0x10, %r13 xorl %r15d, %r15d imulq $0xe0, %r15, %rax vbroadcastf128 0xd0(%r13,%rax), %ymm0 # ymm0 = mem[0,1,0,1] vbroadcastf128 0xc0(%r13,%rax), %ymm1 # ymm1 = mem[0,1,0,1] vmovaps %ymm1, 0x2e0(%rsp) vmovaps %ymm0, 0x2c0(%rsp) vmovaps (%r13,%rax), %xmm0 vmovaps 0x10(%r13,%rax), %xmm1 vmovaps 0x20(%r13,%rax), %xmm2 vinsertf128 $0x1, 0x60(%r13,%rax), %ymm0, %ymm3 vinsertf128 $0x1, 0x70(%r13,%rax), %ymm1, %ymm5 vinsertf128 $0x1, 0x80(%r13,%rax), %ymm2, %ymm7 vbroadcastf128 0x30(%r13,%rax), %ymm0 # ymm0 = mem[0,1,0,1] vbroadcastf128 0x40(%r13,%rax), %ymm1 # ymm1 = mem[0,1,0,1] vbroadcastf128 0x50(%r13,%rax), %ymm2 # ymm2 = mem[0,1,0,1] vbroadcastf128 0x90(%r13,%rax), %ymm9 # ymm9 = mem[0,1,0,1] vbroadcastf128 0xa0(%r13,%rax), %ymm10 # ymm10 = mem[0,1,0,1] vbroadcastf128 0xb0(%r13,%rax), %ymm11 # ymm11 = mem[0,1,0,1] vsubps %ymm0, %ymm3, %ymm4 vsubps %ymm1, %ymm5, %ymm12 vmovaps %ymm12, 0x40(%rsp) vsubps %ymm2, %ymm7, %ymm8 vsubps %ymm3, %ymm9, %ymm9 vsubps %ymm5, %ymm10, %ymm10 vsubps %ymm7, %ymm11, %ymm11 vmulps %ymm11, %ymm12, %ymm0 vmulps %ymm10, %ymm8, %ymm1 vsubps %ymm0, %ymm1, %ymm0 vmovaps %ymm0, 0x140(%rsp) vmulps %ymm9, %ymm8, %ymm0 vmovaps %ymm4, 0xc0(%rsp) vmulps %ymm4, %ymm11, %ymm2 vsubps %ymm0, %ymm2, %ymm6 vmulps %ymm4, %ymm10, %ymm0 vmulps %ymm9, %ymm12, %ymm12 vsubps %ymm0, %ymm12, %ymm4 vbroadcastss (%r14), %ymm12 vbroadcastss 0x4(%r14), %ymm13 vbroadcastss 0x8(%r14), %ymm14 vbroadcastss 0x14(%r14), %ymm15 vsubps %ymm12, %ymm3, %ymm3 vbroadcastss 0x18(%r14), %ymm12 vsubps %ymm13, %ymm5, %ymm5 vsubps %ymm14, %ymm7, %ymm7 vmulps %ymm7, %ymm15, %ymm13 vmulps %ymm5, %ymm12, %ymm14 vsubps %ymm13, %ymm14, %ymm13 vbroadcastss 0x10(%r14), %ymm14 vmulps %ymm3, %ymm12, %ymm0 vmulps %ymm7, %ymm14, %ymm1 vsubps %ymm0, %ymm1, %ymm0 vmulps %ymm5, %ymm14, %ymm1 vmulps %ymm3, %ymm15, %ymm2 vsubps %ymm1, %ymm2, %ymm1 vmovaps %ymm4, 0xa0(%rsp) vmulps %ymm4, %ymm12, %ymm2 vmulps %ymm6, %ymm15, %ymm12 vmovaps %ymm6, %ymm15 vaddps %ymm2, %ymm12, %ymm2 vmovaps 0x140(%rsp), %ymm4 vmulps %ymm4, %ymm14, %ymm12 vmovaps %ymm4, %ymm14 vaddps %ymm2, %ymm12, %ymm2 vmulps %ymm1, %ymm11, %ymm11 vmulps %ymm0, %ymm10, %ymm10 vaddps %ymm11, %ymm10, %ymm10 vmulps %ymm13, %ymm9, %ymm9 vaddps %ymm10, %ymm9, %ymm10 vmulps %ymm1, %ymm8, %ymm1 vmulps 0x40(%rsp), %ymm0, %ymm0 vaddps %ymm1, %ymm0, %ymm0 vbroadcastss 0x1d16823(%rip), %ymm1 # 0x1f20ec0 vandps %ymm1, %ymm2, %ymm9 vxorps %ymm10, %ymm9, %ymm6 vmulps 0xc0(%rsp), %ymm13, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vxorps %ymm0, %ymm9, %ymm8 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %ymm10, %ymm6, %ymm0 vcmpnltps %ymm10, %ymm8, %ymm1 vandps %ymm1, %ymm0, %ymm0 vbroadcastss 0x1d167ef(%rip), %ymm1 # 0x1f20ec4 vandps %ymm1, %ymm2, %ymm4 vcmpneqps %ymm2, %ymm10, %ymm1 vandps %ymm1, %ymm0, %ymm0 vaddps %ymm6, %ymm8, %ymm1 vcmpleps %ymm4, %ymm1, %ymm1 vandps %ymm1, %ymm0, %ymm10 vtestps 0x260(%rsp), %ymm10 jne 0x20a714 incq %r15 cmpq 0x98(%rsp), %r15 setb %bl jne 0x20a502 jmp 0x20aaf5 vandps 0x260(%rsp), %ymm10, %ymm10 vmulps 0xa0(%rsp), %ymm7, %ymm0 vmulps %ymm5, %ymm15, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vmulps %ymm3, %ymm14, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vxorps %ymm0, %ymm9, %ymm3 vbroadcastss 0xc(%r14), %ymm0 vmulps %ymm0, %ymm4, %ymm0 vcmpltps %ymm3, %ymm0, %ymm0 vbroadcastss 0x20(%r14), %ymm1 vmulps %ymm1, %ymm4, %ymm1 vcmpleps %ymm1, %ymm3, %ymm1 vandps %ymm0, %ymm1, %ymm5 vtestps %ymm10, %ymm5 je 0x20a6fb vandps %ymm5, %ymm10, %ymm0 vmovaps %ymm6, 0x300(%rsp) vmovaps %ymm8, 0x320(%rsp) vmovaps %ymm3, 0x340(%rsp) vmovaps %ymm4, 0x360(%rsp) vmovaps %ymm0, 0x3a0(%rsp) vmovaps 0x300(%rsp), %ymm0 vmovaps 0x320(%rsp), %ymm1 vsubps %ymm1, %ymm4, %ymm2 vmovaps 0x2a0(%rsp), %ymm3 vblendvps %ymm3, %ymm2, %ymm0, %ymm2 vmovaps %ymm2, 0x300(%rsp) vsubps %ymm0, %ymm4, %ymm0 vblendvps %ymm3, %ymm0, %ymm1, %ymm0 vmovaps %ymm0, 0x320(%rsp) vmovaps 0x280(%rsp), %ymm3 vmulps %ymm3, %ymm14, %ymm1 vmovaps %ymm1, 0x420(%rsp) vmulps %ymm3, %ymm15, %ymm1 vmovaps %ymm1, 0x440(%rsp) vrcpps %ymm4, %ymm1 vmulps 0xa0(%rsp), %ymm3, %ymm3 vmovaps %ymm3, 0x460(%rsp) vmulps %ymm1, %ymm4, %ymm3 vbroadcastss 0x1ce1ef9(%rip), %ymm4 # 0x1eec714 vsubps %ymm3, %ymm4, %ymm3 vmulps %ymm3, %ymm1, %ymm3 vaddps %ymm3, %ymm1, %ymm1 vmulps 0x340(%rsp), %ymm1, %ymm3 movq (%rdx), %rax movq %rax, 0x90(%rsp) vmovaps %ymm3, 0x400(%rsp) vmulps %ymm1, %ymm2, %ymm2 vmovaps %ymm2, 0x3c0(%rsp) vmulps %ymm1, %ymm0, %ymm0 vmovaps %ymm0, 0x3e0(%rsp) vmovaps 0x3a0(%rsp), %ymm0 vmovmskps %ymm0, %eax movq %rcx, 0x20(%rsp) movq %rax, 0x40(%rsp) bsfq %rax, %rcx movq %rcx, 0xa0(%rsp) movl 0x2e0(%rsp,%rcx,4), %eax movq 0x90(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq %rax, 0x18(%rsp) movq (%rcx,%rax,8), %rax movl 0x24(%r14), %ecx movq %rax, 0x140(%rsp) testl %ecx, 0x34(%rax) je 0x20a8d9 movq 0x10(%rdx), %rax movq %rax, 0x30(%rsp) cmpq $0x0, 0x10(%rax) movq 0x20(%rsp), %rcx jne 0x20a911 movq 0x140(%rsp), %rax cmpq $0x0, 0x48(%rax) jne 0x20a911 xorl %eax, %eax jmp 0x20a8f6 movq 0xa0(%rsp), %rax movq 0x40(%rsp), %rcx btcq %rax, %rcx movq %rcx, 0x40(%rsp) movb $0x1, %al movq 0x20(%rsp), %rcx testb %al, %al je 0x20ab58 movq 0x40(%rsp), %rax testq %rax, %rax jne 0x20a870 jmp 0x20a6fb movq %r11, 0x68(%rsp) movq %r10, 0x70(%rsp) movq %r9, 0x78(%rsp) movq %r8, 0x80(%rsp) movq %rdi, 0x88(%rsp) movq 0xa0(%rsp), %rsi vmovss 0x3c0(%rsp,%rsi,4), %xmm0 vmovss 0x3e0(%rsp,%rsi,4), %xmm1 movq %rdx, 0xc0(%rsp) movq 0x8(%rdx), %rcx movl 0x2c0(%rsp,%rsi,4), %edx vmovss 0x420(%rsp,%rsi,4), %xmm2 vmovss 0x440(%rsp,%rsi,4), %xmm3 vmovss 0x460(%rsp,%rsi,4), %xmm4 vmovss %xmm2, 0xe0(%rsp) vmovss %xmm3, 0xe4(%rsp) vmovss %xmm4, 0xe8(%rsp) vmovss %xmm0, 0xec(%rsp) vmovss %xmm1, 0xf0(%rsp) movl %edx, 0xf4(%rsp) movq 0x18(%rsp), %rax movl %eax, 0xf8(%rsp) movl (%rcx), %eax movl %eax, 0xfc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0x100(%rsp) vmovss 0x20(%r14), %xmm0 vmovss %xmm0, 0x18(%rsp) vmovss 0x400(%rsp,%rsi,4), %xmm0 vmovss %xmm0, 0x20(%r14) orl $-0x1, 0x2c(%rsp) leaq 0x2c(%rsp), %rax movq %rax, 0x110(%rsp) movq 0x140(%rsp), %rdx movq 0x18(%rdx), %rax movq %rax, 0x118(%rsp) movq %rcx, 0x120(%rsp) movq %r14, 0x128(%rsp) leaq 0xe0(%rsp), %rax movq %rax, 0x130(%rsp) movl $0x1, 0x138(%rsp) movq 0x48(%rdx), %rax testq %rax, %rax je 0x20aa5a leaq 0x110(%rsp), %rdi vzeroupper callq %rax movq 0x110(%rsp), %rax cmpl $0x0, (%rax) je 0x20aa9e movq 0x30(%rsp), %rax movq 0x10(%rax), %rax testq %rax, %rax je 0x20aa9a movq 0x30(%rsp), %rcx testb $0x2, (%rcx) jne 0x20aa80 movq 0x140(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x20aa8d leaq 0x110(%rsp), %rdi vzeroupper callq %rax movq 0x110(%rsp), %rax cmpl $0x0, (%rax) je 0x20aa9e xorl %eax, %eax jmp 0x20aac2 vmovss 0x18(%rsp), %xmm0 vmovss %xmm0, 0x20(%r14) movq 0x40(%rsp), %rax movq 0xa0(%rsp), %rcx btcq %rcx, %rax movq %rax, 0x40(%rsp) movb $0x1, %al movq 0xc0(%rsp), %rdx movq 0x88(%rsp), %rdi movq 0x80(%rsp), %r8 movq 0x78(%rsp), %r9 movq 0x70(%rsp), %r10 movq 0x68(%rsp), %r11 movq 0x20(%rsp), %rcx xorl %esi, %esi jmp 0x20a8f6 vmovaps 0x240(%rsp), %ymm6 vmovaps 0x220(%rsp), %ymm7 vmovaps 0x200(%rsp), %ymm8 vmovaps 0x1e0(%rsp), %ymm9 vmovaps 0x1c0(%rsp), %ymm10 vmovaps 0x1a0(%rsp), %ymm3 movq 0x38(%rsp), %r15 vmovaps 0x180(%rsp), %ymm4 vmovaps 0x160(%rsp), %ymm5 leaq 0x480(%rsp), %rbx cmpl $0x3, %esi jne 0x20a392 jmp 0x20a1d8 testb $0x1, %bl vmovaps 0x240(%rsp), %ymm6 vmovaps 0x220(%rsp), %ymm7 vmovaps 0x200(%rsp), %ymm8 vmovaps 0x1e0(%rsp), %ymm9 vmovaps 0x1c0(%rsp), %ymm10 vmovaps 0x1a0(%rsp), %ymm3 movq 0x38(%rsp), %r15 vmovaps 0x180(%rsp), %ymm4 vmovaps 0x160(%rsp), %ymm5 leaq 0x480(%rsp), %rbx je 0x20ab4a movl $0xff800000, 0x20(%r14) # imm = 0xFF800000 pushq $0x3 popq %rsi jmp 0x20ab4a nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, false, embree::avx::ArrayIntersector1<embree::avx::QuadMiIntersector1Moeller<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x1640, %rsp # imm = 0x1640 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x20abf4 leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %rsi, %r14 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x20abe2 movq %rdx, %r8 leaq 0x488(%rsp), %r9 vmovaps 0x10(%r14), %xmm3 vmaxss 0xc(%r14), %xmm2, %xmm1 vbroadcastss 0x1d1629e(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce63b5(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 movq 0x70(%rax), %rax vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1ce1ac1(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss (%r14), %ymm6 vbroadcastss 0x4(%r14), %ymm7 vbroadcastss 0x8(%r14), %ymm8 vaddps %xmm3, %xmm4, %xmm3 xorl %r10d, %r10d vucomiss %xmm2, %xmm3 movq %rax, -0x8(%r9) setb %r10b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 shll $0x5, %r10d xorl %r11d, %r11d vucomiss %xmm2, %xmm4 setb %r11b shll $0x5, %r11d orq $0x40, %r11 xorl %esi, %esi vucomiss %xmm2, %xmm5 setb %sil shll $0x5, %esi orq $0x80, %rsi movq %r10, %rdi xorq $0x20, %rdi movq %r11, %r15 xorq $0x20, %r15 movq %rsi, %rcx xorq $0x20, %rcx vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm4 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm5 leaq 0x1f4527a(%rip), %rax # 0x214ff80 vbroadcastf128 0xf0(%rax), %ymm0 # ymm0 = mem[0,1,0,1] vmovaps %ymm0, 0x260(%rsp) vperm2f128 $0x2, (%rax), %ymm0, %ymm1 # ymm1 = mem[0,1],ymm0[0,1] leaq 0x480(%rsp), %rbx vbroadcastss 0x1ce5c9d(%rip), %ymm0 # 0x1ef09cc vbroadcastss 0x1ce19dc(%rip), %ymm2 # 0x1eec714 vmovaps %ymm1, 0x2a0(%rsp) vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x280(%rsp) vmovaps %ymm6, 0x240(%rsp) vmovaps %ymm7, 0x220(%rsp) vmovaps %ymm8, 0x200(%rsp) vmovaps %ymm9, 0x1e0(%rsp) vmovaps %ymm10, 0x1c0(%rsp) vmovaps %ymm3, 0x1a0(%rsp) movq %rsi, 0x30(%rsp) movq %rdi, 0x28(%rsp) movq %r15, 0x20(%rsp) movq %rcx, 0x18(%rsp) vmovaps %ymm4, 0x180(%rsp) vmovaps %ymm5, 0x160(%rsp) cmpq %rbx, %r9 je 0x20abe2 movq -0x8(%r9), %r13 addq $-0x8, %r9 testb $0x8, %r13b jne 0x20ae40 vmovaps 0x40(%r13,%r10), %ymm0 vsubps %ymm6, %ymm0, %ymm0 vmulps %ymm0, %ymm9, %ymm0 vmovaps 0x40(%r13,%r11), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm10, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r13,%rsi), %ymm1 vsubps %ymm8, %ymm1, %ymm1 vmulps %ymm1, %ymm3, %ymm1 vmaxps %ymm4, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r13,%rdi), %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmovaps 0x40(%r13,%r15), %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm10, %ymm2 vminps %ymm2, %ymm1, %ymm1 vmovaps 0x40(%r13,%rcx), %ymm2 vsubps %ymm8, %ymm2, %ymm2 vmulps %ymm2, %ymm3, %ymm2 vminps %ymm5, %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r12d testb $0x8, %r13b jne 0x20ae86 testq %r12, %r12 je 0x20ae8a andq $-0x10, %r13 bsfq %r12, %rax leaq -0x1(%r12), %rdx movq (%r13,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r12, %rdx jne 0x20ae8f movq %rax, %r13 xorl %eax, %eax testl %eax, %eax je 0x20adbd jmp 0x20aee1 pushq $0x6 jmp 0x20ae8c pushq $0x4 popq %rax jmp 0x20ae7c movq %rcx, %rbx movq %rax, (%r9) addq $0x8, %r9 bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r13,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x20aed1 movq %rcx, (%r9) addq $0x8, %r9 bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x20aea1 movq %rcx, %r13 movq %rbx, %rcx leaq 0x480(%rsp), %rbx jmp 0x20ae7a cmpl $0x6, %eax jne 0x20b648 movl %r13d, %edx andl $0xf, %edx xorl %eax, %eax addq $-0x8, %rdx movq %rdx, 0x78(%rsp) setne %bl je 0x20b640 andq $-0x10, %r13 movq (%r8), %rax movq %rax, 0x38(%rsp) xorl %r15d, %r15d imulq $0x60, %r15, %rax prefetcht0 (%r13,%rax) prefetcht0 0x40(%r13,%rax) movl 0x40(%r13,%rax), %edx movq 0x38(%rsp), %rcx movq 0x228(%rcx), %rcx movq (%rcx,%rdx,8), %rdx movl (%r13,%rax), %esi movl 0x4(%r13,%rax), %edi vmovups (%rdx,%rsi,4), %xmm4 movl 0x10(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm2 movl 0x20(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm1 movl 0x30(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm0 movl 0x44(%r13,%rax), %edx movq (%rcx,%rdx,8), %rdx vmovups (%rdx,%rdi,4), %xmm7 movl 0x14(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm6 movl 0x24(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm5 movl 0x34(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm3 movl 0x48(%r13,%rax), %edx movq (%rcx,%rdx,8), %rdx movl 0x8(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm10 movl 0x18(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm11 movl 0x28(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm9 movl 0x38(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm8 movl 0x4c(%r13,%rax), %edx movq (%rcx,%rdx,8), %rcx movl 0xc(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm12 movl 0x1c(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm13 movl 0x2c(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm14 movl 0x3c(%r13,%rax), %edx vunpcklps %xmm10, %xmm4, %xmm15 # xmm15 = xmm4[0],xmm10[0],xmm4[1],xmm10[1] vunpckhps %xmm10, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm10[2],xmm4[3],xmm10[3] vunpcklps %xmm12, %xmm7, %xmm10 # xmm10 = xmm7[0],xmm12[0],xmm7[1],xmm12[1] vunpckhps %xmm12, %xmm7, %xmm7 # xmm7 = xmm7[2],xmm12[2],xmm7[3],xmm12[3] vmovups (%rcx,%rdx,4), %xmm12 vunpcklps %xmm7, %xmm4, %xmm4 # xmm4 = xmm4[0],xmm7[0],xmm4[1],xmm7[1] vmovaps %ymm4, 0x40(%rsp) vunpcklps %xmm10, %xmm15, %xmm7 # xmm7 = xmm15[0],xmm10[0],xmm15[1],xmm10[1] vunpckhps %xmm10, %xmm15, %xmm10 # xmm10 = xmm15[2],xmm10[2],xmm15[3],xmm10[3] vunpcklps %xmm11, %xmm2, %xmm15 # xmm15 = xmm2[0],xmm11[0],xmm2[1],xmm11[1] vunpckhps %xmm11, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm11[2],xmm2[3],xmm11[3] vunpcklps %xmm13, %xmm6, %xmm11 # xmm11 = xmm6[0],xmm13[0],xmm6[1],xmm13[1] vunpckhps %xmm13, %xmm6, %xmm6 # xmm6 = xmm6[2],xmm13[2],xmm6[3],xmm13[3] vunpcklps %xmm6, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] vunpcklps %xmm11, %xmm15, %xmm6 # xmm6 = xmm15[0],xmm11[0],xmm15[1],xmm11[1] vunpckhps %xmm11, %xmm15, %xmm11 # xmm11 = xmm15[2],xmm11[2],xmm15[3],xmm11[3] vunpcklps %xmm9, %xmm1, %xmm13 # xmm13 = xmm1[0],xmm9[0],xmm1[1],xmm9[1] vunpckhps %xmm9, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm9[2],xmm1[3],xmm9[3] vunpcklps %xmm14, %xmm5, %xmm9 # xmm9 = xmm5[0],xmm14[0],xmm5[1],xmm14[1] vunpckhps %xmm14, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm14[2],xmm5[3],xmm14[3] vunpcklps %xmm5, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm5[0],xmm1[1],xmm5[1] vunpcklps %xmm9, %xmm13, %xmm5 # xmm5 = xmm13[0],xmm9[0],xmm13[1],xmm9[1] vunpckhps %xmm9, %xmm13, %xmm9 # xmm9 = xmm13[2],xmm9[2],xmm13[3],xmm9[3] vunpcklps %xmm8, %xmm0, %xmm13 # xmm13 = xmm0[0],xmm8[0],xmm0[1],xmm8[1] vunpckhps %xmm8, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm8[2],xmm0[3],xmm8[3] vunpcklps %xmm12, %xmm3, %xmm8 # xmm8 = xmm3[0],xmm12[0],xmm3[1],xmm12[1] vunpckhps %xmm12, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm12[2],xmm3[3],xmm12[3] vunpcklps %xmm3, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1] vunpcklps %xmm8, %xmm13, %xmm12 # xmm12 = xmm13[0],xmm8[0],xmm13[1],xmm8[1] vunpckhps %xmm8, %xmm13, %xmm8 # xmm8 = xmm13[2],xmm8[2],xmm13[3],xmm8[3] vbroadcastf128 0x40(%r13,%rax), %ymm3 # ymm3 = mem[0,1,0,1] vmovaps %ymm3, 0x2e0(%rsp) vbroadcastf128 0x50(%r13,%rax), %ymm3 # ymm3 = mem[0,1,0,1] vmovaps %ymm3, 0x2c0(%rsp) vinsertf128 $0x1, %xmm5, %ymm7, %ymm4 vinsertf128 $0x1, %xmm9, %ymm10, %ymm5 vmovaps 0x40(%rsp), %ymm3 vinsertf128 $0x1, %xmm1, %ymm3, %ymm7 vinsertf128 $0x1, %xmm6, %ymm6, %ymm1 vinsertf128 $0x1, %xmm11, %ymm11, %ymm6 vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vinsertf128 $0x1, %xmm12, %ymm12, %ymm9 vinsertf128 $0x1, %xmm8, %ymm8, %ymm10 vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vsubps %ymm1, %ymm4, %ymm3 vsubps %ymm6, %ymm5, %ymm12 vmovaps %ymm12, 0x100(%rsp) vsubps %ymm2, %ymm7, %ymm8 vsubps %ymm4, %ymm9, %ymm9 vsubps %ymm5, %ymm10, %ymm10 vsubps %ymm7, %ymm0, %ymm11 vmulps %ymm11, %ymm12, %ymm0 vmulps %ymm10, %ymm8, %ymm1 vsubps %ymm0, %ymm1, %ymm0 vmovaps %ymm0, 0x140(%rsp) vmulps %ymm9, %ymm8, %ymm0 vmovaps %ymm3, 0x40(%rsp) vmulps %ymm3, %ymm11, %ymm2 vsubps %ymm0, %ymm2, %ymm6 vmulps %ymm3, %ymm10, %ymm0 vmulps %ymm9, %ymm12, %ymm12 vsubps %ymm0, %ymm12, %ymm3 vbroadcastss (%r14), %ymm12 vbroadcastss 0x4(%r14), %ymm13 vbroadcastss 0x8(%r14), %ymm14 vbroadcastss 0x14(%r14), %ymm15 vsubps %ymm12, %ymm4, %ymm2 vbroadcastss 0x18(%r14), %ymm12 vsubps %ymm13, %ymm5, %ymm5 vsubps %ymm14, %ymm7, %ymm7 vmulps %ymm7, %ymm15, %ymm13 vmulps %ymm5, %ymm12, %ymm14 vsubps %ymm13, %ymm14, %ymm13 vbroadcastss 0x10(%r14), %ymm14 vmulps %ymm2, %ymm12, %ymm0 vmulps %ymm7, %ymm14, %ymm1 vsubps %ymm0, %ymm1, %ymm0 vmulps %ymm5, %ymm14, %ymm1 vmovaps %ymm2, 0x120(%rsp) vmulps %ymm2, %ymm15, %ymm2 vsubps %ymm1, %ymm2, %ymm1 vmovaps %ymm3, 0xe0(%rsp) vmulps %ymm3, %ymm12, %ymm2 vmulps %ymm6, %ymm15, %ymm12 vmovaps %ymm6, %ymm15 vaddps %ymm2, %ymm12, %ymm2 vmovaps 0x140(%rsp), %ymm3 vmulps %ymm3, %ymm14, %ymm12 vaddps %ymm2, %ymm12, %ymm2 vmulps %ymm1, %ymm11, %ymm11 vmulps %ymm0, %ymm10, %ymm10 vaddps %ymm11, %ymm10, %ymm10 vmulps %ymm13, %ymm9, %ymm9 vaddps %ymm10, %ymm9, %ymm10 vmulps %ymm1, %ymm8, %ymm1 vmulps 0x100(%rsp), %ymm0, %ymm0 vaddps %ymm1, %ymm0, %ymm0 vbroadcastss 0x1d15cd0(%rip), %ymm1 # 0x1f20ec0 vandps %ymm1, %ymm2, %ymm9 vxorps %ymm10, %ymm9, %ymm6 vmulps 0x40(%rsp), %ymm13, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vxorps %ymm0, %ymm9, %ymm8 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %ymm10, %ymm6, %ymm0 vcmpnltps %ymm10, %ymm8, %ymm1 vandps %ymm1, %ymm0, %ymm0 vbroadcastss 0x1d15c9f(%rip), %ymm1 # 0x1f20ec4 vandps %ymm1, %ymm2, %ymm4 vcmpneqps %ymm2, %ymm10, %ymm1 vandps %ymm1, %ymm0, %ymm0 vaddps %ymm6, %ymm8, %ymm1 vcmpleps %ymm4, %ymm1, %ymm1 vandps %ymm1, %ymm0, %ymm10 vtestps 0x260(%rsp), %ymm10 jne 0x20b261 incq %r15 cmpq 0x78(%rsp), %r15 setb %bl jne 0x20af13 jmp 0x20b5d8 vmovaps %ymm3, %ymm14 vandps 0x260(%rsp), %ymm10, %ymm10 vmulps 0xe0(%rsp), %ymm7, %ymm0 vmulps %ymm5, %ymm15, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vmulps 0x120(%rsp), %ymm3, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vxorps %ymm0, %ymm9, %ymm3 vbroadcastss 0xc(%r14), %ymm0 vmulps %ymm0, %ymm4, %ymm0 vcmpltps %ymm3, %ymm0, %ymm0 vbroadcastss 0x20(%r14), %ymm1 vmulps %ymm1, %ymm4, %ymm1 vcmpleps %ymm1, %ymm3, %ymm1 vandps %ymm0, %ymm1, %ymm5 vtestps %ymm10, %ymm5 je 0x20b24b vandps %ymm5, %ymm10, %ymm0 vmovaps %ymm6, 0x300(%rsp) vmovaps %ymm8, 0x320(%rsp) vmovaps %ymm3, 0x340(%rsp) vmovaps %ymm4, 0x360(%rsp) vmovaps %ymm0, 0x3a0(%rsp) vmovaps 0x300(%rsp), %ymm0 vmovaps 0x320(%rsp), %ymm1 vsubps %ymm1, %ymm4, %ymm2 vmovaps 0x2a0(%rsp), %ymm3 vblendvps %ymm3, %ymm2, %ymm0, %ymm2 vmovaps %ymm2, 0x300(%rsp) vsubps %ymm0, %ymm4, %ymm0 vblendvps %ymm3, %ymm0, %ymm1, %ymm0 vmovaps %ymm0, 0x320(%rsp) vmovaps 0x280(%rsp), %ymm3 vmulps %ymm3, %ymm14, %ymm1 vmovaps %ymm1, 0x420(%rsp) vmulps %ymm3, %ymm15, %ymm1 vmovaps %ymm1, 0x440(%rsp) vrcpps %ymm4, %ymm1 vmulps 0xe0(%rsp), %ymm3, %ymm3 vmovaps %ymm3, 0x460(%rsp) vmulps %ymm1, %ymm4, %ymm3 vbroadcastss 0x1ce13a3(%rip), %ymm4 # 0x1eec714 vsubps %ymm3, %ymm4, %ymm3 vmulps %ymm3, %ymm1, %ymm3 vaddps %ymm3, %ymm1, %ymm1 vmulps 0x340(%rsp), %ymm1, %ymm3 vmovaps %ymm3, 0x400(%rsp) vmulps %ymm1, %ymm2, %ymm2 vmovaps %ymm2, 0x3c0(%rsp) vmulps %ymm1, %ymm0, %ymm0 vmovaps %ymm0, 0x3e0(%rsp) vmovaps 0x3a0(%rsp), %ymm0 vmovmskps %ymm0, %edx bsfq %rdx, %rsi movl 0x2e0(%rsp,%rsi,4), %eax movq 0x38(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rdi movl 0x24(%r14), %ecx testl %ecx, 0x34(%rdi) je 0x20b3f0 movq 0x10(%r8), %rcx cmpq $0x0, 0x10(%rcx) jne 0x20b408 cmpq $0x0, 0x48(%rdi) jne 0x20b408 xorl %eax, %eax jmp 0x20b3f6 btcq %rsi, %rdx movb $0x1, %al testb %al, %al je 0x20b656 testq %rdx, %rdx jne 0x20b3b6 jmp 0x20b24b movq %rcx, 0x100(%rsp) movq %rdx, 0x40(%rsp) movq %r11, 0x70(%rsp) movq %r10, 0x120(%rsp) movq %r9, 0xe0(%rsp) vmovss 0x3c0(%rsp,%rsi,4), %xmm0 vmovss 0x3e0(%rsp,%rsi,4), %xmm1 movq %r8, 0x140(%rsp) movq 0x8(%r8), %rcx movl 0x2c0(%rsp,%rsi,4), %edx vmovss 0x420(%rsp,%rsi,4), %xmm2 vmovss 0x440(%rsp,%rsi,4), %xmm3 vmovss 0x460(%rsp,%rsi,4), %xmm4 vmovss %xmm2, 0x80(%rsp) vmovss %xmm3, 0x84(%rsp) vmovss %xmm4, 0x88(%rsp) vmovss %xmm0, 0x8c(%rsp) vmovss %xmm1, 0x90(%rsp) movl %edx, 0x94(%rsp) movl %eax, 0x98(%rsp) movl (%rcx), %eax movl %eax, 0x9c(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xa0(%rsp) vmovss 0x20(%r14), %xmm0 vmovss %xmm0, 0x10(%rsp) movq %rsi, 0x68(%rsp) vmovss 0x400(%rsp,%rsi,4), %xmm0 vmovss %xmm0, 0x20(%r14) orl $-0x1, 0x14(%rsp) leaq 0x14(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%rdi), %rax movq %rax, 0xb8(%rsp) movq %rcx, 0xc0(%rsp) movq %r14, 0xc8(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x1, 0xd8(%rsp) movq %rdi, 0x60(%rsp) movq 0x48(%rdi), %rax testq %rax, %rax je 0x20b549 leaq 0xb0(%rsp), %rdi vzeroupper callq %rax movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x20b590 movq 0x100(%rsp), %rax movq 0x10(%rax), %rax testq %rax, %rax je 0x20b58c movq 0x100(%rsp), %rcx testb $0x2, (%rcx) jne 0x20b572 movq 0x60(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x20b57f leaq 0xb0(%rsp), %rdi vzeroupper callq %rax movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x20b590 xorl %eax, %eax jmp 0x20b5b1 vmovss 0x10(%rsp), %xmm0 vmovss %xmm0, 0x20(%r14) movq 0x40(%rsp), %rax movq 0x68(%rsp), %rcx btcq %rcx, %rax movq %rax, 0x40(%rsp) movb $0x1, %al movq 0x140(%rsp), %r8 movq 0xe0(%rsp), %r9 movq 0x120(%rsp), %r10 movq 0x70(%rsp), %r11 movq 0x40(%rsp), %rdx jmp 0x20b3f6 vmovaps 0x240(%rsp), %ymm6 vmovaps 0x220(%rsp), %ymm7 vmovaps 0x200(%rsp), %ymm8 vmovaps 0x1e0(%rsp), %ymm9 vmovaps 0x1c0(%rsp), %ymm10 vmovaps 0x1a0(%rsp), %ymm3 movq 0x30(%rsp), %rsi movq 0x28(%rsp), %rdi movq 0x20(%rsp), %r15 movq 0x18(%rsp), %rcx vmovaps 0x180(%rsp), %ymm4 vmovaps 0x160(%rsp), %ymm5 leaq 0x480(%rsp), %rbx xorl %eax, %eax jmp 0x20b648 leaq 0x480(%rsp), %rbx cmpl $0x3, %eax jne 0x20adac jmp 0x20abe2 testb $0x1, %bl vmovaps 0x240(%rsp), %ymm6 vmovaps 0x220(%rsp), %ymm7 vmovaps 0x200(%rsp), %ymm8 vmovaps 0x1e0(%rsp), %ymm9 vmovaps 0x1c0(%rsp), %ymm10 vmovaps 0x1a0(%rsp), %ymm3 movq 0x30(%rsp), %rsi movq 0x28(%rsp), %rdi movq 0x20(%rsp), %r15 movq 0x18(%rsp), %rcx vmovaps 0x180(%rsp), %ymm4 vmovaps 0x160(%rsp), %ymm5 leaq 0x480(%rsp), %rbx movl $0x0, %eax je 0x20b648 movl $0xff800000, 0x20(%r14) # imm = 0xFF800000 pushq $0x3 popq %rax jmp 0x20b648	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, true, embree::avx::ArrayIntersector1<embree::avx::QuadMvIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x17c0, %rsp # imm = 0x17C0 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x20c41d vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x20c41d leaq 0x608(%rsp), %rcx movq 0x70(%rax), %rax movq %rcx, 0x10(%rsp) movq %rax, -0x8(%rcx) vmovaps 0x10(%rsi), %xmm3 vmaxss 0xc(%rsi), %xmm2, %xmm1 vbroadcastss 0x1d1578f(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce58a6(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vbroadcastss 0x1ce0fc4(%rip), %xmm5 # 0x1eec714 vdivps %xmm3, %xmm5, %xmm3 vbroadcastss 0x1d15803(%rip), %xmm5 # 0x1f20f60 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vbroadcastss 0x1d147a4(%rip), %xmm4 # 0x1f1ff10 vmulps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1d1479b(%rip), %xmm5 # 0x1f1ff14 vbroadcastss (%rsi), %ymm7 vbroadcastss 0x4(%rsi), %ymm8 vbroadcastss 0x8(%rsi), %ymm9 vmulps %xmm5, %xmm3, %xmm3 xorl %r9d, %r9d vucomiss %xmm2, %xmm4 setb %r9b vshufps $0x0, %xmm4, %xmm4, %xmm5 # xmm5 = xmm4[0,0,0,0] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vmovshdup %xmm4, %xmm5 # xmm5 = xmm4[1,1,3,3] vshufps $0x55, %xmm4, %xmm4, %xmm6 # xmm6 = xmm4[1,1,1,1] vinsertf128 $0x1, %xmm6, %ymm6, %ymm11 vshufpd $0x1, %xmm4, %xmm4, %xmm6 # xmm6 = xmm4[1,0] vshufps $0xaa, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] vinsertf128 $0x1, %xmm4, %ymm4, %ymm12 vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm13 vshufps $0x55, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm14 vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] shll $0x5, %r9d xorl %r10d, %r10d vucomiss %xmm2, %xmm5 setb %r10b shll $0x5, %r10d orq $0x40, %r10 xorl %r11d, %r11d vucomiss %xmm2, %xmm6 vinsertf128 $0x1, %xmm3, %ymm3, %ymm5 setb %r11b shll $0x5, %r11d orq $0x80, %r11 movq %r9, %r14 xorq $0x20, %r14 movq %r10, %r15 xorq $0x20, %r15 movq %r11, %r12 xorq $0x20, %r12 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm6 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm15 leaq 0x1f4473d(%rip), %rax # 0x214ff80 vmovups (%rax), %ymm0 vinsertf128 $0x1, 0xf0(%rax), %ymm0, %ymm1 vbroadcastss 0x1ce0eba(%rip), %ymm2 # 0x1eec714 vbroadcastss 0x1ce5169(%rip), %ymm0 # 0x1ef09cc vmovaps %ymm1, 0x2e0(%rsp) vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x2c0(%rsp) vmovaps %ymm7, 0x260(%rsp) vmovaps %ymm8, 0x240(%rsp) vmovaps %ymm9, 0x220(%rsp) vmovaps %ymm10, 0x200(%rsp) vmovaps %ymm11, 0x1e0(%rsp) vmovaps %ymm12, 0x1c0(%rsp) vmovaps %ymm13, 0x1a0(%rsp) vmovaps %ymm14, 0x180(%rsp) vmovaps %ymm5, 0x160(%rsp) movq %r14, 0x50(%rsp) movq %r15, 0x48(%rsp) movq %r12, 0x40(%rsp) vmovaps %ymm6, 0x140(%rsp) vmovaps %ymm15, 0x120(%rsp) leaq 0x600(%rsp), %rax cmpq %rax, 0x10(%rsp) je 0x20c41d movq 0x10(%rsp), %rax movq -0x8(%rax), %rbx addq $-0x8, %rax movq %rax, 0x10(%rsp) testb $0x8, %bl jne 0x20b99b vmovaps 0x40(%rbx,%r9), %ymm0 vsubps %ymm7, %ymm0, %ymm0 vmulps %ymm0, %ymm10, %ymm0 vmovaps 0x40(%rbx,%r10), %ymm1 vsubps %ymm8, %ymm1, %ymm1 vmulps %ymm1, %ymm11, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%r11), %ymm1 vsubps %ymm9, %ymm1, %ymm1 vmulps %ymm1, %ymm12, %ymm1 vmaxps %ymm6, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%r14), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm13, %ymm1 vmovaps 0x40(%rbx,%r15), %ymm2 vsubps %ymm8, %ymm2, %ymm2 vmulps %ymm2, %ymm14, %ymm2 vminps %ymm2, %ymm1, %ymm1 vmovaps 0x40(%rbx,%r12), %ymm2 vsubps %ymm9, %ymm2, %ymm2 vmulps %ymm2, %ymm5, %ymm2 vminps %ymm15, %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r13d testb $0x8, %bl jne 0x20b9de testq %r13, %r13 je 0x20b9e2 andq $-0x10, %rbx bsfq %r13, %rax leaq -0x1(%r13), %rdi xorl %ecx, %ecx movq (%rbx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r13, %rdi jne 0x20b9e7 movq %rax, %rbx testl %ecx, %ecx je 0x20b912 jmp 0x20ba40 pushq $0x6 jmp 0x20b9e4 pushq $0x4 popq %rcx jmp 0x20b9d4 movq 0x10(%rsp), %rcx movq %rax, (%rcx) addq $0x8, %rcx movq %rcx, 0x10(%rsp) bsfq %rdi, %rcx leaq -0x1(%rdi), %rax movq (%rbx,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdi, %rax je 0x20ba39 movq 0x10(%rsp), %rdi movq %rcx, (%rdi) addq $0x8, %rdi movq %rdi, 0x10(%rsp) bsfq %rax, %rcx leaq -0x1(%rax), %rdi jmp 0x20ba00 movq %rcx, %rbx xorl %ecx, %ecx jmp 0x20b9d4 cmpl $0x6, %ecx jne 0x20c414 movl %ebx, %edi andl $0xf, %edi xorl %ecx, %ecx addq $-0x8, %rdi setne %al je 0x20c414 andq $-0x10, %rbx xorl %r8d, %r8d movq %r13, 0x18(%rsp) movb %al, 0xf(%rsp) movq %r8, 0x58(%rsp) imulq $0xe0, %r8, %rax vbroadcastf128 0xd0(%rbx,%rax), %ymm0 # ymm0 = mem[0,1,0,1] vbroadcastf128 0xc0(%rbx,%rax), %ymm1 # ymm1 = mem[0,1,0,1] vmovaps %ymm1, 0x420(%rsp) vmovaps %ymm0, 0x400(%rsp) vmovaps (%rbx,%rax), %xmm0 vmovaps 0x10(%rbx,%rax), %xmm1 vmovaps 0x20(%rbx,%rax), %xmm2 vinsertf128 $0x1, 0x60(%rbx,%rax), %ymm0, %ymm0 vinsertf128 $0x1, 0x70(%rbx,%rax), %ymm1, %ymm3 vinsertf128 $0x1, 0x80(%rbx,%rax), %ymm2, %ymm2 vbroadcastf128 0x30(%rbx,%rax), %ymm4 # ymm4 = mem[0,1,0,1] vbroadcastf128 0x40(%rbx,%rax), %ymm5 # ymm5 = mem[0,1,0,1] vbroadcastf128 0x50(%rbx,%rax), %ymm14 # ymm14 = mem[0,1,0,1] vbroadcastf128 0x90(%rbx,%rax), %ymm8 # ymm8 = mem[0,1,0,1] vbroadcastf128 0xa0(%rbx,%rax), %ymm9 # ymm9 = mem[0,1,0,1] vbroadcastf128 0xb0(%rbx,%rax), %ymm10 # ymm10 = mem[0,1,0,1] vbroadcastss (%rsi), %ymm11 vbroadcastss 0x4(%rsi), %ymm12 vbroadcastss 0x8(%rsi), %ymm13 vsubps %ymm11, %ymm0, %ymm1 vsubps %ymm12, %ymm3, %ymm15 vsubps %ymm13, %ymm2, %ymm3 vsubps %ymm11, %ymm4, %ymm0 vmovaps %ymm0, 0x100(%rsp) vsubps %ymm12, %ymm5, %ymm7 vsubps %ymm13, %ymm14, %ymm14 vsubps %ymm11, %ymm8, %ymm2 vsubps %ymm12, %ymm9, %ymm0 vsubps %ymm13, %ymm10, %ymm5 vmovaps %ymm5, 0x60(%rsp) vsubps %ymm1, %ymm2, %ymm11 vmovaps %ymm0, 0xe0(%rsp) vsubps %ymm15, %ymm0, %ymm12 vsubps %ymm3, %ymm5, %ymm6 vaddps %ymm0, %ymm15, %ymm4 vaddps %ymm3, %ymm5, %ymm5 vmulps %ymm4, %ymm6, %ymm8 vmulps %ymm5, %ymm12, %ymm9 vsubps %ymm8, %ymm9, %ymm9 vaddps %ymm1, %ymm2, %ymm8 vmulps %ymm5, %ymm11, %ymm5 vmovaps %ymm6, 0x380(%rsp) vmulps %ymm6, %ymm8, %ymm10 vsubps %ymm5, %ymm10, %ymm5 vmovaps %ymm12, 0x3a0(%rsp) vmulps %ymm8, %ymm12, %ymm8 vmovaps %ymm11, 0x3c0(%rsp) vmulps %ymm4, %ymm11, %ymm4 vsubps %ymm8, %ymm4, %ymm4 vbroadcastss 0x14(%rsi), %ymm10 vbroadcastss 0x18(%rsi), %ymm6 vmulps %ymm4, %ymm6, %ymm4 vmulps %ymm5, %ymm10, %ymm5 vbroadcastss 0x10(%rsi), %ymm13 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm9, %ymm13, %ymm5 vaddps %ymm4, %ymm5, %ymm11 vsubps %ymm7, %ymm15, %ymm9 vsubps %ymm14, %ymm3, %ymm12 vmovaps %ymm15, 0x280(%rsp) vaddps %ymm7, %ymm15, %ymm4 vmovaps %ymm3, 0x3e0(%rsp) vaddps %ymm3, %ymm14, %ymm5 vmulps %ymm4, %ymm12, %ymm15 vmulps %ymm5, %ymm9, %ymm3 vsubps %ymm15, %ymm3, %ymm3 vmovaps 0x100(%rsp), %ymm0 vsubps %ymm0, %ymm1, %ymm15 vmulps %ymm5, %ymm15, %ymm5 vmovaps %ymm1, 0x2a0(%rsp) vaddps %ymm0, %ymm1, %ymm8 vmovaps %ymm12, 0x360(%rsp) vmulps %ymm8, %ymm12, %ymm12 vsubps %ymm5, %ymm12, %ymm5 vmovaps %ymm9, %ymm12 vmulps %ymm8, %ymm9, %ymm8 vmulps %ymm4, %ymm15, %ymm4 vsubps %ymm8, %ymm4, %ymm4 vmulps %ymm4, %ymm6, %ymm4 vmulps %ymm5, %ymm10, %ymm5 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm3, %ymm13, %ymm3 vaddps %ymm4, %ymm3, %ymm4 vsubps %ymm2, %ymm0, %ymm5 vaddps %ymm2, %ymm0, %ymm2 vmovaps 0xe0(%rsp), %ymm0 vsubps %ymm0, %ymm7, %ymm9 vaddps %ymm0, %ymm7, %ymm1 vmovaps 0x60(%rsp), %ymm3 vsubps %ymm3, %ymm14, %ymm0 vaddps %ymm3, %ymm14, %ymm3 vmulps %ymm1, %ymm0, %ymm7 vmulps %ymm3, %ymm9, %ymm8 vsubps %ymm7, %ymm8, %ymm7 vmulps %ymm3, %ymm5, %ymm3 vmulps %ymm2, %ymm0, %ymm8 vsubps %ymm3, %ymm8, %ymm3 vmulps %ymm2, %ymm9, %ymm2 vmulps %ymm1, %ymm5, %ymm1 vsubps %ymm2, %ymm1, %ymm1 vmovaps %ymm6, 0xe0(%rsp) vmulps %ymm1, %ymm6, %ymm1 vmovaps %ymm10, 0x100(%rsp) vmulps %ymm3, %ymm10, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vmovaps %ymm13, 0x60(%rsp) vmulps %ymm7, %ymm13, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vaddps %ymm4, %ymm11, %ymm2 vaddps %ymm2, %ymm1, %ymm7 vminps %ymm4, %ymm11, %ymm2 vminps %ymm1, %ymm2, %ymm2 vbroadcastss 0x1d151d7(%rip), %ymm3 # 0x1f20ec4 vandps %ymm3, %ymm7, %ymm8 vbroadcastss 0x1d151d2(%rip), %ymm3 # 0x1f20ecc vmovaps %ymm8, 0x300(%rsp) vmulps %ymm3, %ymm8, %ymm3 vbroadcastss 0x1d151b0(%rip), %ymm8 # 0x1f20ec0 vxorps %ymm3, %ymm8, %ymm8 vcmpnltps %ymm8, %ymm2, %ymm2 vmovaps %ymm11, 0x340(%rsp) vmovaps %ymm4, 0x320(%rsp) vmaxps %ymm4, %ymm11, %ymm8 vmaxps %ymm1, %ymm8, %ymm1 vcmpleps %ymm3, %ymm1, %ymm1 vorps %ymm1, %ymm2, %ymm1 vtestps %ymm1, %ymm1 je 0x20c38f vmovaps 0x380(%rsp), %ymm6 vmovaps %ymm12, %ymm13 vmulps %ymm6, %ymm12, %ymm2 vmovaps 0x3a0(%rsp), %ymm4 vmovaps 0x360(%rsp), %ymm14 vmulps %ymm4, %ymm14, %ymm3 vsubps %ymm2, %ymm3, %ymm3 vmulps %ymm9, %ymm14, %ymm8 vmulps %ymm0, %ymm12, %ymm12 vsubps %ymm8, %ymm12, %ymm12 vbroadcastss 0x1d15139(%rip), %ymm10 # 0x1f20ec4 vandps %ymm2, %ymm10, %ymm2 vandps %ymm10, %ymm8, %ymm8 vcmpltps %ymm8, %ymm2, %ymm2 vblendvps %ymm2, %ymm3, %ymm12, %ymm12 vmulps %ymm0, %ymm15, %ymm2 vmulps %ymm6, %ymm15, %ymm3 vmovaps 0x3c0(%rsp), %ymm11 vmulps %ymm14, %ymm11, %ymm6 vsubps %ymm6, %ymm3, %ymm3 vmulps %ymm5, %ymm14, %ymm8 vsubps %ymm2, %ymm8, %ymm8 vandps %ymm6, %ymm10, %ymm6 vandps %ymm2, %ymm10, %ymm2 vcmpltps %ymm2, %ymm6, %ymm2 vmovaps %ymm9, %ymm0 vblendvps %ymm2, %ymm3, %ymm8, %ymm9 vmulps %ymm5, %ymm13, %ymm2 vmulps %ymm13, %ymm11, %ymm3 vmulps %ymm4, %ymm15, %ymm5 vmulps %ymm0, %ymm15, %ymm0 vsubps %ymm5, %ymm3, %ymm3 vsubps %ymm2, %ymm0, %ymm0 vandps %ymm5, %ymm10, %ymm5 vandps %ymm2, %ymm10, %ymm2 vcmpltps %ymm2, %ymm5, %ymm2 vblendvps %ymm2, %ymm3, %ymm0, %ymm0 vextractf128 $0x1, %ymm1, %xmm2 vpackssdw %xmm2, %xmm1, %xmm1 vmulps 0xe0(%rsp), %ymm0, %ymm2 vmulps 0x100(%rsp), %ymm9, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vmulps 0x60(%rsp), %ymm12, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vaddps %ymm2, %ymm2, %ymm3 vmulps 0x3e0(%rsp), %ymm0, %ymm2 vmulps 0x280(%rsp), %ymm9, %ymm5 vaddps %ymm2, %ymm5, %ymm2 vmulps 0x2a0(%rsp), %ymm12, %ymm5 vrcpps %ymm3, %ymm6 vaddps %ymm2, %ymm5, %ymm2 vaddps %ymm2, %ymm2, %ymm2 vmulps %ymm6, %ymm3, %ymm5 vbroadcastss 0x1ce08a9(%rip), %ymm8 # 0x1eec714 vsubps %ymm5, %ymm8, %ymm5 vmulps %ymm5, %ymm6, %ymm5 vaddps %ymm5, %ymm6, %ymm5 vbroadcastss 0xc(%rsi), %ymm6 vmulps %ymm5, %ymm2, %ymm2 vcmpleps %ymm2, %ymm6, %ymm5 vbroadcastss 0x20(%rsi), %ymm6 vcmpleps %ymm6, %ymm2, %ymm6 vandps %ymm5, %ymm6, %ymm5 vcmpneqps 0x1d15062(%rip), %ymm3, %ymm3 # 0x1f20f00 vandps %ymm3, %ymm5, %ymm3 vextractf128 $0x1, %ymm3, %xmm5 vpackssdw %xmm5, %xmm3, %xmm3 vpand %xmm1, %xmm3, %xmm1 vpmovsxwd %xmm1, %xmm3 vpshufd $0xee, %xmm1, %xmm5 # xmm5 = xmm1[2,3,2,3] vpmovsxwd %xmm5, %xmm5 vinsertf128 $0x1, %xmm5, %ymm3, %ymm3 vtestps %ymm3, %ymm3 je 0x20c38f vmovaps 0x340(%rsp), %ymm8 vmovaps %ymm8, 0x440(%rsp) vmovaps 0x320(%rsp), %ymm4 vmovaps %ymm4, 0x460(%rsp) vmovaps %ymm7, 0x480(%rsp) vmovaps %ymm12, 0x4a0(%rsp) vmovaps %ymm9, 0x4c0(%rsp) vmovaps %ymm0, 0x4e0(%rsp) vmovaps %ymm3, 0x500(%rsp) vmovaps %ymm2, 0x560(%rsp) vmovaps 0x2e0(%rsp), %ymm6 vmovaps %ymm6, 0x5e0(%rsp) movq (%rdx), %r8 vrcpps %ymm7, %ymm2 vmulps %ymm2, %ymm7, %ymm3 vbroadcastss 0x1ce07c4(%rip), %ymm5 # 0x1eec714 vsubps %ymm3, %ymm5, %ymm3 vmulps %ymm3, %ymm2, %ymm3 vaddps %ymm3, %ymm2, %ymm2 vbroadcastss 0x1ce5083(%rip), %ymm3 # 0x1ef0fe8 vmovaps 0x300(%rsp), %ymm7 vcmpnltps %ymm3, %ymm7, %ymm3 vandps %ymm2, %ymm3, %ymm2 vmulps %ymm2, %ymm8, %ymm3 vminps %ymm5, %ymm3, %ymm3 vmulps %ymm2, %ymm4, %ymm2 vminps %ymm5, %ymm2, %ymm2 vsubps %ymm3, %ymm5, %ymm4 vsubps %ymm2, %ymm5, %ymm5 vblendvps %ymm6, %ymm4, %ymm2, %ymm2 vblendvps %ymm6, %ymm5, %ymm3, %ymm3 vmovaps %ymm2, 0x540(%rsp) vmovaps %ymm3, 0x520(%rsp) vmovaps 0x2c0(%rsp), %ymm4 vmulps %ymm4, %ymm12, %ymm2 vmulps %ymm4, %ymm9, %ymm3 vmulps %ymm0, %ymm4, %ymm0 vmovaps %ymm2, 0x580(%rsp) vmovaps %ymm3, 0x5a0(%rsp) vmovaps %ymm0, 0x5c0(%rsp) vpsllw $0xf, %xmm1, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r12d vmovaps 0x260(%rsp), %ymm7 vmovaps 0x240(%rsp), %ymm8 vmovaps 0x220(%rsp), %ymm9 vmovaps 0x200(%rsp), %ymm10 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x160(%rsp), %ymm5 vmovaps 0x140(%rsp), %ymm6 vmovaps 0x120(%rsp), %ymm15 bsfq %r12, %r15 movl 0x420(%rsp,%r15,4), %eax movq 0x1e8(%r8), %rcx movq (%rcx,%rax,8), %r13 movl 0x24(%rsi), %ecx testl %ecx, 0x34(%r13) je 0x20c091 movq 0x10(%rdx), %r14 cmpq $0x0, 0x10(%r14) movl $0x0, %ecx jne 0x20c0b0 cmpq $0x0, 0x48(%r13) jne 0x20c0b0 xorl %eax, %eax movq 0x18(%rsp), %r13 jmp 0x20c09e btcq %r15, %r12 movb $0x1, %al movq 0x18(%rsp), %r13 xorl %ecx, %ecx testb %al, %al je 0x20c3f4 testq %r12, %r12 jne 0x20c051 jmp 0x20c367 movq %rdi, 0x100(%rsp) vmovss 0x520(%rsp,%r15,4), %xmm0 vmovd 0x540(%rsp,%r15,4), %xmm1 movq 0x8(%rdx), %rcx movl 0x400(%rsp,%r15,4), %edi vmovss 0x580(%rsp,%r15,4), %xmm2 vmovss 0x5a0(%rsp,%r15,4), %xmm3 vmovss 0x5c0(%rsp,%r15,4), %xmm4 vmovss %xmm2, 0x80(%rsp) vmovss %xmm3, 0x84(%rsp) vmovss %xmm4, 0x88(%rsp) vmovss %xmm0, 0x8c(%rsp) vmovd %xmm1, 0x90(%rsp) movl %edi, 0x94(%rsp) movl %eax, 0x98(%rsp) movl (%rcx), %eax movl %eax, 0x9c(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xa0(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0xe0(%rsp) vmovd 0x560(%rsp,%r15,4), %xmm0 vmovd %xmm0, 0x20(%rsi) orl $-0x1, 0x24(%rsp) leaq 0x24(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%r13), %rax movq %rax, 0xb8(%rsp) movq %rcx, 0xc0(%rsp) movq %rsi, 0xc8(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x1, 0xd8(%rsp) movq 0x48(%r13), %rax testq %rax, %rax movq %r9, 0x38(%rsp) movq %r10, 0x30(%rsp) movq %r11, 0x28(%rsp) movq %r8, 0x60(%rsp) je 0x20c280 leaq 0xb0(%rsp), %rdi movq %rdx, 0x2a0(%rsp) movq %rsi, 0x280(%rsp) vzeroupper callq %rax movq 0x60(%rsp), %r8 vmovaps 0x120(%rsp), %ymm15 vmovaps 0x140(%rsp), %ymm6 vmovaps 0x160(%rsp), %ymm5 movq 0x28(%rsp), %r11 movq 0x30(%rsp), %r10 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x200(%rsp), %ymm10 movq 0x38(%rsp), %r9 vmovaps 0x220(%rsp), %ymm9 vmovaps 0x240(%rsp), %ymm8 vmovaps 0x260(%rsp), %ymm7 movq 0x280(%rsp), %rsi movq 0x2a0(%rsp), %rdx movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x20c33f movq 0x10(%r14), %rax testq %rax, %rax je 0x20c33b testb $0x2, (%r14) jne 0x20c29e testb $0x40, 0x3e(%r13) je 0x20c32e leaq 0xb0(%rsp), %rdi movq %rdx, %r14 movq %rsi, %r13 vzeroupper callq %rax movq 0x60(%rsp), %r8 vmovaps 0x120(%rsp), %ymm15 vmovaps 0x140(%rsp), %ymm6 vmovaps 0x160(%rsp), %ymm5 movq 0x28(%rsp), %r11 movq 0x30(%rsp), %r10 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x200(%rsp), %ymm10 movq 0x38(%rsp), %r9 vmovaps 0x220(%rsp), %ymm9 vmovaps 0x240(%rsp), %ymm8 vmovaps 0x260(%rsp), %ymm7 movq %r13, %rsi movq %r14, %rdx movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x20c33f xorl %eax, %eax jmp 0x20c353 vmovd 0xe0(%rsp), %xmm0 vmovd %xmm0, 0x20(%rsi) btcq %r15, %r12 movb $0x1, %al movq 0x18(%rsp), %r13 xorl %ecx, %ecx movq 0x100(%rsp), %rdi jmp 0x20c09e movq 0x50(%rsp), %r14 movq 0x48(%rsp), %r15 movq 0x40(%rsp), %r12 movq 0x58(%rsp), %r8 incq %r8 cmpq %rdi, %r8 setb %al jne 0x20ba69 jmp 0x20c414 vmovaps 0x260(%rsp), %ymm7 vmovaps 0x240(%rsp), %ymm8 vmovaps 0x220(%rsp), %ymm9 vmovaps 0x200(%rsp), %ymm10 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x160(%rsp), %ymm5 vmovaps 0x140(%rsp), %ymm6 vmovaps 0x120(%rsp), %ymm15 jmp 0x20c376 testb $0x1, 0xf(%rsp) movq 0x50(%rsp), %r14 movq 0x48(%rsp), %r15 movq 0x40(%rsp), %r12 je 0x20c414 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rcx cmpl $0x3, %ecx jne 0x20b8ed leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, true, embree::avx::ArrayIntersector1<embree::avx::QuadMiIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x1800, %rsp # imm = 0x1800 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x20d2d1 movq %rsi, %r9 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x20d2d1 movq %rdx, %r8 leaq 0x648(%rsp), %r10 movq 0x70(%rax), %rax movq %rax, -0x8(%r10) vmovaps 0x10(%r9), %xmm3 vmaxss 0xc(%r9), %xmm2, %xmm1 vbroadcastss 0x1d14a30(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce4b47(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vbroadcastss 0x1ce0265(%rip), %xmm5 # 0x1eec714 vdivps %xmm3, %xmm5, %xmm3 vbroadcastss 0x1d14aa4(%rip), %xmm5 # 0x1f20f60 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vbroadcastss 0x1d13a45(%rip), %xmm4 # 0x1f1ff10 vmulps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1d13a3c(%rip), %xmm5 # 0x1f1ff14 vbroadcastss (%r9), %ymm7 vbroadcastss 0x4(%r9), %ymm8 vbroadcastss 0x8(%r9), %ymm9 vmulps %xmm5, %xmm3, %xmm3 xorl %r11d, %r11d vucomiss %xmm2, %xmm4 setb %r11b vshufps $0x0, %xmm4, %xmm4, %xmm5 # xmm5 = xmm4[0,0,0,0] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vmovshdup %xmm4, %xmm5 # xmm5 = xmm4[1,1,3,3] vshufps $0x55, %xmm4, %xmm4, %xmm6 # xmm6 = xmm4[1,1,1,1] vinsertf128 $0x1, %xmm6, %ymm6, %ymm11 vshufpd $0x1, %xmm4, %xmm4, %xmm6 # xmm6 = xmm4[1,0] vshufps $0xaa, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] vinsertf128 $0x1, %xmm4, %ymm4, %ymm12 vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm13 vshufps $0x55, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm14 vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] shll $0x5, %r11d xorl %r14d, %r14d vucomiss %xmm2, %xmm5 setb %r14b shll $0x5, %r14d orq $0x40, %r14 xorl %r15d, %r15d vucomiss %xmm2, %xmm6 vinsertf128 $0x1, %xmm3, %ymm3, %ymm5 setb %r15b shll $0x5, %r15d orq $0x80, %r15 movq %r11, %rsi xorq $0x20, %rsi movq %r14, %rdi xorq $0x20, %rdi movq %r15, %rax xorq $0x20, %rax movq %rax, 0x78(%rsp) vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm6 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm15 leaq 0x1f439d9(%rip), %rax # 0x214ff80 vmovups (%rax), %ymm0 vinsertf128 $0x1, 0xf0(%rax), %ymm0, %ymm1 vbroadcastss 0x1ce0156(%rip), %ymm2 # 0x1eec714 vbroadcastss 0x1ce4405(%rip), %ymm0 # 0x1ef09cc vmovaps %ymm1, 0x2e0(%rsp) vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x2c0(%rsp) vmovaps %ymm7, 0x280(%rsp) vmovaps %ymm8, 0x260(%rsp) vmovaps %ymm9, 0x240(%rsp) vmovaps %ymm10, 0x220(%rsp) vmovaps %ymm11, 0x200(%rsp) vmovaps %ymm12, 0x1e0(%rsp) vmovaps %ymm13, 0x1c0(%rsp) vmovaps %ymm14, 0x1a0(%rsp) movq %r14, 0x60(%rsp) movq %r15, 0x58(%rsp) vmovaps %ymm5, 0x180(%rsp) movq %rsi, 0x28(%rsp) movq %rdi, 0x20(%rsp) vmovaps %ymm6, 0x160(%rsp) vmovaps %ymm15, 0x140(%rsp) leaq 0x640(%rsp), %rax cmpq %rax, %r10 je 0x20d2d1 movq -0x8(%r10), %rbx addq $-0x8, %r10 testb $0x8, %bl jne 0x20c6fa vmovaps 0x40(%rbx,%r11), %ymm0 vsubps %ymm7, %ymm0, %ymm0 vmulps %ymm0, %ymm10, %ymm0 vmovaps 0x40(%rbx,%r14), %ymm1 vsubps %ymm8, %ymm1, %ymm1 vmulps %ymm1, %ymm11, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%r15), %ymm1 vsubps %ymm9, %ymm1, %ymm1 vmulps %ymm1, %ymm12, %ymm1 vmaxps %ymm6, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rbx,%rsi), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm13, %ymm1 vmovaps 0x40(%rbx,%rdi), %ymm2 vsubps %ymm8, %ymm2, %ymm2 vmulps %ymm2, %ymm14, %ymm2 vminps %ymm2, %ymm1, %ymm1 movq 0x78(%rsp), %rax vmovaps 0x40(%rbx,%rax), %ymm2 vsubps %ymm9, %ymm2, %ymm2 vmulps %ymm2, %ymm5, %ymm2 vminps %ymm15, %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r12d testb $0x8, %bl jne 0x20c740 testq %r12, %r12 je 0x20c744 andq $-0x10, %rbx bsfq %r12, %rax leaq -0x1(%r12), %rdx xorl %r13d, %r13d movq (%rbx,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r12, %rdx jne 0x20c74a movq %rax, %rbx testl %r13d, %r13d je 0x20c66f jmp 0x20c78d pushq $0x6 jmp 0x20c746 pushq $0x4 popq %r13 jmp 0x20c735 movq %rax, (%r10) addq $0x8, %r10 bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%rbx,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x20c788 movq %rcx, (%r10) addq $0x8, %r10 bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x20c759 movq %rcx, %rbx jmp 0x20c735 cmpl $0x6, %r13d jne 0x20d2c7 movl %ebx, %eax andl $0xf, %eax xorl %r13d, %r13d addq $-0x8, %rax setne %cl je 0x20d2c7 movb %cl, 0x1f(%rsp) andq $-0x10, %rbx movq (%r8), %rdx xorl %ecx, %ecx movq %r12, 0x30(%rsp) movq %rax, 0x68(%rsp) movq %rdx, 0x38(%rsp) movq %rcx, 0x70(%rsp) imulq $0x60, %rcx, %rax prefetcht0 (%rbx,%rax) prefetcht0 0x40(%rbx,%rax) movq %rdx, %rcx movl 0x40(%rbx,%rax), %edx movq 0x228(%rcx), %rcx movq (%rcx,%rdx,8), %rdx movl (%rbx,%rax), %esi movl 0x4(%rbx,%rax), %edi vmovups (%rdx,%rsi,4), %xmm4 movl 0x10(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm2 movl 0x20(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm1 movl 0x30(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm0 movl 0x44(%rbx,%rax), %edx movq (%rcx,%rdx,8), %rdx vmovups (%rdx,%rdi,4), %xmm7 movl 0x14(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm6 movl 0x24(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm5 movl 0x34(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm3 movl 0x48(%rbx,%rax), %edx movq (%rcx,%rdx,8), %rdx movl 0x8(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm10 movl 0x18(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm11 movl 0x28(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm9 movl 0x38(%rbx,%rax), %esi vmovups (%rdx,%rsi,4), %xmm8 movl 0x4c(%rbx,%rax), %edx movq (%rcx,%rdx,8), %rcx movl 0xc(%rbx,%rax), %edx vmovups (%rcx,%rdx,4), %xmm12 movl 0x1c(%rbx,%rax), %edx vmovups (%rcx,%rdx,4), %xmm13 movl 0x2c(%rbx,%rax), %edx vmovups (%rcx,%rdx,4), %xmm14 movl 0x3c(%rbx,%rax), %edx vunpcklps %xmm10, %xmm4, %xmm15 # xmm15 = xmm4[0],xmm10[0],xmm4[1],xmm10[1] vunpckhps %xmm10, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm10[2],xmm4[3],xmm10[3] vunpcklps %xmm12, %xmm7, %xmm10 # xmm10 = xmm7[0],xmm12[0],xmm7[1],xmm12[1] vunpckhps %xmm12, %xmm7, %xmm7 # xmm7 = xmm7[2],xmm12[2],xmm7[3],xmm12[3] vmovups (%rcx,%rdx,4), %xmm12 vunpcklps %xmm7, %xmm4, %xmm4 # xmm4 = xmm4[0],xmm7[0],xmm4[1],xmm7[1] vunpcklps %xmm10, %xmm15, %xmm7 # xmm7 = xmm15[0],xmm10[0],xmm15[1],xmm10[1] vunpckhps %xmm10, %xmm15, %xmm10 # xmm10 = xmm15[2],xmm10[2],xmm15[3],xmm10[3] vunpcklps %xmm11, %xmm2, %xmm15 # xmm15 = xmm2[0],xmm11[0],xmm2[1],xmm11[1] vunpckhps %xmm11, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm11[2],xmm2[3],xmm11[3] vunpcklps %xmm13, %xmm6, %xmm11 # xmm11 = xmm6[0],xmm13[0],xmm6[1],xmm13[1] vunpckhps %xmm13, %xmm6, %xmm6 # xmm6 = xmm6[2],xmm13[2],xmm6[3],xmm13[3] vunpcklps %xmm6, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] vunpcklps %xmm11, %xmm15, %xmm6 # xmm6 = xmm15[0],xmm11[0],xmm15[1],xmm11[1] vunpckhps %xmm11, %xmm15, %xmm11 # xmm11 = xmm15[2],xmm11[2],xmm15[3],xmm11[3] vunpcklps %xmm9, %xmm1, %xmm13 # xmm13 = xmm1[0],xmm9[0],xmm1[1],xmm9[1] vunpckhps %xmm9, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm9[2],xmm1[3],xmm9[3] vunpcklps %xmm14, %xmm5, %xmm9 # xmm9 = xmm5[0],xmm14[0],xmm5[1],xmm14[1] vunpckhps %xmm14, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm14[2],xmm5[3],xmm14[3] vunpcklps %xmm5, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm5[0],xmm1[1],xmm5[1] vunpcklps %xmm9, %xmm13, %xmm5 # xmm5 = xmm13[0],xmm9[0],xmm13[1],xmm9[1] vunpckhps %xmm9, %xmm13, %xmm9 # xmm9 = xmm13[2],xmm9[2],xmm13[3],xmm9[3] vunpcklps %xmm8, %xmm0, %xmm13 # xmm13 = xmm0[0],xmm8[0],xmm0[1],xmm8[1] vunpckhps %xmm8, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm8[2],xmm0[3],xmm8[3] vunpcklps %xmm12, %xmm3, %xmm8 # xmm8 = xmm3[0],xmm12[0],xmm3[1],xmm12[1] vunpckhps %xmm12, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm12[2],xmm3[3],xmm12[3] vunpcklps %xmm3, %xmm0, %xmm14 # xmm14 = xmm0[0],xmm3[0],xmm0[1],xmm3[1] vunpcklps %xmm8, %xmm13, %xmm3 # xmm3 = xmm13[0],xmm8[0],xmm13[1],xmm8[1] vbroadcastf128 0x40(%rbx,%rax), %ymm12 # ymm12 = mem[0,1,0,1] vunpckhps %xmm8, %xmm13, %xmm8 # xmm8 = xmm13[2],xmm8[2],xmm13[3],xmm8[3] vmovaps %ymm12, 0x460(%rsp) vbroadcastf128 0x50(%rbx,%rax), %ymm12 # ymm12 = mem[0,1,0,1] vmovaps %ymm12, 0x440(%rsp) vinsertf128 $0x1, %xmm5, %ymm7, %ymm15 vinsertf128 $0x1, %xmm9, %ymm10, %ymm7 vinsertf128 $0x1, %xmm1, %ymm4, %ymm0 vinsertf128 $0x1, %xmm6, %ymm6, %ymm4 vinsertf128 $0x1, %xmm11, %ymm11, %ymm5 vinsertf128 $0x1, %xmm2, %ymm2, %ymm1 vinsertf128 $0x1, %xmm3, %ymm3, %ymm9 vinsertf128 $0x1, %xmm8, %ymm8, %ymm10 vinsertf128 $0x1, %xmm14, %ymm14, %ymm3 vbroadcastss (%r9), %ymm12 vbroadcastss 0x4(%r9), %ymm13 vbroadcastss 0x8(%r9), %ymm14 vsubps %ymm12, %ymm15, %ymm2 vsubps %ymm13, %ymm7, %ymm15 vsubps %ymm14, %ymm0, %ymm0 vsubps %ymm12, %ymm4, %ymm4 vmovaps %ymm4, 0x120(%rsp) vsubps %ymm13, %ymm5, %ymm11 vsubps %ymm14, %ymm1, %ymm1 vmovaps %ymm1, 0x2a0(%rsp) vsubps %ymm12, %ymm9, %ymm4 vsubps %ymm13, %ymm10, %ymm1 vmovaps %ymm1, 0x100(%rsp) vsubps %ymm14, %ymm3, %ymm14 vsubps %ymm2, %ymm4, %ymm3 vmovaps %ymm4, %ymm8 vmovaps %ymm4, 0xe0(%rsp) vsubps %ymm15, %ymm1, %ymm12 vsubps %ymm0, %ymm14, %ymm6 vaddps %ymm1, %ymm15, %ymm4 vaddps %ymm0, %ymm14, %ymm5 vmulps %ymm4, %ymm6, %ymm7 vmulps %ymm5, %ymm12, %ymm9 vsubps %ymm7, %ymm9, %ymm1 vaddps %ymm2, %ymm8, %ymm7 vmulps %ymm5, %ymm3, %ymm5 vmovaps %ymm6, 0x380(%rsp) vmulps %ymm7, %ymm6, %ymm10 vsubps %ymm5, %ymm10, %ymm5 vmovaps %ymm12, 0x3a0(%rsp) vmulps %ymm7, %ymm12, %ymm7 vmovaps %ymm3, 0x3c0(%rsp) vmulps %ymm4, %ymm3, %ymm4 vsubps %ymm7, %ymm4, %ymm4 vbroadcastss 0x14(%r9), %ymm13 vbroadcastss 0x18(%r9), %ymm9 vmulps %ymm4, %ymm9, %ymm4 vmulps %ymm5, %ymm13, %ymm5 vbroadcastss 0x10(%r9), %ymm6 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm1, %ymm6, %ymm5 vaddps %ymm4, %ymm5, %ymm10 vmovaps %ymm11, %ymm8 vsubps %ymm11, %ymm15, %ymm11 vmovaps 0x2a0(%rsp), %ymm3 vsubps %ymm3, %ymm0, %ymm12 vmovaps %ymm15, 0x400(%rsp) vaddps %ymm8, %ymm15, %ymm4 vmovaps %ymm0, 0x3e0(%rsp) vaddps %ymm3, %ymm0, %ymm5 vmovaps %ymm3, %ymm0 vmulps %ymm4, %ymm12, %ymm15 vmulps %ymm5, %ymm11, %ymm3 vsubps %ymm15, %ymm3, %ymm3 vmovaps 0x120(%rsp), %ymm1 vsubps %ymm1, %ymm2, %ymm15 vmulps %ymm5, %ymm15, %ymm5 vmovaps %ymm2, 0x420(%rsp) vaddps %ymm1, %ymm2, %ymm7 vmovaps %ymm12, 0x360(%rsp) vmulps %ymm7, %ymm12, %ymm12 vsubps %ymm5, %ymm12, %ymm5 vmovaps %ymm11, %ymm12 vmulps %ymm7, %ymm11, %ymm7 vmulps %ymm4, %ymm15, %ymm4 vsubps %ymm7, %ymm4, %ymm4 vmulps %ymm4, %ymm9, %ymm4 vmulps %ymm5, %ymm13, %ymm5 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm3, %ymm6, %ymm3 vaddps %ymm4, %ymm3, %ymm4 vmovaps 0xe0(%rsp), %ymm2 vsubps %ymm2, %ymm1, %ymm7 vaddps %ymm2, %ymm1, %ymm2 vmovaps 0x100(%rsp), %ymm1 vsubps %ymm1, %ymm8, %ymm11 vaddps %ymm1, %ymm8, %ymm1 vsubps %ymm14, %ymm0, %ymm5 vaddps %ymm0, %ymm14, %ymm3 vmovaps %ymm7, %ymm0 vmulps %ymm1, %ymm5, %ymm7 vmulps %ymm3, %ymm11, %ymm8 vsubps %ymm7, %ymm8, %ymm7 vmulps %ymm3, %ymm0, %ymm3 vmulps %ymm2, %ymm5, %ymm8 vsubps %ymm3, %ymm8, %ymm3 vmulps %ymm2, %ymm11, %ymm2 vmulps %ymm1, %ymm0, %ymm1 vsubps %ymm2, %ymm1, %ymm1 vmovaps %ymm9, 0xe0(%rsp) vmulps %ymm1, %ymm9, %ymm1 vmovaps %ymm13, 0x120(%rsp) vmulps %ymm3, %ymm13, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vmovaps %ymm6, 0x100(%rsp) vmulps %ymm7, %ymm6, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vaddps %ymm4, %ymm10, %ymm2 vaddps %ymm2, %ymm1, %ymm6 vminps %ymm4, %ymm10, %ymm2 vminps %ymm1, %ymm2, %ymm2 vbroadcastss 0x1d1433d(%rip), %ymm3 # 0x1f20ec4 vmovaps %ymm6, 0x300(%rsp) vandps %ymm3, %ymm6, %ymm6 vbroadcastss 0x1d1432f(%rip), %ymm3 # 0x1f20ecc vmovaps %ymm6, 0x320(%rsp) vmulps %ymm3, %ymm6, %ymm3 vbroadcastss 0x1d1430d(%rip), %ymm7 # 0x1f20ec0 vxorps %ymm7, %ymm3, %ymm7 vcmpnltps %ymm7, %ymm2, %ymm2 vmovaps %ymm10, 0x2a0(%rsp) vmovaps %ymm4, 0x340(%rsp) vmaxps %ymm4, %ymm10, %ymm7 vmaxps %ymm1, %ymm7, %ymm1 vcmpleps %ymm3, %ymm1, %ymm1 vorps %ymm1, %ymm2, %ymm1 vtestps %ymm1, %ymm1 je 0x20d233 vmovaps 0x380(%rsp), %ymm6 vmovaps %ymm11, %ymm8 vmovaps %ymm12, %ymm11 vmulps %ymm6, %ymm12, %ymm2 vmovaps 0x3a0(%rsp), %ymm13 vmovaps 0x360(%rsp), %ymm9 vmulps %ymm9, %ymm13, %ymm3 vsubps %ymm2, %ymm3, %ymm3 vmulps %ymm8, %ymm9, %ymm7 vmulps %ymm5, %ymm12, %ymm12 vsubps %ymm7, %ymm12, %ymm12 vbroadcastss 0x1d14292(%rip), %ymm4 # 0x1f20ec4 vandps %ymm4, %ymm2, %ymm2 vandps %ymm4, %ymm7, %ymm7 vcmpltps %ymm7, %ymm2, %ymm2 vblendvps %ymm2, %ymm3, %ymm12, %ymm14 vmulps %ymm5, %ymm15, %ymm2 vmulps %ymm6, %ymm15, %ymm3 vmovaps 0x3c0(%rsp), %ymm10 vmulps %ymm9, %ymm10, %ymm6 vsubps %ymm6, %ymm3, %ymm3 vmulps %ymm0, %ymm9, %ymm7 vsubps %ymm2, %ymm7, %ymm7 vandps %ymm4, %ymm6, %ymm6 vandps %ymm4, %ymm2, %ymm2 vcmpltps %ymm2, %ymm6, %ymm2 vblendvps %ymm2, %ymm3, %ymm7, %ymm9 vmulps %ymm0, %ymm11, %ymm2 vmulps %ymm11, %ymm10, %ymm3 vmulps %ymm15, %ymm13, %ymm5 vmulps %ymm8, %ymm15, %ymm0 vsubps %ymm5, %ymm3, %ymm3 vsubps %ymm2, %ymm0, %ymm0 vandps %ymm4, %ymm5, %ymm5 vandps %ymm4, %ymm2, %ymm2 vcmpltps %ymm2, %ymm5, %ymm2 vblendvps %ymm2, %ymm3, %ymm0, %ymm0 vextractf128 $0x1, %ymm1, %xmm2 vpackssdw %xmm2, %xmm1, %xmm1 vmulps 0xe0(%rsp), %ymm0, %ymm2 vmulps 0x120(%rsp), %ymm9, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vmulps 0x100(%rsp), %ymm14, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vaddps %ymm2, %ymm2, %ymm3 vmulps 0x3e0(%rsp), %ymm0, %ymm2 vmulps 0x400(%rsp), %ymm9, %ymm5 vaddps %ymm2, %ymm5, %ymm2 vmulps 0x420(%rsp), %ymm14, %ymm5 vrcpps %ymm3, %ymm6 vaddps %ymm2, %ymm5, %ymm2 vaddps %ymm2, %ymm2, %ymm2 vmulps %ymm6, %ymm3, %ymm5 vbroadcastss 0x1cdfa03(%rip), %ymm7 # 0x1eec714 vsubps %ymm5, %ymm7, %ymm5 vmulps %ymm5, %ymm6, %ymm5 vaddps %ymm5, %ymm6, %ymm5 vbroadcastss 0xc(%r9), %ymm6 vmulps %ymm5, %ymm2, %ymm2 vcmpleps %ymm2, %ymm6, %ymm5 vbroadcastss 0x20(%r9), %ymm6 vcmpleps %ymm6, %ymm2, %ymm6 vandps %ymm5, %ymm6, %ymm5 vcmpneqps 0x1d141bc(%rip), %ymm3, %ymm3 # 0x1f20f00 vandps %ymm3, %ymm5, %ymm3 vextractf128 $0x1, %ymm3, %xmm5 vpackssdw %xmm5, %xmm3, %xmm3 vpand %xmm1, %xmm3, %xmm1 vpmovsxwd %xmm1, %xmm3 vpshufd $0xee, %xmm1, %xmm5 # xmm5 = xmm1[2,3,2,3] vpmovsxwd %xmm5, %xmm5 vinsertf128 $0x1, %xmm5, %ymm3, %ymm3 vtestps %ymm3, %ymm3 je 0x20d233 vmovaps 0x2a0(%rsp), %ymm4 vmovaps %ymm4, 0x480(%rsp) vmovaps 0x340(%rsp), %ymm7 vmovaps %ymm7, 0x4a0(%rsp) vmovaps 0x300(%rsp), %ymm5 vmovaps %ymm5, 0x4c0(%rsp) vmovaps %ymm14, 0x4e0(%rsp) vmovaps %ymm9, 0x500(%rsp) vmovaps %ymm0, 0x520(%rsp) vmovaps %ymm3, 0x540(%rsp) vmovaps %ymm2, 0x5a0(%rsp) vmovaps 0x2e0(%rsp), %ymm6 vmovaps %ymm6, 0x620(%rsp) vrcpps %ymm5, %ymm2 vmulps %ymm2, %ymm5, %ymm3 vbroadcastss 0x1cdf918(%rip), %ymm5 # 0x1eec714 vsubps %ymm3, %ymm5, %ymm3 vmulps %ymm3, %ymm2, %ymm3 vaddps %ymm3, %ymm2, %ymm2 vbroadcastss 0x1ce41d7(%rip), %ymm3 # 0x1ef0fe8 vmovaps 0x320(%rsp), %ymm8 vcmpnltps %ymm3, %ymm8, %ymm3 vandps %ymm2, %ymm3, %ymm2 vmulps %ymm2, %ymm4, %ymm3 vminps %ymm5, %ymm3, %ymm3 vmulps %ymm2, %ymm7, %ymm2 vminps %ymm5, %ymm2, %ymm2 vsubps %ymm3, %ymm5, %ymm4 vsubps %ymm2, %ymm5, %ymm5 vblendvps %ymm6, %ymm4, %ymm2, %ymm2 vblendvps %ymm6, %ymm5, %ymm3, %ymm3 vmovaps %ymm2, 0x580(%rsp) vmovaps %ymm3, 0x560(%rsp) vmovaps 0x2c0(%rsp), %ymm4 vmulps %ymm4, %ymm14, %ymm2 vmulps %ymm4, %ymm9, %ymm3 vmulps %ymm0, %ymm4, %ymm0 vmovaps %ymm2, 0x5c0(%rsp) vmovaps %ymm3, 0x5e0(%rsp) vmovaps %ymm0, 0x600(%rsp) vpsllw $0xf, %xmm1, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r15d vmovaps 0x280(%rsp), %ymm7 vmovaps 0x260(%rsp), %ymm8 vmovaps 0x240(%rsp), %ymm9 vmovaps 0x220(%rsp), %ymm10 vmovaps 0x200(%rsp), %ymm11 vmovaps 0x1e0(%rsp), %ymm12 vmovaps 0x1c0(%rsp), %ymm13 vmovaps 0x1a0(%rsp), %ymm14 vmovaps 0x180(%rsp), %ymm5 movq 0x28(%rsp), %rsi movq 0x20(%rsp), %rdi vmovaps 0x160(%rsp), %ymm6 vmovaps 0x140(%rsp), %ymm15 movq 0x38(%rsp), %rdx bsfq %r15, %r12 movl 0x460(%rsp,%r12,4), %eax movq 0x1e8(%rdx), %rcx movq (%rcx,%rax,8), %r13 movl 0x24(%r9), %ecx testl %ecx, 0x34(%r13) je 0x20cf43 movq 0x10(%r8), %r14 cmpq $0x0, 0x10(%r14) jne 0x20cf63 cmpq $0x0, 0x48(%r13) jne 0x20cf63 xorl %eax, %eax jmp 0x20cf49 btcq %r12, %r15 movb $0x1, %al movq 0x30(%rsp), %r12 xorl %r13d, %r13d testb %al, %al je 0x20d2aa testq %r15, %r15 jne 0x20cf0c jmp 0x20d209 vmovss 0x560(%rsp,%r12,4), %xmm0 vmovd 0x580(%rsp,%r12,4), %xmm1 movq 0x8(%r8), %rcx movl 0x440(%rsp,%r12,4), %edx vmovss 0x5c0(%rsp,%r12,4), %xmm2 vmovss 0x5e0(%rsp,%r12,4), %xmm3 vmovss 0x600(%rsp,%r12,4), %xmm4 vmovss %xmm2, 0x80(%rsp) vmovss %xmm3, 0x84(%rsp) vmovss %xmm4, 0x88(%rsp) vmovss %xmm0, 0x8c(%rsp) vmovd %xmm1, 0x90(%rsp) movl %edx, 0x94(%rsp) movl %eax, 0x98(%rsp) movl (%rcx), %eax movl %eax, 0x9c(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xa0(%rsp) vmovss 0x20(%r9), %xmm0 vmovss %xmm0, 0x120(%rsp) vmovd 0x5a0(%rsp,%r12,4), %xmm0 vmovd %xmm0, 0x20(%r9) orl $-0x1, 0x44(%rsp) leaq 0x44(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%r13), %rax movq %rax, 0xb8(%rsp) movq %rcx, 0xc0(%rsp) movq %r9, 0xc8(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x1, 0xd8(%rsp) movq 0x48(%r13), %rax testq %rax, %rax movq %r10, 0x50(%rsp) movq %r11, 0x48(%rsp) je 0x20d123 leaq 0xb0(%rsp), %rdi movq %r8, 0x100(%rsp) movq %r9, 0xe0(%rsp) vzeroupper callq %rax vmovaps 0x140(%rsp), %ymm15 vmovaps 0x160(%rsp), %ymm6 movq 0x20(%rsp), %rdi movq 0x28(%rsp), %rsi vmovaps 0x180(%rsp), %ymm5 vmovaps 0x1a0(%rsp), %ymm14 vmovaps 0x1c0(%rsp), %ymm13 vmovaps 0x1e0(%rsp), %ymm12 vmovaps 0x200(%rsp), %ymm11 vmovaps 0x220(%rsp), %ymm10 movq 0x48(%rsp), %r11 vmovaps 0x240(%rsp), %ymm9 vmovaps 0x260(%rsp), %ymm8 vmovaps 0x280(%rsp), %ymm7 movq 0x50(%rsp), %r10 movq 0xe0(%rsp), %r9 movq 0x100(%rsp), %r8 movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x20d1e2 movq 0x10(%r14), %rax testq %rax, %rax je 0x20d1de testb $0x2, (%r14) jne 0x20d141 testb $0x40, 0x3e(%r13) je 0x20d1d1 leaq 0xb0(%rsp), %rdi movq %r8, %r14 movq %r9, %r13 vzeroupper callq %rax vmovaps 0x140(%rsp), %ymm15 vmovaps 0x160(%rsp), %ymm6 movq 0x20(%rsp), %rdi movq 0x28(%rsp), %rsi vmovaps 0x180(%rsp), %ymm5 vmovaps 0x1a0(%rsp), %ymm14 vmovaps 0x1c0(%rsp), %ymm13 vmovaps 0x1e0(%rsp), %ymm12 vmovaps 0x200(%rsp), %ymm11 vmovaps 0x220(%rsp), %ymm10 movq 0x48(%rsp), %r11 vmovaps 0x240(%rsp), %ymm9 vmovaps 0x260(%rsp), %ymm8 vmovaps 0x280(%rsp), %ymm7 movq 0x50(%rsp), %r10 movq %r13, %r9 movq %r14, %r8 movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x20d1e2 xorl %eax, %eax jmp 0x20d1f7 vmovd 0x120(%rsp), %xmm0 vmovd %xmm0, 0x20(%r9) btcq %r12, %r15 movb $0x1, %al movq 0x30(%rsp), %r12 xorl %r13d, %r13d movq 0x38(%rsp), %rdx jmp 0x20cf51 movq 0x60(%rsp), %r14 movq 0x58(%rsp), %r15 movq 0x70(%rsp), %rcx incq %rcx movq 0x68(%rsp), %rax cmpq %rax, %rcx setb 0x1f(%rsp) jne 0x20c7c8 jmp 0x20d2c7 vmovaps 0x280(%rsp), %ymm7 vmovaps 0x260(%rsp), %ymm8 vmovaps 0x240(%rsp), %ymm9 vmovaps 0x220(%rsp), %ymm10 vmovaps 0x200(%rsp), %ymm11 vmovaps 0x1e0(%rsp), %ymm12 vmovaps 0x1c0(%rsp), %ymm13 vmovaps 0x1a0(%rsp), %ymm14 vmovaps 0x180(%rsp), %ymm5 movq 0x28(%rsp), %rsi movq 0x20(%rsp), %rdi vmovaps 0x160(%rsp), %ymm6 vmovaps 0x140(%rsp), %ymm15 movq 0x38(%rsp), %rdx jmp 0x20d213 testb $0x1, 0x1f(%rsp) movq 0x60(%rsp), %r14 movq 0x58(%rsp), %r15 je 0x20d2c7 movl $0xff800000, 0x20(%r9) # imm = 0xFF800000 pushq $0x3 popq %r13 cmpl $0x3, %r13d jne 0x20c656 leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, false, embree::avx::ArrayIntersector1<embree::avx::QuadMiMBIntersector1Moeller<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x1760, %rsp # imm = 0x1760 movq %rdx, 0x20(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x20d31d leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq vmovss 0x20(%rsi), %xmm0 vxorps %xmm5, %xmm5, %xmm5 vucomiss %xmm0, %xmm5 ja 0x20d30b leaq 0x5a8(%rsp), %rdi movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmovaps 0x10(%rsi), %xmm2 vmaxss 0xc(%rsi), %xmm5, %xmm1 vbroadcastss 0x1d13b75(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1ce3c8c(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vblendvps %xmm3, %xmm4, %xmm2, %xmm2 vrcpps %xmm2, %xmm3 vmulps %xmm2, %xmm3, %xmm2 vbroadcastss 0x1cdf39c(%rip), %xmm4 # 0x1eec714 vsubps %xmm2, %xmm4, %xmm2 vmulps %xmm2, %xmm3, %xmm2 vaddps %xmm2, %xmm3, %xmm2 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vbroadcastss 0x8(%rsi), %ymm8 xorl %r8d, %r8d vucomiss %xmm5, %xmm2 setb %r8b vshufps $0x0, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm9 vmovshdup %xmm2, %xmm3 # xmm3 = xmm2[1,1,3,3] vshufps $0x55, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm10 vshufpd $0x1, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[1,0] vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vinsertf128 $0x1, %xmm2, %ymm2, %ymm11 shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm5, %xmm3 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm5, %xmm4 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %rbx xorq $0x20, %rbx movq %r10, %r15 xorq $0x20, %r15 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm12 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm13 leaq 0x1f42b57(%rip), %rax # 0x214ff80 vbroadcastf128 0xf0(%rax), %ymm0 # ymm0 = mem[0,1,0,1] vmovaps %ymm0, 0x380(%rsp) vperm2f128 $0x2, (%rax), %ymm0, %ymm1 # ymm1 = mem[0,1],ymm0[0,1] leaq 0x5a0(%rsp), %r12 vbroadcastss 0x1ce357a(%rip), %ymm0 # 0x1ef09cc vbroadcastss 0x1cdf2b9(%rip), %ymm2 # 0x1eec714 vmovaps %ymm1, 0x3c0(%rsp) vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x3a0(%rsp) movq %rsi, 0x28(%rsp) vmovaps %ymm6, 0x360(%rsp) vmovaps %ymm7, 0x340(%rsp) vmovaps %ymm8, 0x320(%rsp) movq %r8, 0x60(%rsp) vmovaps %ymm9, 0x300(%rsp) vmovaps %ymm10, 0x2e0(%rsp) vmovaps %ymm11, 0x2c0(%rsp) movq %r9, 0x58(%rsp) movq %r10, 0x50(%rsp) movq %r11, 0x48(%rsp) movq %rbx, 0x40(%rsp) movq %r15, 0x38(%rsp) vmovaps %ymm12, 0x2a0(%rsp) vmovaps %ymm13, 0x280(%rsp) cmpq %r12, %rdi je 0x20d30b movq -0x8(%rdi), %r14 addq $-0x8, %rdi testb $0x8, %r14b jne 0x20d5df movq %r14, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rsi), %ymm0 vmulps 0x100(%rax,%r8), %ymm0, %ymm1 vaddps 0x40(%rax,%r8), %ymm1, %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmulps 0x100(%rax,%r9), %ymm0, %ymm2 vaddps 0x40(%rax,%r9), %ymm2, %ymm2 vmaxps %ymm1, %ymm12, %ymm1 vsubps %ymm7, %ymm2, %ymm2 vmulps 0x100(%rax,%r10), %ymm0, %ymm3 vmulps %ymm2, %ymm10, %ymm2 vaddps 0x40(%rax,%r10), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm11, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r11), %ymm3, %ymm2 vsubps %ymm6, %ymm2, %ymm2 vmulps %ymm2, %ymm9, %ymm2 vminps %ymm2, %ymm13, %ymm2 vmulps 0x100(%rax,%rbx), %ymm0, %ymm3 vaddps 0x40(%rax,%rbx), %ymm3, %ymm3 vsubps %ymm7, %ymm3, %ymm3 vmulps %ymm3, %ymm10, %ymm3 vmulps 0x100(%rax,%r15), %ymm0, %ymm4 vaddps 0x40(%rax,%r15), %ymm4, %ymm4 vsubps %ymm8, %ymm4, %ymm4 vmulps %ymm4, %ymm11, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r14d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x20d627 vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r13d testb $0x8, %r14b jne 0x20d623 testq %r13, %r13 je 0x20d654 andq $-0x10, %r14 bsfq %r13, %rax leaq -0x1(%r13), %rdx xorl %ecx, %ecx movq (%r14,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r13, %rdx jne 0x20d659 movq %rax, %r14 testl %ecx, %ecx je 0x20d4ef jmp 0x20d6a1 pushq $0x6 jmp 0x20d656 vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x20d5ce pushq $0x4 popq %rcx jmp 0x20d619 movq %rax, (%rdi) addq $0x8, %rdi bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r14,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x20d697 movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x20d668 movq %rcx, %r14 xorl %ecx, %ecx jmp 0x20d619 cmpl $0x6, %ecx jne 0x20e142 movl %r14d, %eax andl $0xf, %eax xorl %ecx, %ecx addq $-0x8, %rax movq %rax, 0xa8(%rsp) setne %dl je 0x20e142 movq %r13, 0x68(%rsp) movq %rdi, 0x70(%rsp) andq $-0x10, %r14 movq 0x20(%rsp), %rax movq (%rax), %rax movq %rax, 0x78(%rsp) xorl %eax, %eax movb %dl, 0x1f(%rsp) movq %rax, 0xd8(%rsp) imulq $0x60, %rax, %rax vmovss 0x1c(%rsi), %xmm0 movl 0x40(%r14,%rax), %ecx movq 0x78(%rsp), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rcx,8), %rsi vmovss 0x28(%rsi), %xmm1 vmovss 0x2c(%rsi), %xmm2 vmovss 0x30(%rsi), %xmm3 vsubss %xmm2, %xmm0, %xmm0 vsubss %xmm2, %xmm3, %xmm2 vdivss %xmm2, %xmm0, %xmm0 vmulss %xmm0, %xmm1, %xmm0 vroundss $0x9, %xmm0, %xmm0, %xmm2 vaddss 0x1ce3291(%rip), %xmm1, %xmm1 # 0x1ef09cc vminss %xmm1, %xmm2, %xmm1 vmaxss %xmm1, %xmm5, %xmm3 vcvttss2si %xmm3, %ecx movslq %ecx, %rcx movq 0xe0(%rsi), %r15 imulq $0x38, %rcx, %r9 movq (%r15,%r9), %rsi movl (%r14,%rax), %ebx movl 0x4(%r14,%rax), %r8d movq %r8, 0xc0(%rsp) vmovups (%rsi,%rbx,4), %xmm2 movl 0x10(%r14,%rax), %edi movq %rdi, 0x160(%rsp) vmovups (%rsi,%rdi,4), %xmm1 movl 0x20(%r14,%rax), %edi movq %rdi, 0x140(%rsp) vmovups (%rsi,%rdi,4), %xmm7 movl 0x30(%r14,%rax), %edi movq %rdi, 0x80(%rsp) vmovups (%rsi,%rdi,4), %xmm8 movl 0x44(%r14,%rax), %esi movq (%rdx,%rsi,8), %rsi movq 0xe0(%rsi), %rsi movq %rsi, 0xb0(%rsp) movq (%rsi,%r9), %rsi vmovups (%rsi,%r8,4), %xmm4 movl 0x14(%r14,%rax), %edi movq %rdi, 0xd0(%rsp) vmovups (%rsi,%rdi,4), %xmm5 movl 0x24(%r14,%rax), %edi movq %rdi, 0x180(%rsp) vmovups (%rsi,%rdi,4), %xmm9 movl 0x34(%r14,%rax), %edi movq %rdi, 0x1a0(%rsp) vmovups (%rsi,%rdi,4), %xmm10 movl 0x48(%r14,%rax), %esi movq (%rdx,%rsi,8), %rsi movq 0xe0(%rsi), %r8 movq (%r8,%r9), %rsi movl 0x8(%r14,%rax), %edi vmovups (%rsi,%rdi,4), %xmm13 movl 0x18(%r14,%rax), %r13d vmovups (%rsi,%r13,4), %xmm6 movl 0x28(%r14,%rax), %r10d movq %r10, 0xb8(%rsp) vmovups (%rsi,%r10,4), %xmm11 movl 0x38(%r14,%rax), %r10d movq %r10, 0xc8(%rsp) vmovups (%rsi,%r10,4), %xmm12 movl 0x4c(%r14,%rax), %esi movq (%rdx,%rsi,8), %rdx vsubss %xmm3, %xmm0, %xmm0 vmovaps %xmm0, 0x270(%rsp) movq 0xe0(%rdx), %rsi movq (%rsi,%r9), %r12 movl 0xc(%r14,%rax), %edx vmovups (%r12,%rdx,4), %xmm0 movl 0x1c(%r14,%rax), %r11d vmovups (%r12,%r11,4), %xmm3 vunpcklps %xmm13, %xmm2, %xmm14 # xmm14 = xmm2[0],xmm13[0],xmm2[1],xmm13[1] vunpckhps %xmm13, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm13[2],xmm2[3],xmm13[3] vunpcklps %xmm0, %xmm4, %xmm13 # xmm13 = xmm4[0],xmm0[0],xmm4[1],xmm0[1] vunpckhps %xmm0, %xmm4, %xmm0 # xmm0 = xmm4[2],xmm0[2],xmm4[3],xmm0[3] movl 0x2c(%r14,%rax), %r9d vmovups (%r12,%r9,4), %xmm4 vunpcklps %xmm0, %xmm2, %xmm0 # xmm0 = xmm2[0],xmm0[0],xmm2[1],xmm0[1] vmovaps %xmm0, 0x250(%rsp) vunpcklps %xmm13, %xmm14, %xmm0 # xmm0 = xmm14[0],xmm13[0],xmm14[1],xmm13[1] vmovaps %xmm0, 0x210(%rsp) vunpckhps %xmm13, %xmm14, %xmm0 # xmm0 = xmm14[2],xmm13[2],xmm14[3],xmm13[3] vmovaps %xmm0, 0x230(%rsp) vunpcklps %xmm6, %xmm1, %xmm0 # xmm0 = xmm1[0],xmm6[0],xmm1[1],xmm6[1] vunpckhps %xmm6, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm6[2],xmm1[3],xmm6[3] vunpcklps %xmm3, %xmm5, %xmm2 # xmm2 = xmm5[0],xmm3[0],xmm5[1],xmm3[1] vunpckhps %xmm3, %xmm5, %xmm3 # xmm3 = xmm5[2],xmm3[2],xmm5[3],xmm3[3] movl 0x3c(%r14,%rax), %r10d vmovups (%r12,%r10,4), %xmm5 vunpcklps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vmovaps %xmm1, 0x260(%rsp) vunpcklps %xmm2, %xmm0, %xmm1 # xmm1 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vmovaps %xmm1, 0x220(%rsp) vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x240(%rsp) vunpcklps %xmm11, %xmm7, %xmm0 # xmm0 = xmm7[0],xmm11[0],xmm7[1],xmm11[1] vunpckhps %xmm11, %xmm7, %xmm1 # xmm1 = xmm7[2],xmm11[2],xmm7[3],xmm11[3] vunpcklps %xmm4, %xmm9, %xmm2 # xmm2 = xmm9[0],xmm4[0],xmm9[1],xmm4[1] vunpckhps %xmm4, %xmm9, %xmm3 # xmm3 = xmm9[2],xmm4[2],xmm9[3],xmm4[3] incl %ecx movslq %ecx, %rcx imulq $0x38, %rcx, %r12 movq (%r15,%r12), %rcx vmovups (%rcx,%rbx,4), %xmm4 vunpcklps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vmovaps %xmm1, 0x200(%rsp) vunpcklps %xmm2, %xmm0, %xmm1 # xmm1 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vmovaps %xmm1, 0x1c0(%rsp) vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x1e0(%rsp) vunpcklps %xmm12, %xmm8, %xmm0 # xmm0 = xmm8[0],xmm12[0],xmm8[1],xmm12[1] vunpckhps %xmm12, %xmm8, %xmm1 # xmm1 = xmm8[2],xmm12[2],xmm8[3],xmm12[3] vunpcklps %xmm5, %xmm10, %xmm2 # xmm2 = xmm10[0],xmm5[0],xmm10[1],xmm5[1] vunpckhps %xmm5, %xmm10, %xmm3 # xmm3 = xmm10[2],xmm5[2],xmm10[3],xmm5[3] movq (%r8,%r12), %r8 vmovups (%r8,%rdi,4), %xmm5 vunpcklps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vmovaps %xmm1, 0x1f0(%rsp) vunpcklps %xmm2, %xmm0, %xmm12 # xmm12 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x1d0(%rsp) movq 0xb0(%rsp), %rdi movq (%rdi,%r12), %rdi vunpcklps %xmm5, %xmm4, %xmm0 # xmm0 = xmm4[0],xmm5[0],xmm4[1],xmm5[1] vunpckhps %xmm5, %xmm4, %xmm1 # xmm1 = xmm4[2],xmm5[2],xmm4[3],xmm5[3] movq 0xc0(%rsp), %rbx vmovups (%rdi,%rbx,4), %xmm2 movq (%rsi,%r12), %rsi vmovups (%rsi,%rdx,4), %xmm3 vunpcklps %xmm3, %xmm2, %xmm4 # xmm4 = xmm2[0],xmm3[0],xmm2[1],xmm3[1] vunpckhps %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm3[2],xmm2[3],xmm3[3] vunpcklps %xmm2, %xmm1, %xmm13 # xmm13 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] vunpcklps %xmm4, %xmm0, %xmm10 # xmm10 = xmm0[0],xmm4[0],xmm0[1],xmm4[1] vunpckhps %xmm4, %xmm0, %xmm8 # xmm8 = xmm0[2],xmm4[2],xmm0[3],xmm4[3] movq 0x160(%rsp), %rdx vmovups (%rcx,%rdx,4), %xmm0 vmovups (%r8,%r13,4), %xmm1 vunpcklps %xmm1, %xmm0, %xmm3 # xmm3 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpckhps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3] movq 0xd0(%rsp), %rdx vmovups (%rdi,%rdx,4), %xmm1 vmovups (%rsi,%r11,4), %xmm2 vunpcklps %xmm2, %xmm1, %xmm4 # xmm4 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] vunpckhps %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm2[2],xmm1[3],xmm2[3] vunpcklps %xmm1, %xmm0, %xmm14 # xmm14 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm4, %xmm3, %xmm2 # xmm2 = xmm3[0],xmm4[0],xmm3[1],xmm4[1] vunpckhps %xmm4, %xmm3, %xmm7 # xmm7 = xmm3[2],xmm4[2],xmm3[3],xmm4[3] movq 0x140(%rsp), %rdx vmovups (%rcx,%rdx,4), %xmm0 movq 0xb8(%rsp), %rdx vmovups (%r8,%rdx,4), %xmm1 vunpcklps %xmm1, %xmm0, %xmm4 # xmm4 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpckhps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3] movq 0x180(%rsp), %rdx vmovups (%rdi,%rdx,4), %xmm1 vmovups (%rsi,%r9,4), %xmm3 vunpcklps %xmm3, %xmm1, %xmm5 # xmm5 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vunpckhps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm3[2],xmm1[3],xmm3[3] vunpcklps %xmm1, %xmm0, %xmm15 # xmm15 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm5, %xmm4, %xmm3 # xmm3 = xmm4[0],xmm5[0],xmm4[1],xmm5[1] vunpckhps %xmm5, %xmm4, %xmm9 # xmm9 = xmm4[2],xmm5[2],xmm4[3],xmm5[3] movq 0x80(%rsp), %rdx vmovups (%rcx,%rdx,4), %xmm0 movq 0xc8(%rsp), %rcx vmovups (%r8,%rcx,4), %xmm1 vunpcklps %xmm1, %xmm0, %xmm5 # xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpckhps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3] movq 0x1a0(%rsp), %rcx vmovups (%rdi,%rcx,4), %xmm1 vmovups (%rsi,%r10,4), %xmm4 movq 0x28(%rsp), %rsi vunpcklps %xmm4, %xmm1, %xmm11 # xmm11 = xmm1[0],xmm4[0],xmm1[1],xmm4[1] vunpckhps %xmm4, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm4[2],xmm1[3],xmm4[3] vunpcklps %xmm1, %xmm0, %xmm6 # xmm6 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm11, %xmm5, %xmm4 # xmm4 = xmm5[0],xmm11[0],xmm5[1],xmm11[1] vunpckhps %xmm11, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm11[2],xmm5[3],xmm11[3] vmovaps 0x270(%rsp), %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm0 # xmm0 = xmm11[0,0,0,0] vmovss 0x1cdec52(%rip), %xmm1 # 0x1eec714 vsubss %xmm11, %xmm1, %xmm1 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmulps %xmm0, %xmm10, %xmm10 vmulps 0x210(%rsp), %xmm1, %xmm11 vaddps %xmm10, %xmm11, %xmm11 vmulps %xmm0, %xmm8, %xmm8 vmulps 0x230(%rsp), %xmm1, %xmm10 vaddps %xmm8, %xmm10, %xmm10 vmulps %xmm0, %xmm13, %xmm8 vmulps 0x250(%rsp), %xmm1, %xmm13 vaddps %xmm8, %xmm13, %xmm13 vmulps %xmm2, %xmm0, %xmm2 vmulps 0x220(%rsp), %xmm1, %xmm8 vaddps %xmm2, %xmm8, %xmm8 vmulps %xmm7, %xmm0, %xmm2 vmulps 0x240(%rsp), %xmm1, %xmm7 vaddps %xmm2, %xmm7, %xmm2 vmulps %xmm0, %xmm14, %xmm7 vmulps 0x260(%rsp), %xmm1, %xmm14 vaddps %xmm7, %xmm14, %xmm14 vmulps %xmm3, %xmm0, %xmm3 vmulps 0x1c0(%rsp), %xmm1, %xmm7 vaddps %xmm3, %xmm7, %xmm3 vmulps %xmm0, %xmm9, %xmm7 vmulps 0x1e0(%rsp), %xmm1, %xmm9 vaddps %xmm7, %xmm9, %xmm7 vmulps %xmm0, %xmm15, %xmm9 vmulps 0x200(%rsp), %xmm1, %xmm15 vaddps %xmm9, %xmm15, %xmm9 vmulps %xmm4, %xmm0, %xmm4 vmulps %xmm5, %xmm0, %xmm5 vmulps %xmm6, %xmm0, %xmm0 vmulps %xmm1, %xmm12, %xmm6 vaddps %xmm4, %xmm6, %xmm15 vbroadcastf128 0x40(%r14,%rax), %ymm6 # ymm6 = mem[0,1,0,1] vmovaps %ymm6, 0x400(%rsp) vbroadcastf128 0x50(%r14,%rax), %ymm6 # ymm6 = mem[0,1,0,1] movq 0xd8(%rsp), %rax vmulps 0x1d0(%rsp), %xmm1, %xmm12 vaddps %xmm5, %xmm12, %xmm12 vmulps 0x1f0(%rsp), %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmovaps %ymm6, 0x3e0(%rsp) vinsertf128 $0x1, %xmm3, %ymm11, %ymm4 vinsertf128 $0x1, %xmm7, %ymm10, %ymm5 vinsertf128 $0x1, %xmm9, %ymm13, %ymm7 vinsertf128 $0x1, %xmm8, %ymm8, %ymm1 vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vinsertf128 $0x1, %xmm14, %ymm14, %ymm8 vinsertf128 $0x1, %xmm15, %ymm15, %ymm9 vinsertf128 $0x1, %xmm12, %ymm12, %ymm10 vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vsubps %ymm1, %ymm4, %ymm3 vsubps %ymm2, %ymm5, %ymm12 vmovaps %ymm12, 0x140(%rsp) vsubps %ymm8, %ymm7, %ymm8 vsubps %ymm4, %ymm9, %ymm9 vsubps %ymm5, %ymm10, %ymm10 vsubps %ymm7, %ymm0, %ymm11 vmulps %ymm11, %ymm12, %ymm0 vmulps %ymm10, %ymm8, %ymm1 vsubps %ymm0, %ymm1, %ymm0 vmovaps %ymm0, 0x1a0(%rsp) vmulps %ymm9, %ymm8, %ymm0 vmovaps %ymm3, 0x80(%rsp) vmulps %ymm3, %ymm11, %ymm2 vsubps %ymm0, %ymm2, %ymm6 vmulps %ymm3, %ymm10, %ymm0 vmulps %ymm9, %ymm12, %ymm12 vsubps %ymm0, %ymm12, %ymm3 vbroadcastss (%rsi), %ymm12 vbroadcastss 0x4(%rsi), %ymm13 vbroadcastss 0x8(%rsi), %ymm14 vbroadcastss 0x14(%rsi), %ymm15 vsubps %ymm12, %ymm4, %ymm2 vbroadcastss 0x18(%rsi), %ymm12 vsubps %ymm13, %ymm5, %ymm5 vsubps %ymm14, %ymm7, %ymm7 vmulps %ymm7, %ymm15, %ymm13 vmulps %ymm5, %ymm12, %ymm14 vsubps %ymm13, %ymm14, %ymm13 vbroadcastss 0x10(%rsi), %ymm14 vmulps %ymm2, %ymm12, %ymm0 vmulps %ymm7, %ymm14, %ymm1 vsubps %ymm0, %ymm1, %ymm0 vmulps %ymm5, %ymm14, %ymm1 vmovaps %ymm2, 0x180(%rsp) vmulps %ymm2, %ymm15, %ymm2 vsubps %ymm1, %ymm2, %ymm1 vmovaps %ymm3, 0x160(%rsp) vmulps %ymm3, %ymm12, %ymm2 vmulps %ymm6, %ymm15, %ymm12 vmovaps %ymm6, %ymm15 vaddps %ymm2, %ymm12, %ymm2 vmovaps 0x1a0(%rsp), %ymm3 vmulps %ymm3, %ymm14, %ymm12 vaddps %ymm2, %ymm12, %ymm2 vmulps %ymm1, %ymm11, %ymm11 vmulps %ymm0, %ymm10, %ymm10 vaddps %ymm11, %ymm10, %ymm10 vmulps %ymm13, %ymm9, %ymm9 vaddps %ymm10, %ymm9, %ymm10 vmulps %ymm1, %ymm8, %ymm1 vmulps 0x140(%rsp), %ymm0, %ymm0 vaddps %ymm1, %ymm0, %ymm0 vbroadcastss 0x1d131b4(%rip), %ymm1 # 0x1f20ec0 vandps %ymm1, %ymm2, %ymm9 vxorps %ymm10, %ymm9, %ymm6 vmulps 0x80(%rsp), %ymm13, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vxorps %ymm0, %ymm9, %ymm8 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %ymm10, %ymm6, %ymm0 vcmpnltps %ymm10, %ymm8, %ymm1 vandps %ymm1, %ymm0, %ymm0 vbroadcastss 0x1d13180(%rip), %ymm1 # 0x1f20ec4 vandps %ymm1, %ymm2, %ymm4 vcmpneqps %ymm2, %ymm10, %ymm1 vandps %ymm1, %ymm0, %ymm0 vaddps %ymm6, %ymm8, %ymm1 vcmpleps %ymm4, %ymm1, %ymm1 vandps %ymm1, %ymm0, %ymm10 vtestps 0x380(%rsp), %ymm10 jne 0x20dd87 incq %rax cmpq 0xa8(%rsp), %rax setb %dl vxorps %xmm5, %xmm5, %xmm5 jne 0x20d6e4 jmp 0x20e0c8 vmovaps %ymm3, %ymm14 vandps 0x380(%rsp), %ymm10, %ymm10 vmulps 0x160(%rsp), %ymm7, %ymm0 vmulps %ymm5, %ymm15, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vmulps 0x180(%rsp), %ymm3, %ymm1 vaddps %ymm0, %ymm1, %ymm0 vxorps %ymm0, %ymm9, %ymm3 vbroadcastss 0xc(%rsi), %ymm0 vmulps %ymm0, %ymm4, %ymm0 vcmpltps %ymm3, %ymm0, %ymm0 vbroadcastss 0x20(%rsi), %ymm1 vmulps %ymm1, %ymm4, %ymm1 vcmpleps %ymm1, %ymm3, %ymm1 vandps %ymm0, %ymm1, %ymm5 vtestps %ymm10, %ymm5 je 0x20dd6a vandps %ymm5, %ymm10, %ymm0 vmovaps %ymm6, 0x420(%rsp) vmovaps %ymm8, 0x440(%rsp) vmovaps %ymm3, 0x460(%rsp) vmovaps %ymm4, 0x480(%rsp) vmovaps %ymm0, 0x4c0(%rsp) vmovaps 0x420(%rsp), %ymm0 vmovaps 0x440(%rsp), %ymm1 vsubps %ymm1, %ymm4, %ymm2 vmovaps 0x3c0(%rsp), %ymm3 vblendvps %ymm3, %ymm2, %ymm0, %ymm2 vmovaps %ymm2, 0x420(%rsp) vsubps %ymm0, %ymm4, %ymm0 vblendvps %ymm3, %ymm0, %ymm1, %ymm0 vmovaps %ymm0, 0x440(%rsp) vmovaps 0x3a0(%rsp), %ymm3 vmulps %ymm3, %ymm14, %ymm1 vmovaps %ymm1, 0x540(%rsp) vmulps %ymm3, %ymm15, %ymm1 vmovaps %ymm1, 0x560(%rsp) vrcpps %ymm4, %ymm1 vmulps 0x160(%rsp), %ymm3, %ymm3 vmovaps %ymm3, 0x580(%rsp) vmulps %ymm1, %ymm4, %ymm3 vbroadcastss 0x1cde87d(%rip), %ymm4 # 0x1eec714 vsubps %ymm3, %ymm4, %ymm3 vmulps %ymm3, %ymm1, %ymm3 vaddps %ymm3, %ymm1, %ymm1 vmulps 0x460(%rsp), %ymm1, %ymm3 vmovaps %ymm3, 0x520(%rsp) vmulps %ymm1, %ymm2, %ymm2 vmovaps %ymm2, 0x4e0(%rsp) vmulps %ymm1, %ymm0, %ymm0 vmovaps %ymm0, 0x500(%rsp) vmovaps 0x4c0(%rsp), %ymm0 vmovmskps %ymm0, %edx movq %rax, %r13 bsfq %rdx, %rbx movl 0x400(%rsp,%rbx,4), %eax movq 0x78(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %r12 movl 0x24(%rsi), %ecx testl %ecx, 0x34(%r12) je 0x20df20 movq 0x20(%rsp), %rcx movq 0x10(%rcx), %r15 cmpq $0x0, 0x10(%r15) jne 0x20df3b cmpq $0x0, 0x48(%r12) jne 0x20df3b xorl %eax, %eax jmp 0x20df26 btcq %rbx, %rdx movb $0x1, %al testb %al, %al je 0x20e150 testq %rdx, %rdx movq %r13, %rax jne 0x20dedc jmp 0x20dd6a movq %rdx, 0x80(%rsp) vmovss 0x4e0(%rsp,%rbx,4), %xmm0 vmovss 0x500(%rsp,%rbx,4), %xmm1 movq 0x20(%rsp), %rcx movq 0x8(%rcx), %rcx movl 0x3e0(%rsp,%rbx,4), %edx vmovss 0x540(%rsp,%rbx,4), %xmm2 vmovss 0x560(%rsp,%rbx,4), %xmm3 vmovss 0x580(%rsp,%rbx,4), %xmm4 vmovss %xmm2, 0xe0(%rsp) vmovss %xmm3, 0xe4(%rsp) vmovss %xmm4, 0xe8(%rsp) vmovss %xmm0, 0xec(%rsp) vmovss %xmm1, 0xf0(%rsp) movl %edx, 0xf4(%rsp) movl %eax, 0xf8(%rsp) movl (%rcx), %eax movl %eax, 0xfc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0x100(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0x140(%rsp) vmovss 0x520(%rsp,%rbx,4), %xmm0 vmovss %xmm0, 0x20(%rsi) orl $-0x1, 0x34(%rsp) leaq 0x34(%rsp), %rax movq %rax, 0x110(%rsp) movq 0x18(%r12), %rax movq %rax, 0x118(%rsp) movq %rcx, 0x120(%rsp) movq %rsi, 0x128(%rsp) leaq 0xe0(%rsp), %rax movq %rax, 0x130(%rsp) movl $0x1, 0x138(%rsp) movq 0x48(%r12), %rax testq %rax, %rax je 0x20e058 leaq 0x110(%rsp), %rdi vzeroupper callq %rax movq 0x110(%rsp), %rax cmpl $0x0, (%rax) je 0x20e08d movq 0x10(%r15), %rax testq %rax, %rax je 0x20e089 testb $0x2, (%r15) jne 0x20e06f testb $0x40, 0x3e(%r12) je 0x20e07c leaq 0x110(%rsp), %rdi vzeroupper callq %rax movq 0x110(%rsp), %rax cmpl $0x0, (%rax) je 0x20e08d xorl %eax, %eax jmp 0x20e0b6 movq 0x28(%rsp), %rax vmovss 0x140(%rsp), %xmm0 vmovss %xmm0, 0x20(%rax) movq 0x80(%rsp), %rax btcq %rbx, %rax movq %rax, 0x80(%rsp) movb $0x1, %al movq 0x28(%rsp), %rsi movq 0x80(%rsp), %rdx jmp 0x20df26 movq 0x70(%rsp), %rdi vmovaps 0x360(%rsp), %ymm6 vmovaps 0x340(%rsp), %ymm7 vmovaps 0x320(%rsp), %ymm8 movq 0x60(%rsp), %r8 vmovaps 0x300(%rsp), %ymm9 vmovaps 0x2e0(%rsp), %ymm10 vmovaps 0x2c0(%rsp), %ymm11 movq 0x58(%rsp), %r9 movq 0x50(%rsp), %r10 movq 0x48(%rsp), %r11 movq 0x40(%rsp), %rbx movq 0x38(%rsp), %r15 vmovaps 0x2a0(%rsp), %ymm12 vmovaps 0x280(%rsp), %ymm13 leaq 0x5a0(%rsp), %r12 movq 0x68(%rsp), %r13 xorl %ecx, %ecx cmpl $0x3, %ecx jne 0x20d4de jmp 0x20d30b testb $0x1, 0x1f(%rsp) vxorps %xmm5, %xmm5, %xmm5 movq 0x70(%rsp), %rdi vmovaps 0x360(%rsp), %ymm6 vmovaps 0x340(%rsp), %ymm7 vmovaps 0x320(%rsp), %ymm8 movq 0x60(%rsp), %r8 vmovaps 0x300(%rsp), %ymm9 vmovaps 0x2e0(%rsp), %ymm10 vmovaps 0x2c0(%rsp), %ymm11 movq 0x58(%rsp), %r9 movq 0x50(%rsp), %r10 movq 0x48(%rsp), %r11 movq 0x40(%rsp), %rbx movq 0x38(%rsp), %r15 vmovaps 0x2a0(%rsp), %ymm12 vmovaps 0x280(%rsp), %ymm13 leaq 0x5a0(%rsp), %r12 movq 0x68(%rsp), %r13 movl $0x0, %ecx je 0x20e142 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %rcx jmp 0x20e142 nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, true, embree::avx::ArrayIntersector1<embree::avx::QuadMiMBIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x18a0, %rsp # imm = 0x18A0 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x20f446 vmovss 0x20(%rsi), %xmm0 vxorps %xmm6, %xmm6, %xmm6 vucomiss %xmm0, %xmm6 ja 0x20f446 leaq 0x6e8(%rsp), %r8 movq 0x70(%rax), %rax vmovaps 0x10(%rsi), %xmm2 vmaxss 0xc(%rsi), %xmm6, %xmm1 vbroadcastss 0x1d12c80(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1ce2d97(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1cde4b5(%rip), %xmm4 # 0x1eec714 vdivps %xmm2, %xmm4, %xmm2 vbroadcastss 0x1d12cf4(%rip), %xmm4 # 0x1f20f60 vblendvps %xmm3, %xmm4, %xmm2, %xmm2 movq %rax, -0x8(%r8) vbroadcastss 0x1d11c91(%rip), %xmm3 # 0x1f1ff10 vmulps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1d11c88(%rip), %xmm4 # 0x1f1ff14 vbroadcastss (%rsi), %ymm7 vbroadcastss 0x4(%rsi), %ymm8 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss 0x8(%rsi), %ymm9 xorl %r9d, %r9d vucomiss %xmm6, %xmm3 setb %r9b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovaps %ymm4, 0x4c0(%rsp) vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm11 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm12 vshufps $0x0, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm13 vshufps $0x55, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm3, %ymm3, %ymm14 vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] shll $0x5, %r9d xorl %r10d, %r10d vucomiss %xmm6, %xmm4 vinsertf128 $0x1, %xmm2, %ymm2, %ymm15 setb %r10b shll $0x5, %r10d orq $0x40, %r10 xorl %r11d, %r11d vucomiss %xmm6, %xmm5 setb %r11b shll $0x5, %r11d orq $0x80, %r11 movq %r9, %rbx xorq $0x20, %rbx movq %r10, %r15 xorq $0x20, %r15 movq %r11, %rax xorq $0x20, %rax movq %rax, 0xb8(%rsp) vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm5 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovaps %ymm0, 0x4a0(%rsp) leaq 0x1f41c10(%rip), %rax # 0x214ff80 vmovups (%rax), %ymm0 vinsertf128 $0x1, 0xf0(%rax), %ymm0, %ymm1 vbroadcastss 0x1d12b3d(%rip), %ymm6 # 0x1f20ec4 vbroadcastss 0x1cde384(%rip), %ymm10 # 0x1eec714 vbroadcastss 0x1ce2633(%rip), %ymm0 # 0x1ef09cc vmovaps %ymm1, 0x480(%rsp) vblendvps %ymm1, %ymm0, %ymm10, %ymm0 vmovaps %ymm0, 0x460(%rsp) movq %rdx, 0x8(%rsp) movq %rsi, 0x20(%rsp) vmovaps %ymm7, 0x240(%rsp) vmovaps %ymm8, 0x220(%rsp) vmovaps %ymm9, 0x200(%rsp) movq %r9, 0x50(%rsp) vmovaps %ymm11, 0x1e0(%rsp) vmovaps %ymm12, 0x1c0(%rsp) vmovaps %ymm13, 0x1a0(%rsp) vmovaps %ymm14, 0x180(%rsp) movq %r10, 0x48(%rsp) vmovaps %ymm15, 0x160(%rsp) movq %r11, 0x40(%rsp) movq %rbx, 0x38(%rsp) movq %r15, 0x30(%rsp) vmovaps %ymm5, 0x140(%rsp) leaq 0x6e0(%rsp), %rax cmpq %rax, %r8 je 0x20f446 movq -0x8(%r8), %r14 addq $-0x8, %r8 testb $0x8, %r14b jne 0x20e544 movq %r14, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rsi), %ymm0 vmulps 0x100(%rax,%r9), %ymm0, %ymm1 vaddps 0x40(%rax,%r9), %ymm1, %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps 0x4c0(%rsp), %ymm1, %ymm1 vmulps 0x100(%rax,%r10), %ymm0, %ymm2 vaddps 0x40(%rax,%r10), %ymm2, %ymm2 vmaxps %ymm1, %ymm5, %ymm1 vsubps %ymm8, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmulps %ymm2, %ymm11, %ymm2 vaddps 0x40(%rax,%r11), %ymm3, %ymm3 vsubps %ymm9, %ymm3, %ymm3 vmulps %ymm3, %ymm12, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%rbx), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%rbx), %ymm3, %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm13, %ymm2 vmovaps 0x4a0(%rsp), %ymm3 vminps %ymm2, %ymm3, %ymm2 vmulps 0x100(%rax,%r15), %ymm0, %ymm3 vaddps 0x40(%rax,%r15), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm14, %ymm3 movq 0xb8(%rsp), %rcx vmulps 0x100(%rax,%rcx), %ymm0, %ymm4 vaddps 0x40(%rax,%rcx), %ymm4, %ymm4 vsubps %ymm9, %ymm4, %ymm4 vmulps %ymm4, %ymm15, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r14d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x20e58e vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r13d testb $0x8, %r14b jne 0x20e58a testq %r13, %r13 je 0x20e5bb andq $-0x10, %r14 bsfq %r13, %rax leaq -0x1(%r13), %rdi xorl %r12d, %r12d movq (%r14,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r13, %rdi jne 0x20e5c1 movq %rax, %r14 testl %r12d, %r12d je 0x20e43e jmp 0x20e607 pushq $0x6 jmp 0x20e5bd vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x20e533 pushq $0x4 popq %r12 jmp 0x20e57f movq %rax, (%r8) addq $0x8, %r8 bsfq %rdi, %rcx leaq -0x1(%rdi), %rax movq (%r14,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdi, %rax je 0x20e5ff movq %rcx, (%r8) addq $0x8, %r8 bsfq %rax, %rcx leaq -0x1(%rax), %rdi jmp 0x20e5d0 movq %rcx, %r14 jmp 0x20e57f cmpl $0x6, %r12d jne 0x20f43c movl %r14d, %eax andl $0xf, %eax xorl %r12d, %r12d addq $-0x8, %rax setne %cl je 0x20f43c movq %r8, 0x58(%rsp) andq $-0x10, %r14 movq (%rdx), %r8 xorl %edi, %edi movq %r13, 0x18(%rsp) movq %rax, 0x98(%rsp) movq %r8, 0x10(%rsp) movb %cl, 0x7(%rsp) movq %rdi, 0xa0(%rsp) imulq $0x60, %rdi, %rax vmovss 0x1c(%rsi), %xmm0 movl 0x40(%r14,%rax), %ecx movq 0x1e8(%r8), %rdx movq (%rdx,%rcx,8), %rsi vmovss 0x28(%rsi), %xmm1 vmovss 0x2c(%rsi), %xmm2 vmovss 0x30(%rsi), %xmm3 vsubss %xmm2, %xmm0, %xmm0 vsubss %xmm2, %xmm3, %xmm2 vdivss %xmm2, %xmm0, %xmm0 vmulss %xmm0, %xmm1, %xmm0 vroundss $0x9, %xmm0, %xmm0, %xmm2 vaddss 0x1ce2333(%rip), %xmm1, %xmm1 # 0x1ef09cc vminss %xmm1, %xmm2, %xmm1 vxorps %xmm2, %xmm2, %xmm2 vmaxss %xmm1, %xmm2, %xmm3 vcvttss2si %xmm3, %ecx movslq %ecx, %rcx movq 0xe0(%rsi), %r15 imulq $0x38, %rcx, %r9 movq (%r15,%r9), %rsi movl (%r14,%rax), %r11d movl 0x4(%r14,%rax), %r8d movq %r8, 0x2e0(%rsp) vmovups (%rsi,%r11,4), %xmm2 movl 0x10(%r14,%rax), %edi movq %rdi, 0x3c0(%rsp) vmovups (%rsi,%rdi,4), %xmm1 movl 0x20(%r14,%rax), %edi movq %rdi, 0x120(%rsp) vmovups (%rsi,%rdi,4), %xmm7 movl 0x30(%r14,%rax), %edi movq %rdi, 0x60(%rsp) vmovups (%rsi,%rdi,4), %xmm8 movl 0x44(%r14,%rax), %esi movq (%rdx,%rsi,8), %rsi movq 0xe0(%rsi), %rsi movq %rsi, 0xb0(%rsp) movq (%rsi,%r9), %rsi vmovups (%rsi,%r8,4), %xmm4 movl 0x14(%r14,%rax), %edi movq %rdi, 0x360(%rsp) vmovups (%rsi,%rdi,4), %xmm5 movl 0x24(%r14,%rax), %edi movq %rdi, 0x3a0(%rsp) vmovups (%rsi,%rdi,4), %xmm10 movl 0x34(%r14,%rax), %edi movq %rdi, 0x3e0(%rsp) vmovups (%rsi,%rdi,4), %xmm9 movl 0x48(%r14,%rax), %esi movq (%rdx,%rsi,8), %rsi movq 0xe0(%rsi), %r8 movq (%r8,%r9), %rsi movl 0x8(%r14,%rax), %edi vmovups (%rsi,%rdi,4), %xmm13 movl 0x18(%r14,%rax), %r10d movq %r10, 0xa8(%rsp) vmovups (%rsi,%r10,4), %xmm6 movl 0x28(%r14,%rax), %r10d movq %r10, 0x260(%rsp) vmovups (%rsi,%r10,4), %xmm11 movl 0x38(%r14,%rax), %r10d movq %r10, 0x340(%rsp) vmovups (%rsi,%r10,4), %xmm12 movl 0x4c(%r14,%rax), %esi movq (%rdx,%rsi,8), %rdx vsubss %xmm3, %xmm0, %xmm0 vmovaps %xmm0, 0x380(%rsp) movq 0xe0(%rdx), %rsi movq (%rsi,%r9), %rbx movl 0xc(%r14,%rax), %edx vmovups (%rbx,%rdx,4), %xmm0 movl 0x1c(%r14,%rax), %r12d vmovups (%rbx,%r12,4), %xmm3 vunpcklps %xmm13, %xmm2, %xmm14 # xmm14 = xmm2[0],xmm13[0],xmm2[1],xmm13[1] vunpckhps %xmm13, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm13[2],xmm2[3],xmm13[3] vunpcklps %xmm0, %xmm4, %xmm13 # xmm13 = xmm4[0],xmm0[0],xmm4[1],xmm0[1] vunpckhps %xmm0, %xmm4, %xmm0 # xmm0 = xmm4[2],xmm0[2],xmm4[3],xmm0[3] movl 0x2c(%r14,%rax), %r9d vmovups (%rbx,%r9,4), %xmm4 vunpcklps %xmm0, %xmm2, %xmm0 # xmm0 = xmm2[0],xmm0[0],xmm2[1],xmm0[1] vmovaps %xmm0, 0x300(%rsp) vunpcklps %xmm13, %xmm14, %xmm0 # xmm0 = xmm14[0],xmm13[0],xmm14[1],xmm13[1] vmovaps %xmm0, 0x450(%rsp) vunpckhps %xmm13, %xmm14, %xmm0 # xmm0 = xmm14[2],xmm13[2],xmm14[3],xmm13[3] vmovaps %xmm0, 0x2a0(%rsp) vunpcklps %xmm6, %xmm1, %xmm0 # xmm0 = xmm1[0],xmm6[0],xmm1[1],xmm6[1] vunpckhps %xmm6, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm6[2],xmm1[3],xmm6[3] vunpcklps %xmm3, %xmm5, %xmm2 # xmm2 = xmm5[0],xmm3[0],xmm5[1],xmm3[1] vunpckhps %xmm3, %xmm5, %xmm3 # xmm3 = xmm5[2],xmm3[2],xmm5[3],xmm3[3] movl 0x3c(%r14,%rax), %r10d vmovups (%rbx,%r10,4), %xmm5 vunpcklps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vmovaps %xmm1, 0x320(%rsp) vunpcklps %xmm2, %xmm0, %xmm1 # xmm1 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vmovaps %xmm1, 0x280(%rsp) vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x2c0(%rsp) vunpcklps %xmm11, %xmm7, %xmm0 # xmm0 = xmm7[0],xmm11[0],xmm7[1],xmm11[1] vunpckhps %xmm11, %xmm7, %xmm1 # xmm1 = xmm7[2],xmm11[2],xmm7[3],xmm11[3] vunpcklps %xmm4, %xmm10, %xmm2 # xmm2 = xmm10[0],xmm4[0],xmm10[1],xmm4[1] vunpckhps %xmm4, %xmm10, %xmm3 # xmm3 = xmm10[2],xmm4[2],xmm10[3],xmm4[3] incl %ecx movslq %ecx, %rcx imulq $0x38, %rcx, %rbx movq (%r15,%rbx), %rcx vmovups (%rcx,%r11,4), %xmm4 vunpcklps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vmovaps %xmm1, 0x440(%rsp) vunpcklps %xmm2, %xmm0, %xmm1 # xmm1 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vmovaps %xmm1, 0x400(%rsp) vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x410(%rsp) vunpcklps %xmm12, %xmm8, %xmm0 # xmm0 = xmm8[0],xmm12[0],xmm8[1],xmm12[1] vunpckhps %xmm12, %xmm8, %xmm1 # xmm1 = xmm8[2],xmm12[2],xmm8[3],xmm12[3] vunpcklps %xmm5, %xmm9, %xmm2 # xmm2 = xmm9[0],xmm5[0],xmm9[1],xmm5[1] vunpckhps %xmm5, %xmm9, %xmm3 # xmm3 = xmm9[2],xmm5[2],xmm9[3],xmm5[3] movq (%r8,%rbx), %r8 vmovups (%r8,%rdi,4), %xmm5 vunpcklps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vmovaps %xmm1, 0x430(%rsp) vunpcklps %xmm2, %xmm0, %xmm12 # xmm12 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] vunpckhps %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm2[2],xmm0[3],xmm2[3] vmovaps %xmm0, 0x420(%rsp) movq 0xb0(%rsp), %rdi movq (%rdi,%rbx), %rdi vunpcklps %xmm5, %xmm4, %xmm0 # xmm0 = xmm4[0],xmm5[0],xmm4[1],xmm5[1] vunpckhps %xmm5, %xmm4, %xmm1 # xmm1 = xmm4[2],xmm5[2],xmm4[3],xmm5[3] movq 0x2e0(%rsp), %r11 vmovups (%rdi,%r11,4), %xmm2 movq (%rsi,%rbx), %r11 vmovups (%r11,%rdx,4), %xmm3 vunpcklps %xmm3, %xmm2, %xmm4 # xmm4 = xmm2[0],xmm3[0],xmm2[1],xmm3[1] vunpckhps %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm3[2],xmm2[3],xmm3[3] vunpcklps %xmm2, %xmm1, %xmm13 # xmm13 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] vunpcklps %xmm4, %xmm0, %xmm10 # xmm10 = xmm0[0],xmm4[0],xmm0[1],xmm4[1] vunpckhps %xmm4, %xmm0, %xmm7 # xmm7 = xmm0[2],xmm4[2],xmm0[3],xmm4[3] movq 0x3c0(%rsp), %rdx vmovups (%rcx,%rdx,4), %xmm0 movq 0xa8(%rsp), %rdx vmovups (%r8,%rdx,4), %xmm1 vunpcklps %xmm1, %xmm0, %xmm3 # xmm3 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpckhps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3] movq 0x360(%rsp), %rdx vmovups (%rdi,%rdx,4), %xmm1 vmovups (%r11,%r12,4), %xmm2 vunpcklps %xmm2, %xmm1, %xmm4 # xmm4 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] vunpckhps %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm2[2],xmm1[3],xmm2[3] vunpcklps %xmm1, %xmm0, %xmm14 # xmm14 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm4, %xmm3, %xmm2 # xmm2 = xmm3[0],xmm4[0],xmm3[1],xmm4[1] vunpckhps %xmm4, %xmm3, %xmm8 # xmm8 = xmm3[2],xmm4[2],xmm3[3],xmm4[3] movq 0x120(%rsp), %rdx vmovups (%rcx,%rdx,4), %xmm0 movq 0x260(%rsp), %rdx vmovups (%r8,%rdx,4), %xmm1 vunpcklps %xmm1, %xmm0, %xmm4 # xmm4 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpckhps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3] movq 0x3a0(%rsp), %rdx vmovups (%rdi,%rdx,4), %xmm1 vmovups (%r11,%r9,4), %xmm3 vunpcklps %xmm3, %xmm1, %xmm5 # xmm5 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vunpckhps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm3[2],xmm1[3],xmm3[3] vunpcklps %xmm1, %xmm0, %xmm15 # xmm15 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm5, %xmm4, %xmm3 # xmm3 = xmm4[0],xmm5[0],xmm4[1],xmm5[1] vunpckhps %xmm5, %xmm4, %xmm9 # xmm9 = xmm4[2],xmm5[2],xmm4[3],xmm5[3] movq 0x60(%rsp), %rdx vmovups (%rcx,%rdx,4), %xmm0 movq 0x340(%rsp), %rcx vmovups (%r8,%rcx,4), %xmm1 movq 0x20(%rsp), %rsi vunpcklps %xmm1, %xmm0, %xmm5 # xmm5 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpckhps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm1[2],xmm0[3],xmm1[3] movq 0x3e0(%rsp), %rcx vmovups (%rdi,%rcx,4), %xmm1 vmovups (%r11,%r10,4), %xmm4 vunpcklps %xmm4, %xmm1, %xmm11 # xmm11 = xmm1[0],xmm4[0],xmm1[1],xmm4[1] vunpckhps %xmm4, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm4[2],xmm1[3],xmm4[3] vunpcklps %xmm1, %xmm0, %xmm6 # xmm6 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm11, %xmm5, %xmm4 # xmm4 = xmm5[0],xmm11[0],xmm5[1],xmm11[1] vunpckhps %xmm11, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm11[2],xmm5[3],xmm11[3] vmovaps 0x380(%rsp), %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm0 # xmm0 = xmm11[0,0,0,0] vmovss 0x1cddce3(%rip), %xmm1 # 0x1eec714 vsubss %xmm11, %xmm1, %xmm1 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmulps %xmm0, %xmm10, %xmm10 vmulps 0x450(%rsp), %xmm1, %xmm11 vaddps %xmm10, %xmm11, %xmm11 vmulps %xmm7, %xmm0, %xmm7 vmulps 0x2a0(%rsp), %xmm1, %xmm10 vaddps %xmm7, %xmm10, %xmm10 vmulps %xmm0, %xmm13, %xmm7 vmulps 0x300(%rsp), %xmm1, %xmm13 vaddps %xmm7, %xmm13, %xmm13 vmulps %xmm2, %xmm0, %xmm2 vmulps 0x280(%rsp), %xmm1, %xmm7 vaddps %xmm2, %xmm7, %xmm7 vmulps %xmm0, %xmm8, %xmm2 vmulps 0x2c0(%rsp), %xmm1, %xmm8 vaddps %xmm2, %xmm8, %xmm2 vmulps %xmm0, %xmm14, %xmm8 vmulps 0x320(%rsp), %xmm1, %xmm14 vaddps %xmm8, %xmm14, %xmm14 vmulps %xmm3, %xmm0, %xmm3 vmulps 0x400(%rsp), %xmm1, %xmm8 vaddps %xmm3, %xmm8, %xmm3 vmulps %xmm0, %xmm9, %xmm8 vmulps 0x410(%rsp), %xmm1, %xmm9 vaddps %xmm8, %xmm9, %xmm8 vmulps %xmm0, %xmm15, %xmm9 vmulps 0x440(%rsp), %xmm1, %xmm15 vaddps %xmm9, %xmm15, %xmm9 vmulps %xmm4, %xmm0, %xmm4 vmulps %xmm5, %xmm0, %xmm5 vmulps %xmm6, %xmm0, %xmm0 vmulps %xmm1, %xmm12, %xmm6 vbroadcastf128 0x40(%r14,%rax), %ymm12 # ymm12 = mem[0,1,0,1] vaddps %xmm4, %xmm6, %xmm4 vmovaps %ymm12, 0x500(%rsp) vbroadcastf128 0x50(%r14,%rax), %ymm6 # ymm6 = mem[0,1,0,1] vmulps 0x420(%rsp), %xmm1, %xmm12 vaddps %xmm5, %xmm12, %xmm5 vmulps 0x430(%rsp), %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmovaps %ymm6, 0x4e0(%rsp) vinsertf128 $0x1, %xmm3, %ymm11, %ymm1 vinsertf128 $0x1, %xmm8, %ymm10, %ymm3 vinsertf128 $0x1, %xmm9, %ymm13, %ymm6 vinsertf128 $0x1, %xmm7, %ymm7, %ymm7 vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vinsertf128 $0x1, %xmm14, %ymm14, %ymm14 vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vinsertf128 $0x1, %xmm5, %ymm5, %ymm5 vinsertf128 $0x1, %xmm0, %ymm0, %ymm9 vbroadcastss (%rsi), %ymm10 vbroadcastss 0x4(%rsi), %ymm11 vbroadcastss 0x8(%rsi), %ymm12 vsubps %ymm10, %ymm1, %ymm1 vsubps %ymm11, %ymm3, %ymm3 vsubps %ymm12, %ymm6, %ymm15 vsubps %ymm10, %ymm7, %ymm13 vsubps %ymm11, %ymm2, %ymm8 vsubps %ymm12, %ymm14, %ymm14 vsubps %ymm10, %ymm4, %ymm2 vsubps %ymm11, %ymm5, %ymm4 vmovaps %ymm4, 0x120(%rsp) vsubps %ymm12, %ymm9, %ymm6 vmovaps %ymm6, 0x60(%rsp) vsubps %ymm1, %ymm2, %ymm11 vsubps %ymm3, %ymm4, %ymm12 vsubps %ymm15, %ymm6, %ymm0 vaddps %ymm3, %ymm4, %ymm4 vaddps %ymm6, %ymm15, %ymm5 vmulps %ymm4, %ymm0, %ymm7 vmulps %ymm5, %ymm12, %ymm9 vsubps %ymm7, %ymm9, %ymm9 vaddps %ymm1, %ymm2, %ymm7 vmulps %ymm5, %ymm11, %ymm5 vmovaps %ymm0, 0x320(%rsp) vmulps %ymm7, %ymm0, %ymm10 vsubps %ymm5, %ymm10, %ymm5 vmovaps %ymm12, 0x360(%rsp) vmulps %ymm7, %ymm12, %ymm7 vmovaps %ymm11, 0x380(%rsp) vmulps %ymm4, %ymm11, %ymm4 vsubps %ymm7, %ymm4, %ymm4 vbroadcastss 0x14(%rsi), %ymm11 vbroadcastss 0x18(%rsi), %ymm6 vmulps %ymm4, %ymm6, %ymm4 vmulps %ymm5, %ymm11, %ymm5 vbroadcastss 0x10(%rsi), %ymm0 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm0, %ymm9, %ymm5 vaddps %ymm4, %ymm5, %ymm10 vsubps %ymm8, %ymm3, %ymm9 vsubps %ymm14, %ymm15, %ymm12 vmovaps %ymm3, 0x3c0(%rsp) vaddps %ymm3, %ymm8, %ymm4 vmovaps %ymm15, 0x3a0(%rsp) vaddps %ymm14, %ymm15, %ymm5 vmulps %ymm4, %ymm12, %ymm15 vmulps %ymm5, %ymm9, %ymm3 vsubps %ymm15, %ymm3, %ymm3 vsubps %ymm13, %ymm1, %ymm15 vmulps %ymm5, %ymm15, %ymm5 vmovaps %ymm1, 0x3e0(%rsp) vaddps %ymm1, %ymm13, %ymm7 vmovaps %ymm12, 0x300(%rsp) vmulps %ymm7, %ymm12, %ymm12 vsubps %ymm5, %ymm12, %ymm5 vmovaps %ymm9, %ymm12 vmulps %ymm7, %ymm9, %ymm7 vmulps %ymm4, %ymm15, %ymm4 vsubps %ymm7, %ymm4, %ymm4 vmulps %ymm4, %ymm6, %ymm4 vmulps %ymm5, %ymm11, %ymm5 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm3, %ymm0, %ymm3 vaddps %ymm4, %ymm3, %ymm4 vmovaps %ymm13, %ymm3 vsubps %ymm2, %ymm13, %ymm13 vaddps %ymm2, %ymm3, %ymm2 vmovaps 0x120(%rsp), %ymm1 vsubps %ymm1, %ymm8, %ymm9 vaddps %ymm1, %ymm8, %ymm1 vmovaps 0x60(%rsp), %ymm3 vsubps %ymm3, %ymm14, %ymm5 vaddps %ymm3, %ymm14, %ymm3 vmulps %ymm1, %ymm5, %ymm7 vmulps %ymm3, %ymm9, %ymm8 vsubps %ymm7, %ymm8, %ymm7 vmulps %ymm3, %ymm13, %ymm3 vmulps %ymm2, %ymm5, %ymm8 vsubps %ymm3, %ymm8, %ymm3 vmulps %ymm2, %ymm9, %ymm2 vmulps %ymm1, %ymm13, %ymm1 vsubps %ymm2, %ymm1, %ymm1 vmovaps %ymm6, 0x340(%rsp) vmulps %ymm1, %ymm6, %ymm1 vmovaps %ymm11, 0x60(%rsp) vmulps %ymm3, %ymm11, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vmovaps %ymm0, 0x120(%rsp) vmulps %ymm7, %ymm0, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vaddps %ymm4, %ymm10, %ymm2 vaddps %ymm2, %ymm1, %ymm0 vminps %ymm4, %ymm10, %ymm2 vminps %ymm1, %ymm2, %ymm2 vbroadcastss 0x1d1218b(%rip), %ymm3 # 0x1f20ec4 vmovaps %ymm0, 0x280(%rsp) vandps %ymm3, %ymm0, %ymm0 vbroadcastss 0x1d1217d(%rip), %ymm3 # 0x1f20ecc vmovaps %ymm0, 0x2a0(%rsp) vmulps %ymm3, %ymm0, %ymm3 vbroadcastss 0x1d1215b(%rip), %ymm7 # 0x1f20ec0 vxorps %ymm7, %ymm3, %ymm7 vcmpnltps %ymm7, %ymm2, %ymm2 vmovaps %ymm10, 0x2e0(%rsp) vmovaps %ymm4, 0x2c0(%rsp) vmaxps %ymm4, %ymm10, %ymm7 vmaxps %ymm1, %ymm7, %ymm1 vcmpleps %ymm3, %ymm1, %ymm1 vorps %ymm1, %ymm2, %ymm0 vtestps %ymm0, %ymm0 je 0x20f35e vmovaps %ymm0, 0x260(%rsp) vmovaps 0x320(%rsp), %ymm0 vmovaps %ymm12, %ymm11 vmulps %ymm0, %ymm12, %ymm2 vmovaps 0x360(%rsp), %ymm1 vmovaps %ymm9, %ymm8 vmovaps 0x300(%rsp), %ymm9 vmulps %ymm1, %ymm9, %ymm3 vsubps %ymm2, %ymm3, %ymm3 vmulps %ymm8, %ymm9, %ymm7 vmulps %ymm5, %ymm12, %ymm12 vsubps %ymm7, %ymm12, %ymm12 vbroadcastss 0x1d120d8(%rip), %ymm4 # 0x1f20ec4 vandps %ymm4, %ymm2, %ymm2 vandps %ymm4, %ymm7, %ymm7 vcmpltps %ymm7, %ymm2, %ymm2 vblendvps %ymm2, %ymm3, %ymm12, %ymm14 vmulps %ymm5, %ymm15, %ymm2 vmulps %ymm0, %ymm15, %ymm3 vmovaps 0x380(%rsp), %ymm10 vmulps %ymm9, %ymm10, %ymm6 vsubps %ymm6, %ymm3, %ymm3 vmulps %ymm13, %ymm9, %ymm7 vsubps %ymm2, %ymm7, %ymm7 vandps %ymm4, %ymm6, %ymm6 vandps %ymm4, %ymm2, %ymm2 vcmpltps %ymm2, %ymm6, %ymm2 vblendvps %ymm2, %ymm3, %ymm7, %ymm9 vmulps %ymm13, %ymm11, %ymm2 vmulps %ymm11, %ymm10, %ymm3 vmulps %ymm1, %ymm15, %ymm5 vmulps %ymm8, %ymm15, %ymm0 vsubps %ymm5, %ymm3, %ymm3 vsubps %ymm2, %ymm0, %ymm0 vandps %ymm4, %ymm5, %ymm5 vandps %ymm4, %ymm2, %ymm2 vcmpltps %ymm2, %ymm5, %ymm2 vblendvps %ymm2, %ymm3, %ymm0, %ymm0 vmovdqa 0x260(%rsp), %ymm1 vextractf128 $0x1, %ymm1, %xmm2 vpackssdw %xmm2, %xmm1, %xmm1 vmulps 0x340(%rsp), %ymm0, %ymm2 vmulps 0x60(%rsp), %ymm9, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vmulps 0x120(%rsp), %ymm14, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vaddps %ymm2, %ymm2, %ymm3 vmulps 0x3a0(%rsp), %ymm0, %ymm2 vmulps 0x3c0(%rsp), %ymm9, %ymm5 vaddps %ymm2, %ymm5, %ymm2 vmulps 0x3e0(%rsp), %ymm14, %ymm5 vrcpps %ymm3, %ymm6 vaddps %ymm2, %ymm5, %ymm2 vaddps %ymm2, %ymm2, %ymm2 vmulps %ymm6, %ymm3, %ymm5 vbroadcastss 0x1cdd842(%rip), %ymm4 # 0x1eec714 vsubps %ymm5, %ymm4, %ymm5 vmulps %ymm5, %ymm6, %ymm5 vaddps %ymm5, %ymm6, %ymm5 vbroadcastss 0xc(%rsi), %ymm6 vmulps %ymm5, %ymm2, %ymm2 vcmpleps %ymm2, %ymm6, %ymm5 vbroadcastss 0x20(%rsi), %ymm6 vcmpleps %ymm6, %ymm2, %ymm6 vandps %ymm5, %ymm6, %ymm5 vcmpneqps 0x1d11ffb(%rip), %ymm3, %ymm3 # 0x1f20f00 vandps %ymm3, %ymm5, %ymm3 vextractf128 $0x1, %ymm3, %xmm5 vpackssdw %xmm5, %xmm3, %xmm3 vpand %xmm1, %xmm3, %xmm1 vpmovsxwd %xmm1, %xmm3 vpshufd $0xee, %xmm1, %xmm5 # xmm5 = xmm1[2,3,2,3] vpmovsxwd %xmm5, %xmm5 vinsertf128 $0x1, %xmm5, %ymm3, %ymm3 vtestps %ymm3, %ymm3 je 0x20f35e vmovaps 0x2e0(%rsp), %ymm4 vmovaps %ymm4, 0x520(%rsp) vmovaps 0x2c0(%rsp), %ymm5 vmovaps %ymm5, 0x540(%rsp) vmovaps 0x280(%rsp), %ymm7 vmovaps %ymm7, 0x560(%rsp) vmovaps %ymm14, 0x580(%rsp) vmovaps %ymm9, 0x5a0(%rsp) vmovaps %ymm0, 0x5c0(%rsp) vmovaps %ymm3, 0x5e0(%rsp) vmovaps %ymm2, 0x640(%rsp) vmovaps 0x480(%rsp), %ymm6 vmovaps %ymm6, 0x6c0(%rsp) vrcpps %ymm7, %ymm2 vmulps %ymm2, %ymm7, %ymm3 vbroadcastss 0x1cdd757(%rip), %ymm10 # 0x1eec714 vsubps %ymm3, %ymm10, %ymm3 vmulps %ymm3, %ymm2, %ymm3 vaddps %ymm3, %ymm2, %ymm2 vbroadcastss 0x1ce2016(%rip), %ymm3 # 0x1ef0fe8 vmovaps 0x2a0(%rsp), %ymm7 vcmpnltps %ymm3, %ymm7, %ymm3 vandps %ymm2, %ymm3, %ymm2 vmulps %ymm2, %ymm4, %ymm3 vminps %ymm10, %ymm3, %ymm3 vmulps %ymm2, %ymm5, %ymm2 vminps %ymm10, %ymm2, %ymm2 vsubps %ymm3, %ymm10, %ymm4 vsubps %ymm2, %ymm10, %ymm5 vblendvps %ymm6, %ymm4, %ymm2, %ymm2 vblendvps %ymm6, %ymm5, %ymm3, %ymm3 vmovaps %ymm2, 0x620(%rsp) vmovaps %ymm3, 0x600(%rsp) vmovaps 0x460(%rsp), %ymm4 vmulps %ymm4, %ymm14, %ymm2 vmulps %ymm4, %ymm9, %ymm3 vmulps %ymm0, %ymm4, %ymm0 vmovaps %ymm2, 0x660(%rsp) vmovaps %ymm3, 0x680(%rsp) vmovaps %ymm0, 0x6a0(%rsp) vpsllw $0xf, %xmm1, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r15d movq 0x8(%rsp), %rdx vmovaps 0x240(%rsp), %ymm7 vmovaps 0x220(%rsp), %ymm8 vmovaps 0x200(%rsp), %ymm9 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x160(%rsp), %ymm15 vmovaps 0x140(%rsp), %ymm5 vbroadcastss 0x1d11e08(%rip), %ymm6 # 0x1f20ec4 movq 0x10(%rsp), %r8 movb 0x7(%rsp), %dil bsfq %r15, %r12 movl 0x500(%rsp,%r12,4), %eax movq 0x1e8(%r8), %rcx movq (%rcx,%rax,8), %r13 movl 0x24(%rsi), %ecx testl %ecx, 0x34(%r13) je 0x20f0fc movq 0x10(%rdx), %rbx cmpq $0x0, 0x10(%rbx) jne 0x20f11c cmpq $0x0, 0x48(%r13) jne 0x20f11c xorl %eax, %eax jmp 0x20f102 btcq %r12, %r15 movb $0x1, %al movq 0x18(%rsp), %r13 xorl %r12d, %r12d testb %al, %al je 0x20f40d testq %r15, %r15 jne 0x20f0c6 jmp 0x20f3cb vmovss 0x600(%rsp,%r12,4), %xmm0 vmovd 0x620(%rsp,%r12,4), %xmm1 movq 0x8(%rdx), %rcx movl 0x4e0(%rsp,%r12,4), %edi vmovss 0x660(%rsp,%r12,4), %xmm2 vmovss 0x680(%rsp,%r12,4), %xmm3 vmovss 0x6a0(%rsp,%r12,4), %xmm4 vmovss %xmm2, 0xc0(%rsp) vmovss %xmm3, 0xc4(%rsp) vmovss %xmm4, 0xc8(%rsp) vmovss %xmm0, 0xcc(%rsp) vmovd %xmm1, 0xd0(%rsp) movl %edi, 0xd4(%rsp) movl %eax, 0xd8(%rsp) movl (%rcx), %eax movl %eax, 0xdc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xe0(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0x60(%rsp) vmovd 0x640(%rsp,%r12,4), %xmm0 vmovd %xmm0, 0x20(%rsi) orl $-0x1, 0x2c(%rsp) leaq 0x2c(%rsp), %rax movq %rax, 0xf0(%rsp) movq 0x18(%r13), %rax movq %rax, 0xf8(%rsp) movq %rcx, 0x100(%rsp) movq %rsi, 0x108(%rsp) leaq 0xc0(%rsp), %rax movq %rax, 0x110(%rsp) movl $0x1, 0x118(%rsp) movq 0x48(%r13), %rax testq %rax, %rax je 0x20f292 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0x10(%rsp), %r8 vmovaps 0x140(%rsp), %ymm5 vmovaps 0x160(%rsp), %ymm15 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x200(%rsp), %ymm9 vmovaps 0x220(%rsp), %ymm8 vmovaps 0x240(%rsp), %ymm7 movq 0x20(%rsp), %rsi movq 0x8(%rsp), %rdx movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x20f329 movq 0x10(%rbx), %rax testq %rax, %rax je 0x20f325 testb $0x2, (%rbx) jne 0x20f2ab testb $0x40, 0x3e(%r13) je 0x20f318 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0x10(%rsp), %r8 vmovaps 0x140(%rsp), %ymm5 vmovaps 0x160(%rsp), %ymm15 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x200(%rsp), %ymm9 vmovaps 0x220(%rsp), %ymm8 vmovaps 0x240(%rsp), %ymm7 movq 0x20(%rsp), %rsi movq 0x8(%rsp), %rdx movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x20f329 xorl %eax, %eax jmp 0x20f33a vmovd 0x60(%rsp), %xmm0 vmovd %xmm0, 0x20(%rsi) btcq %r12, %r15 movb $0x1, %al vbroadcastss 0x1d11b81(%rip), %ymm6 # 0x1f20ec4 vbroadcastss 0x1cdd3c8(%rip), %ymm10 # 0x1eec714 movq 0x18(%rsp), %r13 xorl %r12d, %r12d movb 0x7(%rsp), %dil jmp 0x20f10a movq 0x8(%rsp), %rdx vmovaps 0x240(%rsp), %ymm7 vmovaps 0x220(%rsp), %ymm8 vmovaps 0x200(%rsp), %ymm9 vmovaps 0x1e0(%rsp), %ymm11 vmovaps 0x1c0(%rsp), %ymm12 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x180(%rsp), %ymm14 vmovaps 0x160(%rsp), %ymm15 vmovaps 0x140(%rsp), %ymm5 vbroadcastss 0x1d11b07(%rip), %ymm6 # 0x1f20ec4 vbroadcastss 0x1cdd34e(%rip), %ymm10 # 0x1eec714 movq 0x10(%rsp), %r8 movq 0xa0(%rsp), %rdi incq %rdi movq 0x98(%rsp), %rax cmpq %rax, %rdi setb %cl jne 0x20e647 movq 0x58(%rsp), %r8 movq 0x50(%rsp), %r9 movq 0x48(%rsp), %r10 movq 0x40(%rsp), %r11 movq 0x38(%rsp), %rbx movq 0x30(%rsp), %r15 xorl %r12d, %r12d jmp 0x20f43c testb $0x1, %dil movq 0x58(%rsp), %r8 movq 0x50(%rsp), %r9 movq 0x48(%rsp), %r10 movq 0x40(%rsp), %r11 movq 0x38(%rsp), %rbx movq 0x30(%rsp), %r15 je 0x20f43c movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %r12 cmpl $0x3, %r12d jne 0x20e425 leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1048576, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMiIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1528, %rsp # imm = 0x1528 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x210125 vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x210125 movq %rdx, %r8 leaq 0x388(%rsp), %r10 movq 0x70(%rax), %rax movq %rax, -0x8(%r10) vmaxss 0xc(%rsi), %xmm1, %xmm2 vmovaps 0x10(%rsi), %xmm3 vbroadcastss 0x1d11a14(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce1b2b(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1cdd23b(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss (%rsi), %ymm5 vmovups %ymm5, 0x360(%rsp) vbroadcastss 0x4(%rsi), %ymm5 vmovups %ymm5, 0x340(%rsp) vaddps %xmm3, %xmm4, %xmm3 vbroadcastss 0x8(%rsi), %ymm4 vmovups %ymm4, 0x320(%rsp) xorl %r11d, %r11d vucomiss %xmm1, %xmm3 setb %r11b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovups %ymm4, 0x300(%rsp) vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm5 vmovups %ymm5, 0x2e0(%rsp) vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmovdqu %ymm3, 0x2c0(%rsp) xorl %eax, %eax vucomiss %xmm1, %xmm4 setb %al xorl %ecx, %ecx vucomiss %xmm1, %xmm5 setb %cl vshufps $0x0, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovups %ymm1, 0x2a0(%rsp) vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovups %ymm0, 0x280(%rsp) shll $0x3, %r11d movq %r11, %rdi xorq $0x8, %rdi leal 0x10(,%rax,8), %r9d movq %r9, %rbx xorq $0x8, %rbx leal 0x20(,%rcx,8), %r14d movq %r14, %r15 xorq $0x8, %r15 leaq 0x1f409b5(%rip), %rax # 0x214ff80 vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x100(%rsp) movq %rdi, 0x88(%rsp) movq %r9, 0x80(%rsp) movq %rbx, 0x78(%rsp) movq %r14, 0x70(%rsp) movq %r15, 0x68(%rsp) leaq 0x380(%rsp), %rax cmpq %rax, %r10 je 0x210125 movq -0x8(%r10), %r13 addq $-0x8, %r10 testb $0x8, %r13b jne 0x20f832 movq %r13, %rcx andq $-0x10, %rcx leaq 0x40(%rcx), %rax testq %rcx, %rcx cmoveq %rcx, %rax vmovq (%rax), %xmm0 vmovq 0x4(%rax), %xmm1 vmovq 0x8(%rax), %xmm2 vmovq 0xc(%rax), %xmm3 vpminub %xmm2, %xmm0, %xmm2 vpcmpeqb %xmm2, %xmm0, %xmm0 vpminub %xmm3, %xmm1, %xmm2 vpcmpeqb %xmm2, %xmm1, %xmm1 vbroadcastss 0x30(%rax), %ymm3 vbroadcastss 0x3c(%rax), %ymm4 vmovq (%rax,%r11), %xmm2 vpmovzxbd %xmm2, %xmm2 # xmm2 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero,xmm2[2],zero,zero,zero,xmm2[3],zero,zero,zero vmovq 0x4(%rax,%r11), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vinsertf128 $0x1, %xmm5, %ymm2, %ymm2 vcvtdq2ps %ymm2, %ymm2 vmulps %ymm2, %ymm4, %ymm2 vaddps %ymm2, %ymm3, %ymm2 vmovq (%rax,%rdi), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vmovq 0x4(%rax,%rdi), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vinsertf128 $0x1, %xmm6, %ymm5, %ymm5 vcvtdq2ps %ymm5, %ymm5 vmulps %ymm5, %ymm4, %ymm4 vaddps %ymm4, %ymm3, %ymm3 vbroadcastss 0x34(%rax), %ymm5 vbroadcastss 0x40(%rax), %ymm6 vmovq (%rax,%r9), %xmm4 vpmovzxbd %xmm4, %xmm4 # xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero,xmm4[2],zero,zero,zero,xmm4[3],zero,zero,zero vmovq 0x4(%rax,%r9), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vinsertf128 $0x1, %xmm7, %ymm4, %ymm4 vcvtdq2ps %ymm4, %ymm4 vmulps %ymm4, %ymm6, %ymm4 vaddps %ymm4, %ymm5, %ymm4 vmovq (%rax,%rbx), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x4(%rax,%rbx), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm6, %ymm6 vaddps %ymm6, %ymm5, %ymm5 vmovq (%rax,%r14), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vmovq 0x4(%rax,%r14), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vinsertf128 $0x1, %xmm7, %ymm6, %ymm6 vmovq (%rax,%r15), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x4(%rax,%r15), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vbroadcastss 0x44(%rax), %ymm8 vcvtdq2ps %ymm6, %ymm6 vmulps %ymm6, %ymm8, %ymm6 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm8, %ymm7 vbroadcastss 0x38(%rax), %ymm8 vaddps %ymm6, %ymm8, %ymm6 vaddps %ymm7, %ymm8, %ymm7 vmovups 0x360(%rsp), %ymm8 vsubps %ymm8, %ymm2, %ymm2 vmovups 0x300(%rsp), %ymm11 vmulps %ymm2, %ymm11, %ymm2 vmovups 0x340(%rsp), %ymm9 vsubps %ymm9, %ymm4, %ymm4 vmovups 0x2e0(%rsp), %ymm12 vmulps %ymm4, %ymm12, %ymm4 vmaxps %ymm4, %ymm2, %ymm2 vmovups 0x320(%rsp), %ymm10 vsubps %ymm10, %ymm6, %ymm4 vmovups 0x2c0(%rsp), %ymm6 vmulps %ymm4, %ymm6, %ymm4 vmaxps 0x2a0(%rsp), %ymm4, %ymm4 vmaxps %ymm4, %ymm2, %ymm2 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm11, %ymm3 vsubps %ymm9, %ymm5, %ymm4 vmulps %ymm4, %ymm12, %ymm4 vminps %ymm4, %ymm3, %ymm3 vsubps %ymm10, %ymm7, %ymm4 vmulps %ymm4, %ymm6, %ymm4 vminps 0x280(%rsp), %ymm4, %ymm4 vminps %ymm4, %ymm3, %ymm3 vcmpleps %ymm3, %ymm2, %ymm2 vpcmpeqd %xmm3, %xmm3, %xmm3 vpxor %xmm3, %xmm0, %xmm0 vpmovsxbd %xmm0, %xmm0 vpxor %xmm3, %xmm1, %xmm1 vpmovsxbd %xmm1, %xmm1 vinsertf128 $0x1, %xmm1, %ymm0, %ymm0 vcvtdq2ps %ymm0, %ymm0 vbroadcastss 0x1ce11a7(%rip), %ymm1 # 0x1ef09cc vcmpltps %ymm0, %ymm1, %ymm0 vandps %ymm0, %ymm2, %ymm0 vmovmskps %ymm0, %ebp testb $0x8, %r13b jne 0x20f872 testq %rbp, %rbp je 0x20f876 andq $-0x10, %r13 bsfq %rbp, %rax leaq -0x1(%rbp), %rdx xorl %r12d, %r12d movq (%r13,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) andq %rbp, %rdx jne 0x20f87c movq %rax, %r13 testl %r12d, %r12d je 0x20f614 jmp 0x20f8b9 pushq $0x6 jmp 0x20f878 pushq $0x4 popq %r12 jmp 0x20f867 movq %rax, (%r10) addq $0x8, %r10 bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r13,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) andq %rdx, %rax je 0x20f8b4 movq %rcx, (%r10) addq $0x8, %r10 bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x20f88b movq %rcx, %r13 jmp 0x20f867 cmpl $0x6, %r12d jne 0x21011b movl %r13d, %eax andl $0xf, %eax xorl %r12d, %r12d addq $-0x8, %rax setne %cl je 0x21011b andq $-0x10, %r13 movq (%r8), %rdi xorl %edx, %edx movq %rax, 0x90(%rsp) movq %rdi, 0x20(%rsp) movb %cl, 0xe(%rsp) movq %rdx, 0x98(%rsp) imulq $0x50, %rdx, %rax movl 0x30(%r13,%rax), %edx movq 0x228(%rdi), %rcx movq (%rcx,%rdx,8), %rdx movl (%r13,%rax), %r9d movl 0x4(%r13,%rax), %edi vmovups (%rdx,%r9,4), %xmm2 movl 0x10(%r13,%rax), %r9d vmovups (%rdx,%r9,4), %xmm1 movl 0x20(%r13,%rax), %r9d vmovups (%rdx,%r9,4), %xmm0 movl 0x34(%r13,%rax), %edx movq (%rcx,%rdx,8), %rdx vmovups (%rdx,%rdi,4), %xmm5 movl 0x14(%r13,%rax), %edi vmovups (%rdx,%rdi,4), %xmm4 movl 0x24(%r13,%rax), %edi vmovups (%rdx,%rdi,4), %xmm3 movl 0x38(%r13,%rax), %edx movq (%rcx,%rdx,8), %rdx movl 0x8(%r13,%rax), %edi vmovups (%rdx,%rdi,4), %xmm6 movl 0x18(%r13,%rax), %edi vmovups (%rdx,%rdi,4), %xmm7 movl 0x28(%r13,%rax), %edi vmovups (%rdx,%rdi,4), %xmm8 movq 0x20(%rsp), %rdi movl 0x3c(%r13,%rax), %edx movq (%rcx,%rdx,8), %rcx movl 0xc(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm9 movl 0x1c(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm10 movl 0x2c(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm11 movb 0xe(%rsp), %dl vunpcklps %xmm6, %xmm2, %xmm12 # xmm12 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] vunpckhps %xmm6, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm6[2],xmm2[3],xmm6[3] vunpcklps %xmm9, %xmm5, %xmm6 # xmm6 = xmm5[0],xmm9[0],xmm5[1],xmm9[1] vunpckhps %xmm9, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm9[2],xmm5[3],xmm9[3] vunpcklps %xmm5, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm5[0],xmm2[1],xmm5[1] vmovaps %xmm2, 0x10(%rsp) vunpcklps %xmm6, %xmm12, %xmm5 # xmm5 = xmm12[0],xmm6[0],xmm12[1],xmm6[1] vunpckhps %xmm6, %xmm12, %xmm6 # xmm6 = xmm12[2],xmm6[2],xmm12[3],xmm6[3] vunpcklps %xmm7, %xmm1, %xmm9 # xmm9 = xmm1[0],xmm7[0],xmm1[1],xmm7[1] vunpckhps %xmm7, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm7[2],xmm1[3],xmm7[3] vunpcklps %xmm10, %xmm4, %xmm7 # xmm7 = xmm4[0],xmm10[0],xmm4[1],xmm10[1] vunpckhps %xmm10, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm10[2],xmm4[3],xmm10[3] vunpcklps %xmm4, %xmm1, %xmm2 # xmm2 = xmm1[0],xmm4[0],xmm1[1],xmm4[1] vunpcklps %xmm7, %xmm9, %xmm10 # xmm10 = xmm9[0],xmm7[0],xmm9[1],xmm7[1] vunpckhps %xmm7, %xmm9, %xmm15 # xmm15 = xmm9[2],xmm7[2],xmm9[3],xmm7[3] vunpcklps %xmm8, %xmm0, %xmm9 # xmm9 = xmm0[0],xmm8[0],xmm0[1],xmm8[1] vunpckhps %xmm8, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm8[2],xmm0[3],xmm8[3] vunpcklps %xmm11, %xmm3, %xmm8 # xmm8 = xmm3[0],xmm11[0],xmm3[1],xmm11[1] vunpckhps %xmm11, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm11[2],xmm3[3],xmm11[3] vunpcklps %xmm3, %xmm0, %xmm4 # xmm4 = xmm0[0],xmm3[0],xmm0[1],xmm3[1] vunpcklps %xmm8, %xmm9, %xmm12 # xmm12 = xmm9[0],xmm8[0],xmm9[1],xmm8[1] vunpckhps %xmm8, %xmm9, %xmm8 # xmm8 = xmm9[2],xmm8[2],xmm9[3],xmm8[3] vmovaps 0x30(%r13,%rax), %xmm0 vmovaps %xmm0, 0x1c0(%rsp) vmovaps 0x40(%r13,%rax), %xmm0 vmovaps %xmm0, 0x1b0(%rsp) vbroadcastss (%rsi), %xmm9 vbroadcastss 0x4(%rsi), %xmm13 vbroadcastss 0x8(%rsi), %xmm14 vsubps %xmm9, %xmm5, %xmm7 vsubps %xmm13, %xmm6, %xmm3 vmovaps 0x10(%rsp), %xmm0 vsubps %xmm14, %xmm0, %xmm1 vsubps %xmm9, %xmm10, %xmm11 vsubps %xmm13, %xmm15, %xmm0 vmovaps %xmm0, 0x30(%rsp) vsubps %xmm14, %xmm2, %xmm0 vmovaps %xmm0, 0x40(%rsp) vsubps %xmm9, %xmm12, %xmm12 vsubps %xmm13, %xmm8, %xmm10 vsubps %xmm14, %xmm4, %xmm2 vmovaps %xmm2, 0x10(%rsp) vsubps %xmm7, %xmm12, %xmm14 vsubps %xmm3, %xmm10, %xmm15 vsubps %xmm1, %xmm2, %xmm0 vaddps %xmm3, %xmm10, %xmm4 vaddps %xmm1, %xmm2, %xmm5 vmulps %xmm4, %xmm0, %xmm8 vmulps %xmm5, %xmm15, %xmm9 vsubps %xmm8, %xmm9, %xmm13 vaddps %xmm7, %xmm12, %xmm8 vmulps %xmm5, %xmm14, %xmm5 vmovaps %xmm0, 0x150(%rsp) vmulps %xmm0, %xmm8, %xmm9 vsubps %xmm5, %xmm9, %xmm5 vmovaps %xmm15, 0x160(%rsp) vmulps %xmm8, %xmm15, %xmm8 vmovaps %xmm14, 0x170(%rsp) vmulps %xmm4, %xmm14, %xmm4 vsubps %xmm8, %xmm4, %xmm4 vbroadcastss 0x18(%rsi), %xmm0 vmulps %xmm4, %xmm0, %xmm4 vbroadcastss 0x14(%rsi), %xmm6 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vbroadcastss 0x10(%rsi), %xmm15 vmulps %xmm13, %xmm15, %xmm5 vaddps %xmm4, %xmm5, %xmm9 vmovaps 0x30(%rsp), %xmm2 vsubps %xmm2, %xmm3, %xmm13 vmovaps 0x40(%rsp), %xmm14 vsubps %xmm14, %xmm1, %xmm8 vmovaps %xmm3, 0x190(%rsp) vaddps %xmm2, %xmm3, %xmm4 vmovaps %xmm1, 0x180(%rsp) vaddps %xmm1, %xmm14, %xmm5 vmovaps %xmm14, %xmm1 vmulps %xmm4, %xmm8, %xmm14 vmulps %xmm5, %xmm13, %xmm3 vsubps %xmm14, %xmm3, %xmm3 vsubps %xmm11, %xmm7, %xmm14 vmulps %xmm5, %xmm14, %xmm5 vmovaps %xmm7, 0x1a0(%rsp) vaddps %xmm7, %xmm11, %xmm7 vmovaps %xmm8, 0x130(%rsp) vmulps %xmm7, %xmm8, %xmm8 vsubps %xmm5, %xmm8, %xmm5 vmovaps %xmm13, 0x140(%rsp) vmulps %xmm7, %xmm13, %xmm7 vmulps %xmm4, %xmm14, %xmm4 vsubps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm0, %xmm4 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vmulps %xmm3, %xmm15, %xmm3 vaddps %xmm4, %xmm3, %xmm4 vsubps %xmm12, %xmm11, %xmm5 vaddps %xmm12, %xmm11, %xmm3 vsubps %xmm10, %xmm2, %xmm12 vaddps %xmm2, %xmm10, %xmm2 vmovaps %xmm0, %xmm11 vmovaps 0x10(%rsp), %xmm0 vsubps %xmm0, %xmm1, %xmm8 vaddps %xmm0, %xmm1, %xmm0 vmulps %xmm2, %xmm8, %xmm1 vmulps %xmm0, %xmm12, %xmm7 vsubps %xmm1, %xmm7, %xmm1 vmulps %xmm0, %xmm5, %xmm0 vmulps %xmm3, %xmm8, %xmm7 vsubps %xmm0, %xmm7, %xmm0 vmulps %xmm3, %xmm12, %xmm3 vmulps %xmm2, %xmm5, %xmm2 vsubps %xmm3, %xmm2, %xmm2 vmovaps %xmm11, 0x30(%rsp) vmulps %xmm2, %xmm11, %xmm2 vmovaps %xmm6, 0x10(%rsp) vmulps %xmm0, %xmm6, %xmm0 vaddps %xmm2, %xmm0, %xmm0 vmovaps %xmm15, 0x40(%rsp) vmulps %xmm1, %xmm15, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm4, %xmm9, %xmm1 vaddps %xmm1, %xmm0, %xmm11 vminps %xmm4, %xmm9, %xmm1 vminps %xmm0, %xmm1, %xmm1 vbroadcastss 0x1d112a3(%rip), %xmm2 # 0x1f20ec4 vandps %xmm2, %xmm11, %xmm15 vbroadcastss 0x1d1129e(%rip), %xmm2 # 0x1f20ecc vmulps %xmm2, %xmm15, %xmm2 vbroadcastss 0x1d11285(%rip), %xmm3 # 0x1f20ec0 vxorps %xmm3, %xmm2, %xmm3 vcmpnltps %xmm3, %xmm1, %xmm1 vmovaps %xmm9, 0x120(%rsp) vmaxps %xmm4, %xmm9, %xmm3 vmaxps %xmm0, %xmm3, %xmm0 vcmpleps %xmm2, %xmm0, %xmm0 vorps %xmm0, %xmm1, %xmm0 movb $0x0, 0xf(%rsp) leaq 0xf(%rsp), %rax movq %rax, 0x200(%rsp) vtestps 0x100(%rsp), %xmm0 je 0x2100ac vmovaps 0x150(%rsp), %xmm6 vmovaps 0x140(%rsp), %xmm10 vmovaps %xmm0, 0x110(%rsp) vmulps %xmm6, %xmm10, %xmm0 vmovaps 0x160(%rsp), %xmm9 vmovaps 0x130(%rsp), %xmm13 vmulps %xmm13, %xmm9, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm12, %xmm13, %xmm2 vmulps %xmm8, %xmm10, %xmm3 vsubps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d111f3(%rip), %xmm7 # 0x1f20ec4 vandps %xmm7, %xmm0, %xmm0 vandps %xmm7, %xmm2, %xmm2 vcmpltps %xmm2, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm3, %xmm0 vmulps %xmm8, %xmm14, %xmm1 vmulps %xmm6, %xmm14, %xmm2 vmovaps 0x170(%rsp), %xmm8 vmulps %xmm13, %xmm8, %xmm3 vsubps %xmm3, %xmm2, %xmm2 vmulps %xmm5, %xmm13, %xmm6 vsubps %xmm1, %xmm6, %xmm6 vandps %xmm7, %xmm3, %xmm3 vandps %xmm7, %xmm1, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm6, %xmm1 vmulps %xmm5, %xmm10, %xmm2 vmulps %xmm10, %xmm8, %xmm3 vmulps %xmm14, %xmm9, %xmm5 vmulps %xmm12, %xmm14, %xmm6 vsubps %xmm5, %xmm3, %xmm3 vsubps %xmm2, %xmm6, %xmm6 vandps %xmm7, %xmm5, %xmm5 vandps %xmm7, %xmm2, %xmm2 vcmpltps %xmm2, %xmm5, %xmm2 vblendvps %xmm2, %xmm3, %xmm6, %xmm2 vmulps 0x30(%rsp), %xmm2, %xmm3 vmulps 0x10(%rsp), %xmm1, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vmulps 0x40(%rsp), %xmm0, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vaddps %xmm3, %xmm3, %xmm5 vmulps 0x180(%rsp), %xmm2, %xmm3 vmulps 0x190(%rsp), %xmm1, %xmm6 vaddps %xmm3, %xmm6, %xmm3 vmulps 0x1a0(%rsp), %xmm0, %xmm6 vaddps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm3, %xmm3 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm7 vbroadcastss 0x1cdc976(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm6, %xmm7 vaddps %xmm7, %xmm6, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0xc(%rsi), %xmm6 vcmpleps %xmm3, %xmm6, %xmm6 vbroadcastss 0x20(%rsi), %xmm7 vcmpleps %xmm7, %xmm3, %xmm7 vandps %xmm7, %xmm6, %xmm6 vcmpneqps 0x1cdbc3f(%rip), %xmm5, %xmm5 # 0x1eeba10 vandps %xmm6, %xmm5, %xmm6 vmovaps 0x110(%rsp), %xmm5 vandps 0x100(%rsp), %xmm5, %xmm5 vpslld $0x1f, %xmm6, %xmm6 vpsrad $0x1f, %xmm6, %xmm6 vtestps %xmm5, %xmm6 je 0x2100ac vandps %xmm5, %xmm6, %xmm5 vmovaps 0x120(%rsp), %xmm6 vmovaps %xmm6, 0x1d0(%rsp) vmovaps %xmm4, 0x1e0(%rsp) vmovaps %xmm11, 0x1f0(%rsp) movq %rax, 0x200(%rsp) vmovaps %xmm5, 0x210(%rsp) vmovaps %xmm3, 0x240(%rsp) vmovaps %xmm0, 0x250(%rsp) vmovaps %xmm1, 0x260(%rsp) vmovaps %xmm2, 0x270(%rsp) vrcpps %xmm11, %xmm0 vmulps %xmm0, %xmm11, %xmm1 vbroadcastss 0x1cdc8a9(%rip), %xmm2 # 0x1eec714 vsubps %xmm1, %xmm2, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1ce1168(%rip), %xmm1 # 0x1ef0fe8 vcmpnltps %xmm1, %xmm15, %xmm1 vandps %xmm0, %xmm1, %xmm0 vmulps %xmm0, %xmm6, %xmm1 vminps %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x220(%rsp) vmulps %xmm0, %xmm4, %xmm0 vminps %xmm2, %xmm0, %xmm0 vmovaps %xmm0, 0x230(%rsp) vmovmskps %xmm5, %r15d bsfq %r15, %r14 movl 0x1c0(%rsp,%r14,4), %eax movq 0x1e8(%rdi), %rcx movq (%rcx,%rax,8), %rbx movl 0x24(%rsi), %ecx testl %ecx, 0x34(%rbx) je 0x20fee8 movq 0x10(%r8), %r12 cmpq $0x0, 0x10(%r12) jne 0x20ff00 cmpq $0x0, 0x48(%rbx) jne 0x20ff00 xorl %eax, %eax xorl %r12d, %r12d jmp 0x20feee btcq %r14, %r15 movb $0x1, %al testb %al, %al je 0x2100ec testq %r15, %r15 jne 0x20feaf jmp 0x2100ac vmovss 0x220(%rsp,%r14,4), %xmm0 vmovss 0x230(%rsp,%r14,4), %xmm1 movq 0x8(%r8), %rcx movl 0x1b0(%rsp,%r14,4), %edx vmovss 0x250(%rsp,%r14,4), %xmm2 vmovss 0x260(%rsp,%r14,4), %xmm3 vmovss 0x270(%rsp,%r14,4), %xmm4 vmovss %xmm2, 0xa0(%rsp) vmovss %xmm3, 0xa4(%rsp) vmovss %xmm4, 0xa8(%rsp) vmovss %xmm0, 0xac(%rsp) vmovss %xmm1, 0xb0(%rsp) movl %edx, 0xb4(%rsp) movl %eax, 0xb8(%rsp) movl (%rcx), %eax movl %eax, 0xbc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xc0(%rsp) vmovss 0x20(%rsi), %xmm0 vmovss %xmm0, 0x10(%rsp) vmovss 0x240(%rsp,%r14,4), %xmm0 vmovss %xmm0, 0x20(%rsi) orl $-0x1, 0x2c(%rsp) leaq 0x2c(%rsp), %rax movq %rax, 0xd0(%rsp) movq 0x18(%rbx), %rax movq %rax, 0xd8(%rsp) movq %rcx, 0xe0(%rsp) movq %rsi, 0xe8(%rsp) leaq 0xa0(%rsp), %rax movq %rax, 0xf0(%rsp) movl $0x1, 0xf8(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax movq %r8, 0x60(%rsp) movq %r10, 0x58(%rsp) je 0x21003f leaq 0xd0(%rsp), %rdi movq %rsi, 0x40(%rsp) movq %r11, 0x30(%rsp) vzeroupper callq %rax movq 0x20(%rsp), %rdi movq 0x30(%rsp), %r11 movq 0x58(%rsp), %r10 movq 0x40(%rsp), %rsi movq 0x60(%rsp), %r8 movq 0xd0(%rsp), %rax cmpl $0x0, (%rax) je 0x21008f movq 0x10(%r12), %rax testq %rax, %rax je 0x21008b testb $0x2, (%r12) jne 0x210056 testb $0x40, 0x3e(%rbx) je 0x21007e leaq 0xd0(%rsp), %rdi movq %rsi, %r12 movq %r11, %rbx vzeroupper callq %rax movq 0x20(%rsp), %rdi movq %rbx, %r11 movq 0x58(%rsp), %r10 movq %r12, %rsi movq 0x60(%rsp), %r8 movq 0xd0(%rsp), %rax cmpl $0x0, (%rax) je 0x21008f xorl %eax, %eax jmp 0x2100a0 vmovss 0x10(%rsp), %xmm0 vmovss %xmm0, 0x20(%rsi) btcq %r14, %r15 movb $0x1, %al xorl %r12d, %r12d movb 0xe(%rsp), %dl jmp 0x20feee movq 0x98(%rsp), %rdx incq %rdx movq 0x90(%rsp), %rax cmpq %rax, %rdx setb %cl jne 0x20f8ef movq 0x88(%rsp), %rdi movq 0x80(%rsp), %r9 movq 0x78(%rsp), %rbx movq 0x70(%rsp), %r14 movq 0x68(%rsp), %r15 jmp 0x21011b testb $0x1, %dl movq 0x88(%rsp), %rdi movq 0x80(%rsp), %r9 movq 0x78(%rsp), %rbx movq 0x70(%rsp), %r14 movq 0x68(%rsp), %r15 je 0x21011b movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %r12 cmpl $0x3, %r12d jne 0x20f5fb addq $0x1528, %rsp # imm = 0x1528 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1048576, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMIntersector1Moeller<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1498, %rsp # imm = 0x1498 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x21016a addq $0x1498, %rsp # imm = 0x1498 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %rsi, %r14 vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x210155 leaq 0x2f8(%rsp), %rdi vmovaps 0x10(%r14), %xmm3 vmaxss 0xc(%r14), %xmm1, %xmm2 vbroadcastss 0x1d10d2b(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce0e42(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 movq 0x70(%rax), %rax vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1cdc54e(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss (%r14), %ymm9 vbroadcastss 0x4(%r14), %ymm10 vbroadcastss 0x8(%r14), %ymm11 vaddps %xmm3, %xmm4, %xmm3 xorl %r8d, %r8d vucomiss %xmm1, %xmm3 movq %rax, -0x8(%rdi) setb %r8b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm12 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm13 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] xorl %eax, %eax vucomiss %xmm1, %xmm4 setb %al xorl %ecx, %ecx vucomiss %xmm1, %xmm5 vinsertf128 $0x1, %xmm3, %ymm3, %ymm14 setb %cl vshufps $0x0, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm15 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovups %ymm0, 0x2d0(%rsp) shll $0x3, %r8d movq %r8, %r9 xorq $0x8, %r9 leal 0x10(,%rax,8), %r10d movq %r10, %r11 xorq $0x8, %r11 leal 0x20(,%rcx,8), %r15d movq %r15, %r12 xorq $0x8, %r12 leaq 0x1f3fd03(%rip), %rax # 0x214ff80 vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x120(%rsp) leaq 0x2f0(%rsp), %r13 vmovups %ymm9, 0x2b0(%rsp) vmovups %ymm10, 0x290(%rsp) vmovups %ymm11, 0x270(%rsp) vmovups %ymm12, 0x250(%rsp) vmovups %ymm13, 0x230(%rsp) vmovups %ymm14, 0x210(%rsp) vmovups %ymm15, 0x1f0(%rsp) movq %r15, 0x58(%rsp) movq %r12, 0x50(%rsp) cmpq %r13, %rdi je 0x210155 movq -0x8(%rdi), %rbp addq $-0x8, %rdi testb $0x8, %bpl jne 0x2104d8 movq %rbp, %rcx andq $-0x10, %rcx leaq 0x40(%rcx), %rax testq %rcx, %rcx cmoveq %rcx, %rax vmovq (%rax), %xmm0 vmovq 0x4(%rax), %xmm1 vmovq 0x8(%rax), %xmm2 vmovq 0xc(%rax), %xmm3 vpminub %xmm2, %xmm0, %xmm2 vpcmpeqb %xmm2, %xmm0, %xmm0 vpminub %xmm3, %xmm1, %xmm2 vpcmpeqb %xmm2, %xmm1, %xmm1 vbroadcastss 0x30(%rax), %ymm3 vbroadcastss 0x3c(%rax), %ymm4 vmovq (%rax,%r8), %xmm2 vpmovzxbd %xmm2, %xmm2 # xmm2 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero,xmm2[2],zero,zero,zero,xmm2[3],zero,zero,zero vmovq 0x4(%rax,%r8), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vinsertf128 $0x1, %xmm5, %ymm2, %ymm2 vcvtdq2ps %ymm2, %ymm2 vmulps %ymm2, %ymm4, %ymm2 vaddps %ymm2, %ymm3, %ymm2 vmovq (%rax,%r9), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vmovq 0x4(%rax,%r9), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vinsertf128 $0x1, %xmm6, %ymm5, %ymm5 vcvtdq2ps %ymm5, %ymm5 vmulps %ymm5, %ymm4, %ymm4 vaddps %ymm4, %ymm3, %ymm3 vbroadcastss 0x34(%rax), %ymm5 vbroadcastss 0x40(%rax), %ymm6 vmovq (%rax,%r10), %xmm4 vpmovzxbd %xmm4, %xmm4 # xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero,xmm4[2],zero,zero,zero,xmm4[3],zero,zero,zero vmovq 0x4(%rax,%r10), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vinsertf128 $0x1, %xmm7, %ymm4, %ymm4 vcvtdq2ps %ymm4, %ymm4 vmulps %ymm4, %ymm6, %ymm4 vaddps %ymm4, %ymm5, %ymm4 vmovq (%rax,%r11), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x4(%rax,%r11), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm6, %ymm6 vaddps %ymm6, %ymm5, %ymm5 vmovq (%rax,%r15), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vmovq 0x4(%rax,%r15), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vinsertf128 $0x1, %xmm7, %ymm6, %ymm6 vmovq (%rax,%r12), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x4(%rax,%r12), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vbroadcastss 0x44(%rax), %ymm8 vcvtdq2ps %ymm6, %ymm6 vmulps %ymm6, %ymm8, %ymm6 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm8, %ymm7 vbroadcastss 0x38(%rax), %ymm8 vaddps %ymm6, %ymm8, %ymm6 vaddps %ymm7, %ymm8, %ymm7 vsubps %ymm9, %ymm2, %ymm2 vmulps %ymm2, %ymm12, %ymm2 vsubps %ymm10, %ymm4, %ymm4 vmulps %ymm4, %ymm13, %ymm4 vmaxps %ymm4, %ymm2, %ymm2 vsubps %ymm11, %ymm6, %ymm4 vmulps %ymm4, %ymm14, %ymm4 vmaxps %ymm15, %ymm4, %ymm4 vmaxps %ymm4, %ymm2, %ymm2 vsubps %ymm9, %ymm3, %ymm3 vmulps %ymm3, %ymm12, %ymm3 vsubps %ymm10, %ymm5, %ymm4 vmulps %ymm4, %ymm13, %ymm4 vminps %ymm4, %ymm3, %ymm3 vsubps %ymm11, %ymm7, %ymm4 vmulps %ymm4, %ymm14, %ymm4 vminps 0x2d0(%rsp), %ymm4, %ymm4 vminps %ymm4, %ymm3, %ymm3 vcmpleps %ymm3, %ymm2, %ymm2 vpcmpeqd %xmm3, %xmm3, %xmm3 vpxor %xmm3, %xmm0, %xmm0 vpmovsxbd %xmm0, %xmm0 vpxor %xmm3, %xmm1, %xmm1 vpmovsxbd %xmm1, %xmm1 vinsertf128 $0x1, %xmm1, %ymm0, %ymm0 vcvtdq2ps %ymm0, %ymm0 vbroadcastss 0x1ce0501(%rip), %ymm1 # 0x1ef09cc vcmpltps %ymm0, %ymm1, %ymm0 vandps %ymm0, %ymm2, %ymm0 vmovmskps %ymm0, %ebx testb $0x8, %bpl jne 0x210516 testq %rbx, %rbx je 0x21051a andq $-0x10, %rbp bsfq %rbx, %rax leaq -0x1(%rbx), %rsi xorl %ecx, %ecx movq (%rbp,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) andq %rbx, %rsi jne 0x21051f movq %rax, %rbp testl %ecx, %ecx je 0x2102f0 jmp 0x21055e pushq $0x6 jmp 0x21051c pushq $0x4 popq %rcx jmp 0x21050c movq %rax, (%rdi) addq $0x8, %rdi bsfq %rsi, %rcx leaq -0x1(%rsi), %rax movq (%rbp,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) andq %rsi, %rax je 0x210557 movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rsi jmp 0x21052e movq %rcx, %rbp xorl %ecx, %ecx jmp 0x21050c cmpl $0x6, %ecx jne 0x210b07 movl %ebp, %esi andl $0xf, %esi xorl %ecx, %ecx addq $-0x8, %rsi setne %r13b je 0x210aff andq $-0x10, %rbp xorl %r12d, %r12d imulq $0xb0, %r12, %r15 vmovaps 0x80(%rbp,%r15), %xmm9 vmovaps 0x40(%rbp,%r15), %xmm6 vmulps %xmm6, %xmm9, %xmm0 vmovaps 0x70(%rbp,%r15), %xmm10 vmovaps 0x50(%rbp,%r15), %xmm7 vmulps %xmm7, %xmm10, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, (%rsp) vmovaps 0x60(%rbp,%r15), %xmm11 vmulps %xmm7, %xmm11, %xmm1 vmovaps (%rbp,%r15), %xmm3 vmovaps 0x10(%rbp,%r15), %xmm13 vmovaps 0x20(%rbp,%r15), %xmm0 vmovaps 0x30(%rbp,%r15), %xmm8 vmulps %xmm8, %xmm9, %xmm2 vsubps %xmm1, %xmm2, %xmm5 vmulps %xmm8, %xmm10, %xmm2 vmulps %xmm6, %xmm11, %xmm12 vsubps %xmm2, %xmm12, %xmm4 vbroadcastss (%r14), %xmm12 vsubps %xmm12, %xmm3, %xmm2 vbroadcastss 0x4(%r14), %xmm12 vsubps %xmm12, %xmm13, %xmm3 vbroadcastss 0x8(%r14), %xmm12 vsubps %xmm12, %xmm0, %xmm1 vbroadcastss 0x14(%r14), %xmm12 vbroadcastss 0x18(%r14), %xmm13 vmulps %xmm1, %xmm12, %xmm14 vmulps %xmm3, %xmm13, %xmm15 vsubps %xmm14, %xmm15, %xmm14 vbroadcastss 0x10(%r14), %xmm15 vmulps %xmm2, %xmm13, %xmm0 vmovaps %xmm1, 0x60(%rsp) vmulps %xmm1, %xmm15, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm3, 0x40(%rsp) vmulps %xmm3, %xmm15, %xmm1 vmovaps %xmm2, 0x70(%rsp) vmulps %xmm2, %xmm12, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm4, 0x20(%rsp) vmulps %xmm4, %xmm13, %xmm2 vmovaps (%rsp), %xmm13 vmovaps %xmm5, 0x30(%rsp) vmulps %xmm5, %xmm12, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm15, %xmm13, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm1, %xmm9, %xmm9 vmulps %xmm0, %xmm10, %xmm10 vaddps %xmm10, %xmm9, %xmm9 vmulps %xmm14, %xmm11, %xmm10 vaddps %xmm9, %xmm10, %xmm10 vmulps %xmm1, %xmm7, %xmm1 vmulps %xmm0, %xmm6, %xmm0 vaddps %xmm0, %xmm1, %xmm0 vmovddup 0x1d108b7(%rip), %xmm1 # xmm1 = mem[0,0] vandps %xmm1, %xmm2, %xmm9 vxorps %xmm10, %xmm9, %xmm7 vmulps %xmm14, %xmm8, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm8 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %xmm10, %xmm7, %xmm0 vcmpnltps %xmm10, %xmm8, %xmm1 vandps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1d107df(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm2, %xmm6 vcmpneqps %xmm2, %xmm10, %xmm1 vandps %xmm1, %xmm0, %xmm0 vaddps %xmm7, %xmm8, %xmm1 vcmpleps %xmm6, %xmm1, %xmm1 vandps %xmm1, %xmm0, %xmm10 vtestps 0x120(%rsp), %xmm10 jne 0x210720 incq %r12 cmpq %rsi, %r12 setb %r13b jne 0x210583 jmp 0x210ab6 vandps 0x120(%rsp), %xmm10, %xmm10 vmovaps 0x60(%rsp), %xmm0 vmulps 0x20(%rsp), %xmm0, %xmm0 vmovaps 0x40(%rsp), %xmm1 vmulps 0x30(%rsp), %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x70(%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm3 vbroadcastss 0xc(%r14), %xmm0 vmulps %xmm0, %xmm6, %xmm0 vcmpltps %xmm3, %xmm0, %xmm0 vbroadcastss 0x20(%r14), %xmm1 vmulps %xmm1, %xmm6, %xmm1 vcmpleps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm4 vtestps %xmm10, %xmm4 je 0x21070b addq %rbp, %r15 vandps %xmm4, %xmm10, %xmm0 vmovaps %xmm7, 0x130(%rsp) vmovaps %xmm8, 0x140(%rsp) vmovaps %xmm3, 0x150(%rsp) vmovaps %xmm6, 0x160(%rsp) vmovaps %xmm0, 0x180(%rsp) vmovaps %xmm13, 0x1c0(%rsp) vmovaps 0x30(%rsp), %xmm1 vmovaps %xmm1, 0x1d0(%rsp) vmovaps 0x20(%rsp), %xmm1 vmovaps %xmm1, 0x1e0(%rsp) movq (%rdx), %rax movq %rax, 0x70(%rsp) vrcpps %xmm6, %xmm1 vmulps %xmm1, %xmm6, %xmm2 vbroadcastss 0x1cdbf24(%rip), %xmm3 # 0x1eec714 vsubps %xmm2, %xmm3, %xmm2 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps 0x150(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x1b0(%rsp) vmulps 0x130(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x190(%rsp) vmulps 0x140(%rsp), %xmm1, %xmm1 vmovaps %xmm1, 0x1a0(%rsp) vmovmskps %xmm0, %eax movq %rax, (%rsp) bsfq %rax, %rcx movq %rcx, 0x20(%rsp) movl 0x90(%r15,%rcx,4), %eax movq 0x70(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq %rax, 0x40(%rsp) movq (%rcx,%rax,8), %rax movl 0x24(%r14), %ecx movq %rax, 0x30(%rsp) testl %ecx, 0x34(%rax) je 0x210893 movq 0x10(%rdx), %rax movq %rax, 0x60(%rsp) cmpq $0x0, 0x10(%rax) movl $0x0, %ecx jne 0x2108c2 movq 0x30(%rsp), %rax cmpq $0x0, 0x48(%rax) jne 0x2108c2 xorl %eax, %eax jmp 0x2108a8 movq 0x20(%rsp), %rax movq (%rsp), %rcx btcq %rax, %rcx movq %rcx, (%rsp) movb $0x1, %al xorl %ecx, %ecx testb %al, %al je 0x210b15 movq (%rsp), %rax testq %rax, %rax jne 0x210836 jmp 0x21070b movq %rsi, 0x88(%rsp) movq %r11, 0x90(%rsp) movq %r10, 0x98(%rsp) movq %r9, 0xa0(%rsp) movq %r8, 0xa8(%rsp) movq %rdi, 0xb0(%rsp) movq 0x20(%rsp), %rsi vmovss 0x190(%rsp,%rsi,4), %xmm0 vmovss 0x1a0(%rsp,%rsi,4), %xmm1 movq %rdx, 0xb8(%rsp) movq 0x8(%rdx), %rcx movl 0xa0(%r15,%rsi,4), %edx vmovss 0x1c0(%rsp,%rsi,4), %xmm2 vmovss 0x1d0(%rsp,%rsi,4), %xmm3 vmovss 0x1e0(%rsp,%rsi,4), %xmm4 vmovss %xmm2, 0xc0(%rsp) vmovss %xmm3, 0xc4(%rsp) vmovss %xmm4, 0xc8(%rsp) vmovss %xmm0, 0xcc(%rsp) vmovss %xmm1, 0xd0(%rsp) movl %edx, 0xd4(%rsp) movq 0x40(%rsp), %rax movl %eax, 0xd8(%rsp) movl (%rcx), %eax movl %eax, 0xdc(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xe0(%rsp) vmovss 0x20(%r14), %xmm0 vmovss %xmm0, 0x40(%rsp) vmovss 0x1b0(%rsp,%rsi,4), %xmm0 vmovss %xmm0, 0x20(%r14) orl $-0x1, 0x1c(%rsp) leaq 0x1c(%rsp), %rax movq %rax, 0xf0(%rsp) movq 0x30(%rsp), %rdx movq 0x18(%rdx), %rax movq %rax, 0xf8(%rsp) movq %rcx, 0x100(%rsp) movq %r14, 0x108(%rsp) leaq 0xc0(%rsp), %rax movq %rax, 0x110(%rsp) movl $0x1, 0x118(%rsp) movq 0x48(%rdx), %rax testq %rax, %rax je 0x210a17 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x210a58 movq 0x60(%rsp), %rax movq 0x10(%rax), %rax testq %rax, %rax je 0x210a54 movq 0x60(%rsp), %rcx testb $0x2, (%rcx) jne 0x210a3a movq 0x30(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x210a47 leaq 0xf0(%rsp), %rdi vzeroupper callq %rax movq 0xf0(%rsp), %rax cmpl $0x0, (%rax) je 0x210a58 xorl %eax, %eax jmp 0x210a77 vmovss 0x40(%rsp), %xmm0 vmovss %xmm0, 0x20(%r14) movq (%rsp), %rax movq 0x20(%rsp), %rcx btcq %rcx, %rax movq %rax, (%rsp) movb $0x1, %al movq 0xb8(%rsp), %rdx movq 0xb0(%rsp), %rdi movq 0xa8(%rsp), %r8 movq 0xa0(%rsp), %r9 movq 0x98(%rsp), %r10 movq 0x90(%rsp), %r11 xorl %ecx, %ecx movq 0x88(%rsp), %rsi jmp 0x2108a8 vmovups 0x2b0(%rsp), %ymm9 vmovups 0x290(%rsp), %ymm10 vmovups 0x270(%rsp), %ymm11 vmovups 0x250(%rsp), %ymm12 vmovups 0x230(%rsp), %ymm13 vmovups 0x210(%rsp), %ymm14 vmovups 0x1f0(%rsp), %ymm15 movq 0x58(%rsp), %r15 movq 0x50(%rsp), %r12 leaq 0x2f0(%rsp), %r13 cmpl $0x3, %ecx jne 0x2102df jmp 0x210155 testb $0x1, %r13b vmovups 0x2b0(%rsp), %ymm9 vmovups 0x290(%rsp), %ymm10 vmovups 0x270(%rsp), %ymm11 vmovups 0x250(%rsp), %ymm12 vmovups 0x230(%rsp), %ymm13 vmovups 0x210(%rsp), %ymm14 vmovups 0x1f0(%rsp), %ymm15 movq 0x58(%rsp), %r15 movq 0x50(%rsp), %r12 leaq 0x2f0(%rsp), %r13 je 0x210b07 movl $0xff800000, 0x20(%r14) # imm = 0xFF800000 pushq $0x3 popq %rcx jmp 0x210b07 nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1048576, false, embree::avx::ArrayIntersector1<embree::avx::QuadMiIntersector1Pluecker<4, true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x1760, %rsp # imm = 0x1760 movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x210bae leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %rsi, %r9 vmovss 0x20(%rsi), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm0, %xmm1 ja 0x210b9c movq %rdx, %r8 leaq 0x5a8(%rsp), %r10 vmovaps 0x10(%r9), %xmm3 vmaxss 0xc(%r9), %xmm1, %xmm2 vbroadcastss 0x1d102e4(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1ce03fb(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 movq 0x70(%rax), %rax vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1cdbb07(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss (%r9), %ymm9 vbroadcastss 0x4(%r9), %ymm10 vbroadcastss 0x8(%r9), %ymm11 vaddps %xmm3, %xmm4, %xmm3 xorl %r11d, %r11d vucomiss %xmm1, %xmm3 movq %rax, -0x8(%r10) setb %r11b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm12 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm13 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] xorl %eax, %eax vucomiss %xmm1, %xmm4 setb %al xorl %ecx, %ecx vucomiss %xmm1, %xmm5 vinsertf128 $0x1, %xmm3, %ymm3, %ymm14 setb %cl vshufps $0x0, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm15 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovaps %ymm0, 0x380(%rsp) shll $0x3, %r11d movq %r11, %rbx xorq $0x8, %rbx leal 0x10(,%rax,8), %r14d movq %r14, %r15 xorq $0x8, %r15 leal 0x20(,%rcx,8), %esi movq %rsi, %rdi xorq $0x8, %rdi leaq 0x1f3f2bd(%rip), %rax # 0x214ff80 vmovups (%rax), %ymm0 vinsertf128 $0x1, 0xf0(%rax), %ymm0, %ymm0 vmovaps %ymm0, 0x220(%rsp) vmovaps %ymm9, 0x200(%rsp) vmovaps %ymm10, 0x1e0(%rsp) vmovaps %ymm11, 0x1c0(%rsp) vmovaps %ymm12, 0x1a0(%rsp) vmovaps %ymm13, 0x180(%rsp) vmovaps %ymm14, 0x160(%rsp) vmovaps %ymm15, 0x140(%rsp) movq %rbx, 0x68(%rsp) movq %r14, 0x60(%rsp) movq %r15, 0x58(%rsp) movq %rsi, 0x28(%rsp) movq %rdi, 0x20(%rsp) leaq 0x5a0(%rsp), %rax cmpq %rax, %r10 je 0x210b9c movq -0x8(%r10), %r13 addq $-0x8, %r10 testb $0x8, %r13b jne 0x210f2d movq %r13, %rcx andq $-0x10, %rcx leaq 0x40(%rcx), %rax testq %rcx, %rcx cmoveq %rcx, %rax vmovq (%rax), %xmm0 vmovq 0x4(%rax), %xmm1 vmovq 0x8(%rax), %xmm2 vmovq 0xc(%rax), %xmm3 vpminub %xmm2, %xmm0, %xmm2 vpcmpeqb %xmm2, %xmm0, %xmm0 vpminub %xmm3, %xmm1, %xmm2 vpcmpeqb %xmm2, %xmm1, %xmm1 vbroadcastss 0x30(%rax), %ymm3 vbroadcastss 0x3c(%rax), %ymm4 vmovq (%rax,%r11), %xmm2 vpmovzxbd %xmm2, %xmm2 # xmm2 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero,xmm2[2],zero,zero,zero,xmm2[3],zero,zero,zero vmovq 0x4(%rax,%r11), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vinsertf128 $0x1, %xmm5, %ymm2, %ymm2 vcvtdq2ps %ymm2, %ymm2 vmulps %ymm2, %ymm4, %ymm2 vaddps %ymm2, %ymm3, %ymm2 vmovq (%rax,%rbx), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vmovq 0x4(%rax,%rbx), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vinsertf128 $0x1, %xmm6, %ymm5, %ymm5 vcvtdq2ps %ymm5, %ymm5 vmulps %ymm5, %ymm4, %ymm4 vaddps %ymm4, %ymm3, %ymm3 vbroadcastss 0x34(%rax), %ymm5 vbroadcastss 0x40(%rax), %ymm6 vmovq (%rax,%r14), %xmm4 vpmovzxbd %xmm4, %xmm4 # xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero,xmm4[2],zero,zero,zero,xmm4[3],zero,zero,zero vmovq 0x4(%rax,%r14), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vinsertf128 $0x1, %xmm7, %ymm4, %ymm4 vcvtdq2ps %ymm4, %ymm4 vmulps %ymm4, %ymm6, %ymm4 vaddps %ymm4, %ymm5, %ymm4 vmovq (%rax,%r15), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x4(%rax,%r15), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm6, %ymm6 vaddps %ymm6, %ymm5, %ymm5 vmovq (%rax,%rsi), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vmovq 0x4(%rax,%rsi), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vinsertf128 $0x1, %xmm7, %ymm6, %ymm6 vmovq (%rax,%rdi), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x4(%rax,%rdi), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vbroadcastss 0x44(%rax), %ymm8 vcvtdq2ps %ymm6, %ymm6 vmulps %ymm6, %ymm8, %ymm6 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm8, %ymm7 vbroadcastss 0x38(%rax), %ymm8 vaddps %ymm6, %ymm8, %ymm6 vaddps %ymm7, %ymm8, %ymm7 vsubps %ymm9, %ymm2, %ymm2 vmulps %ymm2, %ymm12, %ymm2 vsubps %ymm10, %ymm4, %ymm4 vmulps %ymm4, %ymm13, %ymm4 vmaxps %ymm4, %ymm2, %ymm2 vsubps %ymm11, %ymm6, %ymm4 vmulps %ymm4, %ymm14, %ymm4 vmaxps %ymm15, %ymm4, %ymm4 vmaxps %ymm4, %ymm2, %ymm2 vsubps %ymm9, %ymm3, %ymm3 vmulps %ymm3, %ymm12, %ymm3 vsubps %ymm10, %ymm5, %ymm4 vmulps %ymm4, %ymm13, %ymm4 vminps %ymm4, %ymm3, %ymm3 vsubps %ymm11, %ymm7, %ymm4 vmulps %ymm4, %ymm14, %ymm4 vminps 0x380(%rsp), %ymm4, %ymm4 vminps %ymm4, %ymm3, %ymm3 vcmpleps %ymm3, %ymm2, %ymm2 vpcmpeqd %xmm3, %xmm3, %xmm3 vpxor %xmm3, %xmm0, %xmm0 vpmovsxbd %xmm0, %xmm0 vpxor %xmm3, %xmm1, %xmm1 vpmovsxbd %xmm1, %xmm1 vinsertf128 $0x1, %xmm1, %ymm0, %ymm0 vcvtdq2ps %ymm0, %ymm0 vbroadcastss 0x1cdfaac(%rip), %ymm1 # 0x1ef09cc vcmpltps %ymm0, %ymm1, %ymm0 vandps %ymm0, %ymm2, %ymm0 vmovmskps %ymm0, %r12d testb $0x8, %r13b jne 0x210f6c testq %r12, %r12 je 0x210f70 andq $-0x10, %r13 bsfq %r12, %rax leaq -0x1(%r12), %rdx xorl %ecx, %ecx movq (%r13,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) andq %r12, %rdx jne 0x210f75 movq %rax, %r13 testl %ecx, %ecx je 0x210d4b jmp 0x210fb4 pushq $0x6 jmp 0x210f72 pushq $0x4 popq %rcx jmp 0x210f62 movq %rax, (%r10) addq $0x8, %r10 bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r13,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) andq %rdx, %rax je 0x210fad movq %rcx, (%r10) addq $0x8, %r10 bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x210f84 movq %rcx, %r13 xorl %ecx, %ecx jmp 0x210f62 cmpl $0x6, %ecx jne 0x211a5a movl %r13d, %eax andl $0xf, %eax xorl %ecx, %ecx addq $-0x8, %rax setne %dl je 0x211a5a movb %dl, 0x1f(%rsp) andq $-0x10, %r13 movq (%r8), %rdx xorl %ecx, %ecx movq %r12, 0x30(%rsp) movq %rax, 0x70(%rsp) movq %rdx, 0x38(%rsp) movq %rcx, 0x78(%rsp) imulq $0x60, %rcx, %rax prefetcht0 (%r13,%rax) prefetcht0 0x40(%r13,%rax) movq %rdx, %rcx movl 0x40(%r13,%rax), %edx movq 0x228(%rcx), %rcx movq (%rcx,%rdx,8), %rdx movl (%r13,%rax), %esi movl 0x4(%r13,%rax), %edi vmovups (%rdx,%rsi,4), %xmm4 movl 0x10(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm2 movl 0x20(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm1 movl 0x30(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm0 movl 0x44(%r13,%rax), %edx movq (%rcx,%rdx,8), %rdx vmovups (%rdx,%rdi,4), %xmm7 movl 0x14(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm6 movl 0x24(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm5 movl 0x34(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm3 movl 0x48(%r13,%rax), %edx movq (%rcx,%rdx,8), %rdx movl 0x8(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm10 movl 0x18(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm11 movl 0x28(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm9 movl 0x38(%r13,%rax), %esi vmovups (%rdx,%rsi,4), %xmm8 movl 0x4c(%r13,%rax), %edx movq (%rcx,%rdx,8), %rcx movl 0xc(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm12 movl 0x1c(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm13 movl 0x2c(%r13,%rax), %edx vmovups (%rcx,%rdx,4), %xmm14 movl 0x3c(%r13,%rax), %edx vunpcklps %xmm10, %xmm4, %xmm15 # xmm15 = xmm4[0],xmm10[0],xmm4[1],xmm10[1] vunpckhps %xmm10, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm10[2],xmm4[3],xmm10[3] vunpcklps %xmm12, %xmm7, %xmm10 # xmm10 = xmm7[0],xmm12[0],xmm7[1],xmm12[1] vunpckhps %xmm12, %xmm7, %xmm7 # xmm7 = xmm7[2],xmm12[2],xmm7[3],xmm12[3] vmovups (%rcx,%rdx,4), %xmm12 vunpcklps %xmm7, %xmm4, %xmm4 # xmm4 = xmm4[0],xmm7[0],xmm4[1],xmm7[1] vunpcklps %xmm10, %xmm15, %xmm7 # xmm7 = xmm15[0],xmm10[0],xmm15[1],xmm10[1] vunpckhps %xmm10, %xmm15, %xmm10 # xmm10 = xmm15[2],xmm10[2],xmm15[3],xmm10[3] vunpcklps %xmm11, %xmm2, %xmm15 # xmm15 = xmm2[0],xmm11[0],xmm2[1],xmm11[1] vunpckhps %xmm11, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm11[2],xmm2[3],xmm11[3] vunpcklps %xmm13, %xmm6, %xmm11 # xmm11 = xmm6[0],xmm13[0],xmm6[1],xmm13[1] vunpckhps %xmm13, %xmm6, %xmm6 # xmm6 = xmm6[2],xmm13[2],xmm6[3],xmm13[3] vunpcklps %xmm6, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] vunpcklps %xmm11, %xmm15, %xmm6 # xmm6 = xmm15[0],xmm11[0],xmm15[1],xmm11[1] vunpckhps %xmm11, %xmm15, %xmm11 # xmm11 = xmm15[2],xmm11[2],xmm15[3],xmm11[3] vunpcklps %xmm9, %xmm1, %xmm13 # xmm13 = xmm1[0],xmm9[0],xmm1[1],xmm9[1] vunpckhps %xmm9, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm9[2],xmm1[3],xmm9[3] vunpcklps %xmm14, %xmm5, %xmm9 # xmm9 = xmm5[0],xmm14[0],xmm5[1],xmm14[1] vunpckhps %xmm14, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm14[2],xmm5[3],xmm14[3] vunpcklps %xmm5, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm5[0],xmm1[1],xmm5[1] vunpcklps %xmm9, %xmm13, %xmm5 # xmm5 = xmm13[0],xmm9[0],xmm13[1],xmm9[1] vunpckhps %xmm9, %xmm13, %xmm9 # xmm9 = xmm13[2],xmm9[2],xmm13[3],xmm9[3] vunpcklps %xmm8, %xmm0, %xmm13 # xmm13 = xmm0[0],xmm8[0],xmm0[1],xmm8[1] vunpckhps %xmm8, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm8[2],xmm0[3],xmm8[3] vunpcklps %xmm12, %xmm3, %xmm8 # xmm8 = xmm3[0],xmm12[0],xmm3[1],xmm12[1] vunpckhps %xmm12, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm12[2],xmm3[3],xmm12[3] vunpcklps %xmm3, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1] vunpcklps %xmm8, %xmm13, %xmm3 # xmm3 = xmm13[0],xmm8[0],xmm13[1],xmm8[1] vbroadcastf128 0x40(%r13,%rax), %ymm12 # ymm12 = mem[0,1,0,1] vunpckhps %xmm8, %xmm13, %xmm8 # xmm8 = xmm13[2],xmm8[2],xmm13[3],xmm8[3] vmovaps %ymm12, 0x3c0(%rsp) vbroadcastf128 0x50(%r13,%rax), %ymm12 # ymm12 = mem[0,1,0,1] vmovaps %ymm12, 0x3a0(%rsp) vinsertf128 $0x1, %xmm5, %ymm7, %ymm5 vinsertf128 $0x1, %xmm9, %ymm10, %ymm7 vinsertf128 $0x1, %xmm1, %ymm4, %ymm15 vinsertf128 $0x1, %xmm6, %ymm6, %ymm4 vinsertf128 $0x1, %xmm11, %ymm11, %ymm14 vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vinsertf128 $0x1, %xmm3, %ymm3, %ymm9 vinsertf128 $0x1, %xmm8, %ymm8, %ymm8 vinsertf128 $0x1, %xmm0, %ymm0, %ymm10 vbroadcastss (%r9), %ymm11 vbroadcastss 0x4(%r9), %ymm12 vbroadcastss 0x8(%r9), %ymm13 vsubps %ymm11, %ymm5, %ymm1 vsubps %ymm12, %ymm7, %ymm3 vsubps %ymm13, %ymm15, %ymm15 vsubps %ymm11, %ymm4, %ymm0 vmovaps %ymm0, 0x120(%rsp) vsubps %ymm12, %ymm14, %ymm7 vsubps %ymm13, %ymm2, %ymm14 vsubps %ymm11, %ymm9, %ymm2 vsubps %ymm12, %ymm8, %ymm4 vsubps %ymm13, %ymm10, %ymm0 vmovaps %ymm0, 0x100(%rsp) vsubps %ymm1, %ymm2, %ymm11 vmovaps %ymm4, 0xe0(%rsp) vsubps %ymm3, %ymm4, %ymm12 vsubps %ymm15, %ymm0, %ymm6 vaddps %ymm3, %ymm4, %ymm4 vaddps %ymm0, %ymm15, %ymm5 vmulps %ymm4, %ymm6, %ymm8 vmulps %ymm5, %ymm12, %ymm9 vsubps %ymm8, %ymm9, %ymm8 vaddps %ymm1, %ymm2, %ymm9 vmulps %ymm5, %ymm11, %ymm5 vmovaps %ymm6, 0x2c0(%rsp) vmulps %ymm6, %ymm9, %ymm10 vsubps %ymm5, %ymm10, %ymm5 vmovaps %ymm12, 0x2e0(%rsp) vmulps %ymm9, %ymm12, %ymm9 vmovaps %ymm11, 0x300(%rsp) vmulps %ymm4, %ymm11, %ymm4 vsubps %ymm9, %ymm4, %ymm4 vbroadcastss 0x14(%r9), %ymm11 vbroadcastss 0x18(%r9), %ymm6 vmulps %ymm4, %ymm6, %ymm4 vmulps %ymm5, %ymm11, %ymm5 vbroadcastss 0x10(%r9), %ymm13 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm8, %ymm13, %ymm5 vaddps %ymm4, %ymm5, %ymm10 vsubps %ymm7, %ymm3, %ymm8 vsubps %ymm14, %ymm15, %ymm12 vmovaps %ymm3, 0x340(%rsp) vaddps %ymm7, %ymm3, %ymm4 vmovaps %ymm15, 0x320(%rsp) vaddps %ymm14, %ymm15, %ymm5 vmulps %ymm4, %ymm12, %ymm15 vmulps %ymm5, %ymm8, %ymm3 vsubps %ymm15, %ymm3, %ymm3 vmovaps 0x120(%rsp), %ymm0 vsubps %ymm0, %ymm1, %ymm15 vmulps %ymm5, %ymm15, %ymm5 vmovaps %ymm1, 0x360(%rsp) vaddps %ymm0, %ymm1, %ymm9 vmovaps %ymm12, 0x2a0(%rsp) vmulps %ymm9, %ymm12, %ymm12 vsubps %ymm5, %ymm12, %ymm5 vmovaps %ymm8, %ymm12 vmulps %ymm9, %ymm8, %ymm9 vmulps %ymm4, %ymm15, %ymm4 vsubps %ymm9, %ymm4, %ymm4 vmulps %ymm4, %ymm6, %ymm4 vmulps %ymm5, %ymm11, %ymm5 vaddps %ymm4, %ymm5, %ymm4 vmulps %ymm3, %ymm13, %ymm3 vaddps %ymm4, %ymm3, %ymm4 vmovaps %ymm4, 0x260(%rsp) vsubps %ymm2, %ymm0, %ymm5 vaddps %ymm2, %ymm0, %ymm2 vmovaps 0xe0(%rsp), %ymm0 vsubps %ymm0, %ymm7, %ymm8 vaddps %ymm0, %ymm7, %ymm1 vmovaps 0x100(%rsp), %ymm3 vsubps %ymm3, %ymm14, %ymm0 vaddps %ymm3, %ymm14, %ymm3 vmulps %ymm1, %ymm0, %ymm7 vmulps %ymm3, %ymm8, %ymm9 vsubps %ymm7, %ymm9, %ymm7 vmulps %ymm3, %ymm5, %ymm3 vmulps %ymm2, %ymm0, %ymm9 vsubps %ymm3, %ymm9, %ymm3 vmulps %ymm2, %ymm8, %ymm2 vmulps %ymm1, %ymm5, %ymm1 vsubps %ymm2, %ymm1, %ymm1 vmovaps %ymm6, 0x100(%rsp) vmulps %ymm1, %ymm6, %ymm1 vmovaps %ymm11, 0x120(%rsp) vmulps %ymm3, %ymm11, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vmovaps %ymm13, 0xe0(%rsp) vmulps %ymm7, %ymm13, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vaddps %ymm4, %ymm10, %ymm2 vaddps %ymm2, %ymm1, %ymm7 vminps %ymm4, %ymm10, %ymm2 vminps %ymm1, %ymm2, %ymm2 vbroadcastss 0x1d0fb13(%rip), %ymm3 # 0x1f20ec4 vandps %ymm3, %ymm7, %ymm9 vbroadcastss 0x1d0fb0e(%rip), %ymm3 # 0x1f20ecc vmovaps %ymm9, 0x240(%rsp) vmulps %ymm3, %ymm9, %ymm3 vbroadcastss 0x1d0faec(%rip), %ymm9 # 0x1f20ec0 vxorps %ymm3, %ymm9, %ymm9 vcmpnltps %ymm9, %ymm2, %ymm2 vmovaps %ymm10, 0x280(%rsp) vmaxps %ymm4, %ymm10, %ymm9 vmaxps %ymm1, %ymm9, %ymm1 vcmpleps %ymm3, %ymm1, %ymm1 vorps %ymm1, %ymm2, %ymm1 vtestps %ymm1, %ymm1 je 0x2119e9 vmovaps 0x2c0(%rsp), %ymm6 vmovaps %ymm12, %ymm13 vmulps %ymm6, %ymm12, %ymm2 vmovaps 0x2e0(%rsp), %ymm10 vmovaps %ymm8, %ymm11 vmovaps 0x2a0(%rsp), %ymm8 vmulps %ymm8, %ymm10, %ymm3 vsubps %ymm2, %ymm3, %ymm3 vmulps %ymm11, %ymm8, %ymm9 vmulps %ymm0, %ymm12, %ymm12 vsubps %ymm9, %ymm12, %ymm12 vbroadcastss 0x1d0fa78(%rip), %ymm4 # 0x1f20ec4 vandps %ymm4, %ymm2, %ymm2 vandps %ymm4, %ymm9, %ymm9 vcmpltps %ymm9, %ymm2, %ymm2 vblendvps %ymm2, %ymm3, %ymm12, %ymm14 vmulps %ymm0, %ymm15, %ymm2 vmulps %ymm6, %ymm15, %ymm3 vmovaps 0x300(%rsp), %ymm9 vmulps %ymm8, %ymm9, %ymm6 vsubps %ymm6, %ymm3, %ymm3 vmulps %ymm5, %ymm8, %ymm8 vsubps %ymm2, %ymm8, %ymm8 vandps %ymm4, %ymm6, %ymm6 vandps %ymm4, %ymm2, %ymm2 vcmpltps %ymm2, %ymm6, %ymm2 vblendvps %ymm2, %ymm3, %ymm8, %ymm8 vmulps %ymm5, %ymm13, %ymm2 vmulps %ymm13, %ymm9, %ymm3 vmulps %ymm15, %ymm10, %ymm5 vmulps %ymm11, %ymm15, %ymm0 vsubps %ymm5, %ymm3, %ymm3 vsubps %ymm2, %ymm0, %ymm0 vandps %ymm4, %ymm5, %ymm5 vandps %ymm4, %ymm2, %ymm2 vcmpltps %ymm2, %ymm5, %ymm2 vblendvps %ymm2, %ymm3, %ymm0, %ymm5 vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vmulps 0x100(%rsp), %ymm5, %ymm1 vmulps 0x120(%rsp), %ymm8, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vmulps 0xe0(%rsp), %ymm14, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vaddps %ymm1, %ymm1, %ymm2 vmulps 0x320(%rsp), %ymm5, %ymm1 vmulps 0x340(%rsp), %ymm8, %ymm3 vaddps %ymm1, %ymm3, %ymm1 vmulps 0x360(%rsp), %ymm14, %ymm3 vrcpps %ymm2, %ymm6 vaddps %ymm1, %ymm3, %ymm1 vaddps %ymm1, %ymm1, %ymm1 vmulps %ymm6, %ymm2, %ymm3 vbroadcastss 0x1cdb1e8(%rip), %ymm9 # 0x1eec714 vsubps %ymm3, %ymm9, %ymm3 vmulps %ymm3, %ymm6, %ymm3 vaddps %ymm3, %ymm6, %ymm3 vbroadcastss 0xc(%r9), %ymm6 vmulps %ymm3, %ymm1, %ymm1 vcmpleps %ymm1, %ymm6, %ymm3 vbroadcastss 0x20(%r9), %ymm6 vcmpleps %ymm6, %ymm1, %ymm6 vandps %ymm3, %ymm6, %ymm3 vcmpneqps 0x1d0f9a1(%rip), %ymm2, %ymm2 # 0x1f20f00 vandps %ymm2, %ymm3, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vpackssdw %xmm3, %xmm2, %xmm2 vpand %xmm0, %xmm2, %xmm0 vpmovsxwd %xmm0, %xmm2 vpshufd $0xee, %xmm0, %xmm3 # xmm3 = xmm0[2,3,2,3] vpmovsxwd %xmm3, %xmm3 vinsertf128 $0x1, %xmm3, %ymm2, %ymm2 vtestps %ymm2, %ymm2 je 0x2119e9 vmovaps 0x280(%rsp), %ymm3 vmovaps %ymm3, 0x3e0(%rsp) vmovaps 0x260(%rsp), %ymm4 vmovaps %ymm4, 0x400(%rsp) vmovaps %ymm7, 0x420(%rsp) vmovaps %ymm14, 0x440(%rsp) vmovaps %ymm8, 0x460(%rsp) vmovaps %ymm5, 0x480(%rsp) vmovaps %ymm2, 0x4a0(%rsp) vrcpps %ymm7, %ymm2 vmovaps %ymm1, 0x500(%rsp) vmovaps 0x220(%rsp), %ymm9 vmovaps %ymm9, 0x580(%rsp) vmulps %ymm2, %ymm7, %ymm1 vbroadcastss 0x1cdb106(%rip), %ymm7 # 0x1eec714 vsubps %ymm1, %ymm7, %ymm1 vmulps %ymm1, %ymm2, %ymm1 vaddps %ymm1, %ymm2, %ymm1 vbroadcastss 0x1cdf9c5(%rip), %ymm2 # 0x1ef0fe8 vmovaps 0x240(%rsp), %ymm6 vcmpnltps %ymm2, %ymm6, %ymm2 vandps %ymm1, %ymm2, %ymm1 vmulps %ymm1, %ymm3, %ymm2 vminps %ymm7, %ymm2, %ymm2 vmulps %ymm1, %ymm4, %ymm1 vminps %ymm7, %ymm1, %ymm1 vsubps %ymm2, %ymm7, %ymm3 vsubps %ymm1, %ymm7, %ymm4 vbroadcastss 0x1cdf376(%rip), %ymm6 # 0x1ef09cc vblendvps %ymm9, %ymm6, %ymm7, %ymm6 vblendvps %ymm9, %ymm3, %ymm1, %ymm1 vblendvps %ymm9, %ymm4, %ymm2, %ymm2 vmovaps %ymm1, 0x4e0(%rsp) vmovaps %ymm2, 0x4c0(%rsp) vmulps %ymm6, %ymm14, %ymm1 vmulps %ymm6, %ymm8, %ymm2 vmulps %ymm5, %ymm6, %ymm3 vmovaps %ymm1, 0x520(%rsp) vmovaps %ymm2, 0x540(%rsp) vmovaps %ymm3, 0x560(%rsp) vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r15d vmovaps 0x200(%rsp), %ymm9 vmovaps 0x1e0(%rsp), %ymm10 vmovaps 0x1c0(%rsp), %ymm11 vmovaps 0x1a0(%rsp), %ymm12 vmovaps 0x180(%rsp), %ymm13 vmovaps 0x160(%rsp), %ymm14 vmovaps 0x140(%rsp), %ymm15 movq 0x28(%rsp), %rsi movq 0x20(%rsp), %rdi movq 0x38(%rsp), %rdx bsfq %r15, %r14 movl 0x3c0(%rsp,%r14,4), %eax movq 0x1e8(%rdx), %rcx movq (%rcx,%rax,8), %r12 movl 0x24(%r9), %ecx testl %ecx, 0x34(%r12) je 0x211743 movq 0x10(%r8), %rbx cmpq $0x0, 0x10(%rbx) movl $0x0, %ecx jne 0x211762 cmpq $0x0, 0x48(%r12) jne 0x211762 xorl %eax, %eax movq 0x30(%rsp), %r12 jmp 0x211750 btcq %r14, %r15 movb $0x1, %al movq 0x30(%rsp), %r12 xorl %ecx, %ecx testb %al, %al je 0x211a39 testq %r15, %r15 jne 0x211700 jmp 0x2119bd vmovss 0x4c0(%rsp,%r14,4), %xmm0 vmovss 0x4e0(%rsp,%r14,4), %xmm1 movq 0x8(%r8), %rcx movl 0x3a0(%rsp,%r14,4), %edx vmovss 0x520(%rsp,%r14,4), %xmm2 vmovss 0x540(%rsp,%r14,4), %xmm3 vmovss 0x560(%rsp,%r14,4), %xmm4 vmovss %xmm2, 0x80(%rsp) vmovss %xmm3, 0x84(%rsp) vmovss %xmm4, 0x88(%rsp) vmovss %xmm0, 0x8c(%rsp) vmovss %xmm1, 0x90(%rsp) movl %edx, 0x94(%rsp) movl %eax, 0x98(%rsp) movl (%rcx), %eax movl %eax, 0x9c(%rsp) movl 0x4(%rcx), %eax movl %eax, 0xa0(%rsp) vmovss 0x20(%r9), %xmm0 vmovss %xmm0, 0x120(%rsp) vmovd 0x500(%rsp,%r14,4), %xmm0 vmovd %xmm0, 0x20(%r9) orl $-0x1, 0x44(%rsp) leaq 0x44(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%r12), %rax movq %rax, 0xb8(%rsp) movq %rcx, 0xc0(%rsp) movq %r9, 0xc8(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x1, 0xd8(%rsp) movq 0x48(%r12), %rax testq %rax, %rax movq %r10, 0x50(%rsp) movq %r11, 0x48(%rsp) je 0x211900 leaq 0xb0(%rsp), %rdi movq %r8, 0x100(%rsp) movq %r9, 0xe0(%rsp) vzeroupper callq %rax movq 0x20(%rsp), %rdi movq 0x28(%rsp), %rsi vmovaps 0x140(%rsp), %ymm15 vmovaps 0x160(%rsp), %ymm14 vmovaps 0x180(%rsp), %ymm13 vmovaps 0x1a0(%rsp), %ymm12 movq 0x48(%rsp), %r11 vmovaps 0x1c0(%rsp), %ymm11 vmovaps 0x1e0(%rsp), %ymm10 vmovaps 0x200(%rsp), %ymm9 movq 0x50(%rsp), %r10 movq 0xe0(%rsp), %r9 movq 0x100(%rsp), %r8 movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x211997 movq 0x10(%rbx), %rax testq %rax, %rax je 0x211993 testb $0x2, (%rbx) jne 0x21191a testb $0x40, 0x3e(%r12) je 0x211986 leaq 0xb0(%rsp), %rdi movq %r8, %rbx movq %r9, %r12 vzeroupper callq %rax movq 0x20(%rsp), %rdi movq 0x28(%rsp), %rsi vmovaps 0x140(%rsp), %ymm15 vmovaps 0x160(%rsp), %ymm14 vmovaps 0x180(%rsp), %ymm13 vmovaps 0x1a0(%rsp), %ymm12 movq 0x48(%rsp), %r11 vmovaps 0x1c0(%rsp), %ymm11 vmovaps 0x1e0(%rsp), %ymm10 vmovaps 0x200(%rsp), %ymm9 movq 0x50(%rsp), %r10 movq %r12, %r9 movq %rbx, %r8 movq 0xb0(%rsp), %rax cmpl $0x0, (%rax) je 0x211997 xorl %eax, %eax jmp 0x2119ac vmovd 0x120(%rsp), %xmm0 vmovd %xmm0, 0x20(%r9) btcq %r14, %r15 movb $0x1, %al movq 0x30(%rsp), %r12 xorl %ecx, %ecx movq 0x38(%rsp), %rdx jmp 0x211750 movq 0x68(%rsp), %rbx movq 0x60(%rsp), %r14 movq 0x58(%rsp), %r15 movq 0x78(%rsp), %rcx incq %rcx movq 0x70(%rsp), %rax cmpq %rax, %rcx setb 0x1f(%rsp) jne 0x210fee jmp 0x211a68 vmovaps 0x200(%rsp), %ymm9 vmovaps 0x1e0(%rsp), %ymm10 vmovaps 0x1c0(%rsp), %ymm11 vmovaps 0x1a0(%rsp), %ymm12 vmovaps 0x180(%rsp), %ymm13 vmovaps 0x160(%rsp), %ymm14 vmovaps 0x140(%rsp), %ymm15 movq 0x28(%rsp), %rsi movq 0x20(%rsp), %rdi movq 0x38(%rsp), %rdx jmp 0x2119cc testb $0x1, 0x1f(%rsp) movq 0x68(%rsp), %rbx movq 0x60(%rsp), %r14 movq 0x58(%rsp), %r15 je 0x211a5a movl $0xff800000, 0x20(%r9) # imm = 0xFF800000 pushq $0x3 popq %rcx cmpl $0x3, %ecx jne 0x210d32 jmp 0x210b9c xorl %ecx, %ecx jmp 0x211a5a	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, false, embree::avx::ArrayIntersector1<embree::avx::ObjectIntersector1<true>>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1348, %rsp # imm = 0x1348 movq %rdx, 0x40(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x211f31 addq $0x1348, %rsp # imm = 0x1348 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq vmovss 0x20(%rsi), %xmm0 vxorps %xmm5, %xmm5, %xmm5 vucomiss %xmm0, %xmm5 ja 0x211f1c leaq 0x1a8(%rsp), %rdi movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmaxss 0xc(%rsi), %xmm5, %xmm1 vmovaps 0x10(%rsi), %xmm2 vbroadcastss 0x1d0ef61(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1cdf078(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vblendvps %xmm3, %xmm4, %xmm2, %xmm2 vrcpps %xmm2, %xmm3 vmulps %xmm2, %xmm3, %xmm2 vbroadcastss 0x1cda788(%rip), %xmm4 # 0x1eec714 vsubps %xmm2, %xmm4, %xmm2 vmulps %xmm2, %xmm3, %xmm2 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vaddps %xmm2, %xmm3, %xmm2 vbroadcastss 0x8(%rsi), %ymm8 xorl %r8d, %r8d vucomiss %xmm5, %xmm2 setb %r8b vshufps $0x0, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm9 vmovshdup %xmm2, %xmm3 # xmm3 = xmm2[1,1,3,3] vshufps $0x55, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm10 vshufpd $0x1, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[1,0] vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm5, %xmm3 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm5, %xmm4 vinsertf128 $0x1, %xmm2, %ymm2, %ymm5 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %r15 xorq $0x20, %r15 movq %r10, %rbp xorq $0x20, %rbp vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm11 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm12 leaq 0x1a0(%rsp), %r13 vmovups %ymm6, 0x180(%rsp) vmovups %ymm7, 0x160(%rsp) vmovups %ymm8, 0x140(%rsp) movq %r8, 0x30(%rsp) vmovups %ymm9, 0x120(%rsp) vmovups %ymm10, 0x100(%rsp) movq %r9, 0x28(%rsp) movq %r10, 0x20(%rsp) vmovups %ymm5, 0xe0(%rsp) movq %r11, 0x18(%rsp) movq %r15, 0x10(%rsp) movq %rbp, 0x8(%rsp) vmovups %ymm11, 0xc0(%rsp) vmovups %ymm12, 0xa0(%rsp) cmpq %r13, %rdi je 0x211f1c movq -0x8(%rdi), %r12 addq $-0x8, %rdi testb $0x8, %r12b jne 0x2121a4 movq %r12, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rsi), %ymm0 vmulps 0x100(%rax,%r8), %ymm0, %ymm1 vaddps 0x40(%rax,%r8), %ymm1, %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmulps 0x100(%rax,%r9), %ymm0, %ymm2 vaddps 0x40(%rax,%r9), %ymm2, %ymm2 vmaxps %ymm1, %ymm11, %ymm1 vsubps %ymm7, %ymm2, %ymm2 vmulps 0x100(%rax,%r10), %ymm0, %ymm3 vmulps %ymm2, %ymm10, %ymm2 vaddps 0x40(%rax,%r10), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm5, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r11), %ymm3, %ymm2 vsubps %ymm6, %ymm2, %ymm2 vmulps %ymm2, %ymm9, %ymm2 vminps %ymm2, %ymm12, %ymm2 vmulps 0x100(%rax,%r15), %ymm0, %ymm3 vaddps 0x40(%rax,%r15), %ymm3, %ymm3 vsubps %ymm7, %ymm3, %ymm3 vmulps %ymm3, %ymm10, %ymm3 vmulps 0x100(%rax,%rbp), %ymm0, %ymm4 vaddps 0x40(%rax,%rbp), %ymm4, %ymm4 vsubps %ymm8, %ymm4, %ymm4 vmulps %ymm4, %ymm5, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r12d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x2121ee vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %ebx testb $0x8, %r12b jne 0x2121ea testq %rbx, %rbx je 0x21221b andq $-0x10, %r12 bsfq %rbx, %rax leaq -0x1(%rbx), %rdx xorl %r14d, %r14d movq (%r12,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %rbx, %rdx jne 0x212221 movq %rax, %r12 testl %r14d, %r14d je 0x2120b5 jmp 0x212267 pushq $0x6 jmp 0x21221d vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x212194 pushq $0x4 popq %r14 jmp 0x2121df movq %rax, (%rdi) addq $0x8, %rdi bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x21225f movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x212230 movq %rcx, %r12 jmp 0x2121df cmpl $0x6, %r14d jne 0x2123d6 movl %r12d, %eax andl $0xf, %eax xorl %r14d, %r14d addq $-0x8, %rax movq %rax, 0x50(%rsp) setne %r13b je 0x2123ce movq %rdi, 0x38(%rsp) andq $-0x10, %r12 movq 0x40(%rsp), %rax movq (%rax), %rax movq %rax, 0x48(%rsp) xorl %ebp, %ebp movl (%r12,%rbp,8), %ecx movq 0x48(%rsp), %rax movq 0x1e8(%rax), %rax movq (%rax,%rcx,8), %rax movl 0x34(%rax), %edx testl %edx, 0x24(%rsi) je 0x212399 movl 0x4(%r12,%rbp,8), %edx orl $-0x1, 0x4(%rsp) leaq 0x4(%rsp), %rdi movq %rdi, 0x58(%rsp) movq 0x18(%rax), %rdi movq %rdi, 0x60(%rsp) movq 0x40(%rsp), %rdi movq 0x8(%rdi), %r8 movq %r8, 0x70(%rsp) movq %rsi, %r15 movq %rsi, 0x78(%rsp) movl $0x1, 0x80(%rsp) movl %ecx, 0x84(%rsp) movl %edx, 0x68(%rsp) movq %rax, 0x88(%rsp) andq $0x0, 0x90(%rsp) movq 0x10(%rdi), %rcx movq %rcx, 0x98(%rsp) movq 0x18(%rcx), %rcx testq %rcx, %rcx jne 0x212338 movq 0x68(%rax), %rcx leaq 0x58(%rsp), %rdi vzeroupper callq *%rcx movq %r15, %rsi vxorps %xmm0, %xmm0, %xmm0 vucomiss 0x20(%r15), %xmm0 vmovups 0x180(%rsp), %ymm6 vmovups 0x160(%rsp), %ymm7 vmovups 0x140(%rsp), %ymm8 vmovups 0x120(%rsp), %ymm9 vmovups 0x100(%rsp), %ymm10 vmovups 0xe0(%rsp), %ymm5 vmovups 0xc0(%rsp), %ymm11 vmovups 0xa0(%rsp), %ymm12 ja 0x2123e5 incq %rbp cmpq 0x50(%rsp), %rbp setb %r13b jne 0x2122a5 movq 0x38(%rsp), %rdi movq 0x30(%rsp), %r8 movq 0x28(%rsp), %r9 movq 0x20(%rsp), %r10 movq 0x18(%rsp), %r11 movq 0x10(%rsp), %r15 movq 0x8(%rsp), %rbp leaq 0x1a0(%rsp), %r13 cmpl $0x3, %r14d jne 0x2120a4 jmp 0x211f1c testb $0x1, %r13b movq 0x38(%rsp), %rdi movq 0x30(%rsp), %r8 movq 0x28(%rsp), %r9 movq 0x20(%rsp), %r10 movq 0x18(%rsp), %r11 movq 0x10(%rsp), %r15 movq 0x8(%rsp), %rbp leaq 0x1a0(%rsp), %r13 je 0x2123d6 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %r14 jmp 0x2123d6 nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, false, embree::avx::ArrayIntersector1<embree::avx::InstanceIntersector1>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x12f8, %rsp # imm = 0x12F8 movq %rdx, 0x40(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x212866 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x212866 leaq 0x158(%rsp), %rdi movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmovaps 0x10(%rsi), %xmm3 vmaxss 0xc(%rsi), %xmm2, %xmm1 vbroadcastss 0x1d0ea46(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1cdeb5d(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1cda26d(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vbroadcastss 0x8(%rsi), %ymm8 xorl %r8d, %r8d vucomiss %xmm2, %xmm3 setb %r8b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm2, %xmm4 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm2, %xmm5 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %rcx xorq $0x20, %rcx movq %r10, %rbx xorq $0x20, %rbx vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm4 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm5 leaq 0x150(%rsp), %rbp movq %rsi, 0x8(%rsp) vmovups %ymm6, 0x130(%rsp) vmovups %ymm7, 0x110(%rsp) vmovups %ymm8, 0xf0(%rsp) movq %r8, 0x30(%rsp) vmovups %ymm9, 0xd0(%rsp) vmovups %ymm10, 0xb0(%rsp) vmovups %ymm3, 0x90(%rsp) movq %r9, 0x28(%rsp) movq %r10, 0x20(%rsp) movq %r11, 0x18(%rsp) movq %rcx, 0x10(%rsp) vmovups %ymm4, 0x70(%rsp) vmovups %ymm5, 0x50(%rsp) cmpq %rbp, %rdi je 0x212866 movq -0x8(%rdi), %r12 addq $-0x8, %rdi testb $0x8, %r12b jne 0x21264d vmovaps 0x40(%r12,%r8), %ymm0 vsubps %ymm6, %ymm0, %ymm0 vmulps %ymm0, %ymm9, %ymm0 vmovaps 0x40(%r12,%r9), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm10, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r12,%r10), %ymm1 vsubps %ymm8, %ymm1, %ymm1 vmulps %ymm1, %ymm3, %ymm1 vmaxps %ymm4, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r12,%r11), %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmovaps 0x40(%r12,%rcx), %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm10, %ymm2 vminps %ymm2, %ymm1, %ymm1 vmovaps 0x40(%r12,%rbx), %ymm2 vsubps %ymm8, %ymm2, %ymm2 vmulps %ymm2, %ymm3, %ymm2 vminps %ymm5, %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r13d testb $0x8, %r12b jne 0x212693 testq %r13, %r13 je 0x212697 andq $-0x10, %r12 bsfq %r13, %rax leaq -0x1(%r13), %rdx xorl %r14d, %r14d movq (%r12,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r13, %rdx jne 0x21269d movq %rax, %r12 testl %r14d, %r14d je 0x2125ca jmp 0x2126e6 pushq $0x6 jmp 0x212699 pushq $0x4 popq %r14 jmp 0x212688 movq %rcx, %r15 movq %rax, (%rdi) addq $0x8, %rdi bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x2126de movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x2126af movq %rcx, %r12 movq %r15, %rcx jmp 0x212688 cmpl $0x6, %r14d jne 0x21285c movl %r12d, %ebp andl $0xf, %ebp xorl %r14d, %r14d movq %rbp, %rax addq $-0x8, %rax setne %r15b je 0x212854 movq %rax, 0x48(%rsp) movq %rdi, 0x38(%rsp) andq $-0x10, %r12 leaq 0x7(%rsp), %rdi movq 0x40(%rsp), %rdx movq %r12, %rcx vzeroupper callq 0x3f50dc testb %al, %al jne 0x212770 addq $0x10, %r12 addq $-0x9, %rbp xorl %r15d, %r15d cmpq %r15, %rbp je 0x2127ef leaq 0x7(%rsp), %rdi movq 0x8(%rsp), %rsi movq 0x40(%rsp), %rdx movq %r12, %rcx callq 0x3f50dc addq $0x10, %r12 incq %r15 testb %al, %al je 0x21273c cmpq 0x48(%rsp), %r15 setb %r15b testb %r15b, %r15b movq 0x8(%rsp), %rsi movq 0x38(%rsp), %rdi vmovups 0x130(%rsp), %ymm6 vmovups 0x110(%rsp), %ymm7 vmovups 0xf0(%rsp), %ymm8 movq 0x30(%rsp), %r8 vmovups 0xd0(%rsp), %ymm9 vmovups 0xb0(%rsp), %ymm10 vmovups 0x90(%rsp), %ymm3 movq 0x28(%rsp), %r9 movq 0x20(%rsp), %r10 movq 0x18(%rsp), %r11 movq 0x10(%rsp), %rcx vmovups 0x70(%rsp), %ymm4 vmovups 0x50(%rsp), %ymm5 leaq 0x150(%rsp), %rbp je 0x21285c movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %r14 jmp 0x21285c movq 0x8(%rsp), %rsi movq 0x38(%rsp), %rdi vmovups 0x130(%rsp), %ymm6 vmovups 0x110(%rsp), %ymm7 vmovups 0xf0(%rsp), %ymm8 movq 0x30(%rsp), %r8 vmovups 0xd0(%rsp), %ymm9 vmovups 0xb0(%rsp), %ymm10 vmovups 0x90(%rsp), %ymm3 movq 0x28(%rsp), %r9 movq 0x20(%rsp), %r10 movq 0x18(%rsp), %r11 movq 0x10(%rsp), %rcx vmovups 0x70(%rsp), %ymm4 vmovups 0x50(%rsp), %ymm5 leaq 0x150(%rsp), %rbp cmpl $0x3, %r14d jne 0x2125b9 addq $0x12f8, %rsp # imm = 0x12F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, false, embree::avx::ArrayIntersector1<embree::avx::InstanceIntersector1MB>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x12f8, %rsp # imm = 0x12F8 movq %rdx, 0x38(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x212d4b vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x212d4b leaq 0x158(%rsp), %rdi movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmovaps 0x10(%rsi), %xmm3 vmaxss 0xc(%rsi), %xmm2, %xmm1 vbroadcastss 0x1d0e5ee(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1cde705(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1cd9e15(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vbroadcastss 0x8(%rsi), %ymm8 xorl %r8d, %r8d vucomiss %xmm2, %xmm3 setb %r8b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm11 shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm2, %xmm4 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm2, %xmm5 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %r15 xorq $0x20, %r15 movq %r10, %rbx xorq $0x20, %rbx vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm5 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm12 leaq 0x150(%rsp), %rbp movq %rsi, 0x8(%rsp) vmovups %ymm6, 0x130(%rsp) vmovups %ymm7, 0x110(%rsp) vmovups %ymm8, 0xf0(%rsp) movq %r8, 0x28(%rsp) vmovups %ymm9, 0xd0(%rsp) vmovups %ymm10, 0xb0(%rsp) vmovups %ymm11, 0x90(%rsp) movq %r9, 0x20(%rsp) movq %r10, 0x18(%rsp) movq %r11, 0x10(%rsp) vmovups %ymm5, 0x70(%rsp) vmovups %ymm12, 0x50(%rsp) cmpq %rbp, %rdi je 0x212d4b movq -0x8(%rdi), %r12 addq $-0x8, %rdi testb $0x8, %r12b jne 0x212b0d movq %r12, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rsi), %ymm0 vmulps 0x100(%rax,%r8), %ymm0, %ymm1 vaddps 0x40(%rax,%r8), %ymm1, %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmulps 0x100(%rax,%r9), %ymm0, %ymm2 vaddps 0x40(%rax,%r9), %ymm2, %ymm2 vmaxps %ymm1, %ymm5, %ymm1 vsubps %ymm7, %ymm2, %ymm2 vmulps 0x100(%rax,%r10), %ymm0, %ymm3 vmulps %ymm2, %ymm10, %ymm2 vaddps 0x40(%rax,%r10), %ymm3, %ymm3 vsubps %ymm8, %ymm3, %ymm3 vmulps %ymm3, %ymm11, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmulps 0x100(%rax,%r11), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r11), %ymm3, %ymm2 vsubps %ymm6, %ymm2, %ymm2 vmulps %ymm2, %ymm9, %ymm2 vminps %ymm2, %ymm12, %ymm2 vmulps 0x100(%rax,%r15), %ymm0, %ymm3 vaddps 0x40(%rax,%r15), %ymm3, %ymm3 vsubps %ymm7, %ymm3, %ymm3 vmulps %ymm3, %ymm10, %ymm3 vmulps 0x100(%rax,%rbx), %ymm0, %ymm4 vaddps 0x40(%rax,%rbx), %ymm4, %ymm4 vsubps %ymm8, %ymm4, %ymm4 vmulps %ymm4, %ymm11, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r12d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x212b57 vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r14d testb $0x8, %r12b jne 0x212b53 testq %r14, %r14 je 0x212b84 andq $-0x10, %r12 bsfq %r14, %rax leaq -0x1(%r14), %rdx xorl %r13d, %r13d movq (%r12,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r14, %rdx jne 0x212b8a movq %rax, %r12 testl %r13d, %r13d je 0x212a1d jmp 0x212bd0 pushq $0x6 jmp 0x212b86 vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x212afc pushq $0x4 popq %r13 jmp 0x212b48 movq %rax, (%rdi) addq $0x8, %rdi bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x212bc8 movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x212b99 movq %rcx, %r12 jmp 0x212b48 cmpl $0x6, %r13d jne 0x212d41 movl %r12d, %eax andl $0xf, %eax xorl %r13d, %r13d movq %rax, 0x40(%rsp) addq $-0x8, %rax setne %bpl je 0x212d39 movq %rax, 0x48(%rsp) movq %rdi, 0x30(%rsp) andq $-0x10, %r12 leaq 0x7(%rsp), %rdi movq 0x38(%rsp), %rdx movq %r12, %rcx vzeroupper callq 0x3f5b0c testb %al, %al jne 0x212c5f addq $0x10, %r12 addq $-0x9, 0x40(%rsp) xorl %ebp, %ebp cmpq %rbp, 0x40(%rsp) je 0x212cd9 leaq 0x7(%rsp), %rdi movq 0x8(%rsp), %rsi movq 0x38(%rsp), %rdx movq %r12, %rcx callq 0x3f5b0c addq $0x10, %r12 incq %rbp testb %al, %al je 0x212c29 cmpq 0x48(%rsp), %rbp setb %bpl testb %bpl, %bpl movq 0x8(%rsp), %rsi movq 0x30(%rsp), %rdi vmovups 0x130(%rsp), %ymm6 vmovups 0x110(%rsp), %ymm7 vmovups 0xf0(%rsp), %ymm8 movq 0x28(%rsp), %r8 vmovups 0xd0(%rsp), %ymm9 vmovups 0xb0(%rsp), %ymm10 vmovups 0x90(%rsp), %ymm11 movq 0x20(%rsp), %r9 movq 0x18(%rsp), %r10 movq 0x10(%rsp), %r11 vmovups 0x70(%rsp), %ymm5 vmovups 0x50(%rsp), %ymm12 leaq 0x150(%rsp), %rbp je 0x212d41 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %r13 jmp 0x212d41 movq 0x8(%rsp), %rsi movq 0x30(%rsp), %rdi vmovups 0x130(%rsp), %ymm6 vmovups 0x110(%rsp), %ymm7 vmovups 0xf0(%rsp), %ymm8 movq 0x28(%rsp), %r8 vmovups 0xd0(%rsp), %ymm9 vmovups 0xb0(%rsp), %ymm10 vmovups 0x90(%rsp), %ymm11 movq 0x20(%rsp), %r9 movq 0x18(%rsp), %r10 movq 0x10(%rsp), %r11 vmovups 0x70(%rsp), %ymm5 vmovups 0x50(%rsp), %ymm12 leaq 0x150(%rsp), %rbp cmpl $0x3, %r13d jne 0x212a0c addq $0x12f8, %rsp # imm = 0x12F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, false, embree::avx::ArrayIntersector1<embree::avx::InstanceArrayIntersector1>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x12f8, %rsp # imm = 0x12F8 movq %rdx, 0x40(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x2131a2 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x2131a2 leaq 0x158(%rsp), %rdi movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmovaps 0x10(%rsi), %xmm3 vmaxss 0xc(%rsi), %xmm2, %xmm1 vbroadcastss 0x1d0e10a(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1cde221(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vrcpps %xmm3, %xmm4 vmulps %xmm3, %xmm4, %xmm3 vbroadcastss 0x1cd9931(%rip), %xmm5 # 0x1eec714 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vbroadcastss (%rsi), %ymm6 vbroadcastss 0x4(%rsi), %ymm7 vbroadcastss 0x8(%rsi), %ymm8 xorl %r8d, %r8d vucomiss %xmm2, %xmm3 setb %r8b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm10 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 shll $0x5, %r8d xorl %r9d, %r9d vucomiss %xmm2, %xmm4 setb %r9b shll $0x5, %r9d orq $0x40, %r9 xorl %r10d, %r10d vucomiss %xmm2, %xmm5 setb %r10b shll $0x5, %r10d orq $0x80, %r10 movq %r8, %r11 xorq $0x20, %r11 movq %r9, %rcx xorq $0x20, %rcx movq %r10, %rbx xorq $0x20, %rbx vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm4 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm5 leaq 0x150(%rsp), %rbp movq %rsi, 0x8(%rsp) vmovups %ymm6, 0x130(%rsp) vmovups %ymm7, 0x110(%rsp) vmovups %ymm8, 0xf0(%rsp) movq %r8, 0x30(%rsp) vmovups %ymm9, 0xd0(%rsp) vmovups %ymm10, 0xb0(%rsp) vmovups %ymm3, 0x90(%rsp) movq %r9, 0x28(%rsp) movq %r10, 0x20(%rsp) movq %r11, 0x18(%rsp) movq %rcx, 0x10(%rsp) vmovups %ymm4, 0x70(%rsp) vmovups %ymm5, 0x50(%rsp) cmpq %rbp, %rdi je 0x2131a2 movq -0x8(%rdi), %r12 addq $-0x8, %rdi testb $0x8, %r12b jne 0x212f89 vmovaps 0x40(%r12,%r8), %ymm0 vsubps %ymm6, %ymm0, %ymm0 vmulps %ymm0, %ymm9, %ymm0 vmovaps 0x40(%r12,%r9), %ymm1 vsubps %ymm7, %ymm1, %ymm1 vmulps %ymm1, %ymm10, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r12,%r10), %ymm1 vsubps %ymm8, %ymm1, %ymm1 vmulps %ymm1, %ymm3, %ymm1 vmaxps %ymm4, %ymm1, %ymm1 vmaxps %ymm1, %ymm0, %ymm0 vmovaps 0x40(%r12,%r11), %ymm1 vsubps %ymm6, %ymm1, %ymm1 vmulps %ymm1, %ymm9, %ymm1 vmovaps 0x40(%r12,%rcx), %ymm2 vsubps %ymm7, %ymm2, %ymm2 vmulps %ymm2, %ymm10, %ymm2 vminps %ymm2, %ymm1, %ymm1 vmovaps 0x40(%r12,%rbx), %ymm2 vsubps %ymm8, %ymm2, %ymm2 vmulps %ymm2, %ymm3, %ymm2 vminps %ymm5, %ymm2, %ymm2 vminps %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %ymm0 vmovmskps %ymm0, %r13d testb $0x8, %r12b jne 0x212fcf testq %r13, %r13 je 0x212fd3 andq $-0x10, %r12 bsfq %r13, %rax leaq -0x1(%r13), %rdx xorl %r14d, %r14d movq (%r12,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) prefetcht0 0x80(%rax) prefetcht0 0xc0(%rax) andq %r13, %rdx jne 0x212fd9 movq %rax, %r12 testl %r14d, %r14d je 0x212f06 jmp 0x213022 pushq $0x6 jmp 0x212fd5 pushq $0x4 popq %r14 jmp 0x212fc4 movq %rcx, %r15 movq %rax, (%rdi) addq $0x8, %rdi bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %rdx, %rax je 0x21301a movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x212feb movq %rcx, %r12 movq %r15, %rcx jmp 0x212fc4 cmpl $0x6, %r14d jne 0x213198 movl %r12d, %ebp andl $0xf, %ebp xorl %r14d, %r14d movq %rbp, %rax addq $-0x8, %rax setne %r15b je 0x213190 movq %rax, 0x48(%rsp) movq %rdi, 0x38(%rsp) andq $-0x10, %r12 leaq 0x7(%rsp), %rdi movq 0x40(%rsp), %rdx movq %r12, %rcx vzeroupper callq 0x4070ac testb %al, %al jne 0x2130ac addq $0x8, %r12 addq $-0x9, %rbp xorl %r15d, %r15d cmpq %r15, %rbp je 0x21312b leaq 0x7(%rsp), %rdi movq 0x8(%rsp), %rsi movq 0x40(%rsp), %rdx movq %r12, %rcx callq 0x4070ac addq $0x8, %r12 incq %r15 testb %al, %al je 0x213078 cmpq 0x48(%rsp), %r15 setb %r15b testb %r15b, %r15b movq 0x8(%rsp), %rsi movq 0x38(%rsp), %rdi vmovups 0x130(%rsp), %ymm6 vmovups 0x110(%rsp), %ymm7 vmovups 0xf0(%rsp), %ymm8 movq 0x30(%rsp), %r8 vmovups 0xd0(%rsp), %ymm9 vmovups 0xb0(%rsp), %ymm10 vmovups 0x90(%rsp), %ymm3 movq 0x28(%rsp), %r9 movq 0x20(%rsp), %r10 movq 0x18(%rsp), %r11 movq 0x10(%rsp), %rcx vmovups 0x70(%rsp), %ymm4 vmovups 0x50(%rsp), %ymm5 leaq 0x150(%rsp), %rbp je 0x213198 movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %r14 jmp 0x213198 movq 0x8(%rsp), %rsi movq 0x38(%rsp), %rdi vmovups 0x130(%rsp), %ymm6 vmovups 0x110(%rsp), %ymm7 vmovups 0xf0(%rsp), %ymm8 movq 0x30(%rsp), %r8 vmovups 0xd0(%rsp), %ymm9 vmovups 0xb0(%rsp), %ymm10 vmovups 0x90(%rsp), %ymm3 movq 0x28(%rsp), %r9 movq 0x20(%rsp), %r10 movq 0x18(%rsp), %r11 movq 0x10(%rsp), %rcx vmovups 0x70(%rsp), %ymm4 vmovups 0x50(%rsp), %ymm5 leaq 0x150(%rsp), %rbp cmpl $0x3, %r14d jne 0x212ef5 addq $0x12f8, %rsp # imm = 0x12F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 16777232, true, embree::avx::SubGridMBIntersector1Pluecker<8, true>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x17e0, %rsp # imm = 0x17E0 movq %rdx, 0x38(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) jne 0x21453d leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %rsi, %rbx vmovss 0x20(%rsi), %xmm0 vxorps %xmm12, %xmm12, %xmm12 vucomiss %xmm0, %xmm12 ja 0x21452b leaq 0x628(%rsp), %r15 movq 0x70(%rax), %rax vmovaps 0x10(%rbx), %xmm2 vmaxss 0xc(%rbx), %xmm12, %xmm1 vbroadcastss 0x1d0c955(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1cdca6c(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1cd818a(%rip), %xmm4 # 0x1eec714 vdivps %xmm2, %xmm4, %xmm2 vbroadcastss 0x1d0c9c9(%rip), %xmm4 # 0x1f20f60 vblendvps %xmm3, %xmm4, %xmm2, %xmm2 movq %rax, -0x8(%r15) vbroadcastss 0x1d0b966(%rip), %xmm3 # 0x1f1ff10 vmulps %xmm3, %xmm2, %xmm3 vbroadcastss 0x1d0b95d(%rip), %xmm4 # 0x1f1ff14 vbroadcastss (%rbx), %ymm15 vbroadcastss 0x4(%rbx), %ymm13 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss 0x8(%rbx), %ymm6 xorl %r13d, %r13d vucomiss %xmm12, %xmm3 setb %r13b vshufps $0x0, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm4, %ymm4, %ymm7 vmovshdup %xmm3, %xmm4 # xmm4 = xmm3[1,1,3,3] vshufps $0x55, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm5, %ymm5, %ymm8 vshufpd $0x1, %xmm3, %xmm3, %xmm5 # xmm5 = xmm3[1,0] vshufps $0xaa, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm3, %ymm3, %ymm9 vshufps $0x0, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm10 vshufps $0x55, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm3, %ymm3, %ymm11 vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] shll $0x5, %r13d xorl %edi, %edi vucomiss %xmm12, %xmm4 vinsertf128 $0x1, %xmm2, %ymm2, %ymm14 setb %dil shll $0x5, %edi orq $0x40, %rdi xorl %r8d, %r8d vucomiss %xmm12, %xmm5 setb %r8b shll $0x5, %r8d orq $0x80, %r8 movq %r13, %r9 xorq $0x20, %r9 movq %rdi, %r10 xorq $0x20, %r10 movq %r8, %r11 xorq $0x20, %r11 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmovaps %ymm1, 0x400(%rsp) vinsertf128 $0x1, %xmm1, %ymm1, %ymm12 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] leaq 0x1f3b8fb(%rip), %rax # 0x214ff80 vbroadcastf128 0xf0(%rax), %ymm1 # ymm1 = mem[0,1,0,1] vmovaps %ymm1, 0x380(%rsp) vperm2f128 $0x2, (%rax), %ymm1, %ymm1 # ymm1 = mem[0,1],ymm1[0,1] vmovaps %ymm1, 0x3c0(%rsp) vmovaps %ymm0, 0x3e0(%rsp) vinsertf128 $0x1, %xmm0, %ymm0, %ymm5 leaq 0x620(%rsp), %rdx movq %rbx, 0x110(%rsp) vmovaps %ymm15, 0x220(%rsp) vmovaps %ymm13, 0x1a0(%rsp) vmovaps %ymm6, 0x340(%rsp) movq %r13, 0x50(%rsp) vmovaps %ymm7, 0x2e0(%rsp) vmovaps %ymm8, 0x2c0(%rsp) vmovaps %ymm9, 0x2a0(%rsp) vmovaps %ymm10, 0x320(%rsp) vmovaps %ymm11, 0x300(%rsp) movq %rdi, 0x30(%rsp) vmovaps %ymm14, 0x180(%rsp) movq %r8, 0x28(%rsp) movq %r9, 0x20(%rsp) movq %r10, 0x18(%rsp) movq %r11, 0x10(%rsp) vmovaps %ymm12, 0x160(%rsp) vmovaps %ymm5, 0x260(%rsp) cmpq %rdx, %r15 je 0x21452b movq -0x8(%r15), %r14 addq $-0x8, %r15 testb $0x8, %r14b jne 0x214849 movq %r14, %rax andq $-0x10, %rax vbroadcastss 0x1c(%rbx), %ymm0 vmulps 0x100(%rax,%r13), %ymm0, %ymm1 vaddps 0x40(%rax,%r13), %ymm1, %ymm1 vsubps %ymm15, %ymm1, %ymm1 vmulps %ymm1, %ymm7, %ymm1 vmulps 0x100(%rax,%rdi), %ymm0, %ymm2 vaddps 0x40(%rax,%rdi), %ymm2, %ymm2 vsubps %ymm13, %ymm2, %ymm2 vmulps %ymm2, %ymm8, %ymm2 vmulps 0x100(%rax,%r8), %ymm0, %ymm3 vaddps 0x40(%rax,%r8), %ymm3, %ymm3 vsubps %ymm6, %ymm3, %ymm3 vmulps %ymm3, %ymm9, %ymm3 vmaxps %ymm3, %ymm2, %ymm2 vmaxps %ymm1, %ymm12, %ymm1 vmulps 0x100(%rax,%r9), %ymm0, %ymm3 vmaxps %ymm2, %ymm1, %ymm1 vaddps 0x40(%rax,%r9), %ymm3, %ymm2 vsubps %ymm15, %ymm2, %ymm2 vmulps 0x100(%rax,%r10), %ymm0, %ymm3 vaddps 0x40(%rax,%r10), %ymm3, %ymm3 vmulps %ymm2, %ymm10, %ymm2 vsubps %ymm13, %ymm3, %ymm3 vmulps 0x100(%rax,%r11), %ymm0, %ymm4 vmulps %ymm3, %ymm11, %ymm3 vaddps 0x40(%rax,%r11), %ymm4, %ymm4 vsubps %ymm6, %ymm4, %ymm4 vmulps %ymm4, %ymm14, %ymm4 vminps %ymm4, %ymm3, %ymm3 vminps %ymm2, %ymm5, %ymm2 vminps %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r14d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x214895 vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r12d testb $0x8, %r14b jne 0x214891 testq %r12, %r12 je 0x2148c2 andq $-0x10, %r14 bsfq %r12, %rcx leaq -0x1(%r12), %rsi xorl %eax, %eax movq (%r14,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) andq %r12, %rsi jne 0x2148c7 movq %rcx, %r14 testl %eax, %eax je 0x214757 jmp 0x214915 pushq $0x6 jmp 0x2148c4 vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x214838 pushq $0x4 popq %rax jmp 0x214884 movq %rcx, (%r15) addq $0x8, %r15 bsfq %rsi, %rdx leaq -0x1(%rsi), %rcx movq (%r14,%rdx,8), %rdx prefetcht0 (%rdx) prefetcht0 0x40(%rdx) prefetcht0 0x80(%rdx) prefetcht0 0xc0(%rdx) andq %rsi, %rcx je 0x214905 movq %rdx, (%r15) addq $0x8, %r15 bsfq %rcx, %rdx leaq -0x1(%rcx), %rsi jmp 0x2148d6 movq %rdx, %r14 leaq 0x620(%rsp), %rdx jmp 0x214884 cmpl $0x6, %eax jne 0x215c15 movl %r14d, %esi andl $0xf, %esi addq $-0x8, %rsi setne %cl je 0x215c04 andq $-0x10, %r14 xorl %eax, %eax movq %rax, 0x120(%rsp) imulq $0xe0, %rax, %rdx vmovss 0x1c(%rbx), %xmm0 vsubss 0xd0(%r14,%rdx), %xmm0, %xmm0 vmulss 0xd4(%r14,%rdx), %xmm0, %xmm2 vmovq 0x40(%r14,%rdx), %xmm0 vpmovzxbd %xmm0, %xmm1 # xmm1 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vmovq 0x44(%r14,%rdx), %xmm3 vpmovzxbd %xmm3, %xmm4 # xmm4 = xmm3[0],zero,zero,zero,xmm3[1],zero,zero,zero,xmm3[2],zero,zero,zero,xmm3[3],zero,zero,zero vinsertf128 $0x1, %xmm4, %ymm1, %ymm4 vmovq 0x48(%r14,%rdx), %xmm1 vpmovzxbd %xmm1, %xmm5 # xmm5 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero vmovaps %ymm6, %ymm14 vmovq 0x4c(%r14,%rdx), %xmm6 vpmovzxbd %xmm6, %xmm7 # xmm7 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vinsertf128 $0x1, %xmm7, %ymm5, %ymm5 vpminub %xmm1, %xmm0, %xmm1 vpcmpeqb %xmm1, %xmm0, %xmm0 vmovdqa %xmm0, 0x80(%rsp) vpminub %xmm6, %xmm3, %xmm1 vpcmpeqb %xmm1, %xmm3, %xmm0 vmovdqa %xmm0, 0xd0(%rsp) vcvtdq2ps %ymm4, %ymm3 vbroadcastss 0x7c(%r14,%rdx), %ymm4 vbroadcastss 0x70(%r14,%rdx), %ymm6 vmulps %ymm3, %ymm4, %ymm3 vaddps %ymm3, %ymm6, %ymm3 vmovq 0x88(%r14,%rdx), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x8c(%r14,%rdx), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vcvtdq2ps %ymm7, %ymm7 vbroadcastss 0xc4(%r14,%rdx), %ymm8 vbroadcastss 0xb8(%r14,%rdx), %ymm9 vmulps %ymm7, %ymm8, %ymm7 vaddps %ymm7, %ymm9, %ymm7 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vsubps %ymm3, %ymm7, %ymm7 vmulps %ymm7, %ymm2, %ymm7 vaddps %ymm7, %ymm3, %ymm3 vcvtdq2ps %ymm5, %ymm5 vmulps %ymm5, %ymm4, %ymm4 vaddps %ymm4, %ymm6, %ymm4 vmovq 0x90(%r14,%rdx), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vmovq 0x94(%r14,%rdx), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vinsertf128 $0x1, %xmm6, %ymm5, %ymm5 vcvtdq2ps %ymm5, %ymm5 vmulps %ymm5, %ymm8, %ymm5 vaddps %ymm5, %ymm9, %ymm5 vsubps %ymm4, %ymm5, %ymm5 vmulps %ymm5, %ymm2, %ymm5 vaddps %ymm5, %ymm4, %ymm4 vmovq 0x50(%r14,%rdx), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vmovq 0x54(%r14,%rdx), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vinsertf128 $0x1, %xmm6, %ymm5, %ymm5 vbroadcastss 0x80(%r14,%rdx), %ymm6 vcvtdq2ps %ymm5, %ymm5 vbroadcastss 0x74(%r14,%rdx), %ymm7 vmulps %ymm5, %ymm6, %ymm5 vaddps %ymm5, %ymm7, %ymm5 vmovq 0x98(%r14,%rdx), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vmovq 0x9c(%r14,%rdx), %xmm9 vpmovzxbd %xmm9, %xmm9 # xmm9 = xmm9[0],zero,zero,zero,xmm9[1],zero,zero,zero,xmm9[2],zero,zero,zero,xmm9[3],zero,zero,zero vinsertf128 $0x1, %xmm9, %ymm8, %ymm8 vbroadcastss 0xc8(%r14,%rdx), %ymm9 vcvtdq2ps %ymm8, %ymm8 vmovaps %ymm10, %ymm1 vbroadcastss 0xbc(%r14,%rdx), %ymm10 vmulps %ymm8, %ymm9, %ymm8 vaddps %ymm8, %ymm10, %ymm8 vsubps %ymm5, %ymm8, %ymm8 vmulps %ymm2, %ymm8, %ymm8 vaddps %ymm5, %ymm8, %ymm5 vmovq 0x58(%r14,%rdx), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vmovaps %ymm11, %ymm0 vmovq 0x5c(%r14,%rdx), %xmm11 vpmovzxbd %xmm11, %xmm11 # xmm11 = xmm11[0],zero,zero,zero,xmm11[1],zero,zero,zero,xmm11[2],zero,zero,zero,xmm11[3],zero,zero,zero vinsertf128 $0x1, %xmm11, %ymm8, %ymm8 vcvtdq2ps %ymm8, %ymm8 vmulps %ymm6, %ymm8, %ymm6 vaddps %ymm6, %ymm7, %ymm6 vmovq 0xa0(%r14,%rdx), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0xa4(%r14,%rdx), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm9, %ymm7 vaddps %ymm7, %ymm10, %ymm7 vsubps %ymm6, %ymm7, %ymm7 vmulps %ymm7, %ymm2, %ymm7 vaddps %ymm7, %ymm6, %ymm6 vmovq 0x60(%r14,%rdx), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vmovq 0x64(%r14,%rdx), %xmm8 vpmovzxbd %xmm8, %xmm8 # xmm8 = xmm8[0],zero,zero,zero,xmm8[1],zero,zero,zero,xmm8[2],zero,zero,zero,xmm8[3],zero,zero,zero vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vbroadcastss 0x84(%r14,%rdx), %ymm8 vbroadcastss 0x78(%r14,%rdx), %ymm9 vcvtdq2ps %ymm7, %ymm7 vmulps %ymm7, %ymm8, %ymm7 vaddps %ymm7, %ymm9, %ymm7 vmovq 0xa8(%r14,%rdx), %xmm10 vpmovzxbd %xmm10, %xmm10 # xmm10 = xmm10[0],zero,zero,zero,xmm10[1],zero,zero,zero,xmm10[2],zero,zero,zero,xmm10[3],zero,zero,zero vmovq 0xac(%r14,%rdx), %xmm11 vpmovzxbd %xmm11, %xmm11 # xmm11 = xmm11[0],zero,zero,zero,xmm11[1],zero,zero,zero,xmm11[2],zero,zero,zero,xmm11[3],zero,zero,zero vinsertf128 $0x1, %xmm11, %ymm10, %ymm10 vbroadcastss 0xcc(%r14,%rdx), %ymm11 vbroadcastss 0xc0(%r14,%rdx), %ymm12 vcvtdq2ps %ymm10, %ymm10 vmulps %ymm10, %ymm11, %ymm10 vaddps %ymm10, %ymm12, %ymm10 vsubps %ymm7, %ymm10, %ymm10 vmulps %ymm2, %ymm10, %ymm10 vaddps %ymm7, %ymm10, %ymm7 vmovq 0x68(%r14,%rdx), %xmm10 vpmovzxbd %xmm10, %xmm10 # xmm10 = xmm10[0],zero,zero,zero,xmm10[1],zero,zero,zero,xmm10[2],zero,zero,zero,xmm10[3],zero,zero,zero vmovq 0x6c(%r14,%rdx), %xmm13 vpmovzxbd %xmm13, %xmm13 # xmm13 = xmm13[0],zero,zero,zero,xmm13[1],zero,zero,zero,xmm13[2],zero,zero,zero,xmm13[3],zero,zero,zero vinsertf128 $0x1, %xmm13, %ymm10, %ymm10 vmovaps 0x1a0(%rsp), %ymm13 vcvtdq2ps %ymm10, %ymm10 vmulps %ymm10, %ymm8, %ymm8 vaddps %ymm8, %ymm9, %ymm8 vmovq 0xb0(%r14,%rdx), %xmm9 vpmovzxbd %xmm9, %xmm9 # xmm9 = xmm9[0],zero,zero,zero,xmm9[1],zero,zero,zero,xmm9[2],zero,zero,zero,xmm9[3],zero,zero,zero vmovq 0xb4(%r14,%rdx), %xmm10 vpmovzxbd %xmm10, %xmm10 # xmm10 = xmm10[0],zero,zero,zero,xmm10[1],zero,zero,zero,xmm10[2],zero,zero,zero,xmm10[3],zero,zero,zero vinsertf128 $0x1, %xmm10, %ymm9, %ymm9 vcvtdq2ps %ymm9, %ymm9 vmulps %ymm9, %ymm11, %ymm9 vaddps %ymm9, %ymm12, %ymm9 vsubps %ymm8, %ymm9, %ymm9 vmulps %ymm2, %ymm9, %ymm2 vaddps %ymm2, %ymm8, %ymm8 vsubps %ymm15, %ymm3, %ymm2 vmulps 0x2e0(%rsp), %ymm2, %ymm2 vsubps %ymm13, %ymm5, %ymm3 vmulps 0x2c0(%rsp), %ymm3, %ymm3 vsubps %ymm14, %ymm7, %ymm5 vmulps 0x2a0(%rsp), %ymm5, %ymm5 vsubps %ymm15, %ymm4, %ymm4 vmulps %ymm4, %ymm1, %ymm4 vsubps %ymm13, %ymm6, %ymm6 vmulps %ymm6, %ymm0, %ymm6 vsubps %ymm14, %ymm8, %ymm7 vmovaps 0x180(%rsp), %ymm14 vmulps %ymm7, %ymm14, %ymm7 vextractf128 $0x1, %ymm2, %xmm8 vextractf128 $0x1, %ymm4, %xmm9 vpminsd %xmm9, %xmm8, %xmm10 vpmaxsd %xmm9, %xmm8, %xmm8 vextractf128 $0x1, %ymm3, %xmm9 vextractf128 $0x1, %ymm6, %xmm11 vpminsd %xmm11, %xmm9, %xmm12 vpmaxsd %xmm12, %xmm10, %xmm10 vpmaxsd %xmm11, %xmm9, %xmm9 vpminsd %xmm9, %xmm8, %xmm8 vextractf128 $0x1, %ymm5, %xmm9 vextractf128 $0x1, %ymm7, %xmm11 vpminsd %xmm11, %xmm9, %xmm12 vpmaxsd %xmm11, %xmm9, %xmm9 vpmaxsd 0x400(%rsp), %xmm12, %xmm11 vmovdqa 0x160(%rsp), %ymm12 vpmaxsd %xmm11, %xmm10, %xmm10 vpminsd 0x3e0(%rsp), %xmm9, %xmm9 vpminsd %xmm9, %xmm8, %xmm8 vpcmpeqd %xmm0, %xmm0, %xmm0 vpxor 0xd0(%rsp), %xmm0, %xmm1 vpmovsxbd %xmm1, %xmm1 vpcmpgtd %xmm8, %xmm10, %xmm8 vpor %xmm1, %xmm8, %xmm1 vpminsd %xmm4, %xmm2, %xmm8 vpminsd %xmm6, %xmm3, %xmm9 vpmaxsd %xmm9, %xmm8, %xmm8 vpminsd %xmm7, %xmm5, %xmm9 vpmaxsd %xmm12, %xmm9, %xmm9 vpmaxsd %xmm9, %xmm8, %xmm8 vpmaxsd %xmm4, %xmm2, %xmm2 vpmaxsd %xmm6, %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vpmaxsd %xmm7, %xmm5, %xmm3 vmovdqa 0x260(%rsp), %ymm5 vpminsd %xmm5, %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vpcmpgtd %xmm2, %xmm8, %xmm2 vpxor 0x80(%rsp), %xmm0, %xmm0 vpmovsxbd %xmm0, %xmm0 vpor %xmm0, %xmm2, %xmm0 vinsertf128 $0x1, %xmm1, %ymm0, %ymm0 vcvtdq2ps %ymm0, %ymm0 vbroadcastss 0x1cdbc03(%rip), %ymm1 # 0x1ef09cc vcmpltps %ymm0, %ymm1, %ymm0 vmovmskps %ymm0, %eax testl %eax, %eax je 0x215b53 addq %r14, %rdx movzbl %al, %eax movq %r15, 0x58(%rsp) movb %cl, 0x6(%rsp) movq %rsi, 0x8(%rsp) movq %rdx, 0x108(%rsp) movq %rax, 0x118(%rsp) bsfq %rax, %rax movzwl (%rdx,%rax,8), %r8d movl %r8d, 0x140(%rsp) movzwl 0x2(%rdx,%rax,8), %r9d movl %r9d, 0xe0(%rsp) movl 0xd8(%rdx), %ecx movl 0x4(%rdx,%rax,8), %edx movq 0x38(%rsp), %rax movq (%rax), %rax movq %rax, 0x138(%rsp) movq 0x1e8(%rax), %rax movq %rcx, 0xd0(%rsp) movq (%rax,%rcx,8), %rax movq 0x58(%rax), %rsi movq 0x68(%rax), %rdi movq %rdx, 0x100(%rsp) imulq %rdx, %rdi vmovss 0x1c(%rbx), %xmm0 vmovss 0x28(%rax), %xmm1 vmovss 0x2c(%rax), %xmm2 vmovss 0x30(%rax), %xmm3 vsubss %xmm2, %xmm0, %xmm0 vsubss %xmm2, %xmm3, %xmm2 vdivss %xmm2, %xmm0, %xmm0 vmulss %xmm0, %xmm1, %xmm0 vroundss $0x9, %xmm0, %xmm0, %xmm2 vaddss 0x1cdbb3c(%rip), %xmm1, %xmm1 # 0x1ef09cc vminss %xmm1, %xmm2, %xmm1 vxorps %xmm2, %xmm2, %xmm2 vmaxss %xmm1, %xmm2, %xmm5 movl $0x7fff, %edx # imm = 0x7FFF andl %edx, %r8d movl (%rsi,%rdi), %ecx movl %r8d, 0x48(%rsp) addl %r8d, %ecx movl %r9d, %r8d andl %edx, %r8d movq %rsi, 0x130(%rsp) movq %rdi, 0x128(%rsp) movl 0x4(%rsi,%rdi), %r10d movl %r10d, %esi movl %r8d, 0x44(%rsp) imull %r8d, %esi vcvttss2si %xmm5, %edx addl %ecx, %esi movslq %edx, %rcx movq 0xe0(%rax), %rdi movq %rdi, 0xa0(%rsp) imulq $0x38, %rcx, %r8 movq %r8, 0x80(%rsp) movq 0x10(%rdi,%r8), %rcx movq %rcx, %rax imulq %rsi, %rax movq (%rdi,%r8), %rdx vmovups (%rdx,%rax), %xmm3 leaq 0x1(%rsi), %rax movq 0x48(%rdi,%r8), %r9 movq %r9, %rdi imulq %rax, %rdi movq %rdi, 0x60(%rsp) imulq %rcx, %rax vmovups (%rdx,%rax), %xmm2 leaq (%rsi,%r10), %r15 movq %r15, %rax imulq %rcx, %rax vmovups (%rdx,%rax), %xmm4 leaq (%rsi,%r10), %r13 incq %r13 movq %r13, %rax imulq %rcx, %rax vmovups (%rdx,%rax), %xmm1 movq %r9, %r11 imulq %rsi, %r11 movq %r9, %rax imulq %r15, %rax movq %rax, 0x240(%rsp) movq %r9, %r8 imulq %r13, %r8 xorl %ebx, %ebx cmpw $0x0, 0x140(%rsp) setns %bl leaq (%rsi,%rbx), %rax incq %rax movq %r9, %rdi imulq %rax, %rdi imulq %rcx, %rax vmovups (%rdx,%rax), %xmm6 addq %r13, %rbx movq %rbx, %rax imulq %rcx, %rax vmovups (%rdx,%rax), %xmm7 movq %r9, %rsi imulq %rbx, %rsi cmpw $0x0, 0xe0(%rsp) movl $0x0, %eax cmovnsq %r10, %rax addq %rax, %r15 movq %r9, %r10 imulq %r15, %r10 imulq %rcx, %r15 vmovups (%rdx,%r15), %xmm8 addq %rax, %r13 movq %r9, %r15 imulq %r13, %r15 imulq %rcx, %r13 vmovups (%rdx,%r13), %xmm9 addq %rbx, %rax movq 0x110(%rsp), %rbx imulq %rax, %rcx vmovups (%rdx,%rcx), %xmm10 movq 0x80(%rsp), %rcx movq 0xa0(%rsp), %rdx movq 0x38(%rdx,%rcx), %rcx vmovups (%rcx,%r11), %xmm11 movq 0x60(%rsp), %rdx vmovups (%rcx,%rdx), %xmm12 movq 0x240(%rsp), %rdx vmovups (%rcx,%rdx), %xmm13 vmovups (%rcx,%r8), %xmm14 vmovups (%rcx,%rdi), %xmm15 vsubss %xmm5, %xmm0, %xmm0 vmovups (%rcx,%rsi), %xmm5 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vsubps %xmm3, %xmm11, %xmm11 vmulps %xmm0, %xmm11, %xmm11 vaddps %xmm3, %xmm11, %xmm11 vmovups (%rcx,%r10), %xmm3 vsubps %xmm2, %xmm12, %xmm12 vmulps %xmm0, %xmm12, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmovups (%rcx,%r15), %xmm12 imulq %r9, %rax vsubps %xmm4, %xmm13, %xmm13 vmulps %xmm0, %xmm13, %xmm13 vaddps %xmm4, %xmm13, %xmm4 vmovups (%rcx,%rax), %xmm13 vsubps %xmm1, %xmm14, %xmm14 vmulps %xmm0, %xmm14, %xmm14 vaddps %xmm1, %xmm14, %xmm1 vsubps %xmm6, %xmm15, %xmm14 vmulps %xmm0, %xmm14, %xmm14 vaddps %xmm6, %xmm14, %xmm6 vsubps %xmm7, %xmm5, %xmm5 vmulps %xmm5, %xmm0, %xmm5 vaddps %xmm5, %xmm7, %xmm5 vsubps %xmm8, %xmm3, %xmm3 vmulps %xmm3, %xmm0, %xmm3 vaddps %xmm3, %xmm8, %xmm3 vsubps %xmm9, %xmm12, %xmm7 vmulps %xmm7, %xmm0, %xmm7 vaddps %xmm7, %xmm9, %xmm7 vsubps %xmm10, %xmm13, %xmm8 vmulps %xmm0, %xmm8, %xmm0 vaddps %xmm0, %xmm10, %xmm0 vunpcklps %xmm1, %xmm11, %xmm8 # xmm8 = xmm11[0],xmm1[0],xmm11[1],xmm1[1] vunpckhps %xmm1, %xmm11, %xmm9 # xmm9 = xmm11[2],xmm1[2],xmm11[3],xmm1[3] vunpcklps %xmm4, %xmm2, %xmm10 # xmm10 = xmm2[0],xmm4[0],xmm2[1],xmm4[1] vunpckhps %xmm4, %xmm2, %xmm11 # xmm11 = xmm2[2],xmm4[2],xmm2[3],xmm4[3] vunpcklps %xmm11, %xmm9, %xmm9 # xmm9 = xmm9[0],xmm11[0],xmm9[1],xmm11[1] vunpcklps %xmm10, %xmm8, %xmm11 # xmm11 = xmm8[0],xmm10[0],xmm8[1],xmm10[1] vunpckhps %xmm10, %xmm8, %xmm8 # xmm8 = xmm8[2],xmm10[2],xmm8[3],xmm10[3] vunpcklps %xmm5, %xmm2, %xmm10 # xmm10 = xmm2[0],xmm5[0],xmm2[1],xmm5[1] vunpckhps %xmm5, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm5[2],xmm2[3],xmm5[3] vunpcklps %xmm1, %xmm6, %xmm12 # xmm12 = xmm6[0],xmm1[0],xmm6[1],xmm1[1] vunpckhps %xmm1, %xmm6, %xmm6 # xmm6 = xmm6[2],xmm1[2],xmm6[3],xmm1[3] vunpcklps %xmm6, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] vunpcklps %xmm12, %xmm10, %xmm6 # xmm6 = xmm10[0],xmm12[0],xmm10[1],xmm12[1] vunpckhps %xmm12, %xmm10, %xmm10 # xmm10 = xmm10[2],xmm12[2],xmm10[3],xmm12[3] vunpcklps %xmm0, %xmm1, %xmm12 # xmm12 = xmm1[0],xmm0[0],xmm1[1],xmm0[1] vunpckhps %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[2],xmm0[2],xmm1[3],xmm0[3] vunpcklps %xmm7, %xmm5, %xmm13 # xmm13 = xmm5[0],xmm7[0],xmm5[1],xmm7[1] vunpckhps %xmm7, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm7[2],xmm5[3],xmm7[3] vunpcklps %xmm5, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm5[0],xmm0[1],xmm5[1] vunpcklps %xmm13, %xmm12, %xmm5 # xmm5 = xmm12[0],xmm13[0],xmm12[1],xmm13[1] vunpckhps %xmm13, %xmm12, %xmm12 # xmm12 = xmm12[2],xmm13[2],xmm12[3],xmm13[3] vunpcklps %xmm7, %xmm4, %xmm13 # xmm13 = xmm4[0],xmm7[0],xmm4[1],xmm7[1] vunpckhps %xmm7, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm7[2],xmm4[3],xmm7[3] vunpcklps %xmm3, %xmm1, %xmm7 # xmm7 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vunpckhps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm3[2],xmm1[3],xmm3[3] vunpcklps %xmm1, %xmm4, %xmm1 # xmm1 = xmm4[0],xmm1[0],xmm4[1],xmm1[1] vunpcklps %xmm7, %xmm13, %xmm3 # xmm3 = xmm13[0],xmm7[0],xmm13[1],xmm7[1] vunpckhps %xmm7, %xmm13, %xmm4 # xmm4 = xmm13[2],xmm7[2],xmm13[3],xmm7[3] vinsertf128 $0x1, %xmm5, %ymm11, %ymm5 vinsertf128 $0x1, %xmm12, %ymm8, %ymm7 vinsertf128 $0x1, %xmm0, %ymm9, %ymm0 vinsertf128 $0x1, %xmm6, %ymm6, %ymm6 vinsertf128 $0x1, %xmm10, %ymm10, %ymm13 vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vinsertf128 $0x1, %xmm4, %ymm4, %ymm9 vbroadcastss (%rbx), %ymm4 vbroadcastss 0x4(%rbx), %ymm10 vinsertf128 $0x1, %xmm1, %ymm1, %ymm11 vbroadcastss 0x8(%rbx), %ymm12 vsubps %ymm4, %ymm5, %ymm1 vsubps %ymm10, %ymm7, %ymm14 vsubps %ymm12, %ymm0, %ymm5 vsubps %ymm4, %ymm6, %ymm0 vmovaps %ymm0, 0xa0(%rsp) vsubps %ymm10, %ymm13, %ymm13 vsubps %ymm12, %ymm2, %ymm8 vsubps %ymm4, %ymm3, %ymm4 vsubps %ymm10, %ymm9, %ymm3 vsubps %ymm12, %ymm11, %ymm2 vmovaps %ymm2, 0x60(%rsp) vsubps %ymm1, %ymm4, %ymm12 vsubps %ymm14, %ymm3, %ymm15 vsubps %ymm5, %ymm2, %ymm0 vaddps %ymm3, %ymm14, %ymm6 vaddps %ymm5, %ymm2, %ymm7 vmulps %ymm6, %ymm0, %ymm9 vmulps %ymm7, %ymm15, %ymm11 vsubps %ymm9, %ymm11, %ymm10 vaddps %ymm1, %ymm4, %ymm9 vmulps %ymm7, %ymm12, %ymm7 vmovaps %ymm0, 0x4a0(%rsp) vmulps %ymm0, %ymm9, %ymm11 vsubps %ymm7, %ymm11, %ymm7 vmovaps %ymm15, 0x280(%rsp) vmulps %ymm9, %ymm15, %ymm11 vmovaps %ymm12, 0x360(%rsp) vmulps %ymm6, %ymm12, %ymm6 vbroadcastss 0x18(%rbx), %ymm12 vsubps %ymm11, %ymm6, %ymm6 vmulps %ymm6, %ymm12, %ymm6 vbroadcastss 0x14(%rbx), %ymm0 vmovaps %ymm0, 0x80(%rsp) vmulps %ymm7, %ymm0, %ymm7 vaddps %ymm6, %ymm7, %ymm6 vbroadcastss 0x10(%rbx), %ymm0 vmulps %ymm0, %ymm10, %ymm7 vaddps %ymm6, %ymm7, %ymm11 vmovaps %ymm13, %ymm2 vsubps %ymm13, %ymm14, %ymm13 vsubps %ymm8, %ymm5, %ymm15 vmovaps %ymm14, 0x240(%rsp) vaddps %ymm2, %ymm14, %ymm6 vmovaps %ymm5, 0x140(%rsp) vaddps %ymm5, %ymm8, %ymm7 vmulps %ymm6, %ymm15, %ymm14 vmulps %ymm7, %ymm13, %ymm5 vsubps %ymm14, %ymm5, %ymm5 vmovaps 0xa0(%rsp), %ymm9 vsubps %ymm9, %ymm1, %ymm14 vmulps %ymm7, %ymm14, %ymm7 vmovaps %ymm1, 0xe0(%rsp) vaddps %ymm1, %ymm9, %ymm10 vmovaps %ymm9, %ymm1 vmulps %ymm10, %ymm15, %ymm9 vsubps %ymm7, %ymm9, %ymm7 vmovaps %ymm13, 0x480(%rsp) vmulps %ymm10, %ymm13, %ymm9 vmulps %ymm6, %ymm14, %ymm6 vsubps %ymm9, %ymm6, %ymm6 vmulps %ymm6, %ymm12, %ymm6 vmovaps 0x80(%rsp), %ymm9 vmulps %ymm7, %ymm9, %ymm7 vaddps %ymm6, %ymm7, %ymm6 vmulps %ymm5, %ymm0, %ymm5 vaddps %ymm6, %ymm5, %ymm10 vsubps %ymm4, %ymm1, %ymm13 vaddps %ymm4, %ymm1, %ymm4 vsubps %ymm3, %ymm2, %ymm7 vaddps %ymm3, %ymm2, %ymm3 vmovaps 0x60(%rsp), %ymm1 vsubps %ymm1, %ymm8, %ymm6 vaddps %ymm1, %ymm8, %ymm1 vmulps %ymm3, %ymm6, %ymm2 vmulps %ymm1, %ymm7, %ymm5 vsubps %ymm2, %ymm5, %ymm2 vmulps %ymm1, %ymm13, %ymm1 vmulps %ymm4, %ymm6, %ymm5 vsubps %ymm1, %ymm5, %ymm1 vmulps %ymm4, %ymm7, %ymm4 vmulps %ymm3, %ymm13, %ymm3 vsubps %ymm4, %ymm3, %ymm3 vmovaps %ymm12, 0xa0(%rsp) vmulps %ymm3, %ymm12, %ymm3 vmulps %ymm1, %ymm9, %ymm1 vaddps %ymm3, %ymm1, %ymm1 vmovaps %ymm0, 0x60(%rsp) vmulps %ymm2, %ymm0, %ymm2 vaddps %ymm1, %ymm2, %ymm1 vaddps %ymm10, %ymm11, %ymm2 vaddps %ymm2, %ymm1, %ymm9 vminps %ymm10, %ymm11, %ymm2 vminps %ymm1, %ymm2, %ymm2 vbroadcastss 0x1d0bb69(%rip), %ymm3 # 0x1f20ec4 vandps %ymm3, %ymm9, %ymm4 vbroadcastss 0x1d0bb64(%rip), %ymm3 # 0x1f20ecc vmovaps %ymm4, 0x3a0(%rsp) vmulps %ymm3, %ymm4, %ymm3 vbroadcastss 0x1d0bb42(%rip), %ymm4 # 0x1f20ec0 vxorps %ymm4, %ymm3, %ymm4 vcmpnltps %ymm4, %ymm2, %ymm2 vmovaps %ymm11, 0x460(%rsp) vmovaps %ymm10, 0x440(%rsp) vmaxps %ymm10, %ymm11, %ymm4 vmaxps %ymm1, %ymm4, %ymm1 vcmpleps %ymm3, %ymm1, %ymm1 vorps %ymm1, %ymm2, %ymm0 leaq 0x7(%rsp), %rax movq %rax, 0x520(%rsp) vtestps 0x380(%rsp), %ymm0 je 0x215ac4 vmovaps 0x4a0(%rsp), %ymm10 vmovaps %ymm0, 0x420(%rsp) vmovaps 0x480(%rsp), %ymm0 vmulps %ymm0, %ymm10, %ymm1 vmovaps 0x280(%rsp), %ymm12 vmulps %ymm15, %ymm12, %ymm2 vsubps %ymm1, %ymm2, %ymm2 vmulps %ymm7, %ymm15, %ymm3 vmulps %ymm6, %ymm0, %ymm4 vsubps %ymm3, %ymm4, %ymm4 vbroadcastss 0x1d0bab6(%rip), %ymm5 # 0x1f20ec4 vmovaps %ymm9, %ymm11 vmovaps %ymm5, %ymm9 vandps %ymm5, %ymm1, %ymm1 vandps %ymm5, %ymm3, %ymm3 vcmpltps %ymm3, %ymm1, %ymm1 vblendvps %ymm1, %ymm2, %ymm4, %ymm1 vmulps %ymm6, %ymm14, %ymm2 vmulps %ymm14, %ymm10, %ymm3 vmovaps 0x360(%rsp), %ymm8 vmulps %ymm15, %ymm8, %ymm4 vsubps %ymm4, %ymm3, %ymm3 vmulps %ymm13, %ymm15, %ymm5 vsubps %ymm2, %ymm5, %ymm5 vandps %ymm4, %ymm9, %ymm4 vandps %ymm2, %ymm9, %ymm2 vcmpltps %ymm2, %ymm4, %ymm2 vblendvps %ymm2, %ymm3, %ymm5, %ymm10 vmulps %ymm0, %ymm13, %ymm2 vmulps %ymm0, %ymm8, %ymm3 vmulps %ymm14, %ymm12, %ymm4 vmulps %ymm7, %ymm14, %ymm0 vsubps %ymm4, %ymm3, %ymm3 vsubps %ymm2, %ymm0, %ymm0 vandps %ymm4, %ymm9, %ymm4 vandps %ymm2, %ymm9, %ymm2 vcmpltps %ymm2, %ymm4, %ymm2 vblendvps %ymm2, %ymm3, %ymm0, %ymm0 vmulps 0xa0(%rsp), %ymm0, %ymm2 vmulps 0x80(%rsp), %ymm10, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vmulps 0x60(%rsp), %ymm1, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vaddps %ymm2, %ymm2, %ymm2 vmulps 0x140(%rsp), %ymm0, %ymm3 vmulps 0x240(%rsp), %ymm10, %ymm4 vaddps %ymm3, %ymm4, %ymm3 vmulps 0xe0(%rsp), %ymm1, %ymm4 vrcpps %ymm2, %ymm5 vaddps %ymm3, %ymm4, %ymm3 vaddps %ymm3, %ymm3, %ymm3 vmulps %ymm5, %ymm2, %ymm4 vbroadcastss 0x1cd722b(%rip), %ymm7 # 0x1eec714 vsubps %ymm4, %ymm7, %ymm4 vmulps %ymm4, %ymm5, %ymm4 vaddps %ymm4, %ymm5, %ymm4 vbroadcastss 0xc(%rbx), %ymm5 vmulps %ymm4, %ymm3, %ymm4 vcmpleps %ymm4, %ymm5, %ymm3 vbroadcastss 0x20(%rbx), %ymm5 vcmpleps %ymm5, %ymm4, %ymm5 vandps %ymm5, %ymm3, %ymm3 vcmpneqps 0x1d0b9e4(%rip), %ymm2, %ymm2 # 0x1f20f00 vandps %ymm3, %ymm2, %ymm2 vpslld $0x1f, %xmm2, %xmm3 vpsrad $0x1f, %xmm3, %xmm3 vextractf128 $0x1, %ymm2, %xmm2 vpslld $0x1f, %xmm2, %xmm2 vpsrad $0x1f, %xmm2, %xmm2 vinsertf128 $0x1, %xmm2, %ymm3, %ymm2 vmovaps 0x420(%rsp), %ymm3 vandps 0x380(%rsp), %ymm3, %ymm3 vtestps %ymm3, %ymm2 je 0x215ac4 vandps %ymm3, %ymm2, %ymm2 vmovaps %ymm11, 0x500(%rsp) leaq 0x7(%rsp), %rax movq %rax, 0x520(%rsp) vmovaps 0x440(%rsp), %ymm6 vsubps %ymm6, %ymm11, %ymm3 vmovaps 0x3c0(%rsp), %ymm5 vmovaps 0x460(%rsp), %ymm7 vblendvps %ymm5, %ymm3, %ymm7, %ymm3 vmovaps %ymm2, 0x540(%rsp) vsubps %ymm7, %ymm11, %ymm2 vblendvps %ymm5, %ymm2, %ymm6, %ymm2 vmovaps %ymm4, 0x5a0(%rsp) vmovaps %ymm3, 0x4c0(%rsp) vbroadcastss 0x1cdb402(%rip), %ymm4 # 0x1ef09cc vbroadcastss 0x1cd7141(%rip), %ymm6 # 0x1eec714 vblendvps %ymm5, %ymm4, %ymm6, %ymm4 vmovaps %ymm2, 0x4e0(%rsp) vmulps %ymm1, %ymm4, %ymm1 vmovaps %ymm1, 0x5c0(%rsp) vmulps %ymm4, %ymm10, %ymm1 vmovaps %ymm1, 0x5e0(%rsp) vmulps %ymm0, %ymm4, %ymm0 vmovaps %ymm0, 0x600(%rsp) vmovd 0x48(%rsp), %xmm0 vpshufd $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vpaddd 0x1d07374(%rip), %xmm0, %xmm0 # 0x1f1c990 vmovd 0x44(%rsp), %xmm1 movq 0x130(%rsp), %rcx movq 0x128(%rsp), %rdx movzwl 0x8(%rcx,%rdx), %eax decl %eax vcvtsi2ss %eax, %xmm13, %xmm4 vpshufd $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vpaddd 0x1d07356(%rip), %xmm1, %xmm1 # 0x1f1c9a0 vrcpss %xmm4, %xmm4, %xmm5 vmulss %xmm4, %xmm5, %xmm4 vmovss 0x1cdb99e(%rip), %xmm7 # 0x1ef0ff8 vsubss %xmm4, %xmm7, %xmm4 vmulss %xmm4, %xmm5, %xmm4 movzwl 0xa(%rcx,%rdx), %eax decl %eax vcvtsi2ss %eax, %xmm13, %xmm5 vrcpss %xmm5, %xmm5, %xmm6 vmulss %xmm5, %xmm6, %xmm5 vsubss %xmm5, %xmm7, %xmm5 vmulss %xmm5, %xmm6, %xmm5 vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vcvtdq2ps %ymm0, %ymm0 vmulps %ymm0, %ymm11, %ymm0 vaddps %ymm3, %ymm0, %ymm0 vshufps $0x0, %xmm4, %xmm4, %xmm3 # xmm3 = xmm4[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmulps %ymm3, %ymm0, %ymm0 vmovaps %ymm0, 0x4c0(%rsp) vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vcvtdq2ps %ymm1, %ymm1 vmulps %ymm1, %ymm11, %ymm1 vaddps %ymm2, %ymm1, %ymm1 vshufps $0x0, %xmm5, %xmm5, %xmm2 # xmm2 = xmm5[0,0,0,0] vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vmulps %ymm2, %ymm1, %ymm1 vmovaps %ymm1, 0x4e0(%rsp) movq 0x138(%rsp), %rax movq 0x1e8(%rax), %rax movq 0xd0(%rsp), %rcx movq (%rax,%rcx,8), %rsi movl 0x24(%rbx), %eax testl %eax, 0x34(%rsi) je 0x215ac4 movq 0x38(%rsp), %rax movq 0x10(%rax), %rax cmpq $0x0, 0x10(%rax) movq 0x58(%rsp), %r15 movq 0x50(%rsp), %r13 movq 0x30(%rsp), %rdi movq 0x28(%rsp), %r8 movq 0x20(%rsp), %r9 movq 0x18(%rsp), %r10 movq 0x10(%rsp), %r11 vmovdqa 0x260(%rsp), %ymm5 leaq 0x620(%rsp), %rdx movb 0x6(%rsp), %cl jne 0x21574b cmpq $0x0, 0x48(%rsi) je 0x215baa vrcpps %ymm11, %ymm2 vmulps %ymm2, %ymm11, %ymm3 vbroadcastss 0x1cd6fb7(%rip), %ymm4 # 0x1eec714 vsubps %ymm3, %ymm4, %ymm3 vmulps %ymm3, %ymm2, %ymm3 vaddps %ymm3, %ymm2, %ymm2 vbroadcastss 0x1cdb876(%rip), %ymm3 # 0x1ef0fe8 vmovaps 0x3a0(%rsp), %ymm6 vcmpnltps %ymm3, %ymm6, %ymm3 vandps %ymm2, %ymm3, %ymm2 vmulps %ymm2, %ymm0, %ymm0 vminps %ymm4, %ymm0, %ymm0 vmovaps %ymm0, 0x560(%rsp) vmulps %ymm2, %ymm1, %ymm0 vminps %ymm4, %ymm0, %ymm0 vmovaps %ymm0, 0x580(%rsp) vmovaps 0x540(%rsp), %ymm0 vmovmskps %ymm0, %ecx movq %rcx, 0x60(%rsp) movq 0x60(%rsp), %rcx bsfq %rcx, %rcx movq %rcx, 0xe0(%rsp) cmpl $0x0, 0x60(%rsp) setne 0x280(%rsp) vmovaps 0x220(%rsp), %ymm15 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x180(%rsp), %ymm14 vmovdqa 0x160(%rsp), %ymm12 je 0x215b3b movq %rax, 0x140(%rsp) movq %rsi, 0xa0(%rsp) movq 0x38(%rsp), %rax movq 0x8(%rax), %rax movq %rax, 0x80(%rsp) movq 0x8(%rsp), %rsi movq 0xe0(%rsp), %rax movb 0x6(%rsp), %cl vmovss 0x560(%rsp,%rax,4), %xmm0 vmovss 0x580(%rsp,%rax,4), %xmm1 vmovss 0x20(%rbx), %xmm2 vmovss %xmm2, 0x240(%rsp) vmovss 0x5a0(%rsp,%rax,4), %xmm2 vmovss %xmm2, 0x20(%rbx) vmovss 0x5c0(%rsp,%rax,4), %xmm2 vmovss 0x5e0(%rsp,%rax,4), %xmm3 movq %rax, 0xe0(%rsp) vmovss 0x600(%rsp,%rax,4), %xmm4 vmovss %xmm2, 0x1c0(%rsp) vmovss %xmm3, 0x1c4(%rsp) vmovss %xmm4, 0x1c8(%rsp) vmovss %xmm0, 0x1cc(%rsp) vmovss %xmm1, 0x1d0(%rsp) movq 0x100(%rsp), %rax movl %eax, 0x1d4(%rsp) movq 0xd0(%rsp), %rax movl %eax, 0x1d8(%rsp) movq 0x80(%rsp), %rax movl (%rax), %eax movl %eax, 0x1dc(%rsp) movq 0x80(%rsp), %rax movl 0x4(%rax), %eax movl %eax, 0x1e0(%rsp) orl $-0x1, 0x4c(%rsp) leaq 0x4c(%rsp), %rax movq %rax, 0x1f0(%rsp) movq 0xa0(%rsp), %rax movq 0x18(%rax), %rax movq %rax, 0x1f8(%rsp) movq 0x80(%rsp), %rax movq %rax, 0x200(%rsp) movq %rbx, 0x208(%rsp) leaq 0x1c0(%rsp), %rax movq %rax, 0x210(%rsp) movl $0x1, 0x218(%rsp) movq 0xa0(%rsp), %rax movq 0x48(%rax), %rax testq %rax, %rax je 0x2159d0 leaq 0x1f0(%rsp), %rdi vzeroupper callq %rax movq 0x8(%rsp), %rsi movb 0x6(%rsp), %cl leaq 0x620(%rsp), %rdx vmovdqa 0x260(%rsp), %ymm5 vmovdqa 0x160(%rsp), %ymm12 movq 0x10(%rsp), %r11 movq 0x18(%rsp), %r10 movq 0x20(%rsp), %r9 movq 0x28(%rsp), %r8 vmovaps 0x180(%rsp), %ymm14 movq 0x30(%rsp), %rdi vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x220(%rsp), %ymm15 movq 0x1f0(%rsp), %rax cmpl $0x0, (%rax) je 0x215a82 movq 0x140(%rsp), %rax movq 0x10(%rax), %rax testq %rax, %rax je 0x215b44 movq %rax, 0x360(%rsp) movq 0x140(%rsp), %rax testb $0x2, (%rax) jne 0x215a08 movq 0xa0(%rsp), %rax testb $0x40, 0x3e(%rax) je 0x215a71 leaq 0x1f0(%rsp), %rdi vzeroupper callq 0x360(%rsp) movq 0x8(%rsp), %rsi movb 0x6(%rsp), %cl leaq 0x620(%rsp), %rdx vmovdqa 0x260(%rsp), %ymm5 vmovdqa 0x160(%rsp), %ymm12 movq 0x10(%rsp), %r11 movq 0x18(%rsp), %r10 movq 0x20(%rsp), %r9 movq 0x28(%rsp), %r8 vmovaps 0x180(%rsp), %ymm14 movq 0x30(%rsp), %rdi vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x220(%rsp), %ymm15 movq 0x1f0(%rsp), %rax cmpl $0x0, (%rax) jne 0x215b44 vmovss 0x240(%rsp), %xmm0 vmovss %xmm0, 0x20(%rbx) movq 0x60(%rsp), %rcx movq 0xe0(%rsp), %rax btcq %rax, %rcx bsfq %rcx, %rax movq %rcx, 0x60(%rsp) testq %rcx, %rcx movb 0x6(%rsp), %cl setne 0x280(%rsp) jne 0x215832 jmp 0x215b44 movq 0x58(%rsp), %r15 vmovaps 0x220(%rsp), %ymm15 vmovaps 0x1a0(%rsp), %ymm13 movq 0x50(%rsp), %r13 movq 0x30(%rsp), %rdi vmovaps 0x180(%rsp), %ymm14 movq 0x28(%rsp), %r8 movq 0x20(%rsp), %r9 movq 0x18(%rsp), %r10 movq 0x10(%rsp), %r11 vmovdqa 0x160(%rsp), %ymm12 vmovdqa 0x260(%rsp), %ymm5 movq 0x8(%rsp), %rsi movq 0x118(%rsp), %rdx leaq -0x1(%rdx), %rax andq %rax, %rdx movq %rdx, %rax movq 0x108(%rsp), %rdx jne 0x214df6 jmp 0x215b53 movq 0x8(%rsp), %rsi movb 0x6(%rsp), %cl testb $0x1, 0x280(%rsp) je 0x215b19 jmp 0x215c23 movq 0x120(%rsp), %rax incq %rax cmpq %rsi, %rax setb %cl vmovaps 0x340(%rsp), %ymm6 vmovaps 0x2e0(%rsp), %ymm7 vmovaps 0x2c0(%rsp), %ymm8 vmovaps 0x2a0(%rsp), %ymm9 vmovaps 0x320(%rsp), %ymm10 vmovaps 0x300(%rsp), %ymm11 leaq 0x620(%rsp), %rdx jne 0x214937 jmp 0x215c04 vmovaps 0x220(%rsp), %ymm15 vmovaps 0x1a0(%rsp), %ymm13 vmovaps 0x340(%rsp), %ymm6 vmovaps 0x2e0(%rsp), %ymm7 vmovaps 0x2c0(%rsp), %ymm8 vmovaps 0x2a0(%rsp), %ymm9 vmovaps 0x320(%rsp), %ymm10 vmovaps 0x300(%rsp), %ymm11 vmovaps 0x180(%rsp), %ymm14 vmovdqa 0x160(%rsp), %ymm12 xorl %eax, %eax testb $0x1, %cl je 0x215c15 movl $0xff800000, 0x20(%rbx) # imm = 0xFF800000 pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x214746 jmp 0x21452b vmovaps 0x340(%rsp), %ymm6 vmovaps 0x2e0(%rsp), %ymm7 vmovaps 0x2c0(%rsp), %ymm8 vmovaps 0x2a0(%rsp), %ymm9 vmovaps 0x320(%rsp), %ymm10 vmovaps 0x300(%rsp), %ymm11 jmp 0x215c04 nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx::BVHNIntersector1<8, 1, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMIntersector1Moeller<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	__forceinline vfloat4 select(const vboolf4& m, const vfloat4& t, const vfloat4& f) { #if defined(__AVX512VL__) return _mm_mask_blend_ps(m, f, t); #elif defined(__SSE4_1__) \|\| (defined(__aarch64__)) return _mm_blendv_ps(f, t, m); #else return _mm_or_ps(_mm_and_ps(m, t), _mm_andnot_ps(m, f)); #endif }	vbroadcastss 0x1cd4a18(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm0, %xmm3, %xmm0 vshufps $0xb1, %xmm0, %xmm0, %xmm3 # xmm3 = xmm0[1,0,3,2] vminps %xmm0, %xmm3, %xmm3 vshufpd $0x1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0] vminps %xmm3, %xmm4, %xmm3 vcmpeqps %xmm3, %xmm0, %xmm0 vtestps %xmm2, %xmm0 je 0x217030 vandps %xmm2, %xmm0, %xmm1 vmovmskps %xmm1, %eax bsfq %rax, %rax movq %rax, (%rdi) retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat4_sse2.h
embree::avx::BVHNIntersector1<8, 1, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMiIntersector1Moeller<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	__forceinline vfloat4 select(const vboolf4& m, const vfloat4& t, const vfloat4& f) { #if defined(__AVX512VL__) return _mm_mask_blend_ps(m, f, t); #elif defined(__SSE4_1__) \|\| (defined(__aarch64__)) return _mm_blendv_ps(f, t, m); #else return _mm_or_ps(_mm_and_ps(m, t), _mm_andnot_ps(m, f)); #endif }	vbroadcastss 0x1cd49db(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm0, %xmm3, %xmm0 vshufps $0xb1, %xmm0, %xmm0, %xmm3 # xmm3 = xmm0[1,0,3,2] vminps %xmm0, %xmm3, %xmm3 vshufpd $0x1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0] vminps %xmm3, %xmm4, %xmm3 vcmpeqps %xmm3, %xmm0, %xmm0 vtestps %xmm2, %xmm0 je 0x21706d vandps %xmm2, %xmm0, %xmm1 vmovmskps %xmm1, %eax bsfq %rax, %rax movq %rax, (%rdi) retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat4_sse2.h
embree::avx::BVHNIntersector1<8, 1, true, embree::avx::ArrayIntersector1<embree::avx::TriangleMiIntersector1Pluecker<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	__forceinline vfloat4 select(const vboolf4& m, const vfloat4& t, const vfloat4& f) { #if defined(__AVX512VL__) return _mm_mask_blend_ps(m, f, t); #elif defined(__SSE4_1__) \|\| (defined(__aarch64__)) return _mm_blendv_ps(f, t, m); #else return _mm_or_ps(_mm_and_ps(m, t), _mm_andnot_ps(m, f)); #endif }	vbroadcastss 0x1cd4961(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm0, %xmm3, %xmm0 vshufps $0xb1, %xmm0, %xmm0, %xmm3 # xmm3 = xmm0[1,0,3,2] vminps %xmm0, %xmm3, %xmm3 vshufpd $0x1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0] vminps %xmm3, %xmm4, %xmm3 vcmpeqps %xmm3, %xmm0, %xmm0 vtestps %xmm2, %xmm0 je 0x2170e7 vandps %xmm2, %xmm0, %xmm1 vmovmskps %xmm1, %eax bsfq %rax, %rax movq %rax, (%rdi) retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat4_sse2.h
embree::avx::BVHNIntersector1<8, 16777232, true, embree::avx::ArrayIntersector1<embree::avx::QuadMiMBIntersector1Pluecker<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	static __forceinline vfloat8 select(const vboolf8& m, const vfloat8& t, const vfloat8& f) { return _mm256_blendv_ps(f, t, m); }	vbroadcastss 0x1cd468b(%rip), %ymm2 # 0x1eeba20 vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vshufps $0xb1, %ymm0, %ymm0, %ymm2 # ymm2 = ymm0[1,0,3,2,5,4,7,6] vminps %ymm2, %ymm0, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vminps %ymm3, %ymm2, %ymm2 vperm2f128 $0x1, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[2,3,0,1] vminps %ymm3, %ymm2, %ymm2 vcmpeqps %ymm2, %ymm0, %ymm0 vtestps %ymm1, %ymm0 je 0x2173c7 vandps %ymm1, %ymm0, %ymm1 vmovmskps %ymm1, %eax bsfl %eax, %eax movl %eax, (%rdi) vzeroupper retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat8_avx.h
embree::avx::BVHNIntersector1<8, 1048576, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMiIntersector1Pluecker<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	__forceinline vfloat4 select(const vboolf4& m, const vfloat4& t, const vfloat4& f) { #if defined(__AVX512VL__) return _mm_mask_blend_ps(m, f, t); #elif defined(__SSE4_1__) \|\| (defined(__aarch64__)) return _mm_blendv_ps(f, t, m); #else return _mm_or_ps(_mm_and_ps(m, t), _mm_andnot_ps(m, f)); #endif }	vbroadcastss 0x1cd4643(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm0, %xmm3, %xmm0 vshufps $0xb1, %xmm0, %xmm0, %xmm3 # xmm3 = xmm0[1,0,3,2] vminps %xmm0, %xmm3, %xmm3 vshufpd $0x1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0] vminps %xmm3, %xmm4, %xmm3 vcmpeqps %xmm3, %xmm0, %xmm0 vtestps %xmm2, %xmm0 je 0x217405 vandps %xmm2, %xmm0, %xmm1 vmovmskps %xmm1, %eax bsfq %rax, %rax movq %rax, (%rdi) retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat4_sse2.h
embree::avx::BVHNIntersector1<8, 1048576, false, embree::avx::ArrayIntersector1<embree::avx::TriangleMIntersector1Moeller<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	__forceinline vfloat4 select(const vboolf4& m, const vfloat4& t, const vfloat4& f) { #if defined(__AVX512VL__) return _mm_mask_blend_ps(m, f, t); #elif defined(__SSE4_1__) \|\| (defined(__aarch64__)) return _mm_blendv_ps(f, t, m); #else return _mm_or_ps(_mm_and_ps(m, t), _mm_andnot_ps(m, f)); #endif }	vbroadcastss 0x1cd4606(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm0, %xmm3, %xmm0 vshufps $0xb1, %xmm0, %xmm0, %xmm3 # xmm3 = xmm0[1,0,3,2] vminps %xmm0, %xmm3, %xmm3 vshufpd $0x1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0] vminps %xmm3, %xmm4, %xmm3 vcmpeqps %xmm3, %xmm0, %xmm0 vtestps %xmm2, %xmm0 je 0x217442 vandps %xmm2, %xmm0, %xmm1 vmovmskps %xmm1, %eax bsfq %rax, %rax movq %rax, (%rdi) retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat4_sse2.h
embree::avx::BVHNIntersector1<8, 1048576, false, embree::avx::ArrayIntersector1<embree::avx::QuadMiIntersector1Pluecker<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	static __forceinline vfloat8 select(const vboolf8& m, const vfloat8& t, const vfloat8& f) { return _mm256_blendv_ps(f, t, m); }	vbroadcastss 0x1cd45c9(%rip), %ymm2 # 0x1eeba20 vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vshufps $0xb1, %ymm0, %ymm0, %ymm2 # ymm2 = ymm0[1,0,3,2,5,4,7,6] vminps %ymm2, %ymm0, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vminps %ymm3, %ymm2, %ymm2 vperm2f128 $0x1, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[2,3,0,1] vminps %ymm3, %ymm2, %ymm2 vcmpeqps %ymm2, %ymm0, %ymm0 vtestps %ymm1, %ymm0 je 0x217489 vandps %ymm1, %ymm0, %ymm1 vmovmskps %ymm1, %eax bsfl %eax, %eax movl %eax, (%rdi) vzeroupper retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat8_avx.h
embree::avx::PatchEval<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::eval(embree::PatchT<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::Ref, float const&, float const&, float, unsigned long) (.cold.1)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x217546 retq	/embree[P]embree/kernels/common/../../common/sys/array.h
embree::GeneralCatmullClarkPatchT<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::GeneralCatmullClarkPatchT(embree::HalfEdge const, char const, unsigned long) (.cold.1)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq $0x10, 0x38(%rdi) leaq 0x2c0(%rdi), %rax leaq 0x80(%rdi), %rcx xorl %edx, %edx leaq (%rcx,%rdx), %r8 movq %r8, 0x280(%rdi,%rdx) leaq (%rax,%rdx), %r8 movq %r8, 0x340(%rdi,%rdx) andq $0x0, 0x39c(%rdi,%rdx) andl $0x0, 0x3a4(%rdi,%rdx) addq $0x380, %rdx # imm = 0x380 cmpq $0x3800, %rdx # imm = 0x3800 jne 0x217578 addq $0x40, %rdi movq %rdi, (%rsi) retq	/embree[P]embree/kernels/common/../../common/sys/array.h
embree::GeneralCatmullClarkPatchT<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::GeneralCatmullClarkPatchT(embree::HalfEdge const, char const, unsigned long) (.cold.2)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x2175be retq	/embree[P]embree/kernels/common/../../common/sys/array.h
embree::GeneralCatmullClarkPatchT<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::GeneralCatmullClarkPatchT(embree::HalfEdge const, char const, unsigned long) (.cold.3)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x2175dd retq	/embree[P]embree/kernels/common/../../common/sys/array.h
embree::GeneralCatmullClarkPatchT<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::GeneralCatmullClarkPatchT(embree::HalfEdge const, char const, unsigned long) (.cold.4)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x2175fc retq	/embree[P]embree/kernels/common/../../common/sys/array.h
embree::CatmullClark1RingT<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::CatmullClark1RingT(embree::CatmullClark1RingT<embree::vfloat_impl<4>, embree::vfloat_impl<4>> const&) (.cold.2)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x217756 retq	/embree[P]embree/kernels/common/../../common/sys/array.h
embree::avx::PatchEvalSimd<embree::vboolf_impl<4>, embree::vint_impl<4>, embree::vfloat_impl<4>, embree::vfloat_impl<4>, embree::vfloat_impl<4>>::eval(embree::vboolf_impl<4> const&, embree::PatchT<embree::vfloat_impl<4>, embree::vfloat_impl<4>>::Ref, embree::vfloat_impl<4> const&, embree::vfloat_impl<4> const&, float, unsigned long) (.cold.1)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x217846 retq	/embree[P]embree/kernels/common/../../common/sys/array.h
embree::avx::evalGrid(embree::SubdivPatch1Base const&, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, float, float, float, float, float, embree::SubdivMesh const) (.cold.1)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq $0x10, 0x38(%rdi) leaq 0x2c0(%rdi), %rax leaq 0x80(%rdi), %rcx xorl %edx, %edx leaq (%rcx,%rdx), %r8 movq %r8, 0x280(%rdi,%rdx) leaq (%rax,%rdx), %r8 movq %r8, 0x340(%rdi,%rdx) andq $0x0, 0x39c(%rdi,%rdx) andl $0x0, 0x3a4(%rdi,%rdx) addq $0x380, %rdx # imm = 0x380 cmpq $0x3800, %rdx # imm = 0x3800 jne 0x2178d4 addq $0x40, %rdi movq %rdi, (%rsi) retq	/embree[P]embree/kernels/subdiv/../geometry/../common/../../common/sys/array.h
embree::avx::evalGrid(embree::SubdivPatch1Base const&, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, float, float, float, float, float, embree::SubdivMesh const) (.cold.2)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x21791a retq	/embree[P]embree/kernels/subdiv/../geometry/../common/../../common/sys/array.h
embree::avx::evalGrid(embree::SubdivPatch1Base const&, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, float, float, float, float, float, embree::SubdivMesh const) (.cold.3)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x217939 retq	/embree[P]embree/kernels/subdiv/../geometry/../common/../../common/sys/array.h
embree::avx::evalGrid(embree::SubdivPatch1Base const&, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, float, float, float, float, float, embree::SubdivMesh const) (.cold.4)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x217958 retq	/embree[P]embree/kernels/subdiv/../geometry/../common/../../common/sys/array.h
embree::avx::evalGridBounds(embree::SubdivPatch1Base const&, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, embree::SubdivMesh const*) (.cold.2)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x2179d0 retq	/embree[P]embree/kernels/subdiv/../geometry/../common/../../common/sys/array.h
embree::avx::evalGridBounds(embree::SubdivPatch1Base const&, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, embree::SubdivMesh const*) (.cold.3)	__forceinline void resize(size_t M) { assert(M <= max_total_elements); if (likely(M <= max_stack_elements)) return; if (likely(!isStackAllocated())) return; data = new Ty[max_total_elements]; for (size_t i=0; i<max_stack_elements; i++) data[i] = arr[i]; }	movq %rdi, (%rsi) xorl %eax, %eax movq (%rsi), %rcx vmovaps (%rdx,%rax), %xmm0 vmovaps %xmm0, (%rcx,%rax) addq $0x10, %rax cmpq $0x200, %rax # imm = 0x200 jne 0x2179ef retq	/embree[P]embree/kernels/subdiv/../geometry/../common/../../common/sys/array.h
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMvIntersectorKMoeller<4, 4, true>>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x2212de movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 cmpq $0x0, 0x8(%rcx) je 0x221111 movq 0x10(%r15), %rax testb $0x1, 0x2(%rax) jne 0x2212f0 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x2212de movzbl %al, %ebp vmovaps (%r14), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%r14), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%r14), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%r14), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cffd4e(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1ccfe65(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vandps %xmm4, %xmm2, %xmm5 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vrcpps %xmm1, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1ccb557(%rip), %xmm5 # 0x1eec714 vsubps %xmm1, %xmm5, %xmm1 vmulps %xmm1, %xmm4, %xmm1 vaddps %xmm1, %xmm4, %xmm1 vrcpps %xmm2, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm2, %xmm4, %xmm2 vrcpps %xmm3, %xmm4 vmulps %xmm4, %xmm3, %xmm3 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d3974f(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d33771(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d39730(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d39718(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d39713(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%r14), %xmm1 vmovaps 0x80(%r14), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cca795(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1ccb8e2(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x256180 popq %rax popq %rcx andq %rbx, %rbp jne 0x2212b0 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r12, %rsi movq %r14, %rdx movq %r15, %rcx addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x254a6e nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMvIntersectorKMoeller<4, 4, true>>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x19f8, %rsp # imm = 0x19F8 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x22253e movq %rcx, %r10 movq %rdx, %r14 movq 0x8(%rcx), %rax testq %rax, %rax je 0x22134c movq 0x10(%r10), %rcx testb $0x1, 0x2(%rcx) jne 0x222553 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%r14), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x22253e vandps %xmm3, %xmm4, %xmm8 vmovaps (%r14), %xmm3 vmovaps %xmm3, 0x200(%rsp) vmovaps 0x10(%r14), %xmm3 vmovaps %xmm3, 0x210(%rsp) vmovaps 0x20(%r14), %xmm3 vmovaps %xmm3, 0x220(%rsp) vmovaps 0x40(%r14), %xmm3 vmovaps %xmm3, 0x230(%rsp) vmovaps 0x50(%r14), %xmm4 vmovaps %xmm4, 0x240(%rsp) vmovaps 0x60(%r14), %xmm5 vmovaps %xmm5, 0x250(%rsp) vbroadcastss 0x1cffaed(%rip), %xmm9 # 0x1f20ec4 vandps %xmm3, %xmm9, %xmm6 vbroadcastss 0x1ccfc04(%rip), %xmm7 # 0x1ef0fe8 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm3, %xmm3 vandps %xmm4, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm4, %xmm4 vandps %xmm5, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm5, %xmm5 vrcpps %xmm3, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0x1ccb2f6(%rip), %xmm7 # 0x1eec714 vsubps %xmm3, %xmm7, %xmm3 vmulps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm6, %xmm3 vrcpps %xmm4, %xmm6 vmulps %xmm6, %xmm4, %xmm4 vsubps %xmm4, %xmm7, %xmm4 vmulps %xmm4, %xmm6, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm5 vsubps %xmm5, %xmm7, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm6, %xmm5 vmovaps %xmm3, 0x260(%rsp) vmovaps %xmm4, 0x270(%rsp) vmovaps %xmm5, 0x280(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d394e9(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x290(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d3350a(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d394c9(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vcmpnltps %xmm2, %xmm5, %xmm4 vbroadcastss 0x1d394b9(%rip), %xmm5 # 0x1f5a96c vbroadcastss 0x1d394b4(%rip), %xmm6 # 0x1f5a970 vblendvps %xmm4, %xmm5, %xmm6, %xmm4 vmovaps %xmm3, 0x2a0(%rsp) vmovaps %xmm4, 0x2b0(%rsp) vmovaps 0x30(%r14), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cca535(%rip), %xmm10 # 0x1eeba20 vblendvps %xmm8, %xmm3, %xmm10, %xmm1 vmovaps %xmm1, 0x2c0(%rsp) vbroadcastss 0x1ccb681(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm8, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x2d0(%rsp) vmovaps %xmm8, 0x300(%rsp) vxorps %xmm0, %xmm8, %xmm0 vmovaps %xmm0, 0xb0(%rsp) testq %rax, %rax je 0x221541 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r9d, %r9d cmpb $0x1, %al adcq $0x2, %r9 jmp 0x221545 pushq $0x3 popq %r9 leaq 0x80(%r14), %rax movq %rax, 0x98(%rsp) leaq 0x360(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xb00(%rsp), %rbx vmovaps %xmm10, -0x20(%rbx) movq 0x70(%r8), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%rbx) leaq 0x1f2e9fb(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x2f0(%rsp) vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x2e0(%rsp) vpcmpeqd %xmm11, %xmm11, %xmm11 addq $-0x10, %rbx movq -0x8(%r15), %r13 addq $-0x8, %r15 cmpq $-0x8, %r13 je 0x222438 vmovaps (%rbx), %xmm12 vcmpltps 0x2d0(%rsp), %xmm12, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x222449 movzbl %al, %ebp popcntl %ebp, %r12d xorl %eax, %eax cmpq %r9, %r12 jbe 0x22244d cmpq %r9, %r12 jbe 0x22243b testb $0x8, %r13b pushq $0x8 popq %r12 jne 0x22178f movq %r13, %rax movq %r13, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %r12, %r13 vmovaps %xmm10, %xmm12 movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x221728 vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x200(%rsp), %xmm1 vmovaps 0x210(%rsp), %xmm2 vmovaps 0x220(%rsp), %xmm3 vmovaps 0x260(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0x270(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x280(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vpminsd %xmm1, %xmm5, %xmm0 vpminsd %xmm2, %xmm7, %xmm4 vpmaxsd %xmm4, %xmm0, %xmm0 vpminsd %xmm3, %xmm9, %xmm4 vpmaxsd %xmm4, %xmm0, %xmm0 vpmaxsd %xmm1, %xmm5, %xmm1 vpmaxsd %xmm2, %xmm7, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpmaxsd %xmm3, %xmm9, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpmaxsd 0x2c0(%rsp), %xmm0, %xmm2 vpminsd 0x2d0(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vtestps %xmm1, %xmm1 je 0x221728 vblendvps %xmm1, %xmm0, %xmm10, %xmm0 cmpq $0x8, %r13 je 0x221721 movq %r13, (%r15) addq $0x8, %r15 vmovaps %xmm12, (%rbx) addq $0x10, %rbx vmovaps %xmm0, %xmm12 movq %rdi, %r13 cmpq $0x8, %rdi je 0x22173f leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x221616 xorl %eax, %eax cmpq $0x8, %r13 je 0x221788 vmovaps 0x2d0(%rsp), %xmm0 vcmpnleps %xmm12, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r9 jae 0x221779 testb %cl, %cl je 0x22243b testb $0x8, %r13b je 0x221602 jmp 0x22178f movq %r13, (%r15) addq $0x8, %r15 vmovaps %xmm12, (%rbx) addq $0x10, %rbx xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x221765 cmpq $-0x8, %r13 je 0x222438 vmovaps 0x2d0(%rsp), %xmm0 vcmpnleps %xmm12, %xmm0, %xmm0 vtestps %xmm0, %xmm0 movl $0xffffffff, %ecx # imm = 0xFFFFFFFF je 0x222449 movl %r13d, %edx andl $0xf, %edx vmovdqa 0xb0(%rsp), %xmm0 addq $-0x8, %rdx je 0x2223fa andq $-0x10, %r13 vpxor %xmm0, %xmm11, %xmm1 leaq 0xc0(%r13), %r12 xorl %edi, %edi imulq $0xe0, %rdi, %rax addq %r13, %rax movq %rax, 0x138(%rsp) xorl %ebp, %ebp vmovdqa %xmm1, %xmm0 vmovdqa %xmm1, 0x70(%rsp) movl (%r12,%rbp,4), %eax cmpq %rcx, %rax je 0x2223d0 vmovdqa %xmm0, 0x20(%rsp) vbroadcastss -0xc0(%r12,%rbp,4), %xmm0 vbroadcastss -0xb0(%r12,%rbp,4), %xmm1 vbroadcastss -0xa0(%r12,%rbp,4), %xmm2 vbroadcastss -0x90(%r12,%rbp,4), %xmm8 vbroadcastss -0x80(%r12,%rbp,4), %xmm4 vbroadcastss -0x70(%r12,%rbp,4), %xmm6 vbroadcastss -0x60(%r12,%rbp,4), %xmm3 vmovaps %xmm3, 0x110(%rsp) vbroadcastss -0x50(%r12,%rbp,4), %xmm3 vmovaps %xmm3, 0x100(%rsp) vbroadcastss -0x40(%r12,%rbp,4), %xmm3 vmovaps %xmm3, 0xc0(%rsp) vbroadcastss -0x30(%r12,%rbp,4), %xmm9 vbroadcastss -0x20(%r12,%rbp,4), %xmm7 vbroadcastss -0x10(%r12,%rbp,4), %xmm10 movq 0x138(%rsp), %r11 vmovaps 0xd0(%r11), %xmm3 vmovaps %xmm3, 0x160(%rsp) vmovaps %xmm4, 0x120(%rsp) vsubss %xmm4, %xmm1, %xmm5 vmovaps %xmm6, 0x340(%rsp) vsubss %xmm6, %xmm2, %xmm6 vmovaps %xmm7, 0x320(%rsp) vsubss %xmm1, %xmm7, %xmm3 vmovaps %xmm10, 0x310(%rsp) vsubss %xmm2, %xmm10, %xmm11 vmulss %xmm5, %xmm11, %xmm4 vmulss %xmm3, %xmm6, %xmm7 vsubss %xmm4, %xmm7, %xmm4 vmovaps %xmm4, 0x60(%rsp) vmovaps %xmm8, 0xa0(%rsp) vsubss %xmm8, %xmm0, %xmm10 vmovaps %xmm9, 0x330(%rsp) vsubss %xmm0, %xmm9, %xmm12 vmovaps %xmm6, 0x50(%rsp) vmulss %xmm6, %xmm12, %xmm4 vmulss %xmm11, %xmm10, %xmm7 vsubss %xmm4, %xmm7, %xmm13 vmulss %xmm3, %xmm10, %xmm4 vmovaps %xmm5, 0x30(%rsp) vmulss %xmm5, %xmm12, %xmm7 vsubss %xmm4, %xmm7, %xmm6 vsubps (%r14), %xmm0, %xmm5 vsubps 0x10(%r14), %xmm1, %xmm7 vsubps 0x20(%r14), %xmm2, %xmm8 vmovaps 0x50(%r14), %xmm2 vmovaps 0x60(%r14), %xmm1 vmulps %xmm2, %xmm8, %xmm14 vmulps %xmm1, %xmm7, %xmm15 vsubps %xmm14, %xmm15, %xmm4 vmovaps 0x40(%r14), %xmm0 vmulps %xmm1, %xmm5, %xmm14 vmulps %xmm0, %xmm8, %xmm15 vsubps %xmm14, %xmm15, %xmm14 vmulps %xmm0, %xmm7, %xmm15 vmulps %xmm2, %xmm5, %xmm9 vsubps %xmm15, %xmm9, %xmm15 vshufps $0x0, %xmm13, %xmm13, %xmm9 # xmm9 = xmm13[0,0,0,0] vshufps $0x0, %xmm6, %xmm6, %xmm13 # xmm13 = xmm6[0,0,0,0] vmovaps %xmm9, %xmm6 vmulps %xmm1, %xmm13, %xmm1 vmulps %xmm2, %xmm9, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vpermilps $0x0, 0x60(%rsp), %xmm2 # xmm2 = mem[0,0,0,0] vmulps %xmm0, %xmm2, %xmm0 vaddps %xmm1, %xmm0, %xmm9 vshufps $0x0, %xmm3, %xmm3, %xmm0 # xmm0 = xmm3[0,0,0,0] vshufps $0x0, %xmm11, %xmm11, %xmm1 # xmm1 = xmm11[0,0,0,0] vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm0, %xmm14, %xmm0 vaddps %xmm1, %xmm0, %xmm0 vshufps $0x0, %xmm12, %xmm12, %xmm1 # xmm1 = xmm12[0,0,0,0] vmovaps %xmm4, %xmm3 vmulps %xmm4, %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cff4eb(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm9, %xmm11 vxorps %xmm0, %xmm11, %xmm4 vmovaps 0x20(%rsp), %xmm0 vcmpnltps 0x1cca024(%rip), %xmm4, %xmm12 # 0x1eeba10 vtestps %xmm0, %xmm12 jne 0x221bad vtestps %xmm0, %xmm0 je 0x221b79 vmovaps 0x110(%rsp), %xmm7 vsubps 0x330(%rsp), %xmm7, %xmm5 vmovaps 0x100(%rsp), %xmm8 vsubps 0x320(%rsp), %xmm8, %xmm4 vmovaps 0xc0(%rsp), %xmm9 vsubps 0x310(%rsp), %xmm9, %xmm10 vmovaps 0xa0(%rsp), %xmm1 vsubps %xmm7, %xmm1, %xmm3 vmovaps 0x120(%rsp), %xmm1 vsubps %xmm8, %xmm1, %xmm11 vmovaps 0x340(%rsp), %xmm1 vsubps %xmm9, %xmm1, %xmm15 vmovaps %xmm0, 0x20(%rsp) vmulps %xmm4, %xmm15, %xmm0 vmulps %xmm10, %xmm11, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x60(%rsp) vmulps %xmm3, %xmm10, %xmm1 vmulps %xmm5, %xmm15, %xmm2 vsubps %xmm1, %xmm2, %xmm6 vmovaps %xmm5, 0xa0(%rsp) vmulps %xmm5, %xmm11, %xmm2 vmovaps %xmm4, 0x120(%rsp) vmulps %xmm4, %xmm3, %xmm5 vsubps %xmm2, %xmm5, %xmm5 vsubps (%r14), %xmm7, %xmm7 vsubps 0x10(%r14), %xmm8, %xmm8 vsubps 0x20(%r14), %xmm9, %xmm4 vmovaps 0x50(%r14), %xmm9 vmovaps 0x60(%r14), %xmm0 vmulps %xmm4, %xmm9, %xmm12 vmulps %xmm0, %xmm8, %xmm13 vsubps %xmm12, %xmm13, %xmm12 vmovaps 0x40(%r14), %xmm1 vmulps %xmm0, %xmm7, %xmm13 vmulps %xmm1, %xmm4, %xmm14 vsubps %xmm13, %xmm14, %xmm13 vmulps %xmm1, %xmm8, %xmm14 vmulps %xmm7, %xmm9, %xmm2 vsubps %xmm14, %xmm2, %xmm14 vmovaps %xmm5, 0x100(%rsp) vmulps %xmm0, %xmm5, %xmm0 vmovaps %xmm6, 0x110(%rsp) vmulps %xmm6, %xmm9, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vmovaps 0x60(%rsp), %xmm2 vmulps %xmm1, %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm9 vmulps %xmm14, %xmm15, %xmm0 vmulps %xmm13, %xmm11, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cff38f(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm9, %xmm11 vmulps %xmm3, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm3 vmovaps 0x20(%rsp), %xmm0 vcmpnltps 0x1cc9ec0(%rip), %xmm3, %xmm15 # 0x1eeba10 vtestps %xmm0, %xmm15 jne 0x221c0d vbroadcastss 0x1cc9ebc(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x70(%rsp), %xmm1 vtestps %xmm0, %xmm0 setne %al jmp 0x221b8f xorl %eax, %eax vbroadcastss 0x1cc9e9c(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x70(%rsp), %xmm1 testb %al, %al je 0x2223d0 leaq 0x1(%rbp), %rax cmpq $0x3, %rbp movq %rax, %rbp jb 0x221800 jmp 0x2223d0 vmovaps %xmm13, 0x150(%rsp) vmovaps %xmm6, %xmm13 vmovaps %xmm5, 0x60(%rsp) vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vpermilps $0x0, 0x30(%rsp), %xmm1 # xmm1 = mem[0,0,0,0] vpermilps $0x0, 0x50(%rsp), %xmm5 # xmm5 = mem[0,0,0,0] vandps %xmm0, %xmm12, %xmm6 vmulps %xmm5, %xmm15, %xmm5 vmulps %xmm1, %xmm14, %xmm1 vaddps %xmm5, %xmm1, %xmm1 vmulps %xmm3, %xmm10, %xmm0 vaddps %xmm1, %xmm0, %xmm0 vxorps %xmm0, %xmm11, %xmm5 vcmpnltps 0x1cc9e15(%rip), %xmm5, %xmm12 # 0x1eeba10 vtestps %xmm6, %xmm12 jne 0x221c6b vmovdqa 0x20(%rsp), %xmm0 jmp 0x2219f7 vmovaps %xmm4, %xmm5 vandps %xmm0, %xmm15, %xmm15 vmulps %xmm14, %xmm10, %xmm0 vmulps 0x120(%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0xa0(%rsp), %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm4 vcmpnltps 0x1cc9dcf(%rip), %xmm4, %xmm10 # 0x1eeba10 vtestps %xmm15, %xmm10 jne 0x221cd4 vbroadcastss 0x1cc9dcb(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x70(%rsp), %xmm1 vmovdqa 0x20(%rsp), %xmm0 jmp 0x221b6f movq %rdi, 0x18(%rsp) movq %rdx, 0x50(%rsp) movq %r9, 0x30(%rsp) movq %r8, 0x8(%rsp) movq %rsi, 0x10(%rsp) movq %r10, (%rsp) vbroadcastss 0x1cff233(%rip), %xmm0 # 0x1f20ec4 vandps %xmm0, %xmm9, %xmm10 vandps %xmm6, %xmm12, %xmm6 vsubps %xmm4, %xmm10, %xmm0 vcmpnltps %xmm5, %xmm0, %xmm12 vtestps %xmm6, %xmm12 jne 0x221d4c movq (%rsp), %r10 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %r8 movq 0x30(%rsp), %r9 movl $0xffffffff, %ecx # imm = 0xFFFFFFFF movq 0x50(%rsp), %rdx movq 0x18(%rsp), %rdi jmp 0x221c02 movq %rdi, 0x18(%rsp) movq %rdx, 0x50(%rsp) movq %r9, 0x30(%rsp) movq %r8, 0x8(%rsp) movq %rsi, 0x10(%rsp) movq %r10, (%rsp) vbroadcastss 0x1cff1ca(%rip), %xmm0 # 0x1f20ec4 vandps %xmm0, %xmm9, %xmm6 vandps %xmm15, %xmm10, %xmm10 vsubps %xmm3, %xmm6, %xmm0 vcmpnltps %xmm4, %xmm0, %xmm12 vtestps %xmm10, %xmm12 jne 0x221e58 movq (%rsp), %r10 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %r8 vbroadcastss 0x1cc9cf2(%rip), %xmm10 # 0x1eeba20 movq 0x30(%rsp), %r9 vpcmpeqd %xmm11, %xmm11, %xmm11 movl $0xffffffff, %ecx # imm = 0xFFFFFFFF movq 0x50(%rsp), %rdx movq 0x18(%rsp), %rdi jmp 0x221c5a vandps %xmm6, %xmm12, %xmm6 vmulps 0x150(%rsp), %xmm8, %xmm0 vmulps %xmm7, %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x60(%rsp), %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm0 vmulps 0x30(%r14), %xmm10, %xmm1 vcmpltps %xmm0, %xmm1, %xmm1 vmovaps 0x80(%r14), %xmm7 vmovaps %xmm7, 0x140(%rsp) vmulps %xmm7, %xmm10, %xmm7 vcmpleps %xmm7, %xmm0, %xmm7 vandps %xmm1, %xmm7, %xmm7 vtestps %xmm6, %xmm7 je 0x221cad vandps %xmm6, %xmm7, %xmm6 vcmpneqps 0x1cc9c5f(%rip), %xmm9, %xmm7 # 0x1eeba10 vtestps %xmm6, %xmm7 je 0x221cad vandps %xmm6, %xmm7, %xmm6 movq (%rsp), %rcx movq (%rcx), %rcx movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rcx movq %rcx, 0x60(%rsp) vbroadcastss 0x34(%rcx), %xmm1 vandps 0x90(%r14), %xmm1, %xmm1 vpcmpeqd 0x1cc9c22(%rip), %xmm1, %xmm7 # 0x1eeba10 vtestps %xmm6, %xmm7 jb 0x221cad movl 0x160(%rsp,%rbp,4), %ecx vandnps %xmm6, %xmm7, %xmm6 movq (%rsp), %rdx movq 0x10(%rdx), %rdx cmpq $0x0, 0x10(%rdx) jne 0x221f84 movq 0x60(%rsp), %rdx cmpq $0x0, 0x48(%rdx) jne 0x221f84 vmovdqa 0x20(%rsp), %xmm0 vpandn %xmm0, %xmm6, %xmm0 movq (%rsp), %r10 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %r8 movq 0x30(%rsp), %r9 movl $0xffffffff, %ecx # imm = 0xFFFFFFFF movq 0x50(%rsp), %rdx movq 0x18(%rsp), %rdi jmp 0x2219f7 vmovaps %xmm2, %xmm13 vandps %xmm10, %xmm12, %xmm10 vmulps 0x100(%rsp), %xmm5, %xmm0 vmulps 0x110(%rsp), %xmm8, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps %xmm7, %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm5 vmulps 0x30(%r14), %xmm6, %xmm0 vcmpltps %xmm5, %xmm0, %xmm0 vmovaps 0x80(%r14), %xmm1 vmovaps %xmm1, 0xa0(%rsp) vmulps %xmm1, %xmm6, %xmm1 vcmpleps %xmm1, %xmm5, %xmm1 vandps %xmm0, %xmm1, %xmm7 vtestps %xmm10, %xmm7 je 0x221d17 vandps %xmm7, %xmm10, %xmm7 vcmpneqps 0x1cc9b4b(%rip), %xmm9, %xmm8 # 0x1eeba10 vtestps %xmm7, %xmm8 je 0x221d17 vandps %xmm7, %xmm8, %xmm7 movq (%rsp), %rax movq (%rax), %rcx movl (%r12,%rbp,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rcx movq %rcx, 0xc0(%rsp) vbroadcastss 0x34(%rcx), %xmm0 vandps 0x90(%r14), %xmm0, %xmm0 vpcmpeqd 0x1cc9b07(%rip), %xmm0, %xmm8 # 0x1eeba10 vtestps %xmm7, %xmm8 jb 0x221d17 movl 0x160(%rsp,%rbp,4), %ecx vandnps %xmm7, %xmm8, %xmm7 movq (%rsp), %rdx movq 0x10(%rdx), %rdx cmpq $0x0, 0x10(%rdx) jne 0x22216c movq 0xc0(%rsp), %rdx cmpq $0x0, 0x48(%rdx) jne 0x22216c vmovdqa 0x20(%rsp), %xmm0 vpandn %xmm0, %xmm7, %xmm0 movq (%rsp), %r10 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %r8 vbroadcastss 0x1cc9aba(%rip), %xmm10 # 0x1eeba20 movq 0x30(%rsp), %r9 vpcmpeqd %xmm11, %xmm11, %xmm11 movl $0xffffffff, %ecx # imm = 0xFFFFFFFF movq 0x50(%rsp), %rdx movq 0x18(%rsp), %rdi jmp 0x221b69 vrcpps %xmm10, %xmm1 vmulps %xmm1, %xmm10, %xmm7 vbroadcastss 0x1cca77e(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm7, %xmm1, %xmm1 vmulps %xmm4, %xmm1, %xmm3 vminps %xmm8, %xmm3, %xmm3 vmulps %xmm5, %xmm1, %xmm5 vminps %xmm8, %xmm5, %xmm5 vsubps %xmm3, %xmm8, %xmm7 vsubps %xmm5, %xmm8, %xmm8 vmovaps 0x2f0(%rsp), %xmm9 vblendvps %xmm9, %xmm7, %xmm3, %xmm3 vblendvps %xmm9, %xmm8, %xmm5, %xmm5 movq (%rsp), %rsi movq 0x8(%rsi), %rdx vmovd %eax, %xmm7 vpshufd $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmovd %ecx, %xmm8 vpshufd $0x0, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] vmovaps %xmm2, 0x170(%rsp) vmovaps %xmm13, 0x180(%rsp) vmovaps 0x150(%rsp), %xmm2 vmovaps %xmm2, 0x190(%rsp) vmovaps %xmm3, 0x1a0(%rsp) vmovaps %xmm5, 0x1b0(%rsp) vmovdqa %xmm8, 0x1c0(%rsp) vmovdqa %xmm7, 0x1d0(%rsp) vmulps %xmm0, %xmm1, %xmm0 leaq 0x1e0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x1e0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x1f0(%rsp) vmovaps 0x140(%rsp), %xmm1 vblendvps %xmm6, %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm6, 0x80(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xd0(%rsp) movq 0x60(%rsp), %rcx movq 0x18(%rcx), %rax movq %rax, 0xd8(%rsp) movq 0x8(%rsi), %rax movq %rax, 0xe0(%rsp) movq %r14, 0xe8(%rsp) leaq 0x170(%rsp), %rax movq %rax, 0xf0(%rsp) movl $0x4, 0xf8(%rsp) movq 0x48(%rcx), %rax testq %rax, %rax je 0x2220ed leaq 0xd0(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x80(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x22235c movq (%rsp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x22212f testb $0x2, (%rcx) jne 0x222122 movq 0x60(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x22212f leaq 0xd0(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x80(%rsp), %xmm0, %xmm1 vpxor 0x1cc9cdc(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0xe8(%rsp), %rax vbroadcastss 0x1ccaa2f(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x22236c vrcpps %xmm6, %xmm0 vmulps %xmm0, %xmm6, %xmm1 vbroadcastss 0x1cca597(%rip), %xmm6 # 0x1eec714 vsubps %xmm1, %xmm6, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm3, %xmm0, %xmm1 vminps %xmm6, %xmm1, %xmm1 vmulps %xmm4, %xmm0, %xmm2 vminps %xmm6, %xmm2, %xmm2 vsubps %xmm1, %xmm6, %xmm3 vsubps %xmm2, %xmm6, %xmm4 vmovaps 0x2e0(%rsp), %xmm6 vblendvps %xmm6, %xmm3, %xmm1, %xmm1 vblendvps %xmm6, %xmm4, %xmm2, %xmm2 movq (%rsp), %rsi movq 0x8(%rsi), %rdx vmovd %eax, %xmm3 vpshufd $0x0, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmovd %ecx, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovaps %xmm13, 0x170(%rsp) vmovaps 0x110(%rsp), %xmm6 vmovaps %xmm6, 0x180(%rsp) vmovaps 0x100(%rsp), %xmm6 vmovaps %xmm6, 0x190(%rsp) vmovaps %xmm1, 0x1a0(%rsp) vmovaps %xmm2, 0x1b0(%rsp) vmovdqa %xmm4, 0x1c0(%rsp) vmovdqa %xmm3, 0x1d0(%rsp) vmulps %xmm5, %xmm0, %xmm0 leaq 0x1e0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x1e0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x1f0(%rsp) vmovaps 0xa0(%rsp), %xmm1 vblendvps %xmm7, %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm7, 0x80(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xd0(%rsp) movq 0xc0(%rsp), %rcx movq 0x18(%rcx), %rax movq %rax, 0xd8(%rsp) movq 0x8(%rsi), %rax movq %rax, 0xe0(%rsp) movq %r14, 0xe8(%rsp) leaq 0x170(%rsp), %rax movq %rax, 0xf0(%rsp) movl $0x4, 0xf8(%rsp) movq 0x48(%rcx), %rax testq %rax, %rax je 0x2222dd leaq 0xd0(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x80(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x222396 movq (%rsp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x222322 testb $0x2, (%rcx) jne 0x222315 movq 0xc0(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x222322 leaq 0xd0(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x80(%rsp), %xmm0, %xmm1 vpxor 0x1cc9ae9(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0xe8(%rsp), %rax vbroadcastss 0x1cca83c(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x2223a6 vpcmpeqd 0x1cc96ac(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cc9ab4(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm6 movq 0x98(%rsp), %rax vmovaps 0x140(%rsp), %xmm1 vblendvps %xmm0, (%rax), %xmm1, %xmm0 vmovaps %xmm0, (%rax) jmp 0x221e27 vpcmpeqd 0x1cc9672(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cc9a7a(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm7 movq 0x98(%rsp), %rax vmovaps 0xa0(%rsp), %xmm1 vblendvps %xmm0, (%rax), %xmm1, %xmm0 vmovaps %xmm0, (%rax) jmp 0x221f45 vmovaps %xmm0, %xmm2 vpand %xmm1, %xmm0, %xmm0 vtestps %xmm1, %xmm2 je 0x2223f6 incq %rdi addq $0xe0, %r12 vmovdqa %xmm0, %xmm1 cmpq %rdx, %rdi jb 0x2217e2 vpxor %xmm0, %xmm11, %xmm0 vpor 0xb0(%rsp), %xmm0, %xmm0 vmovdqa %xmm0, 0xb0(%rsp) vtestps %xmm11, %xmm0 jb 0x222438 vmovaps 0x2d0(%rsp), %xmm1 vbroadcastss 0x1cca75f(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x2d0(%rsp) xorl %eax, %eax jmp 0x22243b pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x2215a8 jmp 0x222516 pushq $0x2 jmp 0x22243a vmovaps %xmm12, 0x60(%rsp) movq %r9, 0x30(%rsp) movq %r10, (%rsp) movq %rsi, 0x10(%rsp) movq %r8, 0x8(%rsp) bsfq %rbp, %rcx movq %rsi, %rdi movq %r8, %rsi movq %r13, %rdx movq %rcx, 0x20(%rsp) leaq 0x4f(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x208(%rsp), %rax pushq %rax vzeroupper callq 0x258442 popq %rcx popq %rdx testb %al, %al je 0x2224a6 movq 0x20(%rsp), %rax orl $-0x1, 0xb0(%rsp,%rax,4) leaq -0x1(%rbp), %rax andq %rax, %rbp movq (%rsp), %r10 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %r8 jne 0x222466 vmovaps 0xb0(%rsp), %xmm0 vpcmpeqd %xmm11, %xmm11, %xmm11 vtestps %xmm11, %xmm0 pushq $0x3 popq %rax vbroadcastss 0x1cc9544(%rip), %xmm10 # 0x1eeba20 movq 0x30(%rsp), %r9 vmovaps 0x60(%rsp), %xmm12 jb 0x2215eb vmovaps 0x2d0(%rsp), %xmm1 vbroadcastss 0x1cca685(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x2d0(%rsp) pushq $0x2 popq %rax jmp 0x2215eb vmovaps 0x300(%rsp), %xmm0 vandps 0xb0(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cca653(%rip), %xmm1 # 0x1eecb84 movq 0x98(%rsp), %rax vmaskmovps %xmm1, %xmm0, (%rax) addq $0x19f8, %rsp # imm = 0x19F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %r14, %rdx movq %r10, %rcx addq $0x19f8, %rsp # imm = 0x19F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x256eb4 nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMvIntersectorKMoeller<4, 4, false>>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x22277a movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 cmpq $0x0, 0x8(%rcx) je 0x2225ad movq 0x10(%r15), %rax testb $0x1, 0x2(%rax) jne 0x22278c vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22277a movzbl %al, %ebp vmovaps (%r14), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%r14), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%r14), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%r14), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cfe8b2(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1cce9c9(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vandps %xmm4, %xmm2, %xmm5 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vrcpps %xmm1, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1cca0bb(%rip), %xmm5 # 0x1eec714 vsubps %xmm1, %xmm5, %xmm1 vmulps %xmm1, %xmm4, %xmm1 vaddps %xmm1, %xmm4, %xmm1 vrcpps %xmm2, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm2, %xmm4, %xmm2 vrcpps %xmm3, %xmm4 vmulps %xmm4, %xmm3, %xmm3 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d382b3(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d322d5(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d38294(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d3827c(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d38277(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%r14), %xmm1 vmovaps 0x80(%r14), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cc92f9(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cca446(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x25a26c popq %rax popq %rcx andq %rbx, %rbp jne 0x22274c addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r12, %rsi movq %r14, %rdx movq %r15, %rcx addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x258ece nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMvIntersectorKMoeller<4, 4, false>>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	movq (%rsi), %r11 cmpq $0x8, 0x70(%r11) je 0x2234c9 movq %rcx, %r10 movq %rdx, %r9 movq 0x8(%rcx), %rax testq %rax, %rax je 0x2227d7 movq 0x10(%r10), %rcx testb $0x1, 0x2(%rcx) jne 0x2234ca vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%r9), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x2234c9 pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1958, %rsp # imm = 0x1958 vandps %xmm3, %xmm4, %xmm8 vmovaps (%r9), %xmm3 vmovaps %xmm3, 0x150(%rsp) vmovaps 0x10(%r9), %xmm3 vmovaps %xmm3, 0x160(%rsp) vmovaps 0x20(%r9), %xmm3 vmovaps %xmm3, 0x170(%rsp) vmovaps 0x40(%r9), %xmm3 vmovaps %xmm3, 0x180(%rsp) vmovaps 0x50(%r9), %xmm4 vmovaps %xmm4, 0x190(%rsp) vmovaps 0x60(%r9), %xmm5 vmovaps %xmm5, 0x1a0(%rsp) vbroadcastss 0x1cfe651(%rip), %xmm9 # 0x1f20ec4 vandps %xmm3, %xmm9, %xmm6 vbroadcastss 0x1cce768(%rip), %xmm7 # 0x1ef0fe8 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm3, %xmm3 vandps %xmm4, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm4, %xmm4 vandps %xmm5, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm5, %xmm5 vrcpps %xmm3, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0x1cc9e5a(%rip), %xmm7 # 0x1eec714 vsubps %xmm3, %xmm7, %xmm3 vmulps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm6, %xmm3 vrcpps %xmm4, %xmm6 vmulps %xmm6, %xmm4, %xmm4 vsubps %xmm4, %xmm7, %xmm4 vmulps %xmm4, %xmm6, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm5 vsubps %xmm5, %xmm7, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm6, %xmm5 vmovaps %xmm3, 0x1b0(%rsp) vmovaps %xmm4, 0x1c0(%rsp) vmovaps %xmm5, 0x1d0(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d3804d(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x1e0(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d3206e(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d3802d(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vcmpnltps %xmm2, %xmm5, %xmm4 vbroadcastss 0x1d3801d(%rip), %xmm5 # 0x1f5a96c vbroadcastss 0x1d38018(%rip), %xmm6 # 0x1f5a970 vblendvps %xmm4, %xmm5, %xmm6, %xmm4 vmovaps %xmm3, 0x1f0(%rsp) vmovaps %xmm4, 0x200(%rsp) vmovaps 0x30(%r9), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cc9099(%rip), %xmm10 # 0x1eeba20 vblendvps %xmm8, %xmm3, %xmm10, %xmm1 vmovaps %xmm1, 0x210(%rsp) vbroadcastss 0x1cca1e5(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm8, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x220(%rsp) vmovaps %xmm8, 0x250(%rsp) vxorps %xmm0, %xmm8, %xmm0 vmovaps %xmm0, 0x50(%rsp) testq %rax, %rax je 0x2229da movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r13d, %r13d cmpb $0x1, %al adcq $0x2, %r13 jmp 0x2229de pushq $0x3 popq %r13 leaq 0x80(%r9), %r8 leaq 0x2c0(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xa60(%rsp), %rbx vmovaps %xmm10, -0x20(%rbx) movq 0x70(%r11), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%rbx) vpcmpeqd %xmm11, %xmm11, %xmm11 addq $-0x10, %rbx movq -0x8(%r15), %rbp addq $-0x8, %r15 cmpq $-0x8, %rbp je 0x2233a6 vmovaps (%rbx), %xmm12 vcmpltps 0x220(%rsp), %xmm12, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x2233b7 movzbl %al, %r12d popcntl %r12d, %r14d xorl %eax, %eax cmpq %r13, %r14 jbe 0x2233bb cmpq %r13, %r14 jbe 0x2233a9 testb $0x8, %bpl pushq $0x8 popq %r14 jne 0x222bfc movq %rbp, %rax movq %rbp, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %r14, %rbp vmovaps %xmm10, %xmm12 movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x222b95 vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x150(%rsp), %xmm1 vmovaps 0x160(%rsp), %xmm2 vmovaps 0x170(%rsp), %xmm3 vmovaps 0x1b0(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0x1c0(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x1d0(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vpminsd %xmm1, %xmm5, %xmm0 vpminsd %xmm2, %xmm7, %xmm4 vpmaxsd %xmm4, %xmm0, %xmm0 vpminsd %xmm3, %xmm9, %xmm4 vpmaxsd %xmm4, %xmm0, %xmm0 vpmaxsd %xmm1, %xmm5, %xmm1 vpmaxsd %xmm2, %xmm7, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpmaxsd %xmm3, %xmm9, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpmaxsd 0x210(%rsp), %xmm0, %xmm2 vpminsd 0x220(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vtestps %xmm1, %xmm1 je 0x222b95 vblendvps %xmm1, %xmm0, %xmm10, %xmm0 cmpq $0x8, %rbp je 0x222b8e movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm12, (%rbx) addq $0x10, %rbx vmovaps %xmm0, %xmm12 movq %rdi, %rbp cmpq $0x8, %rdi je 0x222bac leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x222a83 xorl %eax, %eax cmpq $0x8, %rbp je 0x222bf5 vmovaps 0x220(%rsp), %xmm0 vcmpnleps %xmm12, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r13 jae 0x222be6 testb %cl, %cl je 0x2233a9 testb $0x8, %bpl je 0x222a6f jmp 0x222bfc movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm12, (%rbx) addq $0x10, %rbx xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x222bd2 cmpq $-0x8, %rbp je 0x2233a6 vmovaps 0x220(%rsp), %xmm0 vmovaps %xmm0, 0x240(%rsp) vcmpnleps %xmm12, %xmm0, %xmm0 vtestps %xmm0, %xmm0 movl $0xffffffff, %r14d # imm = 0xFFFFFFFF je 0x2233b7 movl %ebp, %eax andl $0xf, %eax vmovdqa 0x50(%rsp), %xmm1 vmovdqa %xmm1, %xmm0 addq $-0x8, %rax je 0x223370 movq %r8, %r12 andq $-0x10, %rbp vpxor %xmm1, %xmm11, %xmm2 vmovaps (%r9), %xmm0 vmovaps %xmm0, 0x60(%rsp) vmovaps 0x10(%r9), %xmm0 vmovaps %xmm0, 0x140(%rsp) vmovaps 0x20(%r9), %xmm0 vmovaps %xmm0, 0x130(%rsp) vmovaps 0x30(%r9), %xmm0 vmovaps %xmm0, 0xf0(%rsp) vmovaps 0x50(%r9), %xmm0 vmovaps %xmm0, 0x120(%rsp) vmovaps 0x60(%r9), %xmm0 vmovaps %xmm0, 0x110(%rsp) vmovaps 0x40(%r9), %xmm0 vmovaps %xmm0, 0x100(%rsp) vmovaps 0x80(%r9), %xmm0 vmovaps %xmm0, 0xe0(%rsp) vmovaps 0x90(%r9), %xmm0 vmovaps %xmm0, 0xd0(%rsp) addq $0xc0, %rbp xorl %ecx, %ecx vmovdqa %xmm1, 0x20(%rsp) xorl %edx, %edx vmovdqa %xmm2, %xmm0 vmovdqa %xmm2, 0x30(%rsp) movl (%rbp,%rdx,4), %r8d cmpq %r14, %r8 je 0x223343 vbroadcastss -0xc0(%rbp,%rdx,4), %xmm4 vmovdqa %xmm0, (%rsp) vbroadcastss -0xb0(%rbp,%rdx,4), %xmm0 vbroadcastss -0xa0(%rbp,%rdx,4), %xmm3 vbroadcastss -0x90(%rbp,%rdx,4), %xmm1 vbroadcastss -0x80(%rbp,%rdx,4), %xmm2 vbroadcastss -0x70(%rbp,%rdx,4), %xmm5 vbroadcastss -0x30(%rbp,%rdx,4), %xmm9 vbroadcastss -0x20(%rbp,%rdx,4), %xmm6 vbroadcastss -0x10(%rbp,%rdx,4), %xmm7 vmovaps %xmm1, 0xa0(%rsp) vsubss %xmm1, %xmm4, %xmm1 vmovaps %xmm2, 0x90(%rsp) vsubss %xmm2, %xmm0, %xmm10 vmovaps %xmm5, 0x80(%rsp) vsubss %xmm5, %xmm3, %xmm2 vmovaps %xmm6, 0x2a0(%rsp) vsubss %xmm0, %xmm6, %xmm14 vmovaps %xmm7, 0x290(%rsp) vsubss %xmm3, %xmm7, %xmm5 vmulss %xmm5, %xmm10, %xmm6 vmulss %xmm2, %xmm14, %xmm7 vsubss %xmm6, %xmm7, %xmm6 vmovaps %xmm6, 0x40(%rsp) vmovaps %xmm9, 0x70(%rsp) vsubss %xmm4, %xmm9, %xmm15 vmovaps %xmm2, 0x270(%rsp) vmulss %xmm2, %xmm15, %xmm6 vmulss %xmm5, %xmm1, %xmm7 vsubss %xmm6, %xmm7, %xmm6 vmovaps %xmm1, 0x280(%rsp) vmulss %xmm1, %xmm14, %xmm7 vmovaps %xmm10, 0x260(%rsp) vmulss %xmm15, %xmm10, %xmm9 vsubss %xmm7, %xmm9, %xmm7 vsubps 0x140(%rsp), %xmm0, %xmm2 vsubps 0x130(%rsp), %xmm3, %xmm8 vmovaps 0x120(%rsp), %xmm11 vmulps %xmm11, %xmm8, %xmm9 vmovaps 0x110(%rsp), %xmm0 vmulps %xmm0, %xmm2, %xmm10 vsubps %xmm9, %xmm10, %xmm3 vsubps 0x60(%rsp), %xmm4, %xmm4 vmulps %xmm0, %xmm4, %xmm9 vmovaps 0x100(%rsp), %xmm1 vmulps %xmm1, %xmm8, %xmm10 vsubps %xmm9, %xmm10, %xmm9 vmulps %xmm1, %xmm2, %xmm10 vmulps %xmm4, %xmm11, %xmm12 vsubps %xmm10, %xmm12, %xmm10 vshufps $0x0, %xmm6, %xmm6, %xmm13 # xmm13 = xmm6[0,0,0,0] vshufps $0x0, %xmm7, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmulps %xmm0, %xmm7, %xmm12 vmovaps (%rsp), %xmm0 vmulps %xmm13, %xmm11, %xmm6 vaddps %xmm6, %xmm12, %xmm6 vpermilps $0x0, 0x40(%rsp), %xmm11 # xmm11 = mem[0,0,0,0] vmulps %xmm1, %xmm11, %xmm12 vaddps %xmm6, %xmm12, %xmm12 vshufps $0x0, %xmm14, %xmm14, %xmm6 # xmm6 = xmm14[0,0,0,0] vshufps $0x0, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vmulps %xmm5, %xmm10, %xmm5 vmulps %xmm6, %xmm9, %xmm6 vaddps %xmm5, %xmm6, %xmm5 vshufps $0x0, %xmm15, %xmm15, %xmm6 # xmm6 = xmm15[0,0,0,0] vmulps %xmm3, %xmm6, %xmm6 vaddps %xmm5, %xmm6, %xmm6 vbroadcastss 0x1cfe025(%rip), %xmm5 # 0x1f20ec0 vandps %xmm5, %xmm12, %xmm5 vxorps %xmm6, %xmm5, %xmm14 vcmpnltps 0x1cc8b64(%rip), %xmm14, %xmm15 # 0x1eeba10 vtestps %xmm0, %xmm15 jne 0x223074 vtestps %xmm0, %xmm0 je 0x223039 vbroadcastss -0x60(%rbp,%rdx,4), %xmm2 vbroadcastss -0x50(%rbp,%rdx,4), %xmm5 vbroadcastss -0x40(%rbp,%rdx,4), %xmm6 vsubps 0x70(%rsp), %xmm2, %xmm9 vsubps 0x2a0(%rsp), %xmm5, %xmm12 vsubps 0x290(%rsp), %xmm6, %xmm3 vmovaps 0xa0(%rsp), %xmm1 vsubps %xmm2, %xmm1, %xmm8 vmovaps 0x90(%rsp), %xmm1 vsubps %xmm5, %xmm1, %xmm10 vmovaps 0x80(%rsp), %xmm1 vsubps %xmm6, %xmm1, %xmm15 vmovaps %xmm0, (%rsp) vmulps %xmm12, %xmm15, %xmm0 vmulps %xmm3, %xmm10, %xmm1 vsubps %xmm0, %xmm1, %xmm11 vmovaps %xmm3, 0x80(%rsp) vmulps %xmm3, %xmm8, %xmm1 vmulps %xmm9, %xmm15, %xmm3 vsubps %xmm1, %xmm3, %xmm4 vmovaps %xmm9, 0x40(%rsp) vmulps %xmm9, %xmm10, %xmm3 vmovaps %xmm12, 0xa0(%rsp) vmulps %xmm12, %xmm8, %xmm9 vsubps %xmm3, %xmm9, %xmm7 vsubps 0x60(%rsp), %xmm2, %xmm2 vsubps 0x140(%rsp), %xmm5, %xmm5 vsubps 0x130(%rsp), %xmm6, %xmm6 vmovaps 0x120(%rsp), %xmm0 vmulps %xmm0, %xmm6, %xmm9 vmovaps 0x110(%rsp), %xmm1 vmulps %xmm1, %xmm5, %xmm12 vsubps %xmm9, %xmm12, %xmm12 vmulps %xmm1, %xmm2, %xmm9 vmovaps 0x100(%rsp), %xmm3 vmulps %xmm3, %xmm6, %xmm13 vsubps %xmm9, %xmm13, %xmm13 vmulps %xmm3, %xmm5, %xmm9 vmulps %xmm0, %xmm2, %xmm14 vsubps %xmm9, %xmm14, %xmm14 vmovaps %xmm7, 0x90(%rsp) vmulps %xmm1, %xmm7, %xmm9 vmulps %xmm4, %xmm0, %xmm0 vaddps %xmm0, %xmm9, %xmm0 vmulps %xmm3, %xmm11, %xmm9 vaddps %xmm0, %xmm9, %xmm9 vmulps %xmm14, %xmm15, %xmm0 vmulps %xmm13, %xmm10, %xmm10 vaddps %xmm0, %xmm10, %xmm0 vbroadcastss 0x1cfded7(%rip), %xmm10 # 0x1f20ec0 vandps %xmm10, %xmm9, %xmm10 vmulps %xmm12, %xmm8, %xmm8 vaddps %xmm0, %xmm8, %xmm0 vxorps %xmm0, %xmm10, %xmm8 vmovaps (%rsp), %xmm0 vcmpnltps 0x1cc8a07(%rip), %xmm8, %xmm15 # 0x1eeba10 vtestps %xmm0, %xmm15 jne 0x223110 vbroadcastss 0x1cc8a03(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x20(%rsp), %xmm1 vmovdqa 0x30(%rsp), %xmm2 vtestps %xmm0, %xmm0 setne %dil jmp 0x223055 xorl %edi, %edi vbroadcastss 0x1cc89dc(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x20(%rsp), %xmm1 vmovdqa 0x30(%rsp), %xmm2 testb %dil, %dil je 0x223343 leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x222cf7 jmp 0x223343 vmovaps %xmm3, %xmm1 vmovaps %xmm13, 0x40(%rsp) vmovaps %xmm2, 0x230(%rsp) vpermilps $0x0, 0x280(%rsp), %xmm3 # xmm3 = mem[0,0,0,0] vpermilps $0x0, 0x260(%rsp), %xmm6 # xmm6 = mem[0,0,0,0] vpermilps $0x0, 0x270(%rsp), %xmm13 # xmm13 = mem[0,0,0,0] vandps %xmm0, %xmm15, %xmm2 vmulps %xmm10, %xmm13, %xmm10 vmulps %xmm6, %xmm9, %xmm6 vaddps %xmm6, %xmm10, %xmm6 vmulps %xmm1, %xmm3, %xmm1 vaddps %xmm6, %xmm1, %xmm1 vxorps %xmm1, %xmm5, %xmm9 vcmpnltps 0x1cc8942(%rip), %xmm9, %xmm10 # 0x1eeba10 vtestps %xmm2, %xmm10 je 0x222eb7 vbroadcastss 0x1cfdde2(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm12, %xmm1 vandps %xmm2, %xmm10, %xmm2 vsubps %xmm14, %xmm1, %xmm6 vcmpnltps %xmm9, %xmm6, %xmm9 vtestps %xmm2, %xmm9 jne 0x2231c8 movl $0xffffffff, %r14d # imm = 0xFFFFFFFF vmovaps (%rsp), %xmm0 jmp 0x222eb7 vmovaps %xmm11, %xmm1 vmovaps 0x90(%rsp), %xmm3 vmovaps %xmm4, 0x70(%rsp) vandps %xmm0, %xmm15, %xmm15 vmulps 0x80(%rsp), %xmm14, %xmm0 vmulps 0xa0(%rsp), %xmm13, %xmm7 vaddps %xmm0, %xmm7, %xmm0 vmulps 0x40(%rsp), %xmm12, %xmm4 vaddps %xmm0, %xmm4, %xmm0 vxorps %xmm0, %xmm10, %xmm7 vcmpnltps 0x1cc88bc(%rip), %xmm7, %xmm11 # 0x1eeba10 vtestps %xmm15, %xmm11 jne 0x22317f vbroadcastss 0x1cc88bc(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x20(%rsp), %xmm1 vmovdqa 0x30(%rsp), %xmm2 vmovaps (%rsp), %xmm0 jmp 0x22302e movq %r11, %r14 movq %rsi, %r11 vbroadcastss 0x1cfdd36(%rip), %xmm0 # 0x1f20ec4 vandps %xmm0, %xmm9, %xmm4 vandps %xmm15, %xmm11, %xmm11 vsubps %xmm8, %xmm4, %xmm0 vcmpnltps %xmm7, %xmm0, %xmm7 vtestps %xmm11, %xmm7 jne 0x22327e movq %r11, %rsi movq %r14, %r11 vbroadcastss 0x1cc8865(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 movl $0xffffffff, %r14d # imm = 0xFFFFFFFF jmp 0x223169 vandps %xmm2, %xmm9, %xmm2 vmulps %xmm7, %xmm8, %xmm3 vmovaps 0x230(%rsp), %xmm0 vmulps 0x40(%rsp), %xmm0, %xmm0 vaddps %xmm0, %xmm3, %xmm0 vmulps %xmm4, %xmm11, %xmm3 vaddps %xmm0, %xmm3, %xmm0 vxorps %xmm0, %xmm5, %xmm0 vmulps 0xf0(%rsp), %xmm1, %xmm3 vcmpltps %xmm0, %xmm3, %xmm3 vmulps 0xe0(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm0, %xmm0 vandps %xmm3, %xmm0, %xmm0 vtestps %xmm2, %xmm0 je 0x223100 vandps %xmm2, %xmm0, %xmm0 vcmpneqps 0x1cc87e9(%rip), %xmm12, %xmm1 # 0x1eeba10 vtestps %xmm0, %xmm1 je 0x223100 vandps %xmm0, %xmm1, %xmm0 movq (%r10), %rdi movq 0x1e8(%rdi), %rdi movq (%rdi,%r8,8), %rdi vbroadcastss 0x34(%rdi), %xmm1 vandps 0xd0(%rsp), %xmm1, %xmm1 vpcmpeqd 0x1cc87b5(%rip), %xmm1, %xmm1 # 0x1eeba10 vtestps %xmm0, %xmm1 jb 0x223100 vandnps %xmm0, %xmm1, %xmm0 vmovaps (%rsp), %xmm1 vandnps %xmm1, %xmm0, %xmm0 movl $0xffffffff, %r14d # imm = 0xFFFFFFFF jmp 0x222eb7 vmovaps %xmm1, %xmm12 vandps %xmm7, %xmm11, %xmm7 vmulps %xmm6, %xmm3, %xmm0 vmulps 0x70(%rsp), %xmm5, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm2, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm10, %xmm0 vmulps 0xf0(%rsp), %xmm4, %xmm1 vcmpltps %xmm0, %xmm1, %xmm1 vmulps 0xe0(%rsp), %xmm4, %xmm2 vcmpleps %xmm2, %xmm0, %xmm0 vandps %xmm1, %xmm0, %xmm0 vtestps %xmm7, %xmm0 je 0x2231ac vandps %xmm7, %xmm0, %xmm0 vcmpneqps 0x1cc8738(%rip), %xmm9, %xmm1 # 0x1eeba10 vtestps %xmm0, %xmm1 je 0x2231ac vandps %xmm0, %xmm1, %xmm0 movq (%r10), %rdi movq 0x1e8(%rdi), %rdi movq (%rdi,%r8,8), %rsi vbroadcastss 0x34(%rsi), %xmm1 vandps 0xd0(%rsp), %xmm1, %xmm1 vpcmpeqd 0x1cc8704(%rip), %xmm1, %xmm1 # 0x1eeba10 vtestps %xmm0, %xmm1 jb 0x2231ac vandnps %xmm0, %xmm1, %xmm0 vmovaps (%rsp), %xmm1 vandnps %xmm1, %xmm0, %xmm0 movq %r11, %rsi movq %r14, %r11 vbroadcastss 0x1cc86ed(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 movl $0xffffffff, %r14d # imm = 0xFFFFFFFF jmp 0x223022 vmovaps %xmm0, %xmm3 vandps %xmm2, %xmm0, %xmm0 vtestps %xmm2, %xmm3 je 0x223369 incq %rcx addq $0xe0, %rbp vmovaps %xmm0, %xmm2 cmpq %rax, %rcx jb 0x222ceb vxorps %xmm0, %xmm11, %xmm0 movq %r12, %r8 vpor %xmm0, %xmm1, %xmm0 vmovdqa %xmm0, 0x50(%rsp) vtestps %xmm11, %xmm0 jb 0x2233a6 vbroadcastss 0x1cc97fa(%rip), %xmm1 # 0x1eecb84 vmovaps 0x240(%rsp), %xmm2 vblendvps %xmm0, %xmm1, %xmm2, %xmm0 vmovaps %xmm0, 0x220(%rsp) xorl %eax, %eax jmp 0x2233a9 pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x222a14 jmp 0x22349b pushq $0x2 jmp 0x2233a8 vmovaps %xmm12, 0x60(%rsp) movq %r8, 0x40(%rsp) movq %r10, 0xc8(%rsp) movq %r9, 0xc0(%rsp) movq %rsi, 0xb8(%rsp) movq %r11, 0xb0(%rsp) bsfq %r12, %rcx movq %rsi, %rdi movq %r11, %rsi movq %rbp, %rdx movq %rcx, (%rsp) leaq 0x1f(%rsp), %r8 pushq %r10 leaq 0x158(%rsp), %rax pushq %rax callq 0x25bd0c popq %rcx popq %rdx testb %al, %al je 0x22341b movq (%rsp), %rax orl $-0x1, 0x50(%rsp,%rax,4) leaq -0x1(%r12), %rax andq %rax, %r12 movq 0xc8(%rsp), %r10 movq 0xc0(%rsp), %r9 movq 0xb8(%rsp), %rsi movq 0xb0(%rsp), %r11 jne 0x2233e6 vmovaps 0x50(%rsp), %xmm0 vpcmpeqd %xmm11, %xmm11, %xmm11 vtestps %xmm11, %xmm0 pushq $0x3 popq %rax vbroadcastss 0x1cc85bf(%rip), %xmm10 # 0x1eeba20 movq 0x40(%rsp), %r8 vmovaps 0x60(%rsp), %xmm12 jb 0x222a58 vmovaps 0x220(%rsp), %xmm1 vbroadcastss 0x1cc9700(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x220(%rsp) pushq $0x2 popq %rax jmp 0x222a58 vmovaps 0x250(%rsp), %xmm0 vandps 0x50(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cc96d1(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r8) addq $0x1958, %rsp # imm = 0x1958 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r9, %rdx movq %r10, %rcx jmp 0x25ad06 nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiIntersectorKMoeller<4, 4, true>>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x2236e0 movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 cmpq $0x0, 0x8(%rcx) je 0x223513 movq 0x10(%r15), %rax testb $0x1, 0x2(%rax) jne 0x2236f2 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x2236e0 movzbl %al, %ebp vmovaps (%r14), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%r14), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%r14), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%r14), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cfd94c(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1ccda63(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vandps %xmm4, %xmm2, %xmm5 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vrcpps %xmm1, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1cc9155(%rip), %xmm5 # 0x1eec714 vsubps %xmm1, %xmm5, %xmm1 vmulps %xmm1, %xmm4, %xmm1 vaddps %xmm1, %xmm4, %xmm1 vrcpps %xmm2, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm2, %xmm4, %xmm2 vrcpps %xmm3, %xmm4 vmulps %xmm4, %xmm3, %xmm3 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d3734d(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d3136f(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d3732e(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d37316(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d37311(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%r14), %xmm1 vmovaps 0x80(%r14), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cc8393(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cc94e0(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x25dade popq %rax popq %rcx andq %rbx, %rbp jne 0x2236b2 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r12, %rsi movq %r14, %rdx movq %r15, %rcx addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x25c3c8 nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiIntersectorKMoeller<4, 4, true>>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1a08, %rsp # imm = 0x1A08 movq %rsi, %r9 movq (%rsi), %rsi cmpq $0x8, 0x70(%rsi) je 0x22494f movq %rcx, %r10 movq %rdx, %r14 movq 0x8(%rcx), %rax testq %rax, %rax je 0x223751 movq 0x10(%r10), %rcx testb $0x1, 0x2(%rcx) jne 0x224964 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%r14), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x22494f vandps %xmm3, %xmm4, %xmm8 vmovaps (%r14), %xmm3 vmovaps %xmm3, 0x230(%rsp) vmovaps 0x10(%r14), %xmm3 vmovaps %xmm3, 0x240(%rsp) vmovaps 0x20(%r14), %xmm3 vmovaps %xmm3, 0x250(%rsp) vmovaps 0x40(%r14), %xmm3 vmovaps %xmm3, 0x260(%rsp) vmovaps 0x50(%r14), %xmm4 vmovaps %xmm4, 0x270(%rsp) vmovaps 0x60(%r14), %xmm5 vmovaps %xmm5, 0x280(%rsp) vbroadcastss 0x1cfd6e8(%rip), %xmm9 # 0x1f20ec4 vandps %xmm3, %xmm9, %xmm6 vbroadcastss 0x1ccd7ff(%rip), %xmm7 # 0x1ef0fe8 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm3, %xmm3 vandps %xmm4, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm4, %xmm4 vandps %xmm5, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm5, %xmm5 vrcpps %xmm3, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0x1cc8ef1(%rip), %xmm7 # 0x1eec714 vsubps %xmm3, %xmm7, %xmm3 vmulps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm6, %xmm3 vrcpps %xmm4, %xmm6 vmulps %xmm6, %xmm4, %xmm4 vsubps %xmm4, %xmm7, %xmm4 vmulps %xmm4, %xmm6, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm5 vsubps %xmm5, %xmm7, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm6, %xmm5 vmovaps %xmm3, 0x290(%rsp) vmovaps %xmm4, 0x2a0(%rsp) vmovaps %xmm5, 0x2b0(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d370e4(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x2c0(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d31105(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d370c4(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vcmpnltps %xmm2, %xmm5, %xmm4 vbroadcastss 0x1d370b4(%rip), %xmm5 # 0x1f5a96c vbroadcastss 0x1d370af(%rip), %xmm6 # 0x1f5a970 vblendvps %xmm4, %xmm5, %xmm6, %xmm4 vmovaps %xmm3, 0x2d0(%rsp) vmovaps %xmm4, 0x2e0(%rsp) vmovaps 0x30(%r14), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cc8130(%rip), %xmm10 # 0x1eeba20 vblendvps %xmm8, %xmm3, %xmm10, %xmm1 vmovaps %xmm1, 0x2f0(%rsp) vbroadcastss 0x1cc927c(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm8, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x300(%rsp) vmovaps %xmm8, 0x330(%rsp) vxorps %xmm0, %xmm8, %xmm0 vmovaps %xmm0, 0xe0(%rsp) testq %rax, %rax je 0x223946 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r11d, %r11d cmpb $0x1, %al adcq $0x2, %r11 jmp 0x22394a pushq $0x3 popq %r11 leaq 0x80(%r14), %rax movq %rax, 0xa8(%rsp) leaq 0x370(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xb10(%rsp), %r12 vmovaps %xmm10, -0x20(%r12) movq 0x70(%rsi), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%r12) leaq 0x1f2c5f2(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x320(%rsp) vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x310(%rsp) vpcmpeqd %xmm11, %xmm11, %xmm11 addq $-0x10, %r12 movq -0x8(%r15), %r13 addq $-0x8, %r15 cmpq $-0x8, %r13 je 0x22483f vmovaps (%r12), %xmm12 vcmpltps 0x300(%rsp), %xmm12, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x224850 movzbl %al, %ebp popcntl %ebp, %ebx xorl %eax, %eax cmpq %r11, %rbx jbe 0x224854 cmpq %r11, %rbx jbe 0x224842 testb $0x8, %r13b pushq $0x8 popq %r8 jne 0x223b9d movq %r13, %rax movq %r13, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %r8, %r13 vmovaps %xmm10, %xmm12 movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x223b34 vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x230(%rsp), %xmm1 vmovaps 0x240(%rsp), %xmm2 vmovaps 0x250(%rsp), %xmm3 vmovaps 0x290(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0x2a0(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x2b0(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vpminsd %xmm1, %xmm5, %xmm0 vpminsd %xmm2, %xmm7, %xmm4 vpmaxsd %xmm4, %xmm0, %xmm0 vpminsd %xmm3, %xmm9, %xmm4 vpmaxsd %xmm4, %xmm0, %xmm0 vpmaxsd %xmm1, %xmm5, %xmm1 vpmaxsd %xmm2, %xmm7, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpmaxsd %xmm3, %xmm9, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpmaxsd 0x2f0(%rsp), %xmm0, %xmm2 vpminsd 0x300(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vtestps %xmm1, %xmm1 je 0x223b34 vblendvps %xmm1, %xmm0, %xmm10, %xmm0 cmpq $0x8, %r13 je 0x223b2d movq %r13, (%r15) addq $0x8, %r15 vmovaps %xmm12, (%r12) addq $0x10, %r12 vmovaps %xmm0, %xmm12 movq %rdi, %r13 cmpq $0x8, %rdi je 0x223b4b leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x223a20 xorl %eax, %eax cmpq $0x8, %r13 je 0x223b96 vmovaps 0x300(%rsp), %xmm0 vcmpnleps %xmm12, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r11 jae 0x223b85 testb %cl, %cl je 0x224842 testb $0x8, %r13b je 0x223a0c jmp 0x223b9d movq %r13, (%r15) addq $0x8, %r15 vmovaps %xmm12, (%r12) addq $0x10, %r12 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x223b71 cmpq $-0x8, %r13 je 0x22483f movq %r9, 0x58(%rsp) vmovaps 0x300(%rsp), %xmm0 vcmpnleps %xmm12, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22491a movl %r13d, %r9d andl $0xf, %r9d vmovdqa 0xe0(%rsp), %xmm0 addq $-0x8, %r9 je 0x2247fc andq $-0x10, %r13 vpxor %xmm0, %xmm11, %xmm1 addq $0x50, %r13 xorl %ecx, %ecx movq (%r10), %rax movq %rax, 0x168(%rsp) xorl %ebx, %ebx vmovdqa %xmm1, %xmm0 vmovdqa %xmm1, 0x60(%rsp) movl (%r13,%rbx,4), %eax cmpl $-0x1, %eax je 0x2247d5 movq %rcx, %rbp movl -0x10(%r13,%rbx,4), %ecx movq 0x168(%rsp), %rdx movq 0x228(%rdx), %rdx movq (%rdx,%rcx,8), %rdx movl -0x50(%r13,%rbx,4), %r8d movl -0x40(%r13,%rbx,4), %edi vmovdqa %xmm0, 0x10(%rsp) vbroadcastss (%rdx,%r8,4), %xmm0 vbroadcastss 0x4(%rdx,%r8,4), %xmm1 vbroadcastss 0x8(%rdx,%r8,4), %xmm2 vbroadcastss (%rdx,%rdi,4), %xmm3 vbroadcastss 0x4(%rdx,%rdi,4), %xmm4 vbroadcastss 0x8(%rdx,%rdi,4), %xmm5 movl -0x20(%r13,%rbx,4), %edi vbroadcastss (%rdx,%rdi,4), %xmm9 vbroadcastss 0x4(%rdx,%rdi,4), %xmm10 vbroadcastss 0x8(%rdx,%rdi,4), %xmm11 vmovaps %xmm3, 0x150(%rsp) vsubss %xmm3, %xmm0, %xmm6 vmovaps %xmm4, 0x80(%rsp) vsubss %xmm4, %xmm1, %xmm7 vmovaps %xmm5, 0xd0(%rsp) vsubss %xmm5, %xmm2, %xmm8 vmovaps %xmm10, 0x350(%rsp) vsubss %xmm1, %xmm10, %xmm3 vmovaps %xmm11, 0x340(%rsp) vsubss %xmm2, %xmm11, %xmm13 vmulss %xmm7, %xmm13, %xmm4 vmulss %xmm3, %xmm8, %xmm5 vsubss %xmm4, %xmm5, %xmm4 vmovaps %xmm4, 0x70(%rsp) vmovaps %xmm9, 0x140(%rsp) vsubss %xmm0, %xmm9, %xmm12 vmovaps %xmm8, 0xf0(%rsp) vmulss %xmm12, %xmm8, %xmm4 vmulss %xmm6, %xmm13, %xmm5 vsubss %xmm4, %xmm5, %xmm11 vmovaps %xmm6, 0x20(%rsp) vmulss %xmm3, %xmm6, %xmm4 vmovaps %xmm7, 0x40(%rsp) vmulss %xmm7, %xmm12, %xmm5 vsubss %xmm4, %xmm5, %xmm4 vsubps (%r14), %xmm0, %xmm6 vsubps 0x10(%r14), %xmm1, %xmm7 vsubps 0x20(%r14), %xmm2, %xmm10 vmovaps 0x50(%r14), %xmm2 vmovaps 0x60(%r14), %xmm1 vmulps %xmm2, %xmm10, %xmm14 vmulps %xmm1, %xmm7, %xmm15 vsubps %xmm14, %xmm15, %xmm8 vmovaps 0x40(%r14), %xmm0 vmulps %xmm1, %xmm6, %xmm14 vmulps %xmm0, %xmm10, %xmm15 vsubps %xmm14, %xmm15, %xmm9 vmulps %xmm0, %xmm7, %xmm15 vmulps %xmm2, %xmm6, %xmm14 vsubps %xmm15, %xmm14, %xmm15 vshufps $0x0, %xmm11, %xmm11, %xmm5 # xmm5 = xmm11[0,0,0,0] vshufps $0x0, %xmm4, %xmm4, %xmm11 # xmm11 = xmm4[0,0,0,0] vmulps %xmm1, %xmm11, %xmm1 vmulps %xmm2, %xmm5, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vpermilps $0x0, 0x70(%rsp), %xmm2 # xmm2 = mem[0,0,0,0] vmulps %xmm0, %xmm2, %xmm0 vaddps %xmm1, %xmm0, %xmm4 vshufps $0x0, %xmm3, %xmm3, %xmm0 # xmm0 = xmm3[0,0,0,0] vshufps $0x0, %xmm13, %xmm13, %xmm1 # xmm1 = xmm13[0,0,0,0] vmulps %xmm1, %xmm15, %xmm1 vmovaps %xmm9, 0x130(%rsp) vmulps %xmm0, %xmm9, %xmm0 vmovaps %xmm4, %xmm9 vaddps %xmm1, %xmm0, %xmm0 vshufps $0x0, %xmm12, %xmm12, %xmm1 # xmm1 = xmm12[0,0,0,0] vmovaps %xmm8, %xmm3 vmulps %xmm1, %xmm8, %xmm1 vaddps %xmm0, %xmm1, %xmm0 movl -0x30(%r13,%rbx,4), %edi vbroadcastss 0x1cfd0fe(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm4, %xmm12 vxorps %xmm0, %xmm12, %xmm8 vmovaps 0x10(%rsp), %xmm0 vcmpnltps 0x1cc7c37(%rip), %xmm8, %xmm13 # 0x1eeba10 vtestps %xmm0, %xmm13 vbroadcastss (%rdx,%rdi,4), %xmm14 vbroadcastss 0x4(%rdx,%rdi,4), %xmm1 vbroadcastss 0x8(%rdx,%rdi,4), %xmm4 vmovaps %xmm4, 0x70(%rsp) jne 0x223faa movq %rbp, %rcx vtestps %xmm0, %xmm0 je 0x223f76 vsubps 0x140(%rsp), %xmm14, %xmm5 vsubps 0x350(%rsp), %xmm1, %xmm4 vmovaps 0x70(%rsp), %xmm8 vsubps 0x340(%rsp), %xmm8, %xmm10 vmovaps %xmm1, %xmm7 vmovaps 0x150(%rsp), %xmm1 vsubps %xmm14, %xmm1, %xmm3 vmovaps 0x80(%rsp), %xmm1 vsubps %xmm7, %xmm1, %xmm11 vmovaps 0xd0(%rsp), %xmm1 vsubps %xmm8, %xmm1, %xmm15 vmovaps %xmm0, 0x10(%rsp) vmulps %xmm4, %xmm15, %xmm0 vmulps %xmm10, %xmm11, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x150(%rsp) vmulps %xmm3, %xmm10, %xmm1 vmulps %xmm5, %xmm15, %xmm2 vsubps %xmm1, %xmm2, %xmm6 vmovaps %xmm5, 0x80(%rsp) vmulps %xmm5, %xmm11, %xmm2 vmovaps %xmm4, 0xd0(%rsp) vmulps %xmm4, %xmm3, %xmm5 vsubps %xmm2, %xmm5, %xmm5 vsubps (%r14), %xmm14, %xmm4 vsubps 0x10(%r14), %xmm7, %xmm7 vsubps 0x20(%r14), %xmm8, %xmm8 vmovaps 0x50(%r14), %xmm9 vmovaps 0x60(%r14), %xmm0 vmulps %xmm9, %xmm8, %xmm12 vmulps %xmm0, %xmm7, %xmm13 vsubps %xmm12, %xmm13, %xmm12 vmovaps 0x40(%r14), %xmm1 vmulps %xmm0, %xmm4, %xmm13 vmulps %xmm1, %xmm8, %xmm14 vsubps %xmm13, %xmm14, %xmm13 vmulps %xmm1, %xmm7, %xmm14 vmulps %xmm4, %xmm9, %xmm2 vsubps %xmm14, %xmm2, %xmm14 vmovaps %xmm5, 0x140(%rsp) vmulps %xmm0, %xmm5, %xmm0 vmovaps %xmm6, 0x70(%rsp) vmulps %xmm6, %xmm9, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vmovaps 0x150(%rsp), %xmm2 vmulps %xmm1, %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm9 vmulps %xmm14, %xmm15, %xmm0 vmulps %xmm13, %xmm11, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cfcf92(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm9, %xmm11 vmulps %xmm3, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm3 vmovaps 0x10(%rsp), %xmm0 vcmpnltps 0x1cc7ac3(%rip), %xmm3, %xmm15 # 0x1eeba10 vtestps %xmm0, %xmm15 jne 0x22404c vbroadcastss 0x1cc7abf(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x60(%rsp), %xmm1 vtestps %xmm0, %xmm0 setne %al jmp 0x223f8c xorl %eax, %eax vbroadcastss 0x1cc7a9f(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x60(%rsp), %xmm1 testb %al, %al je 0x2247d5 leaq 0x1(%rbx), %rax cmpq $0x3, %rbx movq %rax, %rbx jb 0x223c05 jmp 0x2247d5 vmovaps %xmm8, 0x190(%rsp) vmovaps %xmm2, 0x170(%rsp) vmovaps %xmm6, %xmm4 vmovaps %xmm5, 0x180(%rsp) vmovaps %xmm1, 0xb0(%rsp) vmovaps %xmm14, 0xc0(%rsp) vmovaps %xmm0, %xmm14 vpermilps $0x0, 0x20(%rsp), %xmm0 # xmm0 = mem[0,0,0,0] vpermilps $0x0, 0x40(%rsp), %xmm1 # xmm1 = mem[0,0,0,0] vpermilps $0x0, 0xf0(%rsp), %xmm6 # xmm6 = mem[0,0,0,0] vandps %xmm14, %xmm13, %xmm8 vmulps %xmm6, %xmm15, %xmm6 vmulps 0x130(%rsp), %xmm1, %xmm1 vaddps %xmm6, %xmm1, %xmm1 vmulps %xmm3, %xmm0, %xmm0 vaddps %xmm1, %xmm0, %xmm0 vxorps %xmm0, %xmm12, %xmm6 vcmpnltps 0x1cc79eb(%rip), %xmm6, %xmm1 # 0x1eeba10 vtestps %xmm8, %xmm1 jne 0x2240a6 movq %rbp, %rcx vmovdqa 0x10(%rsp), %xmm0 vmovaps 0xc0(%rsp), %xmm14 vmovaps 0xb0(%rsp), %xmm1 jmp 0x223e01 vmovaps %xmm4, %xmm5 vandps %xmm0, %xmm15, %xmm15 vmulps %xmm14, %xmm10, %xmm0 vmulps 0xd0(%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x80(%rsp), %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm4 vcmpnltps 0x1cc7990(%rip), %xmm4, %xmm10 # 0x1eeba10 vtestps %xmm15, %xmm10 jne 0x2240f7 vbroadcastss 0x1cc7990(%rip), %xmm10 # 0x1eeba20 vpcmpeqd %xmm11, %xmm11, %xmm11 vmovdqa 0x60(%rsp), %xmm1 vmovdqa 0x10(%rsp), %xmm0 jmp 0x223f6c movq %r9, 0x40(%rsp) movq %r11, 0x20(%rsp) movq %rsi, 0x8(%rsp) movq %r10, (%rsp) vbroadcastss 0x1cfce02(%rip), %xmm0 # 0x1f20ec4 vandps %xmm0, %xmm9, %xmm13 vandps %xmm1, %xmm8, %xmm8 vsubps 0x190(%rsp), %xmm13, %xmm0 vcmpnltps %xmm6, %xmm0, %xmm1 vtestps %xmm8, %xmm1 jne 0x224159 movq (%rsp), %r10 movq 0x8(%rsp), %rsi movq 0x20(%rsp), %r11 movq 0x40(%rsp), %r9 jmp 0x22402c movq %r9, 0x40(%rsp) movq %r11, 0x20(%rsp) movq %rsi, 0x8(%rsp) movq %r10, (%rsp) vbroadcastss 0x1cfcdb1(%rip), %xmm0 # 0x1f20ec4 vandps %xmm0, %xmm9, %xmm6 vandps %xmm15, %xmm10, %xmm10 vsubps %xmm3, %xmm6, %xmm0 vcmpnltps %xmm4, %xmm0, %xmm12 vtestps %xmm10, %xmm12 jne 0x22425c movq (%rsp), %r10 movq 0x8(%rsp), %rsi vbroadcastss 0x1cc78de(%rip), %xmm10 # 0x1eeba20 movq 0x20(%rsp), %r11 vpcmpeqd %xmm11, %xmm11, %xmm11 movq 0x40(%rsp), %r9 movq %rbp, %rcx jmp 0x224095 vandps %xmm1, %xmm8, %xmm8 vmulps %xmm10, %xmm11, %xmm0 vmulps 0x180(%rsp), %xmm7, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x170(%rsp), %xmm4, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm12, %xmm0 vmulps 0x30(%r14), %xmm13, %xmm1 vcmpltps %xmm0, %xmm1, %xmm1 vmovaps 0x80(%r14), %xmm5 vmovaps %xmm5, 0x130(%rsp) vmulps %xmm5, %xmm13, %xmm5 vcmpleps %xmm5, %xmm0, %xmm5 vandps %xmm1, %xmm5, %xmm5 vtestps %xmm8, %xmm5 je 0x2240df vandps %xmm5, %xmm8, %xmm5 vcmpneqps 0x1cc784e(%rip), %xmm9, %xmm8 # 0x1eeba10 vtestps %xmm5, %xmm8 je 0x2240df vandps %xmm5, %xmm8, %xmm5 movq (%rsp), %rdx movq (%rdx), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rcx,8), %rdx movq %rdx, 0xf0(%rsp) vbroadcastss 0x34(%rdx), %xmm1 vandps 0x90(%r14), %xmm1, %xmm1 vpcmpeqd 0x1cc780e(%rip), %xmm1, %xmm8 # 0x1eeba10 vtestps %xmm5, %xmm8 jb 0x2240df vandnps %xmm5, %xmm8, %xmm5 movq (%rsp), %rdx movq 0x10(%rdx), %rdx cmpq $0x0, 0x10(%rdx) jne 0x224378 movq 0xf0(%rsp), %rdx cmpq $0x0, 0x48(%rdx) jne 0x224378 vmovdqa 0x10(%rsp), %xmm0 vpandn %xmm0, %xmm5, %xmm0 movq (%rsp), %r10 movq 0x8(%rsp), %rsi movq 0x20(%rsp), %r11 movq 0x40(%rsp), %r9 movq %rbp, %rcx jmp 0x224035 vmovaps %xmm2, %xmm13 vandps %xmm10, %xmm12, %xmm10 vmulps 0x140(%rsp), %xmm8, %xmm0 vmulps 0x70(%rsp), %xmm7, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps %xmm5, %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm5 vmulps 0x30(%r14), %xmm6, %xmm0 vcmpltps %xmm5, %xmm0, %xmm0 vmovaps 0x80(%r14), %xmm1 vmovaps %xmm1, 0xd0(%rsp) vmulps %xmm1, %xmm6, %xmm1 vcmpleps %xmm1, %xmm5, %xmm1 vandps %xmm0, %xmm1, %xmm7 vtestps %xmm10, %xmm7 je 0x224130 vandps %xmm7, %xmm10, %xmm7 vcmpneqps 0x1cc774a(%rip), %xmm9, %xmm8 # 0x1eeba10 vtestps %xmm7, %xmm8 je 0x224130 vandps %xmm7, %xmm8, %xmm7 movq (%rsp), %rax movq (%rax), %rcx movl -0x10(%r13,%rbx,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rcx movq %rcx, 0x80(%rsp) vbroadcastss 0x34(%rcx), %xmm0 vandps 0x90(%r14), %xmm0, %xmm0 vpcmpeqd 0x1cc7705(%rip), %xmm0, %xmm8 # 0x1eeba10 vtestps %xmm7, %xmm8 jb 0x224130 movl (%r13,%rbx,4), %ecx vandnps %xmm7, %xmm8, %xmm7 movq (%rsp), %rdx movq 0x10(%rdx), %rdx cmpq $0x0, 0x10(%rdx) jne 0x224574 movq 0x80(%rsp), %rdx cmpq $0x0, 0x48(%rdx) jne 0x224574 vmovdqa 0x10(%rsp), %xmm0 vpandn %xmm0, %xmm7, %xmm0 movq (%rsp), %r10 movq 0x8(%rsp), %rsi vbroadcastss 0x1cc76bf(%rip), %xmm10 # 0x1eeba20 movq 0x20(%rsp), %r11 vpcmpeqd %xmm11, %xmm11, %xmm11 movq 0x40(%rsp), %r9 movq %rbp, %rcx jmp 0x223f66 vrcpps %xmm13, %xmm1 vmulps %xmm1, %xmm13, %xmm7 vbroadcastss 0x1cc838a(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm7, %xmm1, %xmm1 vmulps 0x190(%rsp), %xmm1, %xmm3 vminps %xmm8, %xmm3, %xmm3 vmulps %xmm6, %xmm1, %xmm6 vminps %xmm8, %xmm6, %xmm6 vsubps %xmm3, %xmm8, %xmm7 vsubps %xmm6, %xmm8, %xmm8 vmovaps 0x320(%rsp), %xmm9 vblendvps %xmm9, %xmm7, %xmm3, %xmm3 vblendvps %xmm9, %xmm8, %xmm6, %xmm6 movq (%rsp), %rsi movq 0x8(%rsi), %rdx vmovd %ecx, %xmm7 vpshufd $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmovd %eax, %xmm8 vpshufd $0x0, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] vmovaps 0x170(%rsp), %xmm2 vmovaps %xmm2, 0x1a0(%rsp) vmovaps 0x180(%rsp), %xmm2 vmovaps %xmm2, 0x1b0(%rsp) vmovaps %xmm11, 0x1c0(%rsp) vmovaps %xmm3, 0x1d0(%rsp) vmovaps %xmm6, 0x1e0(%rsp) vmovdqa %xmm8, 0x1f0(%rsp) vmovdqa %xmm7, 0x200(%rsp) vmulps %xmm0, %xmm1, %xmm0 leaq 0x210(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x210(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x220(%rsp) vmovaps 0x130(%rsp), %xmm1 vblendvps %xmm5, %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm5, 0x90(%rsp) leaq 0x90(%rsp), %rax movq %rax, 0x100(%rsp) movq 0xf0(%rsp), %rcx movq 0x18(%rcx), %rax movq %rax, 0x108(%rsp) movq 0x8(%rsi), %rax movq %rax, 0x110(%rsp) movq %r14, 0x118(%rsp) leaq 0x1a0(%rsp), %rax movq %rax, 0x120(%rsp) movl $0x4, 0x128(%rsp) movq 0x48(%rcx), %rax testq %rax, %rax je 0x2244f2 leaq 0x100(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x90(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x224761 movq (%rsp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x224537 testb $0x2, (%rcx) jne 0x22452a movq 0xf0(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x224537 leaq 0x100(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x90(%rsp), %xmm0, %xmm1 vpxor 0x1cc78d4(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x118(%rsp), %rax vbroadcastss 0x1cc8627(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x224771 vrcpps %xmm6, %xmm0 vmulps %xmm0, %xmm6, %xmm1 vbroadcastss 0x1cc818f(%rip), %xmm6 # 0x1eec714 vsubps %xmm1, %xmm6, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm3, %xmm0, %xmm1 vminps %xmm6, %xmm1, %xmm1 vmulps %xmm4, %xmm0, %xmm2 vminps %xmm6, %xmm2, %xmm2 vsubps %xmm1, %xmm6, %xmm3 vsubps %xmm2, %xmm6, %xmm4 vmovaps 0x310(%rsp), %xmm6 vblendvps %xmm6, %xmm3, %xmm1, %xmm1 vblendvps %xmm6, %xmm4, %xmm2, %xmm2 movq (%rsp), %rsi movq 0x8(%rsi), %rdx vmovd %eax, %xmm3 vpshufd $0x0, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmovd %ecx, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovaps %xmm13, 0x1a0(%rsp) vmovaps 0x70(%rsp), %xmm6 vmovaps %xmm6, 0x1b0(%rsp) vmovaps 0x140(%rsp), %xmm6 vmovaps %xmm6, 0x1c0(%rsp) vmovaps %xmm1, 0x1d0(%rsp) vmovaps %xmm2, 0x1e0(%rsp) vmovdqa %xmm4, 0x1f0(%rsp) vmovdqa %xmm3, 0x200(%rsp) vmulps %xmm5, %xmm0, %xmm0 leaq 0x210(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x210(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x220(%rsp) vmovaps 0xd0(%rsp), %xmm1 vblendvps %xmm7, %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm7, 0x90(%rsp) leaq 0x90(%rsp), %rax movq %rax, 0x100(%rsp) movq 0x80(%rsp), %rcx movq 0x18(%rcx), %rax movq %rax, 0x108(%rsp) movq 0x8(%rsi), %rax movq %rax, 0x110(%rsp) movq %r14, 0x118(%rsp) leaq 0x1a0(%rsp), %rax movq %rax, 0x120(%rsp) movl $0x4, 0x128(%rsp) movq 0x48(%rcx), %rax testq %rax, %rax je 0x2246e2 leaq 0x100(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x90(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x22479b movq (%rsp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x224727 testb $0x2, (%rcx) jne 0x22471a movq 0x80(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x224727 leaq 0x100(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x90(%rsp), %xmm0, %xmm1 vpxor 0x1cc76e4(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x118(%rsp), %rax vbroadcastss 0x1cc8437(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x2247ab vpcmpeqd 0x1cc72a7(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cc76af(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm5 movq 0xa8(%rsp), %rax vmovaps 0x130(%rsp), %xmm1 vblendvps %xmm0, (%rax), %xmm1, %xmm0 vmovaps %xmm0, (%rax) jmp 0x224237 vpcmpeqd 0x1cc726d(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cc7675(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm7 movq 0xa8(%rsp), %rax vmovaps 0xd0(%rsp), %xmm1 vblendvps %xmm0, (%rax), %xmm1, %xmm0 vmovaps %xmm0, (%rax) jmp 0x224345 vmovaps %xmm0, %xmm2 vpand %xmm1, %xmm0, %xmm0 vtestps %xmm1, %xmm2 je 0x2247f8 incq %rcx addq $0x60, %r13 vmovdqa %xmm0, %xmm1 cmpq %r9, %rcx jb 0x223bee vpxor %xmm0, %xmm11, %xmm0 vpor 0xe0(%rsp), %xmm0, %xmm0 vmovdqa %xmm0, 0xe0(%rsp) vtestps %xmm11, %xmm0 movq 0x58(%rsp), %r9 jb 0x22483f vmovaps 0x300(%rsp), %xmm1 vbroadcastss 0x1cc8358(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x300(%rsp) xorl %eax, %eax jmp 0x224842 pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x2239b1 jmp 0x224927 pushq $0x2 jmp 0x224841 vmovaps %xmm12, 0x70(%rsp) movq %r11, 0x20(%rsp) movq %r10, (%rsp) movq %r9, 0x58(%rsp) movq %rsi, 0x8(%rsp) bsfq %rbp, %rcx movq %r9, %rdi movq %r13, %rdx movq %rcx, 0x10(%rsp) leaq 0x3f(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x238(%rsp), %rax pushq %rax vzeroupper callq 0x25fe68 popq %rcx popq %rdx testb %al, %al je 0x2248aa movq 0x10(%rsp), %rax orl $-0x1, 0xe0(%rsp,%rax,4) leaq -0x1(%rbp), %rax andq %rax, %rbp movq (%rsp), %r10 movq 0x58(%rsp), %r9 movq 0x8(%rsp), %rsi jne 0x22486d vmovaps 0xe0(%rsp), %xmm0 vpcmpeqd %xmm11, %xmm11, %xmm11 vtestps %xmm11, %xmm0 pushq $0x3 popq %rax vbroadcastss 0x1cc7140(%rip), %xmm10 # 0x1eeba20 movq 0x20(%rsp), %r11 vmovaps 0x70(%rsp), %xmm12 jb 0x2239f5 vmovaps 0x300(%rsp), %xmm1 vbroadcastss 0x1cc8281(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x300(%rsp) pushq $0x2 popq %rax jmp 0x2239f5 pushq $0x2 popq %rax movq 0x58(%rsp), %r9 jmp 0x224842 vmovaps 0x330(%rsp), %xmm0 vandps 0xe0(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cc8242(%rip), %xmm1 # 0x1eecb84 movq 0xa8(%rsp), %rax vmaskmovps %xmm1, %xmm0, (%rax) addq $0x1a08, %rsp # imm = 0x1A08 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %r9, %rsi movq %r14, %rdx movq %r10, %rcx addq $0x1a08, %rsp # imm = 0x1A08 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x25e8fe nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMvIntersectorKPluecker<4, 4, true>>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x224b67 movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 cmpq $0x0, 0x8(%rcx) je 0x2249c1 movq 0x10(%r15), %rax testb $0x1, 0x2(%rax) jne 0x224b79 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x224b67 movzbl %al, %ebp vmovaps (%r14), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%r14), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%r14), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%r14), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cfc49e(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1ccc5b5(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1cc7cd3(%rip), %xmm7 # 0x1eec714 vdivps %xmm1, %xmm7, %xmm1 vandps %xmm4, %xmm2, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vdivps %xmm2, %xmm7, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vbroadcastss 0x1cfc4fc(%rip), %xmm6 # 0x1f20f60 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vblendvps %xmm8, %xmm6, %xmm2, %xmm2 vdivps %xmm3, %xmm7, %xmm3 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d35ec6(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d2fee8(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d35ea7(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d35e8f(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d35e8a(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%r14), %xmm1 vmovaps 0x80(%r14), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cc6f0c(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cc8059(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x2624ac popq %rax popq %rcx andq %rbx, %rbp jne 0x224b39 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r12, %rsi movq %r14, %rdx movq %r15, %rcx addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x2609ee	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMvIntersectorKPluecker<4, 4, true>>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x226215 movq %rcx, %r10 movq %rdx, %r9 movq 0x8(%rcx), %rax testq %rax, %rax je 0x224bc3 movq 0x10(%r10), %rcx testb $0x1, 0x2(%rcx) jne 0x226219 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%r9), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x226215 pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1a88, %rsp # imm = 0x1A88 vandps %xmm3, %xmm4, %xmm10 vmovaps (%r9), %xmm3 vmovaps %xmm3, 0x280(%rsp) vmovaps 0x10(%r9), %xmm3 vmovaps %xmm3, 0x290(%rsp) vmovaps 0x20(%r9), %xmm3 vmovaps %xmm3, 0x2a0(%rsp) vmovaps 0x40(%r9), %xmm3 vmovaps %xmm3, 0x2b0(%rsp) vmovaps 0x50(%r9), %xmm4 vmovaps %xmm4, 0x2c0(%rsp) vmovaps 0x60(%r9), %xmm5 vmovaps %xmm5, 0x2d0(%rsp) vbroadcastss 0x1cfc265(%rip), %xmm8 # 0x1f20ec4 vandps %xmm3, %xmm8, %xmm6 vbroadcastss 0x1ccc37c(%rip), %xmm9 # 0x1ef0fe8 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cc7a99(%rip), %xmm11 # 0x1eec714 vdivps %xmm3, %xmm11, %xmm3 vandps %xmm4, %xmm8, %xmm7 vcmpltps %xmm9, %xmm7, %xmm7 vdivps %xmm4, %xmm11, %xmm4 vandps %xmm5, %xmm8, %xmm8 vcmpltps %xmm9, %xmm8, %xmm8 vbroadcastss 0x1cfc2c0(%rip), %xmm9 # 0x1f20f60 vblendvps %xmm6, %xmm9, %xmm3, %xmm3 vblendvps %xmm7, %xmm9, %xmm4, %xmm4 vdivps %xmm5, %xmm11, %xmm5 vblendvps %xmm8, %xmm9, %xmm5, %xmm5 vmovaps %xmm3, 0x2e0(%rsp) vmovaps %xmm4, 0x2f0(%rsp) vmovaps %xmm5, 0x300(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d35c85(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x310(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d2fca6(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d35c65(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vcmpnltps %xmm2, %xmm5, %xmm4 vbroadcastss 0x1d35c55(%rip), %xmm5 # 0x1f5a96c vbroadcastss 0x1d35c50(%rip), %xmm6 # 0x1f5a970 vblendvps %xmm4, %xmm5, %xmm6, %xmm4 vmovaps %xmm3, 0x320(%rsp) vmovaps %xmm4, 0x330(%rsp) vmovaps 0x30(%r9), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cc6cd1(%rip), %xmm11 # 0x1eeba20 vblendvps %xmm10, %xmm3, %xmm11, %xmm1 vmovaps %xmm1, 0x340(%rsp) vbroadcastss 0x1cc7e1d(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm10, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x350(%rsp) vmovaps %xmm10, 0x380(%rsp) vxorps %xmm0, %xmm10, %xmm0 vmovaps %xmm0, 0x90(%rsp) testq %rax, %rax je 0x224da5 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r11d, %r11d cmpb $0x1, %al adcq $0x2, %r11 jmp 0x224da9 pushq $0x3 popq %r11 leaq 0x80(%r9), %r13 leaq 0x3f0(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xb90(%rsp), %r12 vmovaps %xmm11, -0x20(%r12) movq 0x70(%r8), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%r12) leaq 0x1f2b19b(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x370(%rsp) vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x360(%rsp) vbroadcastss 0x1cfb104(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cfb0ff(%rip), %xmm12 # 0x1f1ff14 vpcmpeqd %xmm13, %xmm13, %xmm13 addq $-0x10, %r12 movq -0x8(%r15), %rbp addq $-0x8, %r15 cmpq $-0x8, %rbp je 0x2260d9 vmovaps (%r12), %xmm14 vcmpltps 0x350(%rsp), %xmm14, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x2260ea movzbl %al, %r14d popcntl %r14d, %ebx xorl %eax, %eax cmpq %r11, %rbx jbe 0x2260ee cmpq %r11, %rbx jbe 0x2260dc testb $0x8, %bpl pushq $0x8 popq %rbx jne 0x225003 movq %rbp, %rax movq %rbp, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %rbx, %rbp vmovaps %xmm11, %xmm14 movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x224f9a vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x280(%rsp), %xmm1 vmovaps 0x290(%rsp), %xmm2 vmovaps 0x2a0(%rsp), %xmm3 vmovaps 0x2e0(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0x2f0(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x300(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vminps %xmm1, %xmm5, %xmm0 vminps %xmm2, %xmm7, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vminps %xmm3, %xmm9, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vmaxps %xmm1, %xmm5, %xmm1 vmaxps %xmm2, %xmm7, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmaxps %xmm3, %xmm9, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmulps %xmm1, %xmm12, %xmm1 vmaxps 0x340(%rsp), %xmm0, %xmm2 vminps 0x350(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vtestps %xmm1, %xmm1 je 0x224f9a vblendvps %xmm1, %xmm0, %xmm11, %xmm0 cmpq $0x8, %rbp je 0x224f93 movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm14, (%r12) addq $0x10, %r12 vmovaps %xmm0, %xmm14 movq %rdi, %rbp cmpq $0x8, %rdi je 0x224fb1 leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x224e8a xorl %eax, %eax cmpq $0x8, %rbp je 0x224ffc vmovaps 0x350(%rsp), %xmm0 vcmpnleps %xmm14, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r11 jae 0x224feb testb %cl, %cl je 0x2260dc testb $0x8, %bpl je 0x224e76 jmp 0x225003 movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm14, (%r12) addq $0x10, %r12 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x224fd7 cmpq $-0x8, %rbp je 0x2260d9 movq %r13, 0x38(%rsp) vmovaps 0x350(%rsp), %xmm0 vcmpnleps %xmm14, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x2261d6 movl %ebp, %edi andl $0xf, %edi vmovdqa 0x90(%rsp), %xmm0 addq $-0x8, %rdi je 0x226096 andq $-0x10, %rbp vpxor %xmm0, %xmm13, %xmm1 leaq 0xc0(%rbp), %rbx xorl %eax, %eax movq %rax, 0x1c0(%rsp) imulq $0xe0, %rax, %rax addq %rbp, %rax movq %rax, 0x1c8(%rsp) xorl %r13d, %r13d vmovdqa %xmm1, 0x40(%rsp) vmovdqa %xmm1, 0x1d0(%rsp) movl (%rbx,%r13,4), %eax movl $0xffffffff, %ecx # imm = 0xFFFFFFFF cmpq %rcx, %rax je 0x226062 vbroadcastss -0xc0(%rbx,%r13,4), %xmm14 vbroadcastss -0xb0(%rbx,%r13,4), %xmm9 vbroadcastss -0xa0(%rbx,%r13,4), %xmm3 vbroadcastss -0x90(%rbx,%r13,4), %xmm0 vmovaps %xmm0, 0x100(%rsp) vbroadcastss -0x80(%rbx,%r13,4), %xmm0 vmovaps %xmm0, 0xf0(%rsp) vbroadcastss -0x70(%rbx,%r13,4), %xmm0 vmovaps %xmm0, 0xe0(%rsp) vbroadcastss -0x60(%rbx,%r13,4), %xmm0 vmovaps %xmm0, 0x140(%rsp) vbroadcastss -0x50(%rbx,%r13,4), %xmm0 vmovaps %xmm0, 0x130(%rsp) vbroadcastss -0x40(%rbx,%r13,4), %xmm0 vmovaps %xmm0, 0x120(%rsp) vbroadcastss -0x30(%rbx,%r13,4), %xmm12 vbroadcastss -0x20(%rbx,%r13,4), %xmm5 vbroadcastss -0x10(%rbx,%r13,4), %xmm7 movq 0x1c8(%rsp), %rcx vmovaps 0xd0(%rcx), %xmm0 vmovaps %xmm0, 0x1e0(%rsp) vmovaps 0x10(%r9), %xmm6 vmovaps 0x20(%r9), %xmm11 vsubps %xmm6, %xmm9, %xmm4 vsubps %xmm11, %xmm3, %xmm3 vmovaps %xmm5, 0x3d0(%rsp) vsubps %xmm6, %xmm5, %xmm5 vmovaps %xmm5, 0xa0(%rsp) vmovaps %xmm7, 0x3c0(%rsp) vsubps %xmm11, %xmm7, %xmm0 vmovaps %xmm0, 0x70(%rsp) vsubps %xmm4, %xmm5, %xmm2 vsubps %xmm3, %xmm0, %xmm1 vaddps %xmm4, %xmm5, %xmm7 vaddps %xmm3, %xmm0, %xmm8 vmulps %xmm1, %xmm7, %xmm5 vmulps %xmm2, %xmm8, %xmm10 vsubps %xmm5, %xmm10, %xmm13 vmovaps (%r9), %xmm15 vsubps %xmm15, %xmm14, %xmm0 vmovaps %xmm12, 0x110(%rsp) vsubps %xmm15, %xmm12, %xmm5 vsubps %xmm0, %xmm5, %xmm9 vmulps %xmm8, %xmm9, %xmm8 vaddps %xmm0, %xmm5, %xmm10 vmovaps %xmm1, 0x160(%rsp) vmulps %xmm1, %xmm10, %xmm12 vsubps %xmm8, %xmm12, %xmm12 vmovaps %xmm2, 0x180(%rsp) vmulps %xmm2, %xmm10, %xmm8 vmovaps %xmm9, 0x150(%rsp) vmulps %xmm7, %xmm9, %xmm7 vsubps %xmm8, %xmm7, %xmm7 vmovaps 0x60(%r9), %xmm14 vmulps %xmm7, %xmm14, %xmm7 vmovaps %xmm14, 0x20(%rsp) vmovaps 0x50(%r9), %xmm9 vmulps %xmm12, %xmm9, %xmm12 vaddps %xmm7, %xmm12, %xmm7 vmovaps 0x40(%r9), %xmm1 vmovaps %xmm1, 0x1b0(%rsp) vmulps %xmm1, %xmm13, %xmm13 vaddps %xmm7, %xmm13, %xmm12 vmovaps 0xf0(%rsp), %xmm1 vsubps %xmm6, %xmm1, %xmm1 vmovaps 0xe0(%rsp), %xmm2 vsubps %xmm11, %xmm2, %xmm2 vsubps %xmm1, %xmm4, %xmm13 vsubps %xmm2, %xmm3, %xmm7 vmovaps %xmm4, 0x1a0(%rsp) vaddps %xmm1, %xmm4, %xmm4 vmovaps %xmm3, 0x190(%rsp) vaddps %xmm2, %xmm3, %xmm3 vmulps %xmm7, %xmm4, %xmm6 vmulps %xmm3, %xmm13, %xmm8 vsubps %xmm6, %xmm8, %xmm6 vmovaps 0x100(%rsp), %xmm8 vsubps %xmm15, %xmm8, %xmm8 vsubps %xmm8, %xmm0, %xmm15 vmulps %xmm3, %xmm15, %xmm3 vmovaps %xmm0, 0x170(%rsp) vaddps %xmm0, %xmm8, %xmm11 vmulps %xmm7, %xmm11, %xmm10 vsubps %xmm3, %xmm10, %xmm3 vmulps %xmm13, %xmm11, %xmm10 vmulps %xmm4, %xmm15, %xmm0 vsubps %xmm10, %xmm0, %xmm0 vmulps %xmm0, %xmm14, %xmm0 vmulps %xmm3, %xmm9, %xmm3 vaddps %xmm3, %xmm0, %xmm0 vmovaps 0x1b0(%rsp), %xmm14 vmulps %xmm6, %xmm14, %xmm3 vaddps %xmm0, %xmm3, %xmm11 vsubps %xmm5, %xmm8, %xmm3 vaddps %xmm5, %xmm8, %xmm5 vmovaps 0xa0(%rsp), %xmm4 vsubps %xmm4, %xmm1, %xmm0 vaddps %xmm4, %xmm1, %xmm1 vmovaps 0x70(%rsp), %xmm4 vsubps %xmm4, %xmm2, %xmm10 vaddps %xmm4, %xmm2, %xmm2 vmulps %xmm1, %xmm10, %xmm6 vmulps %xmm2, %xmm0, %xmm8 vsubps %xmm6, %xmm8, %xmm6 vmulps %xmm2, %xmm3, %xmm2 vmulps %xmm5, %xmm10, %xmm8 vsubps %xmm2, %xmm8, %xmm2 vmulps %xmm0, %xmm5, %xmm5 vmulps %xmm1, %xmm3, %xmm1 vsubps %xmm5, %xmm1, %xmm1 vmulps 0x20(%rsp), %xmm1, %xmm1 vmovaps %xmm9, 0x70(%rsp) vmulps %xmm2, %xmm9, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps %xmm6, %xmm14, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm11, %xmm12, %xmm2 vaddps %xmm2, %xmm1, %xmm4 vminps %xmm11, %xmm12, %xmm2 vminps %xmm1, %xmm2, %xmm2 vbroadcastss 0x1cfbb79(%rip), %xmm5 # 0x1f20ec4 vandps %xmm5, %xmm4, %xmm6 vbroadcastss 0x1cfbb74(%rip), %xmm5 # 0x1f20ecc vmovaps %xmm6, 0x3b0(%rsp) vmulps %xmm5, %xmm6, %xmm5 vbroadcastss 0x1cfbb52(%rip), %xmm6 # 0x1f20ec0 vxorps %xmm6, %xmm5, %xmm6 vcmpnltps %xmm6, %xmm2, %xmm2 vmovaps %xmm12, 0xa0(%rsp) vmaxps %xmm11, %xmm12, %xmm6 vmaxps %xmm1, %xmm6, %xmm1 vcmpleps %xmm5, %xmm1, %xmm1 vorps %xmm1, %xmm2, %xmm9 vtestps 0x40(%rsp), %xmm9 je 0x225fe7 vmovaps %xmm4, 0x390(%rsp) vmovaps %xmm11, 0x3a0(%rsp) vmovaps 0x160(%rsp), %xmm11 vmulps %xmm11, %xmm13, %xmm1 vmovaps 0x180(%rsp), %xmm4 vmulps %xmm7, %xmm4, %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmulps %xmm7, %xmm0, %xmm5 vmulps %xmm10, %xmm13, %xmm6 vsubps %xmm5, %xmm6, %xmm6 vbroadcastss 0x1cfbade(%rip), %xmm8 # 0x1f20ec4 vandps %xmm1, %xmm8, %xmm1 vandps %xmm5, %xmm8, %xmm5 vcmpltps %xmm5, %xmm1, %xmm1 vblendvps %xmm1, %xmm2, %xmm6, %xmm12 vmulps %xmm10, %xmm15, %xmm1 vmulps %xmm11, %xmm15, %xmm2 vmovaps 0x150(%rsp), %xmm10 vmulps %xmm7, %xmm10, %xmm5 vsubps %xmm5, %xmm2, %xmm2 vmulps %xmm7, %xmm3, %xmm6 vsubps %xmm1, %xmm6, %xmm6 vandps %xmm5, %xmm8, %xmm5 vandps %xmm1, %xmm8, %xmm1 vcmpltps %xmm1, %xmm5, %xmm1 vblendvps %xmm1, %xmm2, %xmm6, %xmm6 vmulps %xmm3, %xmm13, %xmm1 vmulps %xmm13, %xmm10, %xmm2 vmulps %xmm4, %xmm15, %xmm3 vmulps %xmm0, %xmm15, %xmm0 vsubps %xmm3, %xmm2, %xmm2 vsubps %xmm1, %xmm0, %xmm0 vandps %xmm3, %xmm8, %xmm3 vandps %xmm1, %xmm8, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm0, %xmm5 vmulps 0x20(%rsp), %xmm5, %xmm0 vmulps 0x70(%rsp), %xmm6, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm12, %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm1 vmulps 0x190(%rsp), %xmm5, %xmm0 vmulps 0x1a0(%rsp), %xmm6, %xmm2 vaddps %xmm2, %xmm0, %xmm0 vmulps 0x170(%rsp), %xmm12, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vrcpps %xmm1, %xmm2 vmulps %xmm2, %xmm1, %xmm3 vbroadcastss 0x1cc7264(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vmulps %xmm2, %xmm0, %xmm0 vmovaps 0x30(%r9), %xmm2 vmovaps 0x80(%r9), %xmm15 vcmpleps %xmm0, %xmm2, %xmm2 vcmpleps %xmm15, %xmm0, %xmm3 vandps %xmm3, %xmm2, %xmm2 vxorps %xmm14, %xmm14, %xmm14 vcmpneqps %xmm1, %xmm14, %xmm1 vandps %xmm2, %xmm1, %xmm2 vandps 0x40(%rsp), %xmm9, %xmm1 vpslld $0x1f, %xmm2, %xmm2 vpsrad $0x1f, %xmm2, %xmm2 vtestps %xmm1, %xmm2 vbroadcastss 0x1cc6516(%rip), %xmm11 # 0x1eeba20 je 0x225ff0 vmovaps %xmm12, %xmm9 vandps %xmm1, %xmm2, %xmm1 movq (%r10), %rcx movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %r14 vbroadcastss 0x34(%r14), %xmm2 vandps 0x90(%r9), %xmm2, %xmm2 vpcmpeqd %xmm2, %xmm14, %xmm2 vtestps %xmm1, %xmm2 vbroadcastss 0x1cfa9c8(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cfa9c3(%rip), %xmm12 # 0x1f1ff14 vpcmpeqd %xmm13, %xmm13, %xmm13 jb 0x22558e movl 0x1e0(%rsp,%r13,4), %ecx vandnps %xmm1, %xmm2, %xmm1 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x225a97 cmpq $0x0, 0x48(%r14) jne 0x225a97 vmovdqa 0x40(%rsp), %xmm0 vpandn %xmm0, %xmm1, %xmm0 vmovdqa %xmm0, 0x40(%rsp) vmovaps 0x40(%rsp), %xmm0 vtestps %xmm0, %xmm0 je 0x225a6e vmovaps (%r9), %xmm1 vmovaps 0x10(%r9), %xmm3 vmovaps 0x20(%r9), %xmm5 vmovaps 0x140(%rsp), %xmm0 vsubps %xmm1, %xmm0, %xmm9 vmovaps 0x130(%rsp), %xmm0 vsubps %xmm3, %xmm0, %xmm4 vmovaps 0x120(%rsp), %xmm0 vsubps %xmm5, %xmm0, %xmm15 vmovaps 0x110(%rsp), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0x70(%rsp) vmovaps 0x3d0(%rsp), %xmm2 vsubps %xmm3, %xmm2, %xmm6 vmovaps 0x3c0(%rsp), %xmm2 vsubps %xmm5, %xmm2, %xmm8 vmovaps 0x100(%rsp), %xmm2 vsubps %xmm1, %xmm2, %xmm13 vmovaps 0xf0(%rsp), %xmm2 vsubps %xmm3, %xmm2, %xmm14 vmovaps 0xe0(%rsp), %xmm2 vsubps %xmm5, %xmm2, %xmm1 vmovaps %xmm1, 0x20(%rsp) vsubps %xmm9, %xmm13, %xmm2 vsubps %xmm4, %xmm14, %xmm12 vsubps %xmm15, %xmm1, %xmm0 vaddps %xmm4, %xmm14, %xmm3 vaddps %xmm1, %xmm15, %xmm7 vmulps %xmm0, %xmm3, %xmm10 vmulps %xmm7, %xmm12, %xmm11 vsubps %xmm10, %xmm11, %xmm1 vaddps %xmm9, %xmm13, %xmm10 vmulps %xmm7, %xmm2, %xmm7 vmovaps %xmm0, 0x190(%rsp) vmulps %xmm0, %xmm10, %xmm11 vsubps %xmm7, %xmm11, %xmm7 vmovaps %xmm12, 0x1a0(%rsp) vmulps %xmm12, %xmm10, %xmm10 vmovaps %xmm2, 0x1b0(%rsp) vmulps %xmm3, %xmm2, %xmm3 vsubps %xmm10, %xmm3, %xmm3 vmovaps 0x60(%r9), %xmm12 vmulps %xmm3, %xmm12, %xmm3 vmovaps 0x50(%r9), %xmm0 vmovaps %xmm0, 0x100(%rsp) vmulps %xmm7, %xmm0, %xmm7 vaddps %xmm7, %xmm3, %xmm3 vmovaps 0x40(%r9), %xmm5 vmulps %xmm1, %xmm5, %xmm7 vaddps %xmm3, %xmm7, %xmm11 vmovaps %xmm6, %xmm2 vsubps %xmm6, %xmm4, %xmm7 vmovaps %xmm8, %xmm6 vsubps %xmm8, %xmm15, %xmm1 vmovaps %xmm4, 0xe0(%rsp) vaddps %xmm2, %xmm4, %xmm3 vmovaps %xmm2, %xmm8 vmovaps %xmm15, 0xa0(%rsp) vaddps %xmm6, %xmm15, %xmm2 vmulps %xmm1, %xmm3, %xmm15 vmulps %xmm2, %xmm7, %xmm4 vsubps %xmm15, %xmm4, %xmm4 vmovaps 0x70(%rsp), %xmm15 vsubps %xmm15, %xmm9, %xmm0 vmulps %xmm2, %xmm0, %xmm2 vmovaps %xmm9, 0xf0(%rsp) vaddps %xmm15, %xmm9, %xmm9 vmovaps %xmm1, 0x170(%rsp) vmulps %xmm1, %xmm9, %xmm10 vsubps %xmm2, %xmm10, %xmm2 vmovaps %xmm7, 0x180(%rsp) vmulps %xmm7, %xmm9, %xmm9 vmovaps %xmm0, 0x160(%rsp) vmulps %xmm3, %xmm0, %xmm3 vsubps %xmm9, %xmm3, %xmm3 vmulps %xmm3, %xmm12, %xmm3 vmovaps 0x100(%rsp), %xmm7 vmulps %xmm2, %xmm7, %xmm2 vaddps %xmm2, %xmm3, %xmm2 vmulps %xmm4, %xmm5, %xmm3 vaddps %xmm2, %xmm3, %xmm9 vsubps %xmm13, %xmm15, %xmm3 vaddps %xmm13, %xmm15, %xmm1 vsubps %xmm14, %xmm8, %xmm15 vaddps %xmm14, %xmm8, %xmm2 vmovaps 0x20(%rsp), %xmm0 vsubps %xmm0, %xmm6, %xmm10 vaddps %xmm0, %xmm6, %xmm4 vmulps %xmm2, %xmm10, %xmm0 vmulps %xmm4, %xmm15, %xmm6 vsubps %xmm0, %xmm6, %xmm0 vmulps %xmm4, %xmm3, %xmm4 vmulps %xmm1, %xmm10, %xmm6 vsubps %xmm4, %xmm6, %xmm4 vmovaps 0x40(%rsp), %xmm6 vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm2, %xmm3, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm12, 0x20(%rsp) vmulps %xmm1, %xmm12, %xmm1 vmulps %xmm4, %xmm7, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmovaps %xmm5, 0x70(%rsp) vmulps %xmm0, %xmm5, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm9, %xmm11, %xmm2 vaddps %xmm2, %xmm1, %xmm4 vminps %xmm9, %xmm11, %xmm2 vminps %xmm1, %xmm2, %xmm2 vmovaps %xmm4, 0x130(%rsp) vbroadcastss 0x1cfb6d1(%rip), %xmm0 # 0x1f20ec4 vmovaps %xmm0, %xmm13 vandps %xmm0, %xmm4, %xmm5 vbroadcastss 0x1cfb6c8(%rip), %xmm4 # 0x1f20ecc vmovaps %xmm5, 0x120(%rsp) vmulps %xmm4, %xmm5, %xmm4 vbroadcastss 0x1cfb6a6(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm4, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm11, 0x140(%rsp) vmaxps %xmm9, %xmm11, %xmm5 vmaxps %xmm1, %xmm5, %xmm1 vcmpleps %xmm4, %xmm1, %xmm1 vorps %xmm1, %xmm2, %xmm0 vtestps %xmm6, %xmm0 je 0x225d2e vmovaps %xmm7, %xmm12 vmovaps %xmm9, 0x110(%rsp) vmovaps 0x190(%rsp), %xmm11 vmovaps %xmm0, 0x150(%rsp) vmovaps 0x180(%rsp), %xmm0 vmulps %xmm0, %xmm11, %xmm1 vmovaps 0x1a0(%rsp), %xmm9 vmovaps 0x170(%rsp), %xmm8 vmulps %xmm8, %xmm9, %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmulps %xmm8, %xmm15, %xmm4 vmulps %xmm0, %xmm10, %xmm5 vsubps %xmm4, %xmm5, %xmm5 vandps %xmm1, %xmm13, %xmm1 vandps %xmm4, %xmm13, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm14 vmovaps %xmm15, %xmm7 vmovaps 0x160(%rsp), %xmm15 vmulps %xmm10, %xmm15, %xmm1 vmulps %xmm11, %xmm15, %xmm2 vmovaps 0x1b0(%rsp), %xmm11 vmulps %xmm8, %xmm11, %xmm4 vsubps %xmm4, %xmm2, %xmm2 vmulps %xmm3, %xmm8, %xmm5 vsubps %xmm1, %xmm5, %xmm5 vandps %xmm4, %xmm13, %xmm4 vandps %xmm1, %xmm13, %xmm1 vcmpltps %xmm1, %xmm4, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm8 vmulps %xmm0, %xmm3, %xmm1 vmulps %xmm0, %xmm11, %xmm2 vmulps %xmm9, %xmm15, %xmm3 vmulps %xmm7, %xmm15, %xmm0 vsubps %xmm3, %xmm2, %xmm2 vsubps %xmm1, %xmm0, %xmm0 vandps %xmm3, %xmm13, %xmm3 vandps %xmm1, %xmm13, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm0, %xmm5 vmulps 0x20(%rsp), %xmm5, %xmm0 vmulps %xmm8, %xmm12, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps 0x70(%rsp), %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm1 vmulps 0xa0(%rsp), %xmm5, %xmm0 vmulps 0xe0(%rsp), %xmm8, %xmm2 vaddps %xmm2, %xmm0, %xmm0 vmovaps %xmm14, %xmm9 vmulps 0xf0(%rsp), %xmm14, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vrcpps %xmm1, %xmm2 vmulps %xmm2, %xmm1, %xmm3 vbroadcastss 0x1cc6d9a(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vmulps %xmm2, %xmm0, %xmm0 vmovaps 0x30(%r9), %xmm2 vmovaps 0x80(%r9), %xmm15 vcmpleps %xmm0, %xmm2, %xmm2 vcmpleps %xmm15, %xmm0, %xmm3 vandps %xmm3, %xmm2, %xmm2 vxorps %xmm14, %xmm14, %xmm14 vcmpneqps %xmm1, %xmm14, %xmm1 vandps %xmm2, %xmm1, %xmm2 vandps 0x150(%rsp), %xmm6, %xmm1 vpslld $0x1f, %xmm2, %xmm2 vpsrad $0x1f, %xmm2, %xmm2 vtestps %xmm1, %xmm2 je 0x225d2e vandps %xmm1, %xmm2, %xmm1 movq (%r10), %rcx movl (%rbx,%r13,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %r14 vbroadcastss 0x34(%r14), %xmm2 vandps 0x90(%r9), %xmm2, %xmm2 vpcmpeqd %xmm2, %xmm14, %xmm2 vtestps %xmm1, %xmm2 vbroadcastss 0x1cc6015(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cfa4fc(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cfa4f7(%rip), %xmm12 # 0x1f1ff14 jb 0x225d49 movl 0x1e0(%rsp,%r13,4), %ecx vandnps %xmm1, %xmm2, %xmm1 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x225d53 cmpq $0x0, 0x48(%r14) jne 0x225d53 vpcmpeqd %xmm13, %xmm13, %xmm13 vmovdqa 0x40(%rsp), %xmm0 vpandn %xmm0, %xmm1, %xmm0 vmovdqa %xmm0, 0x40(%rsp) vmovaps 0x40(%rsp), %xmm0 vtestps %xmm0, %xmm0 setne %al jmp 0x225a70 xorl %eax, %eax vmovdqa 0x1d0(%rsp), %xmm1 testb %al, %al je 0x226062 leaq 0x1(%r13), %rax cmpq $0x3, %r13 movq %rax, %r13 jb 0x225081 jmp 0x226062 vmovaps 0x390(%rsp), %xmm3 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cc6c63(%rip), %xmm7 # 0x1eec714 vsubps %xmm3, %xmm7, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1ccb522(%rip), %xmm3 # 0x1ef0fe8 vmovaps 0x3b0(%rsp), %xmm4 vcmpnltps %xmm3, %xmm4, %xmm3 vandps %xmm2, %xmm3, %xmm2 vmulps 0xa0(%rsp), %xmm2, %xmm3 vminps %xmm7, %xmm3, %xmm3 vmulps 0x3a0(%rsp), %xmm2, %xmm2 vminps %xmm7, %xmm2, %xmm2 vsubps %xmm3, %xmm7, %xmm4 vsubps %xmm2, %xmm7, %xmm7 vmovaps 0x370(%rsp), %xmm8 vblendvps %xmm8, %xmm4, %xmm3, %xmm3 vblendvps %xmm8, %xmm7, %xmm2, %xmm2 movq 0x8(%r10), %rdx vmovd %eax, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovd %ecx, %xmm7 vpshufd $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmovaps %xmm9, 0x1f0(%rsp) vmovaps %xmm6, 0x200(%rsp) vmovaps %xmm5, 0x210(%rsp) vmovaps %xmm3, 0x220(%rsp) vmovaps %xmm2, 0x230(%rsp) vmovdqa %xmm7, 0x240(%rsp) vmovdqa %xmm4, 0x250(%rsp) leaq 0x260(%rsp), %rax vcmptrueps %ymm2, %ymm2, %ymm2 vmovups %ymm2, (%rax) vbroadcastss (%rdx), %xmm2 vmovaps %xmm2, 0x260(%rsp) vbroadcastss 0x4(%rdx), %xmm2 vmovaps %xmm2, 0x270(%rsp) vblendvps %xmm1, %xmm0, %xmm15, %xmm0 vmovaps %xmm0, 0x80(%r9) vmovaps %xmm1, 0x80(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%r14), %rax movq %rax, 0xb8(%rsp) movq 0x8(%r10), %rax movq %rax, 0xc0(%rsp) movq %r9, 0xc8(%rsp) leaq 0x1f0(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x4, 0xd8(%rsp) movq 0x48(%r14), %rax testq %rax, %rax movq %r10, 0x68(%rsp) movq %rsi, 0x10(%rsp) movq %r8, 0x8(%rsp) movq %r11, (%rsp) movq %rdi, 0x60(%rsp) vmovaps %xmm15, 0x20(%rsp) je 0x225c71 leaq 0xb0(%rsp), %rdi movq %r9, 0x58(%rsp) vzeroupper callq %rax vmovaps 0x20(%rsp), %xmm15 movq 0x60(%rsp), %rdi vxorps %xmm14, %xmm14, %xmm14 vbroadcastss 0x1cfa2cd(%rip), %xmm12 # 0x1f1ff14 vbroadcastss 0x1cfa2c0(%rip), %xmm10 # 0x1f1ff10 movq (%rsp), %r11 vbroadcastss 0x1cc5dc3(%rip), %xmm11 # 0x1eeba20 movq 0x8(%rsp), %r8 movq 0x10(%rsp), %rsi movq 0x58(%rsp), %r9 movq 0x68(%rsp), %r10 vmovdqa 0x80(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x22600c movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax vpcmpeqd %xmm13, %xmm13, %xmm13 je 0x225cf9 testb $0x2, (%rcx) jne 0x225ca3 testb $0x40, 0x3e(%r14) je 0x225cf9 leaq 0xb0(%rsp), %rdi movq %r9, %r14 vzeroupper callq %rax vmovaps 0x20(%rsp), %xmm15 movq 0x60(%rsp), %rdi vxorps %xmm14, %xmm14, %xmm14 vpcmpeqd %xmm13, %xmm13, %xmm13 vbroadcastss 0x1cfa243(%rip), %xmm12 # 0x1f1ff14 vbroadcastss 0x1cfa236(%rip), %xmm10 # 0x1f1ff10 movq (%rsp), %r11 vbroadcastss 0x1cc5d39(%rip), %xmm11 # 0x1eeba20 movq 0x8(%rsp), %r8 movq 0x10(%rsp), %rsi movq %r14, %r9 movq 0x68(%rsp), %r10 vpcmpeqd 0x80(%rsp), %xmm14, %xmm1 vpxor %xmm1, %xmm13, %xmm0 movq 0xc8(%rsp), %rax vbroadcastss 0x1cc6e6d(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x226019 vbroadcastss 0x1cc5ce9(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cfa1d0(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cfa1cb(%rip), %xmm12 # 0x1f1ff14 vpcmpeqd %xmm13, %xmm13, %xmm13 jmp 0x225a5e vmovaps 0x130(%rsp), %xmm3 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cc69a7(%rip), %xmm6 # 0x1eec714 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1ccb266(%rip), %xmm3 # 0x1ef0fe8 vmovaps 0x120(%rsp), %xmm4 vcmpnltps %xmm3, %xmm4, %xmm3 vandps %xmm2, %xmm3, %xmm2 vmulps 0x140(%rsp), %xmm2, %xmm3 vminps %xmm6, %xmm3, %xmm3 vmulps 0x110(%rsp), %xmm2, %xmm2 vminps %xmm6, %xmm2, %xmm2 vsubps %xmm3, %xmm6, %xmm4 vsubps %xmm2, %xmm6, %xmm6 vmovaps 0x360(%rsp), %xmm7 vblendvps %xmm7, %xmm4, %xmm3, %xmm3 vblendvps %xmm7, %xmm6, %xmm2, %xmm2 movq 0x8(%r10), %rdx vmovd %eax, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovd %ecx, %xmm6 vpshufd $0x0, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmovaps %xmm9, 0x1f0(%rsp) vmovaps %xmm8, 0x200(%rsp) vmovaps %xmm5, 0x210(%rsp) vmovaps %xmm3, 0x220(%rsp) vmovaps %xmm2, 0x230(%rsp) vmovdqa %xmm6, 0x240(%rsp) vmovdqa %xmm4, 0x250(%rsp) leaq 0x260(%rsp), %rax vcmptrueps %ymm2, %ymm2, %ymm2 vmovups %ymm2, (%rax) vbroadcastss (%rdx), %xmm2 vmovaps %xmm2, 0x260(%rsp) vbroadcastss 0x4(%rdx), %xmm2 vmovaps %xmm2, 0x270(%rsp) vblendvps %xmm1, %xmm0, %xmm15, %xmm0 vmovaps %xmm0, 0x80(%r9) vmovaps %xmm1, 0x80(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%r14), %rax movq %rax, 0xb8(%rsp) movq 0x8(%r10), %rax movq %rax, 0xc0(%rsp) movq %r9, 0xc8(%rsp) leaq 0x1f0(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x4, 0xd8(%rsp) movq 0x48(%r14), %rax testq %rax, %rax movq %r10, 0x68(%rsp) movq %rsi, 0x10(%rsp) movq %r8, 0x8(%rsp) movq %r11, (%rsp) movq %rdi, 0x60(%rsp) vmovaps %xmm15, 0x20(%rsp) je 0x225f2d leaq 0xb0(%rsp), %rdi movq %r9, 0x58(%rsp) vzeroupper callq %rax vmovaps 0x20(%rsp), %xmm15 movq 0x60(%rsp), %rdi vxorps %xmm14, %xmm14, %xmm14 vbroadcastss 0x1cfa011(%rip), %xmm12 # 0x1f1ff14 vbroadcastss 0x1cfa004(%rip), %xmm10 # 0x1f1ff10 movq (%rsp), %r11 vbroadcastss 0x1cc5b07(%rip), %xmm11 # 0x1eeba20 movq 0x8(%rsp), %r8 movq 0x10(%rsp), %rsi movq 0x58(%rsp), %r9 movq 0x68(%rsp), %r10 vmovdqa 0x80(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x226037 movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax vpcmpeqd %xmm13, %xmm13, %xmm13 je 0x225fb5 testb $0x2, (%rcx) jne 0x225f5f testb $0x40, 0x3e(%r14) je 0x225fb5 leaq 0xb0(%rsp), %rdi movq %r9, %r14 vzeroupper callq %rax vmovaps 0x20(%rsp), %xmm15 movq 0x60(%rsp), %rdi vxorps %xmm14, %xmm14, %xmm14 vpcmpeqd %xmm13, %xmm13, %xmm13 vbroadcastss 0x1cf9f87(%rip), %xmm12 # 0x1f1ff14 vbroadcastss 0x1cf9f7a(%rip), %xmm10 # 0x1f1ff10 movq (%rsp), %r11 vbroadcastss 0x1cc5a7d(%rip), %xmm11 # 0x1eeba20 movq 0x8(%rsp), %r8 movq 0x10(%rsp), %rsi movq %r14, %r9 movq 0x68(%rsp), %r10 vpcmpeqd 0x80(%rsp), %xmm14, %xmm1 vpxor %xmm1, %xmm13, %xmm0 movq 0xc8(%rsp), %rax vbroadcastss 0x1cc6bb1(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x226044 vbroadcastss 0x1cc5a30(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cf9f17(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cf9f12(%rip), %xmm12 # 0x1f1ff14 vpcmpeqd %xmm13, %xmm13, %xmm13 jmp 0x22558e vpcmpeqd %xmm0, %xmm14, %xmm0 vpcmpeqd %xmm13, %xmm13, %xmm13 vpxor %xmm0, %xmm13, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm1 movq 0x38(%rsp), %rax vblendvps %xmm0, (%rax), %xmm15, %xmm0 vmovaps %xmm0, (%rax) jmp 0x22557e vpcmpeqd %xmm0, %xmm14, %xmm0 vpcmpeqd %xmm13, %xmm13, %xmm13 vpxor %xmm0, %xmm13, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm1 movq 0x38(%rsp), %rax vblendvps %xmm0, (%rax), %xmm15, %xmm0 vmovaps %xmm0, (%rax) jmp 0x225a4e vmovdqa 0x40(%rsp), %xmm2 vpand %xmm1, %xmm2, %xmm0 vtestps %xmm1, %xmm2 movq 0x1c0(%rsp), %rax je 0x226092 incq %rax addq $0xe0, %rbx vmovdqa %xmm0, %xmm1 cmpq %rdi, %rax jb 0x225055 vpxor %xmm0, %xmm13, %xmm0 vpor 0x90(%rsp), %xmm0, %xmm0 vmovdqa %xmm0, 0x90(%rsp) vtestps %xmm13, %xmm0 movq 0x38(%rsp), %r13 jb 0x2260d9 vmovaps 0x350(%rsp), %xmm1 vbroadcastss 0x1cc6abe(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x350(%rsp) xorl %eax, %eax jmp 0x2260dc pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x224e1a jmp 0x2261e3 pushq $0x2 jmp 0x2260db vmovaps %xmm14, 0x20(%rsp) movq %r13, 0x38(%rsp) movq %r11, (%rsp) movq %r9, 0x58(%rsp) movq %rsi, 0x10(%rsp) movq %r8, 0x8(%rsp) bsfq %r14, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, 0x40(%rsp) leaq 0x1f(%rsp), %r8 movq %r10, %r13 pushq %r10 leaq 0x288(%rsp), %rax pushq %rax vzeroupper callq 0x264e02 popq %rcx popq %rdx testb %al, %al je 0x22614c movq 0x40(%rsp), %rax orl $-0x1, 0x90(%rsp,%rax,4) leaq -0x1(%r14), %rax andq %rax, %r14 movq %r13, %r10 movq 0x58(%rsp), %r9 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %r8 jne 0x22610c vmovaps 0x90(%rsp), %xmm0 vpcmpeqd %xmm13, %xmm13, %xmm13 vtestps %xmm13, %xmm0 pushq $0x3 popq %rax vbroadcastss 0x1cc589a(%rip), %xmm11 # 0x1eeba20 movq (%rsp), %r11 movq 0x38(%rsp), %r13 vbroadcastss 0x1cf9d78(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cf9d73(%rip), %xmm12 # 0x1f1ff14 vmovaps 0x20(%rsp), %xmm14 jb 0x224e60 vmovaps 0x350(%rsp), %xmm1 vbroadcastss 0x1cc69c5(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x350(%rsp) pushq $0x2 popq %rax jmp 0x224e60 pushq $0x2 popq %rax movq 0x38(%rsp), %r13 jmp 0x2260dc vmovaps 0x380(%rsp), %xmm0 vandps 0x90(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cc6986(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r13) addq $0x1a88, %rsp # imm = 0x1A88 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %r9, %rdx movq %r10, %rcx jmp 0x263460	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiIntersectorKPluecker<4, 4, true>>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x226407 movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 cmpq $0x0, 0x8(%rcx) je 0x226261 movq 0x10(%r15), %rax testb $0x1, 0x2(%rax) jne 0x226419 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x226407 movzbl %al, %ebp vmovaps (%r14), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%r14), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%r14), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%r14), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cfabfe(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1ccad15(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1cc6433(%rip), %xmm7 # 0x1eec714 vdivps %xmm1, %xmm7, %xmm1 vandps %xmm4, %xmm2, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vdivps %xmm2, %xmm7, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vbroadcastss 0x1cfac5c(%rip), %xmm6 # 0x1f20f60 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vblendvps %xmm8, %xmm6, %xmm2, %xmm2 vdivps %xmm3, %xmm7, %xmm3 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d34626(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d2e648(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d34607(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d345ef(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d345ea(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%r14), %xmm1 vmovaps 0x80(%r14), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cc566c(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cc67b9(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x2675c4 popq %rax popq %rcx andq %rbx, %rbp jne 0x2263d9 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r12, %rsi movq %r14, %rdx movq %r15, %rcx addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x265b8c	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiIntersectorKPluecker<4, 4, true>>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1a78, %rsp # imm = 0x1A78 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x227a76 movq %rcx, %r10 movq %rdx, %r14 movq 0x8(%rcx), %rax testq %rax, %rax je 0x226474 movq 0x10(%r10), %rcx testb $0x1, 0x2(%rcx) jne 0x227a8b vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%r14), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x227a76 vandps %xmm3, %xmm4, %xmm10 vmovaps (%r14), %xmm3 vmovaps %xmm3, 0x240(%rsp) vmovaps 0x10(%r14), %xmm3 vmovaps %xmm3, 0x250(%rsp) vmovaps 0x20(%r14), %xmm3 vmovaps %xmm3, 0x260(%rsp) vmovaps 0x40(%r14), %xmm3 vmovaps %xmm3, 0x270(%rsp) vmovaps 0x50(%r14), %xmm4 vmovaps %xmm4, 0x280(%rsp) vmovaps 0x60(%r14), %xmm5 vmovaps %xmm5, 0x290(%rsp) vbroadcastss 0x1cfa9c5(%rip), %xmm8 # 0x1f20ec4 vandps %xmm3, %xmm8, %xmm6 vbroadcastss 0x1ccaadc(%rip), %xmm9 # 0x1ef0fe8 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cc61f9(%rip), %xmm11 # 0x1eec714 vdivps %xmm3, %xmm11, %xmm3 vandps %xmm4, %xmm8, %xmm7 vcmpltps %xmm9, %xmm7, %xmm7 vdivps %xmm4, %xmm11, %xmm4 vandps %xmm5, %xmm8, %xmm8 vcmpltps %xmm9, %xmm8, %xmm8 vbroadcastss 0x1cfaa20(%rip), %xmm9 # 0x1f20f60 vblendvps %xmm6, %xmm9, %xmm3, %xmm3 vblendvps %xmm7, %xmm9, %xmm4, %xmm4 vdivps %xmm5, %xmm11, %xmm5 vblendvps %xmm8, %xmm9, %xmm5, %xmm5 vmovaps %xmm3, 0x2a0(%rsp) vmovaps %xmm4, 0x2b0(%rsp) vmovaps %xmm5, 0x2c0(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d343e5(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x2d0(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d2e406(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d343c5(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vcmpnltps %xmm2, %xmm5, %xmm4 vbroadcastss 0x1d343b5(%rip), %xmm5 # 0x1f5a96c vbroadcastss 0x1d343b0(%rip), %xmm6 # 0x1f5a970 vblendvps %xmm4, %xmm5, %xmm6, %xmm4 vmovaps %xmm3, 0x2e0(%rsp) vmovaps %xmm4, 0x2f0(%rsp) vmovaps 0x30(%r14), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cc5431(%rip), %xmm11 # 0x1eeba20 vblendvps %xmm10, %xmm3, %xmm11, %xmm1 vmovaps %xmm1, 0x300(%rsp) vbroadcastss 0x1cc657d(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm10, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x310(%rsp) vmovaps %xmm10, 0x340(%rsp) vxorps %xmm0, %xmm10, %xmm0 vmovaps %xmm0, 0xa0(%rsp) testq %rax, %rax je 0x226645 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r9d, %r9d cmpb $0x1, %al adcq $0x2, %r9 jmp 0x226649 pushq $0x3 popq %r9 leaq 0x80(%r14), %r11 leaq 0x3e0(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xb80(%rsp), %r12 vmovaps %xmm11, -0x20(%r12) movq 0x70(%r8), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%r12) leaq 0x1f298fb(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x330(%rsp) vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x320(%rsp) vbroadcastss 0x1cf9864(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf985f(%rip), %xmm13 # 0x1f1ff14 addq $-0x10, %r12 movq -0x8(%r15), %rbp addq $-0x8, %r15 cmpq $-0x8, %rbp je 0x22795d vmovaps (%r12), %xmm14 vcmpltps 0x310(%rsp), %xmm14, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22796e movzbl %al, %r13d popcntl %r13d, %ebx xorl %eax, %eax cmpq %r9, %rbx jbe 0x227972 cmpq %r9, %rbx jbe 0x227960 testb $0x8, %bpl pushq $0x8 popq %rbx jne 0x22689e movq %rbp, %rax movq %rbp, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %rbx, %rbp vmovaps %xmm11, %xmm14 movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x226835 vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x240(%rsp), %xmm1 vmovaps 0x250(%rsp), %xmm2 vmovaps 0x260(%rsp), %xmm3 vmovaps 0x2a0(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0x2b0(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x2c0(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vminps %xmm1, %xmm5, %xmm0 vminps %xmm2, %xmm7, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vminps %xmm3, %xmm9, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vmulps %xmm0, %xmm12, %xmm0 vmaxps %xmm1, %xmm5, %xmm1 vmaxps %xmm2, %xmm7, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmaxps %xmm3, %xmm9, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmulps %xmm1, %xmm13, %xmm1 vmaxps 0x300(%rsp), %xmm0, %xmm2 vminps 0x310(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vtestps %xmm1, %xmm1 je 0x226835 vblendvps %xmm1, %xmm0, %xmm11, %xmm0 cmpq $0x8, %rbp je 0x22682e movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm14, (%r12) addq $0x10, %r12 vmovaps %xmm0, %xmm14 movq %rdi, %rbp cmpq $0x8, %rdi je 0x22684c leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x226725 xorl %eax, %eax cmpq $0x8, %rbp je 0x226897 vmovaps 0x310(%rsp), %xmm0 vcmpnleps %xmm14, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r9 jae 0x226886 testb %cl, %cl je 0x227960 testb $0x8, %bpl je 0x226711 jmp 0x22689e movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm14, (%r12) addq $0x10, %r12 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x226872 cmpq $-0x8, %rbp je 0x22795d vmovaps 0x310(%rsp), %xmm0 vcmpnleps %xmm14, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22796e movl %ebp, %eax andl $0xf, %eax vmovaps 0xa0(%rsp), %xmm0 addq $-0x8, %rax movq %rax, 0x1a0(%rsp) je 0x22791b andq $-0x10, %rbp vxorps 0x1cc5532(%rip), %xmm0, %xmm2 # 0x1eebe20 addq $0x50, %rbp xorl %edi, %edi movq (%r10), %rax movq %rax, 0x1a8(%rsp) xorl %ebx, %ebx vmovaps %xmm2, %xmm10 movq %rdi, 0x28(%rsp) vmovaps %xmm2, 0x90(%rsp) movl (%rbp,%rbx,4), %eax cmpl $-0x1, %eax je 0x2278ed movl -0x10(%rbp,%rbx,4), %ecx movq 0x1a8(%rsp), %rdx movq 0x228(%rdx), %rdx movq (%rdx,%rcx,8), %rdx movl -0x50(%rbp,%rbx,4), %edi vbroadcastss (%rdx,%rdi,4), %xmm4 vbroadcastss 0x4(%rdx,%rdi,4), %xmm1 vbroadcastss 0x8(%rdx,%rdi,4), %xmm2 movl -0x20(%rbp,%rbx,4), %edi vbroadcastss (%rdx,%rdi,4), %xmm13 vbroadcastss 0x4(%rdx,%rdi,4), %xmm5 vbroadcastss 0x8(%rdx,%rdi,4), %xmm7 vmovaps 0x10(%r14), %xmm3 vmovaps 0x20(%r14), %xmm0 vsubps %xmm3, %xmm1, %xmm6 vsubps %xmm0, %xmm2, %xmm15 vmovaps %xmm5, 0xf0(%rsp) vsubps %xmm3, %xmm5, %xmm8 vmovaps %xmm8, 0x70(%rsp) vmovaps %xmm7, 0xe0(%rsp) vsubps %xmm0, %xmm7, %xmm5 vmovaps %xmm5, 0x60(%rsp) vsubps %xmm6, %xmm8, %xmm1 vsubps %xmm15, %xmm5, %xmm2 vaddps %xmm6, %xmm8, %xmm7 vmovaps %xmm10, 0x50(%rsp) vaddps %xmm5, %xmm15, %xmm8 vmulps %xmm2, %xmm7, %xmm5 vmulps %xmm1, %xmm8, %xmm9 vsubps %xmm5, %xmm9, %xmm11 vmovaps (%r14), %xmm9 vsubps %xmm9, %xmm4, %xmm4 vmovaps %xmm13, 0x100(%rsp) vsubps %xmm9, %xmm13, %xmm5 vsubps %xmm4, %xmm5, %xmm12 vmulps %xmm8, %xmm12, %xmm8 vaddps %xmm4, %xmm5, %xmm10 vmovaps %xmm2, 0x150(%rsp) vmulps %xmm2, %xmm10, %xmm13 vsubps %xmm8, %xmm13, %xmm13 vmovaps %xmm1, 0x160(%rsp) vmulps %xmm1, %xmm10, %xmm8 vmovaps %xmm12, 0x140(%rsp) vmulps %xmm7, %xmm12, %xmm7 vsubps %xmm8, %xmm7, %xmm8 vmovaps 0x60(%r14), %xmm12 vmulps %xmm12, %xmm8, %xmm10 vmovaps 0x50(%r14), %xmm14 vmulps %xmm13, %xmm14, %xmm13 vaddps %xmm13, %xmm10, %xmm13 vmovaps 0x40(%r14), %xmm10 vmulps %xmm11, %xmm10, %xmm11 vaddps %xmm13, %xmm11, %xmm1 vmovaps %xmm1, 0x30(%rsp) movl -0x40(%rbp,%rbx,4), %edi vbroadcastss 0x4(%rdx,%rdi,4), %xmm1 vmovaps %xmm1, 0x3c0(%rsp) vsubps %xmm3, %xmm1, %xmm1 vbroadcastss 0x8(%rdx,%rdi,4), %xmm2 vmovaps %xmm2, 0x3b0(%rsp) vsubps %xmm0, %xmm2, %xmm2 vsubps %xmm1, %xmm6, %xmm11 vsubps %xmm2, %xmm15, %xmm13 vmovaps %xmm6, 0x190(%rsp) vaddps %xmm1, %xmm6, %xmm0 vmovaps %xmm15, 0x180(%rsp) vaddps %xmm2, %xmm15, %xmm3 vmulps %xmm0, %xmm13, %xmm6 vmulps %xmm3, %xmm11, %xmm7 vsubps %xmm6, %xmm7, %xmm6 vbroadcastss (%rdx,%rdi,4), %xmm7 vmovaps %xmm7, 0x3a0(%rsp) vsubps %xmm9, %xmm7, %xmm7 vsubps %xmm7, %xmm4, %xmm15 vmulps %xmm3, %xmm15, %xmm3 vmovaps %xmm4, 0x170(%rsp) vaddps %xmm7, %xmm4, %xmm9 vmulps %xmm13, %xmm9, %xmm8 vsubps %xmm3, %xmm8, %xmm3 vmovaps %xmm11, 0x120(%rsp) vmulps %xmm11, %xmm9, %xmm8 vmulps %xmm0, %xmm15, %xmm0 vsubps %xmm8, %xmm0, %xmm0 vmovaps 0x50(%rsp), %xmm4 vmulps %xmm0, %xmm12, %xmm0 vmulps %xmm3, %xmm14, %xmm3 vaddps %xmm3, %xmm0, %xmm0 vmulps %xmm6, %xmm10, %xmm3 vaddps %xmm0, %xmm3, %xmm9 vsubps %xmm5, %xmm7, %xmm3 vaddps %xmm5, %xmm7, %xmm5 vmovaps 0x70(%rsp), %xmm6 vsubps %xmm6, %xmm1, %xmm0 vaddps %xmm6, %xmm1, %xmm1 vmovaps 0x60(%rsp), %xmm6 vsubps %xmm6, %xmm2, %xmm8 vaddps %xmm6, %xmm2, %xmm2 vmulps %xmm1, %xmm8, %xmm6 vmulps %xmm2, %xmm0, %xmm7 vsubps %xmm6, %xmm7, %xmm6 vmulps %xmm2, %xmm3, %xmm2 vmulps %xmm5, %xmm8, %xmm7 vsubps %xmm2, %xmm7, %xmm2 vmulps %xmm0, %xmm5, %xmm5 vmulps %xmm1, %xmm3, %xmm1 vsubps %xmm5, %xmm1, %xmm1 vmovaps %xmm12, 0x60(%rsp) vmulps %xmm1, %xmm12, %xmm1 vbroadcastss 0x1cfa361(%rip), %xmm12 # 0x1f20ec4 vmovaps %xmm14, 0x70(%rsp) vmulps %xmm2, %xmm14, %xmm2 vmovaps %xmm4, %xmm14 vaddps %xmm2, %xmm1, %xmm1 vmovaps %xmm10, 0x130(%rsp) vmulps %xmm6, %xmm10, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vmovaps 0x30(%rsp), %xmm7 vaddps %xmm7, %xmm9, %xmm2 vaddps %xmm2, %xmm1, %xmm4 vminps %xmm9, %xmm7, %xmm2 vminps %xmm1, %xmm2, %xmm2 vandps %xmm4, %xmm12, %xmm10 vbroadcastss 0x1cfa322(%rip), %xmm5 # 0x1f20ecc vmulps %xmm5, %xmm10, %xmm5 vbroadcastss 0x1cfa309(%rip), %xmm6 # 0x1f20ec0 vxorps %xmm6, %xmm5, %xmm6 vcmpnltps %xmm6, %xmm2, %xmm2 vmaxps %xmm9, %xmm7, %xmm6 vmaxps %xmm1, %xmm6, %xmm1 vcmpleps %xmm5, %xmm1, %xmm1 vorps %xmm1, %xmm2, %xmm2 movl -0x30(%rbp,%rbx,4), %edi vtestps %xmm14, %xmm2 vbroadcastss (%rdx,%rdi,4), %xmm1 vmovaps %xmm1, 0x390(%rsp) vbroadcastss 0x4(%rdx,%rdi,4), %xmm1 vmovaps %xmm1, 0x380(%rsp) vbroadcastss 0x8(%rdx,%rdi,4), %xmm1 vmovaps %xmm1, 0x370(%rsp) je 0x2277ff vmovaps %xmm10, 0x350(%rsp) vmovaps %xmm4, 0x360(%rsp) vmovaps 0x150(%rsp), %xmm10 vmovaps 0x120(%rsp), %xmm11 vmulps %xmm10, %xmm11, %xmm1 vmovaps 0x160(%rsp), %xmm4 vmovaps %xmm2, 0x110(%rsp) vmulps %xmm4, %xmm13, %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmulps %xmm0, %xmm13, %xmm5 vmulps %xmm8, %xmm11, %xmm6 vsubps %xmm5, %xmm6, %xmm6 vandps %xmm1, %xmm12, %xmm1 vandps %xmm5, %xmm12, %xmm5 vcmpltps %xmm5, %xmm1, %xmm1 vblendvps %xmm1, %xmm2, %xmm6, %xmm7 vmulps %xmm8, %xmm15, %xmm1 vmulps %xmm10, %xmm15, %xmm2 vmovaps 0x140(%rsp), %xmm8 vmulps %xmm13, %xmm8, %xmm5 vsubps %xmm5, %xmm2, %xmm2 vmulps %xmm3, %xmm13, %xmm6 vsubps %xmm1, %xmm6, %xmm6 vandps %xmm5, %xmm12, %xmm5 vandps %xmm1, %xmm12, %xmm1 vcmpltps %xmm1, %xmm5, %xmm1 vblendvps %xmm1, %xmm2, %xmm6, %xmm6 vmulps %xmm3, %xmm11, %xmm1 vmulps %xmm11, %xmm8, %xmm2 vmulps %xmm4, %xmm15, %xmm3 vmulps %xmm0, %xmm15, %xmm0 vsubps %xmm3, %xmm2, %xmm2 vsubps %xmm1, %xmm0, %xmm0 vandps %xmm3, %xmm12, %xmm3 vandps %xmm1, %xmm12, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm0, %xmm0 vmulps 0x60(%rsp), %xmm0, %xmm1 vmulps 0x70(%rsp), %xmm6, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps 0x130(%rsp), %xmm7, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm1, %xmm1, %xmm2 vmulps 0x180(%rsp), %xmm0, %xmm1 vmulps 0x190(%rsp), %xmm6, %xmm3 vaddps %xmm3, %xmm1, %xmm1 vmovaps %xmm7, 0x70(%rsp) vmulps 0x170(%rsp), %xmm7, %xmm3 vaddps %xmm1, %xmm3, %xmm1 vaddps %xmm1, %xmm1, %xmm1 vrcpps %xmm2, %xmm3 vmulps %xmm3, %xmm2, %xmm4 vbroadcastss 0x1cc59df(%rip), %xmm5 # 0x1eec714 vsubps %xmm4, %xmm5, %xmm4 vmulps %xmm4, %xmm3, %xmm4 vaddps %xmm4, %xmm3, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vmovaps 0x30(%r14), %xmm3 vmovaps 0x80(%r14), %xmm4 vcmpleps %xmm1, %xmm3, %xmm3 vmovaps %xmm4, 0x60(%rsp) vcmpleps %xmm4, %xmm1, %xmm4 vandps %xmm4, %xmm3, %xmm3 vxorps %xmm15, %xmm15, %xmm15 vcmpneqps %xmm2, %xmm15, %xmm2 vandps %xmm3, %xmm2, %xmm3 vandps 0x110(%rsp), %xmm14, %xmm2 vpslld $0x1f, %xmm3, %xmm3 vpsrad $0x1f, %xmm3, %xmm3 vtestps %xmm2, %xmm3 movq 0x28(%rsp), %rdi vmovaps %xmm14, %xmm10 je 0x227829 vmovaps %xmm9, %xmm8 vandps %xmm2, %xmm3, %xmm2 movq (%r10), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rcx,8), %r13 vbroadcastss 0x34(%r13), %xmm3 vandps 0x90(%r14), %xmm3, %xmm3 vpcmpeqd %xmm3, %xmm15, %xmm3 vtestps %xmm2, %xmm3 vbroadcastss 0x1cc4c4a(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cf9131(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf912c(%rip), %xmm13 # 0x1f1ff14 vpcmpeqd %xmm9, %xmm9, %xmm9 jb 0x226e12 vandnps %xmm2, %xmm3, %xmm2 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x2272d7 cmpq $0x0, 0x48(%r13) jne 0x2272d7 vpandn %xmm10, %xmm2, %xmm10 vtestps %xmm10, %xmm10 je 0x2272ae vmovaps (%r14), %xmm1 vmovaps 0x10(%r14), %xmm3 vmovaps 0x20(%r14), %xmm5 vmovaps 0x390(%rsp), %xmm0 vsubps %xmm1, %xmm0, %xmm9 vmovaps 0x380(%rsp), %xmm0 vsubps %xmm3, %xmm0, %xmm4 vmovaps 0x370(%rsp), %xmm0 vsubps %xmm5, %xmm0, %xmm15 vmovaps 0x100(%rsp), %xmm0 vsubps %xmm1, %xmm0, %xmm14 vmovaps 0xf0(%rsp), %xmm2 vsubps %xmm3, %xmm2, %xmm13 vmovaps %xmm10, 0x50(%rsp) vmovaps 0xe0(%rsp), %xmm2 vsubps %xmm5, %xmm2, %xmm8 vmovaps 0x3a0(%rsp), %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps 0x3c0(%rsp), %xmm2 vsubps %xmm3, %xmm2, %xmm6 vmovaps 0x3b0(%rsp), %xmm2 vsubps %xmm5, %xmm2, %xmm5 vmovaps %xmm5, 0x30(%rsp) vsubps %xmm9, %xmm1, %xmm2 vsubps %xmm4, %xmm6, %xmm12 vsubps %xmm15, %xmm5, %xmm0 vaddps %xmm4, %xmm6, %xmm3 vaddps %xmm5, %xmm15, %xmm7 vmulps %xmm0, %xmm3, %xmm10 vmulps %xmm7, %xmm12, %xmm11 vsubps %xmm10, %xmm11, %xmm5 vaddps %xmm1, %xmm9, %xmm10 vmulps %xmm7, %xmm2, %xmm7 vmovaps %xmm0, 0x150(%rsp) vmulps %xmm0, %xmm10, %xmm11 vsubps %xmm7, %xmm11, %xmm7 vmovaps %xmm12, 0x160(%rsp) vmulps %xmm12, %xmm10, %xmm10 vmovaps %xmm2, 0x170(%rsp) vmulps %xmm3, %xmm2, %xmm3 vsubps %xmm10, %xmm3, %xmm3 vmovaps 0x60(%r14), %xmm12 vmulps %xmm3, %xmm12, %xmm3 vmovaps 0x50(%r14), %xmm0 vmovaps %xmm0, 0x60(%rsp) vmulps %xmm7, %xmm0, %xmm7 vaddps %xmm7, %xmm3, %xmm3 vmovaps 0x40(%r14), %xmm0 vmulps %xmm5, %xmm0, %xmm7 vaddps %xmm3, %xmm7, %xmm11 vmovaps %xmm13, %xmm2 vsubps %xmm13, %xmm4, %xmm7 vmovaps %xmm8, %xmm5 vsubps %xmm8, %xmm15, %xmm13 vmovaps %xmm4, 0x190(%rsp) vaddps %xmm2, %xmm4, %xmm3 vmovaps %xmm2, %xmm8 vmovaps %xmm15, 0x180(%rsp) vaddps %xmm5, %xmm15, %xmm2 vmulps %xmm3, %xmm13, %xmm15 vmulps %xmm2, %xmm7, %xmm4 vsubps %xmm15, %xmm4, %xmm4 vsubps %xmm14, %xmm9, %xmm15 vmulps %xmm2, %xmm15, %xmm2 vmovaps %xmm9, 0x70(%rsp) vaddps %xmm14, %xmm9, %xmm9 vmulps %xmm13, %xmm9, %xmm10 vsubps %xmm2, %xmm10, %xmm2 vmovaps %xmm7, 0x130(%rsp) vmulps %xmm7, %xmm9, %xmm9 vmulps %xmm3, %xmm15, %xmm3 vsubps %xmm9, %xmm3, %xmm3 vmulps %xmm3, %xmm12, %xmm3 vmovaps 0x60(%rsp), %xmm10 vmulps %xmm2, %xmm10, %xmm2 vaddps %xmm2, %xmm3, %xmm2 vmulps %xmm4, %xmm0, %xmm3 vaddps %xmm2, %xmm3, %xmm9 vsubps %xmm1, %xmm14, %xmm3 vaddps %xmm1, %xmm14, %xmm1 vsubps %xmm6, %xmm8, %xmm14 vaddps %xmm6, %xmm8, %xmm2 vmovaps 0x30(%rsp), %xmm4 vsubps %xmm4, %xmm5, %xmm7 vaddps %xmm4, %xmm5, %xmm4 vmulps %xmm7, %xmm2, %xmm5 vmulps %xmm4, %xmm14, %xmm6 vsubps %xmm5, %xmm6, %xmm5 vmulps %xmm4, %xmm3, %xmm4 vmulps %xmm7, %xmm1, %xmm6 vsubps %xmm4, %xmm6, %xmm4 vmovaps 0x50(%rsp), %xmm6 vmulps %xmm1, %xmm14, %xmm1 vmulps %xmm2, %xmm3, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm12, 0x30(%rsp) vmulps %xmm1, %xmm12, %xmm1 vmovaps %xmm10, %xmm12 vmulps %xmm4, %xmm10, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmovaps %xmm0, 0x140(%rsp) vmulps %xmm5, %xmm0, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm9, %xmm11, %xmm2 vaddps %xmm2, %xmm1, %xmm4 vminps %xmm9, %xmm11, %xmm2 vminps %xmm1, %xmm2, %xmm2 vbroadcastss 0x1cf9e78(%rip), %xmm10 # 0x1f20ec4 vmovaps %xmm4, 0x100(%rsp) vandps %xmm4, %xmm10, %xmm5 vbroadcastss 0x1cf9e6a(%rip), %xmm4 # 0x1f20ecc vmovaps %xmm5, 0xf0(%rsp) vmulps %xmm4, %xmm5, %xmm4 vbroadcastss 0x1cf9e48(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm4, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm11, 0x110(%rsp) vmaxps %xmm9, %xmm11, %xmm5 vmaxps %xmm1, %xmm5, %xmm1 vcmpleps %xmm4, %xmm1, %xmm1 vorps %xmm1, %xmm2, %xmm0 vtestps %xmm6, %xmm0 je 0x227871 vmovaps %xmm9, 0xe0(%rsp) vmovaps 0x150(%rsp), %xmm11 vmovaps %xmm0, 0x120(%rsp) vmovaps 0x130(%rsp), %xmm0 vmulps %xmm0, %xmm11, %xmm1 vmovaps %xmm10, %xmm9 vmovaps 0x160(%rsp), %xmm10 vmulps %xmm13, %xmm10, %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmulps %xmm13, %xmm14, %xmm4 vmulps %xmm7, %xmm0, %xmm5 vsubps %xmm4, %xmm5, %xmm5 vandps %xmm1, %xmm9, %xmm1 vandps %xmm4, %xmm9, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vmovaps %xmm14, %xmm8 vblendvps %xmm1, %xmm2, %xmm5, %xmm14 vmulps %xmm7, %xmm15, %xmm1 vmulps %xmm11, %xmm15, %xmm2 vmovaps 0x170(%rsp), %xmm11 vmulps %xmm13, %xmm11, %xmm4 vsubps %xmm4, %xmm2, %xmm2 vmulps %xmm3, %xmm13, %xmm5 vsubps %xmm1, %xmm5, %xmm5 vandps %xmm4, %xmm9, %xmm4 vandps %xmm1, %xmm9, %xmm1 vcmpltps %xmm1, %xmm4, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm7 vmulps %xmm0, %xmm3, %xmm1 vmulps %xmm0, %xmm11, %xmm2 vmulps %xmm10, %xmm15, %xmm3 vmulps %xmm8, %xmm15, %xmm0 vsubps %xmm3, %xmm2, %xmm2 vsubps %xmm1, %xmm0, %xmm0 vandps %xmm3, %xmm9, %xmm3 vandps %xmm1, %xmm9, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm0, %xmm5 vmulps 0x30(%rsp), %xmm5, %xmm0 vmulps %xmm7, %xmm12, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps 0x140(%rsp), %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm1 vmulps 0x180(%rsp), %xmm5, %xmm0 vmulps 0x190(%rsp), %xmm7, %xmm2 vaddps %xmm2, %xmm0, %xmm0 vmulps 0x70(%rsp), %xmm14, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vrcpps %xmm1, %xmm2 vmulps %xmm2, %xmm1, %xmm3 vbroadcastss 0x1cc5552(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vmulps %xmm2, %xmm0, %xmm0 vmovaps 0x30(%r14), %xmm2 vmovaps 0x80(%r14), %xmm15 vcmpleps %xmm0, %xmm2, %xmm2 vcmpleps %xmm15, %xmm0, %xmm3 vandps %xmm3, %xmm2, %xmm2 vxorps %xmm9, %xmm9, %xmm9 vcmpneqps %xmm1, %xmm9, %xmm1 vandps %xmm2, %xmm1, %xmm2 vandps 0x120(%rsp), %xmm6, %xmm1 vpslld $0x1f, %xmm2, %xmm2 vpsrad $0x1f, %xmm2, %xmm2 vtestps %xmm1, %xmm2 vmovaps %xmm6, %xmm10 je 0x22789e vandps %xmm1, %xmm2, %xmm1 movq (%r10), %rcx movl -0x10(%rbp,%rbx,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %r13 vbroadcastss 0x34(%r13), %xmm2 vandps 0x90(%r14), %xmm2, %xmm2 vpcmpeqd %xmm2, %xmm9, %xmm2 vtestps %xmm1, %xmm2 vbroadcastss 0x1cc47c9(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cf8cb0(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf8cab(%rip), %xmm13 # 0x1f1ff14 jb 0x2278b9 vmovaps %xmm14, %xmm8 movl (%rbp,%rbx,4), %ecx vandnps %xmm1, %xmm2, %xmm1 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x227567 cmpq $0x0, 0x48(%r13) jne 0x227567 vmovaps 0x90(%rsp), %xmm2 vpandn %xmm10, %xmm1, %xmm10 vtestps %xmm10, %xmm10 setne %al jmp 0x2272b9 xorl %eax, %eax vmovaps 0x90(%rsp), %xmm2 testb %al, %al je 0x2278ed leaq 0x1(%rbx), %rax cmpq $0x3, %rbx movq %rax, %rbx jb 0x226913 jmp 0x2278ed vmovaps 0x360(%rsp), %xmm4 vrcpps %xmm4, %xmm3 vmulps %xmm3, %xmm4, %xmm4 vbroadcastss 0x1cc5423(%rip), %xmm7 # 0x1eec714 vsubps %xmm4, %xmm7, %xmm4 vmulps %xmm4, %xmm3, %xmm4 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1cc9ce2(%rip), %xmm4 # 0x1ef0fe8 vmovaps 0x350(%rsp), %xmm5 vcmpnltps %xmm4, %xmm5, %xmm4 vandps %xmm3, %xmm4, %xmm3 vmulps 0x30(%rsp), %xmm3, %xmm4 vminps %xmm7, %xmm4, %xmm4 vmulps %xmm3, %xmm8, %xmm3 vminps %xmm7, %xmm3, %xmm3 vsubps %xmm4, %xmm7, %xmm5 vsubps %xmm3, %xmm7, %xmm7 vmovaps 0x330(%rsp), %xmm8 vblendvps %xmm8, %xmm5, %xmm4, %xmm4 vblendvps %xmm8, %xmm7, %xmm3, %xmm3 movq 0x8(%r10), %rdx vmovd %ecx, %xmm5 vpshufd $0x0, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vmovd %eax, %xmm7 vpshufd $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmovaps 0x70(%rsp), %xmm8 vmovaps %xmm8, 0x1b0(%rsp) vmovaps %xmm6, 0x1c0(%rsp) vmovaps %xmm0, 0x1d0(%rsp) vmovaps %xmm4, 0x1e0(%rsp) vmovaps %xmm3, 0x1f0(%rsp) vmovdqa %xmm7, 0x200(%rsp) vmovdqa %xmm5, 0x210(%rsp) leaq 0x220(%rsp), %rax vcmptrueps %ymm0, %ymm0, %ymm0 vmovups %ymm0, (%rax) vbroadcastss (%rdx), %xmm0 vmovaps %xmm0, 0x220(%rsp) vbroadcastss 0x4(%rdx), %xmm0 vmovaps %xmm0, 0x230(%rsp) vmovaps 0x60(%rsp), %xmm0 vblendvps %xmm2, %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm2, 0x80(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%r13), %rax movq %rax, 0xb8(%rsp) movq 0x8(%r10), %rax movq %rax, 0xc0(%rsp) movq %r14, 0xc8(%rsp) leaq 0x1b0(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x4, 0xd8(%rsp) movq 0x48(%r13), %rax testq %rax, %rax movq %r10, 0x18(%rsp) movq %r8, 0x10(%rsp) movq %r9, 0x8(%rsp) movq %r11, (%rsp) je 0x2274a9 leaq 0xb0(%rsp), %rdi movq %rsi, 0x48(%rsp) vzeroupper callq %rax movq 0x28(%rsp), %rdi vxorps %xmm15, %xmm15, %xmm15 vpcmpeqd %xmm9, %xmm9, %xmm9 vbroadcastss 0x1cf8a95(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cf8a88(%rip), %xmm12 # 0x1f1ff10 movq (%rsp), %r11 movq 0x8(%rsp), %r9 vbroadcastss 0x1cc4586(%rip), %xmm11 # 0x1eeba20 movq 0x10(%rsp), %r8 movq 0x48(%rsp), %rsi movq 0x18(%rsp), %r10 vmovdqa 0x80(%rsp), %xmm0 vptest %xmm0, %xmm0 vmovaps 0x50(%rsp), %xmm10 je 0x227849 movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x227532 testb $0x2, (%rcx) jne 0x2274dc testb $0x40, 0x3e(%r13) je 0x227532 leaq 0xb0(%rsp), %rdi movq %rsi, %r13 vzeroupper callq %rax vmovaps 0x50(%rsp), %xmm10 movq 0x28(%rsp), %rdi vxorps %xmm15, %xmm15, %xmm15 vpcmpeqd %xmm9, %xmm9, %xmm9 vbroadcastss 0x1cf8a0a(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cf89fd(%rip), %xmm12 # 0x1f1ff10 movq (%rsp), %r11 movq 0x8(%rsp), %r9 vbroadcastss 0x1cc44fb(%rip), %xmm11 # 0x1eeba20 movq 0x10(%rsp), %r8 movq %r13, %rsi movq 0x18(%rsp), %r10 vpcmpeqd 0x80(%rsp), %xmm15, %xmm1 vpxor %xmm1, %xmm9, %xmm0 movq 0xc8(%rsp), %rax vbroadcastss 0x1cc5634(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x227851 vmovaps 0x100(%rsp), %xmm3 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cc5193(%rip), %xmm6 # 0x1eec714 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cc9a52(%rip), %xmm3 # 0x1ef0fe8 vmovaps 0xf0(%rsp), %xmm4 vcmpnltps %xmm3, %xmm4, %xmm3 vandps %xmm2, %xmm3, %xmm2 vmulps 0x110(%rsp), %xmm2, %xmm3 vminps %xmm6, %xmm3, %xmm4 vmulps 0xe0(%rsp), %xmm2, %xmm2 vminps %xmm6, %xmm2, %xmm3 vsubps %xmm4, %xmm6, %xmm14 vsubps %xmm3, %xmm6, %xmm6 vmovaps 0x320(%rsp), %xmm2 vblendvps %xmm2, %xmm14, %xmm4, %xmm14 vblendvps %xmm2, %xmm6, %xmm3, %xmm2 movq 0x8(%r10), %rdx vmovd %eax, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovd %ecx, %xmm6 vpshufd $0x0, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmovaps %xmm8, 0x1b0(%rsp) vmovaps %xmm7, 0x1c0(%rsp) vmovaps %xmm5, 0x1d0(%rsp) vmovaps %xmm14, 0x1e0(%rsp) vmovaps %xmm2, 0x1f0(%rsp) vmovdqa %xmm6, 0x200(%rsp) vmovdqa %xmm4, 0x210(%rsp) leaq 0x220(%rsp), %rax vcmptrueps %ymm2, %ymm2, %ymm2 vmovups %ymm2, (%rax) vbroadcastss (%rdx), %xmm2 vmovaps %xmm2, 0x220(%rsp) vbroadcastss 0x4(%rdx), %xmm2 vmovaps %xmm2, 0x230(%rsp) vblendvps %xmm1, %xmm0, %xmm15, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm1, 0x80(%rsp) leaq 0x80(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%r13), %rax movq %rax, 0xb8(%rsp) movq 0x8(%r10), %rax movq %rax, 0xc0(%rsp) movq %r14, 0xc8(%rsp) leaq 0x1b0(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x4, 0xd8(%rsp) movq 0x48(%r13), %rax testq %rax, %rax movq %r10, 0x18(%rsp) movq %r8, 0x10(%rsp) movq %r9, 0x8(%rsp) movq %r11, (%rsp) vmovaps %xmm15, 0x30(%rsp) je 0x227742 leaq 0xb0(%rsp), %rdi movq %rsi, 0x48(%rsp) vzeroupper callq %rax vmovaps 0x30(%rsp), %xmm15 vmovaps 0x50(%rsp), %xmm10 movq 0x28(%rsp), %rdi vxorps %xmm9, %xmm9, %xmm9 vbroadcastss 0x1cf87fc(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cf87ef(%rip), %xmm12 # 0x1f1ff10 movq (%rsp), %r11 movq 0x8(%rsp), %r9 vbroadcastss 0x1cc42ed(%rip), %xmm11 # 0x1eeba20 movq 0x10(%rsp), %r8 movq 0x48(%rsp), %rsi movq 0x18(%rsp), %r10 vmovdqa 0x80(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x2278c7 movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x2277c6 testb $0x2, (%rcx) jne 0x22776f testb $0x40, 0x3e(%r13) je 0x2277c6 leaq 0xb0(%rsp), %rdi movq %rsi, %r13 vzeroupper callq %rax vmovaps 0x30(%rsp), %xmm15 vmovaps 0x50(%rsp), %xmm10 movq 0x28(%rsp), %rdi vxorps %xmm9, %xmm9, %xmm9 vbroadcastss 0x1cf8776(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cf8769(%rip), %xmm12 # 0x1f1ff10 movq (%rsp), %r11 movq 0x8(%rsp), %r9 vbroadcastss 0x1cc4267(%rip), %xmm11 # 0x1eeba20 movq 0x10(%rsp), %r8 movq %r13, %rsi movq 0x18(%rsp), %r10 vpcmpeqd 0x80(%rsp), %xmm9, %xmm1 vpxor 0x1cc4649(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0xc8(%rsp), %rax vbroadcastss 0x1cc539c(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x2278d3 vbroadcastss 0x1cc4218(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cf86ff(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf86fa(%rip), %xmm13 # 0x1f1ff14 movq 0x28(%rsp), %rdi vmovaps %xmm14, %xmm10 jmp 0x226e12 vbroadcastss 0x1cc41ee(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cf86d5(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf86d0(%rip), %xmm13 # 0x1f1ff14 jmp 0x226e12 vpcmpeqd %xmm0, %xmm15, %xmm0 vpxor %xmm0, %xmm9, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm2 vmovaps 0x60(%rsp), %xmm1 vblendvps %xmm0, (%r11), %xmm1, %xmm0 vmovaps %xmm0, (%r11) jmp 0x226e0d vbroadcastss 0x1cc41a6(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cf868d(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf8688(%rip), %xmm13 # 0x1f1ff14 vmovaps 0x90(%rsp), %xmm2 vmovaps %xmm6, %xmm10 jmp 0x2272a4 vbroadcastss 0x1cc4179(%rip), %xmm11 # 0x1eeba20 vbroadcastss 0x1cf8660(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf865b(%rip), %xmm13 # 0x1f1ff14 vmovaps 0x90(%rsp), %xmm2 jmp 0x2272a4 vpcmpeqd %xmm0, %xmm9, %xmm0 vpxor 0x1cc454d(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm1 vblendvps %xmm0, (%r11), %xmm15, %xmm0 vmovaps %xmm0, (%r11) jmp 0x227296 vandps %xmm2, %xmm10, %xmm0 vtestps %xmm2, %xmm10 je 0x227911 incq %rdi addq $0x60, %rbp vmovaps %xmm0, %xmm2 cmpq 0x1a0(%rsp), %rdi jb 0x2268f4 vpcmpeqd %xmm1, %xmm1, %xmm1 vxorps %xmm1, %xmm0, %xmm0 jmp 0x22791f vpcmpeqd %xmm1, %xmm1, %xmm1 vorps 0xa0(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0xa0(%rsp) vtestps %xmm1, %xmm0 jb 0x22795d vmovaps 0x310(%rsp), %xmm1 vbroadcastss 0x1cc523a(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x310(%rsp) xorl %eax, %eax jmp 0x227960 pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x2266b5 jmp 0x227a56 pushq $0x2 jmp 0x22795f vmovaps %xmm14, 0x30(%rsp) movq %r11, (%rsp) movq %r9, 0x8(%rsp) movq %r10, 0x18(%rsp) movq %rsi, 0x48(%rsp) movq %r8, 0x10(%rsp) bsfq %r13, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, 0x50(%rsp) leaq 0x27(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x248(%rsp), %rax pushq %rax vzeroupper callq 0x26a086 popq %rcx popq %rdx testb %al, %al je 0x2279d0 movq 0x50(%rsp), %rax orl $-0x1, 0xa0(%rsp,%rax,4) leaq -0x1(%r13), %rax andq %rax, %r13 movq 0x18(%rsp), %r10 movq 0x48(%rsp), %rsi movq 0x10(%rsp), %r8 jne 0x227990 vmovaps 0xa0(%rsp), %xmm0 vtestps 0x1cc4426(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax vbroadcastss 0x1cc401a(%rip), %xmm11 # 0x1eeba20 movq 0x8(%rsp), %r9 movq (%rsp), %r11 vbroadcastss 0x1cf84f8(%rip), %xmm12 # 0x1f1ff10 vbroadcastss 0x1cf84f3(%rip), %xmm13 # 0x1f1ff14 vmovaps 0x30(%rsp), %xmm14 jb 0x2266fb vmovaps 0x310(%rsp), %xmm1 vbroadcastss 0x1cc5145(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x310(%rsp) pushq $0x2 popq %rax jmp 0x2266fb vmovaps 0x340(%rsp), %xmm0 vandps 0xa0(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cc5113(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r11) addq $0x1a78, %rsp # imm = 0x1A78 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %r14, %rdx movq %r10, %rcx addq $0x1a78, %rsp # imm = 0x1A78 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x2686c4 nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiMBIntersectorKMoeller<4, 4, true>>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %rbx cmpq $0x8, 0x70(%rbx) je 0x227c95 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x227c95 movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 movzbl %al, %ebp vmovaps (%rdx), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%rdx), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%rdx), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%rdx), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%rdx), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%rdx), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cf9395(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1cc94ac(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vandps %xmm4, %xmm2, %xmm5 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vrcpps %xmm1, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1cc4b9e(%rip), %xmm5 # 0x1eec714 vsubps %xmm1, %xmm5, %xmm1 vmulps %xmm1, %xmm4, %xmm1 vaddps %xmm1, %xmm4, %xmm1 vrcpps %xmm2, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm2, %xmm4, %xmm2 vrcpps %xmm3, %xmm4 vmulps %xmm4, %xmm3, %xmm3 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d32d96(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d2cdb8(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d32d77(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d32d5f(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d32d5a(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%rdx), %xmm1 vmovaps 0x80(%rdx), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cc3dde(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cc4f2b(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %r13 movq 0x70(%rbx), %rdx movq %r12, %rdi movq %rbx, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x26af5e popq %rax popq %rcx andq %r13, %rbp jne 0x227c67 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiMBIntersectorKMoeller<4, 4, true>>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x19f8, %rsp # imm = 0x19F8 movq %rsi, %rax movq (%rsi), %rsi cmpq $0x8, 0x70(%rsi) je 0x229455 movq %rdx, %r14 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%rdx), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x229455 movq %rcx, %r10 movq %rax, 0x128(%rsp) vandps %xmm3, %xmm4, %xmm8 vmovaps (%r14), %xmm3 vmovaps %xmm3, 0x220(%rsp) vmovaps 0x10(%r14), %xmm3 vmovaps %xmm3, 0x230(%rsp) vmovaps 0x20(%r14), %xmm3 vmovaps %xmm3, 0x240(%rsp) vmovaps 0x40(%r14), %xmm3 vmovaps %xmm3, 0x250(%rsp) vmovaps 0x50(%r14), %xmm4 vmovaps %xmm4, 0x260(%rsp) vmovaps 0x60(%r14), %xmm5 vmovaps %xmm5, 0x270(%rsp) vbroadcastss 0x1cf9162(%rip), %xmm9 # 0x1f20ec4 vandps %xmm3, %xmm9, %xmm6 vbroadcastss 0x1cc9279(%rip), %xmm7 # 0x1ef0fe8 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm3, %xmm3 vandps %xmm4, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm4, %xmm4 vandps %xmm5, %xmm9, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vblendvps %xmm6, %xmm7, %xmm5, %xmm5 vrcpps %xmm3, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0x1cc496b(%rip), %xmm7 # 0x1eec714 vsubps %xmm3, %xmm7, %xmm3 vmulps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm6, %xmm3 vrcpps %xmm4, %xmm6 vmulps %xmm6, %xmm4, %xmm4 vsubps %xmm4, %xmm7, %xmm4 vmulps %xmm4, %xmm6, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm5 vsubps %xmm5, %xmm7, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm6, %xmm5 vmovaps %xmm3, 0x280(%rsp) vmovaps %xmm4, 0x290(%rsp) vmovaps %xmm5, 0x2a0(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d32b5e(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x2b0(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d2cb7f(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d32b3e(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vmovaps %xmm3, 0x2c0(%rsp) vcmpnltps %xmm2, %xmm5, %xmm3 vbroadcastss 0x1d32b25(%rip), %xmm4 # 0x1f5a96c vbroadcastss 0x1d32b20(%rip), %xmm5 # 0x1f5a970 vblendvps %xmm3, %xmm4, %xmm5, %xmm3 vmovaps %xmm3, 0x2d0(%rsp) vmovaps 0x30(%r14), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cc3baa(%rip), %xmm1 # 0x1eeba20 vblendvps %xmm8, %xmm3, %xmm1, %xmm1 vmovaps %xmm1, 0x2e0(%rsp) vbroadcastss 0x1cc4cf6(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm8, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x2f0(%rsp) vmovaps %xmm8, 0x320(%rsp) vxorps %xmm0, %xmm8, %xmm0 vmovaps %xmm0, 0x110(%rsp) cmpq $0x0, 0x8(%rcx) je 0x227ece movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r8d, %r8d cmpb $0x1, %al adcq $0x2, %r8 jmp 0x227ed2 pushq $0x3 popq %r8 leaq 0x80(%r14), %rax movq %rax, 0x108(%rsp) leaq 0x360(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xb00(%rsp), %r12 vbroadcastss 0x1cc3b1e(%rip), %xmm0 # 0x1eeba20 vmovaps %xmm0, -0x20(%r12) movq 0x70(%rsi), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%r12) leaq 0x1f28061(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x310(%rsp) vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x300(%rsp) addq $-0x10, %r12 movq -0x8(%r15), %r13 addq $-0x8, %r15 cmpq $-0x8, %r13 je 0x229344 vmovaps (%r12), %xmm1 vcmpltps 0x2f0(%rsp), %xmm1, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x229355 movzbl %al, %ebp popcntl %ebp, %ebx xorl %eax, %eax cmpq %r8, %rbx jbe 0x229359 cmpq %r8, %rbx pushq $0x8 popq %r9 jbe 0x229347 testb $0x8, %r13b jne 0x2281fb vmovaps 0x2f0(%rsp), %xmm0 movq %r13, %rax andq $-0x10, %rax andl $0x7, %r13d movl %r13d, %ecx vcmpnleps %xmm1, %xmm0, %xmm0 xorl %edx, %edx movq %r9, %r13 vbroadcastss 0x1cc3a5e(%rip), %xmm1 # 0x1eeba20 movq (%rax,%rdx,8), %rdi cmpq $0x8, %rdi je 0x228165 vmovaps %xmm1, %xmm15 vbroadcastss 0x80(%rax,%rdx,4), %xmm2 vbroadcastss 0x20(%rax,%rdx,4), %xmm3 vmovaps 0x70(%r14), %xmm1 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0xa0(%rax,%rdx,4), %xmm3 vbroadcastss 0x40(%rax,%rdx,4), %xmm4 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0xc0(%rax,%rdx,4), %xmm4 vbroadcastss 0x60(%rax,%rdx,4), %xmm5 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0x90(%rax,%rdx,4), %xmm5 vbroadcastss 0x30(%rax,%rdx,4), %xmm6 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0xb0(%rax,%rdx,4), %xmm6 vbroadcastss 0x50(%rax,%rdx,4), %xmm7 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0xd0(%rax,%rdx,4), %xmm7 vbroadcastss 0x70(%rax,%rdx,4), %xmm8 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vmovaps 0x220(%rsp), %xmm8 vmovaps 0x230(%rsp), %xmm9 vmovaps 0x240(%rsp), %xmm10 vmovaps 0x280(%rsp), %xmm11 vsubps %xmm8, %xmm2, %xmm2 vmulps %xmm2, %xmm11, %xmm12 vsubps %xmm9, %xmm3, %xmm2 vmovaps 0x290(%rsp), %xmm3 vmulps %xmm3, %xmm2, %xmm13 vsubps %xmm10, %xmm4, %xmm2 vmovaps 0x2a0(%rsp), %xmm4 vmulps %xmm4, %xmm2, %xmm14 vsubps %xmm8, %xmm5, %xmm2 vmulps %xmm2, %xmm11, %xmm5 vsubps %xmm9, %xmm6, %xmm2 vmulps %xmm3, %xmm2, %xmm3 vsubps %xmm10, %xmm7, %xmm2 vmulps %xmm4, %xmm2, %xmm4 vpminsd %xmm5, %xmm12, %xmm2 vpminsd %xmm3, %xmm13, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpminsd %xmm4, %xmm14, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpmaxsd %xmm5, %xmm12, %xmm5 vpmaxsd %xmm3, %xmm13, %xmm3 vpminsd %xmm3, %xmm5, %xmm5 vpmaxsd %xmm4, %xmm14, %xmm4 vpmaxsd 0x2e0(%rsp), %xmm2, %xmm3 vpminsd %xmm4, %xmm5, %xmm4 vpminsd 0x2f0(%rsp), %xmm4, %xmm4 cmpl $0x6, %ecx je 0x22817e vcmpleps %xmm4, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm1 vpslld $0x1f, %xmm1, %xmm1 vtestps %xmm1, %xmm1 je 0x2281ab vbroadcastss 0x1cc38db(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 cmpq $0x8, %r13 je 0x228162 movq %r13, (%r15) addq $0x8, %r15 vmovaps %xmm15, (%r12) addq $0x10, %r12 movq %rdi, %r13 cmpq $0x8, %rdi je 0x2281b1 leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x227fc2 jmp 0x2281b1 vcmpleps %xmm4, %xmm3, %xmm3 vbroadcastss 0xe0(%rax,%rdx,4), %xmm4 vcmpleps %xmm1, %xmm4, %xmm4 vbroadcastss 0xf0(%rax,%rdx,4), %xmm5 vcmpltps %xmm5, %xmm1, %xmm1 vandps %xmm1, %xmm4, %xmm1 vandps %xmm3, %xmm1, %xmm1 jmp 0x22812c vmovaps %xmm15, %xmm1 jmp 0x228165 xorl %eax, %eax cmpq $0x8, %r13 je 0x2281f4 vmovaps 0x2f0(%rsp), %xmm0 vcmpnleps %xmm1, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r8 jae 0x2281e3 testb %cl, %cl jne 0x227f8e jmp 0x229347 movq %r13, (%r15) addq $0x8, %r15 vmovaps %xmm1, (%r12) addq $0x10, %r12 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x2281d6 cmpq $-0x8, %r13 je 0x229344 vmovaps 0x2f0(%rsp), %xmm0 vcmpnleps %xmm1, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x229355 movl %r13d, %eax andl $0xf, %eax vmovaps 0x110(%rsp), %xmm0 addq $-0x8, %rax movq %rax, 0x1f0(%rsp) je 0x229302 movq %r8, 0x100(%rsp) andq $-0x10, %r13 vxorps 0x1cc3bcd(%rip), %xmm0, %xmm4 # 0x1eebe20 xorl %eax, %eax movq %rax, 0x1f8(%rsp) imulq $0x60, %rax, %r11 addq %r13, %r11 vmovmskps %xmm4, %eax movq %rax, 0x28(%rsp) xorl %ebp, %ebp vmovaps %xmm4, %xmm1 vmovaps %xmm4, 0xe0(%rsp) movl 0x50(%r11,%rbp,4), %eax cmpl $-0x1, %eax je 0x2292c8 vmovaps %xmm1, 0x30(%rsp) movq (%r10), %rdx movl 0x40(%r11,%rbp,4), %ecx movq 0x1e8(%rdx), %rdx movq (%rdx,%rcx,8), %rbx vbroadcastss 0x2c(%rbx), %xmm0 vmovss 0x30(%rbx), %xmm1 vbroadcastss 0x28(%rbx), %xmm2 vmovaps 0x70(%r14), %xmm3 vsubps %xmm0, %xmm3, %xmm3 vsubss %xmm0, %xmm1, %xmm0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vdivps %xmm0, %xmm3, %xmm0 vmulps %xmm0, %xmm2, %xmm0 vroundps $0x1, %xmm0, %xmm1 vaddss 0x1cc86ef(%rip), %xmm2, %xmm2 # 0x1ef09cc vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vminps %xmm2, %xmm1, %xmm1 vmaxps 0x1cc3722(%rip), %xmm1, %xmm1 # 0x1eeba10 vsubps %xmm1, %xmm0, %xmm0 vcvtps2dq %xmm1, %xmm1 vmovapd %xmm1, 0x60(%rsp) movq 0x28(%rsp), %rdx bsfq %rdx, %rdx movslq 0x60(%rsp,%rdx,4), %r8 vmovd %r8d, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vpcmpeqd %xmm1, %xmm2, %xmm1 vtestps %xmm4, %xmm1 jae 0x228793 movq 0xe0(%rbx), %rdi imulq $0x38, %r8, %r8 movq (%rdi,%r8), %rdx movq 0x38(%rdi,%r8), %r9 movl (%r11,%rbp,4), %edi movl 0x10(%r11,%rbp,4), %r8d vbroadcastss (%rdx,%rdi,4), %xmm2 vbroadcastss 0x4(%rdx,%rdi,4), %xmm3 vbroadcastss 0x8(%rdx,%rdi,4), %xmm4 vbroadcastss (%r9,%rdi,4), %xmm5 vbroadcastss 0x4(%r9,%rdi,4), %xmm6 vbroadcastss 0x8(%r9,%rdi,4), %xmm7 vbroadcastss 0x1cc43a3(%rip), %xmm1 # 0x1eec714 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm5, %xmm0, %xmm5 vmulps %xmm6, %xmm0, %xmm6 vmulps %xmm7, %xmm0, %xmm7 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm5, %xmm2, %xmm2 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm6, %xmm3, %xmm3 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm7, %xmm4, %xmm14 vbroadcastss (%rdx,%r8,4), %xmm5 vbroadcastss 0x4(%rdx,%r8,4), %xmm6 vbroadcastss 0x8(%rdx,%r8,4), %xmm7 vbroadcastss (%r9,%r8,4), %xmm8 vbroadcastss 0x4(%r9,%r8,4), %xmm9 vbroadcastss 0x8(%r9,%r8,4), %xmm10 vmulps %xmm0, %xmm8, %xmm8 vmulps %xmm0, %xmm9, %xmm9 vmulps %xmm0, %xmm10, %xmm10 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm5, %xmm8, %xmm11 vmulps %xmm6, %xmm1, %xmm5 vaddps %xmm5, %xmm9, %xmm12 vmulps %xmm7, %xmm1, %xmm5 vaddps %xmm5, %xmm10, %xmm13 movl 0x20(%r11,%rbp,4), %edi vbroadcastss (%rdx,%rdi,4), %xmm5 vbroadcastss 0x4(%rdx,%rdi,4), %xmm6 vbroadcastss 0x8(%rdx,%rdi,4), %xmm7 vbroadcastss (%r9,%rdi,4), %xmm8 vbroadcastss 0x4(%r9,%rdi,4), %xmm9 vbroadcastss 0x8(%r9,%rdi,4), %xmm10 vmulps %xmm0, %xmm8, %xmm8 vmulps %xmm0, %xmm9, %xmm9 vmulps %xmm0, %xmm10, %xmm10 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm5, %xmm8, %xmm5 vmovaps %xmm5, 0x1e0(%rsp) vmulps %xmm6, %xmm1, %xmm5 vaddps %xmm5, %xmm9, %xmm5 vmovaps %xmm5, 0x1d0(%rsp) vmulps %xmm7, %xmm1, %xmm5 vaddps %xmm5, %xmm10, %xmm5 vmovaps %xmm5, 0x210(%rsp) movl 0x30(%r11,%rbp,4), %edi vbroadcastss (%r9,%rdi,4), %xmm5 vbroadcastss 0x4(%r9,%rdi,4), %xmm6 vbroadcastss 0x8(%r9,%rdi,4), %xmm7 vmulps %xmm5, %xmm0, %xmm5 vmulps %xmm6, %xmm0, %xmm6 vmulps %xmm7, %xmm0, %xmm0 vbroadcastss (%rdx,%rdi,4), %xmm7 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm5, %xmm7, %xmm8 vbroadcastss 0x4(%rdx,%rdi,4), %xmm5 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm6 vbroadcastss 0x8(%rdx,%rdi,4), %xmm5 vmulps %xmm5, %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmovaps %xmm11, 0xb0(%rsp) vsubps %xmm11, %xmm2, %xmm7 vmovaps %xmm12, 0xc0(%rsp) vsubps %xmm12, %xmm3, %xmm9 vmovaps %xmm13, 0x50(%rsp) vsubps %xmm13, %xmm14, %xmm11 vmovaps %xmm8, 0xa0(%rsp) vsubps %xmm2, %xmm8, %xmm12 vmovaps %xmm6, 0xd0(%rsp) vsubps %xmm3, %xmm6, %xmm13 vmovaps %xmm0, 0x340(%rsp) vsubps %xmm14, %xmm0, %xmm0 vmulps %xmm0, %xmm9, %xmm1 vmulps %xmm11, %xmm13, %xmm5 vsubps %xmm1, %xmm5, %xmm1 vmovaps %xmm1, 0xf0(%rsp) vmulps %xmm11, %xmm12, %xmm1 vmulps %xmm0, %xmm7, %xmm5 vsubps %xmm1, %xmm5, %xmm4 vmovaps %xmm7, 0x10(%rsp) vmulps %xmm7, %xmm13, %xmm5 vmovaps %xmm9, 0x1c0(%rsp) vmulps %xmm9, %xmm12, %xmm6 vsubps %xmm5, %xmm6, %xmm9 vsubps (%r14), %xmm2, %xmm6 vsubps 0x10(%r14), %xmm3, %xmm7 vsubps 0x20(%r14), %xmm14, %xmm8 vmovaps 0x50(%r14), %xmm2 vmovaps 0x60(%r14), %xmm10 vmulps %xmm2, %xmm8, %xmm3 vmulps %xmm7, %xmm10, %xmm14 vsubps %xmm3, %xmm14, %xmm3 vmovaps 0x40(%r14), %xmm1 vmulps %xmm6, %xmm10, %xmm14 vmulps %xmm1, %xmm8, %xmm15 vsubps %xmm14, %xmm15, %xmm14 vmulps %xmm1, %xmm7, %xmm15 vmulps %xmm2, %xmm6, %xmm5 vsubps %xmm15, %xmm5, %xmm15 vmovaps %xmm9, 0x200(%rsp) vmulps %xmm10, %xmm9, %xmm5 vmovaps %xmm4, 0x330(%rsp) vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm5, %xmm2, %xmm2 vmovaps 0xf0(%rsp), %xmm5 vmulps %xmm1, %xmm5, %xmm1 vaddps %xmm2, %xmm1, %xmm10 vmulps %xmm0, %xmm15, %xmm0 vmulps %xmm14, %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cf8901(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm10, %xmm13 vmulps %xmm3, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm13, %xmm2 vcmpnltps 0x1cc3438(%rip), %xmm2, %xmm12 # 0x1eeba10 vmovaps 0x30(%rsp), %xmm1 vtestps %xmm1, %xmm12 jne 0x228bba vmovaps 0xb0(%rsp), %xmm0 vmovaps 0xa0(%rsp), %xmm3 vmovaps 0xd0(%rsp), %xmm6 vtestps %xmm1, %xmm1 vmovaps 0xc0(%rsp), %xmm2 vmovaps 0x50(%rsp), %xmm5 je 0x22876a vmovaps 0x1e0(%rsp), %xmm7 vsubps %xmm3, %xmm7, %xmm12 vmovaps 0x1d0(%rsp), %xmm8 vsubps %xmm6, %xmm8, %xmm4 vmovaps 0x210(%rsp), %xmm9 vsubps 0x340(%rsp), %xmm9, %xmm10 vsubps %xmm7, %xmm0, %xmm3 vsubps %xmm8, %xmm2, %xmm11 vsubps %xmm9, %xmm5, %xmm15 vmulps %xmm4, %xmm15, %xmm0 vmovaps %xmm1, 0x30(%rsp) vmulps %xmm11, %xmm10, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0xc0(%rsp) vmulps %xmm3, %xmm10, %xmm1 vmulps %xmm15, %xmm12, %xmm2 vsubps %xmm1, %xmm2, %xmm6 vmovaps %xmm12, 0x50(%rsp) vmulps %xmm11, %xmm12, %xmm2 vmovaps %xmm4, 0xf0(%rsp) vmulps %xmm3, %xmm4, %xmm5 vsubps %xmm2, %xmm5, %xmm5 vsubps (%r14), %xmm7, %xmm7 vsubps 0x10(%r14), %xmm8, %xmm8 vsubps 0x20(%r14), %xmm9, %xmm4 vmovaps 0x50(%r14), %xmm9 vmovaps 0x60(%r14), %xmm0 vmulps %xmm4, %xmm9, %xmm12 vmulps %xmm0, %xmm8, %xmm13 vsubps %xmm12, %xmm13, %xmm12 vmovaps 0x40(%r14), %xmm1 vmulps %xmm0, %xmm7, %xmm13 vmulps %xmm1, %xmm4, %xmm14 vsubps %xmm13, %xmm14, %xmm13 vmulps %xmm1, %xmm8, %xmm14 vmulps %xmm7, %xmm9, %xmm2 vsubps %xmm14, %xmm2, %xmm14 vmovaps %xmm5, 0xa0(%rsp) vmulps %xmm0, %xmm5, %xmm0 vmovaps %xmm6, 0xb0(%rsp) vmulps %xmm6, %xmm9, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vmovaps 0xc0(%rsp), %xmm2 vmulps %xmm1, %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm9 vmulps %xmm14, %xmm15, %xmm0 vmulps %xmm13, %xmm11, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cf8793(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm9, %xmm11 vmulps %xmm3, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmovaps 0x30(%rsp), %xmm1 vxorps %xmm0, %xmm11, %xmm3 vcmpnltps 0x1cc32c4(%rip), %xmm3, %xmm15 # 0x1eeba10 vtestps %xmm1, %xmm15 jne 0x228c06 vmovaps 0xe0(%rsp), %xmm4 vtestps %xmm1, %xmm1 setne %al jmp 0x228775 xorl %eax, %eax vmovaps 0xe0(%rsp), %xmm4 testb %al, %al je 0x2292c8 leaq 0x1(%rbp), %rax cmpq $0x3, %rbp movq %rax, %rbp jb 0x22827c jmp 0x2292c8 movq %r11, 0x10(%rsp) movq %rsi, (%rsp) movq %r10, 0x8(%rsp) cmpq $0x0, 0x28(%rsp) je 0x22884a movq 0xe0(%rbx), %rsi movq %rsi, 0x50(%rsp) addq $0x38, %rsi movq %rsi, 0xc0(%rsp) movq 0x10(%rsp), %rdi movl (%rdi,%rbp,4), %r8d movq 0x28(%rsp), %r9 movq %rdx, %r10 movslq 0x60(%rsp,%r10,4), %r11 imulq $0x38, %r11, %r11 movq 0x50(%rsp), %rsi movq (%rsi,%r11), %rdi movq 0xc0(%rsp), %rsi movq (%rsi,%r11), %r11 vmovups (%rdi,%r8,4), %xmm1 vmovdqu (%r11,%r8,4), %xmm2 vmovss %xmm1, 0x130(%rsp,%r10,4) vextractps $0x1, %xmm1, 0x140(%rsp,%r10,4) vextractps $0x2, %xmm1, 0x150(%rsp,%r10,4) vmovd %xmm2, 0x70(%rsp,%r10,4) vpextrd $0x1, %xmm2, 0x80(%rsp,%r10,4) vpextrd $0x2, %xmm2, 0x90(%rsp,%r10,4) btcq %r10, %r9 bsfq %r9, %r10 testq %r9, %r9 jne 0x2287d6 vbroadcastss 0x1cc3ec1(%rip), %xmm1 # 0x1eec714 vsubps %xmm0, %xmm1, %xmm1 vmulps 0x130(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0xf0(%rsp) vmulps 0x140(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0xb0(%rsp) vmulps 0x150(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0xa0(%rsp) vmulps 0x70(%rsp), %xmm0, %xmm2 vmovaps %xmm2, 0xd0(%rsp) vmulps 0x80(%rsp), %xmm0, %xmm2 vmovaps %xmm2, 0x1e0(%rsp) vmulps 0x90(%rsp), %xmm0, %xmm14 cmpq $0x0, 0x28(%rsp) je 0x228961 movq 0xe0(%rbx), %rsi movq %rsi, 0x50(%rsp) addq $0x38, %rsi movq %rsi, 0xc0(%rsp) movq 0x10(%rsp), %rdi movl 0x10(%rdi,%rbp,4), %r8d movq 0x28(%rsp), %r9 movq %rdx, %r10 movslq 0x60(%rsp,%r10,4), %r11 imulq $0x38, %r11, %r11 movq 0x50(%rsp), %rsi movq (%rsi,%r11), %rdi movq 0xc0(%rsp), %rsi movq (%rsi,%r11), %r11 vmovups (%rdi,%r8,4), %xmm5 vmovups (%r11,%r8,4), %xmm6 vmovss %xmm5, 0x130(%rsp,%r10,4) vextractps $0x1, %xmm5, 0x140(%rsp,%r10,4) vextractps $0x2, %xmm5, 0x150(%rsp,%r10,4) vmovss %xmm6, 0x70(%rsp,%r10,4) vextractps $0x1, %xmm6, 0x80(%rsp,%r10,4) vextractps $0x2, %xmm6, 0x90(%rsp,%r10,4) btcq %r10, %r9 bsfq %r9, %r10 testq %r9, %r9 jne 0x2288ed vmulps 0x130(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x1d0(%rsp) vmulps 0x140(%rsp), %xmm1, %xmm9 vmulps 0x150(%rsp), %xmm1, %xmm10 vmulps 0x70(%rsp), %xmm0, %xmm11 vmulps 0x80(%rsp), %xmm0, %xmm12 vmulps 0x90(%rsp), %xmm0, %xmm13 cmpq $0x0, 0x28(%rsp) je 0x228a47 movq 0xe0(%rbx), %rsi movq %rsi, 0x50(%rsp) addq $0x38, %rsi movq %rsi, 0xc0(%rsp) movq 0x10(%rsp), %rdi movl 0x20(%rdi,%rbp,4), %r8d movq 0x28(%rsp), %r9 movq %rdx, %r10 movslq 0x60(%rsp,%r10,4), %r11 imulq $0x38, %r11, %r11 movq 0x50(%rsp), %rsi movq (%rsi,%r11), %rdi movq 0xc0(%rsp), %rsi movq (%rsi,%r11), %r11 vmovups (%rdi,%r8,4), %xmm5 vmovups (%r11,%r8,4), %xmm6 vmovss %xmm5, 0x130(%rsp,%r10,4) vextractps $0x1, %xmm5, 0x140(%rsp,%r10,4) vextractps $0x2, %xmm5, 0x150(%rsp,%r10,4) vmovss %xmm6, 0x70(%rsp,%r10,4) vextractps $0x1, %xmm6, 0x80(%rsp,%r10,4) vextractps $0x2, %xmm6, 0x90(%rsp,%r10,4) btcq %r10, %r9 bsfq %r9, %r10 testq %r9, %r9 jne 0x2289d3 vmulps 0x130(%rsp), %xmm1, %xmm5 vmulps 0x140(%rsp), %xmm1, %xmm6 vmulps 0x150(%rsp), %xmm1, %xmm4 vmulps 0x70(%rsp), %xmm0, %xmm7 vmulps 0x80(%rsp), %xmm0, %xmm8 vmulps 0x90(%rsp), %xmm0, %xmm15 cmpq $0x0, 0x28(%rsp) je 0x228b01 movq 0xe0(%rbx), %rsi leaq 0x38(%rsi), %rdi movq 0x10(%rsp), %r8 movl 0x30(%r8,%rbp,4), %r8d movq 0x28(%rsp), %r9 movslq 0x60(%rsp,%rdx,4), %r10 imulq $0x38, %r10, %r10 movq (%rsi,%r10), %r11 movq (%rdi,%r10), %r10 vmovups (%r11,%r8,4), %xmm2 vmovups (%r10,%r8,4), %xmm3 vmovss %xmm2, 0x130(%rsp,%rdx,4) vextractps $0x1, %xmm2, 0x140(%rsp,%rdx,4) vextractps $0x2, %xmm2, 0x150(%rsp,%rdx,4) vmovss %xmm3, 0x70(%rsp,%rdx,4) vextractps $0x1, %xmm3, 0x80(%rsp,%rdx,4) vextractps $0x2, %xmm3, 0x90(%rsp,%rdx,4) btcq %rdx, %r9 bsfq %r9, %rdx testq %r9, %r9 jne 0x228a9c vmovaps 0xd0(%rsp), %xmm2 vaddps 0xf0(%rsp), %xmm2, %xmm2 vmovaps 0x1e0(%rsp), %xmm3 vaddps 0xb0(%rsp), %xmm3, %xmm3 vaddps 0xa0(%rsp), %xmm14, %xmm14 vaddps 0x1d0(%rsp), %xmm11, %xmm11 vaddps %xmm12, %xmm9, %xmm12 vaddps %xmm13, %xmm10, %xmm13 vaddps %xmm7, %xmm5, %xmm5 vmovaps %xmm5, 0x1e0(%rsp) vaddps %xmm6, %xmm8, %xmm5 vmovaps %xmm5, 0x1d0(%rsp) vmulps 0x130(%rsp), %xmm1, %xmm5 vmulps 0x140(%rsp), %xmm1, %xmm6 vmulps 0x150(%rsp), %xmm1, %xmm1 vmulps 0x70(%rsp), %xmm0, %xmm7 vaddps %xmm4, %xmm15, %xmm8 vmovaps %xmm8, 0x210(%rsp) vaddps %xmm7, %xmm5, %xmm8 vmulps 0x80(%rsp), %xmm0, %xmm5 vaddps %xmm5, %xmm6, %xmm6 vmulps 0x90(%rsp), %xmm0, %xmm0 vaddps %xmm0, %xmm1, %xmm0 movq 0x8(%rsp), %r10 movq (%rsp), %rsi movq 0x10(%rsp), %r11 jmp 0x2284a2 vmovaps %xmm8, %xmm4 vmovaps 0x330(%rsp), %xmm8 vandps %xmm1, %xmm12, %xmm12 vmulps %xmm15, %xmm11, %xmm0 vmulps 0x1c0(%rsp), %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x10(%rsp), %xmm3, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm13, %xmm3 vcmpnltps 0x1cc2e1c(%rip), %xmm3, %xmm9 # 0x1eeba10 vtestps %xmm12, %xmm9 jne 0x228c52 vmovdqa 0x30(%rsp), %xmm1 jmp 0x2285e9 vmovaps %xmm4, %xmm5 vandps %xmm1, %xmm15, %xmm15 vmulps %xmm14, %xmm10, %xmm0 vmulps 0xf0(%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x50(%rsp), %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm4 vcmpnltps 0x1cc2dd9(%rip), %xmm4, %xmm10 # 0x1eeba10 vtestps %xmm15, %xmm10 jne 0x228c9f vmovaps 0xe0(%rsp), %xmm4 vmovaps 0x30(%rsp), %xmm1 jmp 0x228760 vmovaps %xmm5, %xmm14 movq %r11, 0x10(%rsp) movq %rsi, (%rsp) movq %r10, 0x8(%rsp) vbroadcastss 0x1cf8257(%rip), %xmm0 # 0x1f20ec4 vandps %xmm0, %xmm10, %xmm5 vandps %xmm12, %xmm9, %xmm9 vsubps %xmm2, %xmm5, %xmm0 vcmpnltps %xmm3, %xmm0, %xmm11 vtestps %xmm9, %xmm11 jne 0x228cf1 movq 0x8(%rsp), %r10 movq (%rsp), %rsi vmovdqa 0x30(%rsp), %xmm1 movq 0x10(%rsp), %r11 jmp 0x2285e9 movq %r11, 0x10(%rsp) movq %rsi, (%rsp) movq %r10, %rbx vbroadcastss 0x1cf8210(%rip), %xmm0 # 0x1f20ec4 vandps %xmm0, %xmm9, %xmm6 vandps %xmm15, %xmm10, %xmm10 vsubps %xmm3, %xmm6, %xmm0 vcmpnltps %xmm4, %xmm0, %xmm12 vtestps %xmm10, %xmm12 jne 0x228dc1 movq %rbx, %r10 movq (%rsp), %rsi vmovaps 0xe0(%rsp), %xmm4 vmovdqa 0x30(%rsp), %xmm1 movq 0x10(%rsp), %r11 jmp 0x228760 vandps %xmm9, %xmm11, %xmm9 vmulps 0x200(%rsp), %xmm4, %xmm0 vmulps %xmm7, %xmm8, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps %xmm6, %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm13, %xmm4 vmulps 0x30(%r14), %xmm5, %xmm0 vcmpltps %xmm4, %xmm0, %xmm0 vmovaps 0x80(%r14), %xmm1 vmovaps %xmm1, 0x1c0(%rsp) vmulps %xmm1, %xmm5, %xmm1 vcmpleps %xmm1, %xmm4, %xmm1 vandps %xmm0, %xmm1, %xmm6 vtestps %xmm9, %xmm6 je 0x228c86 vandps %xmm6, %xmm9, %xmm6 vcmpneqps 0x1cc2cbb(%rip), %xmm10, %xmm0 # 0x1eeba10 vtestps %xmm6, %xmm0 je 0x228c86 vandps %xmm6, %xmm0, %xmm6 vbroadcastss 0x34(%rbx), %xmm0 vandps 0x90(%r14), %xmm0, %xmm0 vpcmpeqd 0x1cc2c95(%rip), %xmm0, %xmm0 # 0x1eeba10 vtestps %xmm6, %xmm0 jb 0x228c86 vandnps %xmm6, %xmm0, %xmm6 movq 0x8(%rsp), %rdx movq 0x10(%rdx), %rdx cmpq $0x0, 0x10(%rdx) jne 0x228ec6 cmpq $0x0, 0x48(%rbx) jne 0x228ec6 vmovdqa 0x30(%rsp), %xmm1 vpandn %xmm1, %xmm6, %xmm1 movq 0x8(%rsp), %r10 movq (%rsp), %rsi jmp 0x228c95 vmovaps %xmm2, %xmm13 vandps %xmm10, %xmm12, %xmm10 vmulps 0xa0(%rsp), %xmm5, %xmm0 vmulps 0xb0(%rsp), %xmm8, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps %xmm7, %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm11, %xmm5 vmulps 0x30(%r14), %xmm6, %xmm0 vcmpltps %xmm5, %xmm0, %xmm0 vmovaps 0x80(%r14), %xmm1 vmovaps %xmm1, 0xf0(%rsp) vmulps %xmm1, %xmm6, %xmm1 vcmpleps %xmm1, %xmm5, %xmm1 vandps %xmm0, %xmm1, %xmm7 vtestps %xmm10, %xmm7 je 0x228cd1 vandps %xmm7, %xmm10, %xmm7 vcmpneqps 0x1cc2be2(%rip), %xmm9, %xmm8 # 0x1eeba10 vtestps %xmm7, %xmm8 je 0x228cd1 vandps %xmm7, %xmm8, %xmm7 movq (%rbx), %rcx movq 0x10(%rsp), %rax movl 0x40(%rax,%rbp,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rcx movq %rcx, 0x50(%rsp) vbroadcastss 0x34(%rcx), %xmm0 vandps 0x90(%r14), %xmm0, %xmm0 vpcmpeqd 0x1cc2ba0(%rip), %xmm0, %xmm8 # 0x1eeba10 vtestps %xmm7, %xmm8 jb 0x228cd1 movq 0x10(%rsp), %rcx movl 0x50(%rcx,%rbp,4), %ecx vandnps %xmm7, %xmm8, %xmm7 movq 0x10(%rbx), %rdx cmpq $0x0, 0x10(%rdx) jne 0x22908a movq 0x50(%rsp), %rdx cmpq $0x0, 0x48(%rdx) jne 0x22908a vmovdqa 0x30(%rsp), %xmm1 vpandn %xmm1, %xmm7, %xmm1 movq %rbx, %r10 movq (%rsp), %rsi vmovaps 0xe0(%rsp), %xmm4 jmp 0x228ce7 vrcpps %xmm5, %xmm0 vmulps %xmm0, %xmm5, %xmm1 vbroadcastss 0x1cc383d(%rip), %xmm5 # 0x1eec714 vsubps %xmm1, %xmm5, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm2, %xmm0, %xmm1 vminps %xmm5, %xmm1, %xmm1 vmulps %xmm3, %xmm0, %xmm2 vminps %xmm5, %xmm2, %xmm2 vsubps %xmm1, %xmm5, %xmm3 vsubps %xmm2, %xmm5, %xmm5 vmovaps 0x310(%rsp), %xmm7 vblendvps %xmm7, %xmm3, %xmm1, %xmm1 vblendvps %xmm7, %xmm5, %xmm2, %xmm2 movq 0x8(%rsp), %rsi movq 0x8(%rsi), %rdx vmovd %ecx, %xmm3 vpshufd $0x0, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmovd %eax, %xmm5 vpshufd $0x0, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vmovaps %xmm14, 0x130(%rsp) vmovaps %xmm8, 0x140(%rsp) vmovaps 0x200(%rsp), %xmm7 vmovaps %xmm7, 0x150(%rsp) vmovaps %xmm1, 0x160(%rsp) vmovaps %xmm2, 0x170(%rsp) vmovdqa %xmm5, 0x180(%rsp) vmovdqa %xmm3, 0x190(%rsp) vmulps %xmm4, %xmm0, %xmm0 leaq 0x1a0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x1a0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x1b0(%rsp) vmovaps 0x1c0(%rsp), %xmm1 vblendvps %xmm6, %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm6, 0x60(%rsp) leaq 0x60(%rsp), %rax movq %rax, 0x70(%rsp) movq 0x18(%rbx), %rax movq %rax, 0x78(%rsp) movq 0x8(%rsi), %rax movq %rax, 0x80(%rsp) movq %r14, 0x88(%rsp) leaq 0x130(%rsp), %rax movq %rax, 0x90(%rsp) movl $0x4, 0x98(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax je 0x229018 leaq 0x70(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x60(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x229254 movq 0x8(%rsp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x229050 testb $0x2, (%rcx) jne 0x229046 testb $0x40, 0x3e(%rbx) je 0x229050 leaq 0x70(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x60(%rsp), %xmm0, %xmm1 vpxor 0x1cc2dbe(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x88(%rsp), %rax vbroadcastss 0x1cc3b11(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x229264 vrcpps %xmm6, %xmm0 vmulps %xmm0, %xmm6, %xmm1 vbroadcastss 0x1cc3679(%rip), %xmm6 # 0x1eec714 vsubps %xmm1, %xmm6, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm3, %xmm0, %xmm1 vminps %xmm6, %xmm1, %xmm1 vmulps %xmm4, %xmm0, %xmm2 vminps %xmm6, %xmm2, %xmm2 vsubps %xmm1, %xmm6, %xmm3 vsubps %xmm2, %xmm6, %xmm4 vmovaps 0x300(%rsp), %xmm6 vblendvps %xmm6, %xmm3, %xmm1, %xmm1 vblendvps %xmm6, %xmm4, %xmm2, %xmm2 movq 0x8(%rbx), %rdx vmovd %eax, %xmm3 vpshufd $0x0, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmovd %ecx, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovaps %xmm13, 0x130(%rsp) vmovaps 0xb0(%rsp), %xmm6 vmovaps %xmm6, 0x140(%rsp) vmovaps 0xa0(%rsp), %xmm6 vmovaps %xmm6, 0x150(%rsp) vmovaps %xmm1, 0x160(%rsp) vmovaps %xmm2, 0x170(%rsp) vmovdqa %xmm4, 0x180(%rsp) vmovdqa %xmm3, 0x190(%rsp) vmulps %xmm5, %xmm0, %xmm0 leaq 0x1a0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x1a0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x1b0(%rsp) vmovaps 0xf0(%rsp), %xmm1 vblendvps %xmm7, %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm7, 0x60(%rsp) leaq 0x60(%rsp), %rax movq %rax, 0x70(%rsp) movq 0x50(%rsp), %rcx movq 0x18(%rcx), %rax movq %rax, 0x78(%rsp) movq 0x8(%rbx), %rax movq %rax, 0x80(%rsp) movq %r14, 0x88(%rsp) leaq 0x130(%rsp), %rax movq %rax, 0x90(%rsp) movl $0x4, 0x98(%rsp) movq 0x48(%rcx), %rax testq %rax, %rax je 0x2291e5 leaq 0x70(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x60(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x22928e movq 0x10(%rbx), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x22921d testb $0x2, (%rcx) jne 0x229213 movq 0x50(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x22921d leaq 0x70(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x60(%rsp), %xmm0, %xmm1 vpxor 0x1cc2bf1(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x88(%rsp), %rax vbroadcastss 0x1cc3944(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x22929e vpcmpeqd 0x1cc27b4(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cc2bbc(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm6 movq 0x108(%rsp), %rax vmovaps 0x1c0(%rsp), %xmm1 vblendvps %xmm0, (%rax), %xmm1, %xmm0 vmovaps %xmm0, (%rax) jmp 0x228da9 vpcmpeqd 0x1cc277a(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cc2b82(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm7 movq 0x108(%rsp), %rax vmovaps 0xf0(%rsp), %xmm1 vblendvps %xmm0, (%rax), %xmm1, %xmm0 vmovaps %xmm0, (%rax) jmp 0x228ea7 vandps %xmm4, %xmm1, %xmm0 vtestps %xmm4, %xmm1 movq 0x1f8(%rsp), %rax je 0x2292f0 incq %rax vmovaps %xmm0, %xmm4 cmpq 0x1f0(%rsp), %rax jb 0x228255 vpcmpeqd %xmm1, %xmm1, %xmm1 vxorps %xmm1, %xmm0, %xmm0 movq 0x100(%rsp), %r8 jmp 0x229306 vpcmpeqd %xmm1, %xmm1, %xmm1 vorps 0x110(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x110(%rsp) vtestps %xmm1, %xmm0 jb 0x229344 vmovaps 0x2f0(%rsp), %xmm1 vbroadcastss 0x1cc3853(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x2f0(%rsp) xorl %eax, %eax jmp 0x229347 pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x227f3d jmp 0x22942d pushq $0x2 jmp 0x229346 vmovaps %xmm1, 0xc0(%rsp) movq %r8, 0x100(%rsp) movq 0x128(%rsp), %rdi movq %r10, 0x8(%rsp) movq %rsi, (%rsp) bsfq %rbp, %rcx movq %r13, %rdx movq %rcx, 0x30(%rsp) leaq 0x4f(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x228(%rsp), %rax pushq %rax vzeroupper callq 0x26c1b2 popq %rcx popq %rdx testb %al, %al je 0x2293b5 movq 0x30(%rsp), %rax orl $-0x1, 0x110(%rsp,%rax,4) leaq -0x1(%rbp), %rax andq %rax, %rbp movq 0x8(%rsp), %r10 movq 0x128(%rsp), %rdi movq (%rsp), %rsi jne 0x22937b vmovaps 0x110(%rsp), %xmm0 vtestps 0x1cc2a3f(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0x100(%rsp), %r8 vmovaps 0xc0(%rsp), %xmm1 jb 0x227f81 vmovaps 0x2f0(%rsp), %xmm1 vbroadcastss 0x1cc3777(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps 0xc0(%rsp), %xmm1 vmovaps %xmm0, 0x2f0(%rsp) pushq $0x2 popq %rax jmp 0x227f81 vmovaps 0x320(%rsp), %xmm0 vandps 0x110(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cc373c(%rip), %xmm1 # 0x1eecb84 movq 0x108(%rsp), %rax vmaskmovps %xmm1, %xmm0, (%rax) addq $0x19f8, %rsp # imm = 0x19F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, true, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiMBIntersectorKPluecker<4, 4, true>>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %rbx cmpq $0x8, 0x70(%rbx) je 0x229630 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x229630 movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 movzbl %al, %ebp vmovaps (%rdx), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%rdx), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%rdx), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%rdx), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%rdx), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%rdx), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cf79d3(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1cc7aea(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1cc3208(%rip), %xmm7 # 0x1eec714 vdivps %xmm1, %xmm7, %xmm1 vandps %xmm4, %xmm2, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vdivps %xmm2, %xmm7, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vbroadcastss 0x1cf7a31(%rip), %xmm6 # 0x1f20f60 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vblendvps %xmm8, %xmm6, %xmm2, %xmm2 vdivps %xmm3, %xmm7, %xmm3 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d313fb(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d2b41d(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d313dc(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d313c4(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d313bf(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%rdx), %xmm1 vmovaps 0x80(%rdx), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cc2443(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cc3590(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %r13 movq 0x70(%rbx), %rdx movq %r12, %rdi movq %rbx, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x26d170 popq %rax popq %rcx andq %r13, %rbp jne 0x229602 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, true, embree::avx::ArrayIntersectorK_1<4, embree::avx::QuadMiMBIntersectorKPluecker<4, 4, true>>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1a88, %rsp # imm = 0x1A88 movq %rsi, %rax movq (%rsi), %rsi cmpq $0x8, 0x70(%rsi) je 0x22b1b5 movq %rdx, %r14 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%rdx), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x22b1b5 movq %rcx, %r10 movq %rax, 0x138(%rsp) vandps %xmm3, %xmm4, %xmm10 vmovaps (%r14), %xmm3 vmovaps %xmm3, 0x260(%rsp) vmovaps 0x10(%r14), %xmm3 vmovaps %xmm3, 0x270(%rsp) vmovaps 0x20(%r14), %xmm3 vmovaps %xmm3, 0x280(%rsp) vmovaps 0x40(%r14), %xmm3 vmovaps %xmm3, 0x290(%rsp) vmovaps 0x50(%r14), %xmm4 vmovaps %xmm4, 0x2a0(%rsp) vmovaps 0x60(%r14), %xmm5 vmovaps %xmm5, 0x2b0(%rsp) vbroadcastss 0x1cf77c8(%rip), %xmm8 # 0x1f20ec4 vandps %xmm3, %xmm8, %xmm6 vbroadcastss 0x1cc78df(%rip), %xmm9 # 0x1ef0fe8 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cc2ffc(%rip), %xmm11 # 0x1eec714 vdivps %xmm3, %xmm11, %xmm3 vandps %xmm4, %xmm8, %xmm7 vcmpltps %xmm9, %xmm7, %xmm7 vdivps %xmm4, %xmm11, %xmm4 vandps %xmm5, %xmm8, %xmm8 vcmpltps %xmm9, %xmm8, %xmm8 vdivps %xmm5, %xmm11, %xmm5 vbroadcastss 0x1cf781f(%rip), %xmm9 # 0x1f20f60 vblendvps %xmm6, %xmm9, %xmm3, %xmm3 vblendvps %xmm7, %xmm9, %xmm4, %xmm4 vblendvps %xmm8, %xmm9, %xmm5, %xmm5 vmovaps %xmm3, 0x2c0(%rsp) vmovaps %xmm4, 0x2d0(%rsp) vmovaps %xmm5, 0x2e0(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d311e8(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x2f0(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d2b209(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d311c8(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vmovaps %xmm3, 0x300(%rsp) vcmpnltps %xmm2, %xmm5, %xmm3 vbroadcastss 0x1d311af(%rip), %xmm4 # 0x1f5a96c vbroadcastss 0x1d311aa(%rip), %xmm5 # 0x1f5a970 vblendvps %xmm3, %xmm4, %xmm5, %xmm3 vmovaps %xmm3, 0x310(%rsp) vmovaps 0x30(%r14), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cc2234(%rip), %xmm1 # 0x1eeba20 vblendvps %xmm10, %xmm3, %xmm1, %xmm1 vmovaps %xmm1, 0x320(%rsp) vbroadcastss 0x1cc3380(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm10, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x330(%rsp) vmovaps %xmm10, 0x360(%rsp) vxorps %xmm0, %xmm10, %xmm0 vmovaps %xmm0, 0xf0(%rsp) cmpq $0x0, 0x8(%rcx) je 0x229844 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r8d, %r8d cmpb $0x1, %al adcq $0x2, %r8 jmp 0x229848 pushq $0x3 popq %r8 leaq 0x80(%r14), %rax movq %rax, 0xd8(%rsp) leaq 0x3f0(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xb90(%rsp), %r12 vbroadcastss 0x1cc21a8(%rip), %xmm0 # 0x1eeba20 vmovaps %xmm0, -0x20(%r12) movq 0x70(%rsi), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%r12) leaq 0x1f266eb(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x350(%rsp) vmovaps 0xf0(%rax), %xmm0 vmovaps %xmm0, 0x340(%rsp) addq $-0x10, %r12 movq -0x8(%r15), %rbp addq $-0x8, %r15 cmpq $-0x8, %rbp je 0x22b0ab vmovaps (%r12), %xmm1 vcmpltps 0x330(%rsp), %xmm1, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22b0bc movzbl %al, %r13d popcntl %r13d, %ebx xorl %eax, %eax cmpq %r8, %rbx jbe 0x22b0c0 cmpq %r8, %rbx pushq $0x8 popq %r9 jbe 0x22b0ae testb $0x8, %bpl jne 0x229b8b vmovaps 0x330(%rsp), %xmm0 movq %rbp, %rax andq $-0x10, %rax andl $0x7, %ebp movl %ebp, %ecx vcmpnleps %xmm1, %xmm0, %xmm0 xorl %edx, %edx movq %r9, %rbp vbroadcastss 0x1cc20e8(%rip), %xmm1 # 0x1eeba20 movq (%rax,%rdx,8), %rdi cmpq $0x8, %rdi je 0x229af5 vmovaps %xmm1, %xmm15 vbroadcastss 0x80(%rax,%rdx,4), %xmm2 vbroadcastss 0x20(%rax,%rdx,4), %xmm3 vmovaps 0x70(%r14), %xmm1 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0xa0(%rax,%rdx,4), %xmm3 vbroadcastss 0x40(%rax,%rdx,4), %xmm4 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0xc0(%rax,%rdx,4), %xmm4 vbroadcastss 0x60(%rax,%rdx,4), %xmm5 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0x90(%rax,%rdx,4), %xmm5 vbroadcastss 0x30(%rax,%rdx,4), %xmm6 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0xb0(%rax,%rdx,4), %xmm6 vbroadcastss 0x50(%rax,%rdx,4), %xmm7 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0xd0(%rax,%rdx,4), %xmm7 vbroadcastss 0x70(%rax,%rdx,4), %xmm8 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vmovaps 0x260(%rsp), %xmm8 vmovaps 0x270(%rsp), %xmm9 vmovaps 0x280(%rsp), %xmm10 vmovaps 0x2c0(%rsp), %xmm11 vsubps %xmm8, %xmm2, %xmm2 vmulps %xmm2, %xmm11, %xmm12 vsubps %xmm9, %xmm3, %xmm2 vmovaps 0x2d0(%rsp), %xmm3 vmulps %xmm3, %xmm2, %xmm13 vsubps %xmm10, %xmm4, %xmm2 vmovaps 0x2e0(%rsp), %xmm4 vmulps %xmm4, %xmm2, %xmm14 vsubps %xmm8, %xmm5, %xmm2 vmulps %xmm2, %xmm11, %xmm5 vsubps %xmm9, %xmm6, %xmm2 vmulps %xmm3, %xmm2, %xmm3 vsubps %xmm10, %xmm7, %xmm2 vmulps %xmm4, %xmm2, %xmm4 vpminsd %xmm5, %xmm12, %xmm2 vpminsd %xmm3, %xmm13, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpminsd %xmm4, %xmm14, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vbroadcastss 0x1cf649c(%rip), %xmm6 # 0x1f1ff10 vmulps %xmm6, %xmm2, %xmm2 vpmaxsd %xmm5, %xmm12, %xmm5 vpmaxsd %xmm3, %xmm13, %xmm3 vpminsd %xmm3, %xmm5, %xmm3 vpmaxsd %xmm4, %xmm14, %xmm4 vpminsd %xmm4, %xmm3, %xmm4 vpmaxsd 0x320(%rsp), %xmm2, %xmm3 vbroadcastss 0x1cf6470(%rip), %xmm5 # 0x1f1ff14 vmulps %xmm5, %xmm4, %xmm4 vpminsd 0x330(%rsp), %xmm4, %xmm4 cmpl $0x6, %ecx je 0x229b0e vcmpleps %xmm4, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm1 vpslld $0x1f, %xmm1, %xmm1 vtestps %xmm1, %xmm1 je 0x229b3b vbroadcastss 0x1cc1f4b(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 cmpq $0x8, %rbp je 0x229af2 movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm15, (%r12) addq $0x10, %r12 movq %rdi, %rbp cmpq $0x8, %rdi je 0x229b41 leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x229938 jmp 0x229b41 vcmpleps %xmm4, %xmm3, %xmm3 vbroadcastss 0xe0(%rax,%rdx,4), %xmm4 vcmpleps %xmm1, %xmm4, %xmm4 vbroadcastss 0xf0(%rax,%rdx,4), %xmm5 vcmpltps %xmm5, %xmm1, %xmm1 vandps %xmm1, %xmm4, %xmm1 vandps %xmm3, %xmm1, %xmm1 jmp 0x229abc vmovaps %xmm15, %xmm1 jmp 0x229af5 xorl %eax, %eax cmpq $0x8, %rbp je 0x229b84 vmovaps 0x330(%rsp), %xmm0 vcmpnleps %xmm1, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r8 jae 0x229b73 testb %cl, %cl jne 0x229906 jmp 0x22b0ae movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm1, (%r12) addq $0x10, %r12 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x229b66 cmpq $-0x8, %rbp je 0x22b0ab vmovaps 0x330(%rsp), %xmm0 vcmpnleps %xmm1, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22b0bc movl %ebp, %eax andl $0xf, %eax vmovaps 0xf0(%rsp), %xmm0 addq $-0x8, %rax movq %rax, 0x240(%rsp) je 0x22b069 movq %r8, 0xd0(%rsp) andq $-0x10, %rbp vxorps 0x1cc223e(%rip), %xmm0, %xmm4 # 0x1eebe20 xorl %eax, %eax movq %r10, 0x40(%rsp) movq %rsi, 0x38(%rsp) movq %rax, 0x248(%rsp) imulq $0x60, %rax, %r11 addq %rbp, %r11 vmovmskps %xmm4, %eax movq %rax, 0x30(%rsp) xorl %r13d, %r13d vmovaps %xmm4, 0x60(%rsp) vmovaps %xmm4, 0xe0(%rsp) movl 0x50(%r11,%r13,4), %eax cmpl $-0x1, %eax je 0x22b029 movq (%r10), %rdx movl 0x40(%r11,%r13,4), %ecx movq 0x1e8(%rdx), %rdx movq (%rdx,%rcx,8), %rbx vbroadcastss 0x2c(%rbx), %xmm0 vmovss 0x30(%rbx), %xmm1 vbroadcastss 0x28(%rbx), %xmm2 vmovaps 0x70(%r14), %xmm3 vsubps %xmm0, %xmm3, %xmm3 vsubss %xmm0, %xmm1, %xmm0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vdivps %xmm0, %xmm3, %xmm0 vmulps %xmm0, %xmm2, %xmm0 vroundps $0x1, %xmm0, %xmm1 vaddss 0x1cc6d59(%rip), %xmm2, %xmm2 # 0x1ef09cc vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vminps %xmm2, %xmm1, %xmm1 vxorps %xmm2, %xmm2, %xmm2 vmaxps %xmm2, %xmm1, %xmm1 vsubps %xmm1, %xmm0, %xmm0 vcvtps2dq %xmm1, %xmm1 vmovapd %xmm1, 0x70(%rsp) movq 0x30(%rsp), %rdx bsfq %rdx, %rdx movslq 0x70(%rsp,%rdx,4), %r8 vmovd %r8d, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vpcmpeqd %xmm1, %xmm2, %xmm1 vtestps %xmm4, %xmm1 jae 0x22aba3 movq 0xe0(%rbx), %rdi imulq $0x38, %r8, %r8 movq (%rdi,%r8), %rdx movq 0x38(%rdi,%r8), %r9 movl (%r11,%r13,4), %edi movl 0x10(%r11,%r13,4), %r8d vbroadcastss (%rdx,%rdi,4), %xmm2 vbroadcastss 0x4(%rdx,%rdi,4), %xmm3 vbroadcastss 0x8(%rdx,%rdi,4), %xmm4 vbroadcastss (%r9,%rdi,4), %xmm5 vbroadcastss 0x4(%r9,%rdi,4), %xmm6 vbroadcastss 0x8(%r9,%rdi,4), %xmm7 vbroadcastss 0x1cc2a0d(%rip), %xmm1 # 0x1eec714 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm5, %xmm0, %xmm5 vmulps %xmm6, %xmm0, %xmm6 vmulps %xmm7, %xmm0, %xmm7 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm5, %xmm2, %xmm12 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm6, %xmm3, %xmm14 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm7, %xmm4, %xmm4 vbroadcastss (%rdx,%r8,4), %xmm5 vbroadcastss 0x4(%rdx,%r8,4), %xmm6 vbroadcastss 0x8(%rdx,%r8,4), %xmm7 vbroadcastss (%r9,%r8,4), %xmm8 vbroadcastss 0x4(%r9,%r8,4), %xmm9 vbroadcastss 0x8(%r9,%r8,4), %xmm10 vmulps %xmm0, %xmm8, %xmm8 vmulps %xmm0, %xmm9, %xmm9 vmulps %xmm0, %xmm10, %xmm10 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm5, %xmm8, %xmm2 vmovaps %xmm2, 0x20(%rsp) vmulps %xmm6, %xmm1, %xmm5 vaddps %xmm5, %xmm9, %xmm2 vmovaps %xmm2, 0x10(%rsp) vmulps %xmm7, %xmm1, %xmm5 vaddps %xmm5, %xmm10, %xmm2 vmovaps %xmm2, 0x50(%rsp) movl 0x20(%r11,%r13,4), %edi vbroadcastss (%rdx,%rdi,4), %xmm5 vbroadcastss 0x4(%rdx,%rdi,4), %xmm6 vbroadcastss 0x8(%rdx,%rdi,4), %xmm7 vbroadcastss (%r9,%rdi,4), %xmm8 vbroadcastss 0x4(%r9,%rdi,4), %xmm9 vbroadcastss 0x8(%r9,%rdi,4), %xmm10 vmulps %xmm0, %xmm8, %xmm8 vmulps %xmm0, %xmm9, %xmm9 vmulps %xmm0, %xmm10, %xmm10 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm5, %xmm8, %xmm5 vmovaps %xmm5, 0x1e0(%rsp) vmulps %xmm6, %xmm1, %xmm5 vaddps %xmm5, %xmm9, %xmm5 vmovaps %xmm5, 0x1d0(%rsp) vmulps %xmm7, %xmm1, %xmm5 vaddps %xmm5, %xmm10, %xmm5 vmovaps %xmm5, 0x250(%rsp) movl 0x30(%r11,%r13,4), %edi vbroadcastss (%r9,%rdi,4), %xmm5 vbroadcastss 0x4(%r9,%rdi,4), %xmm6 vbroadcastss 0x8(%r9,%rdi,4), %xmm7 vmulps %xmm5, %xmm0, %xmm5 vmulps %xmm6, %xmm0, %xmm6 vmulps %xmm7, %xmm0, %xmm0 vbroadcastss (%rdx,%rdi,4), %xmm7 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm5, %xmm7, %xmm8 vbroadcastss 0x4(%rdx,%rdi,4), %xmm5 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm6 vbroadcastss 0x8(%rdx,%rdi,4), %xmm5 vmulps %xmm5, %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm1 vmovaps (%r14), %xmm0 vmovaps 0x10(%r14), %xmm11 vmovaps 0x20(%r14), %xmm15 vsubps %xmm0, %xmm12, %xmm7 vsubps %xmm11, %xmm14, %xmm2 vsubps %xmm15, %xmm4, %xmm3 vmovaps %xmm8, 0x3d0(%rsp) vsubps %xmm0, %xmm8, %xmm14 vmovaps %xmm0, %xmm12 vmovaps %xmm6, 0x3c0(%rsp) vsubps %xmm11, %xmm6, %xmm4 vmovaps %xmm4, 0xc0(%rsp) vmovaps %xmm1, 0x3b0(%rsp) vsubps %xmm15, %xmm1, %xmm5 vmovaps %xmm5, 0xb0(%rsp) vsubps %xmm7, %xmm14, %xmm1 vsubps %xmm2, %xmm4, %xmm6 vsubps %xmm3, %xmm5, %xmm0 vaddps %xmm2, %xmm4, %xmm4 vaddps %xmm3, %xmm5, %xmm8 vmulps %xmm0, %xmm4, %xmm9 vmulps %xmm6, %xmm8, %xmm10 vsubps %xmm9, %xmm10, %xmm5 vaddps %xmm7, %xmm14, %xmm9 vmulps %xmm1, %xmm8, %xmm8 vmovaps %xmm0, 0x210(%rsp) vmulps %xmm0, %xmm9, %xmm10 vsubps %xmm8, %xmm10, %xmm10 vmovaps %xmm6, 0x220(%rsp) vmulps %xmm6, %xmm9, %xmm8 vmovaps %xmm1, 0x230(%rsp) vmulps %xmm4, %xmm1, %xmm4 vsubps %xmm8, %xmm4, %xmm4 vmovaps 0x60(%r14), %xmm13 vmulps %xmm4, %xmm13, %xmm4 vmovaps 0x50(%r14), %xmm0 vmulps %xmm0, %xmm10, %xmm10 vaddps %xmm4, %xmm10, %xmm4 vmovaps 0x40(%r14), %xmm10 vmulps %xmm5, %xmm10, %xmm5 vaddps %xmm4, %xmm5, %xmm9 vmovaps 0x10(%rsp), %xmm1 vsubps %xmm11, %xmm1, %xmm5 vmovaps 0x50(%rsp), %xmm1 vsubps %xmm15, %xmm1, %xmm6 vsubps %xmm5, %xmm2, %xmm15 vsubps %xmm6, %xmm3, %xmm11 vmovaps %xmm2, 0x110(%rsp) vaddps %xmm5, %xmm2, %xmm4 vmovaps %xmm3, 0x100(%rsp) vaddps %xmm6, %xmm3, %xmm1 vmulps %xmm4, %xmm11, %xmm2 vmulps %xmm1, %xmm15, %xmm3 vsubps %xmm2, %xmm3, %xmm2 vmovaps 0x20(%rsp), %xmm3 vsubps %xmm12, %xmm3, %xmm3 vsubps %xmm3, %xmm7, %xmm12 vmulps %xmm1, %xmm12, %xmm1 vmovaps %xmm7, 0x120(%rsp) vaddps %xmm3, %xmm7, %xmm7 vmulps %xmm7, %xmm11, %xmm8 vsubps %xmm1, %xmm8, %xmm1 vmovaps %xmm15, 0x1f0(%rsp) vmulps %xmm7, %xmm15, %xmm7 vmulps %xmm4, %xmm12, %xmm4 vsubps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm13, %xmm4 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm4, %xmm1 vmulps %xmm2, %xmm10, %xmm2 vaddps %xmm1, %xmm2, %xmm7 vsubps %xmm14, %xmm3, %xmm4 vaddps %xmm3, %xmm14, %xmm1 vmovaps 0xc0(%rsp), %xmm2 vsubps %xmm2, %xmm5, %xmm15 vaddps %xmm2, %xmm5, %xmm2 vmovaps 0xb0(%rsp), %xmm3 vsubps %xmm3, %xmm6, %xmm14 vaddps %xmm3, %xmm6, %xmm3 vmulps %xmm2, %xmm14, %xmm5 vmulps %xmm3, %xmm15, %xmm6 vsubps %xmm5, %xmm6, %xmm5 vmulps %xmm3, %xmm4, %xmm3 vmulps %xmm1, %xmm14, %xmm6 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm2, %xmm4, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm13, 0xc0(%rsp) vmulps %xmm1, %xmm13, %xmm1 vmovaps %xmm0, 0xb0(%rsp) vmulps %xmm3, %xmm0, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmovaps %xmm10, 0x200(%rsp) vmulps %xmm5, %xmm10, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm7, %xmm9, %xmm2 vaddps %xmm2, %xmm1, %xmm0 vminps %xmm7, %xmm9, %xmm2 vminps %xmm1, %xmm2, %xmm2 vbroadcastss 0x1cf6e70(%rip), %xmm6 # 0x1f20ec4 vandps %xmm6, %xmm0, %xmm8 vbroadcastss 0x1cf6e6b(%rip), %xmm3 # 0x1f20ecc vmulps %xmm3, %xmm8, %xmm3 vbroadcastss 0x1cf6e52(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm3, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm9, 0x3a0(%rsp) vmaxps %xmm7, %xmm9, %xmm5 vmaxps %xmm1, %xmm5, %xmm1 vcmpleps %xmm3, %xmm1, %xmm1 vorps %xmm1, %xmm2, %xmm13 vtestps 0x60(%rsp), %xmm13 je 0x22a734 vmovaps %xmm8, 0x370(%rsp) vmovaps %xmm0, 0x380(%rsp) vmovaps %xmm7, 0x390(%rsp) vmovaps 0x210(%rsp), %xmm9 vmovaps 0x1f0(%rsp), %xmm0 vmulps %xmm0, %xmm9, %xmm1 vmovaps 0x220(%rsp), %xmm8 vmulps %xmm11, %xmm8, %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmulps %xmm11, %xmm15, %xmm3 vmulps %xmm0, %xmm14, %xmm5 vsubps %xmm3, %xmm5, %xmm5 vandps %xmm6, %xmm1, %xmm1 vandps %xmm6, %xmm3, %xmm3 vcmpltps %xmm3, %xmm1, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm7 vmulps %xmm14, %xmm12, %xmm1 vmulps %xmm9, %xmm12, %xmm2 vmovaps 0x230(%rsp), %xmm9 vmulps %xmm11, %xmm9, %xmm3 vsubps %xmm3, %xmm2, %xmm2 vmulps %xmm4, %xmm11, %xmm5 vsubps %xmm1, %xmm5, %xmm5 vandps %xmm6, %xmm3, %xmm3 vandps %xmm6, %xmm1, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm11 vmulps %xmm0, %xmm4, %xmm1 vmulps %xmm0, %xmm9, %xmm2 vmulps %xmm8, %xmm12, %xmm3 vmovaps 0x60(%rsp), %xmm8 vmulps %xmm15, %xmm12, %xmm0 vmovaps %xmm7, %xmm12 vsubps %xmm3, %xmm2, %xmm2 vsubps %xmm1, %xmm0, %xmm0 vandps %xmm6, %xmm3, %xmm3 vandps %xmm6, %xmm1, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm0, %xmm0 vmulps 0xc0(%rsp), %xmm0, %xmm1 vmulps 0xb0(%rsp), %xmm11, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps 0x200(%rsp), %xmm7, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm1, %xmm1, %xmm2 vmulps 0x100(%rsp), %xmm0, %xmm1 vmulps 0x110(%rsp), %xmm11, %xmm3 vaddps %xmm3, %xmm1, %xmm1 vmulps 0x120(%rsp), %xmm7, %xmm3 vaddps %xmm1, %xmm3, %xmm1 vaddps %xmm1, %xmm1, %xmm1 vrcpps %xmm2, %xmm3 vmulps %xmm3, %xmm2, %xmm4 vbroadcastss 0x1cc2546(%rip), %xmm5 # 0x1eec714 vsubps %xmm4, %xmm5, %xmm4 vmulps %xmm4, %xmm3, %xmm4 vaddps %xmm4, %xmm3, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vmovaps 0x30(%r14), %xmm3 vmovaps 0x80(%r14), %xmm7 vcmpleps %xmm1, %xmm3, %xmm3 vcmpleps %xmm7, %xmm1, %xmm4 vandps %xmm4, %xmm3, %xmm3 vxorps %xmm10, %xmm10, %xmm10 vcmpneqps %xmm2, %xmm10, %xmm2 vandps %xmm3, %xmm2, %xmm3 vandps %xmm8, %xmm13, %xmm2 vpslld $0x1f, %xmm3, %xmm3 vpsrad $0x1f, %xmm3, %xmm3 vtestps %xmm2, %xmm3 je 0x22a734 vandps %xmm2, %xmm3, %xmm2 vbroadcastss 0x34(%rbx), %xmm3 vandps 0x90(%r14), %xmm3, %xmm3 vpcmpeqd %xmm3, %xmm10, %xmm3 vtestps %xmm2, %xmm3 vpcmpeqd %xmm9, %xmm9, %xmm9 jb 0x22a734 vandnps %xmm2, %xmm3, %xmm2 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x22a74b cmpq $0x0, 0x48(%rbx) jne 0x22a74b vmovaps 0x20(%rsp), %xmm6 vmovaps 0x10(%rsp), %xmm7 vmovaps 0x50(%rsp), %xmm10 vmovdqa 0x60(%rsp), %xmm0 vpandn %xmm0, %xmm2, %xmm0 vmovdqa %xmm0, 0x60(%rsp) vmovaps 0x60(%rsp), %xmm0 vtestps %xmm0, %xmm0 je 0x22a70b vmovaps (%r14), %xmm1 vmovaps 0x10(%r14), %xmm3 vmovaps 0x20(%r14), %xmm5 vmovaps 0x1e0(%rsp), %xmm0 vsubps %xmm1, %xmm0, %xmm9 vmovaps 0x1d0(%rsp), %xmm0 vsubps %xmm3, %xmm0, %xmm15 vmovaps 0x250(%rsp), %xmm0 vsubps %xmm5, %xmm0, %xmm4 vmovaps 0x3d0(%rsp), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0x10(%rsp) vmovaps 0x3c0(%rsp), %xmm2 vsubps %xmm3, %xmm2, %xmm0 vmovaps %xmm0, 0xb0(%rsp) vmovaps 0x3b0(%rsp), %xmm2 vsubps %xmm5, %xmm2, %xmm8 vsubps %xmm1, %xmm6, %xmm1 vsubps %xmm3, %xmm7, %xmm6 vsubps %xmm5, %xmm10, %xmm5 vmovaps %xmm5, 0x50(%rsp) vsubps %xmm9, %xmm1, %xmm2 vsubps %xmm15, %xmm6, %xmm12 vsubps %xmm4, %xmm5, %xmm0 vaddps %xmm6, %xmm15, %xmm3 vaddps %xmm4, %xmm5, %xmm7 vmulps %xmm0, %xmm3, %xmm10 vmulps %xmm7, %xmm12, %xmm11 vsubps %xmm10, %xmm11, %xmm14 vaddps %xmm1, %xmm9, %xmm10 vmulps %xmm7, %xmm2, %xmm7 vmovaps %xmm0, 0x220(%rsp) vmulps %xmm0, %xmm10, %xmm11 vsubps %xmm7, %xmm11, %xmm7 vmovaps %xmm12, 0x230(%rsp) vmulps %xmm12, %xmm10, %xmm10 vmovaps %xmm2, 0x100(%rsp) vmulps %xmm3, %xmm2, %xmm3 vsubps %xmm10, %xmm3, %xmm3 vmovaps 0x60(%r14), %xmm0 vmulps %xmm0, %xmm3, %xmm3 vmovaps 0x50(%r14), %xmm11 vmulps %xmm7, %xmm11, %xmm7 vaddps %xmm7, %xmm3, %xmm3 vmovaps 0x40(%r14), %xmm13 vmulps %xmm14, %xmm13, %xmm7 vaddps %xmm3, %xmm7, %xmm2 vmovaps %xmm2, 0x20(%rsp) vmovaps 0xb0(%rsp), %xmm5 vsubps %xmm5, %xmm15, %xmm7 vsubps %xmm8, %xmm4, %xmm10 vmovaps %xmm15, 0x120(%rsp) vaddps %xmm5, %xmm15, %xmm3 vmovaps %xmm4, 0x110(%rsp) vaddps %xmm4, %xmm8, %xmm2 vbroadcastss 0x1cf6aeb(%rip), %xmm14 # 0x1f20ec4 vmulps %xmm3, %xmm10, %xmm15 vmulps %xmm2, %xmm7, %xmm4 vsubps %xmm15, %xmm4, %xmm12 vmovaps 0x10(%rsp), %xmm4 vsubps %xmm4, %xmm9, %xmm15 vmulps %xmm2, %xmm15, %xmm2 vmovaps %xmm9, 0xc0(%rsp) vaddps %xmm4, %xmm9, %xmm9 vmovaps %xmm10, 0x200(%rsp) vmulps %xmm10, %xmm9, %xmm10 vsubps %xmm2, %xmm10, %xmm2 vmovaps 0x20(%rsp), %xmm10 vmovaps %xmm7, 0x210(%rsp) vmulps %xmm7, %xmm9, %xmm9 vmulps %xmm3, %xmm15, %xmm3 vsubps %xmm9, %xmm3, %xmm3 vmulps %xmm0, %xmm3, %xmm3 vmulps %xmm2, %xmm11, %xmm2 vaddps %xmm2, %xmm3, %xmm2 vmulps %xmm12, %xmm13, %xmm3 vaddps %xmm2, %xmm3, %xmm12 vsubps %xmm1, %xmm4, %xmm3 vaddps %xmm1, %xmm4, %xmm1 vsubps %xmm6, %xmm5, %xmm9 vaddps %xmm6, %xmm5, %xmm2 vmovaps 0x50(%rsp), %xmm4 vsubps %xmm4, %xmm8, %xmm7 vaddps %xmm4, %xmm8, %xmm4 vmulps %xmm7, %xmm2, %xmm5 vmulps %xmm4, %xmm9, %xmm6 vsubps %xmm5, %xmm6, %xmm5 vmulps %xmm4, %xmm3, %xmm4 vmulps %xmm7, %xmm1, %xmm6 vsubps %xmm4, %xmm6, %xmm4 vmovaps %xmm9, %xmm6 vmulps %xmm1, %xmm9, %xmm1 vmulps %xmm2, %xmm3, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm0, 0x10(%rsp) vmulps %xmm0, %xmm1, %xmm1 vmovaps %xmm11, 0x50(%rsp) vmulps %xmm4, %xmm11, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps %xmm5, %xmm13, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm12, %xmm10, %xmm2 vaddps %xmm2, %xmm1, %xmm9 vminps %xmm12, %xmm10, %xmm2 vminps %xmm1, %xmm2, %xmm2 vandps %xmm14, %xmm9, %xmm5 vbroadcastss 0x1cf6a02(%rip), %xmm4 # 0x1f20ecc vmovaps %xmm5, 0x1f0(%rsp) vmulps %xmm4, %xmm5, %xmm4 vbroadcastss 0x1cf69e0(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm4, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm12, 0x1e0(%rsp) vmaxps %xmm12, %xmm10, %xmm5 vmaxps %xmm1, %xmm5, %xmm1 vcmpleps %xmm4, %xmm1, %xmm1 vorps %xmm1, %xmm2, %xmm0 vtestps 0x60(%rsp), %xmm0 je 0x22a973 vmovaps %xmm13, %xmm11 vmovaps %xmm9, 0x1d0(%rsp) vmovaps 0x220(%rsp), %xmm9 vmovaps %xmm0, 0xb0(%rsp) vmovaps 0x210(%rsp), %xmm0 vmulps %xmm0, %xmm9, %xmm1 vmovaps 0x230(%rsp), %xmm10 vmovaps 0x200(%rsp), %xmm8 vmulps %xmm8, %xmm10, %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmulps %xmm6, %xmm8, %xmm4 vmulps %xmm7, %xmm0, %xmm5 vsubps %xmm4, %xmm5, %xmm5 vandps %xmm1, %xmm14, %xmm1 vandps %xmm4, %xmm14, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm12 vmulps %xmm7, %xmm15, %xmm1 vmulps %xmm9, %xmm15, %xmm2 vmovaps 0x100(%rsp), %xmm13 vmulps %xmm8, %xmm13, %xmm4 vsubps %xmm4, %xmm2, %xmm2 vmulps %xmm3, %xmm8, %xmm5 vsubps %xmm1, %xmm5, %xmm5 vandps %xmm4, %xmm14, %xmm4 vandps %xmm1, %xmm14, %xmm1 vcmpltps %xmm1, %xmm4, %xmm1 vblendvps %xmm1, %xmm2, %xmm5, %xmm9 vmulps %xmm0, %xmm3, %xmm1 vmulps %xmm0, %xmm13, %xmm2 vmulps %xmm10, %xmm15, %xmm3 vmulps %xmm6, %xmm15, %xmm0 vsubps %xmm3, %xmm2, %xmm2 vsubps %xmm1, %xmm0, %xmm0 vandps %xmm3, %xmm14, %xmm3 vandps %xmm1, %xmm14, %xmm1 vcmpltps %xmm1, %xmm3, %xmm1 vblendvps %xmm1, %xmm2, %xmm0, %xmm5 vmulps 0x10(%rsp), %xmm5, %xmm0 vmulps 0x50(%rsp), %xmm9, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm12, %xmm11, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm1 vmulps 0x110(%rsp), %xmm5, %xmm0 vmulps 0x120(%rsp), %xmm9, %xmm2 vaddps %xmm2, %xmm0, %xmm0 vmulps 0xc0(%rsp), %xmm12, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vrcpps %xmm1, %xmm2 vmulps %xmm2, %xmm1, %xmm3 vbroadcastss 0x1cc20e5(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vmulps %xmm2, %xmm0, %xmm0 vmovaps 0x30(%r14), %xmm2 vmovaps 0x80(%r14), %xmm11 vcmpleps %xmm0, %xmm2, %xmm2 vcmpleps %xmm11, %xmm0, %xmm3 vandps %xmm3, %xmm2, %xmm2 vxorps %xmm3, %xmm3, %xmm3 vcmpneqps %xmm3, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm2 vmovaps 0xb0(%rsp), %xmm1 vandps 0x60(%rsp), %xmm1, %xmm1 vpslld $0x1f, %xmm2, %xmm2 vpsrad $0x1f, %xmm2, %xmm2 vtestps %xmm1, %xmm2 je 0x22a973 vandps %xmm1, %xmm2, %xmm1 movq (%r10), %rcx movl 0x40(%r11,%r13,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %rbx vbroadcastss 0x34(%rbx), %xmm2 vandps 0x90(%r14), %xmm2, %xmm2 vpcmpeqd %xmm3, %xmm2, %xmm2 vtestps %xmm1, %xmm2 vmovaps 0xe0(%rsp), %xmm4 jb 0x22a6fb movl 0x50(%r11,%r13,4), %ecx vandnps %xmm1, %xmm2, %xmm1 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x22a981 cmpq $0x0, 0x48(%rbx) jne 0x22a981 vmovdqa 0x60(%rsp), %xmm0 vpandn %xmm0, %xmm1, %xmm0 vmovdqa %xmm0, 0x60(%rsp) vmovaps 0x60(%rsp), %xmm0 vtestps %xmm0, %xmm0 setne %al jmp 0x22a716 xorl %eax, %eax vmovaps 0xe0(%rsp), %xmm4 testb %al, %al je 0x22b029 leaq 0x1(%r13), %rax cmpq $0x3, %r13 movq %rax, %r13 jb 0x229c18 jmp 0x22b029 vmovaps 0x20(%rsp), %xmm6 vmovaps 0x10(%rsp), %xmm7 vmovaps 0x50(%rsp), %xmm10 jmp 0x22a28a vmovaps 0x380(%rsp), %xmm4 vrcpps %xmm4, %xmm3 vmulps %xmm3, %xmm4, %xmm4 vbroadcastss 0x1cc1faf(%rip), %xmm6 # 0x1eec714 vsubps %xmm4, %xmm6, %xmm4 vmulps %xmm4, %xmm3, %xmm4 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1cc686e(%rip), %xmm4 # 0x1ef0fe8 vmovaps 0x370(%rsp), %xmm5 vcmpnltps %xmm4, %xmm5, %xmm4 vandps %xmm3, %xmm4, %xmm3 vmulps 0x3a0(%rsp), %xmm3, %xmm4 vminps %xmm6, %xmm4, %xmm4 vmulps 0x390(%rsp), %xmm3, %xmm3 vminps %xmm6, %xmm3, %xmm3 vsubps %xmm4, %xmm6, %xmm5 vsubps %xmm3, %xmm6, %xmm6 vmovaps 0x350(%rsp), %xmm8 vblendvps %xmm8, %xmm5, %xmm4, %xmm4 vblendvps %xmm8, %xmm6, %xmm3, %xmm3 movq 0x8(%r10), %rdx vmovd %ecx, %xmm5 vpshufd $0x0, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vmovd %eax, %xmm6 vpshufd $0x0, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmovaps %xmm12, 0x140(%rsp) vmovaps %xmm11, 0x150(%rsp) vmovaps %xmm0, 0x160(%rsp) vmovaps %xmm4, 0x170(%rsp) vmovaps %xmm3, 0x180(%rsp) vmovdqa %xmm6, 0x190(%rsp) vmovdqa %xmm5, 0x1a0(%rsp) leaq 0x1b0(%rsp), %rax vcmptrueps %ymm0, %ymm0, %ymm0 vmovups %ymm0, (%rax) vbroadcastss (%rdx), %xmm0 vmovaps %xmm0, 0x1b0(%rsp) vbroadcastss 0x4(%rdx), %xmm0 vmovaps %xmm0, 0x1c0(%rsp) vblendvps %xmm2, %xmm1, %xmm7, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm2, 0x70(%rsp) leaq 0x70(%rsp), %rax movq %rax, 0x80(%rsp) movq 0x18(%rbx), %rax movq %rax, 0x88(%rsp) movq 0x8(%r10), %rax movq %rax, 0x90(%rsp) movq %r14, 0x98(%rsp) leaq 0x140(%rsp), %rax movq %rax, 0xa0(%rsp) movl $0x4, 0xa8(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax movq %r11, 0x8(%rsp) vmovaps %xmm7, 0xb0(%rsp) je 0x22a8e9 leaq 0x80(%rsp), %rdi vzeroupper callq %rax vmovaps 0xb0(%rsp), %xmm7 movq 0x8(%rsp), %r11 vxorps %xmm10, %xmm10, %xmm10 vpcmpeqd %xmm9, %xmm9, %xmm9 movq 0x38(%rsp), %rsi movq 0x40(%rsp), %r10 vmovdqa 0x70(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x22afcf movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x22a941 testb $0x2, (%rcx) jne 0x22a912 testb $0x40, 0x3e(%rbx) je 0x22a941 leaq 0x80(%rsp), %rdi vzeroupper callq %rax vmovaps 0xb0(%rsp), %xmm7 movq 0x8(%rsp), %r11 vxorps %xmm10, %xmm10, %xmm10 vpcmpeqd %xmm9, %xmm9, %xmm9 movq 0x38(%rsp), %rsi movq 0x40(%rsp), %r10 vpcmpeqd 0x70(%rsp), %xmm10, %xmm1 vpxor %xmm1, %xmm9, %xmm0 movq 0x98(%rsp), %rax vbroadcastss 0x1cc2228(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x22afd7 vmovaps 0xe0(%rsp), %xmm4 jmp 0x22a6fb vmovaps 0x1d0(%rsp), %xmm3 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cc1d79(%rip), %xmm6 # 0x1eec714 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cc6638(%rip), %xmm3 # 0x1ef0fe8 vmovaps 0x1f0(%rsp), %xmm4 vcmpnltps %xmm3, %xmm4, %xmm3 vandps %xmm2, %xmm3, %xmm2 vmulps 0x20(%rsp), %xmm2, %xmm3 vminps %xmm6, %xmm3, %xmm3 vmulps 0x1e0(%rsp), %xmm2, %xmm2 vminps %xmm6, %xmm2, %xmm2 vsubps %xmm3, %xmm6, %xmm4 vsubps %xmm2, %xmm6, %xmm6 vmovaps 0x340(%rsp), %xmm7 vblendvps %xmm7, %xmm4, %xmm3, %xmm3 vblendvps %xmm7, %xmm6, %xmm2, %xmm2 movq 0x8(%r10), %rdx vmovd %eax, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovd %ecx, %xmm6 vpshufd $0x0, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmovaps %xmm12, 0x140(%rsp) vmovaps %xmm9, 0x150(%rsp) vmovaps %xmm5, 0x160(%rsp) vmovaps %xmm3, 0x170(%rsp) vmovaps %xmm2, 0x180(%rsp) vmovdqa %xmm6, 0x190(%rsp) vmovdqa %xmm4, 0x1a0(%rsp) leaq 0x1b0(%rsp), %rax vcmptrueps %ymm2, %ymm2, %ymm2 vmovups %ymm2, (%rax) vbroadcastss (%rdx), %xmm2 vmovaps %xmm2, 0x1b0(%rsp) vbroadcastss 0x4(%rdx), %xmm2 vmovaps %xmm2, 0x1c0(%rsp) vblendvps %xmm1, %xmm0, %xmm11, %xmm0 vmovaps %xmm0, 0x80(%r14) vmovaps %xmm1, 0x70(%rsp) leaq 0x70(%rsp), %rax movq %rax, 0x80(%rsp) movq 0x18(%rbx), %rax movq %rax, 0x88(%rsp) movq 0x8(%r10), %rax movq %rax, 0x90(%rsp) movq %r14, 0x98(%rsp) leaq 0x140(%rsp), %rax movq %rax, 0xa0(%rsp) movl $0x4, 0xa8(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax movq %r11, 0x8(%rsp) vmovaps %xmm11, 0x10(%rsp) je 0x22ab0c leaq 0x80(%rsp), %rdi vzeroupper callq %rax vmovaps 0x10(%rsp), %xmm11 movq 0x8(%rsp), %r11 movq 0x38(%rsp), %rsi movq 0x40(%rsp), %r10 vmovdqa 0x70(%rsp), %xmm0 vptest %xmm0, %xmm0 vmovaps 0xe0(%rsp), %xmm4 je 0x22aff8 movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x22ab69 testb $0x2, (%rcx) jne 0x22ab3e testb $0x40, 0x3e(%rbx) je 0x22ab69 leaq 0x80(%rsp), %rdi vzeroupper callq %rax vmovaps 0x10(%rsp), %xmm11 movq 0x8(%rsp), %r11 vmovaps 0xe0(%rsp), %xmm4 movq 0x38(%rsp), %rsi movq 0x40(%rsp), %r10 vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x70(%rsp), %xmm0, %xmm1 vpxor 0x1cc12a5(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x98(%rsp), %rax vbroadcastss 0x1cc1ff8(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x22b008 movq %r11, 0x8(%rsp) cmpq $0x0, 0x30(%rsp) je 0x22ac4e movq 0xe0(%rbx), %rsi movq %rsi, 0x10(%rsp) addq $0x38, %rsi movq %rsi, 0x20(%rsp) movq 0x8(%rsp), %rdi movl (%rdi,%r13,4), %r8d movq 0x30(%rsp), %r9 movq %rdx, %r10 movslq 0x70(%rsp,%r10,4), %r11 imulq $0x38, %r11, %r11 movq 0x10(%rsp), %rsi movq (%rsi,%r11), %rdi movq 0x20(%rsp), %rsi movq (%rsi,%r11), %r11 vmovups (%rdi,%r8,4), %xmm1 vmovdqu (%r11,%r8,4), %xmm2 vmovss %xmm1, 0x140(%rsp,%r10,4) vextractps $0x1, %xmm1, 0x150(%rsp,%r10,4) vextractps $0x2, %xmm1, 0x160(%rsp,%r10,4) vmovd %xmm2, 0x80(%rsp,%r10,4) vpextrd $0x1, %xmm2, 0x90(%rsp,%r10,4) vpextrd $0x2, %xmm2, 0xa0(%rsp,%r10,4) btcq %r10, %r9 bsfq %r9, %r10 testq %r9, %r9 jne 0x22abda vbroadcastss 0x1cc1abd(%rip), %xmm1 # 0x1eec714 vsubps %xmm0, %xmm1, %xmm1 vmulps 0x140(%rsp), %xmm1, %xmm12 vmulps 0x150(%rsp), %xmm1, %xmm14 vmulps 0x160(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x50(%rsp) vmulps 0x80(%rsp), %xmm0, %xmm2 vmovaps %xmm2, 0xb0(%rsp) vmulps 0x90(%rsp), %xmm0, %xmm2 vmovaps %xmm2, 0xc0(%rsp) vmulps 0xa0(%rsp), %xmm0, %xmm4 cmpq $0x0, 0x30(%rsp) je 0x22ad50 movq 0xe0(%rbx), %rsi movq %rsi, 0x10(%rsp) addq $0x38, %rsi movq %rsi, 0x20(%rsp) movq 0x8(%rsp), %rdi movl 0x10(%rdi,%r13,4), %r8d movq 0x30(%rsp), %r9 movq %rdx, %r10 movslq 0x70(%rsp,%r10,4), %r11 imulq $0x38, %r11, %r11 movq 0x10(%rsp), %rsi movq (%rsi,%r11), %rdi movq 0x20(%rsp), %rsi movq (%rsi,%r11), %r11 vmovups (%rdi,%r8,4), %xmm5 vmovups (%r11,%r8,4), %xmm6 vmovss %xmm5, 0x140(%rsp,%r10,4) vextractps $0x1, %xmm5, 0x150(%rsp,%r10,4) vextractps $0x2, %xmm5, 0x160(%rsp,%r10,4) vmovss %xmm6, 0x80(%rsp,%r10,4) vextractps $0x1, %xmm6, 0x90(%rsp,%r10,4) vextractps $0x2, %xmm6, 0xa0(%rsp,%r10,4) btcq %r10, %r9 bsfq %r9, %r10 testq %r9, %r9 jne 0x22acdc vmulps 0x140(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x120(%rsp) vmulps 0x150(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x110(%rsp) vmulps 0x160(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x100(%rsp) vmulps 0x80(%rsp), %xmm0, %xmm11 vmulps 0x90(%rsp), %xmm0, %xmm9 vmulps 0xa0(%rsp), %xmm0, %xmm13 cmpq $0x0, 0x30(%rsp) je 0x22ae48 movq 0xe0(%rbx), %rsi movq %rsi, 0x10(%rsp) addq $0x38, %rsi movq %rsi, 0x20(%rsp) movq 0x8(%rsp), %rdi movl 0x20(%rdi,%r13,4), %r8d movq 0x30(%rsp), %r9 movq %rdx, %r10 movslq 0x70(%rsp,%r10,4), %r11 imulq $0x38, %r11, %r11 movq 0x10(%rsp), %rsi movq (%rsi,%r11), %rdi movq 0x20(%rsp), %rsi movq (%rsi,%r11), %r11 vmovups (%rdi,%r8,4), %xmm5 vmovups (%r11,%r8,4), %xmm6 vmovss %xmm5, 0x140(%rsp,%r10,4) vextractps $0x1, %xmm5, 0x150(%rsp,%r10,4) vextractps $0x2, %xmm5, 0x160(%rsp,%r10,4) vmovss %xmm6, 0x80(%rsp,%r10,4) vextractps $0x1, %xmm6, 0x90(%rsp,%r10,4) vextractps $0x2, %xmm6, 0xa0(%rsp,%r10,4) btcq %r10, %r9 bsfq %r9, %r10 testq %r9, %r9 jne 0x22add4 vmulps 0x140(%rsp), %xmm1, %xmm5 vmulps 0x150(%rsp), %xmm1, %xmm6 vmulps 0x160(%rsp), %xmm1, %xmm10 vmulps 0x80(%rsp), %xmm0, %xmm7 vmulps 0x90(%rsp), %xmm0, %xmm8 vmulps 0xa0(%rsp), %xmm0, %xmm15 cmpq $0x0, 0x30(%rsp) je 0x22af0c movq 0xe0(%rbx), %rsi leaq 0x38(%rsi), %rdi movq 0x8(%rsp), %r8 movl 0x30(%r8,%r13,4), %r8d movq 0x30(%rsp), %r9 movslq 0x70(%rsp,%rdx,4), %r10 imulq $0x38, %r10, %r10 movq (%rsi,%r10), %r11 movq (%rdi,%r10), %r10 vmovups (%r11,%r8,4), %xmm2 vmovups (%r10,%r8,4), %xmm3 vmovss %xmm2, 0x140(%rsp,%rdx,4) vextractps $0x1, %xmm2, 0x150(%rsp,%rdx,4) vextractps $0x2, %xmm2, 0x160(%rsp,%rdx,4) vmovss %xmm3, 0x80(%rsp,%rdx,4) vextractps $0x1, %xmm3, 0x90(%rsp,%rdx,4) vextractps $0x2, %xmm3, 0xa0(%rsp,%rdx,4) btcq %rdx, %r9 bsfq %r9, %rdx testq %r9, %r9 jne 0x22aea4 vaddps 0xb0(%rsp), %xmm12, %xmm12 vaddps 0xc0(%rsp), %xmm14, %xmm14 vaddps 0x50(%rsp), %xmm4, %xmm4 vaddps 0x120(%rsp), %xmm11, %xmm2 vmovaps %xmm2, 0x20(%rsp) vaddps 0x110(%rsp), %xmm9, %xmm2 vmovaps %xmm2, 0x10(%rsp) vaddps 0x100(%rsp), %xmm13, %xmm2 vmovaps %xmm2, 0x50(%rsp) vaddps %xmm7, %xmm5, %xmm5 vmovaps %xmm5, 0x1e0(%rsp) vaddps %xmm6, %xmm8, %xmm5 vmovaps %xmm5, 0x1d0(%rsp) vmulps 0x140(%rsp), %xmm1, %xmm5 vmulps 0x150(%rsp), %xmm1, %xmm6 vmulps 0x160(%rsp), %xmm1, %xmm1 vmulps 0x80(%rsp), %xmm0, %xmm7 vaddps %xmm15, %xmm10, %xmm8 vmovaps %xmm8, 0x250(%rsp) vaddps %xmm7, %xmm5, %xmm8 vmulps 0x90(%rsp), %xmm0, %xmm5 vaddps %xmm5, %xmm6, %xmm6 vmulps 0xa0(%rsp), %xmm0, %xmm0 vaddps %xmm0, %xmm1, %xmm1 movq 0x40(%rsp), %r10 movq 0x38(%rsp), %rsi movq 0x8(%rsp), %r11 jmp 0x229e4a vpcmpeqd %xmm0, %xmm10, %xmm0 vpxor %xmm0, %xmm9, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm2 movq 0xd8(%rsp), %rax vblendvps %xmm0, (%rax), %xmm7, %xmm0 vmovaps %xmm0, (%rax) jmp 0x22a268 vpcmpeqd 0x1cc0a10(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cc0e18(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm1 movq 0xd8(%rsp), %rax vblendvps %xmm0, (%rax), %xmm11, %xmm0 vmovaps %xmm0, (%rax) jmp 0x22a6eb vmovaps 0x60(%rsp), %xmm1 vandps %xmm4, %xmm1, %xmm0 vtestps %xmm4, %xmm1 movq 0x248(%rsp), %rax je 0x22b057 incq %rax vmovaps %xmm0, %xmm4 cmpq 0x240(%rsp), %rax jb 0x229bee vpcmpeqd %xmm1, %xmm1, %xmm1 vxorps %xmm1, %xmm0, %xmm0 movq 0xd0(%rsp), %r8 jmp 0x22b06d vpcmpeqd %xmm1, %xmm1, %xmm1 vorps 0xf0(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0xf0(%rsp) vtestps %xmm1, %xmm0 jb 0x22b0ab vmovaps 0x330(%rsp), %xmm1 vbroadcastss 0x1cc1aec(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x330(%rsp) xorl %eax, %eax jmp 0x22b0ae pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x2298b3 jmp 0x22b18d pushq $0x2 jmp 0x22b0ad vmovaps %xmm1, 0x20(%rsp) movq %r8, 0xd0(%rsp) movq 0x138(%rsp), %rdi movq %r10, 0x40(%rsp) movq %rsi, 0x38(%rsp) bsfq %r13, %rcx movq %rbp, %rdx movq %rcx, 0x60(%rsp) leaq 0x4f(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x268(%rsp), %rax pushq %rax vzeroupper callq 0x26e66c popq %rcx popq %rdx testb %al, %al je 0x22b11a movq 0x60(%rsp), %rax orl $-0x1, 0xf0(%rsp,%rax,4) leaq -0x1(%r13), %rax andq %rax, %r13 movq 0x40(%rsp), %r10 movq 0x138(%rsp), %rdi movq 0x38(%rsp), %rsi jne 0x22b0e0 vmovaps 0xf0(%rsp), %xmm0 vtestps 0x1cc0cd9(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0xd0(%rsp), %r8 vmovaps 0x20(%rsp), %xmm1 jb 0x2298f9 vmovaps 0x330(%rsp), %xmm1 vbroadcastss 0x1cc1a14(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps 0x20(%rsp), %xmm1 vmovaps %xmm0, 0x330(%rsp) pushq $0x2 popq %rax jmp 0x2298f9 vmovaps 0x360(%rsp), %xmm0 vandps 0xf0(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cc19dc(%rip), %xmm1 # 0x1eecb84 movq 0xd8(%rsp), %rax vmaskmovps %xmm1, %xmm0, (%rax) addq $0x1a88, %rsp # imm = 0x1A88 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 257, false, embree::avx::VirtualCurveIntersectorK<4>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x18e8, %rsp # imm = 0x18E8 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x22bfbc movq %rdx, %r9 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm1 vmovaps 0x80(%rdx), %xmm0 vxorps %xmm4, %xmm4, %xmm4 vcmpnltps %xmm4, %xmm0, %xmm2 vandps %xmm1, %xmm2, %xmm1 vmovaps %xmm1, 0x60(%rsp) vmovmskps %xmm1, %eax testl %eax, %eax je 0x22bfbc movq %rcx, %r10 leaq 0x80(%r9), %r11 movzbl %al, %eax leaq 0x180(%rsp), %rcx vmovaps 0x40(%r9), %xmm2 vmovaps 0x50(%r9), %xmm3 vmovaps 0x60(%r9), %xmm1 vmulps %xmm1, %xmm1, %xmm5 vmulps %xmm3, %xmm3, %xmm6 vaddps %xmm6, %xmm5, %xmm5 vmulps %xmm2, %xmm2, %xmm6 vaddps %xmm5, %xmm6, %xmm5 vrsqrtps %xmm5, %xmm6 vbroadcastss 0x1cc10ff(%rip), %xmm7 # 0x1eec718 vmulps %xmm7, %xmm6, %xmm7 vbroadcastss 0x1cc10f6(%rip), %xmm8 # 0x1eec71c vmulps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vmulps %xmm6, %xmm6, %xmm6 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm7, %xmm5 vmovaps %xmm5, -0x10(%rcx) vbroadcastss 0x1cf5878(%rip), %xmm15 # 0x1f20ec0 vmovss 0x1cc10c8(%rip), %xmm5 # 0x1eec718 vmovss 0x1cc1528(%rip), %xmm6 # 0x1eecb80 bsfq %rax, %rdx vmovss 0x40(%r9,%rdx,4), %xmm7 vinsertps $0x1c, 0x50(%r9,%rdx,4), %xmm7, %xmm7 # xmm7 = xmm7[0],mem[0],zero,zero vinsertps $0x28, 0x60(%r9,%rdx,4), %xmm7, %xmm8 # xmm8 = xmm7[0,1],mem[0],zero vbroadcastss 0x170(%rsp,%rdx,4), %xmm7 vmulps %xmm7, %xmm8, %xmm8 vshufpd $0x1, %xmm8, %xmm8, %xmm9 # xmm9 = xmm8[1,0] vmovshdup %xmm8, %xmm10 # xmm10 = xmm8[1,1,3,3] vxorps %xmm15, %xmm10, %xmm10 vunpckhps %xmm4, %xmm8, %xmm11 # xmm11 = xmm8[2],xmm4[2],xmm8[3],xmm4[3] vmovss %xmm10, %xmm4, %xmm10 # xmm10 = xmm10[0],xmm4[1,2,3] vshufps $0x41, %xmm10, %xmm11, %xmm10 # xmm10 = xmm11[1,0],xmm10[0,1] vxorpd %xmm15, %xmm9, %xmm9 vinsertps $0x2a, %xmm8, %xmm9, %xmm9 # xmm9 = xmm9[0],zero,xmm8[0],zero vdpps $0x7f, %xmm10, %xmm10, %xmm11 vdpps $0x7f, %xmm9, %xmm9, %xmm12 vcmpltps %xmm11, %xmm12, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vblendvps %xmm11, %xmm10, %xmm9, %xmm9 vdpps $0x7f, %xmm9, %xmm9, %xmm10 vrsqrtss %xmm10, %xmm10, %xmm11 vmulss %xmm5, %xmm11, %xmm12 vmulss %xmm6, %xmm10, %xmm10 vmulss %xmm11, %xmm10, %xmm10 vmulss %xmm11, %xmm11, %xmm11 vmulss %xmm11, %xmm10, %xmm10 vsubss %xmm10, %xmm12, %xmm10 vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vmulps %xmm10, %xmm9, %xmm9 vshufps $0xc9, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,2,0,3] vshufps $0xc9, %xmm8, %xmm8, %xmm11 # xmm11 = xmm8[1,2,0,3] vmulps %xmm9, %xmm11, %xmm11 vmulps %xmm10, %xmm8, %xmm10 vsubps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[1,2,0,3] vdpps $0x7f, %xmm10, %xmm10, %xmm11 leaq -0x1(%rax), %rdi imulq $0x30, %rdx, %rdx vrsqrtss %xmm11, %xmm11, %xmm12 vmulss %xmm5, %xmm12, %xmm13 vmulss %xmm6, %xmm11, %xmm11 vmulss %xmm12, %xmm11, %xmm11 vmulss %xmm12, %xmm12, %xmm12 vmulss %xmm12, %xmm11, %xmm11 vsubss %xmm11, %xmm13, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vmulps %xmm10, %xmm11, %xmm10 vmulps %xmm7, %xmm8, %xmm7 vunpcklps %xmm7, %xmm9, %xmm8 # xmm8 = xmm9[0],xmm7[0],xmm9[1],xmm7[1] vunpckhps %xmm7, %xmm9, %xmm7 # xmm7 = xmm9[2],xmm7[2],xmm9[3],xmm7[3] vunpcklps %xmm4, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm4[0],xmm10[1],xmm4[1] vunpckhps %xmm4, %xmm10, %xmm10 # xmm10 = xmm10[2],xmm4[2],xmm10[3],xmm4[3] vunpcklps %xmm10, %xmm7, %xmm7 # xmm7 = xmm7[0],xmm10[0],xmm7[1],xmm10[1] vunpcklps %xmm9, %xmm8, %xmm10 # xmm10 = xmm8[0],xmm9[0],xmm8[1],xmm9[1] vunpckhps %xmm9, %xmm8, %xmm8 # xmm8 = xmm8[2],xmm9[2],xmm8[3],xmm9[3] vmovaps %xmm10, (%rcx,%rdx) vmovaps %xmm8, 0x10(%rcx,%rdx) vmovaps %xmm7, 0x20(%rcx,%rdx) andq %rdi, %rax jne 0x22b658 vbroadcastss 0x1cf5728(%rip), %xmm9 # 0x1f20ec4 vbroadcastss 0x1cc5843(%rip), %xmm10 # 0x1ef0fe8 vandps %xmm3, %xmm9, %xmm4 vcmpltps %xmm10, %xmm4, %xmm4 vblendvps %xmm4, %xmm10, %xmm3, %xmm4 vbroadcastss 0x1cc0f56(%rip), %xmm11 # 0x1eec714 vrcpps %xmm4, %xmm5 vmulps %xmm5, %xmm4, %xmm4 vsubps %xmm4, %xmm11, %xmm4 vmulps %xmm4, %xmm5, %xmm4 vaddps %xmm4, %xmm5, %xmm5 vxorps %xmm4, %xmm4, %xmm4 vcmpnltps %xmm4, %xmm5, %xmm6 vbroadcastss 0x1d291bc(%rip), %xmm7 # 0x1f549a0 vbroadcastss 0x1d2f17b(%rip), %xmm8 # 0x1f5a968 vblendvps %xmm6, %xmm7, %xmm8, %xmm6 vmovaps (%r9), %xmm7 vmovaps %xmm7, 0x80(%rsp) vmovaps 0x10(%r9), %xmm7 vmovaps %xmm7, 0x90(%rsp) vmovdqa 0x20(%r9), %xmm7 vmovdqa %xmm7, 0xa0(%rsp) vmovaps %xmm2, 0xb0(%rsp) vmovaps %xmm3, 0xc0(%rsp) vmovaps %xmm1, 0xd0(%rsp) vandps %xmm2, %xmm9, %xmm3 vcmpltps %xmm10, %xmm3, %xmm3 vblendvps %xmm3, %xmm10, %xmm2, %xmm2 vandps %xmm1, %xmm9, %xmm3 vcmpltps %xmm10, %xmm3, %xmm3 vblendvps %xmm3, %xmm10, %xmm1, %xmm1 vrcpps %xmm2, %xmm3 vmulps %xmm3, %xmm2, %xmm2 vsubps %xmm2, %xmm11, %xmm2 vmulps %xmm2, %xmm3, %xmm2 vaddps %xmm2, %xmm3, %xmm2 vrcpps %xmm1, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vsubps %xmm1, %xmm11, %xmm1 vmulps %xmm1, %xmm3, %xmm1 vaddps %xmm1, %xmm3, %xmm1 vmovaps %xmm2, 0xe0(%rsp) vmovaps %xmm5, 0xf0(%rsp) vmovaps %xmm1, 0x100(%rsp) vcmpltps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d2f0b9(%rip), %xmm3 # 0x1f5a964 vandps %xmm3, %xmm2, %xmm2 vmovaps %xmm2, 0x110(%rsp) vmovaps %xmm6, 0x120(%rsp) vcmpnltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d2f09d(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d2f098(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0x130(%rsp) vmovaps 0x30(%r9), %xmm1 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm0, %xmm2 vbroadcastss 0x1cc0122(%rip), %xmm10 # 0x1eeba20 vmovaps 0x60(%rsp), %xmm3 vblendvps %xmm3, %xmm1, %xmm10, %xmm0 vmovaps %xmm0, 0x140(%rsp) vbroadcastss 0x1cc1268(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm3, %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x150(%rsp) vpcmpeqd %xmm1, %xmm1, %xmm1 vxorps %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0x30(%rsp) cmpq $0x0, 0x8(%r10) je 0x22b954 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r15d, %r15d cmpb $0x1, %al adcq $0x2, %r15 jmp 0x22b958 pushq $0x3 popq %r15 leaq 0x250(%rsp), %r12 movq $-0x8, -0x10(%r12) leaq 0x9f0(%rsp), %rbx vmovaps %xmm10, -0x20(%rbx) movq 0x70(%r8), %rax movq %rax, -0x8(%r12) vmovaps %xmm0, -0x10(%rbx) leaq 0x1f245f5(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x160(%rsp) movq %r10, 0x48(%rsp) movq %r9, 0x40(%rsp) addq $-0x10, %rbx movq -0x8(%r12), %rbp addq $-0x8, %r12 cmpq $-0x8, %rbp je 0x22bec7 vmovaps (%rbx), %xmm11 vcmpltps 0x150(%rsp), %xmm11, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22becb movzbl %al, %r14d popcntl %r14d, %r13d xorl %eax, %eax cmpq %r15, %r13 jbe 0x22bed0 cmpq %r15, %r13 jbe 0x22beb9 testb $0x8, %bpl pushq $0x8 popq %r14 jne 0x22bda1 movq %rbp, %rax movq %rbp, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %r14, %rbp vmovaps %xmm10, %xmm11 movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x22bb2d testb $0x7, %al jne 0x22bb4d vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x80(%rsp), %xmm3 vmovaps 0x90(%rsp), %xmm4 vmovaps 0xa0(%rsp), %xmm5 vmovaps 0xe0(%rsp), %xmm6 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm4, %xmm0, %xmm0 vmovaps 0xf0(%rsp), %xmm7 vmulps %xmm0, %xmm7, %xmm1 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm5, %xmm0, %xmm0 vmovaps 0x100(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm0 vbroadcastss 0x30(%rax,%rdx,4), %xmm9 vsubps %xmm3, %xmm9, %xmm3 vmulps %xmm3, %xmm6, %xmm3 vbroadcastss 0x50(%rax,%rdx,4), %xmm6 vsubps %xmm4, %xmm6, %xmm4 vmulps %xmm4, %xmm7, %xmm4 vbroadcastss 0x70(%rax,%rdx,4), %xmm6 vsubps %xmm5, %xmm6, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vpminsd %xmm3, %xmm2, %xmm6 vpminsd %xmm4, %xmm1, %xmm7 vpmaxsd %xmm7, %xmm6, %xmm6 vpminsd %xmm5, %xmm0, %xmm7 vpmaxsd %xmm7, %xmm6, %xmm6 vpmaxsd %xmm3, %xmm2, %xmm2 vpmaxsd %xmm4, %xmm1, %xmm1 vpminsd %xmm1, %xmm2, %xmm1 vpmaxsd %xmm5, %xmm0, %xmm0 vpminsd %xmm0, %xmm1, %xmm0 vpmaxsd 0x140(%rsp), %xmm6, %xmm1 vpminsd 0x150(%rsp), %xmm0, %xmm0 vcmpleps %xmm0, %xmm1, %xmm0 vtestps %xmm0, %xmm0 je 0x22bb2d vblendvps %xmm0, %xmm6, %xmm10, %xmm0 cmpq $0x8, %rbp je 0x22bb26 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm11, (%rbx) addq $0x10, %rbx vmovaps %xmm0, %xmm11 movq %rdi, %rbp cmpq $0x8, %rdi je 0x22bd50 leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x22ba12 jmp 0x22bd50 vbroadcastss 0x20(%rcx,%rdx,4), %xmm2 vbroadcastss 0x30(%rcx,%rdx,4), %xmm1 vbroadcastss 0x40(%rcx,%rdx,4), %xmm0 vbroadcastss 0x50(%rcx,%rdx,4), %xmm5 vbroadcastss 0x60(%rcx,%rdx,4), %xmm4 vbroadcastss 0x70(%rcx,%rdx,4), %xmm3 vbroadcastss 0x80(%rcx,%rdx,4), %xmm8 vbroadcastss 0x90(%rcx,%rdx,4), %xmm7 vbroadcastss 0xa0(%rcx,%rdx,4), %xmm6 vmovaps 0xc0(%rsp), %xmm9 vmovaps 0xd0(%rsp), %xmm10 vmovaps %xmm11, 0x20(%rsp) vmulps %xmm10, %xmm8, %xmm11 vmulps %xmm7, %xmm10, %xmm12 vmulps %xmm5, %xmm9, %xmm13 vaddps %xmm11, %xmm13, %xmm11 vmulps %xmm4, %xmm9, %xmm13 vaddps %xmm12, %xmm13, %xmm12 vmovaps 0xb0(%rsp), %xmm13 vmulps %xmm6, %xmm10, %xmm10 vmulps %xmm3, %xmm9, %xmm9 vaddps %xmm10, %xmm9, %xmm9 vmulps %xmm2, %xmm13, %xmm10 vaddps %xmm11, %xmm10, %xmm10 vmulps %xmm1, %xmm13, %xmm11 vaddps %xmm12, %xmm11, %xmm11 vmulps %xmm0, %xmm13, %xmm12 vaddps %xmm9, %xmm12, %xmm9 vbroadcastss 0x1cf52c2(%rip), %xmm13 # 0x1f20ec4 vandps %xmm13, %xmm10, %xmm12 vbroadcastss 0x1cc53d8(%rip), %xmm14 # 0x1ef0fe8 vcmpltps %xmm14, %xmm12, %xmm12 vblendvps %xmm12, %xmm14, %xmm10, %xmm10 vandps %xmm13, %xmm11, %xmm12 vcmpltps %xmm14, %xmm12, %xmm12 vblendvps %xmm12, %xmm14, %xmm11, %xmm11 vandps %xmm13, %xmm9, %xmm12 vcmpltps %xmm14, %xmm12, %xmm12 vblendvps %xmm12, %xmm14, %xmm9, %xmm12 vbroadcastss 0xb0(%rcx,%rdx,4), %xmm13 vrcpps %xmm10, %xmm9 vmulps %xmm10, %xmm9, %xmm10 vbroadcastss 0x1cc0ab9(%rip), %xmm14 # 0x1eec714 vsubps %xmm10, %xmm14, %xmm10 vmulps %xmm10, %xmm9, %xmm10 vaddps %xmm10, %xmm9, %xmm9 vrcpps %xmm11, %xmm10 vmulps %xmm11, %xmm10, %xmm11 vsubps %xmm11, %xmm14, %xmm11 vmulps %xmm11, %xmm10, %xmm11 vaddps %xmm11, %xmm10, %xmm10 vrcpps %xmm12, %xmm11 vmulps %xmm12, %xmm11, %xmm12 vsubps %xmm12, %xmm14, %xmm12 vmulps %xmm12, %xmm11, %xmm12 vaddps %xmm12, %xmm11, %xmm11 vmovaps 0xa0(%rsp), %xmm12 vmulps %xmm12, %xmm8, %xmm8 vaddps %xmm8, %xmm13, %xmm8 vbroadcastss 0xc0(%rcx,%rdx,4), %xmm13 vmulps %xmm7, %xmm12, %xmm7 vaddps %xmm7, %xmm13, %xmm7 vmulps %xmm6, %xmm12, %xmm6 vbroadcastss 0xd0(%rcx,%rdx,4), %xmm12 vaddps %xmm6, %xmm12, %xmm6 vmovaps 0x90(%rsp), %xmm12 vmulps %xmm5, %xmm12, %xmm5 vaddps %xmm5, %xmm8, %xmm5 vmulps %xmm4, %xmm12, %xmm4 vaddps %xmm7, %xmm4, %xmm4 vmulps %xmm3, %xmm12, %xmm3 vaddps %xmm6, %xmm3, %xmm3 vmovaps 0x80(%rsp), %xmm6 vmulps %xmm6, %xmm2, %xmm2 vaddps %xmm5, %xmm2, %xmm2 vxorps %xmm15, %xmm9, %xmm5 vmulps %xmm5, %xmm2, %xmm2 vmulps %xmm6, %xmm1, %xmm1 vaddps %xmm4, %xmm1, %xmm1 vxorps %xmm15, %xmm10, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vmulps %xmm6, %xmm0, %xmm0 vaddps %xmm3, %xmm0, %xmm0 vxorps %xmm15, %xmm11, %xmm3 vmulps %xmm3, %xmm0, %xmm0 vaddps %xmm2, %xmm9, %xmm3 vaddps %xmm1, %xmm10, %xmm4 vbroadcastss 0x1cbfcdf(%rip), %xmm10 # 0x1eeba20 vaddps %xmm0, %xmm11, %xmm5 vmovaps 0x20(%rsp), %xmm11 jmp 0x22bab8 xorl %eax, %eax cmpq $0x8, %rbp je 0x22bd9a vmovaps 0x150(%rsp), %xmm0 vcmpnleps %xmm11, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r15 jae 0x22bd8a testb %cl, %cl je 0x22beb9 testb $0x8, %bpl je 0x22b9fe jmp 0x22bda1 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm11, (%rbx) addq $0x10, %rbx xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x22bd76 cmpq $-0x8, %rbp je 0x22bec7 movq %r11, 0x8(%rsp) movq %r8, 0x10(%rsp) vmovaps 0x150(%rsp), %xmm0 vcmpnleps %xmm11, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22bf9a andq $-0x10, %rbp vmovaps 0x30(%rsp), %xmm1 movzbl (%rbp), %eax movq %rsi, 0x18(%rsp) movq 0x8(%rsi), %rcx vmovaps 0x160(%rsp), %xmm0 vmovaps %xmm0, 0x70(%rsp) vmovaps %xmm1, 0x50(%rsp) vmovmskps %xmm1, %edx xorl $0xf, %edx je 0x22be53 movzbl %dl, %r14d shll $0x6, %eax addq %rcx, %rax addq $0x18, %rax movq %rax, 0x20(%rsp) bsfq %r14, %r13 leaq 0x170(%rsp), %rdi movq %r9, %rsi movq %r13, %rdx movq %r10, %rcx movq %rbp, %r8 movq 0x20(%rsp), %rax callq *(%rax) testb %al, %al je 0x22be40 orl $-0x1, 0x70(%rsp,%r13,4) leaq -0x1(%r14), %rax andq %rax, %r14 movq 0x48(%rsp), %r10 movq 0x40(%rsp), %r9 jne 0x22be17 vmovaps 0x50(%rsp), %xmm0 vorps 0x70(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x30(%rsp) vtestps 0x1cbffb2(%rip), %xmm0 # 0x1eebe20 jb 0x22be95 vmovaps 0x150(%rsp), %xmm1 vbroadcastss 0x1cc0d02(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x150(%rsp) xorl %eax, %eax jmp 0x22be98 pushq $0x3 popq %rax movq 0x18(%rsp), %rsi movq 0x10(%rsp), %r8 movq 0x8(%rsp), %r11 vbroadcastss 0x1cf5010(%rip), %xmm15 # 0x1f20ec0 vbroadcastss 0x1cbfb67(%rip), %xmm10 # 0x1eeba20 cmpl $0x3, %eax jne 0x22b9a2 jmp 0x22bfa2 pushq $0x3 jmp 0x22becd pushq $0x2 popq %rax jmp 0x22beb9 vmovaps %xmm11, 0x20(%rsp) movq %r11, 0x8(%rsp) movq %rsi, 0x18(%rsp) movq %r8, 0x10(%rsp) bsfq %r14, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, 0x50(%rsp) leaq 0x170(%rsp), %r8 pushq %r10 leaq 0x88(%rsp), %rax pushq %rax callq 0x27004e popq %rcx popq %rdx testb %al, %al je 0x22bf1f movq 0x50(%rsp), %rax orl $-0x1, 0x30(%rsp,%rax,4) leaq -0x1(%r14), %rax andq %rax, %r14 movq 0x48(%rsp), %r10 movq 0x40(%rsp), %r9 movq 0x18(%rsp), %rsi movq 0x10(%rsp), %r8 jne 0x22bee5 vmovaps 0x30(%rsp), %xmm0 vtestps 0x1cbfed5(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0x8(%rsp), %r11 vbroadcastss 0x1cf4f64(%rip), %xmm15 # 0x1f20ec0 vbroadcastss 0x1cbfabb(%rip), %xmm10 # 0x1eeba20 vmovaps 0x20(%rsp), %xmm11 jb 0x22b9e7 vmovaps 0x150(%rsp), %xmm1 vbroadcastss 0x1cc0c01(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x150(%rsp) pushq $0x2 popq %rax jmp 0x22b9e7 pushq $0x2 popq %rax jmp 0x22be9d vmovaps 0x60(%rsp), %xmm0 vandps 0x30(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cc0bcd(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r11) addq $0x18e8, %rsp # imm = 0x18E8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16781328, false, embree::avx::VirtualCurveIntersectorK<4>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1b8, %rsp # imm = 0x1B8 movq (%rsi), %rbx cmpq $0x8, 0x70(%rbx) je 0x22c372 vmovdqa (%rdi), %xmm3 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd %xmm0, %xmm3, %xmm0 vmovmskps %xmm0, %ebp testl %ebp, %ebp je 0x22c372 movq %rcx, %r13 movq %rdx, %r14 movq %rsi, %r12 leaq 0x10(%rsp), %rax vmovaps 0x40(%rdx), %xmm2 vmovaps 0x50(%rdx), %xmm1 vmovaps 0x60(%rdx), %xmm0 vmulps %xmm0, %xmm0, %xmm4 vmulps %xmm1, %xmm1, %xmm5 vaddps %xmm5, %xmm4, %xmm4 vmulps %xmm2, %xmm2, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vrsqrtps %xmm4, %xmm5 vbroadcastss 0x1cc06d5(%rip), %xmm6 # 0x1eec718 vmulps %xmm6, %xmm5, %xmm6 vbroadcastss 0x1cc06cc(%rip), %xmm7 # 0x1eec71c vmulps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm5, %xmm4 vmulps %xmm5, %xmm5, %xmm5 vmulps %xmm4, %xmm5, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vmovaps %xmm4, -0x10(%rax) vbroadcastss 0x1cf4e4e(%rip), %xmm4 # 0x1f20ec0 vxorps %xmm5, %xmm5, %xmm5 vmovss 0x1cc069a(%rip), %xmm6 # 0x1eec718 vmovss 0x1cc0afa(%rip), %xmm7 # 0x1eecb80 movq %rbp, %rcx bsfq %rcx, %rdx vmovss 0x40(%r14,%rdx,4), %xmm8 vinsertps $0x1c, 0x50(%r14,%rdx,4), %xmm8, %xmm8 # xmm8 = xmm8[0],mem[0],zero,zero vinsertps $0x28, 0x60(%r14,%rdx,4), %xmm8, %xmm9 # xmm9 = xmm8[0,1],mem[0],zero vbroadcastss (%rsp,%rdx,4), %xmm8 vmulps %xmm8, %xmm9, %xmm9 vshufpd $0x1, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,0] vmovshdup %xmm9, %xmm11 # xmm11 = xmm9[1,1,3,3] vxorps %xmm4, %xmm11, %xmm11 vunpckhps %xmm5, %xmm9, %xmm12 # xmm12 = xmm9[2],xmm5[2],xmm9[3],xmm5[3] vmovss %xmm11, %xmm5, %xmm11 # xmm11 = xmm11[0],xmm5[1,2,3] vshufps $0x41, %xmm11, %xmm12, %xmm11 # xmm11 = xmm12[1,0],xmm11[0,1] vxorpd %xmm4, %xmm10, %xmm10 vinsertps $0x2a, %xmm9, %xmm10, %xmm10 # xmm10 = xmm10[0],zero,xmm9[0],zero vdpps $0x7f, %xmm11, %xmm11, %xmm12 vdpps $0x7f, %xmm10, %xmm10, %xmm13 vcmpltps %xmm12, %xmm13, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vblendvps %xmm12, %xmm11, %xmm10, %xmm10 vdpps $0x7f, %xmm10, %xmm10, %xmm11 vrsqrtss %xmm11, %xmm11, %xmm12 vmulss %xmm6, %xmm12, %xmm13 vmulss %xmm7, %xmm11, %xmm11 vmulss %xmm12, %xmm11, %xmm11 vmulss %xmm12, %xmm12, %xmm12 vmulss %xmm12, %xmm11, %xmm11 vsubss %xmm11, %xmm13, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vmulps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm11 # xmm11 = xmm10[1,2,0,3] vshufps $0xc9, %xmm9, %xmm9, %xmm12 # xmm12 = xmm9[1,2,0,3] vmulps %xmm10, %xmm12, %xmm12 vmulps %xmm11, %xmm9, %xmm11 vsubps %xmm12, %xmm11, %xmm11 vshufps $0xc9, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[1,2,0,3] vdpps $0x7f, %xmm11, %xmm11, %xmm12 leaq -0x1(%rcx), %rdi imulq $0x30, %rdx, %rdx vrsqrtss %xmm12, %xmm12, %xmm13 vmulss %xmm6, %xmm13, %xmm14 vmulss %xmm7, %xmm12, %xmm12 vmulss %xmm13, %xmm12, %xmm12 vmulss %xmm13, %xmm13, %xmm13 vmulss %xmm13, %xmm12, %xmm12 vsubss %xmm12, %xmm14, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vmulps %xmm11, %xmm12, %xmm11 vmulps %xmm9, %xmm8, %xmm8 vunpcklps %xmm8, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm8[0],xmm10[1],xmm8[1] vunpckhps %xmm8, %xmm10, %xmm8 # xmm8 = xmm10[2],xmm8[2],xmm10[3],xmm8[3] vunpcklps %xmm5, %xmm11, %xmm10 # xmm10 = xmm11[0],xmm5[0],xmm11[1],xmm5[1] vunpckhps %xmm5, %xmm11, %xmm11 # xmm11 = xmm11[2],xmm5[2],xmm11[3],xmm5[3] vunpcklps %xmm11, %xmm8, %xmm8 # xmm8 = xmm8[0],xmm11[0],xmm8[1],xmm11[1] vunpcklps %xmm10, %xmm9, %xmm11 # xmm11 = xmm9[0],xmm10[0],xmm9[1],xmm10[1] vunpckhps %xmm10, %xmm9, %xmm9 # xmm9 = xmm9[2],xmm10[2],xmm9[3],xmm10[3] vmovaps %xmm11, (%rax,%rdx) vmovaps %xmm9, 0x10(%rax,%rdx) vmovaps %xmm8, 0x20(%rax,%rdx) andq %rdi, %rcx jne 0x22c089 vpcmpeqd %xmm4, %xmm4, %xmm4 vpcmpeqd %xmm4, %xmm3, %xmm3 vmovaps (%r14), %xmm4 leaq 0xd0(%rsp), %rax vmovaps %xmm4, (%rax) vmovaps 0x10(%r14), %xmm4 vmovaps %xmm4, 0x10(%rax) vmovaps 0x20(%r14), %xmm4 vmovaps %xmm4, 0x20(%rax) vmovaps %xmm2, 0x30(%rax) vmovaps %xmm1, 0x40(%rax) vmovaps %xmm0, 0x50(%rax) vbroadcastss 0x1cf4cbb(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm2, %xmm5 vbroadcastss 0x1cc4dd2(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vandps %xmm4, %xmm1, %xmm5 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vandps %xmm4, %xmm0, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vblendvps %xmm4, %xmm6, %xmm0, %xmm0 vrcpps %xmm2, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cc04c4(%rip), %xmm5 # 0x1eec714 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm2, %xmm4, %xmm2 vrcpps %xmm1, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vsubps %xmm1, %xmm5, %xmm1 vmulps %xmm1, %xmm4, %xmm1 vaddps %xmm1, %xmm4, %xmm1 vrcpps %xmm0, %xmm4 vmulps %xmm4, %xmm0, %xmm0 vsubps %xmm0, %xmm5, %xmm0 vmulps %xmm0, %xmm4, %xmm0 vaddps %xmm0, %xmm4, %xmm0 vmovaps %xmm2, 0x60(%rax) vmovaps %xmm1, 0x70(%rax) vmovaps %xmm0, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d2e6bc(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm2, %xmm2 vmovaps %xmm2, 0x90(%rax) vcmpnltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d286de(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d2e69d(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm0, %xmm0 vbroadcastss 0x1d2e685(%rip), %xmm1 # 0x1f5a96c vbroadcastss 0x1d2e680(%rip), %xmm2 # 0x1f5a970 vblendvps %xmm0, %xmm1, %xmm2, %xmm0 vmovaps %xmm0, 0xb0(%rax) vmovaps 0x30(%r14), %xmm0 vmovaps 0x80(%r14), %xmm1 vmaxps %xmm4, %xmm0, %xmm0 vmaxps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1cbf702(%rip), %xmm2 # 0x1eeba20 vblendvps %xmm3, %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0xc0(%rax) vbroadcastss 0x1cc084f(%rip), %xmm0 # 0x1eecb84 vblendvps %xmm3, %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %r15 movq 0x70(%rbx), %rdx movq %r12, %rdi movq %rbx, %rsi movq %rsp, %r8 movq %r14, %r9 pushq %r13 leaq 0xd8(%rsp), %rax pushq %rax callq 0x2705b4 popq %rax popq %rcx andq %r15, %rbp jne 0x22c343 addq $0x1b8, %rsp # imm = 0x1B8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16781328, false, embree::avx::VirtualCurveIntersectorK<4>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1938, %rsp # imm = 0x1938 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x22cfa4 movq %rdx, %r14 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm1 vmovaps 0x80(%rdx), %xmm0 vxorps %xmm4, %xmm4, %xmm4 vcmpnltps %xmm4, %xmm0, %xmm2 vandps %xmm1, %xmm2, %xmm1 vmovaps %xmm1, 0x70(%rsp) vmovmskps %xmm1, %eax testl %eax, %eax je 0x22cfa4 movq %rcx, %r10 leaq 0x80(%r14), %r9 movzbl %al, %eax leaq 0x1d0(%rsp), %rcx vmovaps 0x40(%r14), %xmm2 vmovaps 0x50(%r14), %xmm3 vmovaps 0x60(%r14), %xmm1 vmulps %xmm1, %xmm1, %xmm5 vmulps %xmm3, %xmm3, %xmm6 vaddps %xmm6, %xmm5, %xmm5 vmulps %xmm2, %xmm2, %xmm6 vaddps %xmm5, %xmm6, %xmm5 vrsqrtps %xmm5, %xmm6 vbroadcastss 0x1cc02fb(%rip), %xmm7 # 0x1eec718 vmulps %xmm7, %xmm6, %xmm7 vbroadcastss 0x1cc02f2(%rip), %xmm8 # 0x1eec71c vmulps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vmulps %xmm6, %xmm6, %xmm6 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm7, %xmm5 vmovaps %xmm5, -0x10(%rcx) vbroadcastss 0x1cf4a74(%rip), %xmm5 # 0x1f20ec0 vmovss 0x1cc02c4(%rip), %xmm6 # 0x1eec718 vmovss 0x1cc0724(%rip), %xmm7 # 0x1eecb80 bsfq %rax, %rdx vmovss 0x40(%r14,%rdx,4), %xmm8 vinsertps $0x1c, 0x50(%r14,%rdx,4), %xmm8, %xmm8 # xmm8 = xmm8[0],mem[0],zero,zero vinsertps $0x28, 0x60(%r14,%rdx,4), %xmm8, %xmm9 # xmm9 = xmm8[0,1],mem[0],zero vbroadcastss 0x1c0(%rsp,%rdx,4), %xmm8 vmulps %xmm8, %xmm9, %xmm9 vshufpd $0x1, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,0] vmovshdup %xmm9, %xmm11 # xmm11 = xmm9[1,1,3,3] vxorps %xmm5, %xmm11, %xmm11 vunpckhps %xmm4, %xmm9, %xmm12 # xmm12 = xmm9[2],xmm4[2],xmm9[3],xmm4[3] vmovss %xmm11, %xmm4, %xmm11 # xmm11 = xmm11[0],xmm4[1,2,3] vshufps $0x41, %xmm11, %xmm12, %xmm11 # xmm11 = xmm12[1,0],xmm11[0,1] vxorpd %xmm5, %xmm10, %xmm10 vinsertps $0x2a, %xmm9, %xmm10, %xmm10 # xmm10 = xmm10[0],zero,xmm9[0],zero vdpps $0x7f, %xmm11, %xmm11, %xmm12 vdpps $0x7f, %xmm10, %xmm10, %xmm13 vcmpltps %xmm12, %xmm13, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vblendvps %xmm12, %xmm11, %xmm10, %xmm10 vdpps $0x7f, %xmm10, %xmm10, %xmm11 vrsqrtss %xmm11, %xmm11, %xmm12 vmulss %xmm6, %xmm12, %xmm13 vmulss %xmm7, %xmm11, %xmm11 vmulss %xmm12, %xmm11, %xmm11 vmulss %xmm12, %xmm12, %xmm12 vmulss %xmm12, %xmm11, %xmm11 vsubss %xmm11, %xmm13, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vmulps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm11 # xmm11 = xmm10[1,2,0,3] vshufps $0xc9, %xmm9, %xmm9, %xmm12 # xmm12 = xmm9[1,2,0,3] vmulps %xmm10, %xmm12, %xmm12 vmulps %xmm11, %xmm9, %xmm11 vsubps %xmm12, %xmm11, %xmm11 vshufps $0xc9, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[1,2,0,3] vdpps $0x7f, %xmm11, %xmm11, %xmm12 leaq -0x1(%rax), %rdi imulq $0x30, %rdx, %rdx vrsqrtss %xmm12, %xmm12, %xmm13 vmulss %xmm6, %xmm13, %xmm14 vmulss %xmm7, %xmm12, %xmm12 vmulss %xmm13, %xmm12, %xmm12 vmulss %xmm13, %xmm13, %xmm13 vmulss %xmm13, %xmm12, %xmm12 vsubss %xmm12, %xmm14, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vmulps %xmm11, %xmm12, %xmm11 vmulps %xmm9, %xmm8, %xmm8 vunpcklps %xmm8, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm8[0],xmm10[1],xmm8[1] vunpckhps %xmm8, %xmm10, %xmm8 # xmm8 = xmm10[2],xmm8[2],xmm10[3],xmm8[3] vunpcklps %xmm4, %xmm11, %xmm10 # xmm10 = xmm11[0],xmm4[0],xmm11[1],xmm4[1] vunpckhps %xmm4, %xmm11, %xmm11 # xmm11 = xmm11[2],xmm4[2],xmm11[3],xmm4[3] vunpcklps %xmm11, %xmm8, %xmm8 # xmm8 = xmm8[0],xmm11[0],xmm8[1],xmm11[1] vunpcklps %xmm10, %xmm9, %xmm11 # xmm11 = xmm9[0],xmm10[0],xmm9[1],xmm10[1] vunpckhps %xmm10, %xmm9, %xmm9 # xmm9 = xmm9[2],xmm10[2],xmm9[3],xmm10[3] vmovaps %xmm11, (%rcx,%rdx) vmovaps %xmm9, 0x10(%rcx,%rdx) vmovaps %xmm8, 0x20(%rcx,%rdx) andq %rdi, %rax jne 0x22c45c vbroadcastss 0x1cf4922(%rip), %xmm8 # 0x1f20ec4 vbroadcastss 0x1cc4a3d(%rip), %xmm9 # 0x1ef0fe8 vandps %xmm3, %xmm8, %xmm4 vcmpltps %xmm9, %xmm4, %xmm4 vblendvps %xmm4, %xmm9, %xmm3, %xmm4 vbroadcastss 0x1cc0150(%rip), %xmm10 # 0x1eec714 vrcpps %xmm4, %xmm5 vmulps %xmm5, %xmm4, %xmm4 vsubps %xmm4, %xmm10, %xmm4 vmulps %xmm4, %xmm5, %xmm4 vaddps %xmm4, %xmm5, %xmm4 vxorps %xmm11, %xmm11, %xmm11 vcmpnltps %xmm11, %xmm4, %xmm5 vbroadcastss 0x1d283b4(%rip), %xmm6 # 0x1f549a0 vbroadcastss 0x1d2e373(%rip), %xmm7 # 0x1f5a968 vblendvps %xmm5, %xmm6, %xmm7, %xmm5 vmovaps (%r14), %xmm6 vmovaps %xmm6, 0x90(%rsp) vmovaps 0x10(%r14), %xmm6 vmovaps %xmm6, 0xa0(%rsp) vmovdqa 0x20(%r14), %xmm6 vmovdqa %xmm6, 0xb0(%rsp) vmovaps %xmm2, 0xc0(%rsp) vmovaps %xmm3, 0xd0(%rsp) vmovaps %xmm1, 0xe0(%rsp) vandps %xmm2, %xmm8, %xmm3 vcmpltps %xmm9, %xmm3, %xmm3 vblendvps %xmm3, %xmm9, %xmm2, %xmm2 vandps %xmm1, %xmm8, %xmm3 vcmpltps %xmm9, %xmm3, %xmm3 vblendvps %xmm3, %xmm9, %xmm1, %xmm1 vrcpps %xmm2, %xmm3 vmulps %xmm3, %xmm2, %xmm2 vsubps %xmm2, %xmm10, %xmm2 vmulps %xmm2, %xmm3, %xmm2 vaddps %xmm2, %xmm3, %xmm2 vrcpps %xmm1, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vsubps %xmm1, %xmm10, %xmm1 vmulps %xmm1, %xmm3, %xmm1 vaddps %xmm1, %xmm3, %xmm1 vmovaps %xmm2, 0xf0(%rsp) vmovaps %xmm4, 0x100(%rsp) vmovaps %xmm1, 0x110(%rsp) vcmpltps %xmm11, %xmm2, %xmm2 vbroadcastss 0x1d2e2b0(%rip), %xmm3 # 0x1f5a964 vandps %xmm3, %xmm2, %xmm2 vmovaps %xmm2, 0x120(%rsp) vmovaps %xmm5, 0x130(%rsp) vcmpnltps %xmm11, %xmm1, %xmm1 vbroadcastss 0x1d2e293(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d2e28e(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0x140(%rsp) vmovaps 0x30(%r14), %xmm1 vmaxps %xmm11, %xmm1, %xmm1 vmaxps %xmm11, %xmm0, %xmm2 vbroadcastss 0x1cbf316(%rip), %xmm14 # 0x1eeba20 vmovaps 0x70(%rsp), %xmm3 vblendvps %xmm3, %xmm1, %xmm14, %xmm0 vmovaps %xmm0, 0x150(%rsp) vbroadcastss 0x1cc045c(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm3, %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x160(%rsp) vpcmpeqd %xmm1, %xmm1, %xmm1 vxorps %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0x50(%rsp) cmpq $0x0, 0x8(%r10) je 0x22c760 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r11d, %r11d cmpb $0x1, %al adcq $0x2, %r11 jmp 0x22c764 pushq $0x3 popq %r11 leaq 0x2a0(%rsp), %r12 movq $-0x8, -0x10(%r12) leaq 0xa40(%rsp), %rbx vmovaps %xmm14, -0x20(%rbx) movq 0x70(%r8), %rax movq %rax, -0x8(%r12) vmovaps %xmm0, -0x10(%rbx) leaq 0x1f237e9(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x170(%rsp) movq %r10, 0x68(%rsp) addq $-0x10, %rbx movq -0x8(%r12), %rbp addq $-0x8, %r12 cmpq $-0x8, %rbp je 0x22ceb2 vmovaps (%rbx), %xmm15 vcmpltps 0x160(%rsp), %xmm15, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22ceb6 movzbl %al, %r15d popcntl %r15d, %r13d xorl %eax, %eax cmpq %r11, %r13 jbe 0x22cebb cmpq %r11, %r13 jbe 0x22cea4 pushq $0x8 popq %r15 testb $0x8, %bpl jne 0x22cd89 vmovaps 0x160(%rsp), %xmm0 movq %rbp, %rax andq $-0x10, %rax movl %ebp, %ecx andl $0x7, %ecx vcmpnleps %xmm15, %xmm0, %xmm0 vmovaps %xmm0, 0x40(%rsp) xorl %edx, %edx movq %r15, %rbp vmovaps %xmm14, %xmm15 movq (%rax,%rdx,8), %rdi cmpq $0x8, %rdi je 0x22cd28 cmpl $0x6, %ecx je 0x22c84c cmpl $0x1, %ecx jne 0x22c9d1 vbroadcastss 0x80(%rax,%rdx,4), %xmm0 vbroadcastss 0x20(%rax,%rdx,4), %xmm2 vmovaps 0x70(%r14), %xmm1 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm2, %xmm0, %xmm0 vbroadcastss 0xa0(%rax,%rdx,4), %xmm2 vbroadcastss 0x40(%rax,%rdx,4), %xmm3 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0xc0(%rax,%rdx,4), %xmm3 vbroadcastss 0x60(%rax,%rdx,4), %xmm4 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0x90(%rax,%rdx,4), %xmm4 vbroadcastss 0x30(%rax,%rdx,4), %xmm5 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0xb0(%rax,%rdx,4), %xmm5 vbroadcastss 0x50(%rax,%rdx,4), %xmm6 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0xd0(%rax,%rdx,4), %xmm6 vbroadcastss 0x70(%rax,%rdx,4), %xmm7 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vmovaps 0x90(%rsp), %xmm7 vmovaps 0xa0(%rsp), %xmm8 vmovaps 0xb0(%rsp), %xmm9 vmovaps 0xf0(%rsp), %xmm10 vsubps %xmm7, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vsubps %xmm8, %xmm2, %xmm2 vmovaps 0x100(%rsp), %xmm11 vmulps %xmm2, %xmm11, %xmm12 vsubps %xmm9, %xmm3, %xmm2 vmovaps 0x110(%rsp), %xmm3 vmulps %xmm3, %xmm2, %xmm13 vsubps %xmm7, %xmm4, %xmm2 vmulps %xmm2, %xmm10, %xmm4 vsubps %xmm8, %xmm5, %xmm2 vmulps %xmm2, %xmm11, %xmm5 vsubps %xmm9, %xmm6, %xmm2 vmulps %xmm3, %xmm2, %xmm3 vpminsd %xmm4, %xmm0, %xmm2 vpminsd %xmm5, %xmm12, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpminsd %xmm3, %xmm13, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpmaxsd %xmm4, %xmm0, %xmm0 vpmaxsd %xmm5, %xmm12, %xmm4 vpminsd %xmm4, %xmm0, %xmm4 vpmaxsd %xmm3, %xmm13, %xmm3 vpmaxsd 0x150(%rsp), %xmm2, %xmm0 vpminsd %xmm3, %xmm4, %xmm3 vpminsd 0x160(%rsp), %xmm3, %xmm3 vcmpleps %xmm3, %xmm0, %xmm0 cmpl $0x6, %ecx jne 0x22ccf3 vbroadcastss 0xe0(%rax,%rdx,4), %xmm3 vcmpleps %xmm1, %xmm3, %xmm3 vbroadcastss 0xf0(%rax,%rdx,4), %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vandps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm0 jmp 0x22ccf3 vbroadcastss 0x50(%rax,%rdx,4), %xmm8 vmovaps %xmm8, 0x180(%rsp) vbroadcastss 0x60(%rax,%rdx,4), %xmm4 vbroadcastss 0x70(%rax,%rdx,4), %xmm5 vbroadcastss 0x80(%rax,%rdx,4), %xmm11 vbroadcastss 0x90(%rax,%rdx,4), %xmm10 vbroadcastss 0xe0(%rax,%rdx,4), %xmm0 vbroadcastss 0xf0(%rax,%rdx,4), %xmm1 vbroadcastss 0x100(%rax,%rdx,4), %xmm2 vbroadcastss 0x110(%rax,%rdx,4), %xmm6 vbroadcastss 0x120(%rax,%rdx,4), %xmm7 vbroadcastss 0x130(%rax,%rdx,4), %xmm9 vmovaps 0x70(%r14), %xmm12 vbroadcastss 0x1cbfcc6(%rip), %xmm3 # 0x1eec714 vsubps %xmm12, %xmm3, %xmm13 vmulps %xmm0, %xmm12, %xmm0 vmulps %xmm1, %xmm12, %xmm1 vmulps %xmm2, %xmm12, %xmm2 vmulps 0x1cbefa9(%rip), %xmm13, %xmm14 # 0x1eeba10 vaddps %xmm0, %xmm14, %xmm0 vmovaps %xmm0, 0x1a0(%rsp) vaddps %xmm1, %xmm14, %xmm0 vmovaps %xmm0, 0x10(%rsp) vaddps %xmm2, %xmm14, %xmm0 vmovaps %xmm0, 0x20(%rsp) vmulps %xmm6, %xmm12, %xmm0 vmulps %xmm7, %xmm12, %xmm1 vmulps %xmm9, %xmm12, %xmm2 vaddps %xmm0, %xmm13, %xmm0 vmovaps %xmm0, 0x190(%rsp) vaddps %xmm1, %xmm13, %xmm0 vmovaps %xmm0, 0x1b0(%rsp) vaddps %xmm2, %xmm13, %xmm0 vmovaps %xmm0, 0x30(%rsp) vmovaps 0xd0(%rsp), %xmm0 vmovaps 0xe0(%rsp), %xmm1 vmulps %xmm1, %xmm11, %xmm2 vmulps %xmm1, %xmm10, %xmm12 vmulps %xmm0, %xmm8, %xmm13 vaddps %xmm2, %xmm13, %xmm2 vmulps %xmm0, %xmm4, %xmm13 vmovaps %xmm4, %xmm9 vaddps %xmm12, %xmm13, %xmm14 vmovaps %xmm15, %xmm4 vbroadcastss 0xa0(%rax,%rdx,4), %xmm15 vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm0, %xmm5, %xmm0 vaddps %xmm1, %xmm0, %xmm0 vbroadcastss 0x20(%rax,%rdx,4), %xmm12 vmovaps 0xc0(%rsp), %xmm1 vmulps %xmm1, %xmm12, %xmm13 vaddps %xmm2, %xmm13, %xmm2 vbroadcastss 0x30(%rax,%rdx,4), %xmm13 vmulps %xmm1, %xmm13, %xmm6 vaddps %xmm6, %xmm14, %xmm6 vbroadcastss 0x40(%rax,%rdx,4), %xmm14 vmulps %xmm1, %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cf4383(%rip), %xmm7 # 0x1f20ec4 vandps %xmm7, %xmm2, %xmm1 vbroadcastss 0x1cc449a(%rip), %xmm8 # 0x1ef0fe8 vcmpltps %xmm8, %xmm1, %xmm1 vblendvps %xmm1, %xmm8, %xmm2, %xmm1 vandps %xmm7, %xmm6, %xmm2 vcmpltps %xmm8, %xmm2, %xmm2 vblendvps %xmm2, %xmm8, %xmm6, %xmm6 vandps %xmm7, %xmm0, %xmm2 vcmpltps %xmm8, %xmm2, %xmm2 vblendvps %xmm2, %xmm8, %xmm0, %xmm0 vrcpps %xmm1, %xmm2 vmulps %xmm1, %xmm2, %xmm1 vsubps %xmm1, %xmm3, %xmm1 vmulps %xmm1, %xmm2, %xmm1 vaddps %xmm1, %xmm2, %xmm2 vrcpps %xmm6, %xmm1 vmulps %xmm6, %xmm1, %xmm6 vsubps %xmm6, %xmm3, %xmm6 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm6, %xmm1, %xmm1 vrcpps %xmm0, %xmm6 vmulps %xmm0, %xmm6, %xmm0 vsubps %xmm0, %xmm3, %xmm0 vmulps %xmm0, %xmm6, %xmm0 vaddps %xmm0, %xmm6, %xmm0 vmovaps 0xb0(%rsp), %xmm6 vmulps %xmm6, %xmm11, %xmm11 vbroadcastss 0xb0(%rax,%rdx,4), %xmm7 vaddps %xmm7, %xmm11, %xmm7 vmulps %xmm6, %xmm10, %xmm10 vbroadcastss 0xc0(%rax,%rdx,4), %xmm11 vaddps %xmm10, %xmm11, %xmm10 vmulps %xmm6, %xmm15, %xmm6 vmovaps %xmm4, %xmm15 vbroadcastss 0xd0(%rax,%rdx,4), %xmm11 vaddps %xmm6, %xmm11, %xmm6 vmovaps 0xa0(%rsp), %xmm11 vmulps 0x180(%rsp), %xmm11, %xmm4 vaddps %xmm7, %xmm4, %xmm4 vmulps %xmm11, %xmm9, %xmm3 vaddps %xmm3, %xmm10, %xmm3 vmulps %xmm5, %xmm11, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vmovaps 0x90(%rsp), %xmm6 vmulps %xmm6, %xmm12, %xmm7 vaddps %xmm4, %xmm7, %xmm4 vmulps %xmm6, %xmm13, %xmm7 vaddps %xmm3, %xmm7, %xmm3 vmulps %xmm6, %xmm14, %xmm6 vbroadcastss 0x1cbedd9(%rip), %xmm14 # 0x1eeba20 vaddps %xmm5, %xmm6, %xmm5 vmovaps 0x1a0(%rsp), %xmm6 vsubps %xmm4, %xmm6, %xmm6 vmovaps 0x190(%rsp), %xmm7 vsubps %xmm4, %xmm7, %xmm4 vmulps %xmm6, %xmm2, %xmm6 vmulps %xmm4, %xmm2, %xmm4 vmovaps 0x10(%rsp), %xmm2 vsubps %xmm3, %xmm2, %xmm2 vmovaps 0x1b0(%rsp), %xmm7 vsubps %xmm3, %xmm7, %xmm3 vmulps %xmm2, %xmm1, %xmm7 vmulps %xmm3, %xmm1, %xmm1 vmovaps 0x20(%rsp), %xmm2 vsubps %xmm5, %xmm2, %xmm2 vmulps %xmm2, %xmm0, %xmm3 vmovaps 0x30(%rsp), %xmm2 vsubps %xmm5, %xmm2, %xmm2 vmulps %xmm2, %xmm0, %xmm0 vpminsd %xmm4, %xmm6, %xmm2 vpminsd %xmm1, %xmm7, %xmm5 vpmaxsd %xmm5, %xmm2, %xmm2 vpminsd %xmm0, %xmm3, %xmm5 vpmaxsd %xmm5, %xmm2, %xmm2 vpmaxsd %xmm4, %xmm6, %xmm4 vpmaxsd %xmm1, %xmm7, %xmm1 vpminsd %xmm1, %xmm4, %xmm1 vpmaxsd %xmm0, %xmm3, %xmm0 vpmaxsd 0x150(%rsp), %xmm2, %xmm3 vpminsd %xmm0, %xmm1, %xmm0 vpminsd 0x160(%rsp), %xmm0, %xmm0 vcmpleps %xmm0, %xmm3, %xmm0 vandps 0x40(%rsp), %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22cd28 vblendvps %xmm0, %xmm2, %xmm14, %xmm0 cmpq $0x8, %rbp je 0x22cd21 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm15, (%rbx) addq $0x10, %rbx vmovaps %xmm0, %xmm15 movq %rdi, %rbp cmpq $0x8, %rdi je 0x22cd3f leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x22c830 xorl %eax, %eax cmpq $0x8, %rbp je 0x22cd82 vmovaps 0x160(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r11 jae 0x22cd72 testb %cl, %cl jne 0x22c7fb jmp 0x22cea4 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm15, (%rbx) addq $0x10, %rbx xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x22cd65 cmpq $-0x8, %rbp je 0x22ceb2 movq %r11, 0x10(%rsp) movq %r9, 0x20(%rsp) movq %r8, (%rsp) vmovaps 0x160(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22cf82 andq $-0x10, %rbp vmovaps 0x50(%rsp), %xmm1 movzbl (%rbp), %eax movq %rsi, 0x8(%rsp) movq 0x8(%rsi), %rcx vmovaps 0x170(%rsp), %xmm0 vmovaps %xmm0, 0x80(%rsp) vmovaps %xmm1, 0x30(%rsp) vmovmskps %xmm1, %edx xorl $0xf, %edx je 0x22ce40 movzbl %dl, %r15d shll $0x6, %eax addq %rcx, %rax addq $0x18, %rax movq %rax, 0x40(%rsp) bsfq %r15, %rdx leaq 0x1c0(%rsp), %rdi movq %r14, %rsi movq %rdx, %r13 movq %r10, %rcx movq %rbp, %r8 movq 0x40(%rsp), %rax callq *(%rax) testb %al, %al je 0x22ce32 orl $-0x1, 0x80(%rsp,%r13,4) leaq -0x1(%r15), %rax andq %rax, %r15 movq 0x68(%rsp), %r10 jne 0x22ce06 vmovaps 0x30(%rsp), %xmm0 vorps 0x80(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x50(%rsp) vtestps 0x1cbefc2(%rip), %xmm0 # 0x1eebe20 jb 0x22ce85 vmovaps 0x160(%rsp), %xmm1 vbroadcastss 0x1cbfd12(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x160(%rsp) xorl %eax, %eax jmp 0x22ce88 pushq $0x3 popq %rax movq 0x8(%rsp), %rsi movq (%rsp), %r8 movq 0x20(%rsp), %r9 vbroadcastss 0x1cbeb81(%rip), %xmm14 # 0x1eeba20 movq 0x10(%rsp), %r11 cmpl $0x3, %eax jne 0x22c7a9 jmp 0x22cf8a pushq $0x3 jmp 0x22ceb8 pushq $0x2 popq %rax jmp 0x22cea4 vmovaps %xmm15, 0x30(%rsp) movq %r11, 0x10(%rsp) movq %r9, 0x20(%rsp) movq %rsi, 0x8(%rsp) movq %r8, (%rsp) bsfq %r15, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, 0x40(%rsp) leaq 0x1c0(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x98(%rsp), %rax pushq %rax callq 0x270de2 popq %rcx popq %rdx testb %al, %al je 0x22cf11 movq 0x40(%rsp), %rax orl $-0x1, 0x50(%rsp,%rax,4) leaq -0x1(%r15), %rax andq %rax, %r15 movq 0x68(%rsp), %r10 movq 0x8(%rsp), %rsi movq (%rsp), %r8 jne 0x22ced4 vmovaps 0x50(%rsp), %xmm0 vtestps 0x1cbeee9(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0x20(%rsp), %r9 vbroadcastss 0x1cbead8(%rip), %xmm14 # 0x1eeba20 movq 0x10(%rsp), %r11 vmovaps 0x30(%rsp), %xmm15 jb 0x22c7ee vmovaps 0x160(%rsp), %xmm1 vbroadcastss 0x1cbfc19(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x160(%rsp) pushq $0x2 popq %rax jmp 0x22c7ee pushq $0x2 popq %rax jmp 0x22ce8d vmovaps 0x70(%rsp), %xmm0 vandps 0x50(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cbfbe5(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r9) addq $0x1938, %rsp # imm = 0x1938 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 257, true, embree::avx::VirtualCurveIntersectorK<4>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1b8, %rsp # imm = 0x1B8 movq (%rsi), %rbx cmpq $0x8, 0x70(%rbx) je 0x22d333 vmovdqa (%rdi), %xmm3 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd %xmm0, %xmm3, %xmm0 vmovmskps %xmm0, %ebp testl %ebp, %ebp je 0x22d333 movq %rcx, %r13 movq %rdx, %r14 movq %rsi, %r12 leaq 0x10(%rsp), %rax vmovaps 0x40(%rdx), %xmm2 vmovaps 0x50(%rdx), %xmm1 vmovaps 0x60(%rdx), %xmm0 vmulps %xmm0, %xmm0, %xmm4 vmulps %xmm1, %xmm1, %xmm5 vaddps %xmm5, %xmm4, %xmm4 vmulps %xmm2, %xmm2, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vrsqrtps %xmm4, %xmm5 vbroadcastss 0x1cbf6ed(%rip), %xmm6 # 0x1eec718 vmulps %xmm6, %xmm5, %xmm6 vbroadcastss 0x1cbf6e4(%rip), %xmm7 # 0x1eec71c vmulps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm5, %xmm4 vmulps %xmm5, %xmm5, %xmm5 vmulps %xmm4, %xmm5, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vmovaps %xmm4, -0x10(%rax) vbroadcastss 0x1cf3e66(%rip), %xmm4 # 0x1f20ec0 vxorps %xmm5, %xmm5, %xmm5 vmovss 0x1cbf6b2(%rip), %xmm6 # 0x1eec718 vmovss 0x1cbfb12(%rip), %xmm7 # 0x1eecb80 movq %rbp, %rcx bsfq %rcx, %rdx vmovss 0x40(%r14,%rdx,4), %xmm8 vinsertps $0x1c, 0x50(%r14,%rdx,4), %xmm8, %xmm8 # xmm8 = xmm8[0],mem[0],zero,zero vinsertps $0x28, 0x60(%r14,%rdx,4), %xmm8, %xmm9 # xmm9 = xmm8[0,1],mem[0],zero vbroadcastss (%rsp,%rdx,4), %xmm8 vmulps %xmm8, %xmm9, %xmm9 vshufpd $0x1, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,0] vmovshdup %xmm9, %xmm11 # xmm11 = xmm9[1,1,3,3] vxorps %xmm4, %xmm11, %xmm11 vunpckhps %xmm5, %xmm9, %xmm12 # xmm12 = xmm9[2],xmm5[2],xmm9[3],xmm5[3] vmovss %xmm11, %xmm5, %xmm11 # xmm11 = xmm11[0],xmm5[1,2,3] vshufps $0x41, %xmm11, %xmm12, %xmm11 # xmm11 = xmm12[1,0],xmm11[0,1] vxorpd %xmm4, %xmm10, %xmm10 vinsertps $0x2a, %xmm9, %xmm10, %xmm10 # xmm10 = xmm10[0],zero,xmm9[0],zero vdpps $0x7f, %xmm11, %xmm11, %xmm12 vdpps $0x7f, %xmm10, %xmm10, %xmm13 vcmpltps %xmm12, %xmm13, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vblendvps %xmm12, %xmm11, %xmm10, %xmm10 vdpps $0x7f, %xmm10, %xmm10, %xmm11 vrsqrtss %xmm11, %xmm11, %xmm12 vmulss %xmm6, %xmm12, %xmm13 vmulss %xmm7, %xmm11, %xmm11 vmulss %xmm12, %xmm11, %xmm11 vmulss %xmm12, %xmm12, %xmm12 vmulss %xmm12, %xmm11, %xmm11 vsubss %xmm11, %xmm13, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vmulps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm11 # xmm11 = xmm10[1,2,0,3] vshufps $0xc9, %xmm9, %xmm9, %xmm12 # xmm12 = xmm9[1,2,0,3] vmulps %xmm10, %xmm12, %xmm12 vmulps %xmm11, %xmm9, %xmm11 vsubps %xmm12, %xmm11, %xmm11 vshufps $0xc9, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[1,2,0,3] vdpps $0x7f, %xmm11, %xmm11, %xmm12 leaq -0x1(%rcx), %rdi imulq $0x30, %rdx, %rdx vrsqrtss %xmm12, %xmm12, %xmm13 vmulss %xmm6, %xmm13, %xmm14 vmulss %xmm7, %xmm12, %xmm12 vmulss %xmm13, %xmm12, %xmm12 vmulss %xmm13, %xmm13, %xmm13 vmulss %xmm13, %xmm12, %xmm12 vsubss %xmm12, %xmm14, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vmulps %xmm11, %xmm12, %xmm11 vmulps %xmm9, %xmm8, %xmm8 vunpcklps %xmm8, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm8[0],xmm10[1],xmm8[1] vunpckhps %xmm8, %xmm10, %xmm8 # xmm8 = xmm10[2],xmm8[2],xmm10[3],xmm8[3] vunpcklps %xmm5, %xmm11, %xmm10 # xmm10 = xmm11[0],xmm5[0],xmm11[1],xmm5[1] vunpckhps %xmm5, %xmm11, %xmm11 # xmm11 = xmm11[2],xmm5[2],xmm11[3],xmm5[3] vunpcklps %xmm11, %xmm8, %xmm8 # xmm8 = xmm8[0],xmm11[0],xmm8[1],xmm11[1] vunpcklps %xmm10, %xmm9, %xmm11 # xmm11 = xmm9[0],xmm10[0],xmm9[1],xmm10[1] vunpckhps %xmm10, %xmm9, %xmm9 # xmm9 = xmm9[2],xmm10[2],xmm9[3],xmm10[3] vmovaps %xmm11, (%rax,%rdx) vmovaps %xmm9, 0x10(%rax,%rdx) vmovaps %xmm8, 0x20(%rax,%rdx) andq %rdi, %rcx jne 0x22d071 vpcmpeqd %xmm4, %xmm4, %xmm4 vpcmpeqd %xmm4, %xmm3, %xmm3 vmovaps (%r14), %xmm4 leaq 0xd0(%rsp), %rax vmovaps %xmm4, (%rax) vmovaps 0x10(%r14), %xmm4 vmovaps %xmm4, 0x10(%rax) vmovaps 0x20(%r14), %xmm4 vmovaps %xmm4, 0x20(%rax) vmovaps %xmm2, 0x30(%rax) vmovaps %xmm1, 0x40(%rax) vmovaps %xmm0, 0x50(%rax) vbroadcastss 0x1cf3cd3(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm2, %xmm5 vbroadcastss 0x1cc3dea(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1cbf508(%rip), %xmm7 # 0x1eec714 vdivps %xmm2, %xmm7, %xmm2 vandps %xmm4, %xmm1, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vdivps %xmm1, %xmm7, %xmm1 vandps %xmm4, %xmm0, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vbroadcastss 0x1cf3d31(%rip), %xmm6 # 0x1f20f60 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vblendvps %xmm8, %xmm6, %xmm1, %xmm1 vdivps %xmm0, %xmm7, %xmm0 vblendvps %xmm4, %xmm6, %xmm0, %xmm0 vmovaps %xmm2, 0x60(%rax) vmovaps %xmm1, 0x70(%rax) vmovaps %xmm0, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d2d6fb(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm2, %xmm2 vmovaps %xmm2, 0x90(%rax) vcmpnltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d2771d(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d2d6dc(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm0, %xmm0 vbroadcastss 0x1d2d6c4(%rip), %xmm1 # 0x1f5a96c vbroadcastss 0x1d2d6bf(%rip), %xmm2 # 0x1f5a970 vblendvps %xmm0, %xmm1, %xmm2, %xmm0 vmovaps %xmm0, 0xb0(%rax) vmovaps 0x30(%r14), %xmm0 vmovaps 0x80(%r14), %xmm1 vmaxps %xmm4, %xmm0, %xmm0 vmaxps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1cbe741(%rip), %xmm2 # 0x1eeba20 vblendvps %xmm3, %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0xc0(%rax) vbroadcastss 0x1cbf88e(%rip), %xmm0 # 0x1eecb84 vblendvps %xmm3, %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %r15 movq 0x70(%rbx), %rdx movq %r12, %rdi movq %rbx, %rsi movq %rsp, %r8 movq %r14, %r9 pushq %r13 leaq 0xd8(%rsp), %rax pushq %rax callq 0x271488 popq %rax popq %rcx andq %r15, %rbp jne 0x22d304 addq $0x1b8, %rsp # imm = 0x1B8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 257, true, embree::avx::VirtualCurveIntersectorK<4>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x18e8, %rsp # imm = 0x18E8 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x22ddf8 movq %rdx, %r9 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm1 vmovaps 0x80(%rdx), %xmm0 vxorps %xmm4, %xmm4, %xmm4 vcmpnltps %xmm4, %xmm0, %xmm2 vandps %xmm1, %xmm2, %xmm1 vmovaps %xmm1, 0x60(%rsp) vmovmskps %xmm1, %eax testl %eax, %eax je 0x22ddf8 movq %rcx, %r10 leaq 0x80(%r9), %r11 movzbl %al, %eax leaq 0x180(%rsp), %rcx vmovaps 0x40(%r9), %xmm3 vmovaps 0x50(%r9), %xmm2 vmovaps 0x60(%r9), %xmm1 vmulps %xmm1, %xmm1, %xmm5 vmulps %xmm2, %xmm2, %xmm6 vaddps %xmm6, %xmm5, %xmm5 vmulps %xmm3, %xmm3, %xmm6 vaddps %xmm5, %xmm6, %xmm5 vrsqrtps %xmm5, %xmm6 vbroadcastss 0x1cbf339(%rip), %xmm7 # 0x1eec718 vmulps %xmm7, %xmm6, %xmm7 vbroadcastss 0x1cbf330(%rip), %xmm8 # 0x1eec71c vmulps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vmulps %xmm6, %xmm6, %xmm6 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm7, %xmm5 vmovaps %xmm5, -0x10(%rcx) vbroadcastss 0x1cf3ab2(%rip), %xmm15 # 0x1f20ec0 vmovss 0x1cbf302(%rip), %xmm5 # 0x1eec718 vmovss 0x1cbf762(%rip), %xmm6 # 0x1eecb80 bsfq %rax, %rdx vmovss 0x40(%r9,%rdx,4), %xmm7 vinsertps $0x1c, 0x50(%r9,%rdx,4), %xmm7, %xmm7 # xmm7 = xmm7[0],mem[0],zero,zero vinsertps $0x28, 0x60(%r9,%rdx,4), %xmm7, %xmm8 # xmm8 = xmm7[0,1],mem[0],zero vbroadcastss 0x170(%rsp,%rdx,4), %xmm7 vmulps %xmm7, %xmm8, %xmm8 vshufpd $0x1, %xmm8, %xmm8, %xmm9 # xmm9 = xmm8[1,0] vmovshdup %xmm8, %xmm10 # xmm10 = xmm8[1,1,3,3] vxorps %xmm15, %xmm10, %xmm10 vunpckhps %xmm4, %xmm8, %xmm11 # xmm11 = xmm8[2],xmm4[2],xmm8[3],xmm4[3] vmovss %xmm10, %xmm4, %xmm10 # xmm10 = xmm10[0],xmm4[1,2,3] vshufps $0x41, %xmm10, %xmm11, %xmm10 # xmm10 = xmm11[1,0],xmm10[0,1] vxorpd %xmm15, %xmm9, %xmm9 vinsertps $0x2a, %xmm8, %xmm9, %xmm9 # xmm9 = xmm9[0],zero,xmm8[0],zero vdpps $0x7f, %xmm10, %xmm10, %xmm11 vdpps $0x7f, %xmm9, %xmm9, %xmm12 vcmpltps %xmm11, %xmm12, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vblendvps %xmm11, %xmm10, %xmm9, %xmm9 vdpps $0x7f, %xmm9, %xmm9, %xmm10 vrsqrtss %xmm10, %xmm10, %xmm11 vmulss %xmm5, %xmm11, %xmm12 vmulss %xmm6, %xmm10, %xmm10 vmulss %xmm11, %xmm10, %xmm10 vmulss %xmm11, %xmm11, %xmm11 vmulss %xmm11, %xmm10, %xmm10 vsubss %xmm10, %xmm12, %xmm10 vshufps $0x0, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] vmulps %xmm10, %xmm9, %xmm9 vshufps $0xc9, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,2,0,3] vshufps $0xc9, %xmm8, %xmm8, %xmm11 # xmm11 = xmm8[1,2,0,3] vmulps %xmm9, %xmm11, %xmm11 vmulps %xmm10, %xmm8, %xmm10 vsubps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm10 # xmm10 = xmm10[1,2,0,3] vdpps $0x7f, %xmm10, %xmm10, %xmm11 leaq -0x1(%rax), %rdi imulq $0x30, %rdx, %rdx vrsqrtss %xmm11, %xmm11, %xmm12 vmulss %xmm5, %xmm12, %xmm13 vmulss %xmm6, %xmm11, %xmm11 vmulss %xmm12, %xmm11, %xmm11 vmulss %xmm12, %xmm12, %xmm12 vmulss %xmm12, %xmm11, %xmm11 vsubss %xmm11, %xmm13, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vmulps %xmm10, %xmm11, %xmm10 vmulps %xmm7, %xmm8, %xmm7 vunpcklps %xmm7, %xmm9, %xmm8 # xmm8 = xmm9[0],xmm7[0],xmm9[1],xmm7[1] vunpckhps %xmm7, %xmm9, %xmm7 # xmm7 = xmm9[2],xmm7[2],xmm9[3],xmm7[3] vunpcklps %xmm4, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm4[0],xmm10[1],xmm4[1] vunpckhps %xmm4, %xmm10, %xmm10 # xmm10 = xmm10[2],xmm4[2],xmm10[3],xmm4[3] vunpcklps %xmm10, %xmm7, %xmm7 # xmm7 = xmm7[0],xmm10[0],xmm7[1],xmm10[1] vunpcklps %xmm9, %xmm8, %xmm10 # xmm10 = xmm8[0],xmm9[0],xmm8[1],xmm9[1] vunpckhps %xmm9, %xmm8, %xmm8 # xmm8 = xmm8[2],xmm9[2],xmm8[3],xmm9[3] vmovaps %xmm10, (%rcx,%rdx) vmovaps %xmm8, 0x10(%rcx,%rdx) vmovaps %xmm7, 0x20(%rcx,%rdx) andq %rdi, %rax jne 0x22d41e vmovaps (%r9), %xmm4 vmovaps %xmm4, 0x80(%rsp) vmovaps 0x10(%r9), %xmm4 vmovaps %xmm4, 0x90(%rsp) vmovaps 0x20(%r9), %xmm4 vmovaps %xmm4, 0xa0(%rsp) vmovaps %xmm3, 0xb0(%rsp) vmovaps %xmm2, 0xc0(%rsp) vmovaps %xmm1, 0xd0(%rsp) vbroadcastss 0x1cf391b(%rip), %xmm6 # 0x1f20ec4 vandps %xmm6, %xmm3, %xmm4 vbroadcastss 0x1cc3a32(%rip), %xmm7 # 0x1ef0fe8 vcmpltps %xmm7, %xmm4, %xmm4 vbroadcastss 0x1cbf150(%rip), %xmm8 # 0x1eec714 vdivps %xmm3, %xmm8, %xmm3 vandps %xmm6, %xmm2, %xmm5 vcmpltps %xmm7, %xmm5, %xmm5 vdivps %xmm2, %xmm8, %xmm2 vandps %xmm6, %xmm1, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vdivps %xmm1, %xmm8, %xmm1 vbroadcastss 0x1cf3975(%rip), %xmm7 # 0x1f20f60 vblendvps %xmm4, %xmm7, %xmm3, %xmm3 vblendvps %xmm5, %xmm7, %xmm2, %xmm2 vblendvps %xmm6, %xmm7, %xmm1, %xmm1 vmovaps %xmm3, 0xe0(%rsp) vmovaps %xmm2, 0xf0(%rsp) vmovaps %xmm1, 0x100(%rsp) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1d2d33a(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm3, %xmm3 vcmpnltps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d27364(%rip), %xmm5 # 0x1f549a0 vbroadcastss 0x1d2d323(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm2, %xmm5, %xmm6, %xmm2 vmovaps %xmm3, 0x110(%rsp) vmovaps %xmm2, 0x120(%rsp) vcmpnltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d2d301(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d2d2fc(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0x130(%rsp) vmovaps 0x30(%r9), %xmm1 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm0, %xmm2 vbroadcastss 0x1cbe386(%rip), %xmm10 # 0x1eeba20 vmovaps 0x60(%rsp), %xmm3 vblendvps %xmm3, %xmm1, %xmm10, %xmm0 vmovaps %xmm0, 0x140(%rsp) vbroadcastss 0x1cbf4cc(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm3, %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x150(%rsp) vpcmpeqd %xmm1, %xmm1, %xmm1 vxorps %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0x30(%rsp) cmpq $0x0, 0x8(%r10) je 0x22d6f0 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r15d, %r15d cmpb $0x1, %al adcq $0x2, %r15 jmp 0x22d6f4 pushq $0x3 popq %r15 leaq 0x250(%rsp), %r12 movq $-0x8, -0x10(%r12) leaq 0x9f0(%rsp), %rbx vmovaps %xmm10, -0x20(%rbx) movq 0x70(%r8), %rax movq %rax, -0x8(%r12) vmovaps %xmm0, -0x10(%rbx) leaq 0x1f22859(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x160(%rsp) vbroadcastss 0x1cf27d3(%rip), %xmm11 # 0x1f1ff10 vbroadcastss 0x1cf27ce(%rip), %xmm12 # 0x1f1ff14 movq %r10, 0x48(%rsp) movq %r9, 0x40(%rsp) addq $-0x10, %rbx movq -0x8(%r12), %rbp addq $-0x8, %r12 cmpq $-0x8, %rbp je 0x22dcf3 vmovaps (%rbx), %xmm13 vcmpltps 0x150(%rsp), %xmm13, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22dcf7 movzbl %al, %r14d popcntl %r14d, %r13d xorl %eax, %eax cmpq %r15, %r13 jbe 0x22dcfc cmpq %r15, %r13 jbe 0x22dce5 testb $0x8, %bpl pushq $0x8 popq %r14 jne 0x22dbbd movq %rbp, %rax movq %rbp, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %r14, %rbp vmovaps %xmm10, %xmm13 movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x22d8e4 testb $0x7, %al jne 0x22d904 vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x80(%rsp), %xmm1 vmovaps 0x90(%rsp), %xmm2 vmovaps 0xa0(%rsp), %xmm3 vmovaps 0xe0(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0xf0(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x100(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vminps %xmm1, %xmm5, %xmm0 vminps %xmm2, %xmm7, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vminps %xmm3, %xmm9, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vmulps %xmm0, %xmm11, %xmm0 vmaxps %xmm1, %xmm5, %xmm1 vmaxps %xmm2, %xmm7, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmaxps %xmm3, %xmm9, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmaxps 0x140(%rsp), %xmm0, %xmm2 vmulps %xmm1, %xmm12, %xmm1 vminps 0x150(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vmovddup 0x1cf3633(%rip), %xmm2 # xmm2 = mem[0,0] vptest %xmm2, %xmm1 je 0x22d8e4 vpslld $0x1f, %xmm1, %xmm1 vblendvps %xmm1, %xmm0, %xmm10, %xmm0 cmpq $0x8, %rbp je 0x22d8dd movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm13, (%rbx) addq $0x10, %rbx vmovaps %xmm0, %xmm13 movq %rdi, %rbp cmpq $0x8, %rdi je 0x22db6c leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x22d7c0 jmp 0x22db6c vbroadcastss 0x20(%rcx,%rdx,4), %xmm2 vbroadcastss 0x30(%rcx,%rdx,4), %xmm1 vbroadcastss 0x40(%rcx,%rdx,4), %xmm0 vmovaps %xmm0, 0x10(%rsp) vbroadcastss 0x50(%rcx,%rdx,4), %xmm5 vbroadcastss 0x60(%rcx,%rdx,4), %xmm4 vbroadcastss 0x70(%rcx,%rdx,4), %xmm3 vbroadcastss 0x80(%rcx,%rdx,4), %xmm8 vbroadcastss 0x90(%rcx,%rdx,4), %xmm7 vbroadcastss 0xa0(%rcx,%rdx,4), %xmm6 vmovaps 0xc0(%rsp), %xmm9 vmovaps 0xd0(%rsp), %xmm10 vmulps %xmm10, %xmm8, %xmm11 vmulps %xmm7, %xmm10, %xmm12 vmovaps %xmm13, 0x20(%rsp) vmulps %xmm5, %xmm9, %xmm13 vaddps %xmm11, %xmm13, %xmm11 vmulps %xmm4, %xmm9, %xmm13 vaddps %xmm12, %xmm13, %xmm12 vmovaps 0xb0(%rsp), %xmm13 vmulps %xmm6, %xmm10, %xmm10 vmulps %xmm3, %xmm9, %xmm9 vaddps %xmm10, %xmm9, %xmm9 vmulps %xmm2, %xmm13, %xmm10 vaddps %xmm11, %xmm10, %xmm10 vmulps %xmm1, %xmm13, %xmm11 vaddps %xmm12, %xmm11, %xmm11 vbroadcastss 0xb0(%rcx,%rdx,4), %xmm12 vmulps %xmm0, %xmm13, %xmm13 vaddps %xmm9, %xmm13, %xmm9 vbroadcastss 0x1cf34fb(%rip), %xmm14 # 0x1f20ec4 vandps %xmm14, %xmm10, %xmm13 vbroadcastss 0x1cc3611(%rip), %xmm0 # 0x1ef0fe8 vcmpltps %xmm0, %xmm13, %xmm13 vblendvps %xmm13, %xmm0, %xmm10, %xmm10 vandps %xmm14, %xmm11, %xmm13 vcmpltps %xmm0, %xmm13, %xmm13 vblendvps %xmm13, %xmm0, %xmm11, %xmm11 vandps %xmm14, %xmm9, %xmm13 vcmpltps %xmm0, %xmm13, %xmm13 vblendvps %xmm13, %xmm0, %xmm9, %xmm13 vrcpps %xmm10, %xmm9 vmulps %xmm10, %xmm9, %xmm10 vbroadcastss 0x1cbecff(%rip), %xmm0 # 0x1eec714 vsubps %xmm10, %xmm0, %xmm10 vmulps %xmm10, %xmm9, %xmm10 vaddps %xmm10, %xmm9, %xmm9 vrcpps %xmm11, %xmm10 vmulps %xmm11, %xmm10, %xmm11 vsubps %xmm11, %xmm0, %xmm11 vmulps %xmm11, %xmm10, %xmm11 vaddps %xmm11, %xmm10, %xmm10 vrcpps %xmm13, %xmm11 vmulps %xmm13, %xmm11, %xmm13 vsubps %xmm13, %xmm0, %xmm13 vmulps %xmm13, %xmm11, %xmm13 vaddps %xmm13, %xmm11, %xmm11 vmovaps 0xa0(%rsp), %xmm13 vmulps %xmm13, %xmm8, %xmm8 vaddps %xmm8, %xmm12, %xmm8 vbroadcastss 0xc0(%rcx,%rdx,4), %xmm12 vmulps %xmm7, %xmm13, %xmm7 vaddps %xmm7, %xmm12, %xmm7 vmulps %xmm6, %xmm13, %xmm6 vmovaps 0x20(%rsp), %xmm13 vbroadcastss 0xd0(%rcx,%rdx,4), %xmm12 vaddps %xmm6, %xmm12, %xmm6 vmovaps 0x90(%rsp), %xmm12 vmulps %xmm5, %xmm12, %xmm5 vaddps %xmm5, %xmm8, %xmm5 vmulps %xmm4, %xmm12, %xmm4 vaddps %xmm7, %xmm4, %xmm4 vmulps %xmm3, %xmm12, %xmm3 vbroadcastss 0x1cf245b(%rip), %xmm12 # 0x1f1ff14 vaddps %xmm6, %xmm3, %xmm3 vmovaps 0x80(%rsp), %xmm6 vmulps %xmm6, %xmm2, %xmm2 vaddps %xmm5, %xmm2, %xmm2 vxorps %xmm15, %xmm9, %xmm5 vmulps %xmm5, %xmm2, %xmm2 vmulps %xmm6, %xmm1, %xmm1 vaddps %xmm4, %xmm1, %xmm1 vxorps %xmm15, %xmm10, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vmulps 0x10(%rsp), %xmm6, %xmm0 vxorps %xmm15, %xmm11, %xmm4 vaddps %xmm3, %xmm0, %xmm0 vmulps %xmm4, %xmm0, %xmm0 vaddps %xmm2, %xmm9, %xmm3 vaddps %xmm1, %xmm10, %xmm4 vbroadcastss 0x1cbdf14(%rip), %xmm10 # 0x1eeba20 vaddps %xmm0, %xmm11, %xmm5 vbroadcastss 0x1cf23f7(%rip), %xmm11 # 0x1f1ff10 vpminsd %xmm3, %xmm2, %xmm6 vpminsd %xmm4, %xmm1, %xmm7 vpmaxsd %xmm7, %xmm6, %xmm6 vpminsd %xmm5, %xmm0, %xmm7 vpmaxsd %xmm7, %xmm6, %xmm6 vpmaxsd %xmm3, %xmm2, %xmm2 vpmaxsd %xmm4, %xmm1, %xmm1 vpminsd %xmm1, %xmm2, %xmm1 vpmaxsd %xmm5, %xmm0, %xmm0 vpminsd %xmm0, %xmm1, %xmm1 vmulps %xmm6, %xmm11, %xmm0 vpmaxsd 0x140(%rsp), %xmm0, %xmm2 vmulps %xmm1, %xmm12, %xmm1 vpminsd 0x150(%rsp), %xmm1, %xmm1 jmp 0x22d8a8 xorl %eax, %eax cmpq $0x8, %rbp je 0x22dbb6 vmovaps 0x150(%rsp), %xmm0 vcmpnleps %xmm13, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r15 jae 0x22dba6 testb %cl, %cl je 0x22dce5 testb $0x8, %bpl je 0x22d7ac jmp 0x22dbbd movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm13, (%rbx) addq $0x10, %rbx xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x22db92 cmpq $-0x8, %rbp je 0x22dcf3 movq %r11, 0x10(%rsp) movq %r8, (%rsp) vmovaps 0x150(%rsp), %xmm0 vcmpnleps %xmm13, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22ddd6 andq $-0x10, %rbp vmovaps 0x30(%rsp), %xmm1 movzbl (%rbp), %eax movq %rsi, 0x8(%rsp) movq 0x8(%rsi), %rcx vmovaps 0x160(%rsp), %xmm0 vmovaps %xmm0, 0x70(%rsp) vmovaps %xmm1, 0x50(%rsp) vmovmskps %xmm1, %edx xorl $0xf, %edx je 0x22dc6e movzbl %dl, %r14d shll $0x6, %eax addq %rcx, %rax addq $0x18, %rax movq %rax, 0x20(%rsp) bsfq %r14, %r13 leaq 0x170(%rsp), %rdi movq %r9, %rsi movq %r13, %rdx movq %r10, %rcx movq %rbp, %r8 movq 0x20(%rsp), %rax callq *(%rax) testb %al, %al je 0x22dc5b orl $-0x1, 0x70(%rsp,%r13,4) leaq -0x1(%r14), %rax andq %rax, %r14 movq 0x48(%rsp), %r10 movq 0x40(%rsp), %r9 jne 0x22dc32 vmovaps 0x50(%rsp), %xmm0 vorps 0x70(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x30(%rsp) vtestps 0x1cbe197(%rip), %xmm0 # 0x1eebe20 jb 0x22dcb0 vmovaps 0x150(%rsp), %xmm1 vbroadcastss 0x1cbeee7(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x150(%rsp) xorl %eax, %eax jmp 0x22dcb3 pushq $0x3 popq %rax movq 0x8(%rsp), %rsi movq (%rsp), %r8 movq 0x10(%rsp), %r11 vbroadcastss 0x1cf31f6(%rip), %xmm15 # 0x1f20ec0 vbroadcastss 0x1cbdd4d(%rip), %xmm10 # 0x1eeba20 vbroadcastss 0x1cf2234(%rip), %xmm11 # 0x1f1ff10 vbroadcastss 0x1cf222f(%rip), %xmm12 # 0x1f1ff14 cmpl $0x3, %eax jne 0x22d750 jmp 0x22ddde pushq $0x3 jmp 0x22dcf9 pushq $0x2 popq %rax jmp 0x22dce5 vmovaps %xmm13, 0x20(%rsp) movq %r11, 0x10(%rsp) movq %rsi, 0x8(%rsp) movq %r8, (%rsp) bsfq %r14, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, 0x50(%rsp) leaq 0x170(%rsp), %r8 pushq %r10 leaq 0x88(%rsp), %rax pushq %rax callq 0x271c42 popq %rcx popq %rdx testb %al, %al je 0x22dd4a movq 0x50(%rsp), %rax orl $-0x1, 0x30(%rsp,%rax,4) leaq -0x1(%r14), %rax andq %rax, %r14 movq 0x48(%rsp), %r10 movq 0x40(%rsp), %r9 movq 0x8(%rsp), %rsi movq (%rsp), %r8 jne 0x22dd10 vmovaps 0x30(%rsp), %xmm0 vtestps 0x1cbe0ab(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0x10(%rsp), %r11 vbroadcastss 0x1cf313a(%rip), %xmm15 # 0x1f20ec0 vbroadcastss 0x1cbdc91(%rip), %xmm10 # 0x1eeba20 vbroadcastss 0x1cf2178(%rip), %xmm11 # 0x1f1ff10 vbroadcastss 0x1cf2173(%rip), %xmm12 # 0x1f1ff14 vmovaps 0x20(%rsp), %xmm13 jb 0x22d795 vmovaps 0x150(%rsp), %xmm1 vbroadcastss 0x1cbedc5(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x150(%rsp) pushq $0x2 popq %rax jmp 0x22d795 pushq $0x2 popq %rax jmp 0x22dcb8 vmovaps 0x60(%rsp), %xmm0 vandps 0x30(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cbed91(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r11) addq $0x18e8, %rsp # imm = 0x18E8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16781328, true, embree::avx::VirtualCurveIntersectorK<4>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1b8, %rsp # imm = 0x1B8 movq (%rsi), %rbx cmpq $0x8, 0x70(%rbx) je 0x22e187 vmovdqa (%rdi), %xmm3 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd %xmm0, %xmm3, %xmm0 vmovmskps %xmm0, %ebp testl %ebp, %ebp je 0x22e187 movq %rcx, %r13 movq %rdx, %r14 movq %rsi, %r12 leaq 0x10(%rsp), %rax vmovaps 0x40(%rdx), %xmm2 vmovaps 0x50(%rdx), %xmm1 vmovaps 0x60(%rdx), %xmm0 vmulps %xmm0, %xmm0, %xmm4 vmulps %xmm1, %xmm1, %xmm5 vaddps %xmm5, %xmm4, %xmm4 vmulps %xmm2, %xmm2, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vrsqrtps %xmm4, %xmm5 vbroadcastss 0x1cbe899(%rip), %xmm6 # 0x1eec718 vmulps %xmm6, %xmm5, %xmm6 vbroadcastss 0x1cbe890(%rip), %xmm7 # 0x1eec71c vmulps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm5, %xmm4 vmulps %xmm5, %xmm5, %xmm5 vmulps %xmm4, %xmm5, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vmovaps %xmm4, -0x10(%rax) vbroadcastss 0x1cf3012(%rip), %xmm4 # 0x1f20ec0 vxorps %xmm5, %xmm5, %xmm5 vmovss 0x1cbe85e(%rip), %xmm6 # 0x1eec718 vmovss 0x1cbecbe(%rip), %xmm7 # 0x1eecb80 movq %rbp, %rcx bsfq %rcx, %rdx vmovss 0x40(%r14,%rdx,4), %xmm8 vinsertps $0x1c, 0x50(%r14,%rdx,4), %xmm8, %xmm8 # xmm8 = xmm8[0],mem[0],zero,zero vinsertps $0x28, 0x60(%r14,%rdx,4), %xmm8, %xmm9 # xmm9 = xmm8[0,1],mem[0],zero vbroadcastss (%rsp,%rdx,4), %xmm8 vmulps %xmm8, %xmm9, %xmm9 vshufpd $0x1, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,0] vmovshdup %xmm9, %xmm11 # xmm11 = xmm9[1,1,3,3] vxorps %xmm4, %xmm11, %xmm11 vunpckhps %xmm5, %xmm9, %xmm12 # xmm12 = xmm9[2],xmm5[2],xmm9[3],xmm5[3] vmovss %xmm11, %xmm5, %xmm11 # xmm11 = xmm11[0],xmm5[1,2,3] vshufps $0x41, %xmm11, %xmm12, %xmm11 # xmm11 = xmm12[1,0],xmm11[0,1] vxorpd %xmm4, %xmm10, %xmm10 vinsertps $0x2a, %xmm9, %xmm10, %xmm10 # xmm10 = xmm10[0],zero,xmm9[0],zero vdpps $0x7f, %xmm11, %xmm11, %xmm12 vdpps $0x7f, %xmm10, %xmm10, %xmm13 vcmpltps %xmm12, %xmm13, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vblendvps %xmm12, %xmm11, %xmm10, %xmm10 vdpps $0x7f, %xmm10, %xmm10, %xmm11 vrsqrtss %xmm11, %xmm11, %xmm12 vmulss %xmm6, %xmm12, %xmm13 vmulss %xmm7, %xmm11, %xmm11 vmulss %xmm12, %xmm11, %xmm11 vmulss %xmm12, %xmm12, %xmm12 vmulss %xmm12, %xmm11, %xmm11 vsubss %xmm11, %xmm13, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vmulps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm11 # xmm11 = xmm10[1,2,0,3] vshufps $0xc9, %xmm9, %xmm9, %xmm12 # xmm12 = xmm9[1,2,0,3] vmulps %xmm10, %xmm12, %xmm12 vmulps %xmm11, %xmm9, %xmm11 vsubps %xmm12, %xmm11, %xmm11 vshufps $0xc9, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[1,2,0,3] vdpps $0x7f, %xmm11, %xmm11, %xmm12 leaq -0x1(%rcx), %rdi imulq $0x30, %rdx, %rdx vrsqrtss %xmm12, %xmm12, %xmm13 vmulss %xmm6, %xmm13, %xmm14 vmulss %xmm7, %xmm12, %xmm12 vmulss %xmm13, %xmm12, %xmm12 vmulss %xmm13, %xmm13, %xmm13 vmulss %xmm13, %xmm12, %xmm12 vsubss %xmm12, %xmm14, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vmulps %xmm11, %xmm12, %xmm11 vmulps %xmm9, %xmm8, %xmm8 vunpcklps %xmm8, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm8[0],xmm10[1],xmm8[1] vunpckhps %xmm8, %xmm10, %xmm8 # xmm8 = xmm10[2],xmm8[2],xmm10[3],xmm8[3] vunpcklps %xmm5, %xmm11, %xmm10 # xmm10 = xmm11[0],xmm5[0],xmm11[1],xmm5[1] vunpckhps %xmm5, %xmm11, %xmm11 # xmm11 = xmm11[2],xmm5[2],xmm11[3],xmm5[3] vunpcklps %xmm11, %xmm8, %xmm8 # xmm8 = xmm8[0],xmm11[0],xmm8[1],xmm11[1] vunpcklps %xmm10, %xmm9, %xmm11 # xmm11 = xmm9[0],xmm10[0],xmm9[1],xmm10[1] vunpckhps %xmm10, %xmm9, %xmm9 # xmm9 = xmm9[2],xmm10[2],xmm9[3],xmm10[3] vmovaps %xmm11, (%rax,%rdx) vmovaps %xmm9, 0x10(%rax,%rdx) vmovaps %xmm8, 0x20(%rax,%rdx) andq %rdi, %rcx jne 0x22dec5 vpcmpeqd %xmm4, %xmm4, %xmm4 vpcmpeqd %xmm4, %xmm3, %xmm3 vmovaps (%r14), %xmm4 leaq 0xd0(%rsp), %rax vmovaps %xmm4, (%rax) vmovaps 0x10(%r14), %xmm4 vmovaps %xmm4, 0x10(%rax) vmovaps 0x20(%r14), %xmm4 vmovaps %xmm4, 0x20(%rax) vmovaps %xmm2, 0x30(%rax) vmovaps %xmm1, 0x40(%rax) vmovaps %xmm0, 0x50(%rax) vbroadcastss 0x1cf2e7f(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm2, %xmm5 vbroadcastss 0x1cc2f96(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1cbe6b4(%rip), %xmm7 # 0x1eec714 vdivps %xmm2, %xmm7, %xmm2 vandps %xmm4, %xmm1, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vdivps %xmm1, %xmm7, %xmm1 vandps %xmm4, %xmm0, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vbroadcastss 0x1cf2edd(%rip), %xmm6 # 0x1f20f60 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vblendvps %xmm8, %xmm6, %xmm1, %xmm1 vdivps %xmm0, %xmm7, %xmm0 vblendvps %xmm4, %xmm6, %xmm0, %xmm0 vmovaps %xmm2, 0x60(%rax) vmovaps %xmm1, 0x70(%rax) vmovaps %xmm0, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d2c8a7(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm2, %xmm2 vmovaps %xmm2, 0x90(%rax) vcmpnltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d268c9(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d2c888(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm0, %xmm0 vbroadcastss 0x1d2c870(%rip), %xmm1 # 0x1f5a96c vbroadcastss 0x1d2c86b(%rip), %xmm2 # 0x1f5a970 vblendvps %xmm0, %xmm1, %xmm2, %xmm0 vmovaps %xmm0, 0xb0(%rax) vmovaps 0x30(%r14), %xmm0 vmovaps 0x80(%r14), %xmm1 vmaxps %xmm4, %xmm0, %xmm0 vmaxps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1cbd8ed(%rip), %xmm2 # 0x1eeba20 vblendvps %xmm3, %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0xc0(%rax) vbroadcastss 0x1cbea3a(%rip), %xmm0 # 0x1eecb84 vblendvps %xmm3, %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %r15 movq 0x70(%rbx), %rdx movq %r12, %rdi movq %rbx, %rsi movq %rsp, %r8 movq %r14, %r9 pushq %r13 leaq 0xd8(%rsp), %rax pushq %rax callq 0x272244 popq %rax popq %rcx andq %r15, %rbp jne 0x22e158 addq $0x1b8, %rsp # imm = 0x1B8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16781328, true, embree::avx::VirtualCurveIntersectorK<4>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1948, %rsp # imm = 0x1948 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x22ede5 movq %rdx, %r14 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm1 vmovaps 0x80(%rdx), %xmm0 vxorps %xmm4, %xmm4, %xmm4 vcmpnltps %xmm4, %xmm0, %xmm2 vandps %xmm1, %xmm2, %xmm1 vmovaps %xmm1, 0x70(%rsp) vmovmskps %xmm1, %eax testl %eax, %eax je 0x22ede5 movq %rcx, %r10 leaq 0x80(%r14), %r9 movzbl %al, %eax leaq 0x1e0(%rsp), %rcx vmovaps 0x40(%r14), %xmm3 vmovaps 0x50(%r14), %xmm2 vmovaps 0x60(%r14), %xmm1 vmulps %xmm1, %xmm1, %xmm5 vmulps %xmm2, %xmm2, %xmm6 vaddps %xmm6, %xmm5, %xmm5 vmulps %xmm3, %xmm3, %xmm6 vaddps %xmm5, %xmm6, %xmm5 vrsqrtps %xmm5, %xmm6 vbroadcastss 0x1cbe4e5(%rip), %xmm7 # 0x1eec718 vmulps %xmm7, %xmm6, %xmm7 vbroadcastss 0x1cbe4dc(%rip), %xmm8 # 0x1eec71c vmulps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vmulps %xmm6, %xmm6, %xmm6 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm7, %xmm5 vmovaps %xmm5, -0x10(%rcx) vbroadcastss 0x1cf2c5e(%rip), %xmm5 # 0x1f20ec0 vmovss 0x1cbe4ae(%rip), %xmm6 # 0x1eec718 vmovss 0x1cbe90e(%rip), %xmm7 # 0x1eecb80 bsfq %rax, %rdx vmovss 0x40(%r14,%rdx,4), %xmm8 vinsertps $0x1c, 0x50(%r14,%rdx,4), %xmm8, %xmm8 # xmm8 = xmm8[0],mem[0],zero,zero vinsertps $0x28, 0x60(%r14,%rdx,4), %xmm8, %xmm9 # xmm9 = xmm8[0,1],mem[0],zero vbroadcastss 0x1d0(%rsp,%rdx,4), %xmm8 vmulps %xmm8, %xmm9, %xmm9 vshufpd $0x1, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,0] vmovshdup %xmm9, %xmm11 # xmm11 = xmm9[1,1,3,3] vxorps %xmm5, %xmm11, %xmm11 vunpckhps %xmm4, %xmm9, %xmm12 # xmm12 = xmm9[2],xmm4[2],xmm9[3],xmm4[3] vmovss %xmm11, %xmm4, %xmm11 # xmm11 = xmm11[0],xmm4[1,2,3] vshufps $0x41, %xmm11, %xmm12, %xmm11 # xmm11 = xmm12[1,0],xmm11[0,1] vxorpd %xmm5, %xmm10, %xmm10 vinsertps $0x2a, %xmm9, %xmm10, %xmm10 # xmm10 = xmm10[0],zero,xmm9[0],zero vdpps $0x7f, %xmm11, %xmm11, %xmm12 vdpps $0x7f, %xmm10, %xmm10, %xmm13 vcmpltps %xmm12, %xmm13, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vblendvps %xmm12, %xmm11, %xmm10, %xmm10 vdpps $0x7f, %xmm10, %xmm10, %xmm11 vrsqrtss %xmm11, %xmm11, %xmm12 vmulss %xmm6, %xmm12, %xmm13 vmulss %xmm7, %xmm11, %xmm11 vmulss %xmm12, %xmm11, %xmm11 vmulss %xmm12, %xmm12, %xmm12 vmulss %xmm12, %xmm11, %xmm11 vsubss %xmm11, %xmm13, %xmm11 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vmulps %xmm11, %xmm10, %xmm10 vshufps $0xc9, %xmm10, %xmm10, %xmm11 # xmm11 = xmm10[1,2,0,3] vshufps $0xc9, %xmm9, %xmm9, %xmm12 # xmm12 = xmm9[1,2,0,3] vmulps %xmm10, %xmm12, %xmm12 vmulps %xmm11, %xmm9, %xmm11 vsubps %xmm12, %xmm11, %xmm11 vshufps $0xc9, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[1,2,0,3] vdpps $0x7f, %xmm11, %xmm11, %xmm12 leaq -0x1(%rax), %rdi imulq $0x30, %rdx, %rdx vrsqrtss %xmm12, %xmm12, %xmm13 vmulss %xmm6, %xmm13, %xmm14 vmulss %xmm7, %xmm12, %xmm12 vmulss %xmm13, %xmm12, %xmm12 vmulss %xmm13, %xmm13, %xmm13 vmulss %xmm13, %xmm12, %xmm12 vsubss %xmm12, %xmm14, %xmm12 vshufps $0x0, %xmm12, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] vmulps %xmm11, %xmm12, %xmm11 vmulps %xmm9, %xmm8, %xmm8 vunpcklps %xmm8, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm8[0],xmm10[1],xmm8[1] vunpckhps %xmm8, %xmm10, %xmm8 # xmm8 = xmm10[2],xmm8[2],xmm10[3],xmm8[3] vunpcklps %xmm4, %xmm11, %xmm10 # xmm10 = xmm11[0],xmm4[0],xmm11[1],xmm4[1] vunpckhps %xmm4, %xmm11, %xmm11 # xmm11 = xmm11[2],xmm4[2],xmm11[3],xmm4[3] vunpcklps %xmm11, %xmm8, %xmm8 # xmm8 = xmm8[0],xmm11[0],xmm8[1],xmm11[1] vunpcklps %xmm10, %xmm9, %xmm11 # xmm11 = xmm9[0],xmm10[0],xmm9[1],xmm10[1] vunpckhps %xmm10, %xmm9, %xmm9 # xmm9 = xmm9[2],xmm10[2],xmm9[3],xmm10[3] vmovaps %xmm11, (%rcx,%rdx) vmovaps %xmm9, 0x10(%rcx,%rdx) vmovaps %xmm8, 0x20(%rcx,%rdx) andq %rdi, %rax jne 0x22e272 vmovaps (%r14), %xmm4 vmovaps %xmm4, 0x90(%rsp) vmovaps 0x10(%r14), %xmm4 vmovaps %xmm4, 0xa0(%rsp) vmovaps 0x20(%r14), %xmm4 vmovaps %xmm4, 0xb0(%rsp) vmovaps %xmm3, 0xc0(%rsp) vmovaps %xmm2, 0xd0(%rsp) vmovaps %xmm1, 0xe0(%rsp) vbroadcastss 0x1cf2ac5(%rip), %xmm6 # 0x1f20ec4 vandps %xmm6, %xmm3, %xmm4 vbroadcastss 0x1cc2bdc(%rip), %xmm7 # 0x1ef0fe8 vcmpltps %xmm7, %xmm4, %xmm4 vbroadcastss 0x1cbe2fa(%rip), %xmm8 # 0x1eec714 vdivps %xmm3, %xmm8, %xmm3 vandps %xmm6, %xmm2, %xmm5 vcmpltps %xmm7, %xmm5, %xmm5 vdivps %xmm2, %xmm8, %xmm2 vandps %xmm6, %xmm1, %xmm6 vcmpltps %xmm7, %xmm6, %xmm6 vdivps %xmm1, %xmm8, %xmm1 vbroadcastss 0x1cf2b1f(%rip), %xmm7 # 0x1f20f60 vblendvps %xmm4, %xmm7, %xmm3, %xmm3 vblendvps %xmm5, %xmm7, %xmm2, %xmm2 vblendvps %xmm6, %xmm7, %xmm1, %xmm1 vmovaps %xmm3, 0xf0(%rsp) vmovaps %xmm2, 0x100(%rsp) vmovaps %xmm1, 0x110(%rsp) vxorps %xmm6, %xmm6, %xmm6 vcmpltps %xmm6, %xmm3, %xmm3 vbroadcastss 0x1d2c4e4(%rip), %xmm4 # 0x1f5a964 vandps %xmm4, %xmm3, %xmm3 vcmpnltps %xmm6, %xmm2, %xmm2 vbroadcastss 0x1d2650e(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d2c4cd(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm2, %xmm4, %xmm5, %xmm2 vmovaps %xmm3, 0x120(%rsp) vmovaps %xmm2, 0x130(%rsp) vcmpnltps %xmm6, %xmm1, %xmm1 vbroadcastss 0x1d2c4ab(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d2c4a6(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0x140(%rsp) vmovaps 0x30(%r14), %xmm1 vmaxps %xmm6, %xmm1, %xmm1 vmaxps %xmm6, %xmm0, %xmm2 vbroadcastss 0x1cbd530(%rip), %xmm14 # 0x1eeba20 vmovaps 0x70(%rsp), %xmm3 vblendvps %xmm3, %xmm1, %xmm14, %xmm0 vmovaps %xmm0, 0x150(%rsp) vbroadcastss 0x1cbe676(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm3, %xmm2, %xmm1, %xmm1 vmovaps %xmm1, 0x160(%rsp) vpcmpeqd %xmm1, %xmm1, %xmm1 vxorps %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0x50(%rsp) cmpq $0x0, 0x8(%r10) je 0x22e546 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r11d, %r11d cmpb $0x1, %al adcq $0x2, %r11 jmp 0x22e54a pushq $0x3 popq %r11 leaq 0x2b0(%rsp), %r12 movq $-0x8, -0x10(%r12) leaq 0xa50(%rsp), %rbx vmovaps %xmm14, -0x20(%rbx) movq 0x70(%r8), %rax movq %rax, -0x8(%r12) vmovaps %xmm0, -0x10(%rbx) leaq 0x1f21a03(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x170(%rsp) vbroadcastss 0x1cf197d(%rip), %xmm15 # 0x1f1ff10 movq %r10, 0x68(%rsp) addq $-0x10, %rbx movq -0x8(%r12), %rbp addq $-0x8, %r12 cmpq $-0x8, %rbp je 0x22ece4 vmovaps (%rbx), %xmm1 vcmpltps 0x160(%rsp), %xmm1, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22ece8 movzbl %al, %r15d popcntl %r15d, %r13d xorl %eax, %eax cmpq %r11, %r13 jbe 0x22eced cmpq %r11, %r13 jbe 0x22ecd6 pushq $0x8 popq %r15 testb $0x8, %bpl jne 0x22ebb3 vmovaps 0x160(%rsp), %xmm0 movq %rbp, %rax andq $-0x10, %rax movl %ebp, %ecx andl $0x7, %ecx vcmpnleps %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0x20(%rsp) xorl %edx, %edx movq %r15, %rbp vmovaps %xmm14, %xmm1 movq (%rax,%rdx,8), %rdi cmpq $0x8, %rdi je 0x22eb49 vmovaps %xmm1, 0x30(%rsp) cmpl $0x6, %ecx je 0x22e63f cmpl $0x1, %ecx jne 0x22e7d5 vbroadcastss 0x80(%rax,%rdx,4), %xmm0 vbroadcastss 0x20(%rax,%rdx,4), %xmm2 vmovaps 0x70(%r14), %xmm1 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm2, %xmm0, %xmm0 vbroadcastss 0xa0(%rax,%rdx,4), %xmm2 vbroadcastss 0x40(%rax,%rdx,4), %xmm3 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0xc0(%rax,%rdx,4), %xmm3 vbroadcastss 0x60(%rax,%rdx,4), %xmm4 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0x90(%rax,%rdx,4), %xmm4 vbroadcastss 0x30(%rax,%rdx,4), %xmm5 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0xb0(%rax,%rdx,4), %xmm5 vbroadcastss 0x50(%rax,%rdx,4), %xmm6 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0xd0(%rax,%rdx,4), %xmm6 vbroadcastss 0x70(%rax,%rdx,4), %xmm7 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vmovaps 0x90(%rsp), %xmm7 vmovaps 0xa0(%rsp), %xmm8 vmovaps 0xb0(%rsp), %xmm9 vmovaps 0xf0(%rsp), %xmm10 vsubps %xmm7, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vsubps %xmm8, %xmm2, %xmm2 vmovaps 0x100(%rsp), %xmm11 vmulps %xmm2, %xmm11, %xmm12 vsubps %xmm9, %xmm3, %xmm2 vmovaps 0x110(%rsp), %xmm3 vmulps %xmm3, %xmm2, %xmm13 vsubps %xmm7, %xmm4, %xmm2 vmulps %xmm2, %xmm10, %xmm4 vsubps %xmm8, %xmm5, %xmm2 vmulps %xmm2, %xmm11, %xmm5 vsubps %xmm9, %xmm6, %xmm2 vmulps %xmm3, %xmm2, %xmm3 vpminsd %xmm4, %xmm0, %xmm2 vpminsd %xmm5, %xmm12, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpminsd %xmm3, %xmm13, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vmulps %xmm2, %xmm15, %xmm2 vpmaxsd %xmm4, %xmm0, %xmm0 vpmaxsd %xmm5, %xmm12, %xmm4 vpminsd %xmm4, %xmm0, %xmm0 vpmaxsd %xmm3, %xmm13, %xmm3 vpminsd %xmm3, %xmm0, %xmm3 vpmaxsd 0x150(%rsp), %xmm2, %xmm0 vbroadcastss 0x1cf1786(%rip), %xmm4 # 0x1f1ff14 vmulps %xmm4, %xmm3, %xmm3 vpminsd 0x160(%rsp), %xmm3, %xmm3 vcmpleps %xmm3, %xmm0, %xmm0 cmpl $0x6, %ecx jne 0x22eb0e vbroadcastss 0xe0(%rax,%rdx,4), %xmm3 vcmpleps %xmm1, %xmm3, %xmm3 vbroadcastss 0xf0(%rax,%rdx,4), %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vandps %xmm1, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm0 jmp 0x22eb0e vbroadcastss 0x50(%rax,%rdx,4), %xmm8 vmovaps %xmm8, 0x180(%rsp) vbroadcastss 0x60(%rax,%rdx,4), %xmm4 vbroadcastss 0x70(%rax,%rdx,4), %xmm5 vbroadcastss 0x80(%rax,%rdx,4), %xmm11 vbroadcastss 0x90(%rax,%rdx,4), %xmm10 vbroadcastss 0xe0(%rax,%rdx,4), %xmm0 vbroadcastss 0xf0(%rax,%rdx,4), %xmm1 vbroadcastss 0x100(%rax,%rdx,4), %xmm2 vbroadcastss 0x110(%rax,%rdx,4), %xmm6 vbroadcastss 0x120(%rax,%rdx,4), %xmm7 vbroadcastss 0x130(%rax,%rdx,4), %xmm9 vmovaps 0x70(%r14), %xmm12 vbroadcastss 0x1cbdec2(%rip), %xmm3 # 0x1eec714 vsubps %xmm12, %xmm3, %xmm13 vmulps %xmm0, %xmm12, %xmm0 vmulps %xmm1, %xmm12, %xmm1 vmulps %xmm2, %xmm12, %xmm2 vmulps 0x1cbd1a5(%rip), %xmm13, %xmm14 # 0x1eeba10 vaddps %xmm0, %xmm14, %xmm0 vmovaps %xmm0, 0x1a0(%rsp) vaddps %xmm1, %xmm14, %xmm0 vmovaps %xmm0, 0x40(%rsp) vaddps %xmm2, %xmm14, %xmm0 vmovaps %xmm0, 0x10(%rsp) vmulps %xmm6, %xmm12, %xmm0 vmulps %xmm7, %xmm12, %xmm1 vmulps %xmm9, %xmm12, %xmm2 vaddps %xmm0, %xmm13, %xmm0 vmovaps %xmm0, 0x190(%rsp) vaddps %xmm1, %xmm13, %xmm0 vmovaps %xmm0, 0x1b0(%rsp) vaddps %xmm2, %xmm13, %xmm0 vmovaps %xmm0, 0x1c0(%rsp) vmovaps 0xd0(%rsp), %xmm0 vmovaps 0xe0(%rsp), %xmm1 vmulps %xmm1, %xmm11, %xmm2 vmulps %xmm1, %xmm10, %xmm12 vmulps %xmm0, %xmm8, %xmm13 vaddps %xmm2, %xmm13, %xmm2 vmulps %xmm0, %xmm4, %xmm13 vmovaps %xmm4, %xmm9 vaddps %xmm12, %xmm13, %xmm14 vmovaps %xmm15, %xmm4 vbroadcastss 0xa0(%rax,%rdx,4), %xmm15 vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm0, %xmm5, %xmm0 vaddps %xmm1, %xmm0, %xmm0 vbroadcastss 0x20(%rax,%rdx,4), %xmm12 vmovaps 0xc0(%rsp), %xmm1 vmulps %xmm1, %xmm12, %xmm13 vaddps %xmm2, %xmm13, %xmm2 vbroadcastss 0x30(%rax,%rdx,4), %xmm13 vmulps %xmm1, %xmm13, %xmm6 vaddps %xmm6, %xmm14, %xmm6 vbroadcastss 0x40(%rax,%rdx,4), %xmm14 vmulps %xmm1, %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cf257c(%rip), %xmm7 # 0x1f20ec4 vandps %xmm7, %xmm2, %xmm1 vbroadcastss 0x1cc2693(%rip), %xmm8 # 0x1ef0fe8 vcmpltps %xmm8, %xmm1, %xmm1 vblendvps %xmm1, %xmm8, %xmm2, %xmm1 vandps %xmm7, %xmm6, %xmm2 vcmpltps %xmm8, %xmm2, %xmm2 vblendvps %xmm2, %xmm8, %xmm6, %xmm6 vandps %xmm7, %xmm0, %xmm2 vcmpltps %xmm8, %xmm2, %xmm2 vblendvps %xmm2, %xmm8, %xmm0, %xmm0 vrcpps %xmm1, %xmm2 vmulps %xmm1, %xmm2, %xmm1 vsubps %xmm1, %xmm3, %xmm1 vmulps %xmm1, %xmm2, %xmm1 vaddps %xmm1, %xmm2, %xmm2 vrcpps %xmm6, %xmm1 vmulps %xmm6, %xmm1, %xmm6 vsubps %xmm6, %xmm3, %xmm6 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm6, %xmm1, %xmm1 vrcpps %xmm0, %xmm6 vmulps %xmm0, %xmm6, %xmm0 vsubps %xmm0, %xmm3, %xmm0 vmulps %xmm0, %xmm6, %xmm0 vaddps %xmm0, %xmm6, %xmm0 vmovaps 0xb0(%rsp), %xmm6 vmulps %xmm6, %xmm11, %xmm11 vbroadcastss 0xb0(%rax,%rdx,4), %xmm7 vaddps %xmm7, %xmm11, %xmm7 vmulps %xmm6, %xmm10, %xmm10 vbroadcastss 0xc0(%rax,%rdx,4), %xmm11 vaddps %xmm10, %xmm11, %xmm10 vmulps %xmm6, %xmm15, %xmm6 vmovaps %xmm4, %xmm15 vbroadcastss 0xd0(%rax,%rdx,4), %xmm11 vaddps %xmm6, %xmm11, %xmm6 vmovaps 0xa0(%rsp), %xmm11 vmulps 0x180(%rsp), %xmm11, %xmm4 vaddps %xmm7, %xmm4, %xmm4 vmulps %xmm11, %xmm9, %xmm3 vaddps %xmm3, %xmm10, %xmm3 vmulps %xmm5, %xmm11, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vmovaps 0x90(%rsp), %xmm6 vmulps %xmm6, %xmm12, %xmm7 vaddps %xmm4, %xmm7, %xmm4 vmulps %xmm6, %xmm13, %xmm7 vaddps %xmm3, %xmm7, %xmm3 vmulps %xmm6, %xmm14, %xmm6 vbroadcastss 0x1cbcfd2(%rip), %xmm14 # 0x1eeba20 vaddps %xmm5, %xmm6, %xmm5 vmovaps 0x1a0(%rsp), %xmm6 vsubps %xmm4, %xmm6, %xmm6 vmovaps 0x190(%rsp), %xmm7 vsubps %xmm4, %xmm7, %xmm4 vmulps %xmm6, %xmm2, %xmm6 vmulps %xmm4, %xmm2, %xmm2 vmovaps 0x40(%rsp), %xmm4 vsubps %xmm3, %xmm4, %xmm4 vmovaps 0x1b0(%rsp), %xmm7 vsubps %xmm3, %xmm7, %xmm3 vmulps %xmm4, %xmm1, %xmm4 vmulps %xmm3, %xmm1, %xmm1 vmovaps 0x10(%rsp), %xmm3 vsubps %xmm5, %xmm3, %xmm3 vmulps %xmm3, %xmm0, %xmm3 vmovaps 0x1c0(%rsp), %xmm7 vsubps %xmm5, %xmm7, %xmm5 vmulps %xmm5, %xmm0, %xmm0 vpminsd %xmm2, %xmm6, %xmm5 vpminsd %xmm1, %xmm4, %xmm7 vpmaxsd %xmm7, %xmm5, %xmm5 vpminsd %xmm0, %xmm3, %xmm7 vpmaxsd %xmm7, %xmm5, %xmm5 vpmaxsd %xmm2, %xmm6, %xmm2 vpmaxsd %xmm1, %xmm4, %xmm1 vpminsd %xmm1, %xmm2, %xmm1 vpmaxsd %xmm0, %xmm3, %xmm0 vpminsd %xmm0, %xmm1, %xmm0 vmulps %xmm5, %xmm15, %xmm2 vpmaxsd 0x150(%rsp), %xmm2, %xmm1 vbroadcastss 0x1cf1419(%rip), %xmm3 # 0x1f1ff14 vmulps %xmm3, %xmm0, %xmm0 vpminsd 0x160(%rsp), %xmm0, %xmm0 vcmpleps %xmm0, %xmm1, %xmm0 vandps 0x20(%rsp), %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22eb62 vblendvps %xmm0, %xmm2, %xmm14, %xmm0 cmpq $0x8, %rbp je 0x22eb42 movq %rbp, (%r12) addq $0x8, %r12 vmovaps 0x30(%rsp), %xmm1 vmovaps %xmm1, (%rbx) addq $0x10, %rbx vmovaps %xmm0, %xmm1 movq %rdi, %rbp cmpq $0x8, %rdi je 0x22eb6a leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x22e61d jmp 0x22eb6a vmovaps 0x30(%rsp), %xmm1 jmp 0x22eb49 xorl %eax, %eax cmpq $0x8, %rbp je 0x22ebac vmovaps 0x160(%rsp), %xmm0 vcmpnleps %xmm1, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r11 jae 0x22eb9c testb %cl, %cl jne 0x22e5ea jmp 0x22ecd6 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm1, (%rbx) addq $0x10, %rbx xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x22eb8f cmpq $-0x8, %rbp je 0x22ece4 movq %r11, 0x10(%rsp) movq %r9, 0x20(%rsp) movq %r8, (%rsp) vmovaps 0x160(%rsp), %xmm0 vcmpnleps %xmm1, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x22edc3 andq $-0x10, %rbp vmovaps 0x50(%rsp), %xmm1 movzbl (%rbp), %eax movq %rsi, 0x8(%rsp) movq 0x8(%rsi), %rcx vmovaps 0x170(%rsp), %xmm0 vmovaps %xmm0, 0x80(%rsp) vmovaps %xmm1, 0x40(%rsp) vmovmskps %xmm1, %edx xorl $0xf, %edx je 0x22ec69 movzbl %dl, %r15d shll $0x6, %eax addq %rcx, %rax addq $0x18, %rax movq %rax, 0x30(%rsp) bsfq %r15, %rdx leaq 0x1d0(%rsp), %rdi movq %r14, %rsi movq %rdx, %r13 movq %r10, %rcx movq %rbp, %r8 movq 0x30(%rsp), %rax callq *(%rax) testb %al, %al je 0x22ec5b orl $-0x1, 0x80(%rsp,%r13,4) leaq -0x1(%r15), %rax andq %rax, %r15 movq 0x68(%rsp), %r10 jne 0x22ec2f vmovaps 0x40(%rsp), %xmm0 vorps 0x80(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x50(%rsp) vtestps 0x1cbd199(%rip), %xmm0 # 0x1eebe20 jb 0x22ecae vmovaps 0x160(%rsp), %xmm1 vbroadcastss 0x1cbdee9(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x160(%rsp) xorl %eax, %eax jmp 0x22ecb1 pushq $0x3 popq %rax movq 0x8(%rsp), %rsi movq (%rsp), %r8 movq 0x20(%rsp), %r9 vbroadcastss 0x1cbcd58(%rip), %xmm14 # 0x1eeba20 movq 0x10(%rsp), %r11 vbroadcastss 0x1cf123a(%rip), %xmm15 # 0x1f1ff10 cmpl $0x3, %eax jne 0x22e598 jmp 0x22edcb pushq $0x3 jmp 0x22ecea pushq $0x2 popq %rax jmp 0x22ecd6 vmovaps %xmm1, 0x30(%rsp) movq %r11, 0x10(%rsp) movq %r9, 0x20(%rsp) movq %rsi, 0x8(%rsp) movq %r8, (%rsp) bsfq %r15, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, 0x40(%rsp) leaq 0x1d0(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x98(%rsp), %rax pushq %rax callq 0x272b3a popq %rcx popq %rdx testb %al, %al je 0x22ed43 movq 0x40(%rsp), %rax orl $-0x1, 0x50(%rsp,%rax,4) leaq -0x1(%r15), %rax andq %rax, %r15 movq 0x68(%rsp), %r10 movq 0x8(%rsp), %rsi movq (%rsp), %r8 jne 0x22ed06 vmovaps 0x50(%rsp), %xmm0 vtestps 0x1cbd0b7(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0x20(%rsp), %r9 vbroadcastss 0x1cbcca6(%rip), %xmm14 # 0x1eeba20 movq 0x10(%rsp), %r11 vbroadcastss 0x1cf1188(%rip), %xmm15 # 0x1f1ff10 vmovaps 0x30(%rsp), %xmm1 jb 0x22e5dd vmovaps 0x160(%rsp), %xmm1 vbroadcastss 0x1cbddde(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps 0x30(%rsp), %xmm1 vmovaps %xmm0, 0x160(%rsp) pushq $0x2 popq %rax jmp 0x22e5dd pushq $0x2 popq %rax jmp 0x22ecb6 vmovaps 0x70(%rsp), %xmm0 vandps 0x50(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cbdda4(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r9) addq $0x1948, %rsp # imm = 0x1948 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::SubdivPatch1IntersectorK<4>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x22efe0 movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 cmpq $0x0, 0x8(%rcx) je 0x22ee35 movq 0x10(%r15), %rax testb $0x1, 0x2(%rax) jne 0x22eff2 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x22efe0 movzbl %al, %ebp movq %rsp, %rax andq $0x0, (%rax) vmovaps (%r14), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%r14), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%r14), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%r14), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cf2023(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1cc213a(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1cbd858(%rip), %xmm7 # 0x1eec714 vdivps %xmm1, %xmm7, %xmm1 vandps %xmm4, %xmm2, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vdivps %xmm2, %xmm7, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vdivps %xmm3, %xmm7, %xmm3 vbroadcastss 0x1cf207d(%rip), %xmm6 # 0x1f20f60 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vblendvps %xmm8, %xmm6, %xmm2, %xmm2 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d2ba4b(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vcmpnltps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d25a75(%rip), %xmm5 # 0x1f549a0 vbroadcastss 0x1d2ba34(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm2, %xmm5, %xmm6, %xmm2 vmovaps %xmm1, 0x90(%rax) vmovaps %xmm2, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d2ba14(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d2ba0f(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%r14), %xmm1 vmovaps 0x80(%r14), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cbca91(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cbdbde(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi movq %rsp, %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x2751de popq %rax popq %rcx andq %rbx, %rbp jne 0x22efb4 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r12, %rsi movq %r14, %rdx movq %r15, %rcx addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x2732a0 nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::SubdivPatch1IntersectorK<4>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1b68, %rsp # imm = 0x1B68 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x230972 movq %rcx, %r10 movq %rdx, %r14 movq 0x8(%rcx), %rax testq %rax, %rax je 0x22f04e movq 0x10(%r10), %rcx testb $0x1, 0x2(%rcx) jne 0x230987 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%r14), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x230972 vandps %xmm3, %xmm4, %xmm10 andq $0x0, 0x100(%rsp) vmovaps (%r14), %xmm3 vmovaps %xmm3, 0x280(%rsp) vmovaps 0x10(%r14), %xmm3 vmovaps %xmm3, 0x290(%rsp) vmovaps 0x20(%r14), %xmm3 vmovaps %xmm3, 0x2a0(%rsp) vmovaps 0x40(%r14), %xmm3 vmovaps %xmm3, 0x2b0(%rsp) vmovaps 0x50(%r14), %xmm4 vmovaps %xmm4, 0x2c0(%rsp) vmovaps 0x60(%r14), %xmm5 vmovaps %xmm5, 0x2d0(%rsp) vbroadcastss 0x1cf1de2(%rip), %xmm11 # 0x1f20ec4 vandps %xmm3, %xmm11, %xmm6 vbroadcastss 0x1cc1ef9(%rip), %xmm9 # 0x1ef0fe8 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cbd616(%rip), %xmm12 # 0x1eec714 vdivps %xmm3, %xmm12, %xmm3 vandps %xmm4, %xmm11, %xmm7 vcmpltps %xmm9, %xmm7, %xmm7 vdivps %xmm4, %xmm12, %xmm4 vandps %xmm5, %xmm11, %xmm8 vcmpltps %xmm9, %xmm8, %xmm8 vbroadcastss 0x1cf1e3d(%rip), %xmm9 # 0x1f20f60 vblendvps %xmm6, %xmm9, %xmm3, %xmm3 vblendvps %xmm7, %xmm9, %xmm4, %xmm4 vdivps %xmm5, %xmm12, %xmm5 vblendvps %xmm8, %xmm9, %xmm5, %xmm5 vmovaps %xmm3, 0x2e0(%rsp) vmovaps %xmm4, 0x2f0(%rsp) vmovaps %xmm5, 0x300(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d2b802(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x310(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d25823(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d2b7e2(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vcmpnltps %xmm2, %xmm5, %xmm4 vbroadcastss 0x1d2b7d2(%rip), %xmm5 # 0x1f5a96c vbroadcastss 0x1d2b7cd(%rip), %xmm6 # 0x1f5a970 vblendvps %xmm4, %xmm5, %xmm6, %xmm4 vmovaps %xmm3, 0x320(%rsp) vmovaps %xmm4, 0x330(%rsp) vmovaps 0x30(%r14), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cbc84e(%rip), %xmm12 # 0x1eeba20 vblendvps %xmm10, %xmm3, %xmm12, %xmm1 vmovaps %xmm1, 0x340(%rsp) vbroadcastss 0x1cbd99a(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm10, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x350(%rsp) vmovaps %xmm10, 0x370(%rsp) vxorps %xmm0, %xmm10, %xmm0 vmovaps %xmm0, 0xe0(%rsp) testq %rax, %rax je 0x22f228 movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r9d, %r9d cmpb $0x1, %al adcq $0x2, %r9 jmp 0x22f22c pushq $0x3 popq %r9 leaq 0x80(%r14), %r11 leaq 0x4d0(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0xc70(%rsp), %r12 vmovaps %xmm12, -0x20(%r12) movq 0x70(%r8), %rax movq %rax, -0x8(%r15) vmovaps %xmm1, -0x10(%r12) leaq 0x1f20d18(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x360(%rsp) vbroadcastss 0x1cf0c92(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cf0c8d(%rip), %xmm13 # 0x1f1ff14 vpcmpeqd %xmm14, %xmm14, %xmm14 addq $-0x10, %r12 movq -0x8(%r15), %rbp addq $-0x8, %r15 cmpq $-0x8, %rbp je 0x23081c vmovaps (%r12), %xmm15 vcmpltps 0x350(%rsp), %xmm15, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x23082d movzbl %al, %r13d popcntl %r13d, %ebx xorl %eax, %eax cmpq %r9, %rbx jbe 0x230831 cmpq %r9, %rbx jbe 0x23081f testb $0x8, %bpl pushq $0x8 popq %rbx jne 0x22f475 movq %rbp, %rax movq %rbp, %rcx andq $-0x10, %rcx xorl %edx, %edx vmovaps %xmm12, %xmm15 movq %rbx, %rbp movq (%rcx,%rdx,8), %rdi cmpq $0x8, %rdi je 0x22f40c vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x280(%rsp), %xmm1 vmovaps 0x290(%rsp), %xmm2 vmovaps 0x2a0(%rsp), %xmm3 vmovaps 0x2e0(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0x2f0(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x300(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vminps %xmm1, %xmm5, %xmm0 vminps %xmm2, %xmm7, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vminps %xmm3, %xmm9, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vmaxps %xmm1, %xmm5, %xmm1 vmaxps %xmm2, %xmm7, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmaxps %xmm3, %xmm9, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmulps %xmm1, %xmm13, %xmm1 vmaxps 0x340(%rsp), %xmm0, %xmm2 vminps 0x350(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vtestps %xmm1, %xmm1 je 0x22f40c vblendvps %xmm1, %xmm0, %xmm12, %xmm0 cmpq $0x8, %rbp je 0x22f405 movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm15, (%r12) addq $0x10, %r12 movq %rdi, %rbp vmovaps %xmm0, %xmm15 cmpq $0x8, %rdi je 0x22f423 leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x22f2fc xorl %eax, %eax cmpq $0x8, %rbp je 0x22f46e vmovaps 0x350(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r9 jae 0x22f45d testb %cl, %cl je 0x23081f testb $0x8, %bpl je 0x22f2e8 jmp 0x22f475 movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm15, (%r12) addq $0x10, %r12 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x22f449 cmpq $-0x8, %rbp je 0x23081c vmovaps 0x350(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x23082d movq %rsi, 0xb8(%rsp) movq %r11, 0x8(%rsp) movq %r9, 0x10(%rsp) movq %r8, 0x18(%rsp) movl %ebp, %eax andl $0xf, %eax cmpl $0x8, %eax jne 0x230921 vpxor 0xe0(%rsp), %xmm14, %xmm9 movq 0x100(%rsp), %rsi movl 0x14(%rsi), %eax movl 0xc(%rsi), %ecx movl 0x10(%rsi), %ebx xorl %edx, %edx cmpq $0x2, %rcx setne %dl movl 0x24(%rsi), %r11d imulq $0xc, %rax, %rdi leaq 0x4(,%rdx,4), %rdx movq %rdx, 0x1e0(%rsp) shrq $0x4, %rbp leaq (,%rcx,4), %rdx movq %rdx, 0x1d0(%rsp) leaq (%r11,%rbp,4), %rdx addq %rdx, %rdi leaq (%rdi,%rcx,4), %r8 addq %rsi, %r8 addq $0x2c, %r8 movq %r8, 0xc0(%rsp) leaq (%rdx,%rax,8), %r8 leaq (%r8,%rcx,4), %r9 leaq 0x30(%rsi,%r9), %r9 movq %r9, 0xd8(%rsp) leaq (%rdx,%rax,4), %rax leaq (%rax,%rcx,4), %r9 leaq 0x30(%rsi,%r9), %r9 movq %r9, 0xd0(%rsp) leaq (%rdx,%rcx,4), %rcx leaq 0x30(%rsi,%rcx), %rcx movq %rcx, 0xc8(%rsp) leaq 0x30(%rsi,%rdi), %rcx movq %rcx, 0x98(%rsp) movl %ebx, %edi leaq 0x30(%rsi,%r8), %rcx movq %rcx, 0xf8(%rsp) leaq 0x30(%rsi,%rax), %r13 leaq 0x2c(%r11,%rbp,4), %rdx movq $0x0, 0xf0(%rsp) movq 0x18(%rsp), %r8 movq 0x10(%rsp), %r9 movq 0x8(%rsp), %r11 movq %rdx, 0x1d8(%rsp) leaq (%rsi,%rdx), %rax movq %rax, 0x1e8(%rsp) xorl %ebp, %ebp cmpq %rbp, 0x1e0(%rsp) je 0x230748 movq 0x1e8(%rsp), %rbx vbroadcastss (%rbx,%rbp), %xmm12 vbroadcastss -0x4(%r13,%rbp), %xmm0 movq 0xf8(%rsp), %rcx vbroadcastss -0x4(%rcx,%rbp), %xmm1 movq 0xc8(%rsp), %rax vbroadcastss -0x4(%rax,%rbp), %xmm14 movq 0xd0(%rsp), %rax vbroadcastss -0x4(%rax,%rbp), %xmm2 movq 0xd8(%rsp), %rax vbroadcastss -0x4(%rax,%rbp), %xmm6 vmovaps 0x10(%r14), %xmm3 vmovaps 0x20(%r14), %xmm8 vmovaps %xmm8, 0x60(%rsp) vsubps %xmm3, %xmm0, %xmm7 vsubps %xmm8, %xmm1, %xmm5 vmovaps %xmm2, 0x4a0(%rsp) vsubps %xmm3, %xmm2, %xmm4 vmovaps %xmm4, 0x80(%rsp) vmovaps %xmm6, 0x490(%rsp) vsubps %xmm8, %xmm6, %xmm2 vmovaps %xmm2, 0x70(%rsp) vsubps %xmm7, %xmm4, %xmm0 vsubps %xmm5, %xmm2, %xmm1 vaddps %xmm7, %xmm4, %xmm6 vaddps %xmm5, %xmm2, %xmm8 vmovdqa %xmm9, 0x20(%rsp) vmulps %xmm1, %xmm6, %xmm9 vmulps %xmm0, %xmm8, %xmm10 vsubps %xmm9, %xmm10, %xmm13 vmovaps (%r14), %xmm15 vsubps %xmm15, %xmm12, %xmm2 vmovaps %xmm14, 0x4b0(%rsp) vsubps %xmm15, %xmm14, %xmm12 vsubps %xmm2, %xmm12, %xmm4 vmulps %xmm4, %xmm8, %xmm8 vaddps %xmm2, %xmm12, %xmm10 vmovaps %xmm1, 0x1a0(%rsp) vmulps %xmm1, %xmm10, %xmm14 vsubps %xmm8, %xmm14, %xmm14 vmovaps %xmm0, 0x1c0(%rsp) vmulps %xmm0, %xmm10, %xmm8 vmovaps %xmm4, 0x190(%rsp) vmulps %xmm6, %xmm4, %xmm6 vsubps %xmm8, %xmm6, %xmm8 vmovaps 0x60(%r14), %xmm0 vmovaps %xmm0, 0x40(%rsp) vmulps %xmm0, %xmm8, %xmm10 vmovaps 0x50(%r14), %xmm11 vmulps %xmm14, %xmm11, %xmm14 vaddps %xmm14, %xmm10, %xmm14 vmovaps 0x40(%r14), %xmm4 vmulps %xmm4, %xmm13, %xmm13 vaddps %xmm14, %xmm13, %xmm10 vbroadcastss (%r13,%rbp), %xmm0 vmovaps %xmm0, 0x480(%rsp) vsubps %xmm3, %xmm0, %xmm0 vbroadcastss (%rcx,%rbp), %xmm1 vmovaps %xmm1, 0x470(%rsp) vsubps 0x60(%rsp), %xmm1, %xmm1 vsubps %xmm0, %xmm7, %xmm3 vsubps %xmm1, %xmm5, %xmm14 vmovaps %xmm7, 0x60(%rsp) vaddps %xmm0, %xmm7, %xmm7 vmovaps %xmm5, 0x180(%rsp) vaddps %xmm1, %xmm5, %xmm5 vmulps %xmm7, %xmm14, %xmm13 vmulps %xmm5, %xmm3, %xmm6 vsubps %xmm13, %xmm6, %xmm6 vbroadcastss 0x4(%rbx,%rbp), %xmm8 vmovaps %xmm8, 0x460(%rsp) vsubps %xmm15, %xmm8, %xmm9 vsubps %xmm9, %xmm2, %xmm13 vmulps %xmm5, %xmm13, %xmm5 vmovaps %xmm2, 0x1b0(%rsp) vaddps %xmm2, %xmm9, %xmm15 vmulps %xmm14, %xmm15, %xmm8 vsubps %xmm5, %xmm8, %xmm5 vmovaps %xmm3, 0x170(%rsp) vmulps %xmm3, %xmm15, %xmm8 vmulps %xmm7, %xmm13, %xmm7 vsubps %xmm8, %xmm7, %xmm7 vmovaps 0x40(%rsp), %xmm3 vmulps %xmm3, %xmm7, %xmm7 vmulps %xmm5, %xmm11, %xmm5 vaddps %xmm5, %xmm7, %xmm5 vmulps %xmm6, %xmm4, %xmm6 vaddps %xmm5, %xmm6, %xmm7 vsubps %xmm12, %xmm9, %xmm15 vaddps %xmm12, %xmm9, %xmm5 vmovaps 0x20(%rsp), %xmm8 vmovaps 0x80(%rsp), %xmm2 vsubps %xmm2, %xmm0, %xmm12 vaddps %xmm2, %xmm0, %xmm0 vmovaps 0x70(%rsp), %xmm2 vsubps %xmm2, %xmm1, %xmm9 vaddps %xmm2, %xmm1, %xmm1 vmulps %xmm0, %xmm9, %xmm2 vmulps %xmm1, %xmm12, %xmm6 vsubps %xmm2, %xmm6, %xmm2 vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm5, %xmm9, %xmm6 vsubps %xmm1, %xmm6, %xmm1 vbroadcastss 0x1cf16b7(%rip), %xmm6 # 0x1f20ec4 vmulps %xmm5, %xmm12, %xmm5 vmulps %xmm0, %xmm15, %xmm0 vsubps %xmm5, %xmm0, %xmm0 vmulps %xmm3, %xmm0, %xmm0 vmovaps %xmm11, 0x70(%rsp) vmulps %xmm1, %xmm11, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmovaps %xmm4, 0x80(%rsp) vmulps %xmm2, %xmm4, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm7, %xmm10, %xmm1 vaddps %xmm1, %xmm0, %xmm4 vandps %xmm6, %xmm4, %xmm1 vbroadcastss 0x1cf167b(%rip), %xmm2 # 0x1f20ecc vmulps %xmm2, %xmm1, %xmm1 vminps %xmm7, %xmm10, %xmm2 vminps %xmm0, %xmm2, %xmm2 vbroadcastss 0x1cf165a(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm1, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm10, 0x160(%rsp) vmaxps %xmm7, %xmm10, %xmm5 vmaxps %xmm0, %xmm5, %xmm0 vcmpleps %xmm1, %xmm0, %xmm0 vorps %xmm0, %xmm2, %xmm0 vtestps %xmm8, %xmm0 vandps %xmm0, %xmm8, %xmm0 je 0x230693 vmovaps %xmm4, 0x140(%rsp) vmovaps 0x1a0(%rsp), %xmm10 vmovaps 0x170(%rsp), %xmm3 vmovaps %xmm0, 0x150(%rsp) vmulps %xmm3, %xmm10, %xmm0 vmovaps 0x1c0(%rsp), %xmm8 vmulps %xmm14, %xmm8, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm14, %xmm12, %xmm2 vmulps %xmm3, %xmm9, %xmm5 vsubps %xmm2, %xmm5, %xmm5 vandps %xmm6, %xmm0, %xmm0 vandps %xmm6, %xmm2, %xmm2 vcmpltps %xmm2, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm5, %xmm2 vmulps %xmm9, %xmm13, %xmm0 vmulps %xmm10, %xmm13, %xmm1 vmovaps 0x190(%rsp), %xmm9 vmulps %xmm14, %xmm9, %xmm4 vsubps %xmm4, %xmm1, %xmm1 vmulps %xmm14, %xmm15, %xmm5 vsubps %xmm0, %xmm5, %xmm5 vandps %xmm6, %xmm4, %xmm4 vandps %xmm6, %xmm0, %xmm0 vcmpltps %xmm0, %xmm4, %xmm0 vblendvps %xmm0, %xmm1, %xmm5, %xmm4 vmulps %xmm3, %xmm15, %xmm0 vmulps %xmm3, %xmm9, %xmm1 vmulps %xmm8, %xmm13, %xmm5 vmovaps %xmm6, %xmm11 vmulps %xmm12, %xmm13, %xmm6 vsubps %xmm5, %xmm1, %xmm1 vsubps %xmm0, %xmm6, %xmm6 vandps %xmm5, %xmm11, %xmm5 vandps %xmm0, %xmm11, %xmm0 vcmpltps %xmm0, %xmm5, %xmm0 vblendvps %xmm0, %xmm1, %xmm6, %xmm5 vmulps 0x40(%rsp), %xmm5, %xmm0 vmulps 0x70(%rsp), %xmm4, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps 0x80(%rsp), %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vmulps 0x180(%rsp), %xmm5, %xmm1 vmulps 0x60(%rsp), %xmm4, %xmm6 vaddps %xmm6, %xmm1, %xmm1 vmulps 0x1b0(%rsp), %xmm2, %xmm3 vaddps %xmm1, %xmm3, %xmm1 vaddps %xmm1, %xmm1, %xmm1 vrcpps %xmm0, %xmm3 vmulps %xmm3, %xmm0, %xmm6 vbroadcastss 0x1cbcd63(%rip), %xmm8 # 0x1eec714 vsubps %xmm6, %xmm8, %xmm6 vmulps %xmm6, %xmm3, %xmm6 vaddps %xmm6, %xmm3, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vmovaps 0x30(%r14), %xmm3 vcmpleps 0x80(%r14), %xmm1, %xmm6 vcmpleps %xmm1, %xmm3, %xmm3 vandps %xmm6, %xmm3, %xmm3 vxorps %xmm8, %xmm8, %xmm8 vcmpneqps %xmm0, %xmm8, %xmm0 vandps %xmm3, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm0 vmovaps 0x150(%rsp), %xmm3 vtestps %xmm3, %xmm0 vandps %xmm3, %xmm0, %xmm0 je 0x22fa57 vmovaps 0x160(%rsp), %xmm3 vmovaps %xmm3, 0x450(%rsp) vmovaps %xmm7, 0x440(%rsp) vmovaps 0x140(%rsp), %xmm3 vmovaps %xmm3, 0x430(%rsp) vmovaps %xmm1, 0x420(%rsp) vmovaps %xmm2, 0x410(%rsp) vmovaps %xmm4, 0x400(%rsp) vmovaps %xmm5, 0x3f0(%rsp) vbroadcastss 0x1cbbfc0(%rip), %xmm12 # 0x1eeba20 vbroadcastss 0x1cf04a7(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cf04a2(%rip), %xmm13 # 0x1f1ff14 vmovdqa 0x20(%rsp), %xmm9 movq 0xc8(%rsp), %rax vbroadcastss (%rax,%rbp), %xmm1 vmovaps %xmm1, 0x40(%rsp) movq 0xd0(%rsp), %rax vbroadcastss (%rax,%rbp), %xmm2 movq 0xd8(%rsp), %rax vbroadcastss (%rax,%rbp), %xmm14 movq 0x100(%rsp), %rcx movl 0x18(%rcx), %eax movq (%r10), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rax,8), %rbx vbroadcastss 0x34(%rbx), %xmm1 vandps 0x90(%r14), %xmm1, %xmm1 vpcmpeqd %xmm1, %xmm8, %xmm1 vtestps %xmm0, %xmm1 jb 0x22fb01 movl 0x1c(%rcx), %ecx vandnps %xmm0, %xmm1, %xmm0 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x22ffe2 cmpq $0x0, 0x48(%rbx) jne 0x22ffe2 vpandn %xmm9, %xmm0, %xmm9 vtestps %xmm9, %xmm9 je 0x230743 vmovaps (%r14), %xmm0 vmovaps 0x10(%r14), %xmm5 vmovaps 0x20(%r14), %xmm6 vmovaps 0x4b0(%rsp), %xmm1 vsubps %xmm0, %xmm1, %xmm7 vmovaps 0x4a0(%rsp), %xmm1 vsubps %xmm5, %xmm1, %xmm13 vmovaps 0x490(%rsp), %xmm1 vsubps %xmm6, %xmm1, %xmm15 vmovaps 0x460(%rsp), %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmovaps %xmm1, 0x70(%rsp) vmovaps 0x480(%rsp), %xmm1 vsubps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x1c0(%rsp) vmovaps 0x470(%rsp), %xmm1 vsubps %xmm6, %xmm1, %xmm1 vmovaps %xmm1, 0x60(%rsp) vmovaps 0x40(%rsp), %xmm1 vsubps %xmm0, %xmm1, %xmm0 vsubps %xmm5, %xmm2, %xmm5 vsubps %xmm6, %xmm14, %xmm1 vmovaps %xmm1, 0x40(%rsp) vsubps %xmm7, %xmm0, %xmm2 vmovaps %xmm0, %xmm3 vmovaps %xmm0, 0x80(%rsp) vsubps %xmm13, %xmm5, %xmm14 vsubps %xmm15, %xmm1, %xmm0 vaddps %xmm5, %xmm13, %xmm6 vaddps %xmm1, %xmm15, %xmm8 vmovaps %xmm9, 0x20(%rsp) vmulps %xmm0, %xmm6, %xmm9 vmulps %xmm8, %xmm14, %xmm10 vsubps %xmm9, %xmm10, %xmm12 vaddps %xmm7, %xmm3, %xmm9 vmulps %xmm2, %xmm8, %xmm8 vmovaps %xmm0, 0x160(%rsp) vmulps %xmm0, %xmm9, %xmm10 vsubps %xmm8, %xmm10, %xmm10 vmovaps %xmm14, 0x170(%rsp) vmulps %xmm14, %xmm9, %xmm8 vmovaps %xmm2, 0x180(%rsp) vmulps %xmm6, %xmm2, %xmm6 vsubps %xmm8, %xmm6, %xmm6 vmovaps 0x60(%r14), %xmm4 vmulps %xmm4, %xmm6, %xmm6 vmovaps 0x50(%r14), %xmm14 vmulps %xmm10, %xmm14, %xmm10 vaddps %xmm6, %xmm10, %xmm6 vmovaps 0x40(%r14), %xmm11 vmulps %xmm12, %xmm11, %xmm12 vaddps %xmm6, %xmm12, %xmm9 vmovaps 0x1c0(%rsp), %xmm3 vsubps %xmm3, %xmm13, %xmm1 vmovaps 0x60(%rsp), %xmm12 vsubps %xmm12, %xmm15, %xmm10 vmovaps %xmm13, 0x1a0(%rsp) vaddps %xmm3, %xmm13, %xmm6 vmovaps %xmm15, 0x190(%rsp) vaddps %xmm12, %xmm15, %xmm15 vmulps %xmm6, %xmm10, %xmm13 vmulps %xmm1, %xmm15, %xmm2 vsubps %xmm13, %xmm2, %xmm2 vmovaps 0x70(%rsp), %xmm0 vsubps %xmm0, %xmm7, %xmm13 vmulps %xmm15, %xmm13, %xmm15 vmovaps %xmm7, 0x1b0(%rsp) vaddps %xmm0, %xmm7, %xmm7 vmulps %xmm7, %xmm10, %xmm8 vsubps %xmm15, %xmm8, %xmm8 vmovaps %xmm1, 0x150(%rsp) vmulps %xmm1, %xmm7, %xmm7 vmulps %xmm6, %xmm13, %xmm6 vsubps %xmm7, %xmm6, %xmm6 vmulps %xmm4, %xmm6, %xmm6 vmulps %xmm8, %xmm14, %xmm7 vaddps %xmm7, %xmm6, %xmm6 vmulps %xmm2, %xmm11, %xmm2 vaddps %xmm6, %xmm2, %xmm7 vmovaps %xmm7, 0x140(%rsp) vmovaps 0x80(%rsp), %xmm1 vsubps %xmm1, %xmm0, %xmm15 vaddps %xmm1, %xmm0, %xmm1 vsubps %xmm5, %xmm3, %xmm8 vaddps %xmm5, %xmm3, %xmm0 vmovaps 0x40(%rsp), %xmm2 vsubps %xmm2, %xmm12, %xmm6 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm6, %xmm0, %xmm3 vmulps %xmm2, %xmm8, %xmm5 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm2, %xmm15, %xmm2 vmulps %xmm6, %xmm1, %xmm5 vsubps %xmm2, %xmm5, %xmm2 vbroadcastss 0x1cf11b3(%rip), %xmm5 # 0x1f20ec4 vmulps %xmm1, %xmm8, %xmm1 vmulps %xmm0, %xmm15, %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmovaps %xmm4, 0x40(%rsp) vmulps %xmm4, %xmm0, %xmm0 vmovaps %xmm14, 0x70(%rsp) vmulps %xmm2, %xmm14, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmovaps %xmm11, 0x80(%rsp) vmulps %xmm3, %xmm11, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm7, %xmm9, %xmm1 vaddps %xmm1, %xmm0, %xmm11 vandps %xmm5, %xmm11, %xmm1 vbroadcastss 0x1cf1171(%rip), %xmm2 # 0x1f20ecc vmulps %xmm2, %xmm1, %xmm1 vminps %xmm7, %xmm9, %xmm2 vminps %xmm0, %xmm2, %xmm2 vbroadcastss 0x1cf1150(%rip), %xmm3 # 0x1f20ec0 vxorps %xmm3, %xmm1, %xmm3 vcmpnltps %xmm3, %xmm2, %xmm2 vmovaps %xmm9, 0x60(%rsp) vmaxps %xmm7, %xmm9, %xmm3 vmaxps %xmm0, %xmm3, %xmm0 vcmpleps %xmm1, %xmm0, %xmm0 vorps %xmm0, %xmm2, %xmm0 vmovaps 0x20(%rsp), %xmm1 vtestps %xmm1, %xmm0 vandps %xmm1, %xmm0, %xmm9 je 0x2306bc vmovaps 0x160(%rsp), %xmm7 vmovaps 0x150(%rsp), %xmm12 vmulps %xmm7, %xmm12, %xmm0 vmovaps %xmm8, %xmm14 vmovaps 0x170(%rsp), %xmm8 vmulps %xmm10, %xmm8, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm10, %xmm14, %xmm2 vmulps %xmm6, %xmm12, %xmm3 vsubps %xmm2, %xmm3, %xmm3 vandps %xmm5, %xmm0, %xmm0 vandps %xmm5, %xmm2, %xmm2 vcmpltps %xmm2, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm3, %xmm3 vmulps %xmm6, %xmm13, %xmm0 vmulps %xmm7, %xmm13, %xmm1 vmovaps 0x180(%rsp), %xmm7 vmulps %xmm7, %xmm10, %xmm2 vsubps %xmm2, %xmm1, %xmm1 vmulps %xmm10, %xmm15, %xmm4 vsubps %xmm0, %xmm4, %xmm4 vandps %xmm5, %xmm2, %xmm2 vandps %xmm5, %xmm0, %xmm0 vcmpltps %xmm0, %xmm2, %xmm0 vblendvps %xmm0, %xmm1, %xmm4, %xmm1 vmulps %xmm12, %xmm15, %xmm0 vmulps %xmm7, %xmm12, %xmm2 vmulps %xmm8, %xmm13, %xmm4 vmovaps %xmm5, %xmm12 vmulps %xmm14, %xmm13, %xmm5 vsubps %xmm4, %xmm2, %xmm2 vsubps %xmm0, %xmm5, %xmm5 vandps %xmm4, %xmm12, %xmm4 vandps %xmm0, %xmm12, %xmm0 vcmpltps %xmm0, %xmm4, %xmm0 vblendvps %xmm0, %xmm2, %xmm5, %xmm4 vmulps 0x40(%rsp), %xmm4, %xmm0 vmulps 0x70(%rsp), %xmm1, %xmm2 vaddps %xmm2, %xmm0, %xmm0 vmulps 0x80(%rsp), %xmm3, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vmulps 0x190(%rsp), %xmm4, %xmm2 vmulps 0x1a0(%rsp), %xmm1, %xmm5 vaddps %xmm5, %xmm2, %xmm2 vmulps 0x1b0(%rsp), %xmm3, %xmm5 vaddps %xmm2, %xmm5, %xmm2 vaddps %xmm2, %xmm2, %xmm2 vrcpps %xmm0, %xmm5 vmulps %xmm5, %xmm0, %xmm7 vbroadcastss 0x1cbc862(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm5, %xmm7 vaddps %xmm7, %xmm5, %xmm5 vmulps %xmm5, %xmm2, %xmm2 vmovaps 0x30(%r14), %xmm5 vcmpleps 0x80(%r14), %xmm2, %xmm7 vcmpleps %xmm2, %xmm5, %xmm5 vandps %xmm7, %xmm5, %xmm5 vxorps %xmm7, %xmm7, %xmm7 vcmpneqps %xmm7, %xmm0, %xmm0 vandps %xmm5, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm0 vtestps %xmm9, %xmm0 vandps %xmm0, %xmm9, %xmm0 vmovaps %xmm12, %xmm6 je 0x22ff4f vmovaps 0x60(%rsp), %xmm5 vmovaps %xmm5, 0x3e0(%rsp) vmovaps 0x140(%rsp), %xmm5 vmovaps %xmm5, 0x3d0(%rsp) vmovaps %xmm11, 0x3c0(%rsp) vmovaps %xmm2, 0x3b0(%rsp) vmovaps %xmm3, 0x3a0(%rsp) vmovaps %xmm1, 0x390(%rsp) vmovaps %xmm4, 0x380(%rsp) vbroadcastss 0x1cbbac8(%rip), %xmm12 # 0x1eeba20 vbroadcastss 0x1ceffaf(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1ceffaa(%rip), %xmm13 # 0x1f1ff14 vpcmpeqd %xmm14, %xmm14, %xmm14 vmovdqa 0x20(%rsp), %xmm9 movq 0x100(%rsp), %rcx movl 0x18(%rcx), %eax movq (%r10), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rax,8), %rbx vbroadcastss 0x34(%rbx), %xmm1 vandps 0x90(%r14), %xmm1, %xmm1 vpcmpeqd %xmm7, %xmm1, %xmm1 vtestps %xmm0, %xmm1 jb 0x22ffce movl 0x1c(%rcx), %ecx vandnps %xmm0, %xmm1, %xmm0 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x23034f cmpq $0x0, 0x48(%rbx) jne 0x23034f vpandn %xmm9, %xmm0, %xmm9 addq $0x4, %rbp vtestps %xmm9, %xmm9 jne 0x22f5b9 jmp 0x230748 vmovaps %xmm14, 0x70(%rsp) vmovaps %xmm2, 0x80(%rsp) vmovaps 0x430(%rsp), %xmm3 vandps %xmm3, %xmm11, %xmm1 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cbc705(%rip), %xmm6 # 0x1eec714 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cc0fc4(%rip), %xmm3 # 0x1ef0fe8 vcmpnltps %xmm3, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm2 vmulps 0x450(%rsp), %xmm2, %xmm1 vminps %xmm6, %xmm1, %xmm1 vmulps 0x440(%rsp), %xmm2, %xmm2 vminps %xmm6, %xmm2, %xmm2 movq 0x98(%rsp), %rdx vbroadcastss -0x4(%rdx,%rbp), %xmm3 movq %r11, %rdx movq 0xc0(%rsp), %r11 vbroadcastss (%r11,%rbp), %xmm4 movq %rdx, %r11 vpsrld $0x10, %xmm3, %xmm5 vpblendw $0xaa, %xmm8, %xmm3, %xmm3 # xmm3 = xmm3[0],xmm8[1],xmm3[2],xmm8[3],xmm3[4],xmm8[5],xmm3[6],xmm8[7] vcvtdq2ps %xmm3, %xmm3 vbroadcastss 0x1cee45e(%rip), %xmm7 # 0x1f1e4e0 vmulps %xmm7, %xmm3, %xmm3 vcvtdq2ps %xmm5, %xmm5 vmulps %xmm7, %xmm5, %xmm5 vsubps %xmm1, %xmm6, %xmm6 vsubps %xmm2, %xmm6, %xmm6 vmulps %xmm3, %xmm6, %xmm3 vmulps %xmm5, %xmm6, %xmm5 vpblendw $0xaa, %xmm8, %xmm4, %xmm6 # xmm6 = xmm4[0],xmm8[1],xmm4[2],xmm8[3],xmm4[4],xmm8[5],xmm4[6],xmm8[7] vcvtdq2ps %xmm6, %xmm6 vmulps %xmm7, %xmm6, %xmm6 vmulps %xmm6, %xmm2, %xmm6 vaddps %xmm6, %xmm3, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm2, %xmm2 movq 0x98(%rsp), %rdx vbroadcastss (%rdx,%rbp), %xmm4 vaddps %xmm2, %xmm5, %xmm2 vpblendw $0xaa, %xmm8, %xmm4, %xmm5 # xmm5 = xmm4[0],xmm8[1],xmm4[2],xmm8[3],xmm4[4],xmm8[5],xmm4[6],xmm8[7] vcvtdq2ps %xmm5, %xmm5 vmulps %xmm7, %xmm5, %xmm5 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm7, %xmm4, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vaddps %xmm2, %xmm1, %xmm1 movq 0x8(%r10), %rdx vmovaps 0x410(%rsp), %xmm2 vmovaps %xmm2, 0x1f0(%rsp) vmovaps 0x400(%rsp), %xmm2 vmovaps %xmm2, 0x200(%rsp) vmovaps 0x3f0(%rsp), %xmm2 vmovaps %xmm2, 0x210(%rsp) vmovaps %xmm3, 0x220(%rsp) vmovd %eax, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmovaps %xmm1, 0x230(%rsp) vmovd %ecx, %xmm1 vpshufd $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmovdqa %xmm1, 0x240(%rsp) vmovdqa %xmm2, 0x250(%rsp) leaq 0x260(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x260(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x270(%rsp) vmovaps 0x80(%r14), %xmm3 vblendvps %xmm0, 0x420(%rsp), %xmm3, %xmm1 vmovaps %xmm1, 0x80(%r14) vmovaps %xmm0, 0xa0(%rsp) leaq 0xa0(%rsp), %rax movq %rax, 0x110(%rsp) movq 0x18(%rbx), %rax movq %rax, 0x118(%rsp) movq 0x8(%r10), %rax movq %rax, 0x120(%rsp) movq %r14, 0x128(%rsp) leaq 0x1f0(%rsp), %rax movq %rax, 0x130(%rsp) movl $0x4, 0x138(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax movq %r10, 0x38(%rsp) movq %rsi, 0x58(%rsp) movl %edi, 0x4(%rsp) vmovaps %xmm3, 0x60(%rsp) je 0x230284 leaq 0x110(%rsp), %rdi vzeroupper callq %rax vmovaps 0x60(%rsp), %xmm3 movl 0x4(%rsp), %edi vmovdqa 0x20(%rsp), %xmm9 vxorps %xmm8, %xmm8, %xmm8 vbroadcastss 0x1cefcbb(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cefcae(%rip), %xmm10 # 0x1f1ff10 movq 0x8(%rsp), %r11 movq 0x10(%rsp), %r9 vbroadcastss 0x1cbb7ab(%rip), %xmm12 # 0x1eeba20 movq 0x18(%rsp), %r8 movq 0x58(%rsp), %rsi movq 0x38(%rsp), %r10 vmovdqa 0xa0(%rsp), %xmm0 vptest %xmm0, %xmm0 vmovaps 0x70(%rsp), %xmm14 je 0x2306ed movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax vpcmpeqd %xmm0, %xmm0, %xmm0 je 0x23031a testb $0x2, (%rcx) jne 0x2302ba testb $0x40, 0x3e(%rbx) je 0x23031a leaq 0x110(%rsp), %rdi vzeroupper callq %rax vmovaps 0x60(%rsp), %xmm3 vmovaps 0x70(%rsp), %xmm14 movl 0x4(%rsp), %edi vmovdqa 0x20(%rsp), %xmm9 vxorps %xmm8, %xmm8, %xmm8 vpcmpeqd %xmm0, %xmm0, %xmm0 vbroadcastss 0x1cefc25(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cefc18(%rip), %xmm10 # 0x1f1ff10 movq 0x8(%rsp), %r11 movq 0x10(%rsp), %r9 vbroadcastss 0x1cbb715(%rip), %xmm12 # 0x1eeba20 movq 0x18(%rsp), %r8 movq 0x58(%rsp), %rsi movq 0x38(%rsp), %r10 vpcmpeqd 0xa0(%rsp), %xmm8, %xmm1 vpxor %xmm0, %xmm1, %xmm0 movq 0x128(%rsp), %rax vbroadcastss 0x1cbc84c(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x2306f9 vmovaps 0x3c0(%rsp), %xmm3 vandps %xmm6, %xmm3, %xmm1 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cbc3a7(%rip), %xmm6 # 0x1eec714 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cc0c66(%rip), %xmm3 # 0x1ef0fe8 vcmpnltps %xmm3, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm2 vmulps 0x3e0(%rsp), %xmm2, %xmm1 vminps %xmm6, %xmm1, %xmm1 vmulps 0x3d0(%rsp), %xmm2, %xmm2 vminps %xmm6, %xmm2, %xmm2 movq 0xc0(%rsp), %rdx vbroadcastss (%rdx,%rbp), %xmm3 vbroadcastss 0x4(%rdx,%rbp), %xmm4 vpsrld $0x10, %xmm3, %xmm5 vpblendw $0xaa, %xmm7, %xmm3, %xmm3 # xmm3 = xmm3[0],xmm7[1],xmm3[2],xmm7[3],xmm3[4],xmm7[5],xmm3[6],xmm7[7] vcvtdq2ps %xmm3, %xmm3 vbroadcastss 0x1cee10e(%rip), %xmm8 # 0x1f1e4e0 vmulps %xmm3, %xmm8, %xmm3 vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vsubps %xmm1, %xmm6, %xmm6 vsubps %xmm2, %xmm6, %xmm6 vmulps %xmm3, %xmm6, %xmm3 vmulps %xmm5, %xmm6, %xmm5 vpblendw $0xaa, %xmm7, %xmm4, %xmm6 # xmm6 = xmm4[0],xmm7[1],xmm4[2],xmm7[3],xmm4[4],xmm7[5],xmm4[6],xmm7[7] vcvtdq2ps %xmm6, %xmm6 vmulps %xmm6, %xmm8, %xmm6 vmulps %xmm6, %xmm2, %xmm6 vaddps %xmm6, %xmm3, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm2, %xmm2 movq 0x98(%rsp), %rdx vbroadcastss (%rdx,%rbp), %xmm4 vaddps %xmm2, %xmm5, %xmm2 vpblendw $0xaa, %xmm7, %xmm4, %xmm5 # xmm5 = xmm4[0],xmm7[1],xmm4[2],xmm7[3],xmm4[4],xmm7[5],xmm4[6],xmm7[7] vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vaddps %xmm2, %xmm1, %xmm1 movq 0x8(%r10), %rdx vmovaps 0x3a0(%rsp), %xmm2 vmovaps %xmm2, 0x1f0(%rsp) vmovaps 0x390(%rsp), %xmm2 vmovaps %xmm2, 0x200(%rsp) vmovaps 0x380(%rsp), %xmm2 vmovaps %xmm2, 0x210(%rsp) vmovaps %xmm3, 0x220(%rsp) vmovd %eax, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmovaps %xmm1, 0x230(%rsp) vmovd %ecx, %xmm1 vpshufd $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmovdqa %xmm1, 0x240(%rsp) vmovdqa %xmm2, 0x250(%rsp) leaq 0x260(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x260(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x270(%rsp) vmovaps 0x80(%r14), %xmm3 vblendvps %xmm0, 0x3b0(%rsp), %xmm3, %xmm1 vmovaps %xmm1, 0x80(%r14) vmovaps %xmm0, 0xa0(%rsp) leaq 0xa0(%rsp), %rax movq %rax, 0x110(%rsp) movq 0x18(%rbx), %rax movq %rax, 0x118(%rsp) movq 0x8(%r10), %rax movq %rax, 0x120(%rsp) movq %r14, 0x128(%rsp) leaq 0x1f0(%rsp), %rax movq %rax, 0x130(%rsp) movl $0x4, 0x138(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax movq %r10, 0x38(%rsp) movq %rsi, 0x58(%rsp) movl %edi, 0x4(%rsp) vmovaps %xmm3, 0x40(%rsp) je 0x2305d3 leaq 0x110(%rsp), %rdi vzeroupper callq %rax vmovaps 0x40(%rsp), %xmm3 movl 0x4(%rsp), %edi vmovdqa 0x20(%rsp), %xmm9 vxorps %xmm7, %xmm7, %xmm7 vbroadcastss 0x1cef96c(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cef95f(%rip), %xmm10 # 0x1f1ff10 movq 0x8(%rsp), %r11 movq 0x10(%rsp), %r9 vbroadcastss 0x1cbb45c(%rip), %xmm12 # 0x1eeba20 movq 0x18(%rsp), %r8 movq 0x58(%rsp), %rsi movq 0x38(%rsp), %r10 vmovdqa 0xa0(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x23071c movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax vpcmpeqd %xmm14, %xmm14, %xmm14 je 0x23065e testb $0x2, (%rcx) jne 0x230604 testb $0x40, 0x3e(%rbx) je 0x23065e leaq 0x110(%rsp), %rdi vzeroupper callq %rax vmovaps 0x40(%rsp), %xmm3 movl 0x4(%rsp), %edi vmovdqa 0x20(%rsp), %xmm9 vxorps %xmm7, %xmm7, %xmm7 vpcmpeqd %xmm14, %xmm14, %xmm14 vbroadcastss 0x1cef8e1(%rip), %xmm13 # 0x1f1ff14 vbroadcastss 0x1cef8d4(%rip), %xmm10 # 0x1f1ff10 movq 0x8(%rsp), %r11 movq 0x10(%rsp), %r9 vbroadcastss 0x1cbb3d1(%rip), %xmm12 # 0x1eeba20 movq 0x18(%rsp), %r8 movq 0x58(%rsp), %rsi movq 0x38(%rsp), %r10 vpcmpeqd 0xa0(%rsp), %xmm7, %xmm1 vpxor %xmm1, %xmm14, %xmm0 movq 0x128(%rsp), %rax vbroadcastss 0x1cbc508(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x230729 vmovaps %xmm6, %xmm11 vbroadcastss 0x1cbb380(%rip), %xmm12 # 0x1eeba20 vbroadcastss 0x1cef867(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cef862(%rip), %xmm13 # 0x1f1ff14 vxorps %xmm8, %xmm8, %xmm8 jmp 0x22fa72 vmovaps %xmm9, %xmm0 vmovaps %xmm5, %xmm6 vbroadcastss 0x1cbb353(%rip), %xmm12 # 0x1eeba20 vbroadcastss 0x1cef83a(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cef835(%rip), %xmm13 # 0x1f1ff14 vpcmpeqd %xmm14, %xmm14, %xmm14 vxorps %xmm7, %xmm7, %xmm7 jmp 0x22ff6f vpcmpeqd %xmm0, %xmm8, %xmm0 vpcmpeqd %xmm1, %xmm1, %xmm1 vpxor %xmm1, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm1 vpsrad $0x1f, %xmm1, %xmm0 vblendvps %xmm1, (%r11), %xmm3, %xmm1 vmovaps %xmm1, (%r11) vmovaps 0x80(%rsp), %xmm2 jmp 0x22fafc vpcmpeqd %xmm7, %xmm0, %xmm0 vpcmpeqd %xmm14, %xmm14, %xmm14 vpxor %xmm0, %xmm14, %xmm0 vpslld $0x1f, %xmm0, %xmm1 vpsrad $0x1f, %xmm1, %xmm0 vblendvps %xmm1, (%r11), %xmm3, %xmm1 vmovaps %xmm1, (%r11) jmp 0x22ffc9 vpcmpeqd %xmm14, %xmm14, %xmm14 cmpl $0x2, %edi movq 0x1d8(%rsp), %rdx je 0x2307b0 movq 0xf0(%rsp), %rbx leaq 0x1(%rbx), %rax movq 0x1d0(%rsp), %rcx addq %rcx, 0xc0(%rsp) addq %rcx, 0xd8(%rsp) addq %rcx, 0xd0(%rsp) addq %rcx, 0xc8(%rsp) addq %rcx, 0x98(%rsp) addq %rcx, 0xf8(%rsp) addq %rcx, %r13 addq %rcx, %rdx testq %rbx, %rbx movq %rax, 0xf0(%rsp) je 0x22f5a3 vpxor %xmm14, %xmm9, %xmm0 xorl %ecx, %ecx vpor 0xe0(%rsp), %xmm0, %xmm0 vmovdqa %xmm0, 0xe0(%rsp) vtestps %xmm14, %xmm0 movq 0xb8(%rsp), %rsi jb 0x23081c vmovaps 0x350(%rsp), %xmm1 vbroadcastss 0x1cbc39a(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x350(%rsp) xorl %eax, %eax testq %rcx, %rcx je 0x23081f movq %rcx, (%r15) addq $0x8, %r15 vbroadcastss 0x1cbc374(%rip), %xmm0 # 0x1eecb84 vmovaps %xmm0, (%r12) addq $0x10, %r12 jmp 0x23081f pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x22f28c jmp 0x230952 pushq $0x2 jmp 0x23081e vmovaps %xmm15, 0x40(%rsp) movq %r11, 0x8(%rsp) movq %r9, 0x10(%rsp) movq %r10, 0x38(%rsp) movq %rsi, 0xb8(%rsp) movq %r8, 0x18(%rsp) bsfq %r13, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, 0x20(%rsp) leaq 0x100(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x288(%rsp), %rax pushq %rax vzeroupper callq 0x2780d4 popq %rcx popq %rdx testb %al, %al je 0x230896 movq 0x20(%rsp), %rax orl $-0x1, 0xe0(%rsp,%rax,4) leaq -0x1(%r13), %rax andq %rax, %r13 movq 0x38(%rsp), %r10 movq 0xb8(%rsp), %rsi movq 0x18(%rsp), %r8 jne 0x230853 vmovaps 0xe0(%rsp), %xmm0 vpcmpeqd %xmm14, %xmm14, %xmm14 vtestps %xmm14, %xmm0 pushq $0x3 popq %rax vbroadcastss 0x1cbb150(%rip), %xmm12 # 0x1eeba20 movq 0x10(%rsp), %r9 movq 0x8(%rsp), %r11 vbroadcastss 0x1cef62d(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1cef628(%rip), %xmm13 # 0x1f1ff14 vmovaps 0x40(%rsp), %xmm15 jb 0x22f2d2 vmovaps 0x350(%rsp), %xmm1 vbroadcastss 0x1cbc27a(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x350(%rsp) pushq $0x2 popq %rax jmp 0x22f2d2 andq $-0x10, %rbp movl 0x2c(%rbp), %eax movq 0x30(%rbp,%rax), %rcx movq %rbp, 0x100(%rsp) vmovdqa 0x360(%rsp), %xmm0 movq 0x18(%rsp), %r8 movq 0x10(%rsp), %r9 movq 0x8(%rsp), %r11 jmp 0x2307b7 vmovaps 0x370(%rsp), %xmm0 vandps 0xe0(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cbc217(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r11) addq $0x1b68, %rsp # imm = 0x1B68 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %r14, %rdx movq %r10, %rcx addq $0x1b68, %rsp # imm = 0x1B68 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x276342 nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, false, embree::avx::SubdivPatch1MBIntersectorK<4>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x230b96 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x230b96 movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 movq %rsp, %rcx andq $0x0, (%rcx) movzbl %al, %ebp vmovaps (%rdx), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%rdx), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%rdx), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%rdx), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%rdx), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%rdx), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1cf0492(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1cc05a9(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vandps %xmm4, %xmm2, %xmm5 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm2, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vrcpps %xmm1, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1cbbc9b(%rip), %xmm5 # 0x1eec714 vsubps %xmm1, %xmm5, %xmm1 vmulps %xmm1, %xmm4, %xmm1 vaddps %xmm1, %xmm4, %xmm1 vrcpps %xmm2, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm2, %xmm4, %xmm2 vrcpps %xmm3, %xmm4 vmulps %xmm4, %xmm3, %xmm3 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d29e93(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vcmpnltps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1d23ebd(%rip), %xmm5 # 0x1f549a0 vbroadcastss 0x1d29e7c(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm2, %xmm5, %xmm6, %xmm2 vmovaps %xmm1, 0x90(%rax) vmovaps %xmm2, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d29e5c(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d29e57(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%rdx), %xmm1 vmovaps 0x80(%rdx), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cbaedb(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cbc028(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi movq %rsp, %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x279122 popq %rax popq %rcx andq %rbx, %rbp jne 0x230b6a addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, false, embree::avx::SubdivPatch1MBIntersectorK<4>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1c28, %rsp # imm = 0x1C28 movq (%rsi), %r8 cmpq $0x8, 0x70(%r8) je 0x232888 movq %rdx, %r14 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%rdx), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x232888 movq %rcx, %r10 vandps %xmm3, %xmm4, %xmm7 andq $0x0, 0x168(%rsp) vmovaps (%r14), %xmm3 vmovaps %xmm3, 0x2e0(%rsp) vmovaps 0x10(%r14), %xmm3 vmovaps %xmm3, 0x2f0(%rsp) vmovaps 0x20(%r14), %xmm3 vmovaps %xmm3, 0x300(%rsp) vmovaps 0x40(%r14), %xmm3 vmovaps %xmm3, 0x310(%rsp) vmovaps 0x50(%r14), %xmm4 vmovaps %xmm4, 0x320(%rsp) vmovaps 0x60(%r14), %xmm5 vmovaps %xmm5, 0x330(%rsp) vbroadcastss 0x1cf0264(%rip), %xmm9 # 0x1f20ec4 vandps %xmm3, %xmm9, %xmm6 vbroadcastss 0x1cc037b(%rip), %xmm8 # 0x1ef0fe8 vcmpltps %xmm8, %xmm6, %xmm6 vblendvps %xmm6, %xmm8, %xmm3, %xmm3 vandps %xmm4, %xmm9, %xmm6 vcmpltps %xmm8, %xmm6, %xmm6 vblendvps %xmm6, %xmm8, %xmm4, %xmm4 vandps %xmm5, %xmm9, %xmm6 vcmpltps %xmm8, %xmm6, %xmm6 vblendvps %xmm6, %xmm8, %xmm5, %xmm5 vrcpps %xmm3, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0x1cbba6a(%rip), %xmm8 # 0x1eec714 vsubps %xmm3, %xmm8, %xmm3 vmulps %xmm3, %xmm6, %xmm3 vaddps %xmm3, %xmm6, %xmm3 vrcpps %xmm4, %xmm6 vmulps %xmm6, %xmm4, %xmm4 vsubps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm6, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm5 vsubps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm6, %xmm5 vmovaps %xmm3, 0x340(%rsp) vmovaps %xmm4, 0x350(%rsp) vmovaps %xmm5, 0x360(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d29c5d(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x370(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d23c7e(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d29c3d(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vmovaps %xmm3, 0x380(%rsp) vcmpnltps %xmm2, %xmm5, %xmm3 vbroadcastss 0x1d29c24(%rip), %xmm4 # 0x1f5a96c vbroadcastss 0x1d29c1f(%rip), %xmm5 # 0x1f5a970 vblendvps %xmm3, %xmm4, %xmm5, %xmm3 vmovaps %xmm3, 0x390(%rsp) vmovaps 0x30(%r14), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cbaca9(%rip), %xmm1 # 0x1eeba20 vblendvps %xmm7, %xmm3, %xmm1, %xmm1 vmovaps %xmm1, 0x3a0(%rsp) vbroadcastss 0x1cbbdf5(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm7, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x3b0(%rsp) vmovaps %xmm7, 0x3d0(%rsp) vxorps %xmm0, %xmm7, %xmm0 vmovaps %xmm0, 0x150(%rsp) cmpq $0x0, 0x8(%rcx) je 0x230dcf movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r9d, %r9d cmpb $0x1, %al adcq $0x2, %r9 jmp 0x230dd3 pushq $0x3 popq %r9 leaq 0x80(%r14), %rax movq %rax, 0xa8(%rsp) leaq 0x590(%rsp), %r12 movq $-0x8, -0x10(%r12) leaq 0xd30(%rsp), %r15 vbroadcastss 0x1cbac1c(%rip), %xmm0 # 0x1eeba20 vmovaps %xmm0, -0x20(%r15) movq 0x70(%r8), %rax movq %rax, -0x8(%r12) vmovaps %xmm1, -0x10(%r15) leaq 0x1f1f160(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x3c0(%rsp) movq %r10, 0x40(%rsp) movq %r9, 0x178(%rsp) addq $-0x10, %r15 movq -0x8(%r12), %rbp addq $-0x8, %r12 cmpq $-0x8, %rbp je 0x232763 vmovaps (%r15), %xmm15 vcmpltps 0x3b0(%rsp), %xmm15, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x232774 movzbl %al, %r13d popcntl %r13d, %ebx xorl %eax, %eax cmpq %r9, %rbx jbe 0x232778 cmpq %r9, %rbx pushq $0x8 popq %rbx jbe 0x232766 testb $0x8, %bpl jne 0x2310f6 vmovaps 0x3b0(%rsp), %xmm0 movq %rbp, %rax andq $-0x10, %rax andl $0x7, %ebp movl %ebp, %ecx vcmpnleps %xmm15, %xmm0, %xmm0 xorl %edx, %edx vbroadcastss 0x1cbab64(%rip), %xmm15 # 0x1eeba20 movq %rbx, %rbp movq (%rax,%rdx,8), %rdi cmpq $0x8, %rdi je 0x231062 vbroadcastss 0x80(%rax,%rdx,4), %xmm2 vbroadcastss 0x20(%rax,%rdx,4), %xmm3 vmovaps 0x70(%r14), %xmm1 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0xa0(%rax,%rdx,4), %xmm3 vbroadcastss 0x40(%rax,%rdx,4), %xmm4 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0xc0(%rax,%rdx,4), %xmm4 vbroadcastss 0x60(%rax,%rdx,4), %xmm5 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0x90(%rax,%rdx,4), %xmm5 vbroadcastss 0x30(%rax,%rdx,4), %xmm6 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0xb0(%rax,%rdx,4), %xmm6 vbroadcastss 0x50(%rax,%rdx,4), %xmm7 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0xd0(%rax,%rdx,4), %xmm7 vbroadcastss 0x70(%rax,%rdx,4), %xmm8 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vmovaps 0x2e0(%rsp), %xmm8 vmovaps 0x2f0(%rsp), %xmm9 vmovaps 0x300(%rsp), %xmm10 vmovaps 0x340(%rsp), %xmm11 vsubps %xmm8, %xmm2, %xmm2 vmulps %xmm2, %xmm11, %xmm12 vsubps %xmm9, %xmm3, %xmm2 vmovaps 0x350(%rsp), %xmm3 vmulps %xmm3, %xmm2, %xmm13 vsubps %xmm10, %xmm4, %xmm2 vmovaps 0x360(%rsp), %xmm4 vmulps %xmm4, %xmm2, %xmm14 vsubps %xmm8, %xmm5, %xmm2 vmulps %xmm2, %xmm11, %xmm5 vsubps %xmm9, %xmm6, %xmm2 vmulps %xmm3, %xmm2, %xmm3 vsubps %xmm10, %xmm7, %xmm2 vmulps %xmm4, %xmm2, %xmm4 vpminsd %xmm5, %xmm12, %xmm2 vpminsd %xmm3, %xmm13, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpminsd %xmm4, %xmm14, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpmaxsd %xmm5, %xmm12, %xmm5 vpmaxsd %xmm3, %xmm13, %xmm3 vpminsd %xmm3, %xmm5, %xmm5 vpmaxsd %xmm4, %xmm14, %xmm4 vpmaxsd 0x3a0(%rsp), %xmm2, %xmm3 vpminsd %xmm4, %xmm5, %xmm4 vpminsd 0x3b0(%rsp), %xmm4, %xmm4 cmpl $0x6, %ecx je 0x23107b vcmpleps %xmm4, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm1 vpslld $0x1f, %xmm1, %xmm1 vtestps %xmm1, %xmm1 je 0x231062 vbroadcastss 0x1cba9e2(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 cmpq $0x8, %rbp je 0x23105b movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm15, (%r15) addq $0x10, %r15 movq %rdi, %rbp vmovaps %xmm1, %xmm15 cmpq $0x8, %rdi je 0x2310ab leaq 0x1(%rdx), %rdi cmpq $0x3, %rdx movq %rdi, %rdx jb 0x230ebf jmp 0x2310ab vcmpleps %xmm4, %xmm3, %xmm3 vbroadcastss 0xe0(%rax,%rdx,4), %xmm4 vcmpleps %xmm1, %xmm4, %xmm4 vbroadcastss 0xf0(%rax,%rdx,4), %xmm5 vcmpltps %xmm5, %xmm1, %xmm1 vandps %xmm1, %xmm4, %xmm1 vandps %xmm3, %xmm1, %xmm1 jmp 0x231025 xorl %eax, %eax cmpq $0x8, %rbp je 0x2310ef vmovaps 0x3b0(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r9 jae 0x2310de testb %cl, %cl jne 0x230e8c jmp 0x232766 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm15, (%r15) addq $0x10, %r15 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x2310d1 cmpq $-0x8, %rbp je 0x232763 vmovaps 0x3b0(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x232774 movq %r8, 0x88(%rsp) movl %ebp, %eax andl $0xf, %eax cmpl $0x8, %eax jne 0x23283f vmovaps 0x150(%rsp), %xmm0 movq 0x168(%rsp), %rax movl 0x8(%rax), %eax decl %eax vcvtsi2ss %rax, %xmm7, %xmm2 vshufps $0x0, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[0,0,0,0] vmulps 0x70(%r14), %xmm1, %xmm1 vroundps $0x1, %xmm1, %xmm3 vaddss 0x1cbf868(%rip), %xmm2, %xmm2 # 0x1ef09cc vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vminps %xmm2, %xmm3, %xmm2 vmaxps 0x1cba89b(%rip), %xmm2, %xmm2 # 0x1eeba10 vcvtps2dq %xmm2, %xmm3 vmovapd %xmm3, 0x240(%rsp) vmovmskps %xmm0, %eax xorl $0xf, %eax je 0x2326f8 movq %rsi, 0xa0(%rsp) vsubps %xmm2, %xmm1, %xmm1 vxorps 0x1cbac7d(%rip), %xmm0, %xmm2 # 0x1eebe20 vbroadcastss 0x1cbb568(%rip), %xmm0 # 0x1eec714 vmovaps %xmm1, 0x230(%rsp) vsubps %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0x220(%rsp) vmovaps %xmm2, %xmm0 vmovaps 0x240(%rsp), %xmm1 vmovaps %xmm1, 0x3e0(%rsp) shrq $0x2, %rbp andq $-0x4, %rbp vmovaps %xmm0, 0x410(%rsp) movzbl %al, %eax andl $0xf, %eax bsfq %rax, %rax cltq movslq 0x240(%rsp,%rax,4), %rdi vmovd %edi, %xmm0 movq 0x168(%rsp), %rax movl 0x28(%rax), %ebx movl 0x14(%rax), %r8d movl 0xc(%rax), %r13d xorl %ecx, %ecx cmpq $0x2, %r13 setne %cl movq %rcx, 0x30(%rsp) movl 0x24(%rax), %ecx movq %rax, (%rsp) movq %rbx, %r11 imulq %rdi, %r11 imulq $0xc, %r8, %rsi leaq (%rax,%rbp), %rdi andl $-0x4, %ebx movq %rcx, %rax movq %rcx, 0x28(%rsp) leaq (%rbx,%rcx), %r9 addq %r11, %r9 addq %r9, %rsi leaq (%rsi,%r13,4), %rcx addq %rdi, %rcx movq %rcx, 0x10(%rsp) leaq (%r11,%r8,8), %rdx movq %r8, %rcx movq %r8, 0x140(%rsp) addq %rax, %rdx addq %rdx, %rbx leaq (%rbx,%r13,4), %r8 addq %rdi, %r8 movq %r8, 0xe0(%rsp) leaq (%r9,%rcx,4), %rax movq %rax, 0xb0(%rsp) leaq (%rax,%r13,4), %r8 addq %rdi, %r8 movq %r8, 0xd8(%rsp) leaq (%r9,%r13,4), %r8 addq %rdi, %r8 movq %r8, 0xd0(%rsp) movq (%rsp), %rax leaq (%rax,%rbp), %rdi addq $0x30, %rdi leaq (%r11,%r13,4), %rax movq 0x28(%rsp), %rcx addq %rdi, %rax addq %rcx, %rax movq %rax, 0xc8(%rsp) movq %rsi, %r8 leaq (%rdx,%r13,4), %rax addq %rdi, %rax movq %rax, 0xc0(%rsp) movq 0x140(%rsp), %rax leaq (%r11,%rax,4), %rax movq %rbx, %rsi addq %rcx, %rax leaq (%rax,%r13,4), %rbx addq %rdi, %rbx movq %rbx, 0xb8(%rsp) addq %rdi, %r8 addq %rdi, %rsi addq %rdi, 0xb0(%rsp) addq %rdi, %r9 movq %r9, 0x198(%rsp) movq 0x10(%rsp), %r9 addq %rdi, %rdx movq %rdx, 0x190(%rsp) addq %rdi, %rax movq %rax, 0x188(%rsp) movq %rcx, %rdi vpshufd $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vpcmpeqd 0x3e0(%rsp), %xmm0, %xmm0 vpand %xmm2, %xmm0, %xmm0 vpxor 0x1cbaad2(%rip), %xmm0, %xmm1 # 0x1eebe20 vmovdqa %xmm1, 0x3f0(%rsp) movq (%rsp), %rax addq %rax, %r11 movl 0x10(%rax), %eax movl %eax, 0xec(%rsp) movq 0x30(%rsp), %rax leaq 0x4(,%rax,4), %rax movq %rax, 0x200(%rsp) leaq (,%r13,4), %rax movq %rax, 0x1f8(%rsp) addq $0x2c, %rdi vmovdqa %xmm0, 0x400(%rsp) vmovdqa %xmm0, %xmm9 movq $0x0, 0x180(%rsp) vmovaps %xmm2, %xmm0 vmovaps %xmm2, 0x210(%rsp) leaq (%r11,%rbp), %rax addq %rdi, %rax movq %rax, 0x208(%rsp) xorl %r13d, %r13d cmpq %r13, 0x200(%rsp) je 0x232626 movq 0x208(%rsp), %rdx vbroadcastss (%rdx,%r13), %xmm3 movq 0x188(%rsp), %rcx vbroadcastss -0x4(%rcx,%r13), %xmm5 movq 0x190(%rsp), %rax vbroadcastss -0x4(%rax,%r13), %xmm6 vbroadcastss 0x4(%rdx,%r13), %xmm7 vbroadcastss (%rcx,%r13), %xmm1 vbroadcastss (%rax,%r13), %xmm4 movq 0xc8(%rsp), %rax vbroadcastss -0x4(%rax,%r13), %xmm2 movq 0xb8(%rsp), %rax vbroadcastss -0x4(%rax,%r13), %xmm0 vmovaps %xmm0, 0x60(%rsp) movq 0xc0(%rsp), %rax vbroadcastss -0x4(%rax,%r13), %xmm0 vmovaps %xmm0, 0x70(%rsp) movq 0x198(%rsp), %rax vmovdqa %xmm9, (%rsp) vbroadcastss -0x4(%rax,%r13), %xmm9 movq 0xb0(%rsp), %rcx vbroadcastss -0x4(%rcx,%r13), %xmm10 vbroadcastss -0x4(%rsi,%r13), %xmm11 vbroadcastss (%rax,%r13), %xmm12 vbroadcastss (%rcx,%r13), %xmm13 vbroadcastss (%rsi,%r13), %xmm14 movq 0xd0(%rsp), %rax vbroadcastss 0x2c(%rax,%r13), %xmm15 vmovaps 0x230(%rsp), %xmm0 vmulps %xmm0, %xmm9, %xmm9 vmovaps 0x220(%rsp), %xmm8 vmulps %xmm3, %xmm8, %xmm3 vaddps %xmm3, %xmm9, %xmm3 vmovaps %xmm3, 0x10(%rsp) movq 0xd8(%rsp), %rax vbroadcastss 0x2c(%rax,%r13), %xmm9 vmulps %xmm0, %xmm10, %xmm10 vmulps %xmm5, %xmm8, %xmm5 vaddps %xmm5, %xmm10, %xmm5 movq 0xe0(%rsp), %rax vbroadcastss 0x2c(%rax,%r13), %xmm10 vmulps %xmm0, %xmm11, %xmm11 vmulps %xmm6, %xmm8, %xmm6 vaddps %xmm6, %xmm11, %xmm3 vmulps %xmm0, %xmm12, %xmm6 vmulps %xmm0, %xmm13, %xmm11 vmulps %xmm0, %xmm14, %xmm12 vmulps %xmm7, %xmm8, %xmm7 vaddps %xmm6, %xmm7, %xmm6 vmovaps %xmm6, 0x30(%rsp) vmulps %xmm1, %xmm8, %xmm6 vaddps %xmm6, %xmm11, %xmm1 vmovaps %xmm1, 0x140(%rsp) vmulps %xmm4, %xmm8, %xmm4 vaddps %xmm4, %xmm12, %xmm1 vmovaps %xmm1, 0x1d0(%rsp) vmulps %xmm0, %xmm15, %xmm4 vmulps %xmm0, %xmm9, %xmm6 vmulps %xmm0, %xmm10, %xmm7 vmulps %xmm2, %xmm8, %xmm2 vaddps %xmm4, %xmm2, %xmm10 vmulps 0x60(%rsp), %xmm8, %xmm1 vaddps %xmm6, %xmm1, %xmm0 vmulps 0x70(%rsp), %xmm8, %xmm1 vaddps %xmm7, %xmm1, %xmm1 vmovaps 0x10(%r14), %xmm6 vmovaps 0x20(%r14), %xmm11 vsubps %xmm6, %xmm5, %xmm15 vsubps %xmm11, %xmm3, %xmm3 vmovaps %xmm0, 0x510(%rsp) vsubps %xmm6, %xmm0, %xmm2 vmovaps %xmm2, 0x60(%rsp) vmovaps %xmm1, 0x500(%rsp) vsubps %xmm11, %xmm1, %xmm0 vmovaps %xmm0, 0x70(%rsp) vsubps %xmm15, %xmm2, %xmm1 vsubps %xmm3, %xmm0, %xmm5 vaddps %xmm2, %xmm15, %xmm7 vaddps %xmm3, %xmm0, %xmm8 vmulps %xmm5, %xmm7, %xmm0 vmulps %xmm1, %xmm8, %xmm9 vsubps %xmm0, %xmm9, %xmm12 vmovaps (%r14), %xmm9 vmovaps 0x10(%rsp), %xmm0 vsubps %xmm9, %xmm0, %xmm13 vmovaps %xmm10, 0x520(%rsp) vsubps %xmm9, %xmm10, %xmm0 vsubps %xmm13, %xmm0, %xmm2 vmulps %xmm2, %xmm8, %xmm8 vaddps %xmm0, %xmm13, %xmm10 vmovaps %xmm5, 0xf0(%rsp) vmulps %xmm5, %xmm10, %xmm14 vsubps %xmm8, %xmm14, %xmm14 vmovaps %xmm1, 0x100(%rsp) vmulps %xmm1, %xmm10, %xmm8 vmovaps %xmm2, 0x1e0(%rsp) vmulps %xmm7, %xmm2, %xmm7 vsubps %xmm8, %xmm7, %xmm8 vmovaps 0x60(%r14), %xmm2 vmulps %xmm2, %xmm8, %xmm10 vmovaps 0x50(%r14), %xmm1 vmovaps %xmm1, 0x10(%rsp) vmulps %xmm1, %xmm14, %xmm14 vaddps %xmm14, %xmm10, %xmm14 vmovaps 0x40(%r14), %xmm4 vmulps %xmm4, %xmm12, %xmm12 vaddps %xmm14, %xmm12, %xmm10 vmovaps 0x140(%rsp), %xmm1 vsubps %xmm6, %xmm1, %xmm1 vmovaps 0x1d0(%rsp), %xmm5 vsubps %xmm11, %xmm5, %xmm5 vsubps %xmm1, %xmm15, %xmm12 vsubps %xmm5, %xmm3, %xmm8 vmovaps %xmm15, 0x130(%rsp) vaddps %xmm1, %xmm15, %xmm6 vmovaps %xmm3, 0x120(%rsp) vaddps %xmm5, %xmm3, %xmm11 vmulps %xmm6, %xmm8, %xmm15 vmulps %xmm11, %xmm12, %xmm7 vsubps %xmm15, %xmm7, %xmm7 vmovaps 0x30(%rsp), %xmm3 vsubps %xmm9, %xmm3, %xmm9 vsubps %xmm9, %xmm13, %xmm3 vmulps %xmm3, %xmm11, %xmm11 vmovaps %xmm13, 0x110(%rsp) vaddps %xmm9, %xmm13, %xmm15 vmovaps %xmm2, %xmm14 vmovaps %xmm8, 0x560(%rsp) vmulps %xmm8, %xmm15, %xmm8 vsubps %xmm11, %xmm8, %xmm8 vmovaps %xmm12, 0x570(%rsp) vmulps %xmm12, %xmm15, %xmm11 vmovaps %xmm3, %xmm12 vmulps %xmm6, %xmm3, %xmm6 vsubps %xmm11, %xmm6, %xmm6 vmulps %xmm2, %xmm6, %xmm6 vmovaps 0x10(%rsp), %xmm13 vmulps %xmm8, %xmm13, %xmm8 vaddps %xmm6, %xmm8, %xmm6 vbroadcastss 0x1cef7ce(%rip), %xmm8 # 0x1f20ec4 vmulps %xmm7, %xmm4, %xmm7 vaddps %xmm6, %xmm7, %xmm6 vsubps %xmm0, %xmm9, %xmm15 vaddps %xmm0, %xmm9, %xmm0 vmovaps (%rsp), %xmm3 vmovaps 0x60(%rsp), %xmm2 vsubps %xmm2, %xmm1, %xmm11 vaddps %xmm2, %xmm1, %xmm1 vmovaps 0x70(%rsp), %xmm2 vsubps %xmm2, %xmm5, %xmm9 vaddps %xmm2, %xmm5, %xmm2 vmulps %xmm1, %xmm9, %xmm5 vmulps %xmm2, %xmm11, %xmm7 vsubps %xmm5, %xmm7, %xmm5 vmulps %xmm2, %xmm15, %xmm2 vmulps %xmm0, %xmm9, %xmm7 vsubps %xmm2, %xmm7, %xmm2 vmulps %xmm0, %xmm11, %xmm0 vmulps %xmm1, %xmm15, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm14, 0x60(%rsp) vmulps %xmm0, %xmm14, %xmm0 vmulps %xmm2, %xmm13, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmovaps %xmm4, 0x70(%rsp) vmulps %xmm5, %xmm4, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm6, %xmm10, %xmm1 vaddps %xmm1, %xmm0, %xmm4 vandps %xmm4, %xmm8, %xmm1 vbroadcastss 0x1cef74c(%rip), %xmm2 # 0x1f20ecc vmulps %xmm2, %xmm1, %xmm1 vminps %xmm6, %xmm10, %xmm2 vminps %xmm0, %xmm2, %xmm2 vbroadcastss 0x1cef72b(%rip), %xmm5 # 0x1f20ec0 vxorps %xmm5, %xmm1, %xmm5 vcmpnltps %xmm5, %xmm2, %xmm2 vmovaps %xmm10, 0x550(%rsp) vmaxps %xmm6, %xmm10, %xmm5 vmaxps %xmm0, %xmm5, %xmm0 vcmpleps %xmm1, %xmm0, %xmm0 vorps %xmm0, %xmm2, %xmm0 vtestps %xmm3, %xmm0 vandps %xmm3, %xmm0, %xmm0 je 0x231988 vmovaps %xmm4, 0x530(%rsp) vmovaps 0xf0(%rsp), %xmm7 vmovaps 0x570(%rsp), %xmm10 vmovaps %xmm0, 0x540(%rsp) vmulps %xmm7, %xmm10, %xmm0 vmovaps 0x100(%rsp), %xmm3 vmovaps 0x560(%rsp), %xmm13 vmulps %xmm3, %xmm13, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm13, %xmm11, %xmm2 vmulps %xmm9, %xmm10, %xmm5 vsubps %xmm2, %xmm5, %xmm5 vandps %xmm0, %xmm8, %xmm0 vandps %xmm2, %xmm8, %xmm2 vcmpltps %xmm2, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm5, %xmm2 vmulps %xmm9, %xmm12, %xmm0 vmulps %xmm7, %xmm12, %xmm1 vmovaps 0x1e0(%rsp), %xmm7 vmulps %xmm7, %xmm13, %xmm4 vsubps %xmm4, %xmm1, %xmm1 vmulps %xmm13, %xmm15, %xmm5 vsubps %xmm0, %xmm5, %xmm5 vandps %xmm4, %xmm8, %xmm4 vandps %xmm0, %xmm8, %xmm0 vcmpltps %xmm0, %xmm4, %xmm0 vblendvps %xmm0, %xmm1, %xmm5, %xmm4 vmulps %xmm10, %xmm15, %xmm0 vmulps %xmm7, %xmm10, %xmm1 vmulps %xmm3, %xmm12, %xmm5 vmulps %xmm11, %xmm12, %xmm7 vsubps %xmm5, %xmm1, %xmm1 vsubps %xmm0, %xmm7, %xmm7 vandps %xmm5, %xmm8, %xmm5 vandps %xmm0, %xmm8, %xmm0 vcmpltps %xmm0, %xmm5, %xmm0 vblendvps %xmm0, %xmm1, %xmm7, %xmm5 vmulps 0x60(%rsp), %xmm5, %xmm0 vmulps 0x10(%rsp), %xmm4, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps 0x70(%rsp), %xmm2, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vmulps 0x120(%rsp), %xmm5, %xmm1 vmulps 0x130(%rsp), %xmm4, %xmm7 vaddps %xmm7, %xmm1, %xmm1 vmulps 0x110(%rsp), %xmm2, %xmm3 vaddps %xmm1, %xmm3, %xmm1 vaddps %xmm1, %xmm1, %xmm1 vrcpps %xmm0, %xmm3 vmulps %xmm3, %xmm0, %xmm7 vbroadcastss 0x1cbae31(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm3, %xmm7 vaddps %xmm7, %xmm3, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vmovaps 0x30(%r14), %xmm3 vcmpleps 0x80(%r14), %xmm1, %xmm7 vcmpleps %xmm1, %xmm3, %xmm3 vandps %xmm7, %xmm3, %xmm3 vcmpneqps 0x1cba0fb(%rip), %xmm0, %xmm0 # 0x1eeba10 vandps %xmm3, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm0 vmovaps 0x540(%rsp), %xmm3 vtestps %xmm3, %xmm0 vandps %xmm3, %xmm0, %xmm0 je 0x231988 vmovaps 0x550(%rsp), %xmm3 vmovaps %xmm3, 0x4f0(%rsp) vmovaps %xmm6, 0x4e0(%rsp) vmovaps 0x530(%rsp), %xmm3 vmovaps %xmm3, 0x4d0(%rsp) vmovaps %xmm1, 0x4c0(%rsp) vmovaps %xmm2, 0x4b0(%rsp) vmovaps %xmm4, 0x4a0(%rsp) vmovaps %xmm5, 0x490(%rsp) vmovdqa (%rsp), %xmm9 movq 0xc8(%rsp), %rax vbroadcastss (%rax,%r13), %xmm12 movq 0xb8(%rsp), %rax vbroadcastss (%rax,%r13), %xmm11 movq 0xc0(%rsp), %rax vbroadcastss (%rax,%r13), %xmm10 movq 0xd0(%rsp), %rax vbroadcastss 0x30(%rax,%r13), %xmm13 movq 0xd8(%rsp), %rax vbroadcastss 0x30(%rax,%r13), %xmm14 movq 0xe0(%rsp), %rax vbroadcastss 0x30(%rax,%r13), %xmm15 movq 0x168(%rsp), %rcx movl 0x18(%rcx), %eax movq (%r10), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rax,8), %rbx vbroadcastss 0x34(%rbx), %xmm1 vandps 0x90(%r14), %xmm1, %xmm1 vpcmpeqd 0x1cb9ffc(%rip), %xmm1, %xmm1 # 0x1eeba10 vtestps %xmm0, %xmm1 jb 0x231a41 movl 0x1c(%rcx), %ecx vandnps %xmm0, %xmm1, %xmm0 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x231f12 cmpq $0x0, 0x48(%rbx) jne 0x231f12 vpandn %xmm9, %xmm0, %xmm9 vtestps %xmm9, %xmm9 je 0x231ef6 vmovaps 0x230(%rsp), %xmm2 vmulps %xmm2, %xmm13, %xmm0 vmulps %xmm2, %xmm14, %xmm1 vmulps %xmm2, %xmm15, %xmm2 vmovaps 0x220(%rsp), %xmm4 vmulps %xmm4, %xmm12, %xmm3 vaddps %xmm0, %xmm3, %xmm12 vmulps %xmm4, %xmm11, %xmm3 vaddps %xmm1, %xmm3, %xmm5 vmulps %xmm4, %xmm10, %xmm1 vaddps %xmm2, %xmm1, %xmm6 vmovaps (%r14), %xmm1 vmovaps 0x10(%r14), %xmm7 vmovaps 0x20(%r14), %xmm8 vmovaps 0x520(%rsp), %xmm2 vsubps %xmm1, %xmm2, %xmm0 vmovaps 0x510(%rsp), %xmm2 vsubps %xmm7, %xmm2, %xmm13 vmovaps 0x500(%rsp), %xmm2 vsubps %xmm8, %xmm2, %xmm15 vmovaps 0x30(%rsp), %xmm2 vsubps %xmm1, %xmm2, %xmm2 vmovaps %xmm2, 0x10(%rsp) vmovaps 0x140(%rsp), %xmm2 vsubps %xmm7, %xmm2, %xmm3 vmovaps 0x1d0(%rsp), %xmm2 vsubps %xmm8, %xmm2, %xmm11 vsubps %xmm1, %xmm12, %xmm4 vsubps %xmm7, %xmm5, %xmm5 vsubps %xmm8, %xmm6, %xmm1 vmovaps %xmm1, 0x30(%rsp) vsubps %xmm0, %xmm4, %xmm2 vsubps %xmm13, %xmm5, %xmm7 vsubps %xmm15, %xmm1, %xmm14 vaddps %xmm5, %xmm13, %xmm6 vaddps %xmm1, %xmm15, %xmm8 vmovaps %xmm9, (%rsp) vmulps %xmm6, %xmm14, %xmm9 vmulps %xmm7, %xmm8, %xmm10 vsubps %xmm9, %xmm10, %xmm12 vaddps %xmm0, %xmm4, %xmm9 vmulps %xmm2, %xmm8, %xmm8 vmovaps %xmm14, 0x110(%rsp) vmulps %xmm14, %xmm9, %xmm10 vsubps %xmm8, %xmm10, %xmm10 vmovaps %xmm7, 0x120(%rsp) vmulps %xmm7, %xmm9, %xmm8 vmovaps %xmm2, 0x130(%rsp) vmulps %xmm6, %xmm2, %xmm6 vsubps %xmm8, %xmm6, %xmm6 vmovaps 0x60(%r14), %xmm1 vmulps %xmm1, %xmm6, %xmm6 vmovaps 0x50(%r14), %xmm14 vmulps %xmm10, %xmm14, %xmm10 vaddps %xmm6, %xmm10, %xmm6 vmovaps 0x40(%r14), %xmm2 vmovaps %xmm2, 0x140(%rsp) vmulps %xmm2, %xmm12, %xmm12 vaddps %xmm6, %xmm12, %xmm10 vsubps %xmm3, %xmm13, %xmm12 vmovaps %xmm11, %xmm9 vsubps %xmm11, %xmm15, %xmm11 vmovaps %xmm13, 0x70(%rsp) vaddps %xmm3, %xmm13, %xmm6 vmovaps %xmm15, 0x60(%rsp) vaddps %xmm9, %xmm15, %xmm15 vmulps %xmm6, %xmm11, %xmm13 vmulps %xmm15, %xmm12, %xmm2 vsubps %xmm13, %xmm2, %xmm2 vmovaps 0x10(%rsp), %xmm8 vsubps %xmm8, %xmm0, %xmm13 vmulps %xmm15, %xmm13, %xmm15 vmovaps %xmm0, 0x1d0(%rsp) vaddps %xmm0, %xmm8, %xmm7 vmovaps %xmm8, %xmm0 vmulps %xmm7, %xmm11, %xmm8 vsubps %xmm15, %xmm8, %xmm8 vmovaps %xmm12, 0x100(%rsp) vmulps %xmm7, %xmm12, %xmm7 vmulps %xmm6, %xmm13, %xmm6 vsubps %xmm7, %xmm6, %xmm6 vmulps %xmm1, %xmm6, %xmm6 vmulps %xmm8, %xmm14, %xmm7 vmovaps %xmm1, %xmm8 vaddps %xmm7, %xmm6, %xmm6 vmovaps 0x140(%rsp), %xmm12 vmulps %xmm2, %xmm12, %xmm2 vaddps %xmm6, %xmm2, %xmm7 vmovaps %xmm7, 0x1e0(%rsp) vsubps %xmm4, %xmm0, %xmm15 vaddps %xmm4, %xmm0, %xmm1 vsubps %xmm5, %xmm3, %xmm6 vaddps %xmm5, %xmm3, %xmm0 vmovaps 0x30(%rsp), %xmm2 vsubps %xmm2, %xmm9, %xmm4 vaddps %xmm2, %xmm9, %xmm2 vbroadcastss 0x1cef277(%rip), %xmm9 # 0x1f20ec4 vmulps %xmm4, %xmm0, %xmm3 vmulps %xmm2, %xmm6, %xmm5 vsubps %xmm3, %xmm5, %xmm3 vmulps %xmm2, %xmm15, %xmm2 vmulps %xmm4, %xmm1, %xmm5 vsubps %xmm2, %xmm5, %xmm2 vmulps %xmm6, %xmm1, %xmm1 vmulps %xmm0, %xmm15, %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmovaps %xmm8, 0x30(%rsp) vmulps %xmm0, %xmm8, %xmm0 vmovaps %xmm14, 0x10(%rsp) vmulps %xmm2, %xmm14, %xmm1 vmovaps %xmm12, %xmm14 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm3, %xmm12, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm7, %xmm10, %xmm1 vaddps %xmm1, %xmm0, %xmm12 vandps %xmm9, %xmm12, %xmm1 vbroadcastss 0x1cef220(%rip), %xmm2 # 0x1f20ecc vmulps %xmm2, %xmm1, %xmm1 vminps %xmm7, %xmm10, %xmm2 vminps %xmm0, %xmm2, %xmm2 vbroadcastss 0x1cef1ff(%rip), %xmm3 # 0x1f20ec0 vxorps %xmm3, %xmm1, %xmm3 vcmpnltps %xmm3, %xmm2, %xmm2 vmovaps %xmm10, 0xf0(%rsp) vmaxps %xmm7, %xmm10, %xmm3 vmaxps %xmm0, %xmm3, %xmm0 vcmpleps %xmm1, %xmm0, %xmm0 vorps %xmm0, %xmm2, %xmm0 vmovaps (%rsp), %xmm1 vtestps %xmm1, %xmm0 vandps %xmm1, %xmm0, %xmm10 je 0x231e89 vmovaps 0x110(%rsp), %xmm5 vmovaps 0x100(%rsp), %xmm8 vmulps %xmm5, %xmm8, %xmm0 vmovaps 0x120(%rsp), %xmm7 vmulps %xmm7, %xmm11, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm6, %xmm11, %xmm2 vmulps %xmm4, %xmm8, %xmm3 vsubps %xmm2, %xmm3, %xmm3 vandps %xmm0, %xmm9, %xmm0 vandps %xmm2, %xmm9, %xmm2 vcmpltps %xmm2, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm3, %xmm3 vmulps %xmm4, %xmm13, %xmm0 vmulps %xmm5, %xmm13, %xmm1 vmovaps 0x130(%rsp), %xmm5 vmulps %xmm5, %xmm11, %xmm2 vsubps %xmm2, %xmm1, %xmm1 vmulps %xmm11, %xmm15, %xmm4 vsubps %xmm0, %xmm4, %xmm4 vandps %xmm2, %xmm9, %xmm2 vandps %xmm0, %xmm9, %xmm0 vcmpltps %xmm0, %xmm2, %xmm0 vblendvps %xmm0, %xmm1, %xmm4, %xmm1 vmulps %xmm8, %xmm15, %xmm0 vmulps %xmm5, %xmm8, %xmm2 vmulps %xmm7, %xmm13, %xmm4 vmulps %xmm6, %xmm13, %xmm5 vsubps %xmm4, %xmm2, %xmm2 vsubps %xmm0, %xmm5, %xmm5 vandps %xmm4, %xmm9, %xmm4 vandps %xmm0, %xmm9, %xmm0 vcmpltps %xmm0, %xmm4, %xmm0 vblendvps %xmm0, %xmm2, %xmm5, %xmm4 vmulps 0x30(%rsp), %xmm4, %xmm0 vmulps 0x10(%rsp), %xmm1, %xmm2 vaddps %xmm2, %xmm0, %xmm0 vmulps %xmm3, %xmm14, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vmulps 0x60(%rsp), %xmm4, %xmm2 vmulps 0x70(%rsp), %xmm1, %xmm5 vaddps %xmm5, %xmm2, %xmm2 vmulps 0x1d0(%rsp), %xmm3, %xmm5 vaddps %xmm2, %xmm5, %xmm2 vaddps %xmm2, %xmm2, %xmm2 vrcpps %xmm0, %xmm5 vmulps %xmm5, %xmm0, %xmm7 vbroadcastss 0x1cba927(%rip), %xmm8 # 0x1eec714 vsubps %xmm7, %xmm8, %xmm7 vmulps %xmm7, %xmm5, %xmm7 vaddps %xmm7, %xmm5, %xmm5 vmulps %xmm5, %xmm2, %xmm2 vmovaps 0x30(%r14), %xmm5 vcmpleps 0x80(%r14), %xmm2, %xmm7 vcmpleps %xmm2, %xmm5, %xmm5 vandps %xmm7, %xmm5, %xmm5 vcmpneqps 0x1cb9bf1(%rip), %xmm0, %xmm0 # 0x1eeba10 vandps %xmm5, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm0 vtestps %xmm10, %xmm0 vandps %xmm0, %xmm10, %xmm10 je 0x231e89 vmovaps 0xf0(%rsp), %xmm5 vmovaps %xmm5, 0x480(%rsp) vmovaps 0x1e0(%rsp), %xmm5 vmovaps %xmm5, 0x470(%rsp) vmovaps %xmm12, 0x460(%rsp) vmovaps %xmm2, 0x450(%rsp) vmovaps %xmm3, 0x440(%rsp) vmovaps %xmm1, 0x430(%rsp) vmovaps %xmm4, 0x420(%rsp) vmovdqa (%rsp), %xmm9 movq 0x168(%rsp), %rcx movl 0x18(%rcx), %eax movq (%r10), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rax,8), %rbx vbroadcastss 0x34(%rbx), %xmm1 vandps 0x90(%r14), %xmm1, %xmm1 vpcmpeqd 0x1cb9b52(%rip), %xmm1, %xmm1 # 0x1eeba10 vtestps %xmm10, %xmm1 jb 0x231eec movl 0x1c(%rcx), %ecx vandnps %xmm10, %xmm1, %xmm0 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x2322b9 cmpq $0x0, 0x48(%rbx) jne 0x2322b9 vpandn %xmm9, %xmm0, %xmm9 vtestps %xmm9, %xmm9 setne %al jmp 0x231ef8 xorl %eax, %eax vmovaps 0x210(%rsp), %xmm0 addq $0x4, %r13 testb %al, %al jne 0x2313c9 jmp 0x232626 vmovaps 0x4d0(%rsp), %xmm3 vbroadcastss 0x1ceefa0(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm3, %xmm1 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cba7db(%rip), %xmm6 # 0x1eec714 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cbf09a(%rip), %xmm3 # 0x1ef0fe8 vcmpnltps %xmm3, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm2 vmulps 0x4f0(%rsp), %xmm2, %xmm1 vminps %xmm6, %xmm1, %xmm1 vmulps 0x4e0(%rsp), %xmm2, %xmm2 vminps %xmm6, %xmm2, %xmm2 vbroadcastss -0x4(%r8,%r13), %xmm3 vbroadcastss 0x2c(%r9,%r13), %xmm4 vpsrld $0x10, %xmm3, %xmm5 vxorps %xmm7, %xmm7, %xmm7 vpblendw $0xaa, %xmm7, %xmm3, %xmm3 # xmm3 = xmm3[0],xmm7[1],xmm3[2],xmm7[3],xmm3[4],xmm7[5],xmm3[6],xmm7[7] vcvtdq2ps %xmm3, %xmm3 vbroadcastss 0x1cec545(%rip), %xmm8 # 0x1f1e4e0 vmulps %xmm3, %xmm8, %xmm3 vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vsubps %xmm1, %xmm6, %xmm6 vsubps %xmm2, %xmm6, %xmm6 vmulps %xmm3, %xmm6, %xmm3 vmulps %xmm5, %xmm6, %xmm5 vpblendw $0xaa, %xmm7, %xmm4, %xmm6 # xmm6 = xmm4[0],xmm7[1],xmm4[2],xmm7[3],xmm4[4],xmm7[5],xmm4[6],xmm7[7] vcvtdq2ps %xmm6, %xmm6 vmulps %xmm6, %xmm8, %xmm6 vmulps %xmm6, %xmm2, %xmm6 vaddps %xmm6, %xmm3, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss (%r8,%r13), %xmm4 vaddps %xmm2, %xmm5, %xmm2 vpblendw $0xaa, %xmm7, %xmm4, %xmm5 # xmm5 = xmm4[0],xmm7[1],xmm4[2],xmm7[3],xmm4[4],xmm7[5],xmm4[6],xmm7[7] vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vaddps %xmm2, %xmm1, %xmm1 movq 0x8(%r10), %rdx vmovaps 0x4b0(%rsp), %xmm2 vmovaps %xmm2, 0x250(%rsp) vmovaps 0x4a0(%rsp), %xmm2 vmovaps %xmm2, 0x260(%rsp) vmovaps 0x490(%rsp), %xmm2 vmovaps %xmm2, 0x270(%rsp) vmovaps %xmm3, 0x280(%rsp) vmovd %eax, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmovaps %xmm1, 0x290(%rsp) vmovd %ecx, %xmm1 vpshufd $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmovdqa %xmm1, 0x2a0(%rsp) vmovdqa %xmm2, 0x2b0(%rsp) leaq 0x2c0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x2c0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x2d0(%rsp) vmovaps 0x80(%r14), %xmm3 vblendvps %xmm0, 0x4c0(%rsp), %xmm3, %xmm1 vmovaps %xmm1, 0x80(%r14) vmovaps %xmm0, 0x90(%rsp) leaq 0x90(%rsp), %rax movq %rax, 0x1a0(%rsp) movq 0x18(%rbx), %rax movq %rax, 0x1a8(%rsp) movq 0x8(%r10), %rax movq %rax, 0x1b0(%rsp) movq %r14, 0x1b8(%rsp) leaq 0x250(%rsp), %rax movq %rax, 0x1c0(%rsp) movl $0x4, 0x1c8(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax movq %rsi, 0x58(%rsp) movq %rdi, 0x28(%rsp) movq %r11, 0x50(%rsp) movq %r8, 0x48(%rsp) vmovaps %xmm10, 0x70(%rsp) vmovaps %xmm11, 0x60(%rsp) vmovaps %xmm12, 0x130(%rsp) vmovaps %xmm13, 0x120(%rsp) vmovaps %xmm14, 0x110(%rsp) vmovaps %xmm15, 0x100(%rsp) vmovaps %xmm3, 0xf0(%rsp) je 0x2321e6 leaq 0x1a0(%rsp), %rdi movq %r9, 0x10(%rsp) vzeroupper callq %rax vmovaps 0xf0(%rsp), %xmm3 vmovaps 0x100(%rsp), %xmm15 vmovaps 0x110(%rsp), %xmm14 vmovaps 0x120(%rsp), %xmm13 vmovaps 0x130(%rsp), %xmm12 vmovaps 0x60(%rsp), %xmm11 vmovaps 0x70(%rsp), %xmm10 vmovdqa (%rsp), %xmm9 movq 0x10(%rsp), %r9 movq 0x48(%rsp), %r8 movq 0x50(%rsp), %r11 movq 0x28(%rsp), %rdi movq 0x58(%rsp), %rsi movq 0x40(%rsp), %r10 vmovdqa 0x90(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x2325c4 movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x23227c testb $0x2, (%rcx) jne 0x232212 testb $0x40, 0x3e(%rbx) je 0x23227c leaq 0x1a0(%rsp), %rdi movq %r9, %rbx vzeroupper callq %rax vmovaps 0xf0(%rsp), %xmm3 vmovaps 0x100(%rsp), %xmm15 vmovaps 0x110(%rsp), %xmm14 vmovaps 0x120(%rsp), %xmm13 vmovaps 0x130(%rsp), %xmm12 vmovaps 0x60(%rsp), %xmm11 vmovaps 0x70(%rsp), %xmm10 vmovdqa (%rsp), %xmm9 movq %rbx, %r9 movq 0x48(%rsp), %r8 movq 0x50(%rsp), %r11 movq 0x28(%rsp), %rdi movq 0x58(%rsp), %rsi movq 0x40(%rsp), %r10 vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x90(%rsp), %xmm0, %xmm1 vpxor 0x1cb9b8f(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x1b8(%rsp), %rax vbroadcastss 0x1cba8e2(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x2325d4 vmovaps 0x460(%rsp), %xmm3 vbroadcastss 0x1ceebf9(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm3, %xmm1 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cba434(%rip), %xmm6 # 0x1eec714 vsubps %xmm3, %xmm6, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cbecf3(%rip), %xmm3 # 0x1ef0fe8 vcmpnltps %xmm3, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm2 vmulps 0x480(%rsp), %xmm2, %xmm1 vminps %xmm6, %xmm1, %xmm1 vmulps 0x470(%rsp), %xmm2, %xmm2 vminps %xmm6, %xmm2, %xmm2 vbroadcastss 0x2c(%r9,%r13), %xmm3 vbroadcastss 0x30(%r9,%r13), %xmm4 vpsrld $0x10, %xmm3, %xmm5 vxorps %xmm7, %xmm7, %xmm7 vpblendw $0xaa, %xmm7, %xmm3, %xmm3 # xmm3 = xmm3[0],xmm7[1],xmm3[2],xmm7[3],xmm3[4],xmm7[5],xmm3[6],xmm7[7] vcvtdq2ps %xmm3, %xmm3 vbroadcastss 0x1cec19e(%rip), %xmm8 # 0x1f1e4e0 vmulps %xmm3, %xmm8, %xmm3 vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vsubps %xmm1, %xmm6, %xmm6 vsubps %xmm2, %xmm6, %xmm6 vmulps %xmm3, %xmm6, %xmm3 vmulps %xmm5, %xmm6, %xmm5 vpblendw $0xaa, %xmm7, %xmm4, %xmm6 # xmm6 = xmm4[0],xmm7[1],xmm4[2],xmm7[3],xmm4[4],xmm7[5],xmm4[6],xmm7[7] vcvtdq2ps %xmm6, %xmm6 vmulps %xmm6, %xmm8, %xmm6 vmulps %xmm6, %xmm2, %xmm6 vaddps %xmm6, %xmm3, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm2, %xmm2 vbroadcastss (%r8,%r13), %xmm4 vaddps %xmm2, %xmm5, %xmm2 vpblendw $0xaa, %xmm7, %xmm4, %xmm5 # xmm5 = xmm4[0],xmm7[1],xmm4[2],xmm7[3],xmm4[4],xmm7[5],xmm4[6],xmm7[7] vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm3, %xmm5, %xmm3 vpsrld $0x10, %xmm4, %xmm4 vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm1, %xmm1 vaddps %xmm2, %xmm1, %xmm1 movq 0x8(%r10), %rdx vmovaps 0x440(%rsp), %xmm2 vmovaps %xmm2, 0x250(%rsp) vmovaps 0x430(%rsp), %xmm2 vmovaps %xmm2, 0x260(%rsp) vmovaps 0x420(%rsp), %xmm2 vmovaps %xmm2, 0x270(%rsp) vmovaps %xmm3, 0x280(%rsp) vmovd %eax, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmovaps %xmm1, 0x290(%rsp) vmovd %ecx, %xmm1 vpshufd $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmovdqa %xmm1, 0x2a0(%rsp) vmovdqa %xmm2, 0x2b0(%rsp) leaq 0x2c0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x2c0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x2d0(%rsp) vmovaps 0x80(%r14), %xmm3 vblendvps %xmm0, 0x450(%rsp), %xmm3, %xmm1 vmovaps %xmm1, 0x80(%r14) vmovaps %xmm0, 0x90(%rsp) leaq 0x90(%rsp), %rax movq %rax, 0x1a0(%rsp) movq 0x18(%rbx), %rax movq %rax, 0x1a8(%rsp) movq 0x8(%r10), %rax movq %rax, 0x1b0(%rsp) movq %r14, 0x1b8(%rsp) leaq 0x250(%rsp), %rax movq %rax, 0x1c0(%rsp) movl $0x4, 0x1c8(%rsp) movq 0x48(%rbx), %rax testq %rax, %rax movq %rsi, 0x58(%rsp) movq %rdi, 0x28(%rsp) movq %r11, 0x50(%rsp) movq %r8, 0x48(%rsp) vmovaps %xmm3, 0x30(%rsp) je 0x232527 leaq 0x1a0(%rsp), %rdi movq %r9, 0x10(%rsp) vzeroupper callq %rax vmovaps 0x30(%rsp), %xmm3 vmovdqa (%rsp), %xmm9 movq 0x10(%rsp), %r9 movq 0x48(%rsp), %r8 movq 0x50(%rsp), %r11 movq 0x28(%rsp), %rdi movq 0x58(%rsp), %rsi movq 0x40(%rsp), %r10 vmovdqa 0x90(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x2325f5 movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x23258a testb $0x2, (%rcx) jne 0x232553 testb $0x40, 0x3e(%rbx) je 0x23258a leaq 0x1a0(%rsp), %rdi movq %r9, %rbx vzeroupper callq %rax vmovaps 0x30(%rsp), %xmm3 vmovdqa (%rsp), %xmm9 movq %rbx, %r9 movq 0x48(%rsp), %r8 movq 0x50(%rsp), %r11 movq 0x28(%rsp), %rdi movq 0x58(%rsp), %rsi movq 0x40(%rsp), %r10 vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x90(%rsp), %xmm0, %xmm1 vpxor 0x1cb9881(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x1b8(%rsp), %rax vbroadcastss 0x1cba5d4(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x232605 vpcmpeqd 0x1cb9444(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cb984c(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm1 vpsrad $0x1f, %xmm1, %xmm0 movq 0xa8(%rsp), %rax vblendvps %xmm1, (%rax), %xmm3, %xmm1 vmovaps %xmm1, (%rax) jmp 0x231a3c vpcmpeqd 0x1cb9413(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cb981b(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm1 vpsrad $0x1f, %xmm1, %xmm0 movq 0xa8(%rsp), %rax vblendvps %xmm1, (%rax), %xmm3, %xmm1 vmovaps %xmm1, (%rax) jmp 0x231ee7 cmpl $0x2, 0xec(%rsp) je 0x2326b5 movq 0x1f8(%rsp), %rax addq %rax, %r9 addq %rax, 0xe0(%rsp) addq %rax, 0xd8(%rsp) addq %rax, 0xd0(%rsp) addq %rax, 0xc0(%rsp) addq %rax, 0xb8(%rsp) addq %rax, 0xc8(%rsp) addq %rax, %r8 addq %rax, %rsi addq %rax, 0xb0(%rsp) addq %rax, 0x198(%rsp) addq %rax, 0x190(%rsp) addq %rax, 0x188(%rsp) addq %rax, %r11 movq 0x180(%rsp), %rax testq %rax, %rax leaq 0x1(%rax), %rax movq %rax, 0x180(%rsp) je 0x2313b7 vxorps 0x400(%rsp), %xmm0, %xmm2 vpor 0x3f0(%rsp), %xmm9, %xmm0 vmovdqa 0x410(%rsp), %xmm1 vpand %xmm1, %xmm0, %xmm0 vmovmskps %xmm2, %eax testl %eax, %eax jne 0x2311e0 vpxor 0x1cb9738(%rip), %xmm0, %xmm0 # 0x1eebe20 movq 0xa0(%rsp), %rsi movq 0x178(%rsp), %r9 xorl %ecx, %ecx movq 0x88(%rsp), %r8 vpor 0x150(%rsp), %xmm0, %xmm0 vmovdqa %xmm0, 0x150(%rsp) vtestps 0x1cb9703(%rip), %xmm0 # 0x1eebe20 jb 0x232763 vmovaps 0x3b0(%rsp), %xmm1 vbroadcastss 0x1cba453(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x3b0(%rsp) xorl %eax, %eax testq %rcx, %rcx je 0x232766 movq %rcx, (%r12) addq $0x8, %r12 vbroadcastss 0x1cba42c(%rip), %xmm0 # 0x1eecb84 vmovaps %xmm0, (%r15) addq $0x10, %r15 jmp 0x232766 pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x230e3a jmp 0x232860 pushq $0x2 jmp 0x232765 vmovaps %xmm15, 0x10(%rsp) movq %rsi, 0xa0(%rsp) movq %r8, 0x88(%rsp) bsfq %r13, %rcx movq %rsi, %rdi movq %r8, %rsi movq %rbp, %rdx movq %rcx, (%rsp) leaq 0x168(%rsp), %r8 movq %r14, %r9 pushq %r10 leaq 0x2e8(%rsp), %rax pushq %rax vzeroupper callq 0x27a5b6 popq %rcx popq %rdx testb %al, %al je 0x2327cf movq (%rsp), %rax orl $-0x1, 0x150(%rsp,%rax,4) leaq -0x1(%r13), %rax andq %rax, %r13 movq 0x40(%rsp), %r10 movq 0xa0(%rsp), %rsi movq 0x88(%rsp), %r8 jne 0x23278e vmovaps 0x150(%rsp), %xmm0 vtestps 0x1cb9621(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0x178(%rsp), %r9 vmovaps 0x10(%rsp), %xmm15 jb 0x230e80 vmovaps 0x3b0(%rsp), %xmm1 vbroadcastss 0x1cba35c(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x3b0(%rsp) pushq $0x2 popq %rax jmp 0x230e80 andq $-0x10, %rbp movq (%rbp), %rax movq (%rax), %rcx movq %rax, 0x168(%rsp) vmovdqa 0x3c0(%rsp), %xmm0 jmp 0x2326fa vmovaps 0x3d0(%rsp), %xmm0 vandps 0x150(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cba309(%rip), %xmm1 # 0x1eecb84 movq 0xa8(%rsp), %rax vmaskmovps %xmm1, %xmm0, (%rax) addq $0x1c28, %rsp # imm = 0x1C28 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::ObjectIntersectorK<4, false>>, false>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	movq (%rsi), %rax cmpq $0x8, 0x70(%rax) je 0x232e8e cmpq $0x0, 0x8(%rcx) je 0x2328c2 movq 0x10(%rcx), %r8 testb $0x1, 0x2(%r8) jne 0x27b910 vmovdqa (%rdi), %xmm0 vpcmpeqd %xmm2, %xmm2, %xmm2 vpcmpeqd %xmm2, %xmm0, %xmm4 vmovmskps %xmm4, %esi testl %esi, %esi je 0x232e8e pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x17b8, %rsp # imm = 0x17B8 vmovaps 0x40(%rdx), %xmm6 vbroadcastss 0x1cee5cb(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm6, %xmm3 vbroadcastss 0x1cbe6e2(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm3, %xmm3 vblendvps %xmm3, %xmm5, %xmm6, %xmm3 vmovaps 0x50(%rdx), %xmm7 vandps %xmm1, %xmm7, %xmm8 vcmpltps %xmm5, %xmm8, %xmm8 vblendvps %xmm8, %xmm5, %xmm7, %xmm7 vmovaps 0x60(%rdx), %xmm8 vandps %xmm1, %xmm8, %xmm1 vcmpltps %xmm5, %xmm1, %xmm1 vblendvps %xmm1, %xmm5, %xmm8, %xmm8 vrcpps %xmm3, %xmm1 vmulps %xmm1, %xmm3, %xmm3 vbroadcastss 0x1cb9dca(%rip), %xmm9 # 0x1eec714 vsubps %xmm3, %xmm9, %xmm10 vrcpps %xmm7, %xmm3 vmulps %xmm3, %xmm7, %xmm5 vsubps %xmm5, %xmm9, %xmm7 vrcpps %xmm8, %xmm5 vmulps %xmm5, %xmm8, %xmm8 vsubps %xmm8, %xmm9, %xmm8 vxorps %xmm9, %xmm9, %xmm9 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cdfd88(%rip), %xmm11 # 0x1f12704 vandps %xmm6, %xmm11, %xmm6 vmovaps 0x50(%rdx), %xmm11 vcmpltps %xmm9, %xmm11, %xmm11 vbroadcastss 0x1cee548(%rip), %xmm12 # 0x1f20edc vandps %xmm12, %xmm11, %xmm11 vmovaps (%rdx), %xmm12 vorps %xmm6, %xmm11, %xmm11 vmovaps 0x60(%rdx), %xmm6 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1d27fbf(%rip), %xmm9 # 0x1f5a974 vandps %xmm6, %xmm9, %xmm6 vmovaps 0x10(%rdx), %xmm13 vpxor %xmm2, %xmm4, %xmm2 vmovaps 0x20(%rdx), %xmm14 vpor %xmm6, %xmm2, %xmm4 vmovaps 0x30(%rdx), %xmm2 vmulps %xmm1, %xmm10, %xmm6 vmulps %xmm7, %xmm3, %xmm7 vmulps %xmm5, %xmm8, %xmm8 vpor %xmm4, %xmm11, %xmm9 vmovaps 0x80(%rdx), %xmm4 vmovdqa %xmm9, 0x10(%rsp) bsfq %rsi, %rdi vbroadcastss 0x10(%rsp,%rdi,4), %xmm10 vpcmpeqd %xmm9, %xmm10, %xmm10 vmovmskps %xmm10, %edi notq %rdi andq %rdi, %rsi jne 0x2329ee vpcmpeqd %xmm9, %xmm9, %xmm9 vpcmpeqd %xmm0, %xmm9, %xmm0 vaddps %xmm6, %xmm1, %xmm9 vaddps %xmm7, %xmm3, %xmm10 vaddps %xmm5, %xmm8, %xmm11 vxorps %xmm3, %xmm3, %xmm3 vmaxps %xmm3, %xmm2, %xmm1 vmaxps %xmm3, %xmm4, %xmm2 vbroadcastss 0x1cb8feb(%rip), %xmm15 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm15, %xmm3 vbroadcastss 0x1cba140(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm1 vmovaps %xmm1, (%rsp) leaq 0x120(%rsp), %r12 leaq 0x8c0(%rsp), %r13 vorps 0x10(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x10(%rsp) movq $-0x8, -0x10(%r12) vmovaps %xmm15, -0x20(%r13) movq 0x70(%rax), %rax movq %rax, -0x8(%r12) vmovaps %xmm3, 0xf0(%rsp) vmovaps %xmm3, -0x10(%r13) pushq $0x8 popq %r15 addq $-0x10, %r13 movq -0x8(%r12), %rbp addq $-0x8, %r12 movb $0x1, %al cmpq $-0x8, %rbp je 0x232e75 vmovaps (%r13), %xmm0 vcmpltps (%rsp), %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x232e73 testb $0x8, %bpl jne 0x232cdd movq %rbp, %r11 movq %rbp, %rsi andq $-0x10, %rsi xorl %edi, %edi xorl %r10d, %r10d movq %r15, %rbp vmovaps %xmm15, %xmm0 movq (%rsi,%rdi,8), %r8 cmpq $0x8, %r8 je 0x232bf1 vbroadcastss 0x20(%r11,%rdi,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmulps %xmm1, %xmm9, %xmm2 vbroadcastss 0x40(%r11,%rdi,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmulps %xmm1, %xmm10, %xmm3 vbroadcastss 0x60(%r11,%rdi,4), %xmm1 vsubps %xmm14, %xmm1, %xmm1 vmulps %xmm1, %xmm11, %xmm4 vbroadcastss 0x30(%r11,%rdi,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmulps %xmm1, %xmm9, %xmm5 vbroadcastss 0x50(%r11,%rdi,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmulps %xmm1, %xmm10, %xmm6 vbroadcastss 0x70(%r11,%rdi,4), %xmm1 vsubps %xmm14, %xmm1, %xmm1 vmulps %xmm1, %xmm11, %xmm7 vpminsd %xmm5, %xmm2, %xmm1 vpminsd %xmm6, %xmm3, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpminsd %xmm7, %xmm4, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpmaxsd %xmm5, %xmm2, %xmm2 vpmaxsd %xmm6, %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vpmaxsd %xmm7, %xmm4, %xmm3 vpmaxsd 0xf0(%rsp), %xmm1, %xmm4 vpminsd (%rsp), %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vcmpleps %xmm2, %xmm4, %xmm2 vtestps %xmm2, %xmm2 je 0x232bf1 vblendvps %xmm2, %xmm1, %xmm15, %xmm1 vcmpltps %xmm0, %xmm1, %xmm2 vtestps %xmm2, %xmm2 je 0x232bcd movq %rbp, %r9 vmovaps %xmm0, %xmm2 cmpq $0x8, %rbp movq %r8, %rbp jne 0x232bd8 vmovaps %xmm1, %xmm0 movq %r8, %rbp jmp 0x232bf1 movq %r8, %r9 vmovaps %xmm1, %xmm2 vmovaps %xmm0, %xmm1 incq %r10 movq %r9, (%r12) addq $0x8, %r12 vmovaps %xmm2, (%r13) addq $0x10, %r13 vmovaps %xmm1, %xmm0 cmpq $0x8, %r8 je 0x232c08 leaq 0x1(%rdi), %r8 cmpq $0x3, %rdi movq %r8, %rdi jb 0x232ae6 cmpq $0x8, %rbp setne %dil cmpq $0x2, %r10 setae %sil andb %dil, %sil cmpb $0x1, %sil je 0x232c30 cmpq $0x8, %rbp jne 0x232ac6 jmp 0x232e73 vmovaps -0x20(%r13), %xmm1 vmovaps -0x10(%r13), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x232c63 vmovaps %xmm2, -0x20(%r13) vmovaps %xmm1, -0x10(%r13) vpermilps $0x4e, -0x10(%r12), %xmm1 # xmm1 = mem[2,3,0,1] vmovups %xmm1, -0x10(%r12) cmpq $0x2, %r10 je 0x232c21 vmovaps -0x30(%r13), %xmm1 vmovaps -0x10(%r13), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x232ca1 vmovaps %xmm2, -0x30(%r13) vmovaps %xmm1, -0x10(%r13) movq -0x18(%r12), %rsi movq -0x8(%r12), %rdi movq %rdi, -0x18(%r12) movq %rsi, -0x8(%r12) vmovaps -0x30(%r13), %xmm1 vmovaps -0x20(%r13), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x232c21 vmovaps %xmm2, -0x30(%r13) vmovaps %xmm1, -0x20(%r13) vpermilps $0x4e, -0x18(%r12), %xmm1 # xmm1 = mem[2,3,0,1] vmovups %xmm1, -0x18(%r12) jmp 0x232c21 cmpq $-0x8, %rbp je 0x232e75 vmovaps (%rsp), %xmm1 vcmpnleps %xmm0, %xmm1, %xmm1 vtestps %xmm1, %xmm1 je 0x232e73 movl %ebp, %eax andl $0xf, %eax addq $-0x8, %rax movq %rax, 0x30(%rsp) je 0x232e5f andq $-0x10, %rbp xorl %r14d, %r14d movq %rdx, 0x28(%rsp) vmovaps %xmm12, 0xe0(%rsp) vmovaps %xmm13, 0xd0(%rsp) vmovaps %xmm14, 0xc0(%rsp) vmovaps %xmm9, 0xb0(%rsp) vmovaps %xmm10, 0xa0(%rsp) vmovaps %xmm11, 0x90(%rsp) vmovaps %xmm1, 0x80(%rsp) movq (%rcx), %rax movl (%rbp,%r14,8), %edi movq 0x1e8(%rax), %rax movq (%rax,%rdi,8), %rax vbroadcastss 0x34(%rax), %xmm0 vandps 0x90(%rdx), %xmm0, %xmm0 vpcmpeqd 0x1cb8c8c(%rip), %xmm0, %xmm0 # 0x1eeba10 vtestps %xmm1, %xmm0 jb 0x232e51 vandnps %xmm1, %xmm0, %xmm0 movl 0x4(%rbp,%r14,8), %r8d vmovaps %xmm0, 0x100(%rsp) leaq 0x100(%rsp), %rsi movq %rsi, 0x38(%rsp) movq 0x18(%rax), %rsi movq %rsi, 0x40(%rsp) movq 0x8(%rcx), %rsi movq %rsi, 0x50(%rsp) movq %rdx, 0x58(%rsp) movl $0x4, 0x60(%rsp) movl %edi, 0x64(%rsp) movl %r8d, 0x48(%rsp) movq %rax, 0x68(%rsp) andq $0x0, 0x70(%rsp) movq %rcx, %rbx movq 0x10(%rcx), %rcx movq %rcx, 0x78(%rsp) movq 0x18(%rcx), %rcx testq %rcx, %rcx jne 0x232dfa movq 0x60(%rax), %rcx leaq 0x38(%rsp), %rdi callq *%rcx movq %rbx, %rcx movq 0x28(%rsp), %rdx vmovaps 0xe0(%rsp), %xmm12 vmovaps 0xd0(%rsp), %xmm13 vmovaps 0xc0(%rsp), %xmm14 vmovaps 0xb0(%rsp), %xmm9 vmovaps 0xa0(%rsp), %xmm10 vmovaps 0x90(%rsp), %xmm11 vbroadcastss 0x1cb8bd8(%rip), %xmm15 # 0x1eeba20 vmovaps 0x80(%rsp), %xmm1 incq %r14 cmpq %r14, 0x30(%rsp) jne 0x232d5b vmovaps (%rsp), %xmm0 vblendvps %xmm1, 0x80(%rdx), %xmm0, %xmm0 vmovaps %xmm0, (%rsp) xorl %eax, %eax testb %al, %al je 0x232a96 addq $0x17b8, %rsp # imm = 0x17B8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::ObjectIntersectorK<4, false>>, false>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	movq (%rsi), %rax movq 0x70(%rax), %rax cmpq $0x8, %rax je 0x2333cb cmpq $0x0, 0x8(%rcx) je 0x232eb7 movq 0x10(%rcx), %r8 testb $0x1, 0x2(%r8) jne 0x27c2e6 vpcmpeqd %xmm9, %xmm9, %xmm9 vpcmpeqd (%rdi), %xmm9, %xmm1 vmovaps 0x80(%rdx), %xmm0 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %xmm10, %xmm0, %xmm2 vtestps %xmm1, %xmm2 je 0x2333cb pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x17d8, %rsp # imm = 0x17D8 leaq 0x80(%rdx), %r9 vandps %xmm1, %xmm2, %xmm11 vmovaps (%rdx), %xmm12 vmovaps 0x10(%rdx), %xmm13 vmovaps 0x20(%rdx), %xmm14 vmovaps 0x40(%rdx), %xmm1 vbroadcastss 0x1cedfae(%rip), %xmm2 # 0x1f20ec4 vandps %xmm2, %xmm1, %xmm3 vbroadcastss 0x1cbe0c5(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vblendvps %xmm3, %xmm4, %xmm1, %xmm1 vmovaps 0x30(%rdx), %xmm3 vmovaps 0x50(%rdx), %xmm5 vmovaps 0x60(%rdx), %xmm6 vandps %xmm2, %xmm6, %xmm7 vcmpltps %xmm4, %xmm7, %xmm7 vblendvps %xmm7, %xmm4, %xmm6, %xmm6 vrcpps %xmm1, %xmm7 vmulps %xmm7, %xmm1, %xmm1 vbroadcastss 0x1cb97b7(%rip), %xmm8 # 0x1eec714 vsubps %xmm1, %xmm8, %xmm1 vmulps %xmm1, %xmm7, %xmm1 vandps %xmm2, %xmm5, %xmm2 vcmpltps %xmm4, %xmm2, %xmm2 vblendvps %xmm2, %xmm4, %xmm5, %xmm2 vrcpps %xmm6, %xmm4 vmulps %xmm4, %xmm6, %xmm5 vsubps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm4, %xmm5 vmaxps %xmm10, %xmm3, %xmm3 vrcpps %xmm2, %xmm6 vmulps %xmm6, %xmm2, %xmm2 vsubps %xmm2, %xmm8, %xmm2 vbroadcastss 0x1cb8a82(%rip), %xmm15 # 0x1eeba20 vblendvps %xmm11, %xmm3, %xmm15, %xmm3 vmulps %xmm2, %xmm6, %xmm2 vaddps %xmm2, %xmm6, %xmm2 vmovaps %xmm2, 0x110(%rsp) vxorps %xmm9, %xmm11, %xmm2 vmaxps %xmm10, %xmm0, %xmm0 vaddps %xmm1, %xmm7, %xmm1 vmovaps %xmm1, 0x100(%rsp) vbroadcastss 0x1cb9baf(%rip), %xmm1 # 0x1eecb84 vmovaps %xmm11, 0x80(%rsp) vblendvps %xmm11, %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x10(%rsp) vmovaps %xmm2, %xmm0 vmovaps %xmm3, %xmm11 vaddps %xmm5, %xmm4, %xmm1 vmovaps %xmm1, 0xf0(%rsp) leaq 0x140(%rsp), %r13 movq $-0x8, -0x10(%r13) leaq 0x8e0(%rsp), %rbp vmovaps %xmm15, -0x20(%rbp) movq %rax, -0x8(%r13) vmovaps %xmm3, -0x10(%rbp) leaq 0x1f1cf54(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm1 vmovaps %xmm1, 0xd0(%rsp) pushq $0x8 popq %rbx pushq $0x1 popq %r14 addq $-0x10, %rbp movq -0x8(%r13), %r12 addq $-0x8, %r13 movb $0x1, %r10b cmpq $-0x8, %r12 je 0x2331b3 vmovaps %xmm0, (%rsp) vmovaps (%rbp), %xmm0 vcmpltps 0x10(%rsp), %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x2331ab testb $0x8, %r12b jne 0x2331c1 movq %r12, %rax movq %r12, %rdi andq $-0x10, %rdi xorl %r8d, %r8d movq %rbx, %r12 vmovaps %xmm15, %xmm0 movq (%rdi,%r8,8), %rsi cmpq $0x8, %rsi je 0x23318a vbroadcastss 0x20(%rax,%r8,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmovaps 0x100(%rsp), %xmm5 vmulps %xmm1, %xmm5, %xmm2 vbroadcastss 0x40(%rax,%r8,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmovaps 0x110(%rsp), %xmm6 vmulps %xmm1, %xmm6, %xmm3 vbroadcastss 0x60(%rax,%r8,4), %xmm1 vsubps %xmm14, %xmm1, %xmm1 vmovaps 0xf0(%rsp), %xmm7 vmulps %xmm1, %xmm7, %xmm4 vbroadcastss 0x30(%rax,%r8,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmulps %xmm1, %xmm5, %xmm5 vbroadcastss 0x50(%rax,%r8,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmulps %xmm1, %xmm6, %xmm6 vbroadcastss 0x70(%rax,%r8,4), %xmm1 vsubps %xmm14, %xmm1, %xmm1 vmulps %xmm1, %xmm7, %xmm7 vpminsd %xmm5, %xmm2, %xmm1 vpminsd %xmm6, %xmm3, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpminsd %xmm7, %xmm4, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpmaxsd %xmm5, %xmm2, %xmm2 vpmaxsd %xmm6, %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vpmaxsd %xmm7, %xmm4, %xmm3 vpmaxsd %xmm11, %xmm1, %xmm4 vpminsd 0x10(%rsp), %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vcmpleps %xmm2, %xmm4, %xmm2 vtestps %xmm2, %xmm2 je 0x23318a vblendvps %xmm2, %xmm1, %xmm15, %xmm1 cmpq $0x8, %r12 je 0x233183 movq %r12, (%r13) addq $0x8, %r13 vmovaps %xmm0, (%rbp) addq $0x10, %rbp vmovaps %xmm1, %xmm0 movq %rsi, %r12 cmpq $0x8, %rsi je 0x2331a1 leaq 0x1(%r8), %rsi cmpq $0x3, %r8 movq %rsi, %r8 jb 0x233093 cmpq $0x8, %r12 jne 0x233075 xorl %r10d, %r10d vmovaps (%rsp), %xmm0 testb %r10b, %r10b je 0x233040 jmp 0x2333a3 cmpq $-0x8, %r12 je 0x2331ae vmovaps 0x10(%rsp), %xmm1 vcmpnleps %xmm0, %xmm1, %xmm0 vtestps %xmm0, %xmm0 je 0x2331ab movl %r12d, %eax andl $0xf, %eax vmovaps (%rsp), %xmm1 vmovaps %xmm1, %xmm0 addq $-0x8, %rax movq %rax, 0x30(%rsp) je 0x23336d movq %r9, 0x20(%rsp) andq $-0x10, %r12 vxorps %xmm1, %xmm9, %xmm2 movq %rdx, 0x28(%rsp) vmovaps %xmm12, 0xc0(%rsp) vmovaps %xmm13, 0xb0(%rsp) vmovaps %xmm14, 0xa0(%rsp) vmovaps %xmm11, 0x90(%rsp) movq (%rcx), %rax movl -0x8(%r12,%r14,8), %edi movq 0x1e8(%rax), %rax movq (%rax,%rdi,8), %rax vbroadcastss 0x34(%rax), %xmm0 vandps 0x90(%rdx), %xmm0, %xmm0 vpcmpeqd %xmm0, %xmm10, %xmm0 vtestps %xmm2, %xmm0 vmovaps 0xd0(%rsp), %xmm1 jb 0x23333a vmovaps %xmm2, 0xe0(%rsp) vandnps %xmm2, %xmm0, %xmm0 movl -0x4(%r12,%r14,8), %r8d vmovaps %xmm0, 0x120(%rsp) leaq 0x120(%rsp), %rsi movq %rsi, 0x38(%rsp) movq 0x18(%rax), %rsi movq %rsi, 0x40(%rsp) movq 0x8(%rcx), %rsi movq %rsi, 0x50(%rsp) movq %rdx, 0x58(%rsp) movl $0x4, 0x60(%rsp) movl %edi, 0x64(%rsp) movl %r8d, 0x48(%rsp) movq %rax, 0x68(%rsp) andq $0x0, 0x70(%rsp) movq %rcx, %r15 movq 0x10(%rcx), %rcx movq %rcx, 0x78(%rsp) movq 0x18(%rcx), %rcx testq %rcx, %rcx jne 0x2332da movq 0x68(%rax), %rcx leaq 0x38(%rsp), %rdi callq *%rcx movq 0x28(%rsp), %rdx vmovaps 0x80(%rdx), %xmm0 vxorps %xmm10, %xmm10, %xmm10 vcmpltps %xmm10, %xmm0, %xmm1 movq %r15, %rcx vpcmpeqd %xmm9, %xmm9, %xmm9 vmovaps 0xc0(%rsp), %xmm12 vmovaps 0xb0(%rsp), %xmm13 vmovaps 0xa0(%rsp), %xmm14 vbroadcastss 0x1cb86fb(%rip), %xmm15 # 0x1eeba20 vmovaps 0x90(%rsp), %xmm11 movb $0x1, %r10b vmovaps 0xe0(%rsp), %xmm2 vandnps %xmm2, %xmm1, %xmm0 vtestps %xmm2, %xmm1 jb 0x23335b leaq 0x1(%r14), %rax vmovaps %xmm0, %xmm2 cmpq 0x30(%rsp), %r14 movq %rax, %r14 jb 0x23322d vxorps %xmm0, %xmm9, %xmm0 movq 0x20(%rsp), %r9 vmovaps (%rsp), %xmm1 pushq $0x1 popq %r14 vorps %xmm0, %xmm1, %xmm1 vtestps %xmm9, %xmm1 vmovaps %xmm1, %xmm0 jb 0x2331b3 vbroadcastss 0x1cb97fb(%rip), %xmm1 # 0x1eecb84 vmovaps 0x10(%rsp), %xmm2 vblendvps %xmm0, %xmm1, %xmm2, %xmm2 vmovaps %xmm2, 0x10(%rsp) xorl %r10d, %r10d jmp 0x2331b3 vandps 0x80(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cb97cf(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r9) addq $0x17d8, %rsp # imm = 0x17D8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::ObjectIntersectorK<4, true>>, false>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	movq (%rsi), %rax cmpq $0x8, 0x70(%rax) je 0x233a6e vmovdqa (%rdi), %xmm0 vpcmpeqd %xmm2, %xmm2, %xmm2 vpcmpeqd %xmm2, %xmm0, %xmm4 vmovmskps %xmm4, %esi testl %esi, %esi je 0x233a6e pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x17b8, %rsp # imm = 0x17B8 movq %rdx, %r14 vmovaps 0x40(%rdx), %xmm6 vbroadcastss 0x1cedab0(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm6, %xmm3 vbroadcastss 0x1cbdbc7(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm3, %xmm3 vblendvps %xmm3, %xmm5, %xmm6, %xmm3 vmovaps 0x50(%rdx), %xmm7 vandps %xmm1, %xmm7, %xmm8 vcmpltps %xmm5, %xmm8, %xmm8 vblendvps %xmm8, %xmm5, %xmm7, %xmm8 vmovaps 0x60(%rdx), %xmm7 vandps %xmm1, %xmm7, %xmm1 vcmpltps %xmm5, %xmm1, %xmm1 vblendvps %xmm1, %xmm5, %xmm7, %xmm9 vrcpps %xmm3, %xmm1 vmulps %xmm1, %xmm3, %xmm3 vbroadcastss 0x1cb92af(%rip), %xmm10 # 0x1eec714 vsubps %xmm3, %xmm10, %xmm7 vrcpps %xmm8, %xmm3 vmulps %xmm3, %xmm8, %xmm5 vsubps %xmm5, %xmm10, %xmm8 vrcpps %xmm9, %xmm5 vmulps %xmm5, %xmm9, %xmm9 vsubps %xmm9, %xmm10, %xmm10 vxorps %xmm9, %xmm9, %xmm9 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cdf26c(%rip), %xmm11 # 0x1f12704 vandps %xmm6, %xmm11, %xmm11 vmovaps 0x50(%rdx), %xmm6 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1ceda2c(%rip), %xmm12 # 0x1f20edc vandps %xmm6, %xmm12, %xmm12 vmovaps 0x80(%rdx), %xmm6 vmaxps %xmm9, %xmm6, %xmm6 vorps %xmm11, %xmm12, %xmm11 vmovaps 0x60(%rdx), %xmm12 vcmpltps %xmm9, %xmm12, %xmm9 vbroadcastss 0x1d2749a(%rip), %xmm12 # 0x1f5a974 vandps %xmm12, %xmm9, %xmm9 vmovaps (%rdx), %xmm12 vpxor %xmm2, %xmm4, %xmm2 vmovaps 0x10(%rdx), %xmm13 vpor %xmm2, %xmm9, %xmm2 vmovaps 0x20(%rdx), %xmm14 vpor %xmm2, %xmm11, %xmm9 vmovaps 0x30(%rdx), %xmm2 vmulps %xmm7, %xmm1, %xmm4 vmulps %xmm3, %xmm8, %xmm7 vmulps %xmm5, %xmm10, %xmm8 vmovdqa %xmm9, 0x10(%rsp) bsfq %rsi, %rdx vbroadcastss 0x10(%rsp,%rdx,4), %xmm10 vpcmpeqd %xmm9, %xmm10, %xmm10 vmovmskps %xmm10, %edx notq %rdx andq %rdx, %rsi jne 0x233510 vpcmpeqd %xmm9, %xmm9, %xmm9 vpcmpeqd %xmm0, %xmm9, %xmm0 vaddps %xmm4, %xmm1, %xmm11 vaddps %xmm7, %xmm3, %xmm15 vaddps %xmm5, %xmm8, %xmm1 vmovaps %xmm1, 0xf0(%rsp) vxorps %xmm1, %xmm1, %xmm1 vmaxps %xmm1, %xmm2, %xmm1 vbroadcastss 0x1cb84c4(%rip), %xmm2 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm2, %xmm3 vbroadcastss 0x1cb9619(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm6, %xmm1, %xmm1 vmovaps %xmm1, (%rsp) leaq 0x120(%rsp), %r12 leaq 0x8c0(%rsp), %r13 vorps 0x10(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x10(%rsp) movq $-0x8, -0x10(%r12) vmovaps %xmm2, -0x20(%r13) movq 0x70(%rax), %rax movq %rax, -0x8(%r12) vmovaps %xmm3, 0xe0(%rsp) vmovaps %xmm3, -0x10(%r13) pushq $0x8 popq %r15 addq $-0x10, %r13 movq -0x8(%r12), %rbp addq $-0x8, %r12 movb $0x1, %al cmpq $-0x8, %rbp je 0x233a55 vmovaps (%r13), %xmm0 vcmpltps (%rsp), %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x233a53 vmovaps %xmm0, %xmm1 testb $0x8, %bpl jne 0x2338e1 movq %rbp, %rdx andq $-0x10, %rdx andl $0x7, %ebp movl %ebp, %esi xorl %edi, %edi xorl %r10d, %r10d movq %r15, %rbp vbroadcastss 0x1cb8408(%rip), %xmm0 # 0x1eeba20 movq (%rdx,%rdi,8), %r8 cmpq $0x8, %r8 je 0x2337c4 vbroadcastss 0x80(%rdx,%rdi,4), %xmm3 vbroadcastss 0x20(%rdx,%rdi,4), %xmm4 vmovaps 0x70(%r14), %xmm2 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0xa0(%rdx,%rdi,4), %xmm4 vbroadcastss 0x40(%rdx,%rdi,4), %xmm5 vmulps %xmm4, %xmm2, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0xc0(%rdx,%rdi,4), %xmm5 vbroadcastss 0x60(%rdx,%rdi,4), %xmm6 vmulps %xmm5, %xmm2, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0x90(%rdx,%rdi,4), %xmm6 vbroadcastss 0x30(%rdx,%rdi,4), %xmm7 vmulps %xmm6, %xmm2, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0xb0(%rdx,%rdi,4), %xmm7 vbroadcastss 0x50(%rdx,%rdi,4), %xmm8 vmulps %xmm7, %xmm2, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vbroadcastss 0xd0(%rdx,%rdi,4), %xmm8 vbroadcastss 0x70(%rdx,%rdi,4), %xmm9 vmulps %xmm2, %xmm8, %xmm8 vaddps %xmm9, %xmm8, %xmm8 vsubps %xmm12, %xmm3, %xmm3 vmulps %xmm3, %xmm11, %xmm9 vsubps %xmm13, %xmm4, %xmm3 vmulps %xmm3, %xmm15, %xmm4 vsubps %xmm14, %xmm5, %xmm3 vmovaps 0xf0(%rsp), %xmm10 vmulps %xmm3, %xmm10, %xmm5 vsubps %xmm12, %xmm6, %xmm3 vmulps %xmm3, %xmm11, %xmm6 vsubps %xmm13, %xmm7, %xmm3 vmulps %xmm3, %xmm15, %xmm7 vsubps %xmm14, %xmm8, %xmm3 vmulps %xmm3, %xmm10, %xmm8 vpminsd %xmm6, %xmm9, %xmm3 vpminsd %xmm7, %xmm4, %xmm10 vpmaxsd %xmm10, %xmm3, %xmm3 vpminsd %xmm8, %xmm5, %xmm10 vpmaxsd %xmm10, %xmm3, %xmm3 vpmaxsd %xmm6, %xmm9, %xmm6 vpmaxsd %xmm7, %xmm4, %xmm4 vpminsd %xmm4, %xmm6, %xmm4 vpmaxsd %xmm8, %xmm5, %xmm5 vpminsd (%rsp), %xmm5, %xmm5 vpminsd %xmm5, %xmm4, %xmm4 vpmaxsd 0xe0(%rsp), %xmm3, %xmm5 cmpl $0x6, %esi je 0x2337dd vcmpleps %xmm4, %xmm5, %xmm2 vmovaps (%rsp), %xmm4 vcmpnleps %xmm1, %xmm4, %xmm4 vandps %xmm4, %xmm2, %xmm2 vpslld $0x1f, %xmm2, %xmm2 vtestps %xmm2, %xmm2 je 0x2337c4 vbroadcastss 0x1cb82ab(%rip), %xmm4 # 0x1eeba20 vblendvps %xmm2, %xmm3, %xmm4, %xmm2 vcmpltps %xmm0, %xmm2, %xmm3 vtestps %xmm3, %xmm3 je 0x2337a0 movq %rbp, %r9 vmovaps %xmm0, %xmm3 cmpq $0x8, %rbp movq %r8, %rbp jne 0x2337ab vmovaps %xmm2, %xmm0 movq %r8, %rbp jmp 0x2337c4 movq %r8, %r9 vmovaps %xmm2, %xmm3 vmovaps %xmm0, %xmm2 incq %r10 movq %r9, (%r12) addq $0x8, %r12 vmovaps %xmm3, (%r13) addq $0x10, %r13 vmovaps %xmm2, %xmm0 cmpq $0x8, %r8 je 0x23380d leaq 0x1(%rdi), %r8 cmpq $0x3, %rdi movq %r8, %rdi jb 0x233618 jmp 0x23380d vcmpleps %xmm4, %xmm5, %xmm4 vbroadcastss 0xe0(%rdx,%rdi,4), %xmm5 vcmpleps %xmm2, %xmm5, %xmm5 vbroadcastss 0xf0(%rdx,%rdi,4), %xmm6 vcmpltps %xmm6, %xmm2, %xmm2 vandps %xmm2, %xmm5, %xmm2 vandps %xmm4, %xmm2, %xmm2 jmp 0x233752 cmpq $0x8, %rbp setne %dl cmpq $0x2, %r10 setae %sil andb %dl, %sil cmpb $0x1, %sil je 0x233834 cmpq $0x8, %rbp jne 0x2335ed jmp 0x233a53 vmovaps -0x20(%r13), %xmm1 vmovaps -0x10(%r13), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x233867 vmovaps %xmm2, -0x20(%r13) vmovaps %xmm1, -0x10(%r13) vpermilps $0x4e, -0x10(%r12), %xmm1 # xmm1 = mem[2,3,0,1] vmovups %xmm1, -0x10(%r12) cmpq $0x2, %r10 je 0x233825 vmovaps -0x30(%r13), %xmm1 vmovaps -0x10(%r13), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x2338a5 vmovaps %xmm2, -0x30(%r13) vmovaps %xmm1, -0x10(%r13) movq -0x18(%r12), %rsi movq -0x8(%r12), %rdx movq %rdx, -0x18(%r12) movq %rsi, -0x8(%r12) vmovaps -0x30(%r13), %xmm1 vmovaps -0x20(%r13), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x233825 vmovaps %xmm2, -0x30(%r13) vmovaps %xmm1, -0x20(%r13) vpermilps $0x4e, -0x18(%r12), %xmm1 # xmm1 = mem[2,3,0,1] vmovups %xmm1, -0x18(%r12) jmp 0x233825 cmpq $-0x8, %rbp je 0x233a55 vmovaps (%rsp), %xmm0 vcmpnleps %xmm1, %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x233a53 movl %ebp, %eax andl $0xf, %eax addq $-0x8, %rax movq %rax, 0x30(%rsp) je 0x233a3f andq $-0x10, %rbp xorl %ebx, %ebx movq %rcx, 0x28(%rsp) vmovaps %xmm12, 0xd0(%rsp) vmovaps %xmm13, 0xc0(%rsp) vmovaps %xmm14, 0xb0(%rsp) vmovaps %xmm11, 0xa0(%rsp) vmovaps %xmm15, 0x90(%rsp) vmovaps %xmm1, 0x80(%rsp) movq (%rcx), %rax movl (%rbp,%rbx,8), %edi movq 0x1e8(%rax), %rax movq (%rax,%rdi,8), %rax vbroadcastss 0x34(%rax), %xmm0 vandps 0x90(%r14), %xmm0, %xmm0 vpcmpeqd 0x1cb8092(%rip), %xmm0, %xmm0 # 0x1eeba10 vtestps %xmm1, %xmm0 jb 0x233a31 vandnps %xmm1, %xmm0, %xmm0 movl 0x4(%rbp,%rbx,8), %edx vmovaps %xmm0, 0x100(%rsp) leaq 0x100(%rsp), %rsi movq %rsi, 0x38(%rsp) movq 0x18(%rax), %rsi movq %rsi, 0x40(%rsp) movq 0x8(%rcx), %rsi movq %rsi, 0x50(%rsp) movq %r14, 0x58(%rsp) movl $0x4, 0x60(%rsp) movl %edi, 0x64(%rsp) movl %edx, 0x48(%rsp) movq %rax, 0x68(%rsp) andq $0x0, 0x70(%rsp) movq 0x10(%rcx), %rcx movq %rcx, 0x78(%rsp) movq 0x18(%rcx), %rcx testq %rcx, %rcx jne 0x2339ef movq 0x60(%rax), %rcx leaq 0x38(%rsp), %rdi callq *%rcx movq 0x28(%rsp), %rcx vmovaps 0xd0(%rsp), %xmm12 vmovaps 0xc0(%rsp), %xmm13 vmovaps 0xb0(%rsp), %xmm14 vmovaps 0xa0(%rsp), %xmm11 vmovaps 0x90(%rsp), %xmm15 vmovaps 0x80(%rsp), %xmm1 incq %rbx cmpq %rbx, 0x30(%rsp) jne 0x233955 vmovaps (%rsp), %xmm0 vblendvps %xmm1, 0x80(%r14), %xmm0, %xmm0 vmovaps %xmm0, (%rsp) xorl %eax, %eax testb %al, %al je 0x2335bd addq $0x17b8, %rsp # imm = 0x17B8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::ObjectIntersectorK<4, true>>, false>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x17d8, %rsp # imm = 0x17D8 movq (%rsi), %rax movq 0x70(%rax), %rax cmpq $0x8, %rax je 0x234049 movq %rdx, %r14 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm1 vmovaps 0x80(%rdx), %xmm0 vxorps %xmm11, %xmm11, %xmm11 vcmpnltps %xmm11, %xmm0, %xmm2 vtestps %xmm1, %xmm2 je 0x234049 leaq 0x80(%r14), %r8 vandps %xmm1, %xmm2, %xmm9 vpcmpeqd %xmm10, %xmm10, %xmm10 vmovaps (%r14), %xmm1 vmovaps %xmm1, 0x110(%rsp) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x100(%rsp) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0xf0(%rsp) vmovaps 0x40(%r14), %xmm1 vbroadcastss 0x1ced3be(%rip), %xmm2 # 0x1f20ec4 vandps %xmm2, %xmm1, %xmm3 vbroadcastss 0x1cbd4d5(%rip), %xmm4 # 0x1ef0fe8 vcmpltps %xmm4, %xmm3, %xmm3 vblendvps %xmm3, %xmm4, %xmm1, %xmm1 vmovaps 0x30(%r14), %xmm3 vmovaps 0x50(%r14), %xmm5 vmovaps 0x60(%r14), %xmm6 vandps %xmm2, %xmm6, %xmm7 vcmpltps %xmm4, %xmm7, %xmm7 vblendvps %xmm7, %xmm4, %xmm6, %xmm6 vrcpps %xmm1, %xmm7 vmulps %xmm7, %xmm1, %xmm1 vbroadcastss 0x1cb8bc4(%rip), %xmm8 # 0x1eec714 vsubps %xmm1, %xmm8, %xmm1 vmulps %xmm1, %xmm7, %xmm1 vandps %xmm2, %xmm5, %xmm2 vcmpltps %xmm4, %xmm2, %xmm2 vblendvps %xmm2, %xmm4, %xmm5, %xmm2 vrcpps %xmm6, %xmm4 vmulps %xmm4, %xmm6, %xmm5 vsubps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm4, %xmm5 vmaxps %xmm11, %xmm3, %xmm3 vrcpps %xmm2, %xmm6 vmulps %xmm6, %xmm2, %xmm2 vsubps %xmm2, %xmm8, %xmm2 vbroadcastss 0x1cb7e8f(%rip), %xmm8 # 0x1eeba20 vblendvps %xmm9, %xmm3, %xmm8, %xmm3 vmulps %xmm2, %xmm6, %xmm2 vaddps %xmm2, %xmm6, %xmm2 vmovaps %xmm2, 0xd0(%rsp) vxorps %xmm10, %xmm9, %xmm2 vmaxps %xmm11, %xmm0, %xmm0 vaddps %xmm1, %xmm7, %xmm1 vmovaps %xmm1, 0xc0(%rsp) vbroadcastss 0x1cb8fbc(%rip), %xmm1 # 0x1eecb84 vmovaps %xmm9, 0x80(%rsp) vblendvps %xmm9, %xmm0, %xmm1, %xmm12 vmovaps %xmm2, %xmm0 vaddps %xmm5, %xmm4, %xmm1 vmovaps %xmm1, 0xb0(%rsp) leaq 0x140(%rsp), %r13 movq $-0x8, -0x10(%r13) leaq 0x8e0(%rsp), %rbp vmovaps %xmm8, -0x20(%rbp) movq %rax, -0x8(%r13) vmovaps %xmm3, 0xe0(%rsp) vmovaps %xmm3, -0x10(%rbp) leaq 0x1f1c362(%rip), %rax # 0x214ff80 vmovaps (%rax), %xmm1 vmovaps %xmm1, 0xa0(%rsp) pushq $0x8 popq %rbx pushq $0x1 popq %r15 addq $-0x10, %rbp movq -0x8(%r13), %r12 addq $-0x8, %r13 movb $0x1, %r9b cmpq $-0x8, %r12 je 0x233e97 vmovaps %xmm0, (%rsp) vmovaps (%rbp), %xmm0 vmovaps %xmm0, 0x10(%rsp) vcmpltps %xmm12, %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x233e8f testb $0x8, %r12b jne 0x233ea5 movq %r12, %rax andq $-0x10, %rax andl $0x7, %r12d movl %r12d, %edi xorl %edx, %edx movq %rbx, %r12 vbroadcastss 0x1cb7d8e(%rip), %xmm1 # 0x1eeba20 movq (%rax,%rdx,8), %rsi cmpq $0x8, %rsi je 0x233e36 vbroadcastss 0x80(%rax,%rdx,4), %xmm3 vbroadcastss 0x20(%rax,%rdx,4), %xmm4 vmovaps 0x70(%r14), %xmm2 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0xa0(%rax,%rdx,4), %xmm4 vbroadcastss 0x40(%rax,%rdx,4), %xmm5 vmulps %xmm4, %xmm2, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0xc0(%rax,%rdx,4), %xmm5 vbroadcastss 0x60(%rax,%rdx,4), %xmm6 vmulps %xmm5, %xmm2, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0x90(%rax,%rdx,4), %xmm6 vbroadcastss 0x30(%rax,%rdx,4), %xmm7 vmulps %xmm6, %xmm2, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0xb0(%rax,%rdx,4), %xmm7 vbroadcastss 0x50(%rax,%rdx,4), %xmm8 vmulps %xmm7, %xmm2, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vbroadcastss 0xd0(%rax,%rdx,4), %xmm8 vbroadcastss 0x70(%rax,%rdx,4), %xmm9 vmulps %xmm2, %xmm8, %xmm8 vaddps %xmm9, %xmm8, %xmm8 vmovaps 0x110(%rsp), %xmm14 vsubps %xmm14, %xmm3, %xmm3 vmovaps 0xc0(%rsp), %xmm11 vmulps %xmm3, %xmm11, %xmm9 vmovaps 0x100(%rsp), %xmm15 vsubps %xmm15, %xmm4, %xmm3 vmovaps 0xd0(%rsp), %xmm10 vmulps %xmm3, %xmm10, %xmm4 vmovaps 0xf0(%rsp), %xmm0 vsubps %xmm0, %xmm5, %xmm3 vmovaps 0xb0(%rsp), %xmm13 vmulps %xmm3, %xmm13, %xmm5 vsubps %xmm14, %xmm6, %xmm3 vmulps %xmm3, %xmm11, %xmm6 vsubps %xmm15, %xmm7, %xmm3 vmulps %xmm3, %xmm10, %xmm7 vsubps %xmm0, %xmm8, %xmm3 vmulps %xmm3, %xmm13, %xmm8 vpminsd %xmm6, %xmm9, %xmm3 vpminsd %xmm7, %xmm4, %xmm10 vpmaxsd %xmm10, %xmm3, %xmm3 vpminsd %xmm8, %xmm5, %xmm10 vpmaxsd %xmm10, %xmm3, %xmm3 vpmaxsd %xmm6, %xmm9, %xmm6 vpmaxsd %xmm7, %xmm4, %xmm4 vpminsd %xmm4, %xmm6, %xmm4 vpmaxsd %xmm8, %xmm5, %xmm5 vpminsd %xmm12, %xmm5, %xmm5 vpminsd %xmm5, %xmm4, %xmm4 vpmaxsd 0xe0(%rsp), %xmm3, %xmm5 cmpl $0x6, %edi je 0x233e4f vcmpleps %xmm4, %xmm5, %xmm2 vcmpnleps 0x10(%rsp), %xmm12, %xmm4 vandps %xmm4, %xmm2, %xmm2 vpslld $0x1f, %xmm2, %xmm2 vtestps %xmm2, %xmm2 je 0x233e36 vbroadcastss 0x1cb7c0e(%rip), %xmm4 # 0x1eeba20 vblendvps %xmm2, %xmm3, %xmm4, %xmm2 cmpq $0x8, %r12 je 0x233e2f movq %r12, (%r13) addq $0x8, %r13 vmovaps %xmm1, (%rbp) addq $0x10, %rbp vmovaps %xmm2, %xmm1 movq %rsi, %r12 cmpq $0x8, %rsi je 0x233e7f leaq 0x1(%rdx), %rsi cmpq $0x3, %rdx movq %rsi, %rdx jb 0x233c92 jmp 0x233e7f vcmpleps %xmm4, %xmm5, %xmm4 vbroadcastss 0xe0(%rax,%rdx,4), %xmm5 vcmpleps %xmm2, %xmm5, %xmm5 vbroadcastss 0xf0(%rax,%rdx,4), %xmm6 vcmpltps %xmm6, %xmm2, %xmm2 vandps %xmm2, %xmm5, %xmm2 vandps %xmm4, %xmm2, %xmm2 jmp 0x233df2 vmovaps %xmm1, 0x10(%rsp) cmpq $0x8, %r12 jne 0x233c6c xorl %r9d, %r9d vmovaps (%rsp), %xmm0 testb %r9b, %r9b je 0x233c32 jmp 0x234032 cmpq $-0x8, %r12 je 0x233e92 vcmpnleps 0x10(%rsp), %xmm12, %xmm0 vtestps %xmm0, %xmm0 je 0x233e8f movl %r12d, %eax andl $0xf, %eax vmovdqa (%rsp), %xmm1 vmovdqa %xmm1, %xmm0 addq $-0x8, %rax movq %rax, 0x30(%rsp) vpcmpeqd %xmm2, %xmm2, %xmm2 je 0x234008 movq %r8, 0x20(%rsp) andq $-0x10, %r12 vpxor %xmm2, %xmm1, %xmm3 vpxor %xmm2, %xmm2, %xmm2 movq %rcx, 0x28(%rsp) vmovaps %xmm12, 0x90(%rsp) movq (%rcx), %rax movl -0x8(%r12,%r15,8), %edi movq 0x1e8(%rax), %rax movq (%rax,%rdi,8), %rax vbroadcastss 0x34(%rax), %xmm0 vandps 0x90(%r14), %xmm0, %xmm0 vpcmpeqd %xmm2, %xmm0, %xmm0 vtestps %xmm3, %xmm0 vmovaps 0xa0(%rsp), %xmm1 jb 0x233fd1 vmovaps %xmm3, 0x10(%rsp) vandnps %xmm3, %xmm0, %xmm0 movl -0x4(%r12,%r15,8), %edx vmovaps %xmm0, 0x120(%rsp) leaq 0x120(%rsp), %rsi movq %rsi, 0x38(%rsp) movq 0x18(%rax), %rsi movq %rsi, 0x40(%rsp) movq 0x8(%rcx), %rsi movq %rsi, 0x50(%rsp) movq %r14, 0x58(%rsp) movl $0x4, 0x60(%rsp) movl %edi, 0x64(%rsp) movl %edx, 0x48(%rsp) movq %rax, 0x68(%rsp) andq $0x0, 0x70(%rsp) movq 0x10(%rcx), %rcx movq %rcx, 0x78(%rsp) movq 0x18(%rcx), %rcx testq %rcx, %rcx jne 0x233fa1 movq 0x68(%rax), %rcx leaq 0x38(%rsp), %rdi callq *%rcx vmovaps 0x80(%r14), %xmm0 vpxor %xmm2, %xmm2, %xmm2 vcmpltps %xmm2, %xmm0, %xmm1 movq 0x28(%rsp), %rcx vmovaps 0x90(%rsp), %xmm12 movb $0x1, %r9b vmovaps 0x10(%rsp), %xmm3 vandnps %xmm3, %xmm1, %xmm0 vtestps %xmm3, %xmm1 jb 0x233ff2 leaq 0x1(%r15), %rax vmovaps %xmm0, %xmm3 cmpq 0x30(%rsp), %r15 movq %rax, %r15 jb 0x233efa vpcmpeqd %xmm2, %xmm2, %xmm2 vxorps %xmm2, %xmm0, %xmm0 movq 0x20(%rsp), %r8 vmovdqa (%rsp), %xmm1 pushq $0x1 popq %r15 vpor %xmm0, %xmm1, %xmm1 vtestps %xmm2, %xmm1 vmovaps %xmm1, %xmm0 jb 0x233e97 vbroadcastss 0x1cb8b60(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm1, %xmm12, %xmm12 xorl %r9d, %r9d jmp 0x233e97 vandps 0x80(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cb8b40(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r8) addq $0x17d8, %rsp # imm = 0x17D8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::InstanceIntersectorK<4>>, false>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	movq (%rsi), %rax cmpq $0x8, 0x70(%rax) je 0x234594 cmpq $0x0, 0x8(%rcx) je 0x234080 movq 0x10(%rcx), %r8 testb $0x1, 0x2(%r8) jne 0x27cc7c vmovdqa (%rdi), %xmm0 vpcmpeqd %xmm2, %xmm2, %xmm2 vpcmpeqd %xmm2, %xmm0, %xmm4 vmovmskps %xmm4, %esi testl %esi, %esi je 0x234594 pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1768, %rsp # imm = 0x1768 vmovaps 0x40(%rdx), %xmm6 vbroadcastss 0x1cece0d(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm6, %xmm3 vbroadcastss 0x1cbcf24(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm3, %xmm3 vblendvps %xmm3, %xmm5, %xmm6, %xmm3 vmovaps 0x50(%rdx), %xmm7 vandps %xmm1, %xmm7, %xmm8 vcmpltps %xmm5, %xmm8, %xmm8 vblendvps %xmm8, %xmm5, %xmm7, %xmm7 vmovaps 0x60(%rdx), %xmm8 vandps %xmm1, %xmm8, %xmm1 vcmpltps %xmm5, %xmm1, %xmm1 vblendvps %xmm1, %xmm5, %xmm8, %xmm8 vrcpps %xmm3, %xmm1 vmulps %xmm1, %xmm3, %xmm3 vbroadcastss 0x1cb860c(%rip), %xmm9 # 0x1eec714 vsubps %xmm3, %xmm9, %xmm10 vrcpps %xmm7, %xmm3 vmulps %xmm3, %xmm7, %xmm5 vsubps %xmm5, %xmm9, %xmm7 vrcpps %xmm8, %xmm5 vmulps %xmm5, %xmm8, %xmm8 vsubps %xmm8, %xmm9, %xmm8 vxorps %xmm9, %xmm9, %xmm9 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cde5ca(%rip), %xmm11 # 0x1f12704 vandps %xmm6, %xmm11, %xmm6 vmovaps 0x50(%rdx), %xmm11 vcmpltps %xmm9, %xmm11, %xmm11 vbroadcastss 0x1cecd8a(%rip), %xmm12 # 0x1f20edc vandps %xmm12, %xmm11, %xmm11 vmovaps (%rdx), %xmm12 vorps %xmm6, %xmm11, %xmm11 vmovaps 0x60(%rdx), %xmm6 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1d26801(%rip), %xmm9 # 0x1f5a974 vandps %xmm6, %xmm9, %xmm6 vmovaps 0x10(%rdx), %xmm13 vpxor %xmm2, %xmm4, %xmm2 vmovaps 0x20(%rdx), %xmm14 vpor %xmm6, %xmm2, %xmm4 vmovaps 0x30(%rdx), %xmm2 vmulps %xmm1, %xmm10, %xmm6 vmulps %xmm7, %xmm3, %xmm7 vmulps %xmm5, %xmm8, %xmm8 vpor %xmm4, %xmm11, %xmm9 vmovaps 0x80(%rdx), %xmm4 vmovdqa %xmm9, 0x30(%rsp) bsfq %rsi, %rdi vbroadcastss 0x30(%rsp,%rdi,4), %xmm10 vpcmpeqd %xmm9, %xmm10, %xmm10 vmovmskps %xmm10, %edi notq %rdi andq %rdi, %rsi jne 0x2341ac vpcmpeqd %xmm9, %xmm9, %xmm9 vpcmpeqd %xmm0, %xmm9, %xmm0 vaddps %xmm6, %xmm1, %xmm9 vaddps %xmm7, %xmm3, %xmm10 vaddps %xmm5, %xmm8, %xmm11 vxorps %xmm1, %xmm1, %xmm1 vmaxps %xmm1, %xmm2, %xmm2 vmaxps %xmm1, %xmm4, %xmm1 vbroadcastss 0x1cb782d(%rip), %xmm15 # 0x1eeba20 vblendvps %xmm0, %xmm2, %xmm15, %xmm3 vbroadcastss 0x1cb8982(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm1, %xmm2, %xmm1 vmovaps %xmm1, 0x10(%rsp) leaq 0xd0(%rsp), %rbp leaq 0x870(%rsp), %r15 vorps 0x30(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x30(%rsp) movq $-0x8, -0x10(%rbp) vmovaps %xmm15, -0x20(%r15) movq 0x70(%rax), %rax movq %rax, -0x8(%rbp) vmovaps %xmm3, 0xb0(%rsp) vmovaps %xmm3, -0x10(%r15) pushq $0x8 popq %r12 leaq 0x20(%rsp), %rdi leaq 0xf(%rsp), %rsi addq $-0x10, %r15 movq -0x8(%rbp), %r13 addq $-0x8, %rbp movb $0x1, %al cmpq $-0x8, %r13 je 0x23457b vmovaps (%r15), %xmm0 vcmpltps 0x10(%rsp), %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x234579 testb $0x8, %r13b jne 0x234499 movq %r13, %r11 movq %r13, %r14 andq $-0x10, %r14 xorl %ebx, %ebx xorl %r10d, %r10d movq %r12, %r13 vmovaps %xmm15, %xmm0 movq (%r14,%rbx,8), %r8 cmpq $0x8, %r8 je 0x2343b7 vbroadcastss 0x20(%r11,%rbx,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmulps %xmm1, %xmm9, %xmm2 vbroadcastss 0x40(%r11,%rbx,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmulps %xmm1, %xmm10, %xmm3 vbroadcastss 0x60(%r11,%rbx,4), %xmm1 vsubps %xmm14, %xmm1, %xmm1 vmulps %xmm1, %xmm11, %xmm4 vbroadcastss 0x30(%r11,%rbx,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmulps %xmm1, %xmm9, %xmm5 vbroadcastss 0x50(%r11,%rbx,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmulps %xmm1, %xmm10, %xmm6 vbroadcastss 0x70(%r11,%rbx,4), %xmm1 vsubps %xmm14, %xmm1, %xmm1 vmulps %xmm1, %xmm11, %xmm7 vpminsd %xmm5, %xmm2, %xmm1 vpminsd %xmm6, %xmm3, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpminsd %xmm7, %xmm4, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpmaxsd %xmm5, %xmm2, %xmm2 vpmaxsd %xmm6, %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vpmaxsd %xmm7, %xmm4, %xmm3 vpmaxsd 0xb0(%rsp), %xmm1, %xmm4 vpminsd 0x10(%rsp), %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vcmpleps %xmm2, %xmm4, %xmm2 vtestps %xmm2, %xmm2 je 0x2343b7 vblendvps %xmm2, %xmm1, %xmm15, %xmm1 vcmpltps %xmm0, %xmm1, %xmm2 vtestps %xmm2, %xmm2 je 0x234394 movq %r13, %r9 vmovaps %xmm0, %xmm2 cmpq $0x8, %r13 movq %r8, %r13 jne 0x23439f vmovaps %xmm1, %xmm0 movq %r8, %r13 jmp 0x2343b7 movq %r8, %r9 vmovaps %xmm1, %xmm2 vmovaps %xmm0, %xmm1 incq %r10 movq %r9, (%rbp) addq $0x8, %rbp vmovaps %xmm2, (%r15) addq $0x10, %r15 vmovaps %xmm1, %xmm0 cmpq $0x8, %r8 je 0x2343ce leaq 0x1(%rbx), %r8 cmpq $0x3, %rbx movq %r8, %rbx jb 0x2342ac cmpq $0x8, %r13 setne %r8b cmpq $0x2, %r10 setae %r9b andb %r8b, %r9b cmpb $0x1, %r9b je 0x2343f6 cmpq $0x8, %r13 jne 0x23428c jmp 0x234579 vmovaps -0x20(%r15), %xmm1 vmovaps -0x10(%r15), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x234426 vmovaps %xmm2, -0x20(%r15) vmovaps %xmm1, -0x10(%r15) vpermilps $0x4e, -0x10(%rbp), %xmm1 # xmm1 = mem[2,3,0,1] vmovups %xmm1, -0x10(%rbp) cmpq $0x2, %r10 je 0x2343e7 vmovaps -0x30(%r15), %xmm1 vmovaps -0x10(%r15), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x234460 vmovaps %xmm2, -0x30(%r15) vmovaps %xmm1, -0x10(%r15) movq -0x18(%rbp), %r9 movq -0x8(%rbp), %r8 movq %r8, -0x18(%rbp) movq %r9, -0x8(%rbp) vmovaps -0x30(%r15), %xmm1 vmovaps -0x20(%r15), %xmm2 vcmpltps %xmm2, %xmm1, %xmm3 vtestps %xmm3, %xmm3 je 0x2343e7 vmovaps %xmm2, -0x30(%r15) vmovaps %xmm1, -0x20(%r15) vpermilps $0x4e, -0x18(%rbp), %xmm1 # xmm1 = mem[2,3,0,1] vmovups %xmm1, -0x18(%rbp) jmp 0x2343e7 cmpq $-0x8, %r13 je 0x23457b vmovaps 0x10(%rsp), %xmm1 vcmpnleps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x20(%rsp) vtestps %xmm0, %xmm0 je 0x234579 vmovaps %xmm11, 0x50(%rsp) vmovaps %xmm10, 0x60(%rsp) vmovaps %xmm9, 0x70(%rsp) vmovaps %xmm14, 0x80(%rsp) vmovaps %xmm13, 0x90(%rsp) vmovaps %xmm12, 0xa0(%rsp) movl %r13d, %ebx andl $0xf, %ebx addq $-0x8, %rbx je 0x234527 andq $-0x10, %r13 movq %rcx, 0x48(%rsp) movq %r13, %r8 movq %rdx, %r14 callq 0x3ecaec movq %r14, %rdx leaq 0xf(%rsp), %rsi leaq 0x20(%rsp), %rdi movq 0x48(%rsp), %rcx addq $0x10, %r13 decq %rbx jne 0x234501 vmovaps 0x20(%rsp), %xmm0 vmovaps 0x10(%rsp), %xmm1 vblendvps %xmm0, 0x80(%rdx), %xmm1, %xmm1 vmovaps %xmm1, 0x10(%rsp) vmovaps 0xa0(%rsp), %xmm12 vmovaps 0x90(%rsp), %xmm13 vmovaps 0x80(%rsp), %xmm14 vmovaps 0x70(%rsp), %xmm9 vmovaps 0x60(%rsp), %xmm10 vmovaps 0x50(%rsp), %xmm11 vbroadcastss 0x1cb74a7(%rip), %xmm15 # 0x1eeba20 xorl %eax, %eax testb %al, %al je 0x23425d addq $0x1768, %rsp # imm = 0x1768 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::InstanceIntersectorK<4>>, false>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1798, %rsp # imm = 0x1798 movq (%rsi), %rax movq 0x70(%rax), %rax cmpq $0x8, %rax je 0x234a28 movq %rcx, %rbx movq %rdx, %rcx cmpq $0x0, 0x8(%rbx) je 0x2345d3 movq 0x10(%rbx), %rdx testb $0x1, 0x2(%rdx) jne 0x234a3a vpcmpeqd %xmm9, %xmm9, %xmm9 vpcmpeqd (%rdi), %xmm9, %xmm2 vmovaps 0x80(%rcx), %xmm0 vxorps %xmm1, %xmm1, %xmm1 vcmpnltps %xmm1, %xmm0, %xmm3 vtestps %xmm2, %xmm3 je 0x234a28 leaq 0x80(%rcx), %r9 vandps %xmm2, %xmm3, %xmm10 vmovaps (%rcx), %xmm11 vmovaps 0x10(%rcx), %xmm12 vmovaps 0x40(%rcx), %xmm2 vbroadcastss 0x1cec8aa(%rip), %xmm3 # 0x1f20ec4 vandps %xmm3, %xmm2, %xmm4 vbroadcastss 0x1cbc9c1(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm2, %xmm4 vmovaps 0x50(%rcx), %xmm6 vmovaps 0x60(%rcx), %xmm2 vandps %xmm3, %xmm2, %xmm7 vcmpltps %xmm5, %xmm7, %xmm7 vblendvps %xmm7, %xmm5, %xmm2, %xmm7 vrcpps %xmm4, %xmm2 vmulps %xmm2, %xmm4, %xmm4 vbroadcastss 0x1cb80b8(%rip), %xmm8 # 0x1eec714 vsubps %xmm4, %xmm8, %xmm4 vandps %xmm3, %xmm6, %xmm3 vcmpltps %xmm5, %xmm3, %xmm3 vblendvps %xmm3, %xmm5, %xmm6, %xmm3 vrcpps %xmm7, %xmm5 vmulps %xmm5, %xmm7, %xmm6 vsubps %xmm6, %xmm8, %xmm6 vrcpps %xmm3, %xmm7 vmulps %xmm7, %xmm3, %xmm3 vsubps %xmm3, %xmm8, %xmm3 vmovaps 0x20(%rcx), %xmm13 vmulps %xmm3, %xmm7, %xmm3 vaddps %xmm3, %xmm7, %xmm14 vmovaps 0x30(%rcx), %xmm3 vmulps %xmm4, %xmm2, %xmm4 vmulps %xmm6, %xmm5, %xmm6 vmaxps %xmm1, %xmm3, %xmm3 vmaxps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1cb736e(%rip), %xmm15 # 0x1eeba20 vblendvps %xmm10, %xmm3, %xmm15, %xmm1 vaddps %xmm4, %xmm2, %xmm2 vmovaps %xmm2, 0xc0(%rsp) vbroadcastss 0x1cb84b6(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm10, %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x10(%rsp) vaddps %xmm6, %xmm5, %xmm0 vmovaps %xmm0, 0xb0(%rsp) vxorps %xmm9, %xmm10, %xmm0 vmovaps %xmm0, 0x30(%rsp) leaq 0x100(%rsp), %r15 movq $-0x8, -0x10(%r15) leaq 0x8a0(%rsp), %r12 vmovaps %xmm15, -0x20(%r12) movq %rax, -0x8(%r15) vmovaps %xmm1, 0xd0(%rsp) vmovaps %xmm1, -0x10(%r12) pushq $0x8 popq %r13 addq $-0x10, %r12 movq -0x8(%r15), %rbp addq $-0x8, %r15 movb $0x1, %r10b cmpq $-0x8, %rbp je 0x23488e vmovaps (%r12), %xmm0 vcmpltps 0x10(%rsp), %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x23488b testb $0x8, %bpl jne 0x23489c movq %rbp, %rax movq %rbp, %rdi andq $-0x10, %rdi xorl %edx, %edx movq %r13, %rbp vmovaps %xmm15, %xmm0 movq (%rdi,%rdx,8), %rsi cmpq $0x8, %rsi je 0x23486a vbroadcastss 0x20(%rax,%rdx,4), %xmm1 vsubps %xmm11, %xmm1, %xmm1 vmovaps 0xc0(%rsp), %xmm5 vmulps %xmm1, %xmm5, %xmm2 vbroadcastss 0x40(%rax,%rdx,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmulps %xmm1, %xmm14, %xmm3 vbroadcastss 0x60(%rax,%rdx,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmovaps 0xb0(%rsp), %xmm7 vmulps %xmm1, %xmm7, %xmm4 vbroadcastss 0x30(%rax,%rdx,4), %xmm1 vsubps %xmm11, %xmm1, %xmm1 vmulps %xmm1, %xmm5, %xmm5 vbroadcastss 0x50(%rax,%rdx,4), %xmm1 vsubps %xmm12, %xmm1, %xmm1 vmulps %xmm1, %xmm14, %xmm6 vbroadcastss 0x70(%rax,%rdx,4), %xmm1 vsubps %xmm13, %xmm1, %xmm1 vmulps %xmm1, %xmm7, %xmm7 vpminsd %xmm5, %xmm2, %xmm1 vpminsd %xmm6, %xmm3, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpminsd %xmm7, %xmm4, %xmm8 vpmaxsd %xmm8, %xmm1, %xmm1 vpmaxsd %xmm5, %xmm2, %xmm2 vpmaxsd %xmm6, %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vpmaxsd %xmm7, %xmm4, %xmm3 vpmaxsd 0xd0(%rsp), %xmm1, %xmm4 vpminsd 0x10(%rsp), %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vcmpleps %xmm2, %xmm4, %xmm2 vtestps %xmm2, %xmm2 je 0x23486a vblendvps %xmm2, %xmm1, %xmm15, %xmm1 cmpq $0x8, %rbp je 0x234863 movq %rbp, (%r15) addq $0x8, %r15 vmovaps %xmm0, (%r12) addq $0x10, %r12 vmovaps %xmm1, %xmm0 movq %rsi, %rbp cmpq $0x8, %rsi je 0x234881 leaq 0x1(%rdx), %rsi cmpq $0x3, %rdx movq %rsi, %rdx jb 0x234777 cmpq $0x8, %rbp jne 0x23475a xorl %r10d, %r10d testb %r10b, %r10b je 0x234729 jmp 0x234a14 cmpq $-0x8, %rbp je 0x23488e vmovaps 0x10(%rsp), %xmm1 vcmpnleps %xmm0, %xmm1, %xmm0 vtestps %xmm0, %xmm0 je 0x23488b vmovaps %xmm14, 0x60(%rsp) vmovaps %xmm13, 0x70(%rsp) vmovaps %xmm12, 0x80(%rsp) vmovaps %xmm11, 0x90(%rsp) vmovaps %xmm10, 0xa0(%rsp) movq %r9, 0x28(%rsp) movl %ebp, %eax andl $0xf, %eax vxorps 0x30(%rsp), %xmm9, %xmm0 vmovaps %xmm0, 0x40(%rsp) addq $-0x8, %rax movq %rax, 0x58(%rsp) je 0x23495d andq $-0x10, %rbp pushq $0x1 popq %r14 movq %rcx, 0x50(%rsp) leaq 0xe0(%rsp), %rdi leaq 0x40(%rsp), %rsi leaq 0xf(%rsp), %rdx movq %rbx, %r8 movq %rbp, %r9 callq 0x3ecd0e movq 0x50(%rsp), %rcx vmovaps 0xe0(%rsp), %xmm0 vmovaps 0x40(%rsp), %xmm1 vandnps %xmm1, %xmm0, %xmm2 vmovaps %xmm2, 0x40(%rsp) vtestps %xmm1, %xmm0 jb 0x23495d addq $0x10, %rbp leaq 0x1(%r14), %rax cmpq 0x58(%rsp), %r14 movq %rax, %r14 jb 0x234909 vpcmpeqd %xmm9, %xmm9, %xmm9 vpxor 0x40(%rsp), %xmm9, %xmm0 vmovdqa 0x30(%rsp), %xmm2 vpor %xmm0, %xmm2, %xmm2 vtestps %xmm9, %xmm2 vmovaps %xmm2, 0x30(%rsp) jb 0x2349d7 vbroadcastss 0x1cb81fc(%rip), %xmm0 # 0x1eecb84 vmovaps 0x10(%rsp), %xmm1 vblendvps %xmm2, %xmm0, %xmm1, %xmm1 vmovaps %xmm1, 0x10(%rsp) xorl %r10d, %r10d movq 0x28(%rsp), %r9 vmovaps 0xa0(%rsp), %xmm10 vmovaps 0x90(%rsp), %xmm11 vmovaps 0x80(%rsp), %xmm12 vmovaps 0x70(%rsp), %xmm13 vmovaps 0x60(%rsp), %xmm14 vbroadcastss 0x1cb704e(%rip), %xmm15 # 0x1eeba20 jmp 0x23488e movq 0x28(%rsp), %r9 vmovaps 0xa0(%rsp), %xmm10 vmovaps 0x90(%rsp), %xmm11 vmovaps 0x80(%rsp), %xmm12 vmovaps 0x70(%rsp), %xmm13 vmovaps 0x60(%rsp), %xmm14 vbroadcastss 0x1cb7014(%rip), %xmm15 # 0x1eeba20 movb $0x1, %r10b jmp 0x23488e vandps 0x30(%rsp), %xmm10, %xmm0 vbroadcastss 0x1cb8161(%rip), %xmm1 # 0x1eecb84 vmaskmovps %xmm1, %xmm0, (%r9) addq $0x1798, %rsp # imm = 0x1798 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %rcx, %rdx movq %rbx, %rcx addq $0x1798, %rsp # imm = 0x1798 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x27d5f4	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 16777232, true, embree::avx::SubGridMBIntersectorKPluecker<4, 4, true>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x1da0, %rsp # imm = 0x1DA0 movq (%rsi), %rax cmpq $0x8, 0x70(%rax) je 0x23a249 movq %rdx, %r13 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%rdx), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x23a249 movq %rcx, %r12 movq %rsi, 0x150(%rsp) vandps %xmm3, %xmm4, %xmm11 vmovaps (%r13), %xmm3 vmovaps %xmm3, 0x210(%rsp) vmovaps 0x10(%r13), %xmm3 vmovaps %xmm3, 0x220(%rsp) vmovaps 0x20(%r13), %xmm3 vmovaps %xmm3, 0x230(%rsp) vmovaps 0x40(%r13), %xmm3 vmovaps %xmm3, 0x240(%rsp) vmovaps 0x50(%r13), %xmm4 vmovaps %xmm4, 0x250(%rsp) vmovaps 0x60(%r13), %xmm5 vmovaps %xmm5, 0x260(%rsp) vbroadcastss 0x1ce84fb(%rip), %xmm6 # 0x1f20ec4 vandps %xmm6, %xmm3, %xmm7 vbroadcastss 0x1cb8612(%rip), %xmm8 # 0x1ef0fe8 vcmpltps %xmm8, %xmm7, %xmm7 vbroadcastss 0x1cb3d2f(%rip), %xmm9 # 0x1eec714 vdivps %xmm3, %xmm9, %xmm3 vandps %xmm6, %xmm4, %xmm10 vcmpltps %xmm8, %xmm10, %xmm10 vdivps %xmm4, %xmm9, %xmm4 vandps %xmm6, %xmm5, %xmm6 vcmpltps %xmm8, %xmm6, %xmm6 vdivps %xmm5, %xmm9, %xmm5 vbroadcastss 0x1ce8552(%rip), %xmm8 # 0x1f20f60 vblendvps %xmm7, %xmm8, %xmm3, %xmm3 vblendvps %xmm10, %xmm8, %xmm4, %xmm4 vblendvps %xmm6, %xmm8, %xmm5, %xmm5 vmovaps %xmm3, 0x270(%rsp) vmovaps %xmm4, 0x280(%rsp) vmovaps %xmm5, 0x290(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d21f1b(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x2a0(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d1bf3c(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d21efb(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vmovaps %xmm3, 0x2b0(%rsp) vcmpnltps %xmm2, %xmm5, %xmm3 vbroadcastss 0x1d21ee2(%rip), %xmm4 # 0x1f5a96c vbroadcastss 0x1d21edd(%rip), %xmm5 # 0x1f5a970 vblendvps %xmm3, %xmm4, %xmm5, %xmm3 vmovaps %xmm3, 0x2c0(%rsp) vmovaps 0x30(%r13), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cb2f67(%rip), %xmm1 # 0x1eeba20 vblendvps %xmm11, %xmm3, %xmm1, %xmm1 vmovaps %xmm1, 0x2d0(%rsp) vbroadcastss 0x1cb40b3(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm11, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x2e0(%rsp) vmovaps %xmm11, 0x3c0(%rsp) vxorps %xmm0, %xmm11, %xmm0 vmovaps %xmm0, 0x70(%rsp) cmpq $0x0, 0x8(%rcx) je 0x238b15 movq %rax, %rcx movq 0x10(%r12), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r15d, %r15d cmpb $0x1, %al movq %rcx, %rax adcq $0x2, %r15 jmp 0x238b19 pushq $0x3 popq %r15 leaq 0x80(%r13), %rcx movq %rcx, 0xb8(%rsp) leaq 0x6f0(%rsp), %rdi movq $-0x8, -0x10(%rdi) leaq 0xe90(%rsp), %r8 vbroadcastss 0x1cb2ed7(%rip), %xmm0 # 0x1eeba20 vmovaps %xmm0, -0x20(%r8) movq %rax, 0x148(%rsp) movq 0x70(%rax), %rax movq %rax, -0x8(%rdi) vmovaps %xmm1, -0x10(%r8) leaq 0x1f17414(%rip), %rax # 0x214ff80 vmovups (%rax), %ymm0 vinsertf128 $0x1, 0xf0(%rax), %ymm0, %ymm1 vbroadcastss 0x1cb3b91(%rip), %ymm2 # 0x1eec714 vbroadcastss 0x1cb7e40(%rip), %ymm0 # 0x1ef09cc vmovaps %ymm1, 0x5a0(%rsp) vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x580(%rsp) movq %r15, 0x50(%rsp) addq $-0x10, %r8 movq -0x8(%rdi), %rdx addq $-0x8, %rdi cmpq $-0x8, %rdx je 0x23a149 vmovaps (%r8), %xmm15 vcmpltps 0x2e0(%rsp), %xmm15, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x23a15a movzbl %al, %ebx popcntl %ebx, %r14d xorl %eax, %eax cmpq %r15, %r14 jbe 0x23a15e cmpq %r15, %r14 jbe 0x23a14c testb $0x8, %dl jne 0x238e78 vmovaps 0x2e0(%rsp), %xmm0 movq %rdx, %rax andq $-0x10, %rax andl $0x7, %edx movl %edx, %ecx vcmpnleps %xmm15, %xmm0, %xmm0 xorl %r9d, %r9d pushq $0x8 popq %rdx vbroadcastss 0x1cb2df7(%rip), %xmm15 # 0x1eeba20 movq (%rax,%r9,8), %rsi cmpq $0x8, %rsi je 0x238de5 vbroadcastss 0x80(%rax,%r9,4), %xmm2 vbroadcastss 0x20(%rax,%r9,4), %xmm3 vmovaps 0x70(%r13), %xmm1 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0xa0(%rax,%r9,4), %xmm3 vbroadcastss 0x40(%rax,%r9,4), %xmm4 vmulps %xmm3, %xmm1, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vbroadcastss 0xc0(%rax,%r9,4), %xmm4 vbroadcastss 0x60(%rax,%r9,4), %xmm5 vmulps %xmm4, %xmm1, %xmm4 vaddps %xmm5, %xmm4, %xmm4 vbroadcastss 0x90(%rax,%r9,4), %xmm5 vbroadcastss 0x30(%rax,%r9,4), %xmm6 vmulps %xmm5, %xmm1, %xmm5 vaddps %xmm6, %xmm5, %xmm5 vbroadcastss 0xb0(%rax,%r9,4), %xmm6 vbroadcastss 0x50(%rax,%r9,4), %xmm7 vmulps %xmm6, %xmm1, %xmm6 vaddps %xmm7, %xmm6, %xmm6 vbroadcastss 0xd0(%rax,%r9,4), %xmm7 vbroadcastss 0x70(%rax,%r9,4), %xmm8 vmulps %xmm7, %xmm1, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vmovaps 0x210(%rsp), %xmm8 vmovaps 0x220(%rsp), %xmm9 vmovaps 0x230(%rsp), %xmm10 vmovaps 0x270(%rsp), %xmm11 vsubps %xmm8, %xmm2, %xmm2 vmulps %xmm2, %xmm11, %xmm12 vsubps %xmm9, %xmm3, %xmm2 vmovaps 0x280(%rsp), %xmm3 vmulps %xmm3, %xmm2, %xmm13 vsubps %xmm10, %xmm4, %xmm2 vmovaps 0x290(%rsp), %xmm4 vmulps %xmm4, %xmm2, %xmm14 vsubps %xmm8, %xmm5, %xmm2 vmulps %xmm2, %xmm11, %xmm5 vsubps %xmm9, %xmm6, %xmm2 vmulps %xmm3, %xmm2, %xmm3 vsubps %xmm10, %xmm7, %xmm2 vmulps %xmm4, %xmm2, %xmm4 vpminsd %xmm5, %xmm12, %xmm2 vpminsd %xmm3, %xmm13, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vpminsd %xmm4, %xmm14, %xmm6 vpmaxsd %xmm6, %xmm2, %xmm2 vbroadcastss 0x1ce71af(%rip), %xmm6 # 0x1f1ff10 vmulps %xmm6, %xmm2, %xmm2 vpmaxsd %xmm5, %xmm12, %xmm5 vpmaxsd %xmm3, %xmm13, %xmm3 vpminsd %xmm3, %xmm5, %xmm3 vpmaxsd %xmm4, %xmm14, %xmm4 vpminsd %xmm4, %xmm3, %xmm4 vpmaxsd 0x2d0(%rsp), %xmm2, %xmm3 vbroadcastss 0x1ce7183(%rip), %xmm5 # 0x1f1ff14 vmulps %xmm5, %xmm4, %xmm4 vpminsd 0x2e0(%rsp), %xmm4, %xmm4 cmpl $0x6, %ecx je 0x238dfe vcmpleps %xmm4, %xmm3, %xmm1 vandps %xmm0, %xmm1, %xmm1 vpslld $0x1f, %xmm1, %xmm1 vtestps %xmm1, %xmm1 je 0x238de5 vbroadcastss 0x1cb2c5e(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 cmpq $0x8, %rdx je 0x238dde movq %rdx, (%rdi) addq $0x8, %rdi vmovaps %xmm15, (%r8) addq $0x10, %r8 vmovaps %xmm1, %xmm15 movq %rsi, %rdx cmpq $0x8, %rsi je 0x238e2e leaq 0x1(%r9), %rsi cmpq $0x3, %r9 movq %rsi, %r9 jb 0x238c29 jmp 0x238e2e vcmpleps %xmm4, %xmm3, %xmm3 vbroadcastss 0xe0(%rax,%r9,4), %xmm4 vcmpleps %xmm1, %xmm4, %xmm4 vbroadcastss 0xf0(%rax,%r9,4), %xmm5 vcmpltps %xmm5, %xmm1, %xmm1 vandps %xmm1, %xmm4, %xmm1 vandps %xmm3, %xmm1, %xmm1 jmp 0x238da9 xorl %eax, %eax cmpq $0x8, %rdx je 0x238e71 vmovaps 0x2e0(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %esi movb $0x1, %cl cmpq %rsi, %r15 jae 0x238e61 testb %cl, %cl jne 0x238bf6 jmp 0x23a14c movq %rdx, (%rdi) addq $0x8, %rdi vmovaps %xmm15, (%r8) addq $0x10, %r8 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x238e54 cmpq $-0x8, %rdx vpcmpeqd %xmm1, %xmm1, %xmm1 je 0x23a149 vmovaps 0x2e0(%rsp), %xmm0 vcmpnleps %xmm15, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x23a15a movl %edx, %eax andl $0xf, %eax vmovdqa 0x70(%rsp), %xmm0 addq $-0x8, %rax movq %rax, 0x158(%rsp) je 0x23a111 andq $-0x10, %rdx vpxor %xmm1, %xmm0, %xmm0 vmovdqa %xmm0, 0xd0(%rsp) xorl %eax, %eax movq %r12, 0xc0(%rsp) movq %r13, 0x170(%rsp) movq %rdi, 0x28(%rsp) movq %r8, 0x20(%rsp) movq %rdx, 0x30(%rsp) movq %rax, 0x160(%rsp) imulq $0x90, %rax, %rax vmovq 0x20(%rdx,%rax), %xmm0 vmovq 0x24(%rdx,%rax), %xmm1 vpminub %xmm1, %xmm0, %xmm1 vpcmpeqb %xmm1, %xmm0, %xmm0 vpcmpeqd %xmm1, %xmm1, %xmm1 vpxor %xmm1, %xmm0, %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vpxor %xmm1, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vmovmskps %xmm0, %ecx testl %ecx, %ecx je 0x23a0eb leaq (%rdx,%rax), %rbx vbroadcastss 0x80(%rdx,%rax), %xmm0 vmovaps 0x70(%r13), %xmm1 vsubps %xmm0, %xmm1, %xmm0 vbroadcastss 0x84(%rdx,%rax), %xmm1 vmulps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x3f0(%rsp) movzbl %cl, %eax movq %rbx, 0x58(%rsp) movq %rax, 0x168(%rsp) bsfq %rax, %rax vmovq 0x20(%rbx), %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vcvtdq2ps %xmm0, %xmm0 vbroadcastss 0x44(%rbx), %xmm1 vbroadcastss 0x38(%rbx), %xmm2 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x420(%rsp) vbroadcastss 0x420(%rsp,%rax,4), %xmm0 vmovq 0x50(%rbx), %xmm3 vpmovzxbd %xmm3, %xmm3 # xmm3 = xmm3[0],zero,zero,zero,xmm3[1],zero,zero,zero,xmm3[2],zero,zero,zero,xmm3[3],zero,zero,zero vcvtdq2ps %xmm3, %xmm3 vbroadcastss 0x74(%rbx), %xmm4 vbroadcastss 0x68(%rbx), %xmm5 vmulps %xmm3, %xmm4, %xmm3 vaddps %xmm3, %xmm5, %xmm3 vmovaps %xmm3, 0x2f0(%rsp) vmovss 0x2f0(%rsp,%rax,4), %xmm3 vsubss %xmm0, %xmm3, %xmm3 vshufps $0x0, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmovaps 0x3f0(%rsp), %xmm10 vmulps %xmm3, %xmm10, %xmm3 vaddps %xmm3, %xmm0, %xmm0 vmovq 0x24(%rbx), %xmm3 vpmovzxbd %xmm3, %xmm3 # xmm3 = xmm3[0],zero,zero,zero,xmm3[1],zero,zero,zero,xmm3[2],zero,zero,zero,xmm3[3],zero,zero,zero vcvtdq2ps %xmm3, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vaddps %xmm1, %xmm2, %xmm1 vmovaps %xmm1, 0x420(%rsp) vbroadcastss 0x420(%rsp,%rax,4), %xmm1 vmovq 0x54(%rbx), %xmm2 vpmovzxbd %xmm2, %xmm2 # xmm2 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero,xmm2[2],zero,zero,zero,xmm2[3],zero,zero,zero vcvtdq2ps %xmm2, %xmm2 vmulps %xmm2, %xmm4, %xmm2 vaddps %xmm2, %xmm5, %xmm2 vmovaps %xmm2, 0x2f0(%rsp) vmovss 0x2f0(%rsp,%rax,4), %xmm2 vsubss %xmm1, %xmm2, %xmm2 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmulps %xmm2, %xmm10, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmovq 0x28(%rbx), %xmm2 vpmovzxbd %xmm2, %xmm2 # xmm2 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero,xmm2[2],zero,zero,zero,xmm2[3],zero,zero,zero vcvtdq2ps %xmm2, %xmm2 vbroadcastss 0x48(%rbx), %xmm3 vbroadcastss 0x3c(%rbx), %xmm4 vmulps %xmm2, %xmm3, %xmm2 vaddps %xmm2, %xmm4, %xmm2 vmovaps %xmm2, 0x420(%rsp) vbroadcastss 0x420(%rsp,%rax,4), %xmm2 vmovq 0x58(%rbx), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vcvtdq2ps %xmm5, %xmm5 vbroadcastss 0x78(%rbx), %xmm6 vbroadcastss 0x6c(%rbx), %xmm7 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm7, %xmm5 vmovaps %xmm5, 0x2f0(%rsp) vmovss 0x2f0(%rsp,%rax,4), %xmm5 vsubss %xmm2, %xmm5, %xmm5 vshufps $0x0, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vmulps %xmm5, %xmm10, %xmm5 vaddps %xmm5, %xmm2, %xmm2 vmovq 0x2c(%rbx), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm3, %xmm3 vaddps %xmm3, %xmm4, %xmm3 vmovaps %xmm3, 0x420(%rsp) vbroadcastss 0x420(%rsp,%rax,4), %xmm3 vmovq 0x5c(%rbx), %xmm4 vpmovzxbd %xmm4, %xmm4 # xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero,xmm4[2],zero,zero,zero,xmm4[3],zero,zero,zero vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm6, %xmm4 vaddps %xmm4, %xmm7, %xmm4 vmovaps %xmm4, 0x2f0(%rsp) vmovss 0x2f0(%rsp,%rax,4), %xmm4 vsubss %xmm3, %xmm4, %xmm4 vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmulps %xmm4, %xmm10, %xmm4 vaddps %xmm4, %xmm3, %xmm3 vmovq 0x30(%rbx), %xmm4 vpmovzxbd %xmm4, %xmm4 # xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero,xmm4[2],zero,zero,zero,xmm4[3],zero,zero,zero vcvtdq2ps %xmm4, %xmm4 vbroadcastss 0x4c(%rbx), %xmm5 vbroadcastss 0x40(%rbx), %xmm6 vmulps %xmm4, %xmm5, %xmm4 vaddps %xmm4, %xmm6, %xmm4 vmovaps %xmm4, 0x420(%rsp) vbroadcastss 0x420(%rsp,%rax,4), %xmm4 vmovq 0x60(%rbx), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vcvtdq2ps %xmm7, %xmm7 vbroadcastss 0x7c(%rbx), %xmm8 vbroadcastss 0x70(%rbx), %xmm9 vmulps %xmm7, %xmm8, %xmm7 vaddps %xmm7, %xmm9, %xmm7 vmovaps %xmm7, 0x2f0(%rsp) vmovss 0x2f0(%rsp,%rax,4), %xmm7 vsubss %xmm4, %xmm7, %xmm7 vshufps $0x0, %xmm7, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmulps %xmm7, %xmm10, %xmm7 vaddps %xmm7, %xmm4, %xmm4 vmovq 0x34(%rbx), %xmm7 vpmovzxbd %xmm7, %xmm7 # xmm7 = xmm7[0],zero,zero,zero,xmm7[1],zero,zero,zero,xmm7[2],zero,zero,zero,xmm7[3],zero,zero,zero vcvtdq2ps %xmm7, %xmm7 vmulps %xmm7, %xmm5, %xmm5 vaddps %xmm5, %xmm6, %xmm5 vmovaps %xmm5, 0x420(%rsp) vbroadcastss 0x420(%rsp,%rax,4), %xmm5 vmovq 0x64(%rbx), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vcvtdq2ps %xmm6, %xmm6 vmulps %xmm6, %xmm8, %xmm6 vaddps %xmm6, %xmm9, %xmm6 vmovaps %xmm6, 0x2f0(%rsp) vmovss 0x2f0(%rsp,%rax,4), %xmm6 vsubss %xmm5, %xmm6, %xmm6 vshufps $0x0, %xmm6, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmulps %xmm6, %xmm10, %xmm6 vaddps %xmm6, %xmm5, %xmm5 vmovaps 0x210(%rsp), %xmm6 vmovaps 0x220(%rsp), %xmm7 vmovaps 0x230(%rsp), %xmm8 vmovaps 0x270(%rsp), %xmm9 vsubps %xmm6, %xmm0, %xmm0 vmulps %xmm0, %xmm9, %xmm0 vsubps %xmm7, %xmm2, %xmm2 vmovaps 0x280(%rsp), %xmm10 vmulps %xmm2, %xmm10, %xmm2 vsubps %xmm8, %xmm4, %xmm4 vmovaps 0x290(%rsp), %xmm11 vmulps %xmm4, %xmm11, %xmm4 vsubps %xmm6, %xmm1, %xmm1 vmulps %xmm1, %xmm9, %xmm1 vsubps %xmm7, %xmm3, %xmm3 vmulps %xmm3, %xmm10, %xmm3 vsubps %xmm8, %xmm5, %xmm5 vmulps %xmm5, %xmm11, %xmm5 vminps %xmm1, %xmm0, %xmm6 vminps %xmm3, %xmm2, %xmm7 vmaxps %xmm7, %xmm6, %xmm6 vminps %xmm5, %xmm4, %xmm7 vmaxps %xmm7, %xmm6, %xmm6 vbroadcastss 0x1ce6c77(%rip), %xmm7 # 0x1f1ff10 vmulps %xmm7, %xmm6, %xmm6 vmaxps %xmm1, %xmm0, %xmm0 vmaxps %xmm3, %xmm2, %xmm1 vminps %xmm1, %xmm0, %xmm0 vmaxps %xmm5, %xmm4, %xmm1 vminps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1ce6c5a(%rip), %xmm1 # 0x1f1ff14 vmulps %xmm1, %xmm0, %xmm0 vmaxps 0x2d0(%rsp), %xmm6, %xmm1 vminps 0x2e0(%rsp), %xmm0, %xmm0 vcmpleps %xmm0, %xmm1, %xmm0 vmovaps 0xd0(%rsp), %xmm1 vtestps %xmm1, %xmm0 vmovaps %xmm1, %xmm0 je 0x23a0c4 movzwl (%rbx,%rax,8), %ecx movzwl 0x2(%rbx,%rax,8), %esi movl %esi, 0x4c(%rsp) movl 0x88(%rbx), %r14d movl 0x4(%rbx,%rax,8), %eax movq %rax, 0xc8(%rsp) vmovaps %xmm0, 0x1d0(%rsp) vmovmskps %xmm0, %eax testl %eax, %eax je 0x23a0a4 movl %ecx, %esi movzbl %al, %r9d movl $0x7fff, %eax # imm = 0x7FFF andl %eax, %esi movl 0x4c(%rsp), %r10d andl %eax, %r10d xorl %eax, %eax testw %cx, %cx setns %al movq %rax, 0x178(%rsp) movl %esi, 0x6c(%rsp) vmovd %esi, %xmm0 vpshufd $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vpaddd 0x1ce3634(%rip), %xmm0, %xmm0 # 0x1f1c990 movl %r10d, 0x68(%rsp) vmovd %r10d, %xmm1 vpshufd $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vpaddd 0x1ce362d(%rip), %xmm1, %xmm1 # 0x1f1c9a0 vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vcvtdq2ps %ymm0, %ymm0 vmovaps %ymm0, 0x600(%rsp) vinsertf128 $0x1, %xmm1, %ymm1, %ymm0 vcvtdq2ps %ymm0, %ymm0 vmovaps %ymm0, 0x5e0(%rsp) vmovd %r14d, %xmm0 vpshufd $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmovdqa %xmm0, 0x3e0(%rsp) vmovss 0xc8(%rsp), %xmm0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmovaps %xmm0, 0x3d0(%rsp) movq %r14, 0x180(%rsp) movq %r9, 0x40(%rsp) bsfq %r9, %rcx movq %rcx, 0x60(%rsp) movq (%r12), %rax movq %rax, 0x198(%rsp) movq 0x1e8(%rax), %rax movq (%rax,%r14,8), %rax movq 0x58(%rax), %rsi movq 0x68(%rax), %rdi imulq 0xc8(%rsp), %rdi vbroadcastss 0x2c(%rax), %xmm0 vmovss 0x30(%rax), %xmm1 vbroadcastss 0x28(%rax), %xmm2 vmovaps 0x70(%r13), %xmm3 vsubps %xmm0, %xmm3, %xmm3 vsubss %xmm0, %xmm1, %xmm0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vdivps %xmm0, %xmm3, %xmm0 vmulps %xmm0, %xmm2, %xmm0 vroundps $0x1, %xmm0, %xmm1 vaddss 0x1cb7591(%rip), %xmm2, %xmm2 # 0x1ef09cc vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vminps %xmm2, %xmm1, %xmm1 vmaxps 0x1cb25c4(%rip), %xmm1, %xmm1 # 0x1eeba10 vsubps %xmm1, %xmm0, %xmm0 vmovaps %xmm0, 0x410(%rsp) vcvtps2dq %xmm1, %xmm0 vmovapd %xmm0, 0x400(%rsp) movslq 0x400(%rsp,%rcx,4), %rcx movl (%rsi,%rdi), %edx addl 0x6c(%rsp), %edx movq %rsi, 0x190(%rsp) movq %rdi, 0x188(%rsp) movl 0x4(%rsi,%rdi), %edi movl %edi, %r11d imull 0x68(%rsp), %r11d addl %edx, %r11d leaq 0x1(%r11), %rdx movq 0xe0(%rax), %rsi movq %rsi, 0x100(%rsp) imulq $0x38, %rcx, %rcx movq %rcx, 0x80(%rsp) movq (%rsi,%rcx), %r9 movq 0x10(%rsi,%rcx), %r10 movq %r10, %rax imulq %r11, %rax movq 0x48(%rsi,%rcx), %rsi vmovups (%r9,%rax), %xmm2 movq %rsi, %r8 imulq %rdx, %r8 imulq %r10, %rdx vmovups (%r9,%rdx), %xmm0 leaq (%r11,%rdi), %r15 movq %r15, %rax imulq %r10, %rax vmovups (%r9,%rax), %xmm1 vmovaps %xmm1, 0xe0(%rsp) leaq (%r11,%rdi), %r14 incq %r14 movq %r14, %rax imulq %r10, %rax vmovups (%r9,%rax), %xmm1 movq %rsi, %rdx imulq %r11, %rdx movq 0x178(%rsp), %rcx leaq (%r11,%rcx), %rax incq %rax movq %rsi, %r11 imulq %rax, %r11 imulq %r10, %rax vmovups (%r9,%rax), %xmm4 leaq (%r14,%rcx), %rbx movq %rbx, %rax imulq %r10, %rax vmovups (%r9,%rax), %xmm5 movq %rsi, %r13 imulq %r15, %r13 movq %rsi, %rax imulq %r14, %rax movq %rsi, %rcx imulq %rbx, %rcx cmpw $0x0, 0x4c(%rsp) movl $0x0, %r12d cmovnsq %rdi, %r12 addq %r12, %r15 movq %rsi, %rdi imulq %r15, %rdi imulq %r10, %r15 vmovups (%r9,%r15), %xmm6 addq %r12, %r14 movq %rsi, %r15 imulq %r14, %r15 imulq %r10, %r14 vmovups (%r9,%r14), %xmm7 movq 0x180(%rsp), %r14 addq %rbx, %r12 imulq %r12, %r10 vmovups (%r9,%r10), %xmm8 movq 0x80(%rsp), %r9 movq 0x100(%rsp), %r10 movq 0x38(%r10,%r9), %r9 vmovups (%r9,%rdx), %xmm9 vmovups (%r9,%r8), %xmm10 vmovups (%r9,%r13), %xmm11 movq 0x170(%rsp), %r13 vmovups (%r9,%rax), %xmm12 vmovups (%r9,%r11), %xmm13 vmovups (%r9,%rcx), %xmm14 movq 0x60(%rsp), %rax vbroadcastss 0x410(%rsp,%rax,4), %xmm15 vsubps %xmm2, %xmm9, %xmm9 vmulps %xmm9, %xmm15, %xmm9 vaddps %xmm2, %xmm9, %xmm2 vmovups (%r9,%rdi), %xmm3 vsubps %xmm0, %xmm10, %xmm9 vmulps %xmm9, %xmm15, %xmm9 vaddps %xmm0, %xmm9, %xmm9 vmovups (%r9,%r15), %xmm10 imulq %rsi, %r12 movq %rax, %rsi vmovaps 0xe0(%rsp), %xmm0 vsubps %xmm0, %xmm11, %xmm11 vmulps %xmm11, %xmm15, %xmm11 vaddps %xmm0, %xmm11, %xmm0 vmovups (%r9,%r12), %xmm11 vsubps %xmm1, %xmm12, %xmm12 vmulps %xmm12, %xmm15, %xmm12 vaddps %xmm1, %xmm12, %xmm1 vsubps %xmm4, %xmm13, %xmm12 vmulps %xmm12, %xmm15, %xmm12 vaddps %xmm4, %xmm12, %xmm4 vsubps %xmm5, %xmm14, %xmm12 vmulps %xmm12, %xmm15, %xmm12 vaddps %xmm5, %xmm12, %xmm5 vsubps %xmm6, %xmm3, %xmm3 vmulps %xmm3, %xmm15, %xmm3 vaddps %xmm3, %xmm6, %xmm3 vsubps %xmm7, %xmm10, %xmm6 vmulps %xmm6, %xmm15, %xmm6 vaddps %xmm6, %xmm7, %xmm6 vsubps %xmm8, %xmm11, %xmm7 vmulps %xmm7, %xmm15, %xmm7 vaddps %xmm7, %xmm8, %xmm7 vunpcklps %xmm1, %xmm2, %xmm8 # xmm8 = xmm2[0],xmm1[0],xmm2[1],xmm1[1] vunpckhps %xmm1, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm1[2],xmm2[3],xmm1[3] vunpcklps %xmm0, %xmm9, %xmm10 # xmm10 = xmm9[0],xmm0[0],xmm9[1],xmm0[1] vunpckhps %xmm0, %xmm9, %xmm11 # xmm11 = xmm9[2],xmm0[2],xmm9[3],xmm0[3] vunpcklps %xmm11, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm11[0],xmm2[1],xmm11[1] vunpcklps %xmm10, %xmm8, %xmm11 # xmm11 = xmm8[0],xmm10[0],xmm8[1],xmm10[1] vunpckhps %xmm10, %xmm8, %xmm8 # xmm8 = xmm8[2],xmm10[2],xmm8[3],xmm10[3] vunpcklps %xmm5, %xmm9, %xmm10 # xmm10 = xmm9[0],xmm5[0],xmm9[1],xmm5[1] vunpckhps %xmm5, %xmm9, %xmm9 # xmm9 = xmm9[2],xmm5[2],xmm9[3],xmm5[3] vunpcklps %xmm1, %xmm4, %xmm12 # xmm12 = xmm4[0],xmm1[0],xmm4[1],xmm1[1] vunpckhps %xmm1, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm1[2],xmm4[3],xmm1[3] vunpcklps %xmm4, %xmm9, %xmm4 # xmm4 = xmm9[0],xmm4[0],xmm9[1],xmm4[1] vunpcklps %xmm12, %xmm10, %xmm9 # xmm9 = xmm10[0],xmm12[0],xmm10[1],xmm12[1] vunpckhps %xmm12, %xmm10, %xmm10 # xmm10 = xmm10[2],xmm12[2],xmm10[3],xmm12[3] vunpcklps %xmm7, %xmm1, %xmm12 # xmm12 = xmm1[0],xmm7[0],xmm1[1],xmm7[1] vunpckhps %xmm7, %xmm1, %xmm7 # xmm7 = xmm1[2],xmm7[2],xmm1[3],xmm7[3] vunpcklps %xmm6, %xmm5, %xmm13 # xmm13 = xmm5[0],xmm6[0],xmm5[1],xmm6[1] vunpckhps %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm6[2],xmm5[3],xmm6[3] vunpcklps %xmm5, %xmm7, %xmm5 # xmm5 = xmm7[0],xmm5[0],xmm7[1],xmm5[1] vunpcklps %xmm13, %xmm12, %xmm7 # xmm7 = xmm12[0],xmm13[0],xmm12[1],xmm13[1] vunpckhps %xmm13, %xmm12, %xmm12 # xmm12 = xmm12[2],xmm13[2],xmm12[3],xmm13[3] vunpcklps %xmm6, %xmm0, %xmm13 # xmm13 = xmm0[0],xmm6[0],xmm0[1],xmm6[1] vunpckhps %xmm6, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm6[2],xmm0[3],xmm6[3] vunpcklps %xmm3, %xmm1, %xmm6 # xmm6 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] vunpckhps %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm3[2],xmm1[3],xmm3[3] vunpcklps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm6, %xmm13, %xmm1 # xmm1 = xmm13[0],xmm6[0],xmm13[1],xmm6[1] vunpckhps %xmm6, %xmm13, %xmm3 # xmm3 = xmm13[2],xmm6[2],xmm13[3],xmm6[3] vinsertf128 $0x1, %xmm7, %ymm11, %ymm6 vinsertf128 $0x1, %xmm12, %ymm8, %ymm7 vinsertf128 $0x1, %xmm5, %ymm2, %ymm2 vinsertf128 $0x1, %xmm9, %ymm9, %ymm8 vinsertf128 $0x1, %xmm10, %ymm10, %ymm9 vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vinsertf128 $0x1, %xmm1, %ymm1, %ymm10 vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vinsertf128 $0x1, %xmm0, %ymm0, %ymm15 leaq 0x1f(%rsp), %rax movq %rax, 0x480(%rsp) vbroadcastss (%r13,%rsi,4), %ymm12 vbroadcastss 0x10(%r13,%rsi,4), %ymm13 vbroadcastss 0x20(%r13,%rsi,4), %ymm14 vsubps %ymm12, %ymm6, %ymm1 vsubps %ymm13, %ymm7, %ymm5 vsubps %ymm14, %ymm2, %ymm0 vsubps %ymm12, %ymm8, %ymm2 vmovaps %ymm2, 0x100(%rsp) vsubps %ymm13, %ymm9, %ymm11 vsubps %ymm14, %ymm4, %ymm2 vmovaps %ymm2, 0x1a0(%rsp) vsubps %ymm12, %ymm10, %ymm4 vsubps %ymm13, %ymm3, %ymm8 vsubps %ymm14, %ymm15, %ymm3 vmovaps %ymm3, 0xe0(%rsp) vsubps %ymm1, %ymm4, %ymm12 vsubps %ymm5, %ymm8, %ymm14 vsubps %ymm0, %ymm3, %ymm2 vaddps %ymm5, %ymm8, %ymm6 vaddps %ymm0, %ymm3, %ymm7 vmulps %ymm2, %ymm6, %ymm9 vmulps %ymm7, %ymm14, %ymm10 vsubps %ymm9, %ymm10, %ymm15 vaddps %ymm1, %ymm4, %ymm9 vmulps %ymm7, %ymm12, %ymm7 vmovaps %ymm2, 0x660(%rsp) vmulps %ymm2, %ymm9, %ymm10 vsubps %ymm7, %ymm10, %ymm7 vmovaps %ymm14, 0x680(%rsp) vmulps %ymm14, %ymm9, %ymm9 vmovaps %ymm12, 0x6a0(%rsp) vmulps %ymm6, %ymm12, %ymm6 vsubps %ymm9, %ymm6, %ymm6 vbroadcastss 0x50(%r13,%rsi,4), %ymm13 vbroadcastss 0x60(%r13,%rsi,4), %ymm2 vmulps %ymm2, %ymm6, %ymm6 vmulps %ymm7, %ymm13, %ymm7 vbroadcastss 0x40(%r13,%rsi,4), %ymm3 vmovaps %ymm3, 0x80(%rsp) vaddps %ymm7, %ymm6, %ymm6 vmulps %ymm3, %ymm15, %ymm7 vaddps %ymm6, %ymm7, %ymm10 vmovaps %ymm11, %ymm3 vsubps %ymm11, %ymm5, %ymm11 vmovaps 0x1a0(%rsp), %ymm9 vsubps %ymm9, %ymm0, %ymm15 vmovaps %ymm5, 0x380(%rsp) vaddps %ymm3, %ymm5, %ymm6 vmovaps %ymm0, 0x6c0(%rsp) vaddps %ymm0, %ymm9, %ymm7 vmovaps %ymm9, %ymm0 vmulps %ymm6, %ymm15, %ymm14 vmulps %ymm7, %ymm11, %ymm5 vsubps %ymm14, %ymm5, %ymm5 vmovaps 0x100(%rsp), %ymm12 vsubps %ymm12, %ymm1, %ymm14 vmulps %ymm7, %ymm14, %ymm7 vmovaps %ymm1, 0x3a0(%rsp) vaddps %ymm1, %ymm12, %ymm9 vmovaps %ymm12, %ymm1 vmulps %ymm15, %ymm9, %ymm12 vsubps %ymm7, %ymm12, %ymm7 vmovaps %ymm11, %ymm12 vmulps %ymm11, %ymm9, %ymm9 vmulps %ymm6, %ymm14, %ymm6 vsubps %ymm9, %ymm6, %ymm6 vmulps %ymm2, %ymm6, %ymm6 vmulps %ymm7, %ymm13, %ymm7 vaddps %ymm7, %ymm6, %ymm6 vmovaps 0x80(%rsp), %ymm9 vmulps %ymm5, %ymm9, %ymm5 vaddps %ymm6, %ymm5, %ymm6 vsubps %ymm4, %ymm1, %ymm7 vaddps %ymm4, %ymm1, %ymm4 vsubps %ymm8, %ymm3, %ymm11 vaddps %ymm3, %ymm8, %ymm1 vmovaps 0xe0(%rsp), %ymm3 vsubps %ymm3, %ymm0, %ymm8 vaddps %ymm3, %ymm0, %ymm0 vmulps %ymm1, %ymm8, %ymm3 vmulps %ymm0, %ymm11, %ymm5 vsubps %ymm3, %ymm5, %ymm3 vmulps %ymm0, %ymm7, %ymm0 vmulps %ymm4, %ymm8, %ymm5 vsubps %ymm0, %ymm5, %ymm0 vmulps %ymm4, %ymm11, %ymm4 vmulps %ymm1, %ymm7, %ymm1 vsubps %ymm4, %ymm1, %ymm1 vmovaps %ymm2, 0xe0(%rsp) vmulps %ymm2, %ymm1, %ymm1 vmovaps %ymm13, 0x100(%rsp) vmulps %ymm0, %ymm13, %ymm0 vaddps %ymm0, %ymm1, %ymm0 vmulps %ymm3, %ymm9, %ymm1 vaddps %ymm0, %ymm1, %ymm1 vaddps %ymm6, %ymm10, %ymm0 vaddps %ymm0, %ymm1, %ymm9 vminps %ymm6, %ymm10, %ymm3 vminps %ymm1, %ymm3, %ymm3 vbroadcastss 0x1ce7571(%rip), %ymm4 # 0x1f20ec4 vandps %ymm4, %ymm9, %ymm4 vbroadcastss 0x1ce756c(%rip), %ymm0 # 0x1f20ecc vmovaps %ymm4, 0x5c0(%rsp) vmulps %ymm0, %ymm4, %ymm4 vbroadcastss 0x1ce754a(%rip), %ymm0 # 0x1f20ec0 vxorps %ymm0, %ymm4, %ymm5 vcmpnltps %ymm5, %ymm3, %ymm3 vmovaps %ymm10, 0x640(%rsp) vmovaps %ymm6, 0x620(%rsp) vmaxps %ymm6, %ymm10, %ymm5 vmaxps %ymm1, %ymm5, %ymm1 vcmpleps %ymm4, %ymm1, %ymm1 vorps %ymm1, %ymm3, %ymm3 vtestps %ymm3, %ymm3 je 0x239d03 vmovaps 0x660(%rsp), %ymm2 vmovaps %ymm11, %ymm6 vmovaps %ymm12, %ymm11 vmulps %ymm2, %ymm12, %ymm1 vmovaps 0x680(%rsp), %ymm13 vmulps %ymm15, %ymm13, %ymm4 vsubps %ymm1, %ymm4, %ymm4 vmulps %ymm6, %ymm15, %ymm5 vmovaps %ymm9, 0x1a0(%rsp) vmulps %ymm8, %ymm12, %ymm9 vsubps %ymm5, %ymm9, %ymm9 vbroadcastss 0x1ce74d0(%rip), %ymm0 # 0x1f20ec4 vandps %ymm0, %ymm1, %ymm1 vandps %ymm0, %ymm5, %ymm5 vcmpltps %ymm5, %ymm1, %ymm1 vblendvps %ymm1, %ymm4, %ymm9, %ymm1 vmulps %ymm8, %ymm14, %ymm4 vmulps %ymm2, %ymm14, %ymm5 vmovaps 0x6a0(%rsp), %ymm10 vmulps %ymm15, %ymm10, %ymm8 vsubps %ymm8, %ymm5, %ymm5 vmulps %ymm7, %ymm15, %ymm9 vsubps %ymm4, %ymm9, %ymm9 vandps %ymm0, %ymm8, %ymm8 vandps %ymm0, %ymm4, %ymm4 vcmpltps %ymm4, %ymm8, %ymm4 vblendvps %ymm4, %ymm5, %ymm9, %ymm12 vmovaps 0x1a0(%rsp), %ymm9 vmulps %ymm7, %ymm11, %ymm4 vmulps %ymm11, %ymm10, %ymm5 vmulps %ymm13, %ymm14, %ymm7 vmulps %ymm6, %ymm14, %ymm2 vsubps %ymm7, %ymm5, %ymm5 vsubps %ymm4, %ymm2, %ymm2 vandps %ymm0, %ymm7, %ymm7 vandps %ymm0, %ymm4, %ymm4 vcmpltps %ymm4, %ymm7, %ymm4 vblendvps %ymm4, %ymm5, %ymm2, %ymm2 vextractf128 $0x1, %ymm3, %xmm4 vpackssdw %xmm4, %xmm3, %xmm5 vmulps 0xe0(%rsp), %ymm2, %ymm3 vmulps 0x100(%rsp), %ymm12, %ymm4 vaddps %ymm4, %ymm3, %ymm3 vmulps 0x80(%rsp), %ymm1, %ymm4 vaddps %ymm3, %ymm4, %ymm3 vaddps %ymm3, %ymm3, %ymm4 vmulps 0x6c0(%rsp), %ymm2, %ymm3 vmulps 0x380(%rsp), %ymm12, %ymm7 vaddps %ymm7, %ymm3, %ymm3 vmulps 0x3a0(%rsp), %ymm1, %ymm7 vaddps %ymm3, %ymm7, %ymm3 vaddps %ymm3, %ymm3, %ymm3 vrcpps %ymm4, %ymm7 vmulps %ymm7, %ymm4, %ymm8 vbroadcastss 0x1cb2c37(%rip), %ymm0 # 0x1eec714 vsubps %ymm8, %ymm0, %ymm8 vmulps %ymm7, %ymm8, %ymm8 vaddps %ymm7, %ymm8, %ymm7 vmulps %ymm7, %ymm3, %ymm3 vbroadcastss 0x30(%r13,%rsi,4), %ymm7 vcmpleps %ymm3, %ymm7, %ymm7 vbroadcastss 0x80(%r13,%rsi,4), %ymm8 vcmpleps %ymm8, %ymm3, %ymm8 vandps %ymm7, %ymm8, %ymm7 vextractf128 $0x1, %ymm7, %xmm8 vpackssdw %xmm8, %xmm7, %xmm7 vpand %xmm5, %xmm7, %xmm5 vpmovsxwd %xmm5, %xmm7 vpshufd $0xee, %xmm5, %xmm8 # xmm8 = xmm5[2,3,2,3] vpmovsxwd %xmm8, %xmm8 vinsertf128 $0x1, %xmm8, %ymm7, %ymm7 vtestps %ymm7, %ymm7 je 0x239d03 vcmpneqps 0x1ce73ba(%rip), %ymm4, %ymm4 # 0x1f20f00 vextractf128 $0x1, %ymm4, %xmm7 vpackssdw %xmm7, %xmm4, %xmm4 vpand %xmm4, %xmm5, %xmm4 vpmovsxwd %xmm4, %xmm5 vpunpckhwd %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[4,4,5,5,6,6,7,7] vinsertf128 $0x1, %xmm4, %ymm5, %ymm4 vtestps %ymm4, %ymm4 je 0x239d03 vmovaps %ymm9, 0x460(%rsp) leaq 0x1f(%rsp), %rax movq %rax, 0x480(%rsp) vmovaps %ymm4, 0x4a0(%rsp) vmovaps %ymm3, 0x500(%rsp) vmovaps 0x620(%rsp), %ymm6 vsubps %ymm6, %ymm9, %ymm3 vmovaps 0x5a0(%rsp), %ymm5 vmovaps 0x640(%rsp), %ymm0 vblendvps %ymm5, %ymm3, %ymm0, %ymm4 vsubps %ymm0, %ymm9, %ymm3 vblendvps %ymm5, %ymm3, %ymm6, %ymm3 vmovaps %ymm4, 0x420(%rsp) vmovaps %ymm3, 0x440(%rsp) vmovaps 0x580(%rsp), %ymm0 vmulps %ymm1, %ymm0, %ymm1 vmovaps %ymm1, 0x520(%rsp) vmulps %ymm0, %ymm12, %ymm1 vmovaps %ymm1, 0x540(%rsp) vmulps %ymm2, %ymm0, %ymm1 vmovaps %ymm1, 0x560(%rsp) movq 0x190(%rsp), %rcx movq 0x188(%rsp), %rdx movzwl 0x8(%rcx,%rdx), %eax decl %eax vcvtsi2ss %eax, %xmm14, %xmm1 vrcpss %xmm1, %xmm1, %xmm2 vmulss %xmm1, %xmm2, %xmm1 vmovss 0x1cb73c6(%rip), %xmm0 # 0x1ef0ff8 vsubss %xmm1, %xmm0, %xmm1 vmulss %xmm1, %xmm2, %xmm1 movzwl 0xa(%rcx,%rdx), %eax decl %eax vcvtsi2ss %eax, %xmm14, %xmm2 vrcpss %xmm2, %xmm2, %xmm5 vmulss %xmm2, %xmm5, %xmm2 vsubss %xmm2, %xmm0, %xmm2 vmulss %xmm2, %xmm5, %xmm2 vmulps 0x600(%rsp), %ymm9, %ymm5 vaddps %ymm4, %ymm5, %ymm4 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmulps %ymm1, %ymm4, %ymm1 vmovaps %ymm1, 0x420(%rsp) vmulps 0x5e0(%rsp), %ymm9, %ymm4 vaddps %ymm3, %ymm4, %ymm3 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vmulps %ymm2, %ymm3, %ymm2 vmovaps %ymm2, 0x440(%rsp) movq 0x198(%rsp), %rax movq 0x1e8(%rax), %rax movq (%rax,%r14,8), %r10 movl 0x90(%r13,%rsi,4), %eax testl %eax, 0x34(%r10) je 0x239d03 movq 0xc0(%rsp), %r12 movq 0x10(%r12), %rax cmpq $0x0, 0x10(%rax) jne 0x239d3b cmpq $0x0, 0x48(%r10) jne 0x239d3b movq 0x50(%rsp), %r15 movq 0x28(%rsp), %rdi movq 0x20(%rsp), %r8 movq 0x30(%rsp), %rdx movq 0x58(%rsp), %rbx movq 0x40(%rsp), %r9 andl $0x0, 0x1d0(%rsp,%rsi,4) jmp 0x239d29 movq 0xc0(%rsp), %r12 movq 0x50(%rsp), %r15 movq 0x28(%rsp), %rdi movq 0x20(%rsp), %r8 movq 0x30(%rsp), %rdx movq 0x58(%rsp), %rbx movq 0x40(%rsp), %r9 leaq -0x1(%r9), %rax andq %rax, %r9 jne 0x2393cb jmp 0x23a0a4 vrcpps %ymm9, %ymm3 vmulps %ymm3, %ymm9, %ymm0 vbroadcastss 0x1cb29c7(%rip), %ymm4 # 0x1eec714 vsubps %ymm0, %ymm4, %ymm0 vmulps %ymm0, %ymm3, %ymm0 vaddps %ymm0, %ymm3, %ymm0 vbroadcastss 0x1cb7286(%rip), %ymm3 # 0x1ef0fe8 vmovaps 0x5c0(%rsp), %ymm5 vcmpnltps %ymm3, %ymm5, %ymm3 vandps %ymm0, %ymm3, %ymm0 vmulps %ymm0, %ymm1, %ymm1 vminps %ymm4, %ymm1, %ymm1 vmovaps %ymm1, 0x4c0(%rsp) vmulps %ymm0, %ymm2, %ymm0 vminps %ymm4, %ymm0, %ymm0 vmovaps %ymm0, 0x4e0(%rsp) vmovaps 0x4a0(%rsp), %ymm0 vmovmskps %ymm0, %r11d bsfq %r11, %rcx testl %r11d, %r11d setne 0x1a0(%rsp) movq 0x50(%rsp), %r15 movq 0x28(%rsp), %rdi movq 0x20(%rsp), %r8 movq 0x30(%rsp), %rdx movq 0x58(%rsp), %rbx movq 0x40(%rsp), %r9 je 0x23a091 pushq $0x1 popq %rax movq %rcx, 0x80(%rsp) movl %esi, %ecx shll %cl, %eax cltq shlq $0x4, %rax leaq 0x1f1618e(%rip), %rcx # 0x214ff80 vmovaps (%rcx,%rax), %xmm0 vmovaps %xmm0, 0x3a0(%rsp) movq 0x80(%rsp), %rcx movq %r10, 0x380(%rsp) vmovss 0x80(%r13,%rsi,4), %xmm5 vmovss 0x500(%rsp,%rcx,4), %xmm0 vbroadcastss 0x4c0(%rsp,%rcx,4), %xmm1 vbroadcastss 0x4e0(%rsp,%rcx,4), %xmm2 vmovss %xmm0, 0x80(%r13,%rsi,4) movq 0x8(%r12), %rax vbroadcastss 0x520(%rsp,%rcx,4), %xmm0 vbroadcastss 0x540(%rsp,%rcx,4), %xmm3 movq %rcx, 0x80(%rsp) vbroadcastss 0x560(%rsp,%rcx,4), %xmm4 vmovaps %xmm0, 0x2f0(%rsp) vmovaps %xmm3, 0x300(%rsp) vmovaps %xmm4, 0x310(%rsp) vmovaps %xmm1, 0x320(%rsp) vmovaps %xmm2, 0x330(%rsp) vmovaps 0x3d0(%rsp), %xmm0 vmovaps %xmm0, 0x340(%rsp) vmovaps 0x3e0(%rsp), %xmm0 vmovaps %xmm0, 0x350(%rsp) leaq 0x360(%rsp), %rcx vcmptrueps %ymm0, %ymm0, %ymm0 vmovups %ymm0, (%rcx) vbroadcastss (%rax), %xmm0 vmovaps %xmm0, 0x360(%rsp) vbroadcastss 0x4(%rax), %xmm0 vmovaps %xmm0, 0x370(%rsp) vmovaps 0x3a0(%rsp), %xmm0 vmovaps %xmm0, 0x130(%rsp) leaq 0x130(%rsp), %rcx movq %rcx, 0x1e0(%rsp) movq 0x18(%r10), %rcx movq %rcx, 0x1e8(%rsp) movq %rax, 0x1f0(%rsp) movq %r13, 0x1f8(%rsp) leaq 0x2f0(%rsp), %rax movq %rax, 0x200(%rsp) movl $0x4, 0x208(%rsp) movq 0x48(%r10), %rax testq %rax, %rax movq %r11, 0x100(%rsp) vmovss %xmm5, 0xe0(%rsp) je 0x239f9d leaq 0x1e0(%rsp), %rdi vzeroupper callq %rax vmovss 0xe0(%rsp), %xmm5 movq 0x100(%rsp), %r11 movq 0x380(%rsp), %r10 movq 0x60(%rsp), %rsi movq 0x40(%rsp), %r9 movq 0x30(%rsp), %rdx movq 0x20(%rsp), %r8 movq 0x28(%rsp), %rdi vmovdqa 0x130(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x23a044 movq 0x10(%r12), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x23a00a testb $0x2, (%rcx) jne 0x239fcb testb $0x40, 0x3e(%r10) je 0x23a00a leaq 0x1e0(%rsp), %rdi vzeroupper callq %rax vmovss 0xe0(%rsp), %xmm5 movq 0x100(%rsp), %r11 movq 0x380(%rsp), %r10 movq 0x60(%rsp), %rsi movq 0x40(%rsp), %r9 movq 0x30(%rsp), %rdx movq 0x20(%rsp), %r8 movq 0x28(%rsp), %rdi vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x130(%rsp), %xmm0, %xmm1 vpxor 0x1cb1e01(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x1f8(%rsp), %rax vbroadcastss 0x1cb2b54(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x23a054 vpcmpeqd 0x1cb19c4(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cb1dcc(%rip), %xmm0, %xmm0 # 0x1eebe20 vmovddup 0x1ce6e8c(%rip), %xmm1 # xmm1 = mem[0,0] vptest %xmm1, %xmm0 movq 0x80(%rsp), %rcx jne 0x23a091 movq 0xb8(%rsp), %rax vmovss %xmm5, (%rax,%rsi,4) btcq %rcx, %r11 bsfq %r11, %rcx testq %r11, %r11 setne 0x1a0(%rsp) jne 0x239e10 testb $0x1, 0x1a0(%rsp) jne 0x239cf9 jmp 0x239d29 vmovaps 0x1d0(%rsp), %xmm0 vmovaps 0xd0(%rsp), %xmm1 vtestps %xmm1, %xmm0 vandps %xmm1, %xmm0, %xmm0 sete %al jmp 0x23a0c6 xorl %eax, %eax vmovaps %xmm0, 0xd0(%rsp) testb %al, %al jne 0x23a0eb movq 0x168(%rsp), %rcx leaq -0x1(%rcx), %rax andq %rax, %rcx movq %rcx, %rax jne 0x238f6b movq 0x160(%rsp), %rax incq %rax cmpq 0x158(%rsp), %rax jne 0x238eef vpcmpeqd %xmm1, %xmm1, %xmm1 vpxor 0xd0(%rsp), %xmm1, %xmm0 vpor 0x70(%rsp), %xmm0, %xmm0 vmovdqa %xmm0, 0x70(%rsp) vtestps %xmm1, %xmm0 jb 0x23a149 vmovaps 0x2e0(%rsp), %xmm1 vbroadcastss 0x1cb2a4e(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x2e0(%rsp) xorl %eax, %eax jmp 0x23a14c pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x238ba9 jmp 0x23a224 pushq $0x2 jmp 0x23a14b vmovaps %xmm15, 0x80(%rsp) movq %r8, 0x20(%rsp) movq %rdi, 0x28(%rsp) movq %rdx, 0x30(%rsp) bsfq %rbx, %r15 movq 0x150(%rsp), %rdi movq 0x148(%rsp), %rsi movq %r15, %rcx leaq 0x3f(%rsp), %r8 movq %r13, %r9 pushq %r12 leaq 0x218(%rsp), %rax pushq %rax vzeroupper callq 0x28632c popq %rcx popq %rdx testb %al, %al je 0x23a1b4 orl $-0x1, 0x70(%rsp,%r15,4) leaq -0x1(%rbx), %rax andq %rax, %rbx movq 0x30(%rsp), %rdx jne 0x23a176 vmovaps 0x70(%rsp), %xmm0 vtestps 0x1cb1c4f(%rip), %xmm0 # 0x1eebe20 pushq $0x3 popq %rax movq 0x50(%rsp), %r15 movq 0x28(%rsp), %rdi movq 0x20(%rsp), %r8 vmovaps 0x80(%rsp), %xmm15 jb 0x238bed vmovaps 0x2e0(%rsp), %xmm1 vbroadcastss 0x1cb2980(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps 0x80(%rsp), %xmm15 vmovaps %xmm0, 0x2e0(%rsp) pushq $0x2 popq %rax jmp 0x238bed vmovaps 0x3c0(%rsp), %xmm0 vandps 0x70(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cb2948(%rip), %xmm1 # 0x1eecb84 movq 0xb8(%rsp), %rax vmaskmovps %xmm1, %xmm0, (%rax) leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::SubGridIntersectorKPluecker<4, 4, true>, true>::intersect(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xf8, %rsp movq (%rsi), %r13 cmpq $0x8, 0x70(%r13) je 0x23a43f movq %rcx, %r15 movq %rdx, %r14 movq %rsi, %r12 cmpq $0x0, 0x8(%rcx) je 0x23a299 movq 0x10(%r15), %rax testb $0x1, 0x2(%rax) jne 0x23a451 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x23a43f movzbl %al, %ebp vmovaps (%r14), %xmm1 leaq 0x10(%rsp), %rax vmovaps %xmm1, (%rax) vmovaps 0x10(%r14), %xmm1 vmovaps %xmm1, 0x10(%rax) vmovaps 0x20(%r14), %xmm1 vmovaps %xmm1, 0x20(%rax) vmovaps 0x40(%r14), %xmm1 vmovaps %xmm1, 0x30(%rax) vmovaps 0x50(%r14), %xmm2 vmovaps %xmm2, 0x40(%rax) vmovaps 0x60(%r14), %xmm3 vmovaps %xmm3, 0x50(%rax) vbroadcastss 0x1ce6bc6(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm1, %xmm5 vbroadcastss 0x1cb6cdd(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1cb23fb(%rip), %xmm7 # 0x1eec714 vdivps %xmm1, %xmm7, %xmm1 vandps %xmm4, %xmm2, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vdivps %xmm2, %xmm7, %xmm2 vandps %xmm4, %xmm3, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vbroadcastss 0x1ce6c24(%rip), %xmm6 # 0x1f20f60 vblendvps %xmm5, %xmm6, %xmm1, %xmm1 vblendvps %xmm8, %xmm6, %xmm2, %xmm2 vdivps %xmm3, %xmm7, %xmm3 vblendvps %xmm4, %xmm6, %xmm3, %xmm3 vmovaps %xmm1, 0x60(%rax) vmovaps %xmm2, 0x70(%rax) vmovaps %xmm3, 0x80(%rax) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %xmm4, %xmm1, %xmm1 vbroadcastss 0x1d205ee(%rip), %xmm5 # 0x1f5a964 vandps %xmm5, %xmm1, %xmm1 vmovaps %xmm1, 0x90(%rax) vcmpnltps %xmm4, %xmm2, %xmm1 vbroadcastss 0x1d1a610(%rip), %xmm2 # 0x1f549a0 vbroadcastss 0x1d205cf(%rip), %xmm5 # 0x1f5a968 vblendvps %xmm1, %xmm2, %xmm5, %xmm1 vmovaps %xmm1, 0xa0(%rax) vcmpnltps %xmm4, %xmm3, %xmm1 vbroadcastss 0x1d205b7(%rip), %xmm2 # 0x1f5a96c vbroadcastss 0x1d205b2(%rip), %xmm3 # 0x1f5a970 vblendvps %xmm1, %xmm2, %xmm3, %xmm1 vmovaps %xmm1, 0xb0(%rax) vmovaps 0x30(%r14), %xmm1 vmovaps 0x80(%r14), %xmm2 vmaxps %xmm4, %xmm1, %xmm1 vmaxps %xmm4, %xmm2, %xmm2 vbroadcastss 0x1cb1634(%rip), %xmm3 # 0x1eeba20 vblendvps %xmm0, %xmm1, %xmm3, %xmm1 vmovaps %xmm1, 0xc0(%rax) vbroadcastss 0x1cb2781(%rip), %xmm1 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0xd0(%rax) bsfq %rbp, %rcx leaq -0x1(%rbp), %rbx movq 0x70(%r13), %rdx movq %r12, %rdi movq %r13, %rsi leaq 0xf(%rsp), %r8 movq %r14, %r9 pushq %r15 leaq 0x18(%rsp), %rax pushq %rax callq 0x289aa2 popq %rax popq %rcx andq %rbx, %rbp jne 0x23a411 addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %r12, %rsi movq %r14, %rdx movq %r15, %rcx addq $0xf8, %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x2879ca	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, true, embree::avx::SubGridIntersectorKPluecker<4, 4, true>, true>::occluded(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayK<K>& __restrict__ ray, RayQueryContext* context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { occludedCoherent(valid_i, This, ray, context); return; } #endif / filter out already occluded and invalid rays / vbool<K> valid = (valid_i == -1) & (ray.tfar >= 0.0f); #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / const size_t valid_bits = movemask(valid); if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); vbool<K> terminated = !valid; const vfloat<K> inf = vfloat<K>(pos_inf); / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); if (occluded1(This, bvh, cur, i, pre, ray, tray, context)) set(terminated, i); } if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(shadow.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we push the previously hit node onto the stack, and continue with the currently hit child / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push 'cur' node onto stack and continue with hit child / if (likely(cur != BVH::emptyNode)) { sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } } if (unlikely(cur == BVH::emptyNode)) goto pop; #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(shadow.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive) cur.leaf(items); size_t lazy_node = 0; terminated \|= PrimitiveIntersectorK::occluded(!terminated, This, pre, ray, context, prim, items, tray, lazy_node); if (all(terminated)) break; tray.tfar = select(terminated, vfloat<K>(neg_inf), tray.tfar); // ignore node intersections for terminated rays if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } vfloat<K>::store(valid & terminated, &ray.tfar, neg_inf); }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1c58, %rsp # imm = 0x1C58 movq (%rsi), %rax movq %rax, 0x178(%rsp) cmpq $0x8, 0x70(%rax) je 0x23c19c movq %rcx, %r10 movq %rdx, %r9 movq 0x8(%rcx), %rax testq %rax, %rax je 0x23a4b4 movq 0x10(%r10), %rcx testb $0x1, 0x2(%rcx) jne 0x23c1b1 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm3 vmovaps 0x80(%r9), %xmm1 vxorps %xmm2, %xmm2, %xmm2 vcmpnltps %xmm2, %xmm1, %xmm4 vtestps %xmm3, %xmm4 je 0x23c19c vandps %xmm3, %xmm4, %xmm10 vmovaps (%r9), %xmm3 vmovaps %xmm3, 0x270(%rsp) vmovaps 0x10(%r9), %xmm3 vmovaps %xmm3, 0x280(%rsp) vmovaps 0x20(%r9), %xmm3 vmovaps %xmm3, 0x290(%rsp) vmovaps 0x40(%r9), %xmm3 vmovaps %xmm3, 0x2a0(%rsp) vmovaps 0x50(%r9), %xmm4 vmovaps %xmm4, 0x2b0(%rsp) vmovaps 0x60(%r9), %xmm5 vmovaps %xmm5, 0x2c0(%rsp) vbroadcastss 0x1ce6985(%rip), %xmm8 # 0x1f20ec4 vandps %xmm3, %xmm8, %xmm6 vbroadcastss 0x1cb6a9c(%rip), %xmm9 # 0x1ef0fe8 vcmpltps %xmm9, %xmm6, %xmm6 vbroadcastss 0x1cb21b9(%rip), %xmm11 # 0x1eec714 vdivps %xmm3, %xmm11, %xmm3 vandps %xmm4, %xmm8, %xmm7 vcmpltps %xmm9, %xmm7, %xmm7 vdivps %xmm4, %xmm11, %xmm4 vandps %xmm5, %xmm8, %xmm8 vcmpltps %xmm9, %xmm8, %xmm8 vbroadcastss 0x1ce69e0(%rip), %xmm9 # 0x1f20f60 vblendvps %xmm6, %xmm9, %xmm3, %xmm3 vblendvps %xmm7, %xmm9, %xmm4, %xmm4 vdivps %xmm5, %xmm11, %xmm5 vblendvps %xmm8, %xmm9, %xmm5, %xmm5 vmovaps %xmm3, 0x2d0(%rsp) vmovaps %xmm4, 0x2e0(%rsp) vmovaps %xmm5, 0x2f0(%rsp) vcmpltps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1d203a5(%rip), %xmm6 # 0x1f5a964 vandps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0x300(%rsp) vcmpnltps %xmm2, %xmm4, %xmm3 vbroadcastss 0x1d1a3c6(%rip), %xmm4 # 0x1f549a0 vbroadcastss 0x1d20385(%rip), %xmm6 # 0x1f5a968 vblendvps %xmm3, %xmm4, %xmm6, %xmm3 vcmpnltps %xmm2, %xmm5, %xmm4 vbroadcastss 0x1d20375(%rip), %xmm5 # 0x1f5a96c vbroadcastss 0x1d20370(%rip), %xmm6 # 0x1f5a970 vblendvps %xmm4, %xmm5, %xmm6, %xmm4 vmovaps %xmm3, 0x310(%rsp) vmovaps %xmm4, 0x320(%rsp) vmovaps 0x30(%r9), %xmm3 vmaxps %xmm2, %xmm3, %xmm3 vmaxps %xmm2, %xmm1, %xmm2 vbroadcastss 0x1cb13f1(%rip), %xmm11 # 0x1eeba20 vblendvps %xmm10, %xmm3, %xmm11, %xmm1 vmovaps %xmm1, 0x330(%rsp) vbroadcastss 0x1cb253d(%rip), %xmm3 # 0x1eecb84 vblendvps %xmm10, %xmm2, %xmm3, %xmm2 vmovaps %xmm2, 0x340(%rsp) vmovaps %xmm10, 0x3e0(%rsp) vxorps %xmm0, %xmm10, %xmm0 vmovaps %xmm0, 0x130(%rsp) testq %rax, %rax movq %rsi, 0x240(%rsp) je 0x23a68d movq 0x10(%r10), %rax movb 0x2(%rax), %al andb $0x1, %al xorl %r13d, %r13d cmpb $0x1, %al adcq $0x2, %r13 jmp 0x23a691 pushq $0x3 popq %r13 leaq 0x80(%r9), %rax movq %rax, 0x110(%rsp) leaq 0x5c0(%rsp), %r12 movq $-0x8, -0x10(%r12) leaq 0xd60(%rsp), %r15 vmovaps %xmm11, -0x20(%r15) movq 0x178(%rsp), %rax movq 0x70(%rax), %rax movq %rax, -0x8(%r12) vmovaps %xmm1, -0x10(%r15) pushq $0x8 popq %rdi vbroadcastss 0x1ce582e(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1ce5829(%rip), %xmm12 # 0x1f1ff14 vpcmpeqd %xmm13, %xmm13, %xmm13 movq %r10, 0x48(%rsp) movq %r9, 0x40(%rsp) addq $-0x10, %r15 movq -0x8(%r12), %rbp addq $-0x8, %r12 cmpq $-0x8, %rbp je 0x23c096 vmovaps (%r15), %xmm14 vcmpltps 0x340(%rsp), %xmm14, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x23c0a7 movzbl %al, %r14d popcntl %r14d, %ebx xorl %eax, %eax cmpq %r13, %rbx jbe 0x23c0ab cmpq %r13, %rbx jbe 0x23c099 testb $0x8, %bpl jne 0x23a8d9 movq %rbp, %rax movq %rbp, %rcx andq $-0x10, %rcx xorl %edx, %edx movq %rdi, %rbp vmovaps %xmm11, %xmm14 movq (%rcx,%rdx,8), %rsi cmpq $0x8, %rsi je 0x23a877 vbroadcastss 0x20(%rax,%rdx,4), %xmm0 vmovaps 0x270(%rsp), %xmm1 vmovaps 0x280(%rsp), %xmm2 vmovaps 0x290(%rsp), %xmm3 vmovaps 0x2d0(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm5 vbroadcastss 0x40(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmovaps 0x2e0(%rsp), %xmm6 vmulps %xmm0, %xmm6, %xmm7 vbroadcastss 0x60(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmovaps 0x2f0(%rsp), %xmm8 vmulps %xmm0, %xmm8, %xmm9 vbroadcastss 0x30(%rax,%rdx,4), %xmm0 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm1 vbroadcastss 0x50(%rax,%rdx,4), %xmm0 vsubps %xmm2, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm2 vbroadcastss 0x70(%rax,%rdx,4), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm8, %xmm3 vminps %xmm1, %xmm5, %xmm0 vminps %xmm2, %xmm7, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vminps %xmm3, %xmm9, %xmm4 vmaxps %xmm4, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vmaxps %xmm1, %xmm5, %xmm1 vmaxps %xmm2, %xmm7, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmaxps %xmm3, %xmm9, %xmm2 vminps %xmm2, %xmm1, %xmm1 vmulps %xmm1, %xmm12, %xmm1 vmaxps 0x330(%rsp), %xmm0, %xmm2 vminps 0x340(%rsp), %xmm1, %xmm1 vcmpleps %xmm1, %xmm2, %xmm1 vtestps %xmm1, %xmm1 je 0x23a877 vblendvps %xmm1, %xmm0, %xmm11, %xmm0 cmpq $0x8, %rbp je 0x23a870 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm14, (%r15) addq $0x10, %r15 vmovaps %xmm0, %xmm14 movq %rsi, %rbp cmpq $0x8, %rsi je 0x23a88e leaq 0x1(%rdx), %rsi cmpq $0x3, %rdx movq %rsi, %rdx jb 0x23a767 xorl %eax, %eax cmpq $0x8, %rbp je 0x23a8d2 vmovaps 0x340(%rsp), %xmm0 vcmpnleps %xmm14, %xmm0, %xmm0 vmovmskps %xmm0, %ecx popcntl %ecx, %edx movb $0x1, %cl cmpq %rdx, %r13 jae 0x23a8c1 testb %cl, %cl jne 0x23a749 jmp 0x23c099 movq %rbp, (%r12) addq $0x8, %r12 vmovaps %xmm14, (%r15) addq $0x10, %r15 xorl %ecx, %ecx pushq $0x4 popq %rax jmp 0x23a8b4 cmpq $-0x8, %rbp je 0x23c096 vmovaps 0x340(%rsp), %xmm0 vcmpnleps %xmm14, %xmm0, %xmm0 vtestps %xmm0, %xmm0 je 0x23c0a7 movl %ebp, %eax andl $0xf, %eax vmovdqa 0x130(%rsp), %xmm0 addq $-0x8, %rax movq %rax, 0x248(%rsp) je 0x23c058 andq $-0x10, %rbp vpxor %xmm0, %xmm13, %xmm14 xorl %eax, %eax movq %r13, 0x258(%rsp) movq %rax, 0x250(%rsp) imulq $0x58, %rax, %rdx vmovq 0x20(%rbp,%rdx), %xmm0 vmovq 0x24(%rbp,%rdx), %xmm1 vpminub %xmm1, %xmm0, %xmm1 vpcmpeqb %xmm1, %xmm0, %xmm0 vpxor %xmm0, %xmm13, %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vpxor %xmm0, %xmm13, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vmovmskps %xmm0, %eax testl %eax, %eax je 0x23c03a addq %rbp, %rdx movzbl %al, %esi movq %rdx, 0x260(%rsp) bsfq %rsi, %rcx vmovq 0x20(%rdx), %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vcvtdq2ps %xmm0, %xmm0 vbroadcastss 0x44(%rdx), %xmm1 vbroadcastss 0x38(%rdx), %xmm2 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x4b0(%rsp) vmovq 0x24(%rdx), %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vcvtdq2ps %xmm0, %xmm0 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x350(%rsp) vmovq 0x28(%rdx), %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vcvtdq2ps %xmm0, %xmm0 vbroadcastss 0x48(%rdx), %xmm1 vbroadcastss 0x3c(%rdx), %xmm2 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x140(%rsp) vmovq 0x2c(%rdx), %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vcvtdq2ps %xmm0, %xmm0 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0xc0(%rsp) vmovq 0x30(%rdx), %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vcvtdq2ps %xmm0, %xmm0 vbroadcastss 0x4c(%rdx), %xmm1 vbroadcastss 0x40(%rdx), %xmm2 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x190(%rsp) vmovq 0x34(%rdx), %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vcvtdq2ps %xmm0, %xmm0 vmulps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x180(%rsp) vbroadcastss 0x4b0(%rsp,%rcx,4), %xmm0 vmovaps 0x270(%rsp), %xmm1 vmovaps 0x280(%rsp), %xmm2 vmovaps 0x290(%rsp), %xmm3 vmovaps 0x2d0(%rsp), %xmm4 vsubps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm4, %xmm0 vbroadcastss 0x140(%rsp,%rcx,4), %xmm5 vsubps %xmm2, %xmm5, %xmm5 vmovaps 0x2e0(%rsp), %xmm6 vmulps %xmm5, %xmm6, %xmm5 vbroadcastss 0x190(%rsp,%rcx,4), %xmm7 vsubps %xmm3, %xmm7, %xmm7 vmovaps 0x2f0(%rsp), %xmm8 vmulps %xmm7, %xmm8, %xmm7 vbroadcastss 0x350(%rsp,%rcx,4), %xmm9 vsubps %xmm1, %xmm9, %xmm1 vmulps %xmm1, %xmm4, %xmm1 vbroadcastss 0xc0(%rsp,%rcx,4), %xmm4 vsubps %xmm2, %xmm4, %xmm2 vmulps %xmm2, %xmm6, %xmm2 vbroadcastss 0x180(%rsp,%rcx,4), %xmm4 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm8, %xmm3 vminps %xmm1, %xmm0, %xmm4 vminps %xmm2, %xmm5, %xmm6 vmaxps %xmm6, %xmm4, %xmm4 vminps %xmm3, %xmm7, %xmm6 vmaxps %xmm6, %xmm4, %xmm4 vmulps %xmm4, %xmm10, %xmm4 vmaxps %xmm1, %xmm0, %xmm0 vmaxps %xmm2, %xmm5, %xmm1 vminps %xmm1, %xmm0, %xmm0 vmaxps %xmm3, %xmm7, %xmm1 vminps %xmm1, %xmm0, %xmm0 vmulps %xmm0, %xmm12, %xmm0 vmaxps 0x330(%rsp), %xmm4, %xmm1 vminps 0x340(%rsp), %xmm0, %xmm0 vcmpleps %xmm0, %xmm1, %xmm0 vtestps %xmm14, %xmm0 je 0x23c027 movq %rsi, 0x268(%rsp) movzwl (%rdx,%rcx,8), %r8d movzwl 0x2(%rdx,%rcx,8), %esi movl 0x50(%rdx), %eax movq %rax, 0x10(%rsp) movl 0x4(%rdx,%rcx,8), %r13d movq %r13, 0x20(%rsp) movq (%r10), %rdx movq 0x1e8(%rdx), %rdx movq (%rdx,%rax,8), %r9 movq 0x58(%r9), %r11 movq %r11, 0x118(%rsp) movq 0x90(%r9), %rdx imulq 0x68(%r9), %r13 movl %r8d, %eax movl $0x7fff, %edi # imm = 0x7FFF andl %edi, %eax movl (%r11,%r13), %r10d addl %eax, %r10d movl %esi, %ecx andl %edi, %ecx movl 0x4(%r11,%r13), %edi movl %edi, %r14d imull %ecx, %r14d addl %r10d, %r14d leaq 0x1(%r14), %r10 movq 0xa0(%r9), %r9 movq %r9, %r11 imulq %r14, %r11 vmovups (%rdx,%r11), %xmm2 imulq %r9, %r10 vmovups (%rdx,%r10), %xmm0 leaq (%r14,%rdi), %r11 leaq (%r14,%rdi), %r10 incq %r10 movq %r11, %rbx imulq %r9, %rbx vmovups (%rdx,%rbx), %xmm3 movq %r10, %rbx imulq %r9, %rbx vmovups (%rdx,%rbx), %xmm1 xorl %ebx, %ebx testw %r8w, %r8w setns %bl leaq (%r14,%rbx), %r8 incq %r8 imulq %r9, %r8 vmovups (%rdx,%r8), %xmm5 addq %r10, %rbx movq %rbx, %r8 imulq %r9, %r8 vmovups (%rdx,%r8), %xmm4 movl %eax, %r8d testw %si, %si movl $0x0, %esi cmovsq %rsi, %rdi movq 0x118(%rsp), %rsi addq %rdi, %r11 imulq %r9, %r11 vmovups (%rdx,%r11), %xmm6 movl %ecx, %r11d addq %rdi, %r10 imulq %r9, %r10 vmovups (%rdx,%r10), %xmm7 movq 0x48(%rsp), %r10 addq %rbx, %rdi imulq %r9, %rdi movq 0x40(%rsp), %r9 vmovups (%rdx,%rdi), %xmm8 movq %r13, %rdi vmovss 0x20(%rsp), %xmm9 vmovaps %xmm2, 0x4b0(%rsp) vmovaps %xmm0, 0x4c0(%rsp) vmovaps %xmm1, 0x4d0(%rsp) vmovaps %xmm3, 0x4e0(%rsp) vmovaps %xmm0, 0x4f0(%rsp) vmovaps %xmm5, 0x500(%rsp) vmovaps %xmm4, 0x510(%rsp) vmovaps %xmm1, 0x520(%rsp) vmovaps %xmm3, 0x530(%rsp) vmovaps %xmm1, 0x540(%rsp) vmovaps %xmm7, 0x550(%rsp) vmovaps %xmm6, 0x560(%rsp) vmovaps %xmm1, 0x570(%rsp) vmovaps %xmm4, 0x580(%rsp) vmovaps %xmm8, 0x590(%rsp) vmovaps %xmm7, 0x5a0(%rsp) vmovss 0x10(%rsp), %xmm0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmovaps %xmm0, 0x410(%rsp) vshufps $0x0, %xmm9, %xmm9, %xmm0 # xmm0 = xmm9[0,0,0,0] vmovaps %xmm0, 0x400(%rsp) leaq 0x4e8(%rsp), %r13 xorl %ebx, %ebx vmovaps %xmm14, 0x3f0(%rsp) vmovaps %xmm14, 0x10(%rsp) vbroadcastss -0x38(%r13), %xmm13 vbroadcastss -0x34(%r13), %xmm2 vbroadcastss -0x30(%r13), %xmm1 vbroadcastss -0x8(%r13), %xmm0 vbroadcastss -0x4(%r13), %xmm5 vbroadcastss (%r13), %xmm6 vmovaps 0x410(%rsp), %xmm3 vmovaps %xmm3, 0x190(%rsp) vmovaps 0x400(%rsp), %xmm3 vmovaps %xmm3, 0x180(%rsp) vmovaps 0x10(%r9), %xmm10 vmovaps 0x20(%r9), %xmm3 vmovaps %xmm3, 0xa0(%rsp) vsubps %xmm10, %xmm2, %xmm4 vsubps %xmm3, %xmm1, %xmm9 vmovaps %xmm5, 0x490(%rsp) vsubps %xmm10, %xmm5, %xmm5 vmovaps %xmm5, 0xb0(%rsp) vmovaps %xmm6, 0x480(%rsp) vsubps %xmm3, %xmm6, %xmm2 vmovaps %xmm2, 0x20(%rsp) vsubps %xmm4, %xmm5, %xmm1 vsubps %xmm9, %xmm2, %xmm3 vaddps %xmm4, %xmm5, %xmm5 vaddps %xmm2, %xmm9, %xmm6 vmulps %xmm3, %xmm5, %xmm7 vmulps %xmm6, %xmm1, %xmm8 vsubps %xmm7, %xmm8, %xmm11 vmovaps (%r9), %xmm7 vsubps %xmm7, %xmm13, %xmm14 vmovaps %xmm0, 0x4a0(%rsp) vsubps %xmm7, %xmm0, %xmm13 vsubps %xmm14, %xmm13, %xmm0 vmulps %xmm6, %xmm0, %xmm6 vaddps %xmm14, %xmm13, %xmm8 vmovaps %xmm3, 0x210(%rsp) vmulps %xmm3, %xmm8, %xmm3 vsubps %xmm6, %xmm3, %xmm3 vmovaps %xmm1, 0x220(%rsp) vmulps %xmm1, %xmm8, %xmm6 vmovaps %xmm0, 0x1e0(%rsp) vmulps %xmm5, %xmm0, %xmm5 vsubps %xmm6, %xmm5, %xmm6 vmovaps 0x60(%r9), %xmm15 vmulps %xmm6, %xmm15, %xmm8 vmovaps 0x50(%r9), %xmm12 vmulps %xmm3, %xmm12, %xmm3 vaddps %xmm3, %xmm8, %xmm3 vmovaps 0x40(%r9), %xmm0 vmovaps %xmm0, 0xf0(%rsp) vmulps %xmm0, %xmm11, %xmm11 vaddps %xmm3, %xmm11, %xmm8 vbroadcastss -0x24(%r13), %xmm0 vmovaps %xmm0, 0x470(%rsp) vsubps %xmm10, %xmm0, %xmm3 vbroadcastss -0x20(%r13), %xmm0 vmovaps %xmm0, 0x460(%rsp) vsubps 0xa0(%rsp), %xmm0, %xmm0 vsubps %xmm3, %xmm4, %xmm2 vsubps %xmm0, %xmm9, %xmm11 vmovaps %xmm4, 0xe0(%rsp) vaddps %xmm3, %xmm4, %xmm1 vmovaps %xmm9, 0x1d0(%rsp) vaddps %xmm0, %xmm9, %xmm4 vmulps %xmm1, %xmm11, %xmm10 vmulps %xmm4, %xmm2, %xmm5 vsubps %xmm10, %xmm5, %xmm5 vbroadcastss -0x28(%r13), %xmm6 vmovaps %xmm6, 0x450(%rsp) vsubps %xmm7, %xmm6, %xmm7 vsubps %xmm7, %xmm14, %xmm10 vmulps %xmm4, %xmm10, %xmm4 vmovaps %xmm14, 0x230(%rsp) vaddps %xmm7, %xmm14, %xmm14 vmulps %xmm11, %xmm14, %xmm6 vsubps %xmm4, %xmm6, %xmm4 vmovaps %xmm2, 0x1c0(%rsp) vmulps %xmm2, %xmm14, %xmm6 vmulps %xmm1, %xmm10, %xmm1 vsubps %xmm6, %xmm1, %xmm1 vmovaps %xmm8, %xmm6 vmulps %xmm1, %xmm15, %xmm1 vmulps %xmm4, %xmm12, %xmm4 vaddps %xmm4, %xmm1, %xmm1 vmovaps 0xf0(%rsp), %xmm14 vmulps %xmm5, %xmm14, %xmm4 vaddps %xmm1, %xmm4, %xmm8 vsubps %xmm13, %xmm7, %xmm9 vaddps %xmm7, %xmm13, %xmm1 vmovaps 0x10(%rsp), %xmm7 vmovaps 0xb0(%rsp), %xmm4 vsubps %xmm4, %xmm3, %xmm2 vaddps %xmm4, %xmm3, %xmm3 vmovaps 0x20(%rsp), %xmm4 vsubps %xmm4, %xmm0, %xmm13 vaddps %xmm4, %xmm0, %xmm0 vmulps %xmm3, %xmm13, %xmm4 vmulps %xmm0, %xmm2, %xmm5 vsubps %xmm4, %xmm5, %xmm4 vmulps %xmm0, %xmm9, %xmm0 vmulps %xmm1, %xmm13, %xmm5 vsubps %xmm0, %xmm5, %xmm0 vmovaps %xmm2, 0x1a0(%rsp) vmulps %xmm2, %xmm1, %xmm1 vmovaps %xmm9, 0x1b0(%rsp) vmulps %xmm3, %xmm9, %xmm3 vsubps %xmm1, %xmm3, %xmm1 vmovaps %xmm15, 0x200(%rsp) vmulps %xmm1, %xmm15, %xmm1 vmovaps %xmm12, 0x1f0(%rsp) vmulps %xmm0, %xmm12, %xmm0 vaddps %xmm0, %xmm1, %xmm0 vmulps %xmm4, %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm6, %xmm8, %xmm1 vaddps %xmm1, %xmm0, %xmm1 vbroadcastss 0x1ce5eeb(%rip), %xmm5 # 0x1f20ec4 vmovaps %xmm1, 0xa0(%rsp) vandps %xmm5, %xmm1, %xmm1 vbroadcastss 0x1ce5edd(%rip), %xmm3 # 0x1f20ecc vmulps %xmm3, %xmm1, %xmm1 vminps %xmm8, %xmm6, %xmm3 vminps %xmm0, %xmm3, %xmm3 vbroadcastss 0x1ce5ebb(%rip), %xmm4 # 0x1f20ec0 vxorps %xmm4, %xmm1, %xmm4 vcmpnltps %xmm4, %xmm3, %xmm3 vmaxps %xmm8, %xmm6, %xmm4 vmaxps %xmm0, %xmm4, %xmm0 vcmpleps %xmm1, %xmm0, %xmm0 vorps %xmm0, %xmm3, %xmm0 vtestps %xmm7, %xmm0 vmovaps %xmm7, 0x10(%rsp) vandps %xmm7, %xmm0, %xmm12 vbroadcastss -0x18(%r13), %xmm0 vmovaps %xmm0, 0x440(%rsp) vbroadcastss -0x14(%r13), %xmm0 vmovaps %xmm0, 0x430(%rsp) vbroadcastss -0x10(%r13), %xmm0 vmovaps %xmm0, 0x420(%rsp) je 0x23b93a vmovaps %xmm14, %xmm9 vmovaps %xmm8, 0xb0(%rsp) vmovaps %xmm6, 0x20(%rsp) vmovaps 0x210(%rsp), %xmm8 vmovaps 0x1c0(%rsp), %xmm15 vmulps %xmm8, %xmm15, %xmm0 vmovaps %xmm5, %xmm6 vmovaps 0x220(%rsp), %xmm5 vmulps %xmm5, %xmm11, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmovaps 0x1a0(%rsp), %xmm2 vmulps %xmm2, %xmm11, %xmm3 vmulps %xmm13, %xmm15, %xmm4 vsubps %xmm3, %xmm4, %xmm4 vandps %xmm6, %xmm0, %xmm0 vandps %xmm6, %xmm3, %xmm3 vcmpltps %xmm3, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm4, %xmm7 vmulps %xmm13, %xmm10, %xmm0 vmulps %xmm8, %xmm10, %xmm1 vmovaps 0x1e0(%rsp), %xmm8 vmulps %xmm11, %xmm8, %xmm3 vsubps %xmm3, %xmm1, %xmm1 vmovaps 0x1b0(%rsp), %xmm13 vmulps %xmm11, %xmm13, %xmm4 vsubps %xmm0, %xmm4, %xmm4 vandps %xmm6, %xmm3, %xmm3 vandps %xmm6, %xmm0, %xmm0 vcmpltps %xmm0, %xmm3, %xmm0 vblendvps %xmm0, %xmm1, %xmm4, %xmm3 vmulps %xmm15, %xmm13, %xmm0 vmulps %xmm15, %xmm8, %xmm1 vmovaps %xmm3, %xmm15 vmulps %xmm5, %xmm10, %xmm3 vmovaps %xmm6, %xmm5 vmulps %xmm2, %xmm10, %xmm4 vsubps %xmm3, %xmm1, %xmm1 vsubps %xmm0, %xmm4, %xmm4 vandps %xmm6, %xmm3, %xmm3 vandps %xmm6, %xmm0, %xmm0 vcmpltps %xmm0, %xmm3, %xmm0 vblendvps %xmm0, %xmm1, %xmm4, %xmm14 vmulps 0x200(%rsp), %xmm14, %xmm0 vmulps 0x1f0(%rsp), %xmm15, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps %xmm7, %xmm9, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vmulps 0x1d0(%rsp), %xmm14, %xmm1 vmulps 0xe0(%rsp), %xmm15, %xmm3 vaddps %xmm3, %xmm1, %xmm1 vmovaps %xmm7, 0xf0(%rsp) vmulps 0x230(%rsp), %xmm7, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm1, %xmm1, %xmm1 vrcpps %xmm0, %xmm2 vmulps %xmm2, %xmm0, %xmm3 vbroadcastss 0x1cb1570(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vmulps %xmm2, %xmm1, %xmm10 vmovaps 0x30(%r9), %xmm1 vcmpleps 0x80(%r9), %xmm10, %xmm2 vcmpleps %xmm10, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm1 vcmpneqps 0x1cb0839(%rip), %xmm0, %xmm0 # 0x1eeba10 vandps %xmm1, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm0 vtestps %xmm12, %xmm0 vandps %xmm0, %xmm12, %xmm12 je 0x23b93a vbroadcastss 0x1cb0823(%rip), %xmm11 # 0x1eeba20 vpcmpeqd %xmm13, %xmm13, %xmm13 vmovaps 0x20(%rsp), %xmm6 vmovaps 0xb0(%rsp), %xmm8 vtestps %xmm12, %xmm12 je 0x23b325 movl %ebx, %eax andl $0x1, %eax addl %r8d, %eax movl %ebx, %ecx shrl %ecx addl %r11d, %ecx movzwl 0x8(%rsi,%rdi), %edx decl %edx vcvtsi2ss %edx, %xmm9, %xmm0 vrcpss %xmm0, %xmm0, %xmm1 vmulss %xmm0, %xmm1, %xmm0 vmovss 0x1cb5db2(%rip), %xmm3 # 0x1ef0ff8 vsubss %xmm0, %xmm3, %xmm0 vmulss %xmm0, %xmm1, %xmm0 movzwl 0xa(%rsi,%rdi), %edx decl %edx vcvtsi2ss %edx, %xmm9, %xmm1 vcvtsi2ss %eax, %xmm9, %xmm2 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmovaps 0xa0(%rsp), %xmm4 vmulps %xmm4, %xmm2, %xmm2 vaddps %xmm2, %xmm6, %xmm2 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmulps %xmm0, %xmm2, %xmm0 vrcpss %xmm1, %xmm1, %xmm2 vmulss %xmm1, %xmm2, %xmm1 vblendvps %xmm12, %xmm0, %xmm6, %xmm6 vsubss %xmm1, %xmm3, %xmm0 vcvtsi2ss %ecx, %xmm9, %xmm1 vmulss %xmm0, %xmm2, %xmm0 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmulps %xmm4, %xmm1, %xmm1 vaddps %xmm1, %xmm8, %xmm1 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmulps %xmm0, %xmm1, %xmm0 vblendvps %xmm12, %xmm0, %xmm8, %xmm8 movq (%r10), %rcx movl 0x190(%rsp,%rbx,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %r14 vbroadcastss 0x34(%r14), %xmm0 vandps 0x90(%r9), %xmm0, %xmm0 vpcmpeqd 0x1cb0732(%rip), %xmm0, %xmm0 # 0x1eeba10 vtestps %xmm12, %xmm0 vmovaps 0x10(%rsp), %xmm1 jb 0x23b317 movl 0x180(%rsp,%rbx,4), %ecx vandnps %xmm12, %xmm0, %xmm0 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x23b9f9 cmpq $0x0, 0x48(%r14) jne 0x23b9f9 vpandn 0x10(%rsp), %xmm0, %xmm1 vandnps 0x10(%rsp), %xmm12, %xmm0 vmovaps %xmm1, 0x10(%rsp) jmp 0x23b32b vmovaps 0x10(%rsp), %xmm0 vtestps %xmm0, %xmm0 je 0x23b8be vmovaps %xmm10, 0x60(%rsp) vmovaps %xmm14, 0x70(%rsp) vmovaps %xmm15, 0x80(%rsp) vmovaps %xmm8, 0xb0(%rsp) vmovaps %xmm6, 0x20(%rsp) vmovaps (%r9), %xmm0 vmovaps 0x10(%r9), %xmm3 vmovaps 0x20(%r9), %xmm5 vmovaps 0x440(%rsp), %xmm1 vsubps %xmm0, %xmm1, %xmm4 vmovaps 0x430(%rsp), %xmm1 vsubps %xmm3, %xmm1, %xmm2 vmovaps 0x420(%rsp), %xmm1 vsubps %xmm5, %xmm1, %xmm11 vmovaps 0x4a0(%rsp), %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmovaps %xmm1, 0x50(%rsp) vmovaps 0x490(%rsp), %xmm1 vsubps %xmm3, %xmm1, %xmm15 vmovaps 0x480(%rsp), %xmm1 vsubps %xmm5, %xmm1, %xmm14 vmovaps 0x450(%rsp), %xmm1 vsubps %xmm0, %xmm1, %xmm12 vmovaps 0x470(%rsp), %xmm1 vsubps %xmm3, %xmm1, %xmm13 vmovaps 0x460(%rsp), %xmm1 vsubps %xmm5, %xmm1, %xmm0 vmovaps %xmm0, 0xe0(%rsp) vsubps %xmm4, %xmm12, %xmm1 vsubps %xmm2, %xmm13, %xmm5 vsubps %xmm11, %xmm0, %xmm10 vaddps %xmm2, %xmm13, %xmm6 vaddps %xmm0, %xmm11, %xmm7 vmulps %xmm6, %xmm10, %xmm8 vmulps %xmm7, %xmm5, %xmm9 vsubps %xmm8, %xmm9, %xmm9 vaddps %xmm4, %xmm12, %xmm8 vmulps %xmm7, %xmm1, %xmm7 vmovaps %xmm10, 0x1d0(%rsp) vmulps %xmm10, %xmm8, %xmm10 vsubps %xmm7, %xmm10, %xmm10 vmovaps %xmm5, 0x1e0(%rsp) vmulps %xmm5, %xmm8, %xmm7 vmovaps %xmm1, 0x1f0(%rsp) vmulps %xmm6, %xmm1, %xmm6 vsubps %xmm7, %xmm6, %xmm7 vmovaps 0x60(%r9), %xmm8 vmulps %xmm7, %xmm8, %xmm1 vmovaps 0x50(%r9), %xmm0 vmovaps %xmm0, 0x230(%rsp) vmulps %xmm0, %xmm10, %xmm10 vaddps %xmm1, %xmm10, %xmm10 vmovaps 0x40(%r9), %xmm3 vmulps %xmm3, %xmm9, %xmm9 vaddps %xmm10, %xmm9, %xmm7 vmovaps %xmm15, %xmm9 vsubps %xmm15, %xmm2, %xmm0 vmovaps %xmm14, %xmm5 vsubps %xmm14, %xmm11, %xmm15 vmovaps %xmm2, 0x210(%rsp) vaddps %xmm2, %xmm9, %xmm14 vmovaps %xmm11, 0x200(%rsp) vaddps %xmm5, %xmm11, %xmm1 vmovaps %xmm5, %xmm11 vmulps %xmm15, %xmm14, %xmm10 vmulps %xmm1, %xmm0, %xmm2 vsubps %xmm10, %xmm2, %xmm2 vmovaps 0x50(%rsp), %xmm6 vsubps %xmm6, %xmm4, %xmm10 vmulps %xmm1, %xmm10, %xmm1 vmovaps %xmm4, 0x220(%rsp) vaddps %xmm6, %xmm4, %xmm5 vmovaps %xmm6, %xmm4 vmulps %xmm5, %xmm15, %xmm6 vsubps %xmm1, %xmm6, %xmm1 vmovaps %xmm0, 0x1b0(%rsp) vmulps %xmm0, %xmm5, %xmm5 vmulps %xmm14, %xmm10, %xmm6 vsubps %xmm5, %xmm6, %xmm5 vmulps %xmm5, %xmm8, %xmm5 vmovaps 0x230(%rsp), %xmm6 vmulps %xmm1, %xmm6, %xmm1 vaddps %xmm1, %xmm5, %xmm1 vmulps %xmm2, %xmm3, %xmm2 vaddps %xmm1, %xmm2, %xmm5 vsubps %xmm12, %xmm4, %xmm14 vaddps %xmm4, %xmm12, %xmm0 vsubps %xmm13, %xmm9, %xmm12 vaddps %xmm13, %xmm9, %xmm1 vmovaps 0xe0(%rsp), %xmm2 vsubps %xmm2, %xmm11, %xmm9 vaddps %xmm2, %xmm11, %xmm2 vmulps %xmm1, %xmm9, %xmm11 vmulps %xmm2, %xmm12, %xmm4 vsubps %xmm11, %xmm4, %xmm11 vmulps %xmm2, %xmm14, %xmm2 vmulps %xmm0, %xmm9, %xmm4 vsubps %xmm2, %xmm4, %xmm2 vmulps %xmm0, %xmm12, %xmm0 vmulps %xmm1, %xmm14, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm8, 0xe0(%rsp) vmulps %xmm0, %xmm8, %xmm0 vmovaps %xmm6, %xmm8 vmulps %xmm2, %xmm6, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmovaps %xmm3, 0x1c0(%rsp) vmulps %xmm3, %xmm11, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm5, %xmm7, %xmm1 vaddps %xmm1, %xmm0, %xmm1 vmovaps %xmm1, 0x100(%rsp) vbroadcastss 0x1ce5925(%rip), %xmm2 # 0x1f20ec4 vmovaps %xmm2, %xmm6 vandps %xmm2, %xmm1, %xmm1 vbroadcastss 0x1ce591c(%rip), %xmm2 # 0x1f20ecc vmulps %xmm2, %xmm1, %xmm1 vminps %xmm5, %xmm7, %xmm2 vminps %xmm0, %xmm2, %xmm2 vbroadcastss 0x1ce58fb(%rip), %xmm3 # 0x1f20ec0 vxorps %xmm3, %xmm1, %xmm3 vcmpnltps %xmm3, %xmm2, %xmm2 vmovaps %xmm7, 0xd0(%rsp) vmaxps %xmm5, %xmm7, %xmm3 vmovaps 0x10(%rsp), %xmm7 vmaxps %xmm0, %xmm3, %xmm0 vcmpleps %xmm1, %xmm0, %xmm0 vorps %xmm0, %xmm2, %xmm0 vtestps %xmm7, %xmm0 vmovaps %xmm7, 0x10(%rsp) vandps %xmm7, %xmm0, %xmm0 je 0x23b995 vmovaps %xmm12, %xmm4 vmovaps %xmm5, 0x50(%rsp) vmovaps 0x1d0(%rsp), %xmm11 vmovaps 0x1b0(%rsp), %xmm13 vmovaps %xmm0, 0x1a0(%rsp) vmulps %xmm11, %xmm13, %xmm0 vmovaps %xmm6, %xmm5 vmovaps 0x1e0(%rsp), %xmm6 vmulps %xmm6, %xmm15, %xmm1 vsubps %xmm0, %xmm1, %xmm1 vmulps %xmm4, %xmm15, %xmm2 vmulps %xmm9, %xmm13, %xmm3 vsubps %xmm2, %xmm3, %xmm3 vandps %xmm5, %xmm0, %xmm0 vandps %xmm5, %xmm2, %xmm2 vcmpltps %xmm2, %xmm0, %xmm0 vblendvps %xmm0, %xmm1, %xmm3, %xmm4 vmulps %xmm9, %xmm10, %xmm0 vmulps %xmm11, %xmm10, %xmm1 vmovaps 0x1f0(%rsp), %xmm11 vmulps %xmm15, %xmm11, %xmm2 vsubps %xmm2, %xmm1, %xmm1 vmulps %xmm15, %xmm14, %xmm3 vsubps %xmm0, %xmm3, %xmm3 vandps %xmm5, %xmm2, %xmm2 vandps %xmm5, %xmm0, %xmm0 vcmpltps %xmm0, %xmm2, %xmm0 vblendvps %xmm0, %xmm1, %xmm3, %xmm15 vmulps %xmm13, %xmm14, %xmm0 vmulps %xmm13, %xmm11, %xmm1 vmulps %xmm6, %xmm10, %xmm2 vmulps %xmm12, %xmm10, %xmm3 vsubps %xmm2, %xmm1, %xmm1 vsubps %xmm0, %xmm3, %xmm3 vandps %xmm5, %xmm2, %xmm2 vmovaps %xmm5, %xmm6 vandps %xmm5, %xmm0, %xmm0 vcmpltps %xmm0, %xmm2, %xmm0 vblendvps %xmm0, %xmm1, %xmm3, %xmm9 vmulps 0xe0(%rsp), %xmm9, %xmm0 vmulps %xmm15, %xmm8, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps 0x1c0(%rsp), %xmm4, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vaddps %xmm0, %xmm0, %xmm0 vmulps 0x200(%rsp), %xmm9, %xmm1 vmulps 0x210(%rsp), %xmm15, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmovaps %xmm4, 0x120(%rsp) vmulps 0x220(%rsp), %xmm4, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vaddps %xmm1, %xmm1, %xmm1 vrcpps %xmm0, %xmm2 vmulps %xmm2, %xmm0, %xmm3 vbroadcastss 0x1cb0fe4(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vmulps %xmm2, %xmm1, %xmm10 vmovaps 0x30(%r9), %xmm1 vcmpleps 0x80(%r9), %xmm10, %xmm2 vcmpleps %xmm10, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm1 vcmpneqps 0x1cb02ad(%rip), %xmm0, %xmm0 # 0x1eeba10 vandps %xmm1, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vpsrad $0x1f, %xmm0, %xmm0 vmovaps 0x1a0(%rsp), %xmm1 vtestps %xmm1, %xmm0 vandps %xmm1, %xmm0, %xmm0 je 0x23b995 vbroadcastss 0x1cb028e(%rip), %xmm11 # 0x1eeba20 vpcmpeqd %xmm13, %xmm13, %xmm13 vmovaps 0x50(%rsp), %xmm5 vtestps %xmm0, %xmm0 je 0x23b8f4 vmovaps 0x100(%rsp), %xmm8 vsubps 0xd0(%rsp), %xmm8, %xmm2 vsubps %xmm5, %xmm8, %xmm1 movl %ebx, %ecx andl $0x1, %ecx addl %r8d, %ecx movl %ebx, %eax movzwl 0x8(%rsi,%rdi), %edx decl %edx vcvtsi2ss %edx, %xmm14, %xmm3 vrcpss %xmm3, %xmm3, %xmm4 vmulss %xmm3, %xmm4, %xmm3 vmovss 0x1cb5815(%rip), %xmm7 # 0x1ef0ff8 vsubss %xmm3, %xmm7, %xmm3 movzwl 0xa(%rsi,%rdi), %edx decl %edx vcvtsi2ss %edx, %xmm14, %xmm5 vmulss %xmm3, %xmm4, %xmm3 vcvtsi2ss %ecx, %xmm14, %xmm4 shrl %eax addl %r11d, %eax vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmulps %xmm4, %xmm8, %xmm4 vaddps %xmm4, %xmm2, %xmm4 vshufps $0x0, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmulps %xmm3, %xmm4, %xmm3 vrcpss %xmm5, %xmm5, %xmm4 vmulss %xmm5, %xmm4, %xmm5 vsubss %xmm5, %xmm7, %xmm5 vblendvps %xmm0, %xmm3, %xmm2, %xmm2 vmovaps %xmm2, 0xd0(%rsp) vmulss %xmm5, %xmm4, %xmm2 vcvtsi2ss %eax, %xmm14, %xmm3 vshufps $0x0, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmulps %xmm3, %xmm8, %xmm3 vaddps %xmm3, %xmm1, %xmm3 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmulps %xmm2, %xmm3, %xmm2 vblendvps %xmm0, %xmm2, %xmm1, %xmm5 movq (%r10), %rcx movl 0x190(%rsp,%rbx,4), %eax movq 0x1e8(%rcx), %rcx movq (%rcx,%rax,8), %r14 vbroadcastss 0x34(%r14), %xmm1 vandps 0x90(%r9), %xmm1, %xmm1 vpcmpeqd 0x1cb0190(%rip), %xmm1, %xmm1 # 0x1eeba10 vtestps %xmm0, %xmm1 jb 0x23b8f4 movl 0x180(%rsp,%rbx,4), %ecx vandnps %xmm0, %xmm1, %xmm0 movq 0x10(%r10), %rdx cmpq $0x0, 0x10(%rdx) jne 0x23bcd3 cmpq $0x0, 0x48(%r14) jne 0x23bcd3 vmovdqa 0x10(%rsp), %xmm1 vpandn %xmm1, %xmm0, %xmm1 vmovdqa %xmm1, 0x10(%rsp) jmp 0x23b8f4 vmovaps %xmm6, 0xd0(%rsp) vmovaps %xmm8, %xmm5 vmovaps 0xa0(%rsp), %xmm0 vmovaps %xmm0, 0x100(%rsp) vmovdqa 0xf0(%rsp), %xmm0 vmovdqa %xmm0, 0x120(%rsp) vmovaps %xmm14, %xmm9 vmovaps %xmm10, 0x60(%rsp) vmovaps %xmm9, 0x70(%rsp) vmovaps %xmm15, 0x80(%rsp) vmovaps %xmm5, 0x50(%rsp) vmovaps 0x10(%rsp), %xmm0 vtestps %xmm0, %xmm0 je 0x23bfdd leaq 0x1(%rbx), %rax addq $0x40, %r13 cmpq $0x3, %rbx movq %rax, %rbx jb 0x23ad4d jmp 0x23bfdd vmovaps 0xd0(%rsp), %xmm6 vmovaps 0x50(%rsp), %xmm8 vmovaps 0x100(%rsp), %xmm0 vmovaps %xmm0, 0xa0(%rsp) vmovaps 0x60(%rsp), %xmm10 vmovaps 0x120(%rsp), %xmm0 vmovaps %xmm0, 0xf0(%rsp) vmovaps 0x80(%rsp), %xmm15 vmovaps 0x70(%rsp), %xmm14 vbroadcastss 0x1cb0095(%rip), %xmm11 # 0x1eeba20 vpcmpeqd %xmm13, %xmm13, %xmm13 jmp 0x23b211 vmovaps 0x20(%rsp), %xmm1 vmovaps %xmm1, 0xd0(%rsp) vmovaps 0xb0(%rsp), %xmm5 vmovaps 0xa0(%rsp), %xmm1 vmovaps %xmm1, 0x100(%rsp) vmovaps 0x60(%rsp), %xmm10 vmovaps 0xf0(%rsp), %xmm1 vmovaps %xmm1, 0x120(%rsp) vmovaps 0x80(%rsp), %xmm15 vmovaps 0x70(%rsp), %xmm9 vbroadcastss 0x1cb0031(%rip), %xmm11 # 0x1eeba20 vpcmpeqd %xmm13, %xmm13, %xmm13 jmp 0x23b79d vmovaps 0xa0(%rsp), %xmm3 vandps %xmm5, %xmm3, %xmm1 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cb0cfd(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cb55bc(%rip), %xmm3 # 0x1ef0fe8 vcmpnltps %xmm3, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm1 vmulps %xmm1, %xmm6, %xmm2 vminps %xmm4, %xmm2, %xmm2 vmulps %xmm1, %xmm8, %xmm1 vminps %xmm4, %xmm1, %xmm1 movq 0x8(%r10), %rdx vmovd %eax, %xmm3 vpshufd $0x0, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmovd %ecx, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovaps 0xf0(%rsp), %xmm5 vmovaps %xmm5, 0x350(%rsp) vmovaps %xmm15, 0x360(%rsp) vmovaps %xmm14, 0x370(%rsp) vmovaps %xmm2, 0x380(%rsp) vmovaps %xmm1, 0x390(%rsp) vmovdqa %xmm4, 0x3a0(%rsp) vmovdqa %xmm3, 0x3b0(%rsp) leaq 0x3c0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x3c0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x3d0(%rsp) vmovaps 0x80(%r9), %xmm3 vblendvps %xmm0, %xmm10, %xmm3, %xmm1 vmovaps %xmm1, 0x80(%r9) vmovaps %xmm0, 0xc0(%rsp) leaq 0xc0(%rsp), %rax movq %rax, 0x140(%rsp) movq 0x18(%r14), %rax movq %rax, 0x148(%rsp) movq 0x8(%r10), %rax movq %rax, 0x150(%rsp) movq %r9, 0x158(%rsp) leaq 0x350(%rsp), %rax movq %rax, 0x160(%rsp) movl $0x4, 0x168(%rsp) movq 0x48(%r14), %rax testq %rax, %rax movq %rdi, 0x98(%rsp) movl %r8d, 0xc(%rsp) movl %r11d, 0x8(%rsp) vmovaps %xmm6, 0x20(%rsp) vmovaps %xmm8, 0xb0(%rsp) vmovaps %xmm15, 0x80(%rsp) vmovaps %xmm14, 0x70(%rsp) vmovaps %xmm10, 0x60(%rsp) vmovaps %xmm12, 0x50(%rsp) vmovaps %xmm3, 0xe0(%rsp) je 0x23bbf8 leaq 0x140(%rsp), %rdi vzeroupper callq %rax vmovaps 0xe0(%rsp), %xmm3 vmovaps 0x50(%rsp), %xmm12 vmovaps 0x60(%rsp), %xmm10 vmovaps 0x70(%rsp), %xmm14 vmovaps 0x80(%rsp), %xmm15 vmovaps 0xb0(%rsp), %xmm8 vmovaps 0x20(%rsp), %xmm6 movl 0x8(%rsp), %r11d movl 0xc(%rsp), %r8d movq 0x98(%rsp), %rdi movq 0x118(%rsp), %rsi vbroadcastss 0x1cafe32(%rip), %xmm11 # 0x1eeba20 movq 0x40(%rsp), %r9 movq 0x48(%rsp), %r10 vmovdqa 0xc0(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x23bf79 movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax vpcmpeqd %xmm13, %xmm13, %xmm13 je 0x23bc9a testb $0x2, (%rcx) jne 0x23bc2a testb $0x40, 0x3e(%r14) je 0x23bc9a leaq 0x140(%rsp), %rdi movq %rsi, %r14 vzeroupper callq %rax vmovaps 0xe0(%rsp), %xmm3 vmovaps 0x50(%rsp), %xmm12 vmovaps 0x60(%rsp), %xmm10 vmovaps 0x70(%rsp), %xmm14 vmovaps 0x80(%rsp), %xmm15 vmovaps 0xb0(%rsp), %xmm8 vmovaps 0x20(%rsp), %xmm6 movl 0x8(%rsp), %r11d movl 0xc(%rsp), %r8d movq 0x98(%rsp), %rdi movq %r14, %rsi vpcmpeqd %xmm13, %xmm13, %xmm13 vbroadcastss 0x1cafd90(%rip), %xmm11 # 0x1eeba20 movq 0x40(%rsp), %r9 movq 0x48(%rsp), %r10 vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0xc0(%rsp), %xmm0, %xmm1 vpxor %xmm1, %xmm13, %xmm0 movq 0x158(%rsp), %rax vbroadcastss 0x1cb0ec8(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x23bf8a vmovaps 0x100(%rsp), %xmm3 vandps %xmm6, %xmm3, %xmm1 vrcpps %xmm3, %xmm2 vmulps %xmm2, %xmm3, %xmm3 vbroadcastss 0x1cb0a23(%rip), %xmm4 # 0x1eec714 vsubps %xmm3, %xmm4, %xmm3 vmulps %xmm3, %xmm2, %xmm3 vaddps %xmm3, %xmm2, %xmm2 vbroadcastss 0x1cb52e2(%rip), %xmm3 # 0x1ef0fe8 vcmpnltps %xmm3, %xmm1, %xmm1 vandps %xmm2, %xmm1, %xmm1 vmulps 0xd0(%rsp), %xmm1, %xmm2 vminps %xmm4, %xmm2, %xmm2 vmulps %xmm1, %xmm5, %xmm1 vminps %xmm4, %xmm1, %xmm1 movq 0x8(%r10), %rdx vmovd %eax, %xmm3 vpshufd $0x0, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vmovd %ecx, %xmm4 vpshufd $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmovaps 0x120(%rsp), %xmm6 vmovaps %xmm6, 0x350(%rsp) vmovaps %xmm15, 0x360(%rsp) vmovaps %xmm9, 0x370(%rsp) vmovaps %xmm2, 0x380(%rsp) vmovaps %xmm1, 0x390(%rsp) vmovdqa %xmm4, 0x3a0(%rsp) vmovdqa %xmm3, 0x3b0(%rsp) leaq 0x3c0(%rsp), %rax vcmptrueps %ymm1, %ymm1, %ymm1 vmovups %ymm1, (%rax) vbroadcastss (%rdx), %xmm1 vmovaps %xmm1, 0x3c0(%rsp) vbroadcastss 0x4(%rdx), %xmm1 vmovaps %xmm1, 0x3d0(%rsp) vmovaps 0x80(%r9), %xmm3 vblendvps %xmm0, %xmm10, %xmm3, %xmm1 vmovaps %xmm1, 0x80(%r9) vmovaps %xmm0, 0xc0(%rsp) leaq 0xc0(%rsp), %rax movq %rax, 0x140(%rsp) movq 0x18(%r14), %rax movq %rax, 0x148(%rsp) movq 0x8(%r10), %rax movq %rax, 0x150(%rsp) movq %r9, 0x158(%rsp) leaq 0x350(%rsp), %rax movq %rax, 0x160(%rsp) movl $0x4, 0x168(%rsp) movq 0x48(%r14), %rax testq %rax, %rax vmovaps %xmm5, 0x50(%rsp) vmovaps %xmm15, 0x80(%rsp) vmovaps %xmm9, 0x70(%rsp) vmovaps %xmm10, 0x60(%rsp) movq %rdi, 0x98(%rsp) movl %r8d, 0xc(%rsp) movl %r11d, 0x8(%rsp) vmovaps %xmm3, 0x20(%rsp) je 0x23beb3 leaq 0x140(%rsp), %rdi vzeroupper callq %rax vmovaps 0x20(%rsp), %xmm3 movl 0x8(%rsp), %r11d movl 0xc(%rsp), %r8d movq 0x98(%rsp), %rdi movq 0x118(%rsp), %rsi vmovaps 0x60(%rsp), %xmm10 vmovaps 0x70(%rsp), %xmm9 vmovaps 0x80(%rsp), %xmm15 vmovaps 0x50(%rsp), %xmm5 vbroadcastss 0x1cafb77(%rip), %xmm11 # 0x1eeba20 movq 0x40(%rsp), %r9 movq 0x48(%rsp), %r10 vmovdqa 0xc0(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x23bfab movq 0x10(%r10), %rcx movq 0x10(%rcx), %rax testq %rax, %rax vpcmpeqd %xmm13, %xmm13, %xmm13 je 0x23bf43 testb $0x2, (%rcx) jne 0x23bee5 testb $0x40, 0x3e(%r14) je 0x23bf43 leaq 0x140(%rsp), %rdi movq %rsi, %r14 vzeroupper callq %rax vmovaps 0x20(%rsp), %xmm3 movl 0x8(%rsp), %r11d movl 0xc(%rsp), %r8d movq 0x98(%rsp), %rdi movq %r14, %rsi vmovaps 0x60(%rsp), %xmm10 vmovaps 0x70(%rsp), %xmm9 vmovaps 0x80(%rsp), %xmm15 vmovaps 0x50(%rsp), %xmm5 vpcmpeqd %xmm13, %xmm13, %xmm13 vbroadcastss 0x1cafae7(%rip), %xmm11 # 0x1eeba20 movq 0x40(%rsp), %r9 movq 0x48(%rsp), %r10 vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0xc0(%rsp), %xmm0, %xmm1 vpxor %xmm1, %xmm13, %xmm0 movq 0x158(%rsp), %rax vbroadcastss 0x1cb0c1f(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x23bfbc vpcmpeqd 0x1cafa8f(%rip), %xmm0, %xmm0 # 0x1eeba10 vpcmpeqd %xmm13, %xmm13, %xmm13 vpxor %xmm0, %xmm13, %xmm0 vpslld $0x1f, %xmm0, %xmm1 vpsrad $0x1f, %xmm1, %xmm0 movq 0x110(%rsp), %rax vblendvps %xmm1, (%rax), %xmm3, %xmm1 vmovaps %xmm1, (%rax) jmp 0x23b311 vpcmpeqd 0x1cafa5d(%rip), %xmm0, %xmm0 # 0x1eeba10 vpcmpeqd %xmm13, %xmm13, %xmm13 vpxor %xmm0, %xmm13, %xmm0 vpslld $0x1f, %xmm0, %xmm1 vpsrad $0x1f, %xmm1, %xmm0 movq 0x110(%rsp), %rax vblendvps %xmm1, (%rax), %xmm3, %xmm1 vmovaps %xmm1, (%rax) jmp 0x23b8ac vmovaps 0x3f0(%rsp), %xmm0 vmovaps 0x10(%rsp), %xmm1 vtestps %xmm0, %xmm1 vandps %xmm0, %xmm1, %xmm14 sete %al movq 0x258(%rsp), %r13 pushq $0x8 popq %rdi vbroadcastss 0x1ce3f04(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1ce3eff(%rip), %xmm12 # 0x1f1ff14 movq 0x260(%rsp), %rdx movq 0x268(%rsp), %rsi jmp 0x23c029 xorl %eax, %eax testb %al, %al jne 0x23c03a leaq -0x1(%rsi), %rax andq %rax, %rsi jne 0x23a97b movq 0x250(%rsp), %rax incq %rax cmpq 0x248(%rsp), %rax jne 0x23a92f vxorps %xmm13, %xmm14, %xmm0 vorps 0x130(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x130(%rsp) vtestps %xmm13, %xmm0 jb 0x23c096 vmovaps 0x340(%rsp), %xmm1 vbroadcastss 0x1cb0b01(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x340(%rsp) xorl %eax, %eax jmp 0x23c099 pushq $0x3 popq %rax cmpl $0x3, %eax jne 0x23a6fa jmp 0x23c174 pushq $0x2 jmp 0x23c098 vmovaps %xmm14, 0x20(%rsp) bsfq %r14, %rcx movq 0x240(%rsp), %rdi movq 0x178(%rsp), %rsi movq %rbp, %rdx movq %rcx, 0x10(%rsp) leaq 0x3f(%rsp), %r8 pushq %r10 leaq 0x278(%rsp), %rax pushq %rax vzeroupper callq 0x28ce16 popq %rcx popq %rdx testb %al, %al je 0x23c0f8 movq 0x10(%rsp), %rax orl $-0x1, 0x130(%rsp,%rax,4) leaq -0x1(%r14), %rax andq %rax, %r14 movq 0x48(%rsp), %r10 movq 0x40(%rsp), %r9 jne 0x23c0b1 vmovaps 0x130(%rsp), %xmm0 vpcmpeqd %xmm13, %xmm13, %xmm13 vtestps %xmm13, %xmm0 pushq $0x3 popq %rax vbroadcastss 0x1caf8f6(%rip), %xmm11 # 0x1eeba20 pushq $0x8 popq %rdi vbroadcastss 0x1ce3dda(%rip), %xmm10 # 0x1f1ff10 vbroadcastss 0x1ce3dd5(%rip), %xmm12 # 0x1f1ff14 vmovaps 0x20(%rsp), %xmm14 jb 0x23a740 vmovaps 0x340(%rsp), %xmm1 vbroadcastss 0x1cb0a27(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm0, %xmm2, %xmm1, %xmm0 vmovaps %xmm0, 0x340(%rsp) pushq $0x2 popq %rax jmp 0x23a740 vmovaps 0x3e0(%rsp), %xmm0 vandps 0x130(%rsp), %xmm0, %xmm0 vbroadcastss 0x1cb09f5(%rip), %xmm1 # 0x1eecb84 movq 0x110(%rsp), %rax vmaskmovps %xmm1, %xmm0, (%rax) addq $0x1c58, %rsp # imm = 0x1C58 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq movq %r9, %rdx movq %r10, %rcx addq $0x1c58, %rsp # imm = 0x1C58 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp jmp 0x28aeca nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::TriangleMIntersectorKMoeller<4, 4, true>>, true>::occluded1(embree::Accel::Intersectors, embree::BVHN<4> const, embree::NodeRefPtr<4>, unsigned long, embree::avx::MoellerTrumboreIntersectorK<4, 4>&, embree::RayK<4>&, embree::avx::TravRayK<4, false> const&, embree::RayQueryContext*)	bool BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded1(Accel::Intersectors* This, const BVH* bvh, NodeRef root, size_t k, Precalculations& pre, RayK<K>& ray, const TravRayK<K, robust>& tray, RayQueryContext* context) { /* stack state / NodeRef stack[stackSizeSingle]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSizeSingle; stack[0] = root; /* load the ray into SIMD registers / TravRay<N,robust> tray1; tray1.template init<K>(k, tray.org, tray.dir, tray.rdir, tray.nearXYZ, tray.tnear[k], tray.tfar[k]); / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes, 1, 1, 1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray1, ray.time()[k], tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / BVHNNodeTraverser1Hit<N, types>::traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves, 1, 1, 1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersectorK::occluded(This, pre, ray, k, context, prim, num, tray1, lazy_node)) { ray.tfar[k] = neg_inf; return true; } if (unlikely(lazy_node)) { stackPtr = lazy_node; stackPtr++; } } return false; }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0xa88, %rsp # imm = 0xA88 movq %r9, %rbx movq %rcx, %r14 movq 0xac0(%rsp), %rax leaq 0x2e8(%rsp), %rdi movq %rdx, -0x8(%rdi) vbroadcastss (%rax,%rcx,4), %xmm3 vbroadcastss 0x10(%rax,%rcx,4), %xmm4 vbroadcastss 0x20(%rax,%rcx,4), %xmm5 vbroadcastss 0x60(%rax,%rcx,4), %xmm6 vbroadcastss 0x70(%rax,%rcx,4), %xmm7 vbroadcastss 0x80(%rax,%rcx,4), %xmm8 movslq 0x90(%rax,%rcx,4), %r8 movslq 0xa0(%rax,%rcx,4), %r9 movslq 0xb0(%rax,%rcx,4), %r10 movq %r8, %r11 xorq $0x10, %r11 movq %r9, %rdx xorq $0x10, %rdx movq %r10, %r12 xorq $0x10, %r12 vbroadcastss 0xc0(%rax,%rcx,4), %xmm9 vbroadcastss 0xd0(%rax,%rcx,4), %xmm10 pushq $0x1 popq %rax shll %cl, %eax movq %rdx, %rcx cltq shlq $0x4, %rax addq 0x1ee5865(%rip), %rax # 0x2124730 movq %rax, 0x150(%rsp) leaq 0x2e0(%rsp), %r15 pushq $0x6 popq %rsi vmovaps %xmm3, 0xd0(%rsp) vmovaps %xmm4, 0xc0(%rsp) vmovaps %xmm5, 0xb0(%rsp) vmovaps %xmm6, 0xa0(%rsp) vmovaps %xmm7, 0x90(%rsp) vmovaps %xmm8, 0x80(%rsp) movq %r12, 0x58(%rsp) vmovaps %xmm9, 0x70(%rsp) vmovaps %xmm10, 0x60(%rsp) movq %rdi, %rax cmpq %r15, %rdi je 0x23f7d9 leaq -0x8(%rax), %rdi movq %rax, 0x158(%rsp) movq -0x8(%rax), %rbp testb $0x8, %bpl jne 0x23efd0 vmovaps 0x20(%rbp,%r8), %xmm0 vsubps %xmm3, %xmm0, %xmm0 vmulps %xmm0, %xmm6, %xmm0 vmovaps 0x20(%rbp,%r9), %xmm1 vsubps %xmm4, %xmm1, %xmm1 vmulps %xmm1, %xmm7, %xmm1 vpmaxsd %xmm1, %xmm0, %xmm0 vmovaps 0x20(%rbp,%r10), %xmm1 vsubps %xmm5, %xmm1, %xmm1 vmulps %xmm1, %xmm8, %xmm1 vpmaxsd %xmm9, %xmm1, %xmm1 vpmaxsd %xmm1, %xmm0, %xmm0 vmovaps 0x20(%rbp,%r11), %xmm1 vsubps %xmm3, %xmm1, %xmm1 vmulps %xmm1, %xmm6, %xmm1 vmovaps 0x20(%rbp,%rcx), %xmm2 vsubps %xmm4, %xmm2, %xmm2 vmulps %xmm2, %xmm7, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vmovaps 0x20(%rbp,%r12), %xmm2 vsubps %xmm5, %xmm2, %xmm2 vmulps %xmm2, %xmm8, %xmm2 vpminsd %xmm10, %xmm2, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpcmpgtd %xmm1, %xmm0, %xmm0 vmovmskps %xmm0, %eax xorb $0xf, %al movzbl %al, %r13d testb $0x8, %bpl jne 0x23f007 testq %r13, %r13 je 0x23f00b andq $-0x10, %rbp bsfq %r13, %rax leaq -0x1(%r13), %rdx movq (%rbp,%rax,8), %rax prefetcht0 (%rax) prefetcht0 0x40(%rax) andq %r13, %rdx jne 0x23f010 movq %rax, %rbp xorl %edx, %edx testl %edx, %edx je 0x23ef41 jmp 0x23f055 movl %esi, %edx jmp 0x23effd pushq $0x4 popq %rdx jmp 0x23effd movq %r15, %rsi movq %rcx, %r15 movq %rax, (%rdi) addq $0x8, %rdi bsfq %rdx, %rcx leaq -0x1(%rdx), %rax movq (%rbp,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) andq %rdx, %rax je 0x23f047 movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rax, %rcx leaq -0x1(%rax), %rdx jmp 0x23f025 movq %rcx, %rbp movq %r15, %rcx movq %rsi, %r15 pushq $0x6 popq %rsi jmp 0x23effb cmpl $0x6, %edx jne 0x23f7a1 movl %ebp, %eax andl $0xf, %eax xorl %edx, %edx addq $-0x8, %rax setne 0xf(%rsp) je 0x23f7a1 andq $-0x10, %rbp xorl %r12d, %r12d imulq $0xb0, %r12, %r15 vmovaps 0x80(%rbp,%r15), %xmm8 vmovaps 0x40(%rbp,%r15), %xmm5 vmulps %xmm5, %xmm8, %xmm0 vmovaps 0x70(%rbp,%r15), %xmm10 vmovaps 0x50(%rbp,%r15), %xmm7 vmulps %xmm7, %xmm10, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x10(%rsp) vmovaps 0x60(%rbp,%r15), %xmm11 vmulps %xmm7, %xmm11, %xmm1 vmovaps (%rbp,%r15), %xmm3 vmovaps 0x10(%rbp,%r15), %xmm13 vmovaps 0x20(%rbp,%r15), %xmm0 vmovaps 0x30(%rbp,%r15), %xmm9 vmulps %xmm9, %xmm8, %xmm2 vsubps %xmm1, %xmm2, %xmm6 vmulps %xmm9, %xmm10, %xmm2 vmulps %xmm5, %xmm11, %xmm12 vsubps %xmm2, %xmm12, %xmm4 vbroadcastss (%rbx,%r14,4), %xmm12 vsubps %xmm12, %xmm3, %xmm2 vbroadcastss 0x10(%rbx,%r14,4), %xmm12 vsubps %xmm12, %xmm13, %xmm3 vbroadcastss 0x20(%rbx,%r14,4), %xmm12 vsubps %xmm12, %xmm0, %xmm1 vbroadcastss 0x50(%rbx,%r14,4), %xmm12 vbroadcastss 0x60(%rbx,%r14,4), %xmm13 vmulps %xmm1, %xmm12, %xmm14 vmulps %xmm3, %xmm13, %xmm15 vsubps %xmm14, %xmm15, %xmm14 vbroadcastss 0x40(%rbx,%r14,4), %xmm15 vmulps %xmm2, %xmm13, %xmm0 vmovaps %xmm1, 0xe0(%rsp) vmulps %xmm1, %xmm15, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm3, 0xf0(%rsp) vmulps %xmm3, %xmm15, %xmm1 vmovaps %xmm2, 0x100(%rsp) vmulps %xmm2, %xmm12, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm4, 0x30(%rsp) vmulps %xmm4, %xmm13, %xmm2 vmovaps 0x10(%rsp), %xmm13 vmovaps %xmm6, 0x40(%rsp) vmulps %xmm6, %xmm12, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm15, %xmm13, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm1, %xmm8, %xmm8 vmulps %xmm0, %xmm10, %xmm10 vaddps %xmm10, %xmm8, %xmm8 vmulps %xmm14, %xmm11, %xmm10 vaddps %xmm8, %xmm10, %xmm8 vmulps %xmm1, %xmm7, %xmm1 vmulps %xmm0, %xmm5, %xmm0 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1ce1d05(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm2, %xmm10 vxorps %xmm8, %xmm10, %xmm7 vmulps %xmm14, %xmm9, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm10, %xmm8 vxorps %xmm9, %xmm9, %xmm9 vcmpnltps %xmm9, %xmm7, %xmm0 vcmpnltps %xmm9, %xmm8, %xmm1 vandps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1ce1cd5(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm2, %xmm5 vcmpneqps %xmm2, %xmm9, %xmm1 vandps %xmm1, %xmm0, %xmm9 vaddps %xmm7, %xmm8, %xmm0 vcmpleps %xmm5, %xmm0, %xmm11 vtestps %xmm11, %xmm9 jne 0x23f222 incq %r12 cmpq %rax, %r12 setb 0xf(%rsp) jne 0x23f07b jmp 0x23f752 vandps %xmm11, %xmm9, %xmm9 vmovaps 0xe0(%rsp), %xmm0 vmulps 0x30(%rsp), %xmm0, %xmm0 vmovaps 0xf0(%rsp), %xmm1 vmulps 0x40(%rsp), %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x100(%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm10, %xmm3 vbroadcastss 0x30(%rbx,%r14,4), %xmm0 vmulps %xmm0, %xmm5, %xmm0 vcmpltps %xmm3, %xmm0, %xmm0 vbroadcastss 0x80(%rbx,%r14,4), %xmm1 vmulps %xmm1, %xmm5, %xmm1 vcmpleps %xmm1, %xmm3, %xmm1 vandps %xmm1, %xmm0, %xmm4 vtestps %xmm9, %xmm4 je 0x23f20c addq %rbp, %r15 vandps %xmm4, %xmm9, %xmm0 vmovaps %xmm7, 0x190(%rsp) vmovaps %xmm8, 0x1a0(%rsp) vmovaps %xmm3, 0x1b0(%rsp) vmovaps %xmm5, 0x1c0(%rsp) vmovaps %xmm0, 0x1e0(%rsp) vmovaps %xmm13, 0x220(%rsp) vmovaps 0x40(%rsp), %xmm1 vmovaps %xmm1, 0x230(%rsp) vmovaps 0x30(%rsp), %xmm1 vmovaps %xmm1, 0x240(%rsp) movq %rax, 0xe0(%rsp) movq 0xac8(%rsp), %rax movq (%rax), %rax movq %rax, 0x100(%rsp) vrcpps %xmm5, %xmm1 vmulps %xmm1, %xmm5, %xmm2 vbroadcastss 0x1cad405(%rip), %xmm3 # 0x1eec714 vsubps %xmm2, %xmm3, %xmm2 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps 0x1b0(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x210(%rsp) vmulps 0x190(%rsp), %xmm1, %xmm2 vmovaps %xmm2, 0x1f0(%rsp) vmulps 0x1a0(%rsp), %xmm1, %xmm1 vmovaps %xmm1, 0x200(%rsp) vmovmskps %xmm0, %eax movq %rax, 0x10(%rsp) movq %rcx, 0x20(%rsp) movq 0x10(%rsp), %rax bsfq %rax, %rax movq %rax, 0x30(%rsp) movl 0x90(%r15,%rax,4), %eax movq 0x100(%rsp), %rcx movq 0x1e8(%rcx), %rcx movq %rax, 0xf0(%rsp) movq (%rcx,%rax,8), %rax movl 0x90(%rbx,%r14,4), %ecx movq %rax, 0x40(%rsp) testl %ecx, 0x34(%rax) je 0x23f414 movq 0xac8(%rsp), %rcx movq 0x10(%rcx), %rcx cmpq $0x0, 0x10(%rcx) vmovaps 0xd0(%rsp), %xmm3 vmovaps 0xc0(%rsp), %xmm4 vmovaps 0xb0(%rsp), %xmm5 vmovaps 0xa0(%rsp), %xmm6 vmovaps 0x90(%rsp), %xmm7 vmovaps 0x80(%rsp), %xmm8 vmovaps 0x70(%rsp), %xmm9 vmovaps 0x60(%rsp), %xmm10 jne 0x23f491 movq 0x40(%rsp), %rax cmpq $0x0, 0x48(%rax) jne 0x23f491 xorl %eax, %eax movq 0x20(%rsp), %rcx jmp 0x23f470 movq 0x10(%rsp), %rax movq 0x30(%rsp), %rcx btcq %rcx, %rax movq %rax, 0x10(%rsp) movb $0x1, %al vmovaps 0xd0(%rsp), %xmm3 vmovaps 0xc0(%rsp), %xmm4 vmovaps 0xb0(%rsp), %xmm5 vmovaps 0xa0(%rsp), %xmm6 vmovaps 0x90(%rsp), %xmm7 vmovaps 0x80(%rsp), %xmm8 movq 0x20(%rsp), %rcx vmovaps 0x70(%rsp), %xmm9 vmovaps 0x60(%rsp), %xmm10 testb %al, %al je 0x23f7b4 cmpq $0x0, 0x10(%rsp) movq 0xe0(%rsp), %rax jne 0x23f35f jmp 0x23f20c movq %r11, 0x128(%rsp) movq %r10, 0x130(%rsp) movq %r9, 0x138(%rsp) movq %r8, 0x140(%rsp) movq %rdi, 0x148(%rsp) vmovss 0x80(%rbx,%r14,4), %xmm0 vmovss %xmm0, 0x2c(%rsp) movq 0x30(%rsp), %rax vmovss 0x210(%rsp,%rax,4), %xmm0 vbroadcastss 0x1f0(%rsp,%rax,4), %xmm1 vbroadcastss 0x200(%rsp,%rax,4), %xmm2 vmovss %xmm0, 0x80(%rbx,%r14,4) movq 0xac8(%rsp), %rcx movq 0x8(%rcx), %rcx vmovss 0xf0(%rsp), %xmm0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vbroadcastss 0xa0(%r15,%rax,4), %xmm3 vbroadcastss 0x220(%rsp,%rax,4), %xmm4 vbroadcastss 0x230(%rsp,%rax,4), %xmm5 vbroadcastss 0x240(%rsp,%rax,4), %xmm6 vmovaps %xmm4, 0x250(%rsp) vmovaps %xmm5, 0x260(%rsp) vmovaps %xmm6, 0x270(%rsp) vmovaps %xmm1, 0x280(%rsp) vmovaps %xmm2, 0x290(%rsp) vmovaps %xmm3, 0x2a0(%rsp) vmovaps %xmm0, 0x2b0(%rsp) leaq 0x2c0(%rsp), %rax vcmptrueps %ymm0, %ymm0, %ymm0 vmovups %ymm0, (%rax) vbroadcastss (%rcx), %xmm0 vmovaps %xmm0, 0x2c0(%rsp) vbroadcastss 0x4(%rcx), %xmm0 vmovaps %xmm0, 0x2d0(%rsp) movq 0x150(%rsp), %rax vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x110(%rsp) leaq 0x110(%rsp), %rax movq %rax, 0x160(%rsp) movq 0x40(%rsp), %rdx movq 0x18(%rdx), %rax movq %rax, 0x168(%rsp) movq %rcx, 0x170(%rsp) movq %rbx, 0x178(%rsp) leaq 0x250(%rsp), %rax movq %rax, 0x180(%rsp) movl $0x4, 0x188(%rsp) movq 0x48(%rdx), %rax testq %rax, %rax je 0x23f61b leaq 0x160(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x110(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x23f697 movq 0xac8(%rsp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x23f65d testb $0x2, (%rcx) jne 0x23f650 movq 0x40(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x23f65d leaq 0x160(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0x110(%rsp), %xmm0, %xmm1 vpxor 0x1cac7ae(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x178(%rsp), %rax vbroadcastss 0x1cad501(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x23f6a7 vpcmpeqd 0x1cac371(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1cac779(%rip), %xmm0, %xmm0 # 0x1eebe20 vpslld $0x1f, %xmm0, %xmm0 vtestps %xmm0, %xmm0 sete %al jne 0x23f6d9 vmovss 0x2c(%rsp), %xmm0 vmovss %xmm0, 0x80(%rbx,%r14,4) movq 0x10(%rsp), %rcx movq 0x30(%rsp), %rdx btcq %rdx, %rcx movq %rcx, 0x10(%rsp) vmovaps 0xd0(%rsp), %xmm3 vmovaps 0xc0(%rsp), %xmm4 vmovaps 0xb0(%rsp), %xmm5 vmovaps 0xa0(%rsp), %xmm6 vmovaps 0x90(%rsp), %xmm7 movq 0x148(%rsp), %rdi vmovaps 0x80(%rsp), %xmm8 movq 0x140(%rsp), %r8 movq 0x138(%rsp), %r9 movq 0x130(%rsp), %r10 movq 0x128(%rsp), %r11 movq 0x20(%rsp), %rcx vmovaps 0x70(%rsp), %xmm9 vmovaps 0x60(%rsp), %xmm10 pushq $0x6 popq %rsi xorl %edx, %edx jmp 0x23f470 vmovaps 0xd0(%rsp), %xmm3 vmovaps 0xc0(%rsp), %xmm4 vmovaps 0xb0(%rsp), %xmm5 vmovaps 0xa0(%rsp), %xmm6 vmovaps 0x90(%rsp), %xmm7 vmovaps 0x80(%rsp), %xmm8 movq 0x58(%rsp), %r12 vmovaps 0x70(%rsp), %xmm9 vmovaps 0x60(%rsp), %xmm10 leaq 0x2e0(%rsp), %r15 testb $0x3, %dl movq 0x158(%rsp), %rax je 0x23ef25 jmp 0x23f7d9 testb $0x1, 0xf(%rsp) movq 0x58(%rsp), %r12 leaq 0x2e0(%rsp), %r15 je 0x23f7a1 movl $0xff800000, 0x80(%rbx,%r14,4) # imm = 0xFF800000 pushq $0x1 popq %rdx jmp 0x23f7a1 cmpq %r15, %rax setne %al addq $0xa88, %rsp # imm = 0xA88 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::TriangleMIntersectorKMoeller<4, 4, false>>, true>::intersect1(embree::Accel::Intersectors, embree::BVHN<4> const, embree::NodeRefPtr<4>, unsigned long, embree::avx::MoellerTrumboreIntersectorK<4, 4>&, embree::RayHitK<4>&, embree::avx::TravRayK<4, false> const&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect1(Accel::Intersectors* This, const BVH* bvh, NodeRef root, size_t k, Precalculations& pre, RayHitK<K>& ray, const TravRayK<K, robust>& tray, RayQueryContext* context) { /* stack state / StackItemT<NodeRef> stack[stackSizeSingle]; // stack of nodes StackItemT<NodeRef> stackPtr = stack + 1; // current stack pointer StackItemT<NodeRef>* stackEnd = stack + stackSizeSingle; stack[0].ptr = root; stack[0].dist = neg_inf; /* load the ray into SIMD registers / TravRay<N,robust> tray1; tray1.template init<K>(k, tray.org, tray.dir, tray.rdir, tray.nearXYZ, tray.tnear[k], tray.tfar[k]); / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = NodeRef(stackPtr->ptr); / if popped node is too far, pop next one / if (unlikely((float)&stackPtr->dist > ray.tfar[k])) continue; / downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(normal.trav_nodes, 1, 1, 1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray1, ray.time()[k], tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(normal.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / BVHNNodeTraverser1Hit<N, types>::traverseClosestHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(normal.trav_leaves, 1, 1, 1); size_t num; Primitive prim = (Primitive*)cur.leaf(num); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(This, pre, ray, k, context, prim, num, tray1, lazy_node); tray1.tfar = ray.tfar[k]; if (unlikely(lazy_node)) { stackPtr->ptr = lazy_node; stackPtr->dist = neg_inf; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x10c8, %rsp # imm = 0x10C8 movq 0x1100(%rsp), %rax leaq 0x190(%rsp), %rsi movq %rdx, -0x10(%rsi) andl $0x0, -0x8(%rsi) vbroadcastss (%rax,%rcx,4), %xmm5 vbroadcastss 0x10(%rax,%rcx,4), %xmm7 vbroadcastss 0x20(%rax,%rcx,4), %xmm8 vbroadcastss 0x60(%rax,%rcx,4), %xmm9 vbroadcastss 0x70(%rax,%rcx,4), %xmm10 vbroadcastss 0x80(%rax,%rcx,4), %xmm11 movslq 0x90(%rax,%rcx,4), %rdx movslq 0xa0(%rax,%rcx,4), %rdi movslq 0xb0(%rax,%rcx,4), %r8 movq %rdx, %r10 xorq $0x10, %r10 movq %rdi, %r11 xorq $0x10, %r11 movq %r8, %rbx xorq $0x10, %rbx vbroadcastss 0xd0(%rax,%rcx,4), %xmm12 vbroadcastss 0xc0(%rax,%rcx,4), %xmm13 vmovaps %xmm5, 0x140(%rsp) vmovaps %xmm7, 0x130(%rsp) vmovaps %xmm8, 0x120(%rsp) vmovaps %xmm9, 0x110(%rsp) vmovaps %xmm10, 0x100(%rsp) vmovaps %xmm11, 0xf0(%rsp) vmovaps %xmm13, 0xe0(%rsp) leaq 0x180(%rsp), %rax cmpq %rax, %rsi je 0x240ede vmovss -0x8(%rsi), %xmm0 addq $-0x10, %rsi vucomiss 0x80(%r9,%rcx,4), %xmm0 ja 0x2406ed movq (%rsi), %r13 testb $0x8, %r13b jne 0x2407ae vmovaps 0x20(%r13,%rdx), %xmm0 vsubps %xmm5, %xmm0, %xmm0 vmulps %xmm0, %xmm9, %xmm0 vmovaps 0x20(%r13,%rdi), %xmm1 vsubps %xmm7, %xmm1, %xmm1 vmulps %xmm1, %xmm10, %xmm1 vpmaxsd %xmm1, %xmm0, %xmm0 vmovaps 0x20(%r13,%r8), %xmm1 vsubps %xmm8, %xmm1, %xmm1 vmulps %xmm1, %xmm11, %xmm1 vpmaxsd %xmm13, %xmm1, %xmm1 vpmaxsd %xmm1, %xmm0, %xmm0 vmovaps 0x20(%r13,%r10), %xmm1 vsubps %xmm5, %xmm1, %xmm1 vmulps %xmm1, %xmm9, %xmm1 vmovaps 0x20(%r13,%r11), %xmm2 vsubps %xmm7, %xmm2, %xmm2 vmulps %xmm2, %xmm10, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vmovaps 0x20(%r13,%rbx), %xmm2 vsubps %xmm8, %xmm2, %xmm2 vmulps %xmm2, %xmm11, %xmm2 vpminsd %xmm12, %xmm2, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpcmpgtd %xmm1, %xmm0, %xmm1 vmovmskps %xmm1, %eax xorb $0xf, %al movzbl %al, %r12d vmovdqa %xmm0, 0x20(%rsp) testb $0x8, %r13b jne 0x2407eb testq %r12, %r12 je 0x2407f2 andq $-0x10, %r13 bsfq %r12, %r14 leaq -0x1(%r12), %rbp xorl %eax, %eax movq (%r13,%r14,8), %r15 prefetcht0 (%r15) prefetcht0 0x40(%r15) andq %r12, %rbp jne 0x2407f9 movq %r15, %r13 testl %eax, %eax je 0x240716 jmp 0x2409d1 movl $0x6, %eax jmp 0x2407de movl $0x4, %eax jmp 0x2407de movl 0x20(%rsp,%r14,4), %r14d movq %r14, -0x10(%rsp) bsfq %rbp, %r14 movq %r14, -0x80(%rsp) leaq -0x1(%rbp), %r14 movq %r14, (%rsp) movq -0x80(%rsp), %r14 movq (%r13,%r14,8), %r14 prefetcht0 (%r14) movq %r14, -0x50(%rsp) prefetcht0 0x40(%r14) movq -0x80(%rsp), %r14 movl 0x20(%rsp,%r14,4), %r14d movq %r14, -0x80(%rsp) movq (%rsp), %r14 andq %rbp, %r14 jne 0x24087e leaq 0x10(%rsi), %r14 movq -0x10(%rsp), %r13 movq -0x80(%rsp), %rbp cmpl %ebp, %r13d jae 0x24086a movq -0x50(%rsp), %r13 movq %r13, (%rsi) movl %ebp, 0x8(%rsi) movq %r14, %rsi jmp 0x2407db movq %r15, (%rsi) movl %r13d, 0x8(%rsi) movq %r14, %rsi movq -0x50(%rsp), %r13 jmp 0x2407de vmovq %r15, %xmm0 vmovd -0x10(%rsp), %xmm1 vpunpcklqdq %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm1[0] vmovq -0x50(%rsp), %xmm1 vmovd -0x80(%rsp), %xmm2 vpunpcklqdq %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm2[0] movq %r14, %r15 bsfq %r14, %r14 movq %r14, -0x80(%rsp) leaq -0x1(%r15), %r14 movq -0x80(%rsp), %rbp movq (%r13,%rbp,8), %rbp prefetcht0 (%rbp) prefetcht0 0x40(%rbp) vmovq %rbp, %xmm2 movq -0x80(%rsp), %rbp vmovd 0x20(%rsp,%rbp,4), %xmm3 vpunpcklqdq %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm3[0] vpcmpgtd %xmm0, %xmm1, %xmm3 andq %r15, %r14 jne 0x24092e vpshufd $0xaa, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vblendvps %xmm3, %xmm0, %xmm1, %xmm4 vblendvps %xmm3, %xmm1, %xmm0, %xmm0 vpcmpgtd %xmm4, %xmm2, %xmm1 vpshufd $0xaa, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vblendvps %xmm1, %xmm4, %xmm2, %xmm3 vblendvps %xmm1, %xmm2, %xmm4, %xmm1 vpcmpgtd %xmm0, %xmm1, %xmm2 vpshufd $0xaa, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vblendvps %xmm2, %xmm0, %xmm1, %xmm4 vblendvps %xmm2, %xmm1, %xmm0, %xmm0 vmovaps %xmm0, (%rsi) vmovaps %xmm4, 0x10(%rsi) vmovq %xmm3, %r13 addq $0x20, %rsi jmp 0x2407de bsfq %r14, %r14 movq (%r13,%r14,8), %r15 prefetcht0 (%r15) prefetcht0 0x40(%r15) vmovq %r15, %xmm4 vmovd 0x20(%rsp,%r14,4), %xmm6 vpunpcklqdq %xmm6, %xmm4, %xmm4 # xmm4 = xmm4[0],xmm6[0] vpshufd $0xaa, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vblendvps %xmm3, %xmm0, %xmm1, %xmm6 vblendvps %xmm3, %xmm1, %xmm0, %xmm0 vpcmpgtd %xmm2, %xmm4, %xmm1 vpshufd $0xaa, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vblendvps %xmm1, %xmm2, %xmm4, %xmm3 vblendvps %xmm1, %xmm4, %xmm2, %xmm1 vpcmpgtd %xmm0, %xmm1, %xmm2 vpshufd $0xaa, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vblendvps %xmm2, %xmm0, %xmm1, %xmm4 vblendvps %xmm2, %xmm1, %xmm0, %xmm0 vpcmpgtd %xmm6, %xmm3, %xmm1 vpshufd $0xaa, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vblendvps %xmm1, %xmm6, %xmm3, %xmm2 vblendvps %xmm1, %xmm3, %xmm6, %xmm1 vpcmpgtd %xmm1, %xmm4, %xmm3 vpshufd $0xaa, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vblendvps %xmm3, %xmm1, %xmm4, %xmm6 vblendvps %xmm3, %xmm4, %xmm1, %xmm1 vmovaps %xmm0, (%rsi) vmovaps %xmm1, 0x10(%rsi) vmovaps %xmm6, 0x20(%rsi) vmovq %xmm2, %r13 addq $0x30, %rsi jmp 0x2407de cmpl $0x6, %eax jne 0x2406ed movl %r13d, %eax andl $0xf, %eax addq $-0x8, %rax je 0x240e90 andq $-0x10, %r13 vbroadcastss (%r9,%rcx,4), %xmm0 vmovaps %xmm0, -0x50(%rsp) vbroadcastss 0x10(%r9,%rcx,4), %xmm0 vmovaps %xmm0, (%rsp) vbroadcastss 0x20(%r9,%rcx,4), %xmm0 vmovaps %xmm0, -0x10(%rsp) vbroadcastss 0x40(%r9,%rcx,4), %xmm0 vmovaps %xmm0, 0x170(%rsp) vbroadcastss 0x50(%r9,%rcx,4), %xmm0 vmovaps %xmm0, 0x160(%rsp) vbroadcastss 0x60(%r9,%rcx,4), %xmm0 vmovaps %xmm0, 0x150(%rsp) xorl %ebp, %ebp imulq $0xb0, %rbp, %r15 vmovaps 0x80(%r13,%r15), %xmm1 vmovaps 0x40(%r13,%r15), %xmm14 vmulps %xmm1, %xmm14, %xmm3 vmovaps 0x70(%r13,%r15), %xmm2 vmovaps 0x50(%r13,%r15), %xmm6 vmulps %xmm6, %xmm2, %xmm4 vsubps %xmm3, %xmm4, %xmm12 vmovaps 0x60(%r13,%r15), %xmm3 vmulps %xmm3, %xmm6, %xmm8 vmovaps (%r13,%r15), %xmm4 vsubps -0x50(%rsp), %xmm4, %xmm11 vmovaps 0x10(%r13,%r15), %xmm4 vsubps (%rsp), %xmm4, %xmm0 vmovaps 0x20(%r13,%r15), %xmm4 vsubps -0x10(%rsp), %xmm4, %xmm10 vmovaps 0x30(%r13,%r15), %xmm4 vmulps %xmm4, %xmm1, %xmm9 vsubps %xmm8, %xmm9, %xmm5 vmovaps %xmm5, -0x80(%rsp) vmulps %xmm4, %xmm2, %xmm9 vmulps %xmm3, %xmm14, %xmm5 vsubps %xmm9, %xmm5, %xmm13 vmovaps 0x160(%rsp), %xmm7 vmulps %xmm7, %xmm10, %xmm5 vmovaps 0x150(%rsp), %xmm9 vmulps %xmm0, %xmm9, %xmm15 vsubps %xmm5, %xmm15, %xmm5 vmulps %xmm9, %xmm11, %xmm15 vmovaps 0x170(%rsp), %xmm8 vmovaps %xmm10, -0x40(%rsp) vmulps %xmm10, %xmm8, %xmm10 vsubps %xmm15, %xmm10, %xmm10 vmovaps %xmm0, -0x30(%rsp) vmulps %xmm0, %xmm8, %xmm15 vmovaps %xmm11, -0x20(%rsp) vmulps %xmm7, %xmm11, %xmm11 vsubps %xmm15, %xmm11, %xmm11 vmovaps %xmm13, -0x60(%rsp) vmulps %xmm9, %xmm13, %xmm15 vmovaps -0x80(%rsp), %xmm9 vmulps %xmm7, %xmm9, %xmm7 vaddps %xmm7, %xmm15, %xmm7 vmulps %xmm8, %xmm12, %xmm15 vaddps %xmm7, %xmm15, %xmm7 vmulps %xmm1, %xmm11, %xmm1 vmulps %xmm2, %xmm10, %xmm2 vaddps %xmm2, %xmm1, %xmm1 vmulps %xmm5, %xmm3, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vmulps %xmm6, %xmm11, %xmm0 vmulps %xmm10, %xmm14, %xmm2 vaddps %xmm2, %xmm0, %xmm3 vbroadcastss 0x1ce0354(%rip), %xmm0 # 0x1f20ec0 vandps %xmm0, %xmm7, %xmm2 vxorps %xmm1, %xmm2, %xmm0 vmulps %xmm5, %xmm4, %xmm1 vaddps %xmm3, %xmm1, %xmm1 vxorps %xmm1, %xmm2, %xmm1 vxorps %xmm5, %xmm5, %xmm5 vcmpnltps %xmm5, %xmm0, %xmm3 vcmpnltps %xmm5, %xmm1, %xmm4 vandps %xmm4, %xmm3, %xmm3 vbroadcastss 0x1ce0329(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm7, %xmm14 vcmpneqps %xmm5, %xmm7, %xmm4 vandps %xmm4, %xmm3, %xmm3 vaddps %xmm1, %xmm0, %xmm4 vcmpleps %xmm14, %xmm4, %xmm4 vtestps %xmm4, %xmm3 jne 0x240bca incq %rbp cmpq %rax, %rbp jne 0x240a45 jmp 0x240e90 vandps %xmm4, %xmm3, %xmm3 vmovaps -0x40(%rsp), %xmm4 vmulps -0x60(%rsp), %xmm4, %xmm4 vmulps -0x30(%rsp), %xmm9, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vmulps -0x20(%rsp), %xmm12, %xmm5 vaddps %xmm4, %xmm5, %xmm4 vxorps %xmm4, %xmm2, %xmm2 vbroadcastss 0x30(%r9,%rcx,4), %xmm4 vmulps %xmm4, %xmm14, %xmm4 vcmpltps %xmm2, %xmm4, %xmm4 vbroadcastss 0x80(%r9,%rcx,4), %xmm5 vmulps %xmm5, %xmm14, %xmm5 vcmpleps %xmm5, %xmm2, %xmm5 vandps %xmm5, %xmm4, %xmm4 vtestps %xmm3, %xmm4 je 0x240bb9 movq %rsi, -0x40(%rsp) vandps %xmm3, %xmm4, %xmm6 vmovaps %xmm0, 0x20(%rsp) vmovaps %xmm1, 0x30(%rsp) vmovaps %xmm2, 0x40(%rsp) vmovaps %xmm14, 0x50(%rsp) vmovaps %xmm6, 0x70(%rsp) vmovaps %xmm12, 0xb0(%rsp) vmovaps %xmm9, 0xc0(%rsp) vmovaps -0x60(%rsp), %xmm0 vmovaps %xmm0, 0xd0(%rsp) movq 0x1108(%rsp), %r14 movq (%r14), %rsi vmovaps %xmm6, 0x10(%rsp) vrcpps %xmm14, %xmm0 vmulps %xmm0, %xmm14, %xmm1 vbroadcastss 0x1caba89(%rip), %xmm2 # 0x1eec714 vsubps %xmm1, %xmm2, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps 0x40(%rsp), %xmm0, %xmm7 vmovaps %xmm7, 0xa0(%rsp) vmulps 0x20(%rsp), %xmm0, %xmm1 vmovaps %xmm1, 0x80(%rsp) vmulps 0x30(%rsp), %xmm0, %xmm0 vbroadcastss 0x1caad5c(%rip), %xmm4 # 0x1eeba20 vblendvps %xmm6, %xmm7, %xmm4, %xmm1 vmovaps %xmm0, 0x90(%rsp) vshufps $0xb1, %xmm1, %xmm1, %xmm0 # xmm0 = xmm1[1,0,3,2] vminps %xmm1, %xmm0, %xmm0 vshufpd $0x1, %xmm0, %xmm0, %xmm2 # xmm2 = xmm0[1,0] vminps %xmm0, %xmm2, %xmm0 vcmpeqps %xmm0, %xmm1, %xmm1 vtestps %xmm6, %xmm1 vpcmpeqd %xmm0, %xmm0, %xmm0 je 0x240cf9 vmovaps %xmm1, %xmm0 addq %r13, %r15 vandps %xmm0, %xmm6, %xmm0 vmovmskps %xmm0, %r14d bsfq %r14, %r14 movq 0x1e8(%rsi), %rsi movq %rsi, -0x60(%rsp) movq %r9, -0x20(%rsp) movq %rcx, -0x30(%rsp) movl 0x90(%r9,%rcx,4), %esi movq %r14, -0x80(%rsp) movl 0x90(%r15,%r14,4), %ecx movq -0x60(%rsp), %r9 movq (%r9,%rcx,8), %r14 movl %esi, -0x64(%rsp) testl %esi, 0x34(%r14) jne 0x240dbd movq -0x80(%rsp), %r14 andl $0x0, 0x10(%rsp,%r14,4) vmovaps 0x10(%rsp), %xmm0 vtestps %xmm0, %xmm0 movq -0x20(%rsp), %r9 movq -0x30(%rsp), %rcx movq -0x40(%rsp), %rsi je 0x240bb9 vblendvps %xmm0, %xmm7, %xmm4, %xmm1 vshufps $0xb1, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,0,3,2] vminps %xmm1, %xmm2, %xmm2 vshufpd $0x1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0] vminps %xmm2, %xmm3, %xmm2 vcmpeqps %xmm2, %xmm1, %xmm1 vtestps %xmm0, %xmm1 je 0x240d99 vandps %xmm0, %xmm1, %xmm0 vmovmskps %xmm0, %r14d bsfq %r14, %r14 movq %r14, -0x80(%rsp) movl 0x90(%r15,%r14,4), %ecx movq -0x60(%rsp), %r14 movq (%r14,%rcx,8), %r14 movl -0x64(%rsp), %esi jmp 0x240d40 movq -0x80(%rsp), %r14 vmovss 0x80(%rsp,%r14,4), %xmm0 vmovss 0x90(%rsp,%r14,4), %xmm1 vmovss 0xa0(%rsp,%r14,4), %xmm2 movq -0x20(%rsp), %r9 movq %rcx, %rsi movq -0x30(%rsp), %rcx vmovss %xmm2, 0x80(%r9,%rcx,4) vmovss 0xb0(%rsp,%r14,4), %xmm2 vmovss %xmm2, 0xc0(%r9,%rcx,4) vmovss 0xc0(%rsp,%r14,4), %xmm2 vmovss %xmm2, 0xd0(%r9,%rcx,4) vmovss 0xd0(%rsp,%r14,4), %xmm2 vmovss %xmm2, 0xe0(%r9,%rcx,4) vmovss %xmm0, 0xf0(%r9,%rcx,4) vmovss %xmm1, 0x100(%r9,%rcx,4) movl 0xa0(%r15,%r14,4), %r14d movl %r14d, 0x110(%r9,%rcx,4) movl %esi, 0x120(%r9,%rcx,4) movq 0x1108(%rsp), %r15 movq 0x8(%r15), %r14 movl (%r14), %r14d movl %r14d, 0x130(%r9,%rcx,4) movq 0x8(%r15), %r14 movl 0x4(%r14), %r14d movl %r14d, 0x140(%r9,%rcx,4) movq -0x40(%rsp), %rsi jmp 0x240bb9 vbroadcastss 0x80(%r9,%rcx,4), %xmm12 vmovaps 0x140(%rsp), %xmm5 vmovaps 0x130(%rsp), %xmm7 vmovaps 0x120(%rsp), %xmm8 vmovaps 0x110(%rsp), %xmm9 vmovaps 0x100(%rsp), %xmm10 vmovaps 0xf0(%rsp), %xmm11 vmovaps 0xe0(%rsp), %xmm13 jmp 0x2406ed addq $0x10c8, %rsp # imm = 0x10C8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::TriangleMIntersectorKMoeller<4, 4, false>>, true>::occluded1(embree::Accel::Intersectors, embree::BVHN<4> const, embree::NodeRefPtr<4>, unsigned long, embree::avx::MoellerTrumboreIntersectorK<4, 4>&, embree::RayK<4>&, embree::avx::TravRayK<4, false> const&, embree::RayQueryContext*)	bool BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded1(Accel::Intersectors* This, const BVH* bvh, NodeRef root, size_t k, Precalculations& pre, RayK<K>& ray, const TravRayK<K, robust>& tray, RayQueryContext* context) { /* stack state / NodeRef stack[stackSizeSingle]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSizeSingle; stack[0] = root; /* load the ray into SIMD registers / TravRay<N,robust> tray1; tray1.template init<K>(k, tray.org, tray.dir, tray.rdir, tray.nearXYZ, tray.tnear[k], tray.tfar[k]); / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes, 1, 1, 1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray1, ray.time()[k], tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / BVHNNodeTraverser1Hit<N, types>::traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves, 1, 1, 1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersectorK::occluded(This, pre, ray, k, context, prim, num, tray1, lazy_node)) { ray.tfar[k] = neg_inf; return true; } if (unlikely(lazy_node)) { stackPtr = lazy_node; stackPtr++; } } return false; }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x898, %rsp # imm = 0x898 movq %r9, -0x78(%rsp) movq 0x8d0(%rsp), %rax leaq 0xf8(%rsp), %r14 movq %rdx, -0x8(%r14) vbroadcastss (%rax,%rcx,4), %xmm0 vbroadcastss 0x10(%rax,%rcx,4), %xmm1 vbroadcastss 0x20(%rax,%rcx,4), %xmm2 vbroadcastss 0x60(%rax,%rcx,4), %xmm3 vbroadcastss 0x70(%rax,%rcx,4), %xmm4 vbroadcastss 0x80(%rax,%rcx,4), %xmm5 movslq 0x90(%rax,%rcx,4), %r9 movslq 0xa0(%rax,%rcx,4), %rdi movslq 0xb0(%rax,%rcx,4), %r8 movq %r9, %r10 xorq $0x10, %r10 movq %rdi, %r11 xorq $0x10, %r11 movq %r8, %rbx xorq $0x10, %rbx vbroadcastss 0xc0(%rax,%rcx,4), %xmm14 vbroadcastss 0xd0(%rax,%rcx,4), %xmm7 movq %r14, %rdx leaq 0xf0(%rsp), %rax cmpq %rax, %r14 je 0x2420aa leaq -0x8(%rdx), %r14 movq %rdx, -0x70(%rsp) movq -0x8(%rdx), %r15 testb $0x8, %r15b jne 0x241c51 vmovaps 0x20(%r15,%r9), %xmm8 vsubps %xmm0, %xmm8, %xmm8 vmulps %xmm3, %xmm8, %xmm8 vmovaps 0x20(%r15,%rdi), %xmm9 vsubps %xmm1, %xmm9, %xmm9 vmulps %xmm4, %xmm9, %xmm9 vpmaxsd %xmm9, %xmm8, %xmm8 vmovaps 0x20(%r15,%r8), %xmm9 vsubps %xmm2, %xmm9, %xmm9 vmulps %xmm5, %xmm9, %xmm9 vpmaxsd %xmm14, %xmm9, %xmm9 vpmaxsd %xmm9, %xmm8, %xmm8 vmovaps 0x20(%r15,%r10), %xmm9 vsubps %xmm0, %xmm9, %xmm9 vmulps %xmm3, %xmm9, %xmm9 vmovaps 0x20(%r15,%r11), %xmm10 vsubps %xmm1, %xmm10, %xmm10 vmulps %xmm4, %xmm10, %xmm10 vpminsd %xmm10, %xmm9, %xmm9 vmovaps 0x20(%r15,%rbx), %xmm10 vsubps %xmm2, %xmm10, %xmm10 vmulps %xmm5, %xmm10, %xmm10 vpminsd %xmm7, %xmm10, %xmm10 vpminsd %xmm10, %xmm9, %xmm9 vpcmpgtd %xmm9, %xmm8, %xmm8 vmovmskps %xmm8, %eax xorb $0xf, %al movzbl %al, %ebp testb $0x8, %r15b jne 0x241c87 testq %rbp, %rbp je 0x241c8e andq $-0x10, %r15 bsfq %rbp, %rsi leaq -0x1(%rbp), %r12 xorl %eax, %eax movq (%r15,%rsi,8), %rsi prefetcht0 (%rsi) prefetcht0 0x40(%rsi) andq %rbp, %r12 jne 0x241c95 movq %rsi, %r15 testl %eax, %eax je 0x241bc0 jmp 0x241cd3 movl $0x6, %eax jmp 0x241c7d movl $0x4, %eax jmp 0x241c7d movq %rsi, (%r14) addq $0x8, %r14 movq %r14, %rdx bsfq %r12, %r14 leaq -0x1(%r12), %rsi movq (%r15,%r14,8), %r14 prefetcht0 (%r14) prefetcht0 0x40(%r14) andq %r12, %rsi je 0x241ccb movq %r14, (%rdx) addq $0x8, %rdx bsfq %rsi, %r14 leaq -0x1(%rsi), %r12 jmp 0x241ca8 movq %r14, %r15 movq %rdx, %r14 jmp 0x241c7d cmpl $0x6, %eax jne 0x24209d movl %r15d, %esi andl $0xf, %esi addq $-0x8, %rsi setne %dl je 0x242081 andq $-0x10, %r15 movq -0x78(%rsp), %rax vbroadcastss (%rax,%rcx,4), %xmm6 vmovaps %xmm6, 0xc0(%rsp) vbroadcastss 0x10(%rax,%rcx,4), %xmm9 vbroadcastss 0x20(%rax,%rcx,4), %xmm12 vbroadcastss 0x40(%rax,%rcx,4), %xmm15 vbroadcastss 0x50(%rax,%rcx,4), %xmm11 vbroadcastss 0x60(%rax,%rcx,4), %xmm6 xorl %r12d, %r12d movq %r14, -0x68(%rsp) movq %rsi, -0x20(%rsp) vmovaps %xmm9, -0x40(%rsp) vmovaps %xmm12, -0x50(%rsp) vmovaps %xmm15, 0x60(%rsp) vmovaps %xmm11, 0x50(%rsp) vmovaps %xmm6, 0xb0(%rsp) imulq $0xb0, %r12, %r14 vmovaps 0x80(%r15,%r14), %xmm6 vmovaps %xmm6, 0x10(%rsp) vmovaps 0x40(%r15,%r14), %xmm13 vmovaps %xmm13, 0x30(%rsp) vmulps %xmm6, %xmm13, %xmm10 vmovaps 0x70(%r15,%r14), %xmm8 vmovaps %xmm8, 0x20(%rsp) vmovaps 0x50(%r15,%r14), %xmm6 vmovaps %xmm6, 0xe0(%rsp) vmovaps %xmm12, %xmm13 vmulps %xmm6, %xmm8, %xmm12 vsubps %xmm10, %xmm12, %xmm8 vmovaps %xmm8, -0x10(%rsp) vmovaps 0x60(%r15,%r14), %xmm8 vmovaps %xmm8, (%rsp) vmulps %xmm6, %xmm8, %xmm6 vmovaps %xmm6, -0x60(%rsp) vmovaps (%r15,%r14), %xmm10 vsubps 0xc0(%rsp), %xmm10, %xmm12 vmovaps 0x10(%r15,%r14), %xmm10 vmovaps %xmm15, %xmm6 vsubps %xmm9, %xmm10, %xmm9 vmovaps 0x20(%r15,%r14), %xmm10 vsubps %xmm13, %xmm10, %xmm8 vmovaps 0x30(%r15,%r14), %xmm10 vmulps 0x10(%rsp), %xmm10, %xmm13 vmovaps %xmm10, 0xd0(%rsp) vsubps -0x60(%rsp), %xmm13, %xmm13 vmovaps %xmm13, -0x60(%rsp) vmulps 0x20(%rsp), %xmm10, %xmm15 vmovaps (%rsp), %xmm10 vmulps 0x30(%rsp), %xmm10, %xmm13 vsubps %xmm15, %xmm13, %xmm10 vmulps %xmm8, %xmm11, %xmm13 vmovaps %xmm13, -0x30(%rsp) vmovaps 0xb0(%rsp), %xmm15 vmulps %xmm9, %xmm15, %xmm13 vsubps -0x30(%rsp), %xmm13, %xmm13 vmovaps %xmm13, -0x30(%rsp) vmulps %xmm15, %xmm12, %xmm13 vmovaps %xmm8, 0x80(%rsp) vmulps %xmm6, %xmm8, %xmm8 vsubps %xmm13, %xmm8, %xmm8 vmovaps %xmm9, 0x90(%rsp) vmulps %xmm6, %xmm9, %xmm13 vmovaps %xmm12, 0xa0(%rsp) vmulps %xmm12, %xmm11, %xmm12 vsubps %xmm13, %xmm12, %xmm12 vmovaps %xmm10, 0x70(%rsp) vmulps %xmm15, %xmm10, %xmm13 vmulps -0x60(%rsp), %xmm11, %xmm9 vaddps %xmm13, %xmm9, %xmm9 vmulps -0x10(%rsp), %xmm6, %xmm13 vaddps %xmm9, %xmm13, %xmm13 vmulps 0x10(%rsp), %xmm12, %xmm9 vmulps 0x20(%rsp), %xmm8, %xmm11 vaddps %xmm11, %xmm9, %xmm9 vmovaps -0x30(%rsp), %xmm6 vmulps (%rsp), %xmm6, %xmm11 vaddps %xmm9, %xmm11, %xmm9 vmulps 0xe0(%rsp), %xmm12, %xmm11 vmulps 0x30(%rsp), %xmm8, %xmm8 vaddps %xmm8, %xmm11, %xmm11 vbroadcastss 0x1cdefdb(%rip), %xmm8 # 0x1f20ec0 vandps %xmm8, %xmm13, %xmm8 vxorps %xmm9, %xmm8, %xmm12 vmulps 0xd0(%rsp), %xmm6, %xmm9 vaddps %xmm11, %xmm9, %xmm9 vxorps %xmm9, %xmm8, %xmm10 vxorps %xmm6, %xmm6, %xmm6 vcmpnltps %xmm6, %xmm12, %xmm9 vcmpnltps %xmm6, %xmm10, %xmm11 vandps %xmm11, %xmm9, %xmm9 vcmpneqps %xmm6, %xmm13, %xmm11 vandps %xmm11, %xmm9, %xmm9 vaddps %xmm10, %xmm12, %xmm11 vbroadcastss 0x1cdef97(%rip), %xmm10 # 0x1f20ec4 vandps %xmm10, %xmm13, %xmm10 vcmpleps %xmm10, %xmm11, %xmm11 vtestps %xmm11, %xmm9 movb %dl, -0x79(%rsp) movq %r12, -0x18(%rsp) jne 0x241fe7 xorl %esi, %esi vmovdqa -0x40(%rsp), %xmm9 vmovaps -0x50(%rsp), %xmm12 testb %sil, %sil je 0x241fb8 addq %r15, %r14 movq 0x8d8(%rsp), %rax movq (%rax), %r12 vmovaps 0x40(%rsp), %xmm8 vmovmskps %xmm8, %esi vmovaps -0x50(%rsp), %xmm12 bsfq %rsi, %r13 movl 0x90(%r14,%r13,4), %eax movq 0x1e8(%r12), %rdx movq (%rdx,%rax,8), %rdx movq -0x78(%rsp), %rax movl 0x90(%rax,%rcx,4), %eax andl 0x34(%rdx), %eax jne 0x241fab btcq %r13, %rsi testl %eax, %eax jne 0x242078 testq %rsi, %rsi jne 0x241f7e movq -0x18(%rsp), %r12 incq %r12 movq -0x20(%rsp), %rsi cmpq %rsi, %r12 setb %dl movq -0x68(%rsp), %r14 vmovaps 0x60(%rsp), %xmm15 vmovaps 0x50(%rsp), %xmm11 jne 0x241d58 jmp 0x242081 vandps %xmm11, %xmm9, %xmm9 vmovaps 0x70(%rsp), %xmm11 vmulps 0x80(%rsp), %xmm11, %xmm11 vmovaps -0x60(%rsp), %xmm12 vmulps 0x90(%rsp), %xmm12, %xmm12 vaddps %xmm11, %xmm12, %xmm11 vmovaps -0x10(%rsp), %xmm6 vmulps 0xa0(%rsp), %xmm6, %xmm12 vaddps %xmm11, %xmm12, %xmm11 vxorps %xmm11, %xmm8, %xmm8 movq -0x78(%rsp), %rax vbroadcastss 0x30(%rax,%rcx,4), %xmm11 vmulps %xmm11, %xmm10, %xmm11 vcmpltps %xmm8, %xmm11, %xmm11 vbroadcastss 0x80(%rax,%rcx,4), %xmm12 vmulps %xmm12, %xmm10, %xmm10 vcmpleps %xmm10, %xmm8, %xmm8 vandps %xmm8, %xmm11, %xmm8 vtestps %xmm9, %xmm8 setne %sil je 0x241f4e vandps %xmm9, %xmm8, %xmm8 vmovaps %xmm8, 0x40(%rsp) jmp 0x241f4e movq -0x68(%rsp), %r14 movb -0x79(%rsp), %dl xorl %eax, %eax testb $0x1, %dl je 0x24209d movq -0x78(%rsp), %rax movl $0xff800000, 0x80(%rax,%rcx,4) # imm = 0xFF800000 movl $0x1, %eax movq -0x70(%rsp), %rdx testb $0x3, %al je 0x241b9f leaq 0xf0(%rsp), %rax cmpq %rax, %rdx setne %al addq $0x898, %rsp # imm = 0x898 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::TriangleMiIntersectorKMoeller<4, 4, true>>, true>::intersect1(embree::Accel::Intersectors, embree::BVHN<4> const, embree::NodeRefPtr<4>, unsigned long, embree::avx::MoellerTrumboreIntersectorK<4, 4>&, embree::RayHitK<4>&, embree::avx::TravRayK<4, false> const&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect1(Accel::Intersectors* This, const BVH* bvh, NodeRef root, size_t k, Precalculations& pre, RayHitK<K>& ray, const TravRayK<K, robust>& tray, RayQueryContext* context) { /* stack state / StackItemT<NodeRef> stack[stackSizeSingle]; // stack of nodes StackItemT<NodeRef> stackPtr = stack + 1; // current stack pointer StackItemT<NodeRef>* stackEnd = stack + stackSizeSingle; stack[0].ptr = root; stack[0].dist = neg_inf; /* load the ray into SIMD registers / TravRay<N,robust> tray1; tray1.template init<K>(k, tray.org, tray.dir, tray.rdir, tray.nearXYZ, tray.tnear[k], tray.tfar[k]); / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = NodeRef(stackPtr->ptr); / if popped node is too far, pop next one / if (unlikely((float)&stackPtr->dist > ray.tfar[k])) continue; / downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(normal.trav_nodes, 1, 1, 1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray1, ray.time()[k], tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(normal.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / BVHNNodeTraverser1Hit<N, types>::traverseClosestHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(normal.trav_leaves, 1, 1, 1); size_t num; Primitive prim = (Primitive*)cur.leaf(num); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(This, pre, ray, k, context, prim, num, tray1, lazy_node); tray1.tfar = ray.tfar[k]; if (unlikely(lazy_node)) { stackPtr->ptr = lazy_node; stackPtr->dist = neg_inf; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x1208, %rsp # imm = 0x1208 movq %r9, %rbx movq %rcx, %r14 movq 0x1248(%rsp), %r10 movq 0x1240(%rsp), %rax leaq 0x2d0(%rsp), %r11 movq %rdx, -0x10(%r11) andl $0x0, -0x8(%r11) vbroadcastss (%rax,%rcx,4), %xmm7 vbroadcastss 0x10(%rax,%rcx,4), %xmm8 vbroadcastss 0x20(%rax,%rcx,4), %xmm9 vbroadcastss 0x60(%rax,%rcx,4), %xmm10 vbroadcastss 0x70(%rax,%rcx,4), %xmm11 vbroadcastss 0x80(%rax,%rcx,4), %xmm12 movslq 0x90(%rax,%rcx,4), %r9 movslq 0xa0(%rax,%rcx,4), %r12 movslq 0xb0(%rax,%rcx,4), %rsi movq %r9, %rdi xorq $0x10, %rdi movq %r12, %r15 xorq $0x10, %r15 movq %rsi, %rcx xorq $0x10, %rcx movq %rcx, 0xa8(%rsp) vbroadcastss 0xd0(%rax,%r14,4), %xmm0 vbroadcastss 0xc0(%rax,%r14,4), %xmm13 pushq $0x1 popq %rax movl %r14d, %ecx shll %cl, %eax cltq shlq $0x4, %rax addq 0x1ee143a(%rip), %rax # 0x2124730 movq %rax, 0x98(%rsp) movq %rsi, 0x10(%rsp) movq %rdi, 0x8(%rsp) movq %r15, (%rsp) vmovaps %xmm7, 0x210(%rsp) vmovaps %xmm8, 0x200(%rsp) vmovaps %xmm9, 0x1f0(%rsp) vmovaps %xmm10, 0x1e0(%rsp) vmovaps %xmm11, 0x1d0(%rsp) vmovaps %xmm12, 0x1c0(%rsp) movq %r12, 0xa0(%rsp) vmovaps %xmm13, 0x1b0(%rsp) leaq 0x2c0(%rsp), %rax cmpq %rax, %r11 je 0x243e72 vmovss -0x8(%r11), %xmm1 addq $-0x10, %r11 vucomiss 0x80(%rbx,%r14,4), %xmm1 ja 0x243353 movq (%r11), %rbp testb $0x8, %bpl jne 0x24341f vmovaps 0x20(%rbp,%r9), %xmm1 vsubps %xmm7, %xmm1, %xmm1 vmulps %xmm1, %xmm10, %xmm1 vmovaps 0x20(%rbp,%r12), %xmm2 vsubps %xmm8, %xmm2, %xmm2 vmulps %xmm2, %xmm11, %xmm2 vpmaxsd %xmm2, %xmm1, %xmm1 vmovaps 0x20(%rbp,%rsi), %xmm2 vsubps %xmm9, %xmm2, %xmm2 vmulps %xmm2, %xmm12, %xmm2 vpmaxsd %xmm13, %xmm2, %xmm2 vpmaxsd %xmm2, %xmm1, %xmm1 vmovaps 0x20(%rbp,%rdi), %xmm2 vsubps %xmm7, %xmm2, %xmm2 vmulps %xmm2, %xmm10, %xmm2 vmovaps 0x20(%rbp,%r15), %xmm3 vsubps %xmm8, %xmm3, %xmm3 vmulps %xmm3, %xmm11, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 movq 0xa8(%rsp), %rax vmovaps 0x20(%rbp,%rax), %xmm3 vsubps %xmm9, %xmm3, %xmm3 vmulps %xmm3, %xmm12, %xmm3 vpminsd %xmm0, %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vpcmpgtd %xmm2, %xmm1, %xmm2 vmovmskps %xmm2, %eax xorb $0xf, %al movzbl %al, %r13d vmovdqa %xmm1, 0xf0(%rsp) testb $0x8, %bpl jne 0x243459 testq %r13, %r13 je 0x24345d andq $-0x10, %rbp bsfq %r13, %rdx leaq -0x1(%r13), %r8 xorl %eax, %eax movq (%rbp,%rdx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) andq %r13, %r8 jne 0x243462 movq %rcx, %rbp testl %eax, %eax je 0x24337d jmp 0x243609 pushq $0x6 jmp 0x24345f pushq $0x4 popq %rax jmp 0x24344c movq %r9, %r15 movl 0xf0(%rsp,%rdx,4), %esi bsfq %r8, %r9 leaq -0x1(%r8), %rdx movq (%rbp,%r9,8), %rdi prefetcht0 (%rdi) prefetcht0 0x40(%rdi) movl 0xf0(%rsp,%r9,4), %r9d andq %r8, %rdx jne 0x2434c5 leaq 0x10(%r11), %rdx cmpl %r9d, %esi jae 0x2434a5 movq %rdi, (%r11) movl %r9d, 0x8(%r11) movq %rdx, %r11 movq %rcx, %rbp jmp 0x2434b2 movq %rcx, (%r11) movl %esi, 0x8(%r11) movq %rdx, %r11 movq %rdi, %rbp movq %r15, %r9 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %rdi movq (%rsp), %r15 jmp 0x24344c vmovq %rcx, %xmm1 vmovd %esi, %xmm2 vpunpcklqdq %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm2[0] vmovq %rdi, %xmm2 vmovd %r9d, %xmm3 vpunpcklqdq %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm3[0] bsfq %rdx, %rsi leaq -0x1(%rdx), %rcx movq (%rbp,%rsi,8), %rdi prefetcht0 (%rdi) prefetcht0 0x40(%rdi) vmovq %rdi, %xmm3 vmovd 0xf0(%rsp,%rsi,4), %xmm4 vpunpcklqdq %xmm4, %xmm3, %xmm3 # xmm3 = xmm3[0],xmm4[0] vpcmpgtd %xmm1, %xmm2, %xmm4 andq %rdx, %rcx jne 0x243563 vpshufd $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] vblendvps %xmm4, %xmm1, %xmm2, %xmm5 vblendvps %xmm4, %xmm2, %xmm1, %xmm1 vpcmpgtd %xmm5, %xmm3, %xmm2 vpshufd $0xaa, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vblendvps %xmm2, %xmm5, %xmm3, %xmm4 vblendvps %xmm2, %xmm3, %xmm5, %xmm2 vpcmpgtd %xmm1, %xmm2, %xmm3 vpshufd $0xaa, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vblendvps %xmm3, %xmm1, %xmm2, %xmm5 vblendvps %xmm3, %xmm2, %xmm1, %xmm1 vmovaps %xmm1, (%r11) vmovaps %xmm5, 0x10(%r11) vmovq %xmm4, %rbp addq $0x20, %r11 jmp 0x2434b2 bsfq %rcx, %rcx movq (%rbp,%rcx,8), %rdx prefetcht0 (%rdx) prefetcht0 0x40(%rdx) vmovq %rdx, %xmm5 vmovd 0xf0(%rsp,%rcx,4), %xmm6 vpunpcklqdq %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[0],xmm6[0] vpshufd $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] vblendvps %xmm4, %xmm1, %xmm2, %xmm6 vblendvps %xmm4, %xmm2, %xmm1, %xmm1 vpcmpgtd %xmm3, %xmm5, %xmm2 vpshufd $0xaa, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vblendvps %xmm2, %xmm3, %xmm5, %xmm4 vblendvps %xmm2, %xmm5, %xmm3, %xmm2 vpcmpgtd %xmm1, %xmm2, %xmm3 vpshufd $0xaa, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] vblendvps %xmm3, %xmm1, %xmm2, %xmm5 vblendvps %xmm3, %xmm2, %xmm1, %xmm1 vpcmpgtd %xmm6, %xmm4, %xmm2 vpshufd $0xaa, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vblendvps %xmm2, %xmm6, %xmm4, %xmm3 vblendvps %xmm2, %xmm4, %xmm6, %xmm2 vpcmpgtd %xmm2, %xmm5, %xmm4 vpshufd $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] vblendvps %xmm4, %xmm2, %xmm5, %xmm6 vblendvps %xmm4, %xmm5, %xmm2, %xmm2 vmovaps %xmm1, (%r11) vmovaps %xmm2, 0x10(%r11) vmovaps %xmm6, 0x20(%r11) vmovq %xmm3, %rbp addq $0x30, %r11 jmp 0x2434b2 cmpl $0x6, %eax jne 0x243353 movl %ebp, %r8d andl $0xf, %r8d addq $-0x8, %r8 je 0x243e0e andq $-0x10, %rbp xorl %r12d, %r12d imulq $0x50, %r12, %rax movq (%r10), %r15 movl 0x30(%rbp,%rax), %edx movq 0x228(%r15), %rcx movq (%rcx,%rdx,8), %rdx movl (%rbp,%rax), %esi movl 0x4(%rbp,%rax), %edi vmovups (%rdx,%rsi,4), %xmm2 movl 0x10(%rbp,%rax), %esi vmovups (%rdx,%rsi,4), %xmm1 movl 0x20(%rbp,%rax), %esi vmovups (%rdx,%rsi,4), %xmm0 movl 0x34(%rbp,%rax), %edx movq (%rcx,%rdx,8), %rdx vmovups (%rdx,%rdi,4), %xmm4 movl 0x14(%rbp,%rax), %esi vmovups (%rdx,%rsi,4), %xmm6 movl 0x24(%rbp,%rax), %esi vmovups (%rdx,%rsi,4), %xmm3 movl 0x38(%rbp,%rax), %edx movq (%rcx,%rdx,8), %rdx movl 0x8(%rbp,%rax), %esi vmovups (%rdx,%rsi,4), %xmm5 movl 0x18(%rbp,%rax), %esi vmovups (%rdx,%rsi,4), %xmm8 movl 0x28(%rbp,%rax), %esi vmovups (%rdx,%rsi,4), %xmm9 movl 0x3c(%rbp,%rax), %edx movq (%rcx,%rdx,8), %rcx movl 0xc(%rbp,%rax), %edx vmovups (%rcx,%rdx,4), %xmm7 movl 0x1c(%rbp,%rax), %edx vmovups (%rcx,%rdx,4), %xmm10 movl 0x2c(%rbp,%rax), %edx vmovups (%rcx,%rdx,4), %xmm11 vunpcklps %xmm5, %xmm2, %xmm12 # xmm12 = xmm2[0],xmm5[0],xmm2[1],xmm5[1] vunpckhps %xmm5, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm5[2],xmm2[3],xmm5[3] vunpcklps %xmm7, %xmm4, %xmm5 # xmm5 = xmm4[0],xmm7[0],xmm4[1],xmm7[1] vunpckhps %xmm7, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm7[2],xmm4[3],xmm7[3] vunpcklps %xmm4, %xmm2, %xmm13 # xmm13 = xmm2[0],xmm4[0],xmm2[1],xmm4[1] vunpcklps %xmm5, %xmm12, %xmm4 # xmm4 = xmm12[0],xmm5[0],xmm12[1],xmm5[1] vunpckhps %xmm5, %xmm12, %xmm7 # xmm7 = xmm12[2],xmm5[2],xmm12[3],xmm5[3] vunpcklps %xmm8, %xmm1, %xmm2 # xmm2 = xmm1[0],xmm8[0],xmm1[1],xmm8[1] vunpckhps %xmm8, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm8[2],xmm1[3],xmm8[3] vunpcklps %xmm10, %xmm6, %xmm8 # xmm8 = xmm6[0],xmm10[0],xmm6[1],xmm10[1] vunpckhps %xmm10, %xmm6, %xmm6 # xmm6 = xmm6[2],xmm10[2],xmm6[3],xmm10[3] vunpcklps %xmm6, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm6[0],xmm1[1],xmm6[1] vunpcklps %xmm8, %xmm2, %xmm6 # xmm6 = xmm2[0],xmm8[0],xmm2[1],xmm8[1] vunpckhps %xmm8, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm8[2],xmm2[3],xmm8[3] vunpcklps %xmm9, %xmm0, %xmm8 # xmm8 = xmm0[0],xmm9[0],xmm0[1],xmm9[1] vunpckhps %xmm9, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm9[2],xmm0[3],xmm9[3] vunpcklps %xmm11, %xmm3, %xmm9 # xmm9 = xmm3[0],xmm11[0],xmm3[1],xmm11[1] vunpckhps %xmm11, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm11[2],xmm3[3],xmm11[3] vunpcklps %xmm3, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1] vunpcklps %xmm9, %xmm8, %xmm10 # xmm10 = xmm8[0],xmm9[0],xmm8[1],xmm9[1] vunpckhps %xmm9, %xmm8, %xmm11 # xmm11 = xmm8[2],xmm9[2],xmm8[3],xmm9[3] vmovaps 0x30(%rbp,%rax), %xmm3 vmovaps %xmm3, 0x220(%rsp) vmovaps 0x40(%rbp,%rax), %xmm3 vmovaps %xmm3, 0xe0(%rsp) vsubps %xmm6, %xmm4, %xmm3 vmovaps %xmm3, 0x60(%rsp) vsubps %xmm2, %xmm7, %xmm6 vmovaps %xmm13, 0x20(%rsp) vsubps %xmm1, %xmm13, %xmm8 vsubps %xmm4, %xmm10, %xmm9 vsubps %xmm7, %xmm11, %xmm10 vsubps %xmm13, %xmm0, %xmm11 vmulps %xmm6, %xmm11, %xmm0 vmulps %xmm10, %xmm8, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmovaps %xmm0, 0x50(%rsp) vmulps %xmm9, %xmm8, %xmm1 vmulps %xmm3, %xmm11, %xmm2 vsubps %xmm1, %xmm2, %xmm5 vmulps %xmm3, %xmm10, %xmm2 vmulps %xmm6, %xmm9, %xmm12 vsubps %xmm2, %xmm12, %xmm3 vbroadcastss (%rbx,%r14,4), %xmm12 vbroadcastss 0x10(%rbx,%r14,4), %xmm13 vbroadcastss 0x20(%rbx,%r14,4), %xmm14 vbroadcastss 0x50(%rbx,%r14,4), %xmm15 vsubps %xmm12, %xmm4, %xmm2 vbroadcastss 0x60(%rbx,%r14,4), %xmm12 vsubps %xmm13, %xmm7, %xmm4 vmovaps 0x20(%rsp), %xmm0 vsubps %xmm14, %xmm0, %xmm7 vmulps %xmm7, %xmm15, %xmm13 vmulps %xmm4, %xmm12, %xmm14 vsubps %xmm13, %xmm14, %xmm13 vbroadcastss 0x40(%rbx,%r14,4), %xmm14 vmulps %xmm2, %xmm12, %xmm0 vmulps %xmm7, %xmm14, %xmm1 vsubps %xmm0, %xmm1, %xmm0 vmulps %xmm4, %xmm14, %xmm1 vmovaps %xmm2, 0x40(%rsp) vmulps %xmm2, %xmm15, %xmm2 vsubps %xmm1, %xmm2, %xmm1 vmovaps %xmm3, 0x20(%rsp) vmulps %xmm3, %xmm12, %xmm2 vmulps %xmm5, %xmm15, %xmm12 vaddps %xmm2, %xmm12, %xmm2 vmovaps 0x50(%rsp), %xmm3 vmulps %xmm3, %xmm14, %xmm12 vmovaps %xmm3, %xmm14 vaddps %xmm2, %xmm12, %xmm2 vmulps %xmm1, %xmm11, %xmm11 vmulps %xmm0, %xmm10, %xmm10 vaddps %xmm10, %xmm11, %xmm10 vmulps %xmm13, %xmm9, %xmm9 vaddps %xmm10, %xmm9, %xmm10 vmulps %xmm1, %xmm8, %xmm1 vmulps %xmm0, %xmm6, %xmm0 vaddps %xmm0, %xmm1, %xmm0 vbroadcastss 0x1cdd674(%rip), %xmm1 # 0x1f20ec0 vandps %xmm1, %xmm2, %xmm9 vxorps %xmm10, %xmm9, %xmm6 vmulps 0x60(%rsp), %xmm13, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm8 vxorps %xmm10, %xmm10, %xmm10 vcmpnltps %xmm10, %xmm6, %xmm0 vcmpnltps %xmm10, %xmm8, %xmm1 vandps %xmm1, %xmm0, %xmm0 vbroadcastss 0x1cdd643(%rip), %xmm1 # 0x1f20ec4 vandps %xmm1, %xmm2, %xmm3 vcmpneqps %xmm2, %xmm10, %xmm1 vandps %xmm1, %xmm0, %xmm10 vaddps %xmm6, %xmm8, %xmm0 vcmpleps %xmm3, %xmm0, %xmm11 vtestps %xmm11, %xmm10 jne 0x2438af incq %r12 cmpq %r8, %r12 jne 0x24362a jmp 0x243e0e vmovaps %xmm5, %xmm15 vandps %xmm11, %xmm10, %xmm10 vmulps 0x20(%rsp), %xmm7, %xmm0 vmulps %xmm4, %xmm5, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vmulps 0x40(%rsp), %xmm14, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vxorps %xmm0, %xmm9, %xmm5 vbroadcastss 0x30(%rbx,%r14,4), %xmm0 vmulps %xmm0, %xmm3, %xmm0 vcmpltps %xmm5, %xmm0, %xmm0 vbroadcastss 0x80(%rbx,%r14,4), %xmm1 vmulps %xmm1, %xmm3, %xmm1 vcmpleps %xmm1, %xmm5, %xmm1 vandps %xmm1, %xmm0, %xmm4 vtestps %xmm10, %xmm4 je 0x24389e vandps %xmm4, %xmm10, %xmm4 vmovaps %xmm6, 0xf0(%rsp) vmovaps %xmm8, 0x100(%rsp) vmovaps %xmm5, 0x110(%rsp) vmovaps %xmm3, 0x120(%rsp) vmovaps %xmm4, 0x140(%rsp) vmovaps %xmm14, 0x180(%rsp) vmovaps %xmm15, 0x190(%rsp) vmovaps 0x20(%rsp), %xmm0 vmovaps %xmm0, 0x1a0(%rsp) vmovaps %xmm4, 0x30(%rsp) vrcpps %xmm3, %xmm0 vmulps %xmm0, %xmm3, %xmm1 vbroadcastss 0x1ca8da9(%rip), %xmm2 # 0x1eec714 vsubps %xmm1, %xmm2, %xmm1 vmulps %xmm1, %xmm0, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vmulps 0x110(%rsp), %xmm0, %xmm5 vmovaps %xmm5, 0x170(%rsp) vmulps 0xf0(%rsp), %xmm0, %xmm1 vmovaps %xmm1, 0x150(%rsp) vmulps 0x100(%rsp), %xmm0, %xmm0 vbroadcastss 0x1ca8073(%rip), %xmm1 # 0x1eeba20 vblendvps %xmm4, %xmm5, %xmm1, %xmm1 vmovaps %xmm0, 0x160(%rsp) vshufps $0xb1, %xmm1, %xmm1, %xmm0 # xmm0 = xmm1[1,0,3,2] vminps %xmm1, %xmm0, %xmm0 vshufpd $0x1, %xmm0, %xmm0, %xmm2 # xmm2 = xmm0[1,0] vminps %xmm0, %xmm2, %xmm0 vcmpeqps %xmm0, %xmm1, %xmm0 vtestps %xmm4, %xmm0 vpcmpeqd %xmm1, %xmm1, %xmm1 je 0x2439e2 vmovaps %xmm0, %xmm1 vandps %xmm1, %xmm4, %xmm0 vmovmskps %xmm0, %eax bsfq %rax, %rdx movl 0x220(%rsp,%rdx,4), %eax movq 0x1e8(%r15), %rcx movq (%rcx,%rax,8), %rsi movl 0x90(%rbx,%r14,4), %ecx testl %ecx, 0x34(%rsi) je 0x243c96 movq 0x10(%r10), %rcx cmpq $0x0, 0x10(%rcx) jne 0x243a27 cmpq $0x0, 0x40(%rsi) je 0x243d63 vmovaps %xmm5, 0x40(%rsp) movq %r8, 0x20(%rsp) movq %r9, 0x50(%rsp) movq %r11, 0x60(%rsp) vbroadcastss 0x150(%rsp,%rdx,4), %xmm0 vbroadcastss 0x160(%rsp,%rdx,4), %xmm1 movq 0x8(%r10), %rcx vmovd %eax, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vbroadcastss 0xe0(%rsp,%rdx,4), %xmm3 vbroadcastss 0x180(%rsp,%rdx,4), %xmm4 vbroadcastss 0x190(%rsp,%rdx,4), %xmm5 vbroadcastss 0x1a0(%rsp,%rdx,4), %xmm6 vmovaps %xmm4, 0x230(%rsp) vmovaps %xmm5, 0x240(%rsp) vmovaps %xmm6, 0x250(%rsp) vmovaps %xmm0, 0x260(%rsp) vmovaps %xmm1, 0x270(%rsp) vmovaps %xmm3, 0x280(%rsp) vmovdqa %xmm2, 0x290(%rsp) leaq 0x2a0(%rsp), %rax vcmptrueps %ymm0, %ymm0, %ymm0 vmovups %ymm0, (%rax) vbroadcastss (%rcx), %xmm0 vmovaps %xmm0, 0x2a0(%rsp) vbroadcastss 0x4(%rcx), %xmm0 vmovaps %xmm0, 0x2b0(%rsp) vmovss 0x80(%rbx,%r14,4), %xmm0 vmovss %xmm0, 0x1c(%rsp) movq %rdx, 0x90(%rsp) vmovss 0x170(%rsp,%rdx,4), %xmm0 vmovss %xmm0, 0x80(%rbx,%r14,4) movq 0x98(%rsp), %rax vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x70(%rsp) leaq 0x70(%rsp), %rax movq %rax, 0xb0(%rsp) movq 0x18(%rsi), %rax movq %rax, 0xb8(%rsp) movq 0x8(%r10), %rax movq %rax, 0xc0(%rsp) movq %rbx, 0xc8(%rsp) leaq 0x230(%rsp), %rax movq %rax, 0xd0(%rsp) movl $0x4, 0xd8(%rsp) movq %rsi, 0x88(%rsp) movq 0x40(%rsi), %rax testq %rax, %rax je 0x243b95 leaq 0xb0(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x70(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x243cea movq 0x1248(%rsp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x243bdb testb $0x2, (%rcx) jne 0x243bce movq 0x88(%rsp), %rcx testb $0x40, 0x3e(%rcx) je 0x243bdb leaq 0xb0(%rsp), %rdi vzeroupper callq %rax vmovdqa 0x70(%rsp), %xmm2 vpcmpeqd 0x1ca7e27(%rip), %xmm2, %xmm1 # 0x1eeba10 vpxor 0x1ca822f(%rip), %xmm1, %xmm0 # 0x1eebe20 vptest %xmm2, %xmm2 je 0x243cfa vpxor 0x1ca821c(%rip), %xmm1, %xmm1 # 0x1eebe20 movq 0xc8(%rsp), %rax movq 0xd0(%rsp), %rcx vmovaps (%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0xc0(%rax) vmovaps 0x10(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0xd0(%rax) vmovaps 0x20(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0xe0(%rax) vmovaps 0x30(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0xf0(%rax) vmovaps 0x40(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0x100(%rax) vmovaps 0x50(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0x110(%rax) vmovaps 0x60(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0x120(%rax) vmovaps 0x70(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0x130(%rax) vmovaps 0x80(%rcx), %xmm2 vmaskmovps %xmm2, %xmm1, 0x140(%rax) jmp 0x243cfa andl $0x0, 0x30(%rsp,%rdx,4) vmovaps 0x30(%rsp), %xmm0 vtestps %xmm0, %xmm0 je 0x24389e vbroadcastss 0x1ca7d6b(%rip), %xmm1 # 0x1eeba20 vblendvps %xmm0, %xmm5, %xmm1, %xmm1 vshufps $0xb1, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,0,3,2] vminps %xmm1, %xmm2, %xmm2 vshufpd $0x1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0] vminps %xmm2, %xmm3, %xmm2 vcmpeqps %xmm2, %xmm1, %xmm1 vtestps %xmm0, %xmm1 je 0x243cdd vandps %xmm0, %xmm1, %xmm0 vmovmskps %xmm0, %eax bsfq %rax, %rdx jmp 0x2439ee vpcmpeqd 0x1ca7d1e(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1ca8126(%rip), %xmm0, %xmm0 # 0x1eebe20 vmovddup 0x1cdd1e6(%rip), %xmm1 # xmm1 = mem[0,0] vptest %xmm1, %xmm0 jne 0x243d19 vmovd 0x1c(%rsp), %xmm0 vmovd %xmm0, 0x80(%rbx,%r14,4) movq 0x90(%rsp), %rax andl $0x0, 0x30(%rsp,%rax,4) vbroadcastss 0x80(%rbx,%r14,4), %xmm0 vmovaps 0x40(%rsp), %xmm5 vcmpleps %xmm0, %xmm5, %xmm0 vandps 0x30(%rsp), %xmm0, %xmm0 vmovaps %xmm0, 0x30(%rsp) movq 0x1248(%rsp), %r10 movq 0x60(%rsp), %r11 movq 0x50(%rsp), %r9 movq 0x20(%rsp), %r8 jmp 0x243c9b vmovss 0x150(%rsp,%rdx,4), %xmm0 vmovss 0x160(%rsp,%rdx,4), %xmm1 vmovss 0x170(%rsp,%rdx,4), %xmm2 vmovss %xmm2, 0x80(%rbx,%r14,4) vmovss 0x180(%rsp,%rdx,4), %xmm2 vmovss %xmm2, 0xc0(%rbx,%r14,4) vmovss 0x190(%rsp,%rdx,4), %xmm2 vmovss %xmm2, 0xd0(%rbx,%r14,4) vmovss 0x1a0(%rsp,%rdx,4), %xmm2 vmovss %xmm2, 0xe0(%rbx,%r14,4) vmovss %xmm0, 0xf0(%rbx,%r14,4) vmovss %xmm1, 0x100(%rbx,%r14,4) movl 0xe0(%rsp,%rdx,4), %ecx movl %ecx, 0x110(%rbx,%r14,4) movl %eax, 0x120(%rbx,%r14,4) movq 0x8(%r10), %rax movl (%rax), %eax movl %eax, 0x130(%rbx,%r14,4) movq 0x8(%r10), %rax movl 0x4(%rax), %eax movl %eax, 0x140(%rbx,%r14,4) jmp 0x24389e vbroadcastss 0x80(%rbx,%r14,4), %xmm0 vmovaps 0x210(%rsp), %xmm7 vmovaps 0x200(%rsp), %xmm8 vmovaps 0x1f0(%rsp), %xmm9 vmovaps 0x1e0(%rsp), %xmm10 vmovaps 0x1d0(%rsp), %xmm11 vmovaps 0x1c0(%rsp), %xmm12 movq 0xa0(%rsp), %r12 movq 0x10(%rsp), %rsi movq 0x8(%rsp), %rdi movq (%rsp), %r15 vmovaps 0x1b0(%rsp), %xmm13 jmp 0x243353 addq $0x1208, %rsp # imm = 0x1208 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<4, 4, 1, false, embree::avx::SubGridIntersectorKMoeller<4, 4, true>, true>::occluded1(embree::Accel::Intersectors, embree::BVHN<4> const, embree::NodeRefPtr<4>, unsigned long, embree::avx::SubGridQuadMIntersectorKMoellerTrumbore<4, 4, true>&, embree::RayK<4>&, embree::avx::TravRayK<4, false> const&, embree::RayQueryContext*)	bool BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::occluded1(Accel::Intersectors* This, const BVH* bvh, NodeRef root, size_t k, Precalculations& pre, RayK<K>& ray, const TravRayK<K, robust>& tray, RayQueryContext* context) { /* stack state / NodeRef stack[stackSizeSingle]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSizeSingle; stack[0] = root; /* load the ray into SIMD registers / TravRay<N,robust> tray1; tray1.template init<K>(k, tray.org, tray.dir, tray.rdir, tray.nearXYZ, tray.tnear[k], tray.tfar[k]); / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes, 1, 1, 1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray1, ray.time()[k], tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / BVHNNodeTraverser1Hit<N, types>::traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves, 1, 1, 1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersectorK::occluded(This, pre, ray, k, context, prim, num, tray1, lazy_node)) { ray.tfar[k] = neg_inf; return true; } if (unlikely(lazy_node)) { stackPtr = lazy_node; stackPtr++; } } return false; }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0xc40, %rsp # imm = 0xC40 movq %r9, 0x10(%rsp) movq 0x10(%rbp), %rax leaq 0x488(%rsp), %rdi movq %rdx, -0x8(%rdi) vbroadcastss (%rax,%rcx,4), %xmm7 vbroadcastss 0x10(%rax,%rcx,4), %xmm8 vbroadcastss 0x20(%rax,%rcx,4), %xmm9 vbroadcastss 0x60(%rax,%rcx,4), %xmm10 vbroadcastss 0x70(%rax,%rcx,4), %xmm11 vbroadcastss 0x80(%rax,%rcx,4), %xmm12 movslq 0x90(%rax,%rcx,4), %r8 movslq 0xa0(%rax,%rcx,4), %r9 movslq 0xb0(%rax,%rcx,4), %r10 movq %r8, %r11 xorq $0x10, %r11 movq %r9, %rbx xorq $0x10, %rbx movq %r10, %r14 xorq $0x10, %r14 vbroadcastss 0xc0(%rax,%rcx,4), %xmm13 vbroadcastss 0xd0(%rax,%rcx,4), %xmm14 movq %r8, %rax shrq $0x2, %rax movq %rax, 0x148(%rsp) leaq 0x1ecc1ba(%rip), %rax # 0x214ff80 vmovups (%rax), %ymm0 vinsertf128 $0x1, 0xf0(%rax), %ymm0, %ymm1 pushq $0x1 popq %rdx movq %rcx, 0x8(%rsp) shll %cl, %edx movslq %edx, %rcx shlq $0x4, %rcx addq %rax, %rcx movq %rcx, 0xd0(%rsp) movq %r11, %rax shrq $0x2, %rax movq %rax, 0x140(%rsp) movq %r9, %rax shrq $0x2, %rax movq %rax, 0x138(%rsp) movq %rbx, %rax shrq $0x2, %rax movq %rax, 0x130(%rsp) movq %r10, %rax shrq $0x2, %rax movq %rax, 0x128(%rsp) movq %r14, %rax shrq $0x2, %rax movq %rax, 0x120(%rsp) vbroadcastss 0x1c6cb88(%rip), %ymm0 # 0x1ef09cc vbroadcastss 0x1c688c7(%rip), %ymm2 # 0x1eec714 vmovaps %ymm1, 0x460(%rsp) vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x440(%rsp) leaq 0x480(%rsp), %r15 vmovaps %xmm7, 0xb0(%rsp) vmovaps %xmm8, 0xa0(%rsp) vmovaps %xmm9, 0x90(%rsp) vmovaps %xmm10, 0x80(%rsp) vmovaps %xmm11, 0x70(%rsp) vmovaps %xmm12, 0x60(%rsp) movq %r8, 0x100(%rsp) movq %r9, 0xf8(%rsp) movq %r10, 0xf0(%rsp) movq %r11, 0xe8(%rsp) movq %rbx, 0xe0(%rsp) movq %r14, 0xd8(%rsp) vmovaps %xmm13, 0x50(%rsp) vmovaps %xmm14, 0x40(%rsp) movq %rdi, %r12 cmpq %r15, %rdi je 0x284b88 leaq -0x8(%r12), %rdi movq -0x8(%r12), %rdx testb $0x8, %dl jne 0x283f81 vmovaps 0x20(%rdx,%r8), %xmm0 vsubps %xmm7, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vmovaps 0x20(%rdx,%r9), %xmm1 vsubps %xmm8, %xmm1, %xmm1 vmulps %xmm1, %xmm11, %xmm1 vpmaxsd %xmm1, %xmm0, %xmm0 vmovaps 0x20(%rdx,%r10), %xmm1 vsubps %xmm9, %xmm1, %xmm1 vmulps %xmm1, %xmm12, %xmm1 vmovaps 0x20(%rdx,%r11), %xmm2 vsubps %xmm7, %xmm2, %xmm2 vmulps %xmm2, %xmm10, %xmm2 vmovaps 0x20(%rdx,%rbx), %xmm3 vsubps %xmm8, %xmm3, %xmm3 vmulps %xmm3, %xmm11, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vmovaps 0x20(%rdx,%r14), %xmm3 vsubps %xmm9, %xmm3, %xmm3 vmulps %xmm3, %xmm12, %xmm3 vpmaxsd %xmm13, %xmm1, %xmm1 vpmaxsd %xmm1, %xmm0, %xmm0 vpminsd %xmm14, %xmm3, %xmm1 vpminsd %xmm1, %xmm2, %xmm1 vpcmpgtd %xmm1, %xmm0, %xmm0 vmovmskps %xmm0, %eax xorb $0xf, %al movzbl %al, %r13d testb $0x8, %dl jne 0x283fb9 testq %r13, %r13 je 0x283fbd andq $-0x10, %rdx bsfq %r13, %rcx leaq -0x1(%r13), %rsi xorl %eax, %eax movq (%rdx,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) andq %r13, %rsi jne 0x283fc2 movq %rcx, %rdx testl %eax, %eax je 0x283eef jmp 0x284039 pushq $0x6 jmp 0x283fbf pushq $0x4 popq %rax jmp 0x283fac movq %rcx, (%rdi) addq $0x8, %rdi movq %rdx, 0x18(%rsp) bsfq %rsi, %rcx movq %rcx, 0x20(%rsp) leaq -0x1(%rsi), %rcx movq %rcx, 0x180(%rsp) movq 0x20(%rsp), %rcx movq 0x18(%rsp), %rdx movq (%rdx,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) movq 0x180(%rsp), %rdx andq %rsi, %rdx je 0x283fa9 movq %rdx, %rsi movq %rcx, (%rdi) addq $0x8, %rdi bsfq %rsi, %rdx leaq -0x1(%rsi), %rcx movq %rcx, 0x20(%rsp) movq 0x18(%rsp), %rcx movq (%rcx,%rdx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) andq 0x20(%rsp), %rsi je 0x283fa9 jmp 0x284008 cmpl $0x6, %eax jne 0x284b39 movq %r13, 0x108(%rsp) movq %r12, 0x110(%rsp) movq %rdi, 0x118(%rsp) movl %edx, %eax andl $0xf, %eax addq $-0x8, %rax movq %rax, 0x150(%rsp) setne %cl je 0x284aca andq $-0x10, %rdx xorl %eax, %eax movq %rdx, 0x18(%rsp) movb %cl, 0x7(%rsp) movq %rax, 0x158(%rsp) imulq $0x58, %rax, %rcx leaq (%rdx,%rcx), %rdi vmovq 0x20(%rdx,%rcx), %xmm0 vmovq 0x24(%rdx,%rcx), %xmm1 vpminub %xmm1, %xmm0, %xmm1 vpcmpeqb %xmm1, %xmm0, %xmm0 vpcmpeqd %xmm1, %xmm1, %xmm1 vpxor %xmm1, %xmm0, %xmm0 vpmovzxbd %xmm0, %xmm0 # xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero vpxor %xmm1, %xmm0, %xmm0 vpslld $0x1f, %xmm0, %xmm0 vmovmskps %xmm0, %eax vbroadcastss 0x38(%rdx,%rcx), %xmm0 vbroadcastss 0x44(%rdx,%rcx), %xmm1 movq %rdx, %rsi movq 0x148(%rsp), %rdx vmovq 0x20(%rdx,%rdi), %xmm2 vpmovzxbd %xmm2, %xmm2 # xmm2 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero,xmm2[2],zero,zero,zero,xmm2[3],zero,zero,zero vcvtdq2ps %xmm2, %xmm2 vmulps %xmm2, %xmm1, %xmm2 vaddps %xmm2, %xmm0, %xmm2 movq 0x140(%rsp), %rdx vmovq 0x20(%rdx,%rdi), %xmm3 vpmovzxbd %xmm3, %xmm3 # xmm3 = xmm3[0],zero,zero,zero,xmm3[1],zero,zero,zero,xmm3[2],zero,zero,zero,xmm3[3],zero,zero,zero vcvtdq2ps %xmm3, %xmm3 vmulps %xmm3, %xmm1, %xmm1 vaddps %xmm1, %xmm0, %xmm0 vbroadcastss 0x3c(%rsi,%rcx), %xmm1 vbroadcastss 0x48(%rsi,%rcx), %xmm3 movq 0x138(%rsp), %rdx vmovq 0x20(%rdx,%rdi), %xmm4 vpmovzxbd %xmm4, %xmm4 # xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero,xmm4[2],zero,zero,zero,xmm4[3],zero,zero,zero vcvtdq2ps %xmm4, %xmm4 vmulps %xmm4, %xmm3, %xmm4 vaddps %xmm4, %xmm1, %xmm4 movq 0x130(%rsp), %rdx vmovq 0x20(%rdx,%rdi), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm3, %xmm3 vaddps %xmm3, %xmm1, %xmm1 vbroadcastss 0x4c(%rsi,%rcx), %xmm3 movq 0x128(%rsp), %rdx vmovq 0x20(%rdx,%rdi), %xmm5 vpmovzxbd %xmm5, %xmm5 # xmm5 = xmm5[0],zero,zero,zero,xmm5[1],zero,zero,zero,xmm5[2],zero,zero,zero,xmm5[3],zero,zero,zero vcvtdq2ps %xmm5, %xmm5 vmulps %xmm5, %xmm3, %xmm5 movq 0x120(%rsp), %rdx movq %rdi, 0x178(%rsp) vmovq 0x20(%rdx,%rdi), %xmm6 vpmovzxbd %xmm6, %xmm6 # xmm6 = xmm6[0],zero,zero,zero,xmm6[1],zero,zero,zero,xmm6[2],zero,zero,zero,xmm6[3],zero,zero,zero vcvtdq2ps %xmm6, %xmm6 vmulps %xmm6, %xmm3, %xmm3 vbroadcastss 0x40(%rsi,%rcx), %xmm6 vaddps %xmm5, %xmm6, %xmm5 vaddps %xmm3, %xmm6, %xmm3 vsubps %xmm7, %xmm2, %xmm2 vmulps %xmm2, %xmm10, %xmm2 vsubps %xmm8, %xmm4, %xmm4 vmulps %xmm4, %xmm11, %xmm4 vpmaxsd %xmm4, %xmm2, %xmm2 vsubps %xmm7, %xmm0, %xmm0 vmulps %xmm0, %xmm10, %xmm0 vsubps %xmm8, %xmm1, %xmm1 vmulps %xmm1, %xmm11, %xmm1 vpminsd %xmm1, %xmm0, %xmm0 vsubps %xmm9, %xmm5, %xmm1 vmulps %xmm1, %xmm12, %xmm1 vpmaxsd %xmm13, %xmm1, %xmm1 vpmaxsd %xmm1, %xmm2, %xmm1 vsubps %xmm9, %xmm3, %xmm2 vmulps %xmm2, %xmm12, %xmm2 vpminsd %xmm14, %xmm2, %xmm2 vpminsd %xmm2, %xmm0, %xmm0 vpcmpgtd %xmm0, %xmm1, %xmm0 vmovmskps %xmm0, %ecx notb %cl andb %al, %cl je 0x284a6d movzbl %cl, %r15d bsfq %r15, %rax movq 0x178(%rsp), %rcx movzwl (%rcx,%rax,8), %r10d movzwl 0x2(%rcx,%rax,8), %r9d movl 0x50(%rcx), %edx movl 0x4(%rcx,%rax,8), %ecx movq 0x18(%rbp), %rax movq (%rax), %rax movq 0x1e8(%rax), %rax movq %rdx, 0x168(%rsp) movq (%rax,%rdx,8), %r12 movq 0x58(%r12), %rdx movq 0x68(%r12), %rsi movq %rcx, 0x160(%rsp) imulq %rcx, %rsi movl %r10d, %edi movl $0x7fff, %ecx # imm = 0x7FFF andl %ecx, %edi movl (%rdx,%rsi), %eax addl %edi, %eax movl %r9d, %r8d andl %ecx, %r8d movq %rdx, 0x170(%rsp) movl 0x4(%rdx,%rsi), %edx movl %edx, %ecx imull %r8d, %ecx addl %eax, %ecx movq 0xa0(%r12), %r11 movq %r11, %rbx imulq %rcx, %rbx movq 0x90(%r12), %rax vmovups (%rax,%rbx), %xmm3 leaq 0x1(%rcx), %rbx imulq %r11, %rbx vmovups (%rax,%rbx), %xmm2 leaq (%rcx,%rdx), %rbx movq %rbx, %r14 imulq %r11, %r14 vmovups (%rax,%r14), %xmm0 leaq (%rcx,%rdx), %r14 incq %r14 movq %r14, %r13 imulq %r11, %r13 vmovups (%rax,%r13), %xmm1 xorl %r13d, %r13d testw %r10w, %r10w setns %r13b addq %r13, %rcx incq %rcx imulq %r11, %rcx vmovups (%rax,%rcx), %xmm5 addq %r14, %r13 movq %r13, %rcx imulq %r11, %rcx vmovups (%rax,%rcx), %xmm4 testw %r9w, %r9w movl $0x0, %ecx cmovnsq %rdx, %rcx addq %rcx, %rbx imulq %r11, %rbx vmovups (%rax,%rbx), %xmm6 addq %rcx, %r14 imulq %r11, %r14 vmovups (%rax,%r14), %xmm7 addq %r13, %rcx imulq %r11, %rcx vmovups (%rax,%rcx), %xmm8 vunpcklps %xmm1, %xmm3, %xmm9 # xmm9 = xmm3[0],xmm1[0],xmm3[1],xmm1[1] vunpckhps %xmm1, %xmm3, %xmm3 # xmm3 = xmm3[2],xmm1[2],xmm3[3],xmm1[3] vunpcklps %xmm0, %xmm2, %xmm10 # xmm10 = xmm2[0],xmm0[0],xmm2[1],xmm0[1] vunpckhps %xmm0, %xmm2, %xmm11 # xmm11 = xmm2[2],xmm0[2],xmm2[3],xmm0[3] vunpcklps %xmm11, %xmm3, %xmm3 # xmm3 = xmm3[0],xmm11[0],xmm3[1],xmm11[1] vunpcklps %xmm10, %xmm9, %xmm11 # xmm11 = xmm9[0],xmm10[0],xmm9[1],xmm10[1] vunpckhps %xmm10, %xmm9, %xmm9 # xmm9 = xmm9[2],xmm10[2],xmm9[3],xmm10[3] vunpcklps %xmm4, %xmm2, %xmm10 # xmm10 = xmm2[0],xmm4[0],xmm2[1],xmm4[1] vunpckhps %xmm4, %xmm2, %xmm2 # xmm2 = xmm2[2],xmm4[2],xmm2[3],xmm4[3] vunpcklps %xmm1, %xmm5, %xmm12 # xmm12 = xmm5[0],xmm1[0],xmm5[1],xmm1[1] vunpckhps %xmm1, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm1[2],xmm5[3],xmm1[3] vunpcklps %xmm5, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm5[0],xmm2[1],xmm5[1] vunpcklps %xmm12, %xmm10, %xmm13 # xmm13 = xmm10[0],xmm12[0],xmm10[1],xmm12[1] vunpckhps %xmm12, %xmm10, %xmm10 # xmm10 = xmm10[2],xmm12[2],xmm10[3],xmm12[3] vunpcklps %xmm8, %xmm1, %xmm5 # xmm5 = xmm1[0],xmm8[0],xmm1[1],xmm8[1] vunpckhps %xmm8, %xmm1, %xmm8 # xmm8 = xmm1[2],xmm8[2],xmm1[3],xmm8[3] vunpcklps %xmm7, %xmm4, %xmm12 # xmm12 = xmm4[0],xmm7[0],xmm4[1],xmm7[1] vunpckhps %xmm7, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm7[2],xmm4[3],xmm7[3] vunpcklps %xmm4, %xmm8, %xmm8 # xmm8 = xmm8[0],xmm4[0],xmm8[1],xmm4[1] vunpcklps %xmm12, %xmm5, %xmm4 # xmm4 = xmm5[0],xmm12[0],xmm5[1],xmm12[1] vunpckhps %xmm12, %xmm5, %xmm5 # xmm5 = xmm5[2],xmm12[2],xmm5[3],xmm12[3] vunpcklps %xmm7, %xmm0, %xmm12 # xmm12 = xmm0[0],xmm7[0],xmm0[1],xmm7[1] vunpckhps %xmm7, %xmm0, %xmm0 # xmm0 = xmm0[2],xmm7[2],xmm0[3],xmm7[3] vunpcklps %xmm6, %xmm1, %xmm7 # xmm7 = xmm1[0],xmm6[0],xmm1[1],xmm6[1] vunpckhps %xmm6, %xmm1, %xmm1 # xmm1 = xmm1[2],xmm6[2],xmm1[3],xmm6[3] vunpcklps %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] vunpcklps %xmm7, %xmm12, %xmm1 # xmm1 = xmm12[0],xmm7[0],xmm12[1],xmm7[1] vunpckhps %xmm7, %xmm12, %xmm6 # xmm6 = xmm12[2],xmm7[2],xmm12[3],xmm7[3] vinsertf128 $0x1, %xmm4, %ymm11, %ymm4 vinsertf128 $0x1, %xmm5, %ymm9, %ymm5 vinsertf128 $0x1, %xmm8, %ymm3, %ymm7 vinsertf128 $0x1, %xmm13, %ymm13, %ymm3 vinsertf128 $0x1, %xmm10, %ymm10, %ymm8 vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vinsertf128 $0x1, %xmm6, %ymm6, %ymm10 vinsertf128 $0x1, %xmm0, %ymm0, %ymm11 vsubps %ymm3, %ymm4, %ymm0 vmovaps %ymm0, 0x20(%rsp) vsubps %ymm8, %ymm5, %ymm6 vsubps %ymm2, %ymm7, %ymm8 vsubps %ymm4, %ymm1, %ymm9 vsubps %ymm5, %ymm10, %ymm10 vsubps %ymm7, %ymm11, %ymm11 vmulps %ymm6, %ymm11, %ymm1 vmulps %ymm8, %ymm10, %ymm2 vsubps %ymm1, %ymm2, %ymm1 vmovaps %ymm1, 0x180(%rsp) vmulps %ymm8, %ymm9, %ymm2 vmulps %ymm0, %ymm11, %ymm3 vsubps %ymm2, %ymm3, %ymm1 vmovaps %ymm1, 0x200(%rsp) vmulps %ymm0, %ymm10, %ymm3 vmulps %ymm6, %ymm9, %ymm12 movq 0x10(%rsp), %rax movq 0x8(%rsp), %rcx vbroadcastss (%rax,%rcx,4), %ymm13 vbroadcastss 0x10(%rax,%rcx,4), %ymm14 vbroadcastss 0x20(%rax,%rcx,4), %ymm15 vbroadcastss 0x50(%rax,%rcx,4), %ymm1 vsubps %ymm3, %ymm12, %ymm0 vsubps %ymm13, %ymm4, %ymm4 vbroadcastss 0x60(%rax,%rcx,4), %ymm12 vsubps %ymm14, %ymm5, %ymm5 vsubps %ymm15, %ymm7, %ymm7 vmulps %ymm1, %ymm7, %ymm13 vmulps %ymm5, %ymm12, %ymm14 vsubps %ymm13, %ymm14, %ymm13 vbroadcastss 0x40(%rax,%rcx,4), %ymm14 vmulps %ymm4, %ymm12, %ymm15 vmulps %ymm7, %ymm14, %ymm2 vsubps %ymm15, %ymm2, %ymm2 vmulps %ymm5, %ymm14, %ymm15 vmulps %ymm1, %ymm4, %ymm3 vsubps %ymm15, %ymm3, %ymm3 vmovaps 0x200(%rsp), %ymm15 vmovaps %ymm0, 0x1e0(%rsp) vmulps %ymm0, %ymm12, %ymm12 vmulps %ymm1, %ymm15, %ymm1 vaddps %ymm1, %ymm12, %ymm1 vmovaps 0x180(%rsp), %ymm0 vmulps %ymm0, %ymm14, %ymm12 vmovaps %ymm0, %ymm14 vaddps %ymm1, %ymm12, %ymm1 vmulps %ymm3, %ymm11, %ymm11 vmulps %ymm2, %ymm10, %ymm10 vaddps %ymm10, %ymm11, %ymm10 vmulps %ymm13, %ymm9, %ymm9 vaddps %ymm10, %ymm9, %ymm10 vmulps %ymm3, %ymm8, %ymm3 vmulps %ymm2, %ymm6, %ymm2 vaddps %ymm2, %ymm3, %ymm2 vbroadcastss 0x1c9c9ba(%rip), %ymm3 # 0x1f20ec0 vandps %ymm3, %ymm1, %ymm9 vxorps %ymm10, %ymm9, %ymm6 vmulps 0x20(%rsp), %ymm13, %ymm0 vaddps %ymm2, %ymm0, %ymm0 vxorps %ymm0, %ymm9, %ymm8 vxorps %xmm3, %xmm3, %xmm3 vcmpnltps %ymm3, %ymm6, %ymm0 vcmpnltps %ymm3, %ymm8, %ymm2 vandps %ymm2, %ymm0, %ymm2 vbroadcastss 0x1c9c98c(%rip), %ymm0 # 0x1f20ec4 vandps %ymm0, %ymm1, %ymm0 vcmpneqps %ymm3, %ymm1, %ymm1 vandps %ymm1, %ymm2, %ymm10 vaddps %ymm6, %ymm8, %ymm1 vcmpleps %ymm0, %ymm1, %ymm11 vtestps %ymm11, %ymm10 jne 0x284567 leaq -0x1(%r15), %rax andq %rax, %r15 jne 0x284218 jmp 0x284a6d vandps %ymm11, %ymm10, %ymm1 vextractf128 $0x1, %ymm1, %xmm2 vpackssdw %xmm2, %xmm1, %xmm1 vmulps 0x1e0(%rsp), %ymm7, %ymm2 vmulps %ymm5, %ymm15, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vmulps %ymm4, %ymm14, %ymm3 vaddps %ymm2, %ymm3, %ymm2 vxorps %ymm2, %ymm9, %ymm4 movq 0x10(%rsp), %rax movq 0x8(%rsp), %rcx vbroadcastss 0x30(%rax,%rcx,4), %ymm2 vmulps %ymm2, %ymm0, %ymm2 vcmpltps %ymm4, %ymm2, %ymm2 vbroadcastss 0x80(%rax,%rcx,4), %ymm3 vmulps %ymm3, %ymm0, %ymm3 vcmpleps %ymm3, %ymm4, %ymm3 vandps %ymm3, %ymm2, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vpackssdw %xmm3, %xmm2, %xmm2 vpand %xmm1, %xmm2, %xmm1 vpmovzxwd %xmm1, %xmm2 # xmm2 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero vpslld $0x1f, %xmm2, %xmm2 vpsrad $0x1f, %xmm2, %xmm2 vpunpckhwd %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[4,4,5,5,6,6,7,7] vpslld $0x1f, %xmm1, %xmm1 vpsrad $0x1f, %xmm1, %xmm1 vinsertf128 $0x1, %xmm1, %ymm2, %ymm5 vtestps %ymm5, %ymm5 je 0x284555 vmovaps %ymm6, 0x2c0(%rsp) vmovaps %ymm8, 0x2e0(%rsp) vmovaps %ymm4, 0x300(%rsp) vmovaps %ymm0, 0x320(%rsp) vmovaps %ymm5, 0x360(%rsp) vmovaps 0x2e0(%rsp), %ymm1 vsubps %ymm1, %ymm0, %ymm2 vmovaps 0x2c0(%rsp), %ymm3 vmovaps 0x460(%rsp), %ymm4 vblendvps %ymm4, %ymm2, %ymm3, %ymm5 vsubps %ymm3, %ymm0, %ymm2 vblendvps %ymm4, %ymm2, %ymm1, %ymm4 vmovaps %ymm5, 0x2c0(%rsp) vmovaps %ymm4, 0x2e0(%rsp) vmovaps 0x440(%rsp), %ymm2 vmulps %ymm2, %ymm14, %ymm1 vmovaps %ymm1, 0x3e0(%rsp) vmulps %ymm2, %ymm15, %ymm1 vmovaps %ymm1, 0x400(%rsp) vmulps 0x1e0(%rsp), %ymm2, %ymm1 vmovaps %ymm1, 0x420(%rsp) vmovd %edi, %xmm1 vpshufd $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vpaddd 0x1c982dc(%rip), %xmm1, %xmm1 # 0x1f1c990 vmovd %r8d, %xmm2 vpshufd $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vpaddd 0x1c982da(%rip), %xmm2, %xmm2 # 0x1f1c9a0 movq 0x170(%rsp), %rcx movzwl 0x8(%rcx,%rsi), %eax decl %eax vcvtsi2ss %eax, %xmm7, %xmm3 vrcpss %xmm3, %xmm3, %xmm6 vmulss %xmm3, %xmm6, %xmm3 vmovss 0x1c6c90f(%rip), %xmm8 # 0x1ef0ff8 vsubss %xmm3, %xmm8, %xmm3 movzwl 0xa(%rcx,%rsi), %eax decl %eax vcvtsi2ss %eax, %xmm7, %xmm7 vmulss %xmm3, %xmm6, %xmm3 vrcpss %xmm7, %xmm7, %xmm6 vmulss %xmm7, %xmm6, %xmm7 vsubss %xmm7, %xmm8, %xmm7 vmulss %xmm7, %xmm6, %xmm6 vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vcvtdq2ps %ymm1, %ymm1 vmulps %ymm1, %ymm0, %ymm1 vaddps %ymm5, %ymm1, %ymm1 vshufps $0x0, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmulps %ymm3, %ymm1, %ymm1 vmovaps %ymm1, 0x2c0(%rsp) vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vcvtdq2ps %ymm2, %ymm2 vmulps %ymm2, %ymm0, %ymm2 vaddps %ymm4, %ymm2, %ymm2 vshufps $0x0, %xmm6, %xmm6, %xmm3 # xmm3 = xmm6[0,0,0,0] vinsertf128 $0x1, %xmm3, %ymm3, %ymm3 vmulps %ymm3, %ymm2, %ymm2 vmovaps %ymm2, 0x2e0(%rsp) movq 0x10(%rsp), %rax movq 0x8(%rsp), %rcx movl 0x90(%rax,%rcx,4), %eax testl %eax, 0x34(%r12) je 0x284555 movq 0x160(%rsp), %rdx movq 0x168(%rsp), %rcx movq 0x18(%rbp), %rax movq 0x10(%rax), %rax cmpq $0x0, 0x10(%rax) jne 0x2847a7 cmpq $0x0, 0x48(%r12) je 0x284b43 vrcpps %ymm0, %ymm3 vmulps %ymm3, %ymm0, %ymm0 vbroadcastss 0x1c67f5c(%rip), %ymm4 # 0x1eec714 vsubps %ymm0, %ymm4, %ymm0 vmulps %ymm0, %ymm3, %ymm0 vaddps %ymm0, %ymm3, %ymm0 vmulps 0x300(%rsp), %ymm0, %ymm3 vmovaps %ymm3, 0x3c0(%rsp) vmulps %ymm0, %ymm1, %ymm1 vmovaps %ymm1, 0x380(%rsp) vmulps %ymm0, %ymm2, %ymm0 vmovaps %ymm0, 0x3a0(%rsp) vmovaps 0x360(%rsp), %ymm0 vmovmskps %ymm0, %ebx bsfq %rbx, %r14 testl %ebx, %ebx setne %r13b je 0x284a5e vmovd %ecx, %xmm0 vpshufd $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmovdqa %xmm0, 0x180(%rsp) vmovd %edx, %xmm0 vpshufd $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmovdqa %xmm0, 0x200(%rsp) movq 0xd0(%rsp), %rax vmovaps (%rax), %xmm0 vmovaps %xmm0, 0x1e0(%rsp) movq 0x10(%rsp), %rdx movq 0x8(%rsp), %rax vmovss 0x80(%rdx,%rax,4), %xmm0 vmovss %xmm0, 0x20(%rsp) vmovss 0x3c0(%rsp,%r14,4), %xmm0 vbroadcastss 0x380(%rsp,%r14,4), %xmm1 vbroadcastss 0x3a0(%rsp,%r14,4), %xmm2 vmovss %xmm0, 0x80(%rdx,%rax,4) movq 0x18(%rbp), %rax movq 0x8(%rax), %rax vbroadcastss 0x3e0(%rsp,%r14,4), %xmm0 vbroadcastss 0x400(%rsp,%r14,4), %xmm3 vbroadcastss 0x420(%rsp,%r14,4), %xmm4 vmovaps %xmm0, 0x230(%rsp) vmovaps %xmm3, 0x240(%rsp) vmovaps %xmm4, 0x250(%rsp) vmovaps %xmm1, 0x260(%rsp) vmovaps %xmm2, 0x270(%rsp) vmovaps 0x200(%rsp), %xmm0 vmovaps %xmm0, 0x280(%rsp) vmovaps 0x180(%rsp), %xmm0 vmovaps %xmm0, 0x290(%rsp) leaq 0x2a0(%rsp), %rcx vcmptrueps %ymm0, %ymm0, %ymm0 vmovups %ymm0, (%rcx) vbroadcastss (%rax), %xmm0 vmovaps %xmm0, 0x2a0(%rsp) vbroadcastss 0x4(%rax), %xmm0 vmovaps %xmm0, 0x2b0(%rsp) vmovaps 0x1e0(%rsp), %xmm0 vmovaps %xmm0, 0xc0(%rsp) leaq 0xc0(%rsp), %rcx movq %rcx, 0x1b0(%rsp) movq 0x18(%r12), %rcx movq %rcx, 0x1b8(%rsp) movq %rax, 0x1c0(%rsp) movq %rdx, 0x1c8(%rsp) leaq 0x230(%rsp), %rax movq %rax, 0x1d0(%rsp) movl $0x4, 0x1d8(%rsp) movq 0x48(%r12), %rax testq %rax, %rax je 0x28499c leaq 0x1b0(%rsp), %rdi vzeroupper callq %rax vmovdqa 0xc0(%rsp), %xmm0 vptest %xmm0, %xmm0 je 0x284a11 movq 0x18(%rbp), %rax movq 0x10(%rax), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0x2849d7 testb $0x2, (%rcx) jne 0x2849ca testb $0x40, 0x3e(%r12) je 0x2849d7 leaq 0x1b0(%rsp), %rdi vzeroupper callq %rax vpxor %xmm0, %xmm0, %xmm0 vpcmpeqd 0xc0(%rsp), %xmm0, %xmm1 vpxor 0x1c67434(%rip), %xmm1, %xmm0 # 0x1eebe20 movq 0x1c8(%rsp), %rax vbroadcastss 0x1c68187(%rip), %xmm2 # 0x1eecb84 vblendvps %xmm1, 0x80(%rax), %xmm2, %xmm1 vmovaps %xmm1, 0x80(%rax) jmp 0x284a21 vpcmpeqd 0x1c66ff7(%rip), %xmm0, %xmm0 # 0x1eeba10 vpxor 0x1c673ff(%rip), %xmm0, %xmm0 # 0x1eebe20 vmovddup 0x1c9c4bf(%rip), %xmm1 # xmm1 = mem[0,0] vptest %xmm1, %xmm0 jne 0x284a5e movq 0x10(%rsp), %rax movq 0x8(%rsp), %rcx vmovss 0x20(%rsp), %xmm0 vmovss %xmm0, 0x80(%rax,%rcx,4) btcq %r14, %rbx bsfq %rbx, %r14 testq %rbx, %rbx setne %r13b jne 0x284846 testb $0x1, %r13b je 0x284555 jmp 0x284b43 movq 0x158(%rsp), %rax incq %rax cmpq 0x150(%rsp), %rax setb %cl vmovaps 0xb0(%rsp), %xmm7 vmovaps 0xa0(%rsp), %xmm8 vmovaps 0x90(%rsp), %xmm9 vmovaps 0x80(%rsp), %xmm10 vmovaps 0x70(%rsp), %xmm11 vmovaps 0x60(%rsp), %xmm12 vmovaps 0x50(%rsp), %xmm13 vmovaps 0x40(%rsp), %xmm14 movq 0x18(%rsp), %rdx jne 0x28407f xorl %eax, %eax testb $0x1, %cl movq 0x118(%rsp), %rdi movq 0x100(%rsp), %r8 movq 0xf8(%rsp), %r9 movq 0xf0(%rsp), %r10 movq 0xe8(%rsp), %r11 movq 0xe0(%rsp), %rbx movq 0xd8(%rsp), %r14 leaq 0x480(%rsp), %r15 movq 0x110(%rsp), %r12 movq 0x108(%rsp), %r13 je 0x284b39 movq 0x10(%rsp), %rax movq 0x8(%rsp), %rcx movl $0xff800000, 0x80(%rax,%rcx,4) # imm = 0xFF800000 pushq $0x1 popq %rax testb $0x3, %al je 0x283ed9 jmp 0x284b88 vmovaps 0xb0(%rsp), %xmm7 vmovaps 0xa0(%rsp), %xmm8 vmovaps 0x90(%rsp), %xmm9 vmovaps 0x80(%rsp), %xmm10 vmovaps 0x70(%rsp), %xmm11 vmovaps 0x60(%rsp), %xmm12 vmovaps 0x50(%rsp), %xmm13 vmovaps 0x40(%rsp), %xmm14 movb 0x7(%rsp), %cl jmp 0x284aca cmpq %r15, %r12 setne %al leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::BVHNIntersectorKHybrid<8, 4, 1, false, embree::avx::ArrayIntersectorK_1<4, embree::avx::ObjectIntersectorK<4, false>>, false>::intersectCoherent(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayHitK<4>&, embree::RayQueryContext*)	__forceinline vboolf4 operator ==(const vint4& a, const vint4& b) { return _mm_castsi128_ps(_mm_cmpeq_epi32(a, b)); }	vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm1 vmovmskps %xmm1, %eax testl %eax, %eax je 0x3695fa pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x26a0, %rsp # imm = 0x26A0 movq (%rsi), %rsi movq %rsi, 0x28(%rsp) movzbl %al, %eax vmovaps (%rdx), %xmm10 vmovaps 0x10(%rdx), %xmm11 vmovaps 0x20(%rdx), %xmm12 vmovaps 0x30(%rdx), %xmm3 vmovaps 0x40(%rdx), %xmm2 vbroadcastss 0x1bb8322(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm2, %xmm5 vbroadcastss 0x1b88439(%rip), %xmm6 # 0x1ef0fe8 vcmpltps %xmm6, %xmm5, %xmm5 vblendvps %xmm5, %xmm6, %xmm2, %xmm5 vmovaps 0x50(%rdx), %xmm7 vandps %xmm4, %xmm7, %xmm8 vcmpltps %xmm6, %xmm8, %xmm8 vblendvps %xmm8, %xmm6, %xmm7, %xmm7 vmovaps 0x60(%rdx), %xmm8 vandps %xmm4, %xmm8, %xmm4 vcmpltps %xmm6, %xmm4, %xmm4 vblendvps %xmm4, %xmm6, %xmm8, %xmm4 vrcpps %xmm5, %xmm6 vmulps %xmm6, %xmm5, %xmm5 vbroadcastss 0x1b83b21(%rip), %xmm8 # 0x1eec714 vsubps %xmm5, %xmm8, %xmm5 vmulps %xmm5, %xmm6, %xmm5 vaddps %xmm5, %xmm6, %xmm13 vrcpps %xmm7, %xmm5 vmulps %xmm5, %xmm7, %xmm6 vsubps %xmm6, %xmm8, %xmm6 vmulps %xmm6, %xmm5, %xmm6 vaddps %xmm6, %xmm5, %xmm14 vrcpps %xmm4, %xmm5 vmulps %xmm5, %xmm4, %xmm4 vsubps %xmm4, %xmm8, %xmm4 vmulps %xmm4, %xmm5, %xmm4 vaddps %xmm4, %xmm5, %xmm15 vxorps %xmm6, %xmm6, %xmm6 vmaxps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0xf0(%rsp) vmovaps 0x80(%rdx), %xmm3 vmaxps %xmm6, %xmm3, %xmm3 vmovaps %xmm3, 0xe0(%rsp) vcmpltps %xmm6, %xmm2, %xmm2 vbroadcastss 0x1ba9aa9(%rip), %xmm3 # 0x1f12704 vandps %xmm3, %xmm2, %xmm2 vmovaps 0x50(%rdx), %xmm3 vmovaps 0x60(%rdx), %xmm4 vcmpltps %xmm6, %xmm3, %xmm3 vbroadcastss 0x1bb8265(%rip), %xmm5 # 0x1f20edc vandps %xmm5, %xmm3, %xmm3 vorps %xmm2, %xmm3, %xmm2 vcmpltps %xmm6, %xmm4, %xmm3 vbroadcastss 0x1bf1ce7(%rip), %xmm4 # 0x1f5a974 vandps %xmm4, %xmm3, %xmm3 vpxor %xmm0, %xmm1, %xmm0 vpor %xmm3, %xmm0, %xmm0 vpor %xmm2, %xmm0, %xmm0 vmovdqa %xmm0, 0x70(%rsp) vbroadcastss 0x1b82d74(%rip), %xmm9 # 0x1eeba20 leaq 0x340(%rsp), %r11 pushq $0x8 popq %r8 movq %rcx, 0x40(%rsp) movq %rdx, 0x38(%rsp) vmovaps %xmm10, 0x150(%rsp) vmovaps %xmm11, 0x140(%rsp) vmovaps %xmm12, 0x130(%rsp) vmovaps %xmm13, 0x120(%rsp) vmovaps %xmm14, 0x110(%rsp) vmovaps %xmm15, 0x100(%rsp) bsfq %rax, %rsi vbroadcastss 0x70(%rsp,%rsi,4), %xmm0 vpcmpeqd 0x70(%rsp), %xmm0, %xmm0 vmovmskps %xmm0, %esi notq %rsi vblendvps %xmm0, %xmm13, %xmm9, %xmm1 andq %rax, %rsi movq %rsi, 0x30(%rsp) vshufps $0xb1, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,0,3,2] vminps %xmm1, %xmm2, %xmm1 vshufpd $0x1, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,0] vminps %xmm1, %xmm2, %xmm1 vblendvps %xmm0, %xmm14, %xmm9, %xmm2 vshufps $0xb1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0,3,2] vminps %xmm2, %xmm3, %xmm2 vshufpd $0x1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0] vminps %xmm2, %xmm3, %xmm2 vinsertps $0x1c, %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm2[0],zero,zero vblendvps %xmm0, %xmm15, %xmm9, %xmm2 vshufps $0xb1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0,3,2] vminps %xmm2, %xmm3, %xmm2 vshufpd $0x1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0] vminps %xmm2, %xmm3, %xmm2 vbroadcastss 0x1b83e15(%rip), %xmm8 # 0x1eecb84 vblendvps %xmm0, %xmm13, %xmm8, %xmm3 vinsertps $0x20, %xmm2, %xmm1, %xmm2 # xmm2 = xmm1[0,1],xmm2[0],xmm1[3] vshufps $0xb1, %xmm3, %xmm3, %xmm1 # xmm1 = xmm3[1,0,3,2] vmaxps %xmm3, %xmm1, %xmm1 vshufpd $0x1, %xmm1, %xmm1, %xmm3 # xmm3 = xmm1[1,0] vmaxps %xmm1, %xmm3, %xmm1 vblendvps %xmm0, %xmm14, %xmm8, %xmm3 vshufps $0xb1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0,3,2] vmaxps %xmm3, %xmm4, %xmm3 vshufpd $0x1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0] vmaxps %xmm3, %xmm4, %xmm3 vinsertps $0x1c, %xmm3, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm3[0],zero,zero vblendvps %xmm0, %xmm15, %xmm8, %xmm3 vshufps $0xb1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0,3,2] vmaxps %xmm3, %xmm4, %xmm3 vshufpd $0x1, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,0] vmaxps %xmm3, %xmm4, %xmm3 vinsertps $0x20, %xmm3, %xmm1, %xmm3 # xmm3 = xmm1[0,1],xmm3[0],xmm1[3] vcmpnltps 0x1b82c3e(%rip), %xmm2, %xmm4 # 0x1eeba10 vblendvps %xmm4, %xmm2, %xmm3, %xmm1 vmovshdup %xmm1, %xmm5 # xmm5 = xmm1[1,1,3,3] xorl %esi, %esi vxorps %xmm6, %xmm6, %xmm6 vucomiss %xmm5, %xmm6 seta %sil shll $0x5, %esi orq $0x40, %rsi vshufpd $0x1, %xmm1, %xmm1, %xmm5 # xmm5 = xmm1[1,0] xorl %r12d, %r12d vucomiss %xmm5, %xmm6 seta %r12b shll $0x5, %r12d orq $0x80, %r12 movq %rsi, 0x68(%rsp) xorq $0x20, %rsi movq %rsi, 0x60(%rsp) xorl %ebx, %ebx vucomiss %xmm1, %xmm6 seta %bl vblendvps %xmm0, %xmm10, %xmm9, %xmm5 vshufps $0xb1, %xmm5, %xmm5, %xmm6 # xmm6 = xmm5[1,0,3,2] vminps %xmm5, %xmm6, %xmm5 vshufpd $0x1, %xmm5, %xmm5, %xmm6 # xmm6 = xmm5[1,0] vminps %xmm5, %xmm6, %xmm5 vblendvps %xmm0, %xmm11, %xmm9, %xmm6 vshufps $0xb1, %xmm6, %xmm6, %xmm7 # xmm7 = xmm6[1,0,3,2] vminps %xmm6, %xmm7, %xmm6 vshufpd $0x1, %xmm6, %xmm6, %xmm7 # xmm7 = xmm6[1,0] vminps %xmm6, %xmm7, %xmm6 vblendvps %xmm0, %xmm12, %xmm9, %xmm7 vinsertps $0x1c, %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[0],xmm6[0],zero,zero vshufps $0xb1, %xmm7, %xmm7, %xmm6 # xmm6 = xmm7[1,0,3,2] vminps %xmm7, %xmm6, %xmm6 vshufpd $0x1, %xmm6, %xmm6, %xmm7 # xmm7 = xmm6[1,0] vminps %xmm6, %xmm7, %xmm6 vinsertps $0x20, %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[0,1],xmm6[0],xmm5[3] vmovaps %xmm8, %xmm9 vblendvps %xmm0, %xmm10, %xmm8, %xmm6 vshufps $0xb1, %xmm6, %xmm6, %xmm7 # xmm7 = xmm6[1,0,3,2] vmaxps %xmm6, %xmm7, %xmm6 vshufpd $0x1, %xmm6, %xmm6, %xmm7 # xmm7 = xmm6[1,0] vmaxps %xmm6, %xmm7, %xmm6 vblendvps %xmm0, %xmm11, %xmm8, %xmm7 vshufps $0xb1, %xmm7, %xmm7, %xmm8 # xmm8 = xmm7[1,0,3,2] vmaxps %xmm7, %xmm8, %xmm7 vshufpd $0x1, %xmm7, %xmm7, %xmm8 # xmm8 = xmm7[1,0] vmaxps %xmm7, %xmm8, %xmm7 vblendvps %xmm0, %xmm12, %xmm9, %xmm8 vinsertps $0x1c, %xmm7, %xmm6, %xmm6 # xmm6 = xmm6[0],xmm7[0],zero,zero vshufps $0xb1, %xmm8, %xmm8, %xmm7 # xmm7 = xmm8[1,0,3,2] vmaxps %xmm8, %xmm7, %xmm7 vshufpd $0x1, %xmm7, %xmm7, %xmm8 # xmm8 = xmm7[1,0] vmaxps %xmm7, %xmm8, %xmm7 vinsertps $0x20, %xmm7, %xmm6, %xmm6 # xmm6 = xmm6[0,1],xmm7[0],xmm6[3] vblendvps %xmm4, %xmm5, %xmm6, %xmm7 vblendvps %xmm4, %xmm6, %xmm5, %xmm6 vblendvps %xmm0, 0xe0(%rsp), %xmm9, %xmm8 vshufps $0xb1, %xmm8, %xmm8, %xmm5 # xmm5 = xmm8[1,0,3,2] vmovaps %xmm8, 0x10(%rsp) vmaxps %xmm8, %xmm5, %xmm8 vshufpd $0x1, %xmm8, %xmm8, %xmm9 # xmm9 = xmm8[1,0] vbroadcastss 0x1b82b17(%rip), %xmm5 # 0x1eeba20 vblendvps %xmm0, 0xf0(%rsp), %xmm5, %xmm5 vshufps $0xb1, %xmm5, %xmm5, %xmm0 # xmm0 = xmm5[1,0,3,2] vmovaps %xmm5, 0x170(%rsp) vminps %xmm5, %xmm0, %xmm5 vmaxss %xmm8, %xmm9, %xmm0 vbroadcastss 0x1b82aec(%rip), %xmm9 # 0x1eeba20 vblendvps %xmm4, %xmm3, %xmm2, %xmm2 vmulps %xmm1, %xmm6, %xmm4 vmulps %xmm2, %xmm7, %xmm3 shll $0x5, %ebx movq %rbx, 0x58(%rsp) xorq $0x20, %rbx movq %r12, %r13 xorq $0x20, %r12 movq 0x28(%rsp), %rax movq 0x70(%rax), %rax movq %rax, 0x340(%rsp) andl $0x0, 0x348(%rsp) vshufps $0x0, %xmm1, %xmm1, %xmm6 # xmm6 = xmm1[0,0,0,0] vinsertf128 $0x1, %xmm6, %ymm6, %ymm6 vmovaps %ymm6, 0x300(%rsp) vshufps $0x0, %xmm4, %xmm4, %xmm6 # xmm6 = xmm4[0,0,0,0] vinsertf128 $0x1, %xmm6, %ymm6, %ymm6 vmovaps %ymm6, 0x2e0(%rsp) vshufps $0x55, %xmm1, %xmm1, %xmm6 # xmm6 = xmm1[1,1,1,1] vinsertf128 $0x1, %xmm6, %ymm6, %ymm6 vmovaps %ymm6, 0x2c0(%rsp) vshufps $0x55, %xmm4, %xmm4, %xmm6 # xmm6 = xmm4[1,1,1,1] vinsertf128 $0x1, %xmm6, %ymm6, %ymm6 vmovaps %ymm6, 0x2a0(%rsp) vshufps $0xaa, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovaps %ymm1, 0x280(%rsp) vshufps $0xaa, %xmm4, %xmm4, %xmm1 # xmm1 = xmm4[2,2,2,2] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovaps %ymm1, 0x260(%rsp) vshufps $0x0, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovaps %ymm1, 0x240(%rsp) vshufps $0x0, %xmm3, %xmm3, %xmm1 # xmm1 = xmm3[0,0,0,0] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovaps %ymm1, 0x220(%rsp) vshufps $0x55, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[1,1,1,1] vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovaps %ymm1, 0x200(%rsp) vshufps $0x0, %xmm5, %xmm5, %xmm1 # xmm1 = xmm5[0,0,0,0] vshufps $0xaa, %xmm5, %xmm5, %xmm4 # xmm4 = xmm5[2,2,2,2] vminps %xmm1, %xmm4, %xmm1 vshufps $0x55, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,1,1,1] vinsertf128 $0x1, %xmm4, %ymm4, %ymm4 vmovaps %ymm4, 0x1e0(%rsp) vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vmovaps %ymm2, 0x1c0(%rsp) vshufps $0xaa, %xmm3, %xmm3, %xmm2 # xmm2 = xmm3[2,2,2,2] vinsertf128 $0x1, %xmm2, %ymm2, %ymm2 vmovaps %ymm2, 0x1a0(%rsp) vinsertf128 $0x1, %xmm1, %ymm1, %ymm1 vmovaps %ymm1, 0x180(%rsp) leaq 0x350(%rsp), %r14 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vinsertf128 $0x1, %xmm0, %ymm0, %ymm0 vmovdqa %ymm0, 0x320(%rsp) cmpq %r11, %r14 je 0x3695de movq %r14, %rax addq $-0x10, %r14 vbroadcastss -0x8(%rax), %xmm0 vcmpltps 0x10(%rsp), %xmm0, %xmm1 vtestps %xmm1, %xmm1 je 0x369097 movq -0x10(%rax), %r15 testb $0x8, %r15b jne 0x36942e movq 0x58(%rsp), %rax vmovaps 0x300(%rsp), %ymm1 vmulps 0x40(%r15,%rax), %ymm1, %ymm1 movq 0x68(%rsp), %rax vmovaps 0x2c0(%rsp), %ymm2 vmulps 0x40(%r15,%rax), %ymm2, %ymm2 vsubps 0x2e0(%rsp), %ymm1, %ymm1 vsubps 0x2a0(%rsp), %ymm2, %ymm2 vmovaps 0x280(%rsp), %ymm3 vmulps 0x40(%r15,%r13), %ymm3, %ymm3 vsubps 0x260(%rsp), %ymm3, %ymm3 vmovaps 0x240(%rsp), %ymm4 vmulps 0x40(%r15,%rbx), %ymm4, %ymm4 vsubps 0x220(%rsp), %ymm4, %ymm4 vpmaxsd %xmm2, %xmm1, %xmm5 vextractf128 $0x1, %ymm2, %xmm2 vextractf128 $0x1, %ymm1, %xmm1 vpmaxsd %xmm2, %xmm1, %xmm1 vmovdqa 0x180(%rsp), %ymm6 vpmaxsd %xmm6, %xmm3, %xmm2 vpmaxsd %xmm2, %xmm5, %xmm2 vextractf128 $0x1, %ymm3, %xmm3 vextractf128 $0x1, %ymm6, %xmm5 vpmaxsd %xmm5, %xmm3, %xmm3 movq 0x60(%rsp), %rax vmovaps 0x200(%rsp), %ymm5 vmulps 0x40(%r15,%rax), %ymm5, %ymm5 vsubps 0x1e0(%rsp), %ymm5, %ymm5 vpmaxsd %xmm3, %xmm1, %xmm1 vmovaps 0x1c0(%rsp), %ymm3 vmulps 0x40(%r15,%r12), %ymm3, %ymm3 vsubps 0x1a0(%rsp), %ymm3, %ymm3 vinsertf128 $0x1, %xmm1, %ymm2, %ymm1 vpminsd %xmm5, %xmm4, %xmm2 vextractf128 $0x1, %ymm5, %xmm5 vextractf128 $0x1, %ymm4, %xmm4 vpminsd %xmm5, %xmm4, %xmm4 vmovdqa 0x320(%rsp), %ymm6 vpminsd %xmm6, %xmm3, %xmm5 vpminsd %xmm5, %xmm2, %xmm2 vextractf128 $0x1, %ymm3, %xmm3 vextractf128 $0x1, %ymm6, %xmm5 vpminsd %xmm5, %xmm3, %xmm3 vpminsd %xmm3, %xmm4, %xmm3 vinsertf128 $0x1, %xmm3, %ymm2, %ymm2 vmovaps %ymm1, 0x80(%rsp) vcmpleps %ymm2, %ymm1, %ymm1 vmovmskps %ymm1, %eax testl %eax, %eax je 0x36936f movzbl %al, %edi vmovaps %xmm9, %xmm0 movq %r8, %rax xorl %r9d, %r9d bsfq %rdi, %rsi vbroadcastss 0x40(%r15,%rsi,4), %xmm1 vsubps %xmm10, %xmm1, %xmm1 vmulps %xmm1, %xmm13, %xmm1 vbroadcastss 0x80(%r15,%rsi,4), %xmm2 vsubps %xmm11, %xmm2, %xmm2 vmulps %xmm2, %xmm14, %xmm2 vbroadcastss 0xc0(%r15,%rsi,4), %xmm3 vsubps %xmm12, %xmm3, %xmm3 vmulps %xmm3, %xmm15, %xmm3 vbroadcastss 0x60(%r15,%rsi,4), %xmm4 vsubps %xmm10, %xmm4, %xmm4 vmulps %xmm4, %xmm13, %xmm4 vbroadcastss 0xa0(%r15,%rsi,4), %xmm5 vsubps %xmm11, %xmm5, %xmm5 vmulps %xmm5, %xmm14, %xmm5 vbroadcastss 0xe0(%r15,%rsi,4), %xmm6 vsubps %xmm12, %xmm6, %xmm6 vmulps %xmm6, %xmm15, %xmm6 vpminsd %xmm4, %xmm1, %xmm7 vpminsd %xmm5, %xmm2, %xmm8 vpmaxsd %xmm8, %xmm7, %xmm7 vpminsd %xmm6, %xmm3, %xmm8 vpmaxsd %xmm4, %xmm1, %xmm1 vpmaxsd %xmm5, %xmm2, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vpmaxsd %xmm6, %xmm3, %xmm2 vpmaxsd 0x170(%rsp), %xmm8, %xmm3 vpmaxsd %xmm3, %xmm7, %xmm3 vpminsd 0x10(%rsp), %xmm2, %xmm2 vpminsd %xmm2, %xmm1, %xmm1 vcmpleps %xmm1, %xmm3, %xmm1 vtestps %xmm1, %xmm1 je 0x369337 vbroadcastss 0x80(%rsp,%rsi,4), %xmm1 movq (%r15,%rsi,8), %rsi prefetcht0 (%rsi) prefetcht0 0x40(%rsi) prefetcht0 0x80(%rsi) prefetcht0 0xc0(%rsi) vcmpltps %xmm0, %xmm1, %xmm2 vtestps %xmm2, %xmm2 je 0x369323 cmpq $0x8, %rax je 0x369346 movq %rax, (%r14) vmovaps %xmm0, %xmm2 vmovaps %xmm1, %xmm0 movq %rsi, %rax jmp 0x36932a movq %rsi, (%r14) vmovaps %xmm1, %xmm2 incq %r9 vmovss %xmm2, 0x8(%r14) addq $0x10, %r14 leaq -0x1(%rdi), %rsi andq %rsi, %rdi jne 0x36921f jmp 0x36934f vmovaps %xmm1, %xmm0 movq %rsi, %rax jmp 0x369337 cmpq $0x8, %rax je 0x369374 movb $0x1, %r10b cmpq $0x2, %r9 jae 0x369379 movq %rax, %r15 testb %r10b, %r10b jne 0x3690bf jmp 0x369097 xorl %r10d, %r10d jmp 0x369361 xorl %r10d, %r10d jmp 0x36935e leaq -0x20(%r14), %rsi movl -0x18(%r14), %r8d leaq -0x10(%r14), %rdi cmpl -0x8(%r14), %r8d jae 0x3693b2 vmovaps (%rsi), %xmm1 vmovaps %xmm1, (%rsp) movl 0x8(%rdi), %r8d movl %r8d, 0x8(%rsi) movq (%rdi), %r8 movq %r8, (%rsi) movq (%rsp), %r8 movq %r8, (%rdi) movl 0x8(%rsp), %r8d movl %r8d, 0x8(%rdi) cmpq $0x2, %r9 pushq $0x8 popq %r8 je 0x36935e leaq -0x30(%r14), %r9 movl -0x28(%r14), %r8d cmpl -0x8(%r14), %r8d jae 0x3693f2 vmovaps (%r9), %xmm1 vmovaps %xmm1, (%rsp) movl 0x8(%rdi), %r8d movl %r8d, 0x8(%r9) movq (%rdi), %r8 movq %r8, (%r9) movq (%rsp), %r8 movq %r8, (%rdi) movl 0x8(%rsp), %r8d movl %r8d, 0x8(%rdi) movl -0x28(%r14), %edi cmpl -0x18(%r14), %edi pushq $0x8 popq %r8 jae 0x36935e vmovaps (%r9), %xmm1 vmovaps %xmm1, (%rsp) movl 0x8(%rsi), %edi movl %edi, 0x8(%r9) movq (%rsi), %rdi movq %rdi, (%r9) movq (%rsp), %rdi movq %rdi, (%rsi) movl 0x8(%rsp), %edi movl %edi, 0x8(%rsi) jmp 0x36935e vmovaps 0x10(%rsp), %xmm1 vcmpnleps %xmm0, %xmm1, %xmm2 vtestps %xmm2, %xmm2 je 0x369097 movl %r15d, %eax andl $0xf, %eax addq $-0x8, %rax movq %rax, 0x48(%rsp) je 0x36959b andq $-0x10, %r15 xorl %esi, %esi vmovaps %xmm2, 0x160(%rsp) movq (%rcx), %rax movl (%r15,%rsi,8), %edi movq 0x1e8(%rax), %rax movq (%rax,%rdi,8), %rax vbroadcastss 0x34(%rax), %xmm0 vandps 0x90(%rdx), %xmm0, %xmm0 vpcmpeqd 0x1b82580(%rip), %xmm0, %xmm0 # 0x1eeba10 vtestps %xmm2, %xmm0 jb 0x36958d vandnps %xmm2, %xmm0, %xmm0 movq %rsi, 0x50(%rsp) movl 0x4(%r15,%rsi,8), %r8d vmovaps %xmm0, (%rsp) movq %rsp, %rsi movq %rsi, 0x80(%rsp) movq 0x18(%rax), %rsi movq %rsi, 0x88(%rsp) movq 0x8(%rcx), %rsi movq %rsi, 0x98(%rsp) movq %rdx, 0xa0(%rsp) movl $0x4, 0xa8(%rsp) movl %edi, 0xac(%rsp) movl %r8d, 0x90(%rsp) movq %rax, 0xb0(%rsp) andq $0x0, 0xb8(%rsp) movq 0x10(%rcx), %rcx movq %rcx, 0xc0(%rsp) movq 0x18(%rcx), %rcx testq %rcx, %rcx jne 0x36951d movq 0x60(%rax), %rcx leaq 0x80(%rsp), %rdi vzeroupper callq *%rcx movq 0x40(%rsp), %rcx movq 0x38(%rsp), %rdx vmovaps 0x150(%rsp), %xmm10 vmovaps 0x140(%rsp), %xmm11 vmovaps 0x130(%rsp), %xmm12 vmovaps 0x120(%rsp), %xmm13 vmovaps 0x110(%rsp), %xmm14 vmovaps 0x100(%rsp), %xmm15 vbroadcastss 0x1b824ad(%rip), %xmm9 # 0x1eeba20 leaq 0x340(%rsp), %r11 pushq $0x8 popq %r8 vmovaps 0x160(%rsp), %xmm2 movq 0x50(%rsp), %rsi incq %rsi cmpq %rsi, 0x48(%rsp) jne 0x369468 vmovaps 0x80(%rdx), %xmm0 vcmpltps 0x10(%rsp), %xmm0, %xmm1 vtestps %xmm2, %xmm1 je 0x369097 vmovaps 0x10(%rsp), %xmm1 vblendvps %xmm2, %xmm0, %xmm1, %xmm1 vshufps $0xb1, %xmm1, %xmm1, %xmm0 # xmm0 = xmm1[1,0,3,2] vmovaps %xmm1, 0x10(%rsp) vmaxps %xmm1, %xmm0, %xmm0 vshufpd $0x1, %xmm0, %xmm0, %xmm1 # xmm1 = xmm0[1,0] vmaxss %xmm0, %xmm1, %xmm0 jmp 0x369083 movq 0x30(%rsp), %rax testq %rax, %rax jne 0x368cf8 leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vint4_sse2.h
embree::avx::BVHNIntersectorKHybrid<8, 8, 1, false, embree::avx::ArrayIntersectorK_1<8, embree::avx::QuadMvIntersectorKMoeller<4, 8, false>>, true>::intersect(embree::vint_impl<8>, embree::Accel::Intersectors, embree::RayHitK<8>&, embree::RayQueryContext*)	void BVHNIntersectorKHybrid<N, K, types, robust, PrimitiveIntersectorK, single>::intersect(vint<K>* __restrict__ valid_i, Accel::Intersectors* __restrict__ This, RayHitK<K>& __restrict__ ray, RayQueryContext* __restrict__ context) { BVH* __restrict__ bvh = (BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; #if ENABLE_FAST_COHERENT_CODEPATHS == 1 assert(context); if (unlikely(types == BVH_AN1 && context->user && context->isCoherent())) { intersectCoherent(valid_i, This, ray, context); return; } #endif / filter out invalid rays / vbool<K> valid = valid_i == -1; #if defined(EMBREE_IGNORE_INVALID_RAYS) valid &= ray.valid(); #endif /* return if there are no valid rays / size_t valid_bits = movemask(valid); #if defined(__AVX__) STAT3(normal.trav_hit_boxes[popcnt(movemask(valid))], 1, 1, 1); #endif if (unlikely(valid_bits == 0)) return; / verify correct input / assert(all(valid, ray.valid())); assert(all(valid, ray.tnear() >= 0.0f)); assert(!(types & BVH_MB) \|\| all(valid, (ray.time() >= 0.0f) & (ray.time() <= 1.0f))); Precalculations pre(valid, ray); / load ray / TravRayK<K, robust> tray(ray.org, ray.dir, single ? N : 0); const vfloat<K> org_ray_tnear = max(ray.tnear(), 0.0f); const vfloat<K> org_ray_tfar = max(ray.tfar , 0.0f); if (single) { tray.tnear = select(valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(valid, org_ray_tfar , vfloat<K>(neg_inf)); for (; valid_bits!=0; ) { const size_t i = bscf(valid_bits); intersect1(This, bvh, bvh->root, i, pre, ray, tray, context); } return; } / determine switch threshold based on flags / const size_t switchThreshold = (context->user && context->isCoherent()) ? 2 : switchThresholdIncoherent; vint<K> octant = ray.octant(); octant = select(valid, octant, vint<K>(0xffffffff)); / test whether we have ray with opposing direction signs in the packet / bool split = false; { size_t bits = valid_bits; vbool<K> vsplit( false ); do { const size_t valid_index = bsf(bits); vbool<K> octant_valid = octant[valid_index] == octant; bits &= ~(size_t)movemask(octant_valid); vsplit \|= vint<K>(octant[valid_index]) == (octant^vint<K>(0x7)); } while (bits); if (any(vsplit)) split = true; } do { const size_t valid_index = bsf(valid_bits); const vint<K> diff_octant = vint<K>(octant[valid_index])^octant; const vint<K> count_diff_octant = \ ((diff_octant >> 2) & 1) + ((diff_octant >> 1) & 1) + ((diff_octant >> 0) & 1); vbool<K> octant_valid = (count_diff_octant <= 1) & (octant != vint<K>(0xffffffff)); if (!single \|\| !split) octant_valid = valid; // deactivate octant sorting in pure chunk mode, otherwise instance traversal performance goes down octant = select(octant_valid,vint<K>(0xffffffff),octant); valid_bits &= ~(size_t)movemask(octant_valid); tray.tnear = select(octant_valid, org_ray_tnear, vfloat<K>(pos_inf)); tray.tfar = select(octant_valid, org_ray_tfar , vfloat<K>(neg_inf)); / allocate stack and push root node / vfloat<K> stack_near[stackSizeChunk]; NodeRef stack_node[stackSizeChunk]; stack_node[0] = BVH::invalidNode; stack_near[0] = inf; stack_node[1] = bvh->root; stack_near[1] = tray.tnear; NodeRef stackEnd MAYBE_UNUSED = stack_node+stackSizeChunk; NodeRef* __restrict__ sptr_node = stack_node + 2; vfloat<K>* __restrict__ sptr_near = stack_near + 2; while (1) pop: { /* pop next node from stack / assert(sptr_node > stack_node); sptr_node--; sptr_near--; NodeRef cur = sptr_node; if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } /* cull node if behind closest hit point / vfloat<K> curDist = sptr_near; const vbool<K> active = curDist < tray.tfar; if (unlikely(none(active))) continue; /* switch to single ray traversal / #if (!defined(__WIN32__) \|\| defined(__X86_64__)) && ((defined(__aarch64__)) \|\| defined(__SSE4_2__)) #if FORCE_SINGLE_MODE == 0 if (single) #endif { size_t bits = movemask(active); #if FORCE_SINGLE_MODE == 0 if (unlikely(popcnt(bits) <= switchThreshold)) #endif { for (; bits!=0; ) { const size_t i = bscf(bits); intersect1(This, bvh, cur, i, pre, ray, tray, context); } tray.tfar = min(tray.tfar, ray.tfar); continue; } } #endif while (likely(!cur.isLeaf())) { / process nodes / const vbool<K> valid_node = tray.tfar > curDist; STAT3(normal.trav_nodes, 1, popcnt(valid_node), K); const NodeRef nodeRef = cur; const BaseNode __restrict__ const node = nodeRef.baseNode(); /* set cur to invalid / cur = BVH::emptyNode; curDist = pos_inf; size_t num_child_hits = 0; for (unsigned i = 0; i < N; i++) { const NodeRef child = node->children[i]; if (unlikely(child == BVH::emptyNode)) break; vfloat<K> lnearP; vbool<K> lhit = valid_node; BVHNNodeIntersectorK<N, K, types, robust>::intersect(nodeRef, i, tray, ray.time(), lnearP, lhit); / if we hit the child we choose to continue with that child if it is closer than the current next child, or we push it onto the stack / if (likely(any(lhit))) { assert(sptr_node < stackEnd); assert(child != BVH::emptyNode); const vfloat<K> childDist = select(lhit, lnearP, inf); / push cur node onto stack and continue with hit child / if (any(childDist < curDist)) { if (likely(cur != BVH::emptyNode)) { num_child_hits++; sptr_node = cur; sptr_node++; sptr_near = curDist; sptr_near++; } curDist = childDist; cur = child; } / push hit child onto stack / else { num_child_hits++; sptr_node = child; sptr_node++; sptr_near = childDist; sptr_near++; } } } #if defined(__AVX__) //STAT3(normal.trav_hit_boxes[num_child_hits], 1, 1, 1); #endif if (unlikely(cur == BVH::emptyNode)) goto pop; / improved distance sorting for 3 or more hits / if (unlikely(num_child_hits >= 2)) { if (any(sptr_near[-2] < sptr_near[-1])) { std::swap(sptr_near[-2],sptr_near[-1]); std::swap(sptr_node[-2],sptr_node[-1]); } if (unlikely(num_child_hits >= 3)) { if (any(sptr_near[-3] < sptr_near[-1])) { std::swap(sptr_near[-3],sptr_near[-1]); std::swap(sptr_node[-3],sptr_node[-1]); } if (any(sptr_near[-3] < sptr_near[-2])) { std::swap(sptr_near[-3],sptr_near[-2]); std::swap(sptr_node[-3],sptr_node[-2]); } } } #if SWITCH_DURING_DOWN_TRAVERSAL == 1 if (single) { // seems to be the best place for testing utilization if (unlikely(popcnt(tray.tfar > curDist) <= switchThreshold)) { sptr_node++ = cur; sptr_near++ = curDist; goto pop; } } #endif } / return if stack is empty / if (unlikely(cur == BVH::invalidNode)) { assert(sptr_node == stack_node); break; } / intersect leaf / assert(cur != BVH::emptyNode); const vbool<K> valid_leaf = tray.tfar > curDist; STAT3(normal.trav_leaves, 1, popcnt(valid_leaf), K); if (unlikely(none(valid_leaf))) continue; size_t items; const Primitive prim = (Primitive)cur.leaf(items); size_t lazy_node = 0; PrimitiveIntersectorK::intersect(valid_leaf, This, pre, ray, context, prim, items, tray, lazy_node); tray.tfar = select(valid_leaf, ray.tfar, tray.tfar); if (unlikely(lazy_node)) { sptr_node = lazy_node; sptr_node++; *sptr_near = neg_inf; sptr_near++; } } } while(valid_bits); }	pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x200, %rsp # imm = 0x200 movq (%rsi), %r14 cmpq $0x8, 0x70(%r14) je 0x385464 movq %rcx, %r12 movq %rdx, %r15 movq %rsi, %rbx cmpq $0x0, 0x8(%rcx) je 0x385257 movq 0x10(%r12), %rax testb $0x1, 0x2(%rax) jne 0x385473 vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %xmm1 vpcmpeqd 0x10(%rdi), %xmm0, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpmovmskb %xmm0, %eax testl %eax, %eax je 0x385464 vpacksswb %xmm0, %xmm0, %xmm1 vpmovmskb %xmm1, %eax vmovaps (%r15), %ymm1 leaq 0x20(%rsp), %rcx vmovaps %ymm1, (%rcx) vmovaps 0x20(%r15), %ymm1 vmovaps %ymm1, 0x20(%rcx) vmovaps 0x40(%r15), %ymm1 vmovaps %ymm1, 0x40(%rcx) vmovaps 0x80(%r15), %ymm1 vmovaps %ymm1, 0x60(%rcx) vmovaps 0xa0(%r15), %ymm2 vmovaps %ymm2, 0x80(%rcx) vmovaps 0xc0(%r15), %ymm3 vbroadcastss 0x1b9bbf3(%rip), %ymm4 # 0x1f20ec4 vbroadcastss 0x1b6bd0e(%rip), %ymm5 # 0x1ef0fe8 vandps %ymm4, %ymm1, %ymm6 vcmpltps %ymm5, %ymm6, %ymm6 vblendvps %ymm6, %ymm5, %ymm1, %ymm1 vandps %ymm4, %ymm2, %ymm6 vcmpltps %ymm5, %ymm6, %ymm6 vblendvps %ymm6, %ymm5, %ymm2, %ymm2 vandps %ymm4, %ymm3, %ymm4 vcmpltps %ymm5, %ymm4, %ymm4 vblendvps %ymm4, %ymm5, %ymm3, %ymm4 movzbl %al, %r13d vrcpps %ymm1, %ymm5 vmovaps %ymm3, 0xa0(%rcx) vmulps %ymm5, %ymm1, %ymm1 vbroadcastss 0x1b673f0(%rip), %ymm3 # 0x1eec714 vsubps %ymm1, %ymm3, %ymm1 vrcpps %ymm2, %ymm6 vmulps %ymm1, %ymm5, %ymm1 vaddps %ymm1, %ymm5, %ymm1 vmulps %ymm6, %ymm2, %ymm2 vsubps %ymm2, %ymm3, %ymm2 vmulps %ymm2, %ymm6, %ymm2 vrcpps %ymm4, %ymm5 vaddps %ymm2, %ymm6, %ymm2 vmulps %ymm5, %ymm4, %ymm4 vsubps %ymm4, %ymm3, %ymm3 vmulps %ymm3, %ymm5, %ymm3 vaddps %ymm3, %ymm5, %ymm3 vmovaps %ymm1, 0xc0(%rcx) vmovaps %ymm2, 0xe0(%rcx) vmovaps %ymm3, 0x100(%rcx) vxorps %xmm4, %xmm4, %xmm4 vcmpltps %ymm4, %ymm1, %ymm1 vbroadcastss 0x1bcf61e(%rip), %ymm5 # 0x1f549a0 vandps %ymm5, %ymm1, %ymm1 vcmpnltps %ymm4, %ymm2, %ymm2 vbroadcastss 0x1bd55d8(%rip), %ymm5 # 0x1f5a96c vbroadcastss 0x1bd6063(%rip), %ymm6 # 0x1f5b400 vblendvps %ymm2, %ymm5, %ymm6, %ymm2 vmovaps %ymm1, 0x120(%rcx) vcmpnltps %ymm4, %ymm3, %ymm1 vbroadcastss 0x1bd604b(%rip), %ymm3 # 0x1f5b404 vbroadcastss 0x1bd6046(%rip), %ymm5 # 0x1f5b408 vblendvps %ymm1, %ymm3, %ymm5, %ymm1 vmovaps %ymm2, 0x140(%rcx) vmovaps %ymm1, 0x160(%rcx) vmovaps 0x60(%r15), %ymm1 vmovaps 0x100(%r15), %ymm2 vmaxps %ymm4, %ymm1, %ymm1 vpmovsxwd %xmm0, %xmm3 vpunpckhwd %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[4,4,5,5,6,6,7,7] vinsertf128 $0x1, %xmm0, %ymm3, %ymm0 vbroadcastss 0x1b6661d(%rip), %ymm3 # 0x1eeba20 vblendvps %ymm0, %ymm1, %ymm3, %ymm1 vmaxps %ymm4, %ymm2, %ymm2 vmovaps %ymm1, 0x180(%rcx) vbroadcastss 0x1b67766(%rip), %ymm1 # 0x1eecb84 vblendvps %ymm0, %ymm2, %ymm1, %ymm0 vmovaps %ymm0, 0x1a0(%rcx) bsfq %r13, %rcx leaq -0x1(%r13), %rax movq %rax, 0x18(%rsp) movq 0x70(%r14), %rdx movq %rbx, %rdi movq %r14, %rsi leaq 0x17(%rsp), %r8 movq %r15, %r9 pushq %r12 leaq 0x28(%rsp), %rax pushq %rax vzeroupper callq 0x3be6a4 popq %rax popq %rcx andq 0x18(%rsp), %r13 jne 0x38542c leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movq %rbx, %rsi movq %r15, %rdx movq %r12, %rcx callq 0x3bd0f2 jmp 0x385464 nop	/embree[P]embree/kernels/bvh/bvh_intersector_hybrid.cpp
embree::avx::InstanceArrayIntersector1MB::occluded(embree::avx::InstanceArrayIntersector1MB::Precalculations const&, embree::RayK<1>&, embree::RayQueryContext*, embree::InstanceArrayPrimitive const&)	bool InstanceArrayIntersector1MB::occluded(const Precalculations& pre, Ray& ray, RayQueryContext* context, const Primitive& prim) { const InstanceArray* instance = context->scene->get<InstanceArray>(prim.instID_); Accel* object = instance->getObject(prim.primID_); if (!object) return false; /* perform ray mask test / #if defined(EMBREE_RAY_MASK) if ((ray.mask & instance->mask) == 0) return false; #endif RTCRayQueryContext user_context = context->user; bool occluded = false; if (likely(instance_id_stack::push(user_context, prim.instID_, prim.primID_))) { const AffineSpace3fa world2local = instance->getWorld2Local(prim.primID_, ray.time()); const Vec3ff ray_org = ray.org; const Vec3ff ray_dir = ray.dir; ray.org = Vec3ff(xfmPoint(world2local, ray_org), ray.tnear()); ray.dir = Vec3ff(xfmVector(world2local, ray_dir), ray.time()); RayQueryContext newcontext((Scene*)object, user_context, context->args); object->intersectors.occluded((RTCRay&)ray, &newcontext); ray.org = ray_org; ray.dir = ray_dir; occluded = ray.tfar < 0.0f; instance_id_stack::pop(user_context); } return occluded; }	pushq %r15 pushq %r14 pushq %r12 pushq %rbx subq $0x178, %rsp # imm = 0x178 movq %rdx, %r14 movq %rsi, %rbx movq (%rdx), %rdx movl (%rcx), %eax movl 0x4(%rcx), %esi movq 0x1e8(%rdx), %rdx movq (%rdx,%rsi,8), %rdx movq 0x58(%rdx), %r15 testq %r15, %r15 jne 0x40912e movq 0x90(%rdx), %rdi movq 0xa0(%rdx), %r8 imulq %rax, %r8 movl (%rdi,%r8), %edi movl $0xffffffff, %r8d # imm = 0xFFFFFFFF cmpq %r8, %rdi je 0x40912b movq 0x60(%rdx), %r8 movq (%r8,%rdi,8), %r15 jmp 0x40912e xorl %r15d, %r15d testq %r15, %r15 je 0x4091d4 movl 0x34(%rdx), %edi testl %edi, 0x24(%rbx) je 0x4091d4 movq 0x8(%r14), %r12 cmpl $-0x1, (%r12) jne 0x4091d4 movl %esi, (%r12) movl %eax, 0x4(%r12) movl (%rcx), %eax cmpl $0x1, 0x24(%rdx) jne 0x4091db movzbl 0x3d(%rdx), %esi shll $0x8, %esi movq 0x88(%rdx), %rcx movl 0x20(%rcx), %edx cmpl $0x100, %esi # imm = 0x100 je 0x4099ba cmpl $0x9134, %edx # imm = 0x9134 je 0x409533 cmpl $0x9234, %edx # imm = 0x9234 je 0x4092a1 cmpl $0xb001, %edx # imm = 0xB001 je 0x40936c cmpl $0x9244, %edx # imm = 0x9244 jne 0x40984c movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovaps (%rdx,%rax), %xmm0 vmovaps 0x10(%rdx,%rax), %xmm4 vmovaps 0x20(%rdx,%rax), %xmm2 vmovaps 0x30(%rdx,%rax), %xmm1 jmp 0x40984c xorl %eax, %eax jmp 0x4099ab vmovss 0x1c(%rbx), %xmm0 vmovss 0x28(%rdx), %xmm1 vmovss 0x2c(%rdx), %xmm2 vmovss 0x30(%rdx), %xmm3 vsubss %xmm2, %xmm0, %xmm0 vsubss %xmm2, %xmm3, %xmm2 vdivss %xmm2, %xmm0, %xmm0 vmulss %xmm0, %xmm1, %xmm0 vroundss $0x9, %xmm0, %xmm0, %xmm2 vaddss 0x1ae77bf(%rip), %xmm1, %xmm1 # 0x1ef09cc vminss %xmm1, %xmm2, %xmm1 vxorps %xmm2, %xmm2, %xmm2 vmaxss %xmm1, %xmm2, %xmm1 vcvttss2si %xmm1, %ecx vsubss %xmm1, %xmm0, %xmm15 movzbl 0x3d(%rdx), %r8d shll $0x8, %r8d movq 0x88(%rdx), %rdx imulq $0x38, %rcx, %rdi leaq (%rdx,%rdi), %rsi movl 0x20(%rdx,%rdi), %edi cmpl $0x100, %r8d # imm = 0x100 je 0x409a0e cmpl $0x9134, %edi # imm = 0x9134 je 0x409597 cmpl $0x9234, %edi # imm = 0x9234 je 0x409305 cmpl $0xb001, %edi # imm = 0xB001 je 0x40944e cmpl $0x9244, %edi # imm = 0x9244 jne 0x4095f9 movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovaps (%rdi,%rsi), %xmm0 vmovaps 0x10(%rdi,%rsi), %xmm1 vmovaps 0x20(%rdi,%rsi), %xmm2 vmovaps 0x30(%rdi,%rsi), %xmm3 jmp 0x4095f9 movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovsd 0x4(%rdx,%rax), %xmm0 vmovss (%rdx,%rax), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,2,3,1] vmovsd 0x10(%rdx,%rax), %xmm1 vmovss 0xc(%rdx,%rax), %xmm2 vshufps $0x4c, %xmm1, %xmm2, %xmm1 # xmm1 = xmm2[0,3],xmm1[0,1] vshufps $0x78, %xmm1, %xmm1, %xmm4 # xmm4 = xmm1[0,2,3,1] vmovsd 0x1c(%rdx,%rax), %xmm1 vmovss 0x18(%rdx,%rax), %xmm2 vshufps $0x4c, %xmm1, %xmm2, %xmm1 # xmm1 = xmm2[0,3],xmm1[0,1] vshufps $0x78, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[0,2,3,1] vmovsd 0x28(%rdx,%rax), %xmm1 vmovss 0x24(%rdx,%rax), %xmm3 vshufps $0x4c, %xmm1, %xmm3, %xmm1 # xmm1 = xmm3[0,3],xmm1[0,1] vshufps $0x78, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,2,3,1] jmp 0x40984c movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovsd 0x4(%rdi,%rsi), %xmm0 vmovss (%rdi,%rsi), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,2,3,1] vmovsd 0x10(%rdi,%rsi), %xmm1 vmovss 0xc(%rdi,%rsi), %xmm2 vshufps $0x4c, %xmm1, %xmm2, %xmm1 # xmm1 = xmm2[0,3],xmm1[0,1] vshufps $0x78, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,2,3,1] vmovsd 0x1c(%rdi,%rsi), %xmm2 vmovss 0x18(%rdi,%rsi), %xmm3 vshufps $0x4c, %xmm2, %xmm3, %xmm2 # xmm2 = xmm3[0,3],xmm2[0,1] vshufps $0x78, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,2,3,1] vmovsd 0x28(%rdi,%rsi), %xmm3 vmovss 0x24(%rdi,%rsi), %xmm4 vshufps $0x4c, %xmm3, %xmm4, %xmm3 # xmm3 = xmm4[0,3],xmm3[0,1] vshufps $0x78, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[0,2,3,1] jmp 0x4095f9 movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovsd 0x10(%rdx,%rax), %xmm0 vinsertps $0x20, 0x8(%rdx,%rax), %xmm0, %xmm2 # xmm2 = xmm0[0,1],mem[0],xmm0[3] vmovsd 0x34(%rdx,%rax), %xmm0 vmovss (%rdx,%rax), %xmm1 vmovss 0xc(%rdx,%rax), %xmm3 vmovlhps %xmm0, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm0[0] vshufps $0xd8, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,2],xmm0[1,3] vmovss 0x18(%rdx,%rax), %xmm1 vmovsd 0x1c(%rdx,%rax), %xmm4 vmovlhps %xmm4, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm4[0] vshufps $0xd8, %xmm4, %xmm1, %xmm1 # xmm1 = xmm1[0,2],xmm4[1,3] vmovss 0x24(%rdx,%rax), %xmm4 vmovss 0x28(%rdx,%rax), %xmm5 vmovss 0x2c(%rdx,%rax), %xmm6 vmovss 0x30(%rdx,%rax), %xmm7 vmulss %xmm5, %xmm5, %xmm8 vmulss %xmm4, %xmm4, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm6, %xmm6, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm7, %xmm7, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vrsqrtss %xmm8, %xmm8, %xmm9 vmulss 0x1ae3323(%rip), %xmm9, %xmm10 # 0x1eec718 vmulss 0x1ae331f(%rip), %xmm8, %xmm8 # 0x1eec71c vmulss %xmm8, %xmm9, %xmm8 vmulss %xmm9, %xmm9, %xmm9 vmulss %xmm8, %xmm9, %xmm8 vaddss %xmm8, %xmm10, %xmm8 vmulss %xmm4, %xmm8, %xmm4 vinsertps $0x30, %xmm4, %xmm1, %xmm1 # xmm1 = xmm1[0,1,2],xmm4[0] vmulss %xmm5, %xmm8, %xmm4 vinsertps $0x30, %xmm4, %xmm0, %xmm0 # xmm0 = xmm0[0,1,2],xmm4[0] vmulss %xmm6, %xmm8, %xmm4 vmulss %xmm7, %xmm8, %xmm5 vinsertps $0x10, 0x4(%rdx,%rax), %xmm3, %xmm3 # xmm3 = xmm3[0],mem[0],xmm3[2,3] vinsertps $0x30, %xmm5, %xmm2, %xmm2 # xmm2 = xmm2[0,1,2],xmm5[0] vinsertps $0x20, 0x3c(%rdx,%rax), %xmm3, %xmm3 # xmm3 = xmm3[0,1],mem[0],xmm3[3] vinsertps $0x30, %xmm4, %xmm3, %xmm4 # xmm4 = xmm3[0,1,2],xmm4[0] jmp 0x40984c movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovsd 0x10(%rdi,%rsi), %xmm0 vinsertps $0x20, 0x8(%rdi,%rsi), %xmm0, %xmm1 # xmm1 = xmm0[0,1],mem[0],xmm0[3] vmovsd 0x34(%rdi,%rsi), %xmm0 vmovss (%rdi,%rsi), %xmm2 vmovss 0xc(%rdi,%rsi), %xmm4 vmovlhps %xmm0, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm0[0] vshufps $0xd8, %xmm0, %xmm2, %xmm0 # xmm0 = xmm2[0,2],xmm0[1,3] vmovss 0x18(%rdi,%rsi), %xmm2 vmovsd 0x1c(%rdi,%rsi), %xmm3 vmovlhps %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm3[0] vshufps $0xd8, %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[0,2],xmm3[1,3] vmovss 0x24(%rdi,%rsi), %xmm3 vmovss 0x28(%rdi,%rsi), %xmm5 vmovss 0x2c(%rdi,%rsi), %xmm6 vmovss 0x30(%rdi,%rsi), %xmm7 vmulss %xmm5, %xmm5, %xmm8 vmulss %xmm3, %xmm3, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm6, %xmm6, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm7, %xmm7, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vrsqrtss %xmm8, %xmm8, %xmm9 vmulss 0x1ae323e(%rip), %xmm9, %xmm10 # 0x1eec718 vmulss 0x1ae323a(%rip), %xmm8, %xmm8 # 0x1eec71c vmulss %xmm8, %xmm9, %xmm8 vmulss %xmm9, %xmm9, %xmm9 vmulss %xmm8, %xmm9, %xmm8 vaddss %xmm8, %xmm10, %xmm8 vmulss %xmm3, %xmm8, %xmm3 vinsertps $0x30, %xmm3, %xmm2, %xmm3 # xmm3 = xmm2[0,1,2],xmm3[0] vmulss %xmm5, %xmm8, %xmm2 vinsertps $0x30, %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[0,1,2],xmm2[0] vmulss %xmm6, %xmm8, %xmm5 vmulss %xmm7, %xmm8, %xmm2 vinsertps $0x10, 0x4(%rdi,%rsi), %xmm4, %xmm4 # xmm4 = xmm4[0],mem[0],xmm4[2,3] vinsertps $0x30, %xmm2, %xmm1, %xmm2 # xmm2 = xmm1[0,1,2],xmm2[0] vinsertps $0x20, 0x3c(%rdi,%rsi), %xmm4, %xmm1 # xmm1 = xmm4[0,1],mem[0],xmm4[3] vinsertps $0x30, %xmm5, %xmm1, %xmm1 # xmm1 = xmm1[0,1,2],xmm5[0] jmp 0x4095f9 movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovss (%rdx,%rax), %xmm0 vmovss 0x4(%rdx,%rax), %xmm1 vmovss 0x8(%rdx,%rax), %xmm2 vmovss 0xc(%rdx,%rax), %xmm3 vinsertps $0x1c, 0x10(%rdx,%rax), %xmm0, %xmm0 # xmm0 = xmm0[0],mem[0],zero,zero vinsertps $0x28, 0x20(%rdx,%rax), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],zero vinsertps $0x1c, 0x14(%rdx,%rax), %xmm1, %xmm1 # xmm1 = xmm1[0],mem[0],zero,zero vinsertps $0x28, 0x24(%rdx,%rax), %xmm1, %xmm4 # xmm4 = xmm1[0,1],mem[0],zero vinsertps $0x1c, 0x18(%rdx,%rax), %xmm2, %xmm1 # xmm1 = xmm2[0],mem[0],zero,zero vinsertps $0x28, 0x28(%rdx,%rax), %xmm1, %xmm2 # xmm2 = xmm1[0,1],mem[0],zero vinsertps $0x1c, 0x1c(%rdx,%rax), %xmm3, %xmm1 # xmm1 = xmm3[0],mem[0],zero,zero vinsertps $0x28, 0x2c(%rdx,%rax), %xmm1, %xmm1 # xmm1 = xmm1[0,1],mem[0],zero jmp 0x40984c movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovss (%rdi,%rsi), %xmm0 vmovss 0x4(%rdi,%rsi), %xmm1 vmovss 0x8(%rdi,%rsi), %xmm2 vmovss 0xc(%rdi,%rsi), %xmm3 vinsertps $0x1c, 0x10(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0],mem[0],zero,zero vinsertps $0x28, 0x20(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],zero vinsertps $0x1c, 0x14(%rdi,%rsi), %xmm1, %xmm1 # xmm1 = xmm1[0],mem[0],zero,zero vinsertps $0x28, 0x24(%rdi,%rsi), %xmm1, %xmm1 # xmm1 = xmm1[0,1],mem[0],zero vinsertps $0x1c, 0x18(%rdi,%rsi), %xmm2, %xmm2 # xmm2 = xmm2[0],mem[0],zero,zero vinsertps $0x28, 0x28(%rdi,%rsi), %xmm2, %xmm2 # xmm2 = xmm2[0,1],mem[0],zero vinsertps $0x1c, 0x1c(%rdi,%rsi), %xmm3, %xmm3 # xmm3 = xmm3[0],mem[0],zero,zero vinsertps $0x28, 0x2c(%rdi,%rsi), %xmm3, %xmm3 # xmm3 = xmm3[0,1],mem[0],zero incl %ecx imulq $0x38, %rcx, %rcx leaq (%rdx,%rcx), %rsi movl 0x20(%rdx,%rcx), %ecx cmpl $0x9134, %ecx # imm = 0x9134 je 0x4097a4 cmpl $0x9234, %ecx # imm = 0x9234 je 0x409657 cmpl $0xb001, %ecx # imm = 0xB001 je 0x4096bb cmpl $0x9244, %ecx # imm = 0x9244 jne 0x409803 movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovaps (%rcx,%rax), %xmm4 vmovaps 0x10(%rcx,%rax), %xmm6 vmovaps 0x20(%rcx,%rax), %xmm7 vmovaps 0x30(%rcx,%rax), %xmm5 jmp 0x409803 movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovsd 0x4(%rcx,%rax), %xmm4 vmovss (%rcx,%rax), %xmm5 vshufps $0x4c, %xmm4, %xmm5, %xmm4 # xmm4 = xmm5[0,3],xmm4[0,1] vshufps $0x78, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,2,3,1] vmovsd 0x10(%rcx,%rax), %xmm5 vmovss 0xc(%rcx,%rax), %xmm6 vshufps $0x4c, %xmm5, %xmm6, %xmm5 # xmm5 = xmm6[0,3],xmm5[0,1] vshufps $0x78, %xmm5, %xmm5, %xmm6 # xmm6 = xmm5[0,2,3,1] vmovsd 0x1c(%rcx,%rax), %xmm5 vmovss 0x18(%rcx,%rax), %xmm7 vshufps $0x4c, %xmm5, %xmm7, %xmm5 # xmm5 = xmm7[0,3],xmm5[0,1] vshufps $0x78, %xmm5, %xmm5, %xmm7 # xmm7 = xmm5[0,2,3,1] vmovsd 0x28(%rcx,%rax), %xmm5 vmovss 0x24(%rcx,%rax), %xmm8 vshufps $0x4c, %xmm5, %xmm8, %xmm5 # xmm5 = xmm8[0,3],xmm5[0,1] vshufps $0x78, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[0,2,3,1] jmp 0x409803 movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovsd 0x10(%rcx,%rax), %xmm4 vinsertps $0x20, 0x8(%rcx,%rax), %xmm4, %xmm6 # xmm6 = xmm4[0,1],mem[0],xmm4[3] vmovsd 0x34(%rcx,%rax), %xmm4 vmovss (%rcx,%rax), %xmm5 vmovss 0xc(%rcx,%rax), %xmm7 vmovlhps %xmm4, %xmm5, %xmm5 # xmm5 = xmm5[0],xmm4[0] vshufps $0xd8, %xmm4, %xmm5, %xmm4 # xmm4 = xmm5[0,2],xmm4[1,3] vmovss 0x18(%rcx,%rax), %xmm5 vmovsd 0x1c(%rcx,%rax), %xmm8 vmovlhps %xmm8, %xmm5, %xmm5 # xmm5 = xmm5[0],xmm8[0] vshufps $0xd8, %xmm8, %xmm5, %xmm5 # xmm5 = xmm5[0,2],xmm8[1,3] vmovss 0x24(%rcx,%rax), %xmm8 vmovss 0x28(%rcx,%rax), %xmm9 vmovss 0x2c(%rcx,%rax), %xmm10 vmovss 0x30(%rcx,%rax), %xmm11 vmulss %xmm9, %xmm9, %xmm12 vmulss %xmm8, %xmm8, %xmm13 vaddss %xmm12, %xmm13, %xmm12 vmulss %xmm10, %xmm10, %xmm13 vaddss %xmm12, %xmm13, %xmm12 vmulss %xmm11, %xmm11, %xmm13 vaddss %xmm12, %xmm13, %xmm12 vrsqrtss %xmm12, %xmm12, %xmm13 vmulss 0x1ae2fce(%rip), %xmm13, %xmm14 # 0x1eec718 vmulss 0x1ae2fca(%rip), %xmm12, %xmm12 # 0x1eec71c vmulss %xmm12, %xmm13, %xmm12 vmulss %xmm13, %xmm13, %xmm13 vmulss %xmm12, %xmm13, %xmm12 vaddss %xmm12, %xmm14, %xmm12 vmulss %xmm12, %xmm8, %xmm8 vinsertps $0x30, %xmm8, %xmm5, %xmm5 # xmm5 = xmm5[0,1,2],xmm8[0] vmulss %xmm12, %xmm9, %xmm8 vinsertps $0x30, %xmm8, %xmm4, %xmm4 # xmm4 = xmm4[0,1,2],xmm8[0] vmulss %xmm12, %xmm10, %xmm8 vmulss %xmm12, %xmm11, %xmm9 vinsertps $0x10, 0x4(%rcx,%rax), %xmm7, %xmm10 # xmm10 = xmm7[0],mem[0],xmm7[2,3] vinsertps $0x30, %xmm9, %xmm6, %xmm7 # xmm7 = xmm6[0,1,2],xmm9[0] vinsertps $0x20, 0x3c(%rcx,%rax), %xmm10, %xmm6 # xmm6 = xmm10[0,1],mem[0],xmm10[3] vinsertps $0x30, %xmm8, %xmm6, %xmm6 # xmm6 = xmm6[0,1,2],xmm8[0] jmp 0x409803 movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovss (%rcx,%rax), %xmm4 vmovss 0x4(%rcx,%rax), %xmm5 vmovss 0x8(%rcx,%rax), %xmm7 vmovss 0xc(%rcx,%rax), %xmm8 vinsertps $0x1c, 0x10(%rcx,%rax), %xmm4, %xmm4 # xmm4 = xmm4[0],mem[0],zero,zero vinsertps $0x28, 0x20(%rcx,%rax), %xmm4, %xmm4 # xmm4 = xmm4[0,1],mem[0],zero vinsertps $0x1c, 0x14(%rcx,%rax), %xmm5, %xmm5 # xmm5 = xmm5[0],mem[0],zero,zero vinsertps $0x28, 0x24(%rcx,%rax), %xmm5, %xmm6 # xmm6 = xmm5[0,1],mem[0],zero vinsertps $0x1c, 0x18(%rcx,%rax), %xmm7, %xmm5 # xmm5 = xmm7[0],mem[0],zero,zero vinsertps $0x28, 0x28(%rcx,%rax), %xmm5, %xmm7 # xmm7 = xmm5[0,1],mem[0],zero vinsertps $0x1c, 0x1c(%rcx,%rax), %xmm8, %xmm5 # xmm5 = xmm8[0],mem[0],zero,zero vinsertps $0x28, 0x2c(%rcx,%rax), %xmm5, %xmm5 # xmm5 = xmm5[0,1],mem[0],zero vmovss 0x1ae2f09(%rip), %xmm8 # 0x1eec714 vsubss %xmm15, %xmm8, %xmm8 vshufps $0x0, %xmm15, %xmm15, %xmm9 # xmm9 = xmm15[0,0,0,0] vmulps %xmm4, %xmm9, %xmm4 vshufps $0x0, %xmm8, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] vmulps %xmm0, %xmm8, %xmm0 vaddps %xmm4, %xmm0, %xmm0 vmulps %xmm6, %xmm9, %xmm4 vmulps %xmm1, %xmm8, %xmm1 vaddps %xmm4, %xmm1, %xmm4 vmulps %xmm7, %xmm9, %xmm1 vmulps %xmm2, %xmm8, %xmm2 vaddps %xmm1, %xmm2, %xmm2 vmulps %xmm5, %xmm9, %xmm1 vmulps %xmm3, %xmm8, %xmm3 vaddps %xmm1, %xmm3, %xmm1 vshufps $0xc9, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,2,0,3] vshufps $0xc9, %xmm4, %xmm4, %xmm5 # xmm5 = xmm4[1,2,0,3] vmulps %xmm5, %xmm2, %xmm6 vmulps %xmm3, %xmm4, %xmm7 vsubps %xmm6, %xmm7, %xmm6 vshufps $0xc9, %xmm6, %xmm6, %xmm7 # xmm7 = xmm6[1,2,0,3] vshufps $0xc9, %xmm0, %xmm0, %xmm8 # xmm8 = xmm0[1,2,0,3] vmulps %xmm0, %xmm3, %xmm3 vmulps %xmm2, %xmm8, %xmm2 vsubps %xmm3, %xmm2, %xmm2 vmulps %xmm4, %xmm8, %xmm3 vmulps %xmm5, %xmm0, %xmm4 vsubps %xmm3, %xmm4, %xmm3 vshufps $0xc9, %xmm3, %xmm3, %xmm4 # xmm4 = xmm3[1,2,0,3] vunpcklps %xmm4, %xmm7, %xmm4 # xmm4 = xmm7[0],xmm4[0],xmm7[1],xmm4[1] vunpcklps %xmm3, %xmm6, %xmm3 # xmm3 = xmm6[0],xmm3[0],xmm6[1],xmm3[1] vinsertps $0x4a, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[1],zero,xmm2[2],zero vxorps %xmm6, %xmm6, %xmm6 vmovss %xmm2, %xmm6, %xmm2 # xmm2 = xmm2[0],xmm6[1,2,3] vunpcklps %xmm2, %xmm3, %xmm2 # xmm2 = xmm3[0],xmm2[0],xmm3[1],xmm2[1] vunpcklps %xmm5, %xmm4, %xmm3 # xmm3 = xmm4[0],xmm5[0],xmm4[1],xmm5[1] vunpckhps %xmm5, %xmm4, %xmm4 # xmm4 = xmm4[2],xmm5[2],xmm4[3],xmm5[3] vdpps $0x7f, %xmm7, %xmm0, %xmm0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vdivps %xmm0, %xmm3, %xmm3 vdivps %xmm0, %xmm4, %xmm4 vdivps %xmm0, %xmm2, %xmm0 vshufps $0x0, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[0,0,0,0] vshufps $0x55, %xmm1, %xmm1, %xmm5 # xmm5 = xmm1[1,1,1,1] vshufps $0xaa, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vmulps %xmm0, %xmm1, %xmm1 vmulps %xmm4, %xmm5, %xmm5 vaddps %xmm1, %xmm5, %xmm1 vmulps %xmm3, %xmm2, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vmovaps (%rbx), %xmm2 vmovaps %xmm2, 0x20(%rsp) vbroadcastss (%rbx), %xmm2 vbroadcastss 0x4(%rbx), %xmm5 vbroadcastss 0x8(%rbx), %xmm6 vmulps %xmm0, %xmm6, %xmm6 vsubps %xmm1, %xmm6, %xmm1 vmulps %xmm4, %xmm5, %xmm5 vaddps %xmm1, %xmm5, %xmm1 vmulps %xmm3, %xmm2, %xmm2 vaddps %xmm1, %xmm2, %xmm1 vinsertps $0x30, 0xc(%rbx), %xmm1, %xmm1 # xmm1 = xmm1[0,1,2],mem[0] vmovaps 0x10(%rbx), %xmm2 vmovaps %xmm2, 0x10(%rsp) vmovaps %xmm1, (%rbx) vbroadcastss 0x10(%rbx), %xmm1 vbroadcastss 0x14(%rbx), %xmm2 vbroadcastss 0x18(%rbx), %xmm5 vmulps %xmm0, %xmm5, %xmm0 vmulps %xmm4, %xmm2, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vmulps %xmm3, %xmm1, %xmm1 vaddps %xmm0, %xmm1, %xmm0 vinsertps $0x30, 0x1c(%rbx), %xmm0, %xmm0 # xmm0 = xmm0[0,1,2],mem[0] vmovaps %xmm0, 0x10(%rbx) movq 0x10(%r14), %rax leaq 0x160(%rsp), %rdx movq %r15, (%rdx) movq %r12, 0x8(%rdx) movq %rax, 0x10(%rdx) leaq 0x58(%r15), %rdi movq %rbx, %rsi callq *0x80(%r15) vmovaps 0x20(%rsp), %xmm0 vmovaps %xmm0, (%rbx) vmovaps 0x10(%rsp), %xmm0 vmovaps %xmm0, 0x10(%rbx) vxorps %xmm0, %xmm0, %xmm0 vucomiss 0x20(%rbx), %xmm0 seta %al orq $-0x1, (%r12) addq $0x178, %rsp # imm = 0x178 popq %rbx popq %r12 popq %r14 popq %r15 retq cmpl $0x9134, %edx # imm = 0x9134 je 0x409d39 cmpl $0x9234, %edx # imm = 0x9234 je 0x409a7d cmpl $0xb001, %edx # imm = 0xB001 je 0x409b5a cmpl $0x9244, %edx # imm = 0x9244 jne 0x409d98 movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovaps (%rdx,%rax), %xmm9 vmovaps 0x10(%rdx,%rax), %xmm6 vmovaps 0x20(%rdx,%rax), %xmm2 vmovaps 0x30(%rdx,%rax), %xmm0 jmp 0x409d98 vmovaps %xmm15, 0x20(%rsp) cmpl $0x9134, %edi # imm = 0x9134 je 0x409f75 cmpl $0x9234, %edi # imm = 0x9234 je 0x409ae1 cmpl $0xb001, %edi # imm = 0xB001 je 0x409c3c cmpl $0x9244, %edi # imm = 0x9244 jne 0x409fef movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovaps (%rdi,%rsi), %xmm0 vmovaps %xmm0, 0x70(%rsp) vmovaps 0x10(%rdi,%rsi), %xmm0 vmovaps %xmm0, 0x60(%rsp) vmovaps 0x20(%rdi,%rsi), %xmm0 vmovaps %xmm0, 0x40(%rsp) vmovaps 0x30(%rdi,%rsi), %xmm0 jmp 0x409fe9 movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovsd 0x4(%rdx,%rax), %xmm0 vmovss (%rdx,%rax), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm9 # xmm9 = xmm0[0,2,3,1] vmovsd 0x10(%rdx,%rax), %xmm1 vmovss 0xc(%rdx,%rax), %xmm2 vshufps $0x4c, %xmm1, %xmm2, %xmm1 # xmm1 = xmm2[0,3],xmm1[0,1] vshufps $0x78, %xmm1, %xmm1, %xmm6 # xmm6 = xmm1[0,2,3,1] vmovsd 0x1c(%rdx,%rax), %xmm1 vmovss 0x18(%rdx,%rax), %xmm2 vshufps $0x4c, %xmm1, %xmm2, %xmm1 # xmm1 = xmm2[0,3],xmm1[0,1] vshufps $0x78, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[0,2,3,1] vmovsd 0x28(%rdx,%rax), %xmm1 vmovss 0x24(%rdx,%rax), %xmm4 vshufps $0x4c, %xmm1, %xmm4, %xmm1 # xmm1 = xmm4[0,3],xmm1[0,1] vshufps $0x78, %xmm1, %xmm1, %xmm0 # xmm0 = xmm1[0,2,3,1] jmp 0x409d98 movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovsd 0x4(%rdi,%rsi), %xmm0 vmovss (%rdi,%rsi), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,2,3,1] vmovaps %xmm0, 0x70(%rsp) vmovsd 0x10(%rdi,%rsi), %xmm0 vmovss 0xc(%rdi,%rsi), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,2,3,1] vmovaps %xmm0, 0x60(%rsp) vmovsd 0x1c(%rdi,%rsi), %xmm0 vmovss 0x18(%rdi,%rsi), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,2,3,1] vmovaps %xmm0, 0x40(%rsp) vmovsd 0x28(%rdi,%rsi), %xmm0 vmovss 0x24(%rdi,%rsi), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,2,3,1] jmp 0x409fe9 movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovsd 0x10(%rdx,%rax), %xmm0 vinsertps $0x20, 0x8(%rdx,%rax), %xmm0, %xmm2 # xmm2 = xmm0[0,1],mem[0],xmm0[3] vmovsd 0x34(%rdx,%rax), %xmm0 vmovss (%rdx,%rax), %xmm1 vmovss 0xc(%rdx,%rax), %xmm3 vmovlhps %xmm0, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm0[0] vshufps $0xd8, %xmm0, %xmm1, %xmm11 # xmm11 = xmm1[0,2],xmm0[1,3] vmovss 0x18(%rdx,%rax), %xmm1 vmovsd 0x1c(%rdx,%rax), %xmm4 vmovlhps %xmm4, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm4[0] vshufps $0xd8, %xmm4, %xmm1, %xmm1 # xmm1 = xmm1[0,2],xmm4[1,3] vmovss 0x24(%rdx,%rax), %xmm4 vmovss 0x28(%rdx,%rax), %xmm5 vmovss 0x2c(%rdx,%rax), %xmm6 vmovss 0x30(%rdx,%rax), %xmm7 vmulss %xmm5, %xmm5, %xmm8 vmulss %xmm4, %xmm4, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm6, %xmm6, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm7, %xmm7, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vrsqrtss %xmm8, %xmm8, %xmm9 vmulss 0x1ae2b35(%rip), %xmm9, %xmm10 # 0x1eec718 vmulss 0x1ae2b31(%rip), %xmm8, %xmm8 # 0x1eec71c vmulss %xmm8, %xmm9, %xmm8 vmulss %xmm9, %xmm9, %xmm9 vmulss %xmm8, %xmm9, %xmm8 vaddss %xmm8, %xmm10, %xmm8 vmulss %xmm4, %xmm8, %xmm4 vinsertps $0x30, %xmm4, %xmm1, %xmm0 # xmm0 = xmm1[0,1,2],xmm4[0] vmulss %xmm5, %xmm8, %xmm4 vinsertps $0x30, %xmm4, %xmm11, %xmm9 # xmm9 = xmm11[0,1,2],xmm4[0] vmulss %xmm6, %xmm8, %xmm4 vmulss %xmm7, %xmm8, %xmm5 vinsertps $0x10, 0x4(%rdx,%rax), %xmm3, %xmm3 # xmm3 = xmm3[0],mem[0],xmm3[2,3] vinsertps $0x30, %xmm5, %xmm2, %xmm2 # xmm2 = xmm2[0,1,2],xmm5[0] vinsertps $0x20, 0x3c(%rdx,%rax), %xmm3, %xmm3 # xmm3 = xmm3[0,1],mem[0],xmm3[3] vinsertps $0x30, %xmm4, %xmm3, %xmm6 # xmm6 = xmm3[0,1,2],xmm4[0] jmp 0x409d98 movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovsd 0x10(%rdi,%rsi), %xmm0 vinsertps $0x20, 0x8(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],xmm0[3] vmovsd 0x34(%rdi,%rsi), %xmm1 vmovss (%rdi,%rsi), %xmm2 vmovss 0xc(%rdi,%rsi), %xmm3 vmovlhps %xmm1, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm1[0] vshufps $0xd8, %xmm1, %xmm2, %xmm1 # xmm1 = xmm2[0,2],xmm1[1,3] vmovss 0x18(%rdi,%rsi), %xmm2 vmovsd 0x1c(%rdi,%rsi), %xmm4 vmovlhps %xmm4, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm4[0] vshufps $0xd8, %xmm4, %xmm2, %xmm2 # xmm2 = xmm2[0,2],xmm4[1,3] vmovss 0x24(%rdi,%rsi), %xmm4 vmovss 0x28(%rdi,%rsi), %xmm5 vmovss 0x2c(%rdi,%rsi), %xmm6 vmovss 0x30(%rdi,%rsi), %xmm7 vmulss %xmm5, %xmm5, %xmm8 vmulss %xmm4, %xmm4, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm6, %xmm6, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm7, %xmm7, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vrsqrtss %xmm8, %xmm8, %xmm9 vmulss 0x1ae2a50(%rip), %xmm9, %xmm10 # 0x1eec718 vmulss 0x1ae2a4c(%rip), %xmm8, %xmm8 # 0x1eec71c vmulss %xmm8, %xmm9, %xmm8 vmulss %xmm9, %xmm9, %xmm9 vmulss %xmm8, %xmm9, %xmm8 vaddss %xmm8, %xmm10, %xmm8 vmulss %xmm4, %xmm8, %xmm4 vinsertps $0x30, %xmm4, %xmm2, %xmm2 # xmm2 = xmm2[0,1,2],xmm4[0] vmovaps %xmm2, 0x50(%rsp) vmulss %xmm5, %xmm8, %xmm2 vinsertps $0x30, %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0,1,2],xmm2[0] vmovaps %xmm1, 0x70(%rsp) vmulss %xmm6, %xmm8, %xmm1 vmulss %xmm7, %xmm8, %xmm2 vinsertps $0x10, 0x4(%rdi,%rsi), %xmm3, %xmm3 # xmm3 = xmm3[0],mem[0],xmm3[2,3] vinsertps $0x30, %xmm2, %xmm0, %xmm0 # xmm0 = xmm0[0,1,2],xmm2[0] vmovaps %xmm0, 0x40(%rsp) vinsertps $0x20, 0x3c(%rdi,%rsi), %xmm3, %xmm0 # xmm0 = xmm3[0,1],mem[0],xmm3[3] vinsertps $0x30, %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[0,1,2],xmm1[0] vmovaps %xmm0, 0x60(%rsp) jmp 0x409fef movq (%rcx), %rdx imulq 0x10(%rcx), %rax vmovss (%rdx,%rax), %xmm0 vmovss 0x4(%rdx,%rax), %xmm1 vmovss 0x8(%rdx,%rax), %xmm2 vmovss 0xc(%rdx,%rax), %xmm4 vinsertps $0x1c, 0x10(%rdx,%rax), %xmm0, %xmm0 # xmm0 = xmm0[0],mem[0],zero,zero vinsertps $0x28, 0x20(%rdx,%rax), %xmm0, %xmm9 # xmm9 = xmm0[0,1],mem[0],zero vinsertps $0x1c, 0x14(%rdx,%rax), %xmm1, %xmm1 # xmm1 = xmm1[0],mem[0],zero,zero vinsertps $0x28, 0x24(%rdx,%rax), %xmm1, %xmm6 # xmm6 = xmm1[0,1],mem[0],zero vinsertps $0x1c, 0x18(%rdx,%rax), %xmm2, %xmm1 # xmm1 = xmm2[0],mem[0],zero,zero vinsertps $0x28, 0x28(%rdx,%rax), %xmm1, %xmm2 # xmm2 = xmm1[0,1],mem[0],zero vinsertps $0x1c, 0x1c(%rdx,%rax), %xmm4, %xmm1 # xmm1 = xmm4[0],mem[0],zero,zero vinsertps $0x28, 0x2c(%rdx,%rax), %xmm1, %xmm0 # xmm0 = xmm1[0,1],mem[0],zero vshufps $0xff, %xmm0, %xmm0, %xmm5 # xmm5 = xmm0[3,3,3,3] vmovaps %xmm9, %xmm3 vshufps $0xff, %xmm9, %xmm9, %xmm9 # xmm9 = xmm9[3,3,3,3] vshufps $0xff, %xmm6, %xmm6, %xmm7 # xmm7 = xmm6[3,3,3,3] vmovaps %xmm2, %xmm1 vshufps $0xff, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[3,3,3,3] vmulss %xmm9, %xmm9, %xmm10 vmulss %xmm5, %xmm5, %xmm11 vmovaps %xmm6, %xmm2 vaddss %xmm10, %xmm11, %xmm6 vbroadcastss 0x1b170f0(%rip), %xmm12 # 0x1f20ec0 vxorps %xmm7, %xmm12, %xmm8 vmulss %xmm7, %xmm8, %xmm8 vaddss %xmm6, %xmm8, %xmm6 vxorps %xmm4, %xmm12, %xmm12 vmulss %xmm4, %xmm12, %xmm12 vaddss %xmm6, %xmm12, %xmm13 vmulss %xmm5, %xmm4, %xmm6 vmulss %xmm7, %xmm9, %xmm14 vaddss %xmm6, %xmm14, %xmm15 vsubss %xmm6, %xmm14, %xmm6 vmovss %xmm6, 0x20(%rsp) vmulss %xmm4, %xmm9, %xmm14 vsubss %xmm10, %xmm11, %xmm10 vmulss %xmm7, %xmm7, %xmm11 vaddss %xmm10, %xmm11, %xmm11 vaddss %xmm11, %xmm12, %xmm6 vmulss %xmm5, %xmm7, %xmm12 vmulss %xmm5, %xmm9, %xmm5 vmovaps %xmm0, %xmm11 vsubss %xmm12, %xmm14, %xmm0 vmulss %xmm4, %xmm7, %xmm9 vaddss %xmm12, %xmm14, %xmm7 vaddss %xmm5, %xmm9, %xmm12 vsubss %xmm5, %xmm9, %xmm9 vaddss %xmm15, %xmm15, %xmm5 vaddss %xmm0, %xmm0, %xmm0 vaddss %xmm10, %xmm8, %xmm8 vmulss %xmm4, %xmm4, %xmm4 vaddss %xmm4, %xmm8, %xmm4 vshufps $0x0, %xmm13, %xmm13, %xmm8 # xmm8 = xmm13[0,0,0,0] vshufps $0x0, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmovaps 0x1ae289b(%rip), %xmm10 # 0x1eec700 vmulps %xmm0, %xmm10, %xmm0 vmovsd 0x1ae287f(%rip), %xmm13 # 0x1eec6f0 vmulps %xmm5, %xmm13, %xmm5 vaddps %xmm0, %xmm5, %xmm0 vmovss 0x1ae2893(%rip), %xmm14 # 0x1eec714 vmulps %xmm14, %xmm8, %xmm5 vaddps %xmm0, %xmm5, %xmm5 vaddss %xmm12, %xmm12, %xmm0 vshufps $0x0, %xmm6, %xmm6, %xmm8 # xmm8 = xmm6[0,0,0,0] vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmulps %xmm0, %xmm10, %xmm0 vmulps %xmm13, %xmm8, %xmm8 vaddps %xmm0, %xmm8, %xmm0 vmovss 0x20(%rsp), %xmm6 vaddss %xmm6, %xmm6, %xmm6 vshufps $0x0, %xmm6, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmulps %xmm6, %xmm14, %xmm6 vaddps %xmm0, %xmm6, %xmm6 vshufps $0x0, %xmm4, %xmm4, %xmm0 # xmm0 = xmm4[0,0,0,0] vmulps %xmm0, %xmm10, %xmm0 vaddss %xmm9, %xmm9, %xmm4 vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmulps %xmm4, %xmm13, %xmm4 vaddps %xmm0, %xmm4, %xmm0 vxorps %xmm4, %xmm4, %xmm4 vaddss %xmm7, %xmm7, %xmm7 vshufps $0x0, %xmm7, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmulps %xmm7, %xmm14, %xmm7 vshufps $0xe9, %xmm4, %xmm3, %xmm8 # xmm8 = xmm3[1,2],xmm4[2,3] vblendps $0x4, %xmm2, %xmm8, %xmm8 # xmm8 = xmm8[0,1],xmm2[2],xmm8[3] vaddps %xmm0, %xmm7, %xmm7 vaddps %xmm4, %xmm8, %xmm8 vshufps $0x0, %xmm3, %xmm3, %xmm0 # xmm0 = xmm3[0,0,0,0] vmulps %xmm4, %xmm7, %xmm9 vmulps %xmm4, %xmm6, %xmm4 vaddps %xmm4, %xmm9, %xmm4 vmulps %xmm5, %xmm0, %xmm0 vaddps %xmm4, %xmm0, %xmm0 vshufps $0x0, %xmm2, %xmm2, %xmm4 # xmm4 = xmm2[0,0,0,0] vshufps $0x55, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,1,1,1] vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm3, %xmm9, %xmm3 vmulps %xmm5, %xmm4, %xmm4 vaddps %xmm3, %xmm4, %xmm4 vshufps $0x55, %xmm1, %xmm1, %xmm3 # xmm3 = xmm1[1,1,1,1] vshufps $0xaa, %xmm1, %xmm1, %xmm9 # xmm9 = xmm1[2,2,2,2] vmulps %xmm7, %xmm9, %xmm9 vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm3, %xmm9, %xmm3 vshufps $0x0, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[0,0,0,0] vmulps %xmm5, %xmm2, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vshufps $0xaa, %xmm11, %xmm11, %xmm3 # xmm3 = xmm11[2,2,2,2] vmulps %xmm7, %xmm3, %xmm3 vshufps $0x55, %xmm11, %xmm11, %xmm7 # xmm7 = xmm11[1,1,1,1] vmulps %xmm6, %xmm7, %xmm6 vaddps %xmm3, %xmm6, %xmm3 vshufps $0x0, %xmm11, %xmm11, %xmm1 # xmm1 = xmm11[0,0,0,0] jmp 0x40a8fb movq (%rsi), %rdi movq 0x10(%rsi), %rsi imulq %rax, %rsi vmovss (%rdi,%rsi), %xmm0 vmovss 0x4(%rdi,%rsi), %xmm1 vmovss 0x8(%rdi,%rsi), %xmm2 vmovss 0xc(%rdi,%rsi), %xmm3 vinsertps $0x1c, 0x10(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0],mem[0],zero,zero vinsertps $0x28, 0x20(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],zero vmovaps %xmm0, 0x70(%rsp) vinsertps $0x1c, 0x14(%rdi,%rsi), %xmm1, %xmm0 # xmm0 = xmm1[0],mem[0],zero,zero vinsertps $0x28, 0x24(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],zero vmovaps %xmm0, 0x60(%rsp) vinsertps $0x1c, 0x18(%rdi,%rsi), %xmm2, %xmm0 # xmm0 = xmm2[0],mem[0],zero,zero vinsertps $0x28, 0x28(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],zero vmovaps %xmm0, 0x40(%rsp) vinsertps $0x1c, 0x1c(%rdi,%rsi), %xmm3, %xmm0 # xmm0 = xmm3[0],mem[0],zero,zero vinsertps $0x28, 0x2c(%rdi,%rsi), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],zero vmovaps %xmm0, 0x50(%rsp) incl %ecx imulq $0x38, %rcx, %rcx leaq (%rdx,%rcx), %rsi movl 0x20(%rdx,%rcx), %ecx cmpl $0x9134, %ecx # imm = 0x9134 je 0x40a190 cmpl $0x9234, %ecx # imm = 0x9234 je 0x40a04d cmpl $0xb001, %ecx # imm = 0xB001 je 0x40a0b1 cmpl $0x9244, %ecx # imm = 0x9244 jne 0x40a1ef movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovaps (%rcx,%rax), %xmm4 vmovaps 0x10(%rcx,%rax), %xmm5 vmovaps 0x20(%rcx,%rax), %xmm2 vmovaps 0x30(%rcx,%rax), %xmm9 jmp 0x40a1ef movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovsd 0x4(%rcx,%rax), %xmm0 vmovss (%rcx,%rax), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm4 # xmm4 = xmm0[0,2,3,1] vmovsd 0x10(%rcx,%rax), %xmm0 vmovss 0xc(%rcx,%rax), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm5 # xmm5 = xmm0[0,2,3,1] vmovsd 0x1c(%rcx,%rax), %xmm0 vmovss 0x18(%rcx,%rax), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm2 # xmm2 = xmm0[0,2,3,1] vmovsd 0x28(%rcx,%rax), %xmm0 vmovss 0x24(%rcx,%rax), %xmm1 vshufps $0x4c, %xmm0, %xmm1, %xmm0 # xmm0 = xmm1[0,3],xmm0[0,1] vshufps $0x78, %xmm0, %xmm0, %xmm9 # xmm9 = xmm0[0,2,3,1] jmp 0x40a1ef movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovsd 0x10(%rcx,%rax), %xmm0 vinsertps $0x20, 0x8(%rcx,%rax), %xmm0, %xmm0 # xmm0 = xmm0[0,1],mem[0],xmm0[3] vmovsd 0x34(%rcx,%rax), %xmm1 vmovss (%rcx,%rax), %xmm2 vmovss 0xc(%rcx,%rax), %xmm3 vmovlhps %xmm1, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm1[0] vshufps $0xd8, %xmm1, %xmm2, %xmm1 # xmm1 = xmm2[0,2],xmm1[1,3] vmovss 0x18(%rcx,%rax), %xmm2 vmovsd 0x1c(%rcx,%rax), %xmm4 vmovlhps %xmm4, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm4[0] vshufps $0xd8, %xmm4, %xmm2, %xmm2 # xmm2 = xmm2[0,2],xmm4[1,3] vmovss 0x24(%rcx,%rax), %xmm4 vmovss 0x28(%rcx,%rax), %xmm5 vmovss 0x2c(%rcx,%rax), %xmm6 vmovss 0x30(%rcx,%rax), %xmm7 vmulss %xmm5, %xmm5, %xmm8 vmulss %xmm4, %xmm4, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm6, %xmm6, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vmulss %xmm7, %xmm7, %xmm9 vaddss %xmm8, %xmm9, %xmm8 vrsqrtss %xmm8, %xmm8, %xmm9 vmulss 0x1ae25de(%rip), %xmm9, %xmm10 # 0x1eec718 vmulss 0x1ae25da(%rip), %xmm8, %xmm8 # 0x1eec71c vmulss %xmm8, %xmm9, %xmm8 vmulss %xmm9, %xmm9, %xmm9 vmulss %xmm8, %xmm9, %xmm8 vaddss %xmm8, %xmm10, %xmm8 vmulss %xmm4, %xmm8, %xmm4 vinsertps $0x30, %xmm4, %xmm2, %xmm9 # xmm9 = xmm2[0,1,2],xmm4[0] vmulss %xmm5, %xmm8, %xmm2 vinsertps $0x30, %xmm2, %xmm1, %xmm4 # xmm4 = xmm1[0,1,2],xmm2[0] vmulss %xmm6, %xmm8, %xmm1 vmulss %xmm7, %xmm8, %xmm2 vinsertps $0x10, 0x4(%rcx,%rax), %xmm3, %xmm3 # xmm3 = xmm3[0],mem[0],xmm3[2,3] vinsertps $0x30, %xmm2, %xmm0, %xmm2 # xmm2 = xmm0[0,1,2],xmm2[0] vinsertps $0x20, 0x3c(%rcx,%rax), %xmm3, %xmm0 # xmm0 = xmm3[0,1],mem[0],xmm3[3] vinsertps $0x30, %xmm1, %xmm0, %xmm5 # xmm5 = xmm0[0,1,2],xmm1[0] jmp 0x40a1ef movq (%rsi), %rcx imulq 0x10(%rsi), %rax vmovss (%rcx,%rax), %xmm0 vmovss 0x4(%rcx,%rax), %xmm1 vmovss 0x8(%rcx,%rax), %xmm2 vmovss 0xc(%rcx,%rax), %xmm3 vinsertps $0x1c, 0x10(%rcx,%rax), %xmm0, %xmm0 # xmm0 = xmm0[0],mem[0],zero,zero vinsertps $0x28, 0x20(%rcx,%rax), %xmm0, %xmm4 # xmm4 = xmm0[0,1],mem[0],zero vinsertps $0x1c, 0x14(%rcx,%rax), %xmm1, %xmm0 # xmm0 = xmm1[0],mem[0],zero,zero vinsertps $0x28, 0x24(%rcx,%rax), %xmm0, %xmm5 # xmm5 = xmm0[0,1],mem[0],zero vinsertps $0x1c, 0x18(%rcx,%rax), %xmm2, %xmm0 # xmm0 = xmm2[0],mem[0],zero,zero vinsertps $0x28, 0x28(%rcx,%rax), %xmm0, %xmm2 # xmm2 = xmm0[0,1],mem[0],zero vinsertps $0x1c, 0x1c(%rcx,%rax), %xmm3, %xmm0 # xmm0 = xmm3[0],mem[0],zero,zero vinsertps $0x28, 0x2c(%rcx,%rax), %xmm0, %xmm9 # xmm9 = xmm0[0,1],mem[0],zero vpermilps $0xff, 0x50(%rsp), %xmm7 # xmm7 = mem[3,3,3,3] vpermilps $0xff, 0x70(%rsp), %xmm1 # xmm1 = mem[3,3,3,3] vpermilps $0xff, 0x60(%rsp), %xmm6 # xmm6 = mem[3,3,3,3] vpermilps $0xff, 0x40(%rsp), %xmm3 # xmm3 = mem[3,3,3,3] vmovaps %xmm9, 0x130(%rsp) vshufps $0xff, %xmm9, %xmm9, %xmm8 # xmm8 = xmm9[3,3,3,3] vmovaps %xmm4, 0x150(%rsp) vshufps $0xff, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[3,3,3,3] vmovaps %xmm5, 0x140(%rsp) vshufps $0xff, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[3,3,3,3] vmovaps %xmm2, 0x120(%rsp) vshufps $0xff, %xmm2, %xmm2, %xmm0 # xmm0 = xmm2[3,3,3,3] vmovaps %xmm1, 0xe0(%rsp) vmulss %xmm4, %xmm1, %xmm1 vmovaps %xmm7, 0x110(%rsp) vmulss %xmm7, %xmm8, %xmm2 vaddss %xmm1, %xmm2, %xmm1 vmovaps %xmm6, 0xf0(%rsp) vmovaps %xmm5, 0x90(%rsp) vmulss %xmm5, %xmm6, %xmm2 vaddss %xmm1, %xmm2, %xmm1 vmovaps %xmm3, 0x100(%rsp) vmulss %xmm0, %xmm3, %xmm2 vaddss %xmm1, %xmm2, %xmm2 vbroadcastss 0x1b16c26(%rip), %xmm1 # 0x1f20ec0 vxorps %xmm1, %xmm2, %xmm3 vmovaps %xmm2, 0x10(%rsp) vmovaps %xmm3, 0x80(%rsp) vucomiss %xmm2, %xmm3 vxorps %xmm1, %xmm0, %xmm1 vmovaps %xmm1, 0xd0(%rsp) ja 0x40a2cb vmovaps %xmm8, 0xb0(%rsp) jmp 0x40a2e5 vbroadcastss 0x1b16bec(%rip), %xmm1 # 0x1f20ec0 vxorps %xmm1, %xmm8, %xmm8 vmovaps %xmm8, 0xb0(%rsp) vxorps %xmm1, %xmm4, %xmm4 vmovaps %xmm4, 0xc0(%rsp) ja 0x40a2f9 vmovaps %xmm0, 0xd0(%rsp) vbroadcastss 0x1b16bc2(%rip), %xmm0 # 0x1f20ec4 vmovaps 0x10(%rsp), %xmm1 vandps %xmm0, %xmm1, %xmm3 vmulss 0x1ae666c(%rip), %xmm3, %xmm0 # 0x1ef0980 vaddss 0x1ae6668(%rip), %xmm0, %xmm0 # 0x1ef0984 vmulss %xmm0, %xmm3, %xmm0 vaddss 0x1ae6660(%rip), %xmm0, %xmm0 # 0x1ef0988 vmulss %xmm0, %xmm3, %xmm0 vaddss 0x1ae6658(%rip), %xmm0, %xmm0 # 0x1ef098c vmovaps 0x80(%rsp), %xmm2 vmaxss %xmm1, %xmm2, %xmm1 vmovss %xmm1, 0xc(%rsp) vmulss %xmm0, %xmm3, %xmm0 vaddss 0x1ae663d(%rip), %xmm0, %xmm0 # 0x1ef0990 vmulss %xmm0, %xmm3, %xmm0 vaddss 0x1ae6635(%rip), %xmm0, %xmm0 # 0x1ef0994 vmovss %xmm0, 0x3c(%rsp) vmovss 0x1ae23a7(%rip), %xmm0 # 0x1eec714 vmovaps %xmm3, 0xa0(%rsp) vsubss %xmm3, %xmm0, %xmm0 vxorps %xmm1, %xmm1, %xmm1 vucomiss %xmm1, %xmm0 jb 0x40a38a vsqrtss %xmm0, %xmm0, %xmm0 jmp 0x40a38f callq 0x6aa20 vmulss 0x3c(%rsp), %xmm0, %xmm0 vmovss 0x1ae65fb(%rip), %xmm1 # 0x1ef0998 vsubss %xmm0, %xmm1, %xmm0 vxorps %xmm5, %xmm5, %xmm5 vmaxss %xmm0, %xmm5, %xmm0 vbroadcastss 0x1b16b0e(%rip), %xmm4 # 0x1f20ec0 vxorps %xmm4, %xmm0, %xmm2 vmovss 0xc(%rsp), %xmm3 vcmpltss %xmm5, %xmm3, %xmm3 vblendvps %xmm3, %xmm2, %xmm0, %xmm0 vsubss %xmm0, %xmm1, %xmm0 vmovss 0x1ae2341(%rip), %xmm7 # 0x1eec714 vcmpltss 0xa0(%rsp), %xmm7, %xmm2 vbroadcastss 0x1ae65b6(%rip), %xmm3 # 0x1ef099c vblendvps %xmm2, %xmm3, %xmm0, %xmm0 vmulss 0x20(%rsp), %xmm0, %xmm0 vmulss 0x1ae65a6(%rip), %xmm0, %xmm2 # 0x1ef09a0 vroundss $0x9, %xmm2, %xmm2, %xmm2 vcvttss2si %xmm2, %eax vmulss %xmm1, %xmm2, %xmm1 vsubss %xmm1, %xmm0, %xmm0 vmulss %xmm0, %xmm0, %xmm1 vmulss 0x1ae658c(%rip), %xmm1, %xmm2 # 0x1ef09a4 vaddss 0x1ae6588(%rip), %xmm2, %xmm2 # 0x1ef09a8 vmulss 0x1ae6584(%rip), %xmm1, %xmm3 # 0x1ef09ac vaddss 0x1ae6580(%rip), %xmm3, %xmm5 # 0x1ef09b0 vxorps 0x90(%rsp), %xmm4, %xmm3 vmulss %xmm2, %xmm1, %xmm2 vaddss 0x1ae656f(%rip), %xmm2, %xmm2 # 0x1ef09b4 vmovaps 0x10(%rsp), %xmm4 vcmpltss 0x80(%rsp), %xmm4, %xmm4 vmulss %xmm5, %xmm1, %xmm5 vaddss 0x1ae6557(%rip), %xmm5, %xmm5 # 0x1ef09b8 movl %eax, %ecx vmulss %xmm2, %xmm1, %xmm2 vaddss 0x1ae654d(%rip), %xmm2, %xmm2 # 0x1ef09bc andl $0x3, %ecx vmulss %xmm5, %xmm1, %xmm5 vaddss 0x1ae6542(%rip), %xmm5, %xmm5 # 0x1ef09c0 vmulss %xmm2, %xmm1, %xmm2 vaddss 0x1ae653a(%rip), %xmm2, %xmm6 # 0x1ef09c4 vmulss %xmm5, %xmm1, %xmm5 vaddss 0x1ae2286(%rip), %xmm5, %xmm5 # 0x1eec71c vmulss %xmm6, %xmm1, %xmm6 vaddss %xmm7, %xmm6, %xmm6 vmulss %xmm5, %xmm1, %xmm1 vaddss %xmm7, %xmm1, %xmm5 vmulss %xmm6, %xmm0, %xmm14 testb $0x1, %al je 0x40a4ba vmovaps %xmm14, 0x10(%rsp) vmovaps %xmm5, %xmm14 jmp 0x40a4c0 vmovaps %xmm5, 0x10(%rsp) vmovaps 0x90(%rsp), %xmm1 vblendvps %xmm4, %xmm3, %xmm1, %xmm2 leal -0x1(%rcx), %eax cmpl $0x2, %ecx jb 0x40a4e4 vbroadcastss 0x1b169e0(%rip), %xmm4 # 0x1f20ec0 vxorps %xmm4, %xmm14, %xmm14 cmpl $0x2, %eax jae 0x40a502 vbroadcastss 0x1b169ce(%rip), %xmm4 # 0x1f20ec0 vmovaps 0x10(%rsp), %xmm0 vxorps %xmm4, %xmm0, %xmm0 vmovaps %xmm0, 0x10(%rsp) vmovaps 0x110(%rsp), %xmm11 vmovss 0xc(%rsp), %xmm1 vmulss %xmm1, %xmm11, %xmm4 vmovaps 0xb0(%rsp), %xmm13 vsubss %xmm13, %xmm4, %xmm4 vmovaps 0xe0(%rsp), %xmm3 vmulss %xmm3, %xmm1, %xmm5 vmovaps 0xc0(%rsp), %xmm15 vsubss %xmm15, %xmm5, %xmm5 vmovaps 0xf0(%rsp), %xmm7 vmulss %xmm7, %xmm1, %xmm6 vsubss %xmm2, %xmm6, %xmm10 vmovaps 0x100(%rsp), %xmm8 vmulss %xmm1, %xmm8, %xmm6 vmovaps 0xd0(%rsp), %xmm0 vsubss %xmm0, %xmm6, %xmm12 vmovss %xmm5, 0x90(%rsp) vmulss %xmm5, %xmm5, %xmm6 vmulss %xmm4, %xmm4, %xmm9 vaddss %xmm6, %xmm9, %xmm6 vmovss %xmm10, 0x80(%rsp) vmulss %xmm10, %xmm10, %xmm9 vaddss %xmm6, %xmm9, %xmm6 vmovss %xmm12, 0xa0(%rsp) vmulss %xmm12, %xmm12, %xmm9 vaddss %xmm6, %xmm9, %xmm6 vrsqrtss %xmm6, %xmm6, %xmm9 vmovss 0x1ae216e(%rip), %xmm12 # 0x1eec71c vmulss %xmm6, %xmm12, %xmm6 vmulss %xmm6, %xmm9, %xmm6 vmulss %xmm9, %xmm9, %xmm10 vmulss %xmm6, %xmm10, %xmm6 vmulss 0x1ae2151(%rip), %xmm9, %xmm9 # 0x1eec718 vaddss %xmm6, %xmm9, %xmm9 vmulss %xmm4, %xmm9, %xmm4 vmovaps %xmm14, %xmm5 vmulss %xmm4, %xmm14, %xmm4 vmulss 0x10(%rsp), %xmm11, %xmm6 vsubss %xmm4, %xmm6, %xmm4 vmovaps 0x20(%rsp), %xmm14 vmovss 0x1ae2125(%rip), %xmm6 # 0x1eec714 vsubss %xmm14, %xmm6, %xmm6 vmulss %xmm13, %xmm14, %xmm10 vmulss %xmm6, %xmm11, %xmm11 vaddss %xmm10, %xmm11, %xmm13 vmulss %xmm15, %xmm14, %xmm10 vmulss %xmm3, %xmm6, %xmm11 vaddss %xmm10, %xmm11, %xmm10 vmulss %xmm2, %xmm14, %xmm3 vmulss %xmm7, %xmm6, %xmm11 vaddss %xmm3, %xmm11, %xmm3 vmulss %xmm0, %xmm14, %xmm11 vmulss %xmm6, %xmm8, %xmm14 vaddss %xmm11, %xmm14, %xmm11 vmulss %xmm10, %xmm10, %xmm14 vmulss %xmm13, %xmm13, %xmm15 vaddss %xmm14, %xmm15, %xmm14 vmulss %xmm3, %xmm3, %xmm15 vaddss %xmm14, %xmm15, %xmm14 vmulss %xmm11, %xmm11, %xmm15 vaddss %xmm14, %xmm15, %xmm14 vmulss %xmm12, %xmm14, %xmm2 vrsqrtss %xmm14, %xmm14, %xmm14 vmulss 0x1ae20bb(%rip), %xmm14, %xmm12 # 0x1eec718 vmulss %xmm2, %xmm14, %xmm2 vmulss %xmm14, %xmm14, %xmm14 vmulss %xmm2, %xmm14, %xmm2 vaddss %xmm2, %xmm12, %xmm2 vmovss 0x1ae6352(%rip), %xmm12 # 0x1ef09c8 vucomiss %xmm12, %xmm1 vcmpltss %xmm1, %xmm12, %xmm12 vmulss %xmm2, %xmm13, %xmm13 vblendvps %xmm12, %xmm13, %xmm4, %xmm4 ja 0x40a6e2 vmulss 0x90(%rsp), %xmm9, %xmm2 vmulss 0x80(%rsp), %xmm9, %xmm3 vmulss 0xa0(%rsp), %xmm9, %xmm9 vmulss %xmm5, %xmm2, %xmm2 vmovaps 0x10(%rsp), %xmm0 vmulss 0xe0(%rsp), %xmm0, %xmm7 vsubss %xmm2, %xmm7, %xmm7 vmulss %xmm5, %xmm3, %xmm2 vmulss 0xf0(%rsp), %xmm0, %xmm3 vsubss %xmm2, %xmm3, %xmm8 vmulss %xmm5, %xmm9, %xmm1 vmulss 0x100(%rsp), %xmm0, %xmm0 vsubss %xmm1, %xmm0, %xmm5 jmp 0x40a6ee vmulss %xmm2, %xmm10, %xmm7 vmulss %xmm2, %xmm3, %xmm8 vmulss %xmm2, %xmm11, %xmm5 vpermilps $0x0, 0x20(%rsp), %xmm1 # xmm1 = mem[0,0,0,0] vmulps 0x150(%rsp), %xmm1, %xmm0 vshufps $0x0, %xmm6, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmulps 0x70(%rsp), %xmm6, %xmm2 vaddps %xmm0, %xmm2, %xmm0 vmovaps %xmm0, 0x20(%rsp) vmulps 0x140(%rsp), %xmm1, %xmm2 vmulps 0x60(%rsp), %xmm6, %xmm3 vaddps %xmm2, %xmm3, %xmm3 vmulps 0x120(%rsp), %xmm1, %xmm2 vmulps 0x40(%rsp), %xmm6, %xmm9 vaddps %xmm2, %xmm9, %xmm2 vmulps 0x130(%rsp), %xmm1, %xmm1 vmulps 0x50(%rsp), %xmm6, %xmm6 vaddps %xmm1, %xmm6, %xmm1 vmulss %xmm7, %xmm7, %xmm10 vmulss %xmm4, %xmm4, %xmm11 vaddss %xmm10, %xmm11, %xmm6 vbroadcastss 0x1b1675d(%rip), %xmm0 # 0x1f20ec0 vxorps %xmm0, %xmm8, %xmm9 vmulss %xmm8, %xmm9, %xmm9 vaddss %xmm6, %xmm9, %xmm6 vxorps %xmm0, %xmm5, %xmm12 vmulss %xmm5, %xmm12, %xmm12 vaddss %xmm6, %xmm12, %xmm13 vmulss %xmm5, %xmm4, %xmm6 vmulss %xmm7, %xmm8, %xmm14 vaddss %xmm6, %xmm14, %xmm15 vsubss %xmm6, %xmm14, %xmm6 vmulss %xmm5, %xmm7, %xmm14 vsubss %xmm10, %xmm11, %xmm10 vmulss %xmm8, %xmm8, %xmm11 vaddss %xmm10, %xmm11, %xmm11 vaddss %xmm11, %xmm12, %xmm11 vmulss %xmm4, %xmm8, %xmm12 vmulss %xmm7, %xmm4, %xmm7 vsubss %xmm12, %xmm14, %xmm0 vmulss %xmm5, %xmm8, %xmm8 vaddss %xmm12, %xmm14, %xmm4 vaddss %xmm7, %xmm8, %xmm12 vsubss %xmm7, %xmm8, %xmm7 vaddss %xmm15, %xmm15, %xmm8 vaddss %xmm0, %xmm0, %xmm0 vaddss %xmm10, %xmm9, %xmm9 vmulss %xmm5, %xmm5, %xmm5 vaddss %xmm5, %xmm9, %xmm9 vshufps $0x0, %xmm13, %xmm13, %xmm5 # xmm5 = xmm13[0,0,0,0] vshufps $0x0, %xmm8, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmovaps 0x1ae1f0f(%rip), %xmm10 # 0x1eec700 vmulps %xmm0, %xmm10, %xmm0 vmovsd 0x1ae1ef3(%rip), %xmm13 # 0x1eec6f0 vmulps %xmm13, %xmm8, %xmm8 vaddps %xmm0, %xmm8, %xmm0 vmovss 0x1ae1f06(%rip), %xmm8 # 0x1eec714 vmulps %xmm5, %xmm8, %xmm5 vaddps %xmm0, %xmm5, %xmm5 vaddss %xmm12, %xmm12, %xmm0 vshufps $0x0, %xmm11, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vmulps %xmm0, %xmm10, %xmm0 vmulps %xmm13, %xmm11, %xmm11 vaddps %xmm0, %xmm11, %xmm0 vaddss %xmm6, %xmm6, %xmm6 vshufps $0x0, %xmm6, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] vmulps %xmm6, %xmm8, %xmm6 vaddps %xmm0, %xmm6, %xmm6 vshufps $0x0, %xmm9, %xmm9, %xmm0 # xmm0 = xmm9[0,0,0,0] vmulps %xmm0, %xmm10, %xmm0 vaddss %xmm7, %xmm7, %xmm7 vshufps $0x0, %xmm7, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] vmulps %xmm7, %xmm13, %xmm7 vaddps %xmm0, %xmm7, %xmm0 vxorps %xmm7, %xmm7, %xmm7 vaddss %xmm4, %xmm4, %xmm4 vshufps $0x0, %xmm4, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] vmulps %xmm4, %xmm8, %xmm4 vmovaps 0x20(%rsp), %xmm10 vshufps $0xe9, %xmm7, %xmm10, %xmm8 # xmm8 = xmm10[1,2],xmm7[2,3] vblendps $0x4, %xmm3, %xmm8, %xmm8 # xmm8 = xmm8[0,1],xmm3[2],xmm8[3] vaddps %xmm0, %xmm4, %xmm9 vaddps %xmm7, %xmm8, %xmm8 vshufps $0x0, %xmm10, %xmm10, %xmm0 # xmm0 = xmm10[0,0,0,0] vmulps %xmm7, %xmm9, %xmm4 vmulps %xmm7, %xmm6, %xmm7 vaddps %xmm4, %xmm7, %xmm7 vmulps %xmm5, %xmm0, %xmm0 vaddps %xmm7, %xmm0, %xmm0 vshufps $0x0, %xmm3, %xmm3, %xmm7 # xmm7 = xmm3[0,0,0,0] vshufps $0x55, %xmm3, %xmm3, %xmm3 # xmm3 = xmm3[1,1,1,1] vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm4, %xmm3, %xmm3 vmulps %xmm5, %xmm7, %xmm4 vaddps %xmm3, %xmm4, %xmm4 vshufps $0x55, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,1,1,1] vshufps $0xaa, %xmm2, %xmm2, %xmm7 # xmm7 = xmm2[2,2,2,2] vmulps %xmm7, %xmm9, %xmm7 vmulps %xmm6, %xmm3, %xmm3 vaddps %xmm7, %xmm3, %xmm3 vshufps $0x0, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] vmulps %xmm5, %xmm2, %xmm2 vaddps %xmm3, %xmm2, %xmm2 vshufps $0xaa, %xmm1, %xmm1, %xmm3 # xmm3 = xmm1[2,2,2,2] vmulps %xmm3, %xmm9, %xmm3 vshufps $0x55, %xmm1, %xmm1, %xmm7 # xmm7 = xmm1[1,1,1,1] vmulps %xmm6, %xmm7, %xmm6 vaddps %xmm3, %xmm6, %xmm3 vshufps $0x0, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] vmulps %xmm5, %xmm1, %xmm1 vaddps %xmm3, %xmm1, %xmm1 vaddps %xmm1, %xmm8, %xmm1 jmp 0x40984c	/embree[P]embree/kernels/geometry/instance_array_intersector.cpp
embree::avx::InstanceArrayIntersector1::pointQuery(embree::PointQueryK<1>, embree::PointQueryContext, embree::InstanceArrayPrimitive const&) (.cold.1)	__forceinline void updateAABB() { if (likely(query_ws->radius == (float)inf \|\| userContext->instStackSize == 0)) { query_radius = Vec3fa(query_ws->radius); return; } const AffineSpace3fa m = AffineSpace3fa_load_unaligned((AffineSpace3fa)userContext->world2inst[userContext->instStackSize-1]); BBox3fa bbox(Vec3fa(-query_ws->radius), Vec3fa(query_ws->radius)); bbox = xfmBounds(m, bbox); query_radius = 0.5f (bbox.upper - bbox.lower); }	decl %edi shlq $0x6, %rdi vmovups (%rsi,%rdi), %xmm1 vmovups 0x10(%rsi,%rdi), %xmm2 vmovups 0x30(%rsi,%rdi), %xmm3 vbroadcastss 0x1b16594(%rip), %xmm4 # 0x1f20ec0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vxorps %xmm4, %xmm0, %xmm4 vmulps 0x20(%rsi,%rdi), %xmm0, %xmm5 vsubps %xmm5, %xmm3, %xmm6 vmulps %xmm2, %xmm4, %xmm7 vaddps %xmm6, %xmm7, %xmm8 vmulps %xmm1, %xmm4, %xmm4 vaddps %xmm4, %xmm8, %xmm9 vbroadcastss 0x1ae10c8(%rip), %xmm10 # 0x1eeba20 vminps %xmm9, %xmm10, %xmm10 vbroadcastss 0x1ae221e(%rip), %xmm11 # 0x1eecb84 vmaxps %xmm9, %xmm11, %xmm9 vaddps %xmm3, %xmm5, %xmm3 vaddps %xmm3, %xmm7, %xmm5 vaddps %xmm5, %xmm4, %xmm7 vminps %xmm7, %xmm10, %xmm10 vmaxps %xmm7, %xmm9, %xmm7 vmulps %xmm2, %xmm0, %xmm2 vaddps %xmm6, %xmm2, %xmm6 vaddps %xmm6, %xmm4, %xmm9 vminps %xmm9, %xmm10, %xmm10 vmaxps %xmm9, %xmm7, %xmm7 vaddps %xmm3, %xmm2, %xmm2 vaddps %xmm2, %xmm4, %xmm3 vminps %xmm3, %xmm10, %xmm4 vmaxps %xmm3, %xmm7, %xmm3 vmulps %xmm1, %xmm0, %xmm0 vaddps %xmm0, %xmm8, %xmm1 vminps %xmm1, %xmm4, %xmm4 vmaxps %xmm1, %xmm3, %xmm1 vaddps %xmm5, %xmm0, %xmm3 vminps %xmm3, %xmm4, %xmm4 vmaxps %xmm3, %xmm1, %xmm1 vaddps %xmm6, %xmm0, %xmm3 vminps %xmm3, %xmm4, %xmm4 vmaxps %xmm3, %xmm1, %xmm1 vaddps %xmm2, %xmm0, %xmm0 vminps %xmm0, %xmm4, %xmm2 vmaxps %xmm0, %xmm1, %xmm0 vsubps %xmm2, %xmm0, %xmm0 vbroadcastss 0x1ae219a(%rip), %xmm1 # 0x1eecb80 vmulps %xmm1, %xmm0, %xmm0 vmovaps %xmm0, (%rdx) retq	/embree[P]embree/kernels/geometry/../common/context.h
embree::avx::InstanceArrayIntersector1MB::pointQuery(embree::PointQueryK<1>, embree::PointQueryContext, embree::InstanceArrayPrimitive const&) (.cold.1)	__forceinline void updateAABB() { if (likely(query_ws->radius == (float)inf \|\| userContext->instStackSize == 0)) { query_radius = Vec3fa(query_ws->radius); return; } const AffineSpace3fa m = AffineSpace3fa_load_unaligned((AffineSpace3fa)userContext->world2inst[userContext->instStackSize-1]); BBox3fa bbox(Vec3fa(-query_ws->radius), Vec3fa(query_ws->radius)); bbox = xfmBounds(m, bbox); query_radius = 0.5f (bbox.upper - bbox.lower); }	decl %edi shlq $0x6, %rdi vmovups (%rsi,%rdi), %xmm1 vmovups 0x10(%rsi,%rdi), %xmm2 vmovups 0x30(%rsi,%rdi), %xmm3 vbroadcastss 0x1b164b1(%rip), %xmm4 # 0x1f20ec0 vshufps $0x0, %xmm0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] vxorps %xmm4, %xmm0, %xmm4 vmulps 0x20(%rsi,%rdi), %xmm0, %xmm5 vsubps %xmm5, %xmm3, %xmm6 vmulps %xmm2, %xmm4, %xmm7 vaddps %xmm6, %xmm7, %xmm8 vmulps %xmm1, %xmm4, %xmm4 vaddps %xmm4, %xmm8, %xmm9 vbroadcastss 0x1ae0fe5(%rip), %xmm10 # 0x1eeba20 vminps %xmm9, %xmm10, %xmm10 vbroadcastss 0x1ae213b(%rip), %xmm11 # 0x1eecb84 vmaxps %xmm9, %xmm11, %xmm9 vaddps %xmm3, %xmm5, %xmm3 vaddps %xmm3, %xmm7, %xmm5 vaddps %xmm5, %xmm4, %xmm7 vminps %xmm7, %xmm10, %xmm10 vmaxps %xmm7, %xmm9, %xmm7 vmulps %xmm2, %xmm0, %xmm2 vaddps %xmm6, %xmm2, %xmm6 vaddps %xmm6, %xmm4, %xmm9 vminps %xmm9, %xmm10, %xmm10 vmaxps %xmm9, %xmm7, %xmm7 vaddps %xmm3, %xmm2, %xmm2 vaddps %xmm2, %xmm4, %xmm3 vminps %xmm3, %xmm10, %xmm4 vmaxps %xmm3, %xmm7, %xmm3 vmulps %xmm1, %xmm0, %xmm0 vaddps %xmm0, %xmm8, %xmm1 vminps %xmm1, %xmm4, %xmm4 vmaxps %xmm1, %xmm3, %xmm1 vaddps %xmm5, %xmm0, %xmm3 vminps %xmm3, %xmm4, %xmm4 vmaxps %xmm3, %xmm1, %xmm1 vaddps %xmm6, %xmm0, %xmm3 vminps %xmm3, %xmm4, %xmm4 vmaxps %xmm3, %xmm1, %xmm1 vaddps %xmm2, %xmm0, %xmm0 vminps %xmm0, %xmm4, %xmm2 vmaxps %xmm0, %xmm1, %xmm0 vsubps %xmm2, %xmm0, %xmm0 vbroadcastss 0x1ae20b7(%rip), %xmm1 # 0x1eecb80 vmulps %xmm1, %xmm0, %xmm0 vmovaps %xmm0, (%rdx) retq	/embree[P]embree/kernels/geometry/../common/context.h
embree::avx2::BVHNIntersector1<8, 16777232, false, embree::avx2::ArrayIntersector1<embree::avx2::InstanceArrayIntersector1MB>>::occluded(embree::Accel::Intersectors const, embree::RayK<1>&, embree::RayQueryContext)	void BVHNIntersector1<N, types, robust, PrimitiveIntersector1>::occluded(const Accel::Intersectors* __restrict__ This, Ray& __restrict__ ray, RayQueryContext* __restrict__ context) { const BVH* __restrict__ bvh = (const BVH)This->ptr; / we may traverse an empty BVH in case all geometry was invalid / if (bvh->root == BVH::emptyNode) return; / early out for already occluded rays / if (unlikely(ray.tfar < 0.0f)) return; / perform per ray precalculations required by the primitive intersector / Precalculations pre(ray, bvh); / stack state / NodeRef stack[stackSize]; // stack of nodes that still need to get traversed NodeRef stackPtr = stack+1; // current stack pointer NodeRef* stackEnd = stack+stackSize; stack[0] = bvh->root; /* filter out invalid rays / #if defined(EMBREE_IGNORE_INVALID_RAYS) if (!ray.valid()) return; #endif / verify correct input / assert(ray.valid()); assert(ray.tnear() >= 0.0f); assert(!(types & BVH_MB) \|\| (ray.time() >= 0.0f && ray.time() <= 1.0f)); / load the ray into SIMD registers / TravRay<N,robust> tray(ray.org, ray.dir, max(ray.tnear(), 0.0f), max(ray.tfar, 0.0f)); / initialize the node traverser / BVHNNodeTraverser1Hit<N, types> nodeTraverser; / pop loop / while (true) pop: { / pop next node / if (unlikely(stackPtr == stack)) break; stackPtr--; NodeRef cur = (NodeRef)stackPtr; /* downtraversal loop / while (true) { / intersect node / size_t mask; vfloat<N> tNear; STAT3(shadow.trav_nodes,1,1,1); bool nodeIntersected = BVHNNodeIntersector1<N, types, robust>::intersect(cur, tray, ray.time(), tNear, mask); if (unlikely(!nodeIntersected)) { STAT3(shadow.trav_nodes,-1,-1,-1); break; } / if no child is hit, pop next node / if (unlikely(mask == 0)) goto pop; / select next child and push other children / nodeTraverser.traverseAnyHit(cur, mask, tNear, stackPtr, stackEnd); } / this is a leaf node / assert(cur != BVH::emptyNode); STAT3(shadow.trav_leaves,1,1,1); size_t num; Primitive prim = (Primitive)cur.leaf(num); size_t lazy_node = 0; if (PrimitiveIntersector1::occluded(This, pre, ray, context, prim, num, tray, lazy_node)) { ray.tfar = neg_inf; break; } / push lazy node onto stack / if (unlikely(lazy_node)) { stackPtr = (NodeRef)lazy_node; stackPtr++; } } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x12f8, %rsp # imm = 0x12F8 movq %rdx, 0x38(%rsp) movq (%rdi), %rax cmpq $0x8, 0x70(%rax) je 0x4336d8 vmovss 0x20(%rsi), %xmm0 vxorps %xmm2, %xmm2, %xmm2 vucomiss %xmm0, %xmm2 ja 0x4336d8 leaq 0x158(%rsp), %rdx movq 0x70(%rax), %rax movq %rax, -0x8(%rdx) vmaxss 0xc(%rsi), %xmm2, %xmm1 vmovaps 0x10(%rsi), %xmm3 vbroadcastss 0x1aedc1e(%rip), %xmm4 # 0x1f20ec4 vandps %xmm4, %xmm3, %xmm4 vbroadcastss 0x1abdd35(%rip), %xmm5 # 0x1ef0fe8 vcmpltps %xmm5, %xmm4, %xmm4 vblendvps %xmm4, %xmm5, %xmm3, %xmm3 vrcpps %xmm3, %xmm4 vbroadcastss 0x1ab9449(%rip), %xmm5 # 0x1eec714 vfnmadd231ps %xmm3, %xmm4, %xmm5 # xmm5 = -(xmm4 * xmm3) + xmm5 vfmadd132ps %xmm4, %xmm4, %xmm5 # xmm5 = (xmm5 * xmm4) + xmm4 xorl %edi, %edi vucomiss %xmm2, %xmm5 setb %dil vbroadcastss %xmm5, %ymm8 vmovshdup %xmm5, %xmm6 # xmm6 = xmm5[1,1,3,3] vbroadcastsd %xmm6, %ymm9 vshufpd $0x1, %xmm5, %xmm5, %xmm7 # xmm7 = xmm5[1,0] vshufps $0xaa, %xmm5, %xmm5, %xmm3 # xmm3 = xmm5[2,2,2,2] vbroadcastsd %xmm3, %ymm10 vmulps (%rsi), %xmm5, %xmm5 vbroadcastss %xmm5, %ymm3 vmovshdup %xmm5, %xmm4 # xmm4 = xmm5[1,1,3,3] vbroadcastsd %xmm4, %ymm4 vshufps $0xaa, %xmm5, %xmm5, %xmm5 # xmm5 = xmm5[2,2,2,2] shll $0x5, %edi xorl %r8d, %r8d vucomiss %xmm2, %xmm6 vbroadcastsd %xmm5, %ymm5 setb %r8b shll $0x5, %r8d orq $0x40, %r8 xorl %r9d, %r9d vucomiss %xmm2, %xmm7 setb %r9b shll $0x5, %r9d orq $0x80, %r9 movq %rdi, %r10 xorq $0x20, %r10 movq %r8, %r15 xorq $0x20, %r15 movq %r9, %rbx xorq $0x20, %rbx vbroadcastss 0x1aedb5e(%rip), %ymm2 # 0x1f20ec0 vxorps %ymm2, %ymm3, %ymm6 vxorps %ymm2, %ymm4, %ymm7 vxorps %ymm2, %ymm5, %ymm5 vbroadcastss %xmm1, %ymm11 vbroadcastss %xmm0, %ymm12 leaq 0x150(%rsp), %r11 movq %rsi, 0x8(%rsp) movq %rdi, 0x28(%rsp) vmovups %ymm8, 0x130(%rsp) vmovups %ymm9, 0x110(%rsp) vmovups %ymm10, 0xf0(%rsp) movq %r8, 0x20(%rsp) movq %r9, 0x18(%rsp) movq %r10, 0x10(%rsp) vmovups %ymm6, 0xd0(%rsp) vmovups %ymm7, 0xb0(%rsp) vmovups %ymm5, 0x90(%rsp) vmovups %ymm11, 0x70(%rsp) vmovups %ymm12, 0x50(%rsp) cmpq %r11, %rdx je 0x4336d8 movq -0x8(%rdx), %r12 addq $-0x8, %rdx testb $0x8, %r12b jne 0x4334ce movq %r12, %rax vbroadcastss 0x1c(%rsi), %ymm0 andq $-0x10, %rax vmovaps 0x100(%rax,%rdi), %ymm1 vfmadd213ps 0x40(%rax,%rdi), %ymm0, %ymm1 # ymm1 = (ymm0 * ymm1) + mem vmovaps 0x100(%rax,%r8), %ymm2 vfmadd213ps 0x40(%rax,%r8), %ymm0, %ymm2 # ymm2 = (ymm0 * ymm2) + mem vfmadd213ps %ymm6, %ymm8, %ymm1 # ymm1 = (ymm8 * ymm1) + ymm6 vfmadd213ps %ymm7, %ymm9, %ymm2 # ymm2 = (ymm9 * ymm2) + ymm7 vpmaxsd %ymm2, %ymm1, %ymm1 vmovaps 0x100(%rax,%r9), %ymm2 vfmadd213ps 0x40(%rax,%r9), %ymm0, %ymm2 # ymm2 = (ymm0 * ymm2) + mem vfmadd213ps %ymm5, %ymm10, %ymm2 # ymm2 = (ymm10 * ymm2) + ymm5 vpmaxsd %ymm11, %ymm2, %ymm2 vpmaxsd %ymm2, %ymm1, %ymm1 vmovaps 0x100(%rax,%r10), %ymm2 vfmadd213ps 0x40(%rax,%r10), %ymm0, %ymm2 # ymm2 = (ymm0 * ymm2) + mem vfmadd213ps %ymm6, %ymm8, %ymm2 # ymm2 = (ymm8 * ymm2) + ymm6 vmovaps 0x100(%rax,%r15), %ymm3 vfmadd213ps 0x40(%rax,%r15), %ymm0, %ymm3 # ymm3 = (ymm0 * ymm3) + mem vfmadd213ps %ymm7, %ymm9, %ymm3 # ymm3 = (ymm9 * ymm3) + ymm7 vmovaps 0x100(%rax,%rbx), %ymm4 vfmadd213ps 0x40(%rax,%rbx), %ymm0, %ymm4 # ymm4 = (ymm0 * ymm4) + mem vpminsd %ymm3, %ymm2, %ymm2 vfmadd213ps %ymm5, %ymm10, %ymm4 # ymm4 = (ymm10 * ymm4) + ymm5 vpminsd %ymm12, %ymm4, %ymm3 vpminsd %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm1, %ymm1 movl %r12d, %ecx andl $0x7, %ecx cmpl $0x6, %ecx je 0x43352a vextractf128 $0x1, %ymm1, %xmm0 vpackssdw %xmm0, %xmm1, %xmm0 vpsllw $0xf, %xmm0, %xmm0 vpacksswb %xmm0, %xmm0, %xmm0 vpmovmskb %xmm0, %eax movzbl %al, %r14d testb $0x8, %r12b jne 0x433526 testq %r14, %r14 je 0x433557 andq $-0x10, %r12 tzcntq %r14, %rcx xorl %r13d, %r13d blsrq %r14, %rax movq (%r12,%rcx,8), %rcx prefetcht0 (%rcx) prefetcht0 0x40(%rcx) prefetcht0 0x80(%rcx) prefetcht0 0xc0(%rcx) je 0x433518 movq %rcx, (%rdx) addq $0x8, %rdx tzcntq %rax, %rcx blsrq %rax, %rax jmp 0x4334ea movq %rcx, %r12 testl %r13d, %r13d je 0x4333ec jmp 0x43355d pushq $0x6 jmp 0x433559 vmovaps 0x1c0(%rax), %ymm2 vcmpleps %ymm0, %ymm2, %ymm2 vcmpltps 0x1e0(%rax), %ymm0, %ymm0 vandps %ymm0, %ymm2, %ymm0 vandps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vpackssdw %xmm1, %xmm0, %xmm0 jmp 0x4334bd pushq $0x4 popq %r13 jmp 0x43351b cmpl $0x6, %r13d jne 0x4336ce movl %r12d, %eax andl $0xf, %eax xorl %r13d, %r13d movq %rax, 0x40(%rsp) addq $-0x8, %rax setne %bpl je 0x4336ce movq %rax, 0x48(%rsp) movq %rdx, 0x30(%rsp) andq $-0x10, %r12 leaq 0x7(%rsp), %rdi movq 0x38(%rsp), %rdx movq %r12, %rcx vzeroupper callq 0x60a670 testb %al, %al jne 0x4335ec addq $0x8, %r12 addq $-0x9, 0x40(%rsp) xorl %ebp, %ebp cmpq %rbp, 0x40(%rsp) je 0x433666 leaq 0x7(%rsp), %rdi movq 0x8(%rsp), %rsi movq 0x38(%rsp), %rdx movq %r12, %rcx callq 0x60a670 addq $0x8, %r12 incq %rbp testb %al, %al je 0x4335b6 cmpq 0x48(%rsp), %rbp setb %bpl testb %bpl, %bpl movq 0x8(%rsp), %rsi movq 0x30(%rsp), %rdx movq 0x28(%rsp), %rdi vmovups 0x130(%rsp), %ymm8 vmovups 0x110(%rsp), %ymm9 vmovups 0xf0(%rsp), %ymm10 movq 0x20(%rsp), %r8 movq 0x18(%rsp), %r9 movq 0x10(%rsp), %r10 vmovups 0xd0(%rsp), %ymm6 vmovups 0xb0(%rsp), %ymm7 vmovups 0x90(%rsp), %ymm5 vmovups 0x70(%rsp), %ymm11 vmovups 0x50(%rsp), %ymm12 leaq 0x150(%rsp), %r11 je 0x4336ce movl $0xff800000, 0x20(%rsi) # imm = 0xFF800000 pushq $0x3 popq %r13 jmp 0x4336ce movq 0x8(%rsp), %rsi movq 0x30(%rsp), %rdx movq 0x28(%rsp), %rdi vmovups 0x130(%rsp), %ymm8 vmovups 0x110(%rsp), %ymm9 vmovups 0xf0(%rsp), %ymm10 movq 0x20(%rsp), %r8 movq 0x18(%rsp), %r9 movq 0x10(%rsp), %r10 vmovups 0xd0(%rsp), %ymm6 vmovups 0xb0(%rsp), %ymm7 vmovups 0x90(%rsp), %ymm5 vmovups 0x70(%rsp), %ymm11 vmovups 0x50(%rsp), %ymm12 leaq 0x150(%rsp), %r11 cmpl $0x3, %r13d jne 0x4333db addq $0x12f8, %rsp # imm = 0x12F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/bvh_intersector1.cpp
embree::avx2::BVHNIntersector1<8, 1, false, embree::avx2::ArrayIntersector1<embree::avx2::QuadMiIntersector1Moeller<4, true>>>::intersect(embree::Accel::Intersectors const, embree::RayHitK<1>&, embree::RayQueryContext) (.cold.1)	static __forceinline vfloat8 select(const vboolf8& m, const vfloat8& t, const vfloat8& f) { return _mm256_blendv_ps(f, t, m); }	vbroadcastss 0x1ab5132(%rip), %ymm2 # 0x1eeba20 vblendvps %ymm1, %ymm0, %ymm2, %ymm0 vshufps $0xb1, %ymm0, %ymm0, %ymm2 # ymm2 = ymm0[1,0,3,2,5,4,7,6] vminps %ymm2, %ymm0, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vminps %ymm3, %ymm2, %ymm2 vpermpd $0x4e, %ymm2, %ymm3 # ymm3 = ymm2[2,3,0,1] vminps %ymm3, %ymm2, %ymm2 vcmpeqps %ymm2, %ymm0, %ymm0 vtestps %ymm1, %ymm0 je 0x436920 vandps %ymm1, %ymm0, %ymm1 vmovmskps %ymm1, %eax tzcntl %eax, %eax movl %eax, (%rdi) vzeroupper retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vfloat8_avx.h
embree::avx2::BVHNIntersectorKHybrid<4, 8, 1, false, embree::avx2::ArrayIntersectorK_1<8, embree::avx2::ObjectIntersectorK<8, false>>, false>::intersectCoherent(embree::vint_impl<8>, embree::Accel::Intersectors, embree::RayHitK<8>&, embree::RayQueryContext*)	static __forceinline vboolf8 operator ==(const vint8& a, const vint8& b) { return _mm256_castsi256_ps(_mm256_cmpeq_epi32(a, b)); }	vpcmpeqd %ymm0, %ymm0, %ymm0 vpcmpeqd (%rdi), %ymm0, %ymm1 vmovmskps %ymm1, %eax testl %eax, %eax je 0x50e0f4 pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x1320, %rsp # imm = 0x1320 movq (%rsi), %rsi movq %rsi, 0x58(%rsp) movzbl %al, %esi vmovaps (%rdx), %ymm7 vmovaps 0x20(%rdx), %ymm8 vmovaps 0x40(%rdx), %ymm9 vmovaps 0x80(%rdx), %ymm2 vbroadcastss 0x1a138ca(%rip), %ymm3 # 0x1f20ec4 vandps %ymm3, %ymm2, %ymm4 vbroadcastss 0x19e39e1(%rip), %ymm5 # 0x1ef0fe8 vcmpltps %ymm5, %ymm4, %ymm4 vblendvps %ymm4, %ymm5, %ymm2, %ymm11 vmovaps 0xa0(%rdx), %ymm4 vandps %ymm3, %ymm4, %ymm6 vcmpltps %ymm5, %ymm6, %ymm6 vblendvps %ymm6, %ymm5, %ymm4, %ymm12 vmovaps 0xc0(%rdx), %ymm4 vandps %ymm3, %ymm4, %ymm3 vcmpltps %ymm5, %ymm3, %ymm3 vblendvps %ymm3, %ymm5, %ymm4, %ymm13 vmovaps 0x60(%rdx), %ymm3 vrcpps %ymm11, %ymm4 vbroadcastss 0x19df0c1(%rip), %ymm5 # 0x1eec714 vfnmadd213ps %ymm5, %ymm4, %ymm11 # ymm11 = -(ymm4 * ymm11) + ymm5 vfmadd132ps %ymm4, %ymm4, %ymm11 # ymm11 = (ymm11 * ymm4) + ymm4 vrcpps %ymm12, %ymm4 vfnmadd213ps %ymm5, %ymm4, %ymm12 # ymm12 = -(ymm4 * ymm12) + ymm5 vfmadd132ps %ymm4, %ymm4, %ymm12 # ymm12 = (ymm12 * ymm4) + ymm4 vrcpps %ymm13, %ymm4 vfnmadd213ps %ymm5, %ymm4, %ymm13 # ymm13 = -(ymm4 * ymm13) + ymm5 vfmadd132ps %ymm4, %ymm4, %ymm13 # ymm13 = (ymm13 * ymm4) + ymm4 vxorps %xmm14, %xmm14, %xmm14 vmaxps %ymm14, %ymm3, %ymm3 vmovaps %ymm3, 0x2c0(%rsp) vmovaps 0x100(%rdx), %ymm3 vmaxps %ymm14, %ymm3, %ymm3 vmovaps %ymm3, 0x2a0(%rsp) vbroadcastss 0x1a05057(%rip), %ymm3 # 0x1f12704 vcmpltps %ymm14, %ymm2, %ymm2 vandps %ymm3, %ymm2, %ymm2 vmovaps 0xa0(%rdx), %ymm3 vmovaps 0xc0(%rdx), %ymm4 vcmpltps %ymm14, %ymm3, %ymm3 vbroadcastss 0x1a13806(%rip), %ymm5 # 0x1f20edc vandps %ymm5, %ymm3, %ymm3 vorps %ymm2, %ymm3, %ymm2 vcmpltps %ymm14, %ymm4, %ymm3 vbroadcastss 0x1a4d287(%rip), %ymm4 # 0x1f5a974 vandps %ymm4, %ymm3, %ymm3 vpxor %ymm0, %ymm1, %ymm0 vpor %ymm3, %ymm0, %ymm0 vpor %ymm2, %ymm0, %ymm0 vmovdqa %ymm0, 0x260(%rsp) vbroadcastss 0x1a137b1(%rip), %ymm0 # 0x1f20ec0 vmovaps %ymm7, 0x320(%rsp) vxorps %ymm0, %ymm7, %ymm1 vmulps %ymm1, %ymm11, %ymm10 vmovaps %ymm8, 0x300(%rsp) vxorps %ymm0, %ymm8, %ymm1 vmulps %ymm1, %ymm12, %ymm15 vmovaps %ymm9, 0x2e0(%rsp) vxorps %ymm0, %ymm9, %ymm0 vmulps %ymm0, %ymm13, %ymm9 leaq 0x3c0(%rsp), %r11 pushq $0x8 popq %r8 vbroadcastss 0x1a13769(%rip), %xmm0 # 0x1f20ec0 vmovaps %xmm0, 0x110(%rsp) vbroadcastss 0x19de2b7(%rip), %ymm0 # 0x1eeba20 vmovaps %ymm0, 0x280(%rsp) movq %rcx, 0x70(%rsp) movq %rdx, 0x68(%rsp) vmovaps %ymm11, 0x380(%rsp) vmovaps %ymm12, 0x360(%rsp) vmovaps %ymm13, 0x340(%rsp) vmovaps %ymm10, 0x240(%rsp) vmovaps %ymm15, 0x220(%rsp) vmovaps %ymm9, 0x200(%rsp) tzcntq %rsi, %rax vpbroadcastd 0x260(%rsp,%rax,4), %ymm0 vpcmpeqd 0x260(%rsp), %ymm0, %ymm0 vmovmskps %ymm0, %eax andnq %rsi, %rax, %rsi movq %rsi, 0x60(%rsp) vbroadcastss 0x19de23f(%rip), %ymm4 # 0x1eeba20 vblendvps %ymm0, %ymm11, %ymm4, %ymm1 vshufps $0xb1, %ymm1, %ymm1, %ymm2 # ymm2 = ymm1[1,0,3,2,5,4,7,6] vminps %ymm2, %ymm1, %ymm1 vshufpd $0x5, %ymm1, %ymm1, %ymm2 # ymm2 = ymm1[1,0,3,2] vminps %ymm2, %ymm1, %ymm1 vextractf128 $0x1, %ymm1, %xmm2 vminps %xmm2, %xmm1, %xmm1 vblendvps %ymm0, %ymm12, %ymm4, %ymm2 vshufps $0xb1, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2,5,4,7,6] vminps %ymm3, %ymm2, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vminps %ymm3, %ymm2, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vminps %xmm3, %xmm2, %xmm2 vunpcklps %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] vblendvps %ymm0, %ymm13, %ymm4, %ymm2 vmovaps %ymm4, %ymm7 vshufps $0xb1, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2,5,4,7,6] vminps %ymm3, %ymm2, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vminps %ymm3, %ymm2, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vminps %xmm3, %xmm2, %xmm2 vbroadcastss 0x19df32c(%rip), %ymm15 # 0x1eecb84 vblendvps %ymm0, %ymm11, %ymm15, %ymm3 vinsertps $0x28, %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0,1],xmm2[0],zero vshufps $0xb1, %ymm3, %ymm3, %ymm2 # ymm2 = ymm3[1,0,3,2,5,4,7,6] vmaxps %ymm2, %ymm3, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vmaxps %ymm3, %ymm2, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vblendvps %ymm0, %ymm12, %ymm15, %ymm4 vmaxps %xmm3, %xmm2, %xmm2 vshufps $0xb1, %ymm4, %ymm4, %ymm3 # ymm3 = ymm4[1,0,3,2,5,4,7,6] vmaxps %ymm3, %ymm4, %ymm3 vshufpd $0x5, %ymm3, %ymm3, %ymm4 # ymm4 = ymm3[1,0,3,2] vmaxps %ymm4, %ymm3, %ymm3 vextractf128 $0x1, %ymm3, %xmm4 vmaxps %xmm4, %xmm3, %xmm3 vunpcklps %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1] vblendvps %ymm0, %ymm13, %ymm15, %ymm3 vshufps $0xb1, %ymm3, %ymm3, %ymm4 # ymm4 = ymm3[1,0,3,2,5,4,7,6] vmaxps %ymm4, %ymm3, %ymm3 vshufpd $0x5, %ymm3, %ymm3, %ymm4 # ymm4 = ymm3[1,0,3,2] vmaxps %ymm4, %ymm3, %ymm3 vextractf128 $0x1, %ymm3, %xmm4 vmaxps %xmm4, %xmm3, %xmm3 vinsertps $0x28, %xmm3, %xmm2, %xmm3 # xmm3 = xmm2[0,1],xmm3[0],zero vmovaps %xmm1, 0xc0(%rsp) vcmpnltps 0x19de130(%rip), %xmm1, %xmm4 # 0x1eeba10 vblendvps %xmm4, %xmm1, %xmm3, %xmm2 vmovshdup %xmm2, %xmm5 # xmm5 = xmm2[1,1,3,3] xorl %esi, %esi vxorps %xmm6, %xmm6, %xmm6 vucomiss %xmm5, %xmm6 seta %sil shll $0x4, %esi orq $0x20, %rsi vshufpd $0x1, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[1,0] xorl %r14d, %r14d vucomiss %xmm5, %xmm6 seta %r14b shll $0x4, %r14d orq $0x40, %r14 movq %rsi, 0x98(%rsp) xorq $0x10, %rsi movq %rsi, 0x90(%rsp) xorl %r12d, %r12d vucomiss %xmm2, %xmm6 seta %r12b vmovaps 0x320(%rsp), %ymm8 vmovaps %ymm7, %ymm1 vblendvps %ymm0, %ymm8, %ymm7, %ymm5 vshufps $0xb1, %ymm5, %ymm5, %ymm6 # ymm6 = ymm5[1,0,3,2,5,4,7,6] vminps %ymm6, %ymm5, %ymm5 vshufpd $0x5, %ymm5, %ymm5, %ymm6 # ymm6 = ymm5[1,0,3,2] vminps %ymm6, %ymm5, %ymm5 vextractf128 $0x1, %ymm5, %xmm6 vminps %xmm6, %xmm5, %xmm5 vmovaps 0x300(%rsp), %ymm9 vblendvps %ymm0, %ymm9, %ymm7, %ymm6 vshufps $0xb1, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2,5,4,7,6] vminps %ymm7, %ymm6, %ymm6 vshufpd $0x5, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2] vminps %ymm7, %ymm6, %ymm6 vextractf128 $0x1, %ymm6, %xmm7 vminps %xmm7, %xmm6, %xmm6 vunpcklps %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[0],xmm6[0],xmm5[1],xmm6[1] vmovaps 0x2e0(%rsp), %ymm10 vblendvps %ymm0, %ymm10, %ymm1, %ymm6 vshufps $0xb1, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2,5,4,7,6] vminps %ymm7, %ymm6, %ymm6 vshufpd $0x5, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2] vminps %ymm7, %ymm6, %ymm6 vextractf128 $0x1, %ymm6, %xmm7 vminps %xmm7, %xmm6, %xmm6 vinsertps $0x28, %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[0,1],xmm6[0],zero vblendvps %ymm0, %ymm8, %ymm15, %ymm6 vshufps $0xb1, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2,5,4,7,6] vmaxps %ymm7, %ymm6, %ymm6 vshufpd $0x5, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2] vmaxps %ymm7, %ymm6, %ymm6 vextractf128 $0x1, %ymm6, %xmm7 vmaxps %xmm7, %xmm6, %xmm6 vblendvps %ymm0, %ymm9, %ymm15, %ymm7 vshufps $0xb1, %ymm7, %ymm7, %ymm8 # ymm8 = ymm7[1,0,3,2,5,4,7,6] vmaxps %ymm8, %ymm7, %ymm7 vshufpd $0x5, %ymm7, %ymm7, %ymm8 # ymm8 = ymm7[1,0,3,2] vmaxps %ymm8, %ymm7, %ymm7 vextractf128 $0x1, %ymm7, %xmm8 vmaxps %xmm8, %xmm7, %xmm7 vblendvps %ymm0, %ymm10, %ymm15, %ymm8 vunpcklps %xmm7, %xmm6, %xmm6 # xmm6 = xmm6[0],xmm7[0],xmm6[1],xmm7[1] vshufps $0xb1, %ymm8, %ymm8, %ymm7 # ymm7 = ymm8[1,0,3,2,5,4,7,6] vmaxps %ymm7, %ymm8, %ymm7 vshufpd $0x5, %ymm7, %ymm7, %ymm8 # ymm8 = ymm7[1,0,3,2] vmaxps %ymm8, %ymm7, %ymm7 vextractf128 $0x1, %ymm7, %xmm8 vmaxps %xmm8, %xmm7, %xmm7 vinsertps $0x28, %xmm7, %xmm6, %xmm6 # xmm6 = xmm6[0,1],xmm7[0],zero vblendvps %xmm4, %xmm5, %xmm6, %xmm7 vblendvps %xmm4, %xmm6, %xmm5, %xmm8 vblendvps %ymm0, 0x2a0(%rsp), %ymm15, %ymm6 vshufps $0xb1, %ymm6, %ymm6, %ymm5 # ymm5 = ymm6[1,0,3,2,5,4,7,6] vmovaps %ymm6, 0xa0(%rsp) vmaxps %ymm5, %ymm6, %ymm5 vshufpd $0x5, %ymm5, %ymm5, %ymm6 # ymm6 = ymm5[1,0,3,2] vmaxps %ymm6, %ymm5, %ymm5 vmovaps 0x2c0(%rsp), %ymm15 vblendvps %ymm0, %ymm15, %ymm1, %ymm6 vextractf128 $0x1, %ymm5, %xmm9 vshufps $0xb1, %ymm6, %ymm6, %ymm10 # ymm10 = ymm6[1,0,3,2,5,4,7,6] vminps %ymm10, %ymm6, %ymm6 vshufpd $0x5, %ymm6, %ymm6, %ymm10 # ymm10 = ymm6[1,0,3,2] vminps %ymm10, %ymm6, %ymm6 vextractf128 $0x1, %ymm6, %xmm10 vminps %xmm10, %xmm6, %xmm6 vmovaps 0x240(%rsp), %ymm10 vmaxss %xmm9, %xmm5, %xmm5 vmovaps 0x200(%rsp), %ymm9 vmovaps 0xc0(%rsp), %xmm1 vblendvps %xmm4, %xmm3, %xmm1, %xmm1 vmulps %xmm2, %xmm8, %xmm3 vmulps %xmm1, %xmm7, %xmm4 shll $0x4, %r12d movq %r12, 0x88(%rsp) xorq $0x10, %r12 movq 0x58(%rsp), %rax movq 0x70(%rax), %rax movq %rax, 0x3c0(%rsp) andl $0x0, 0x3c8(%rsp) movq %r14, %r13 xorq $0x10, %r14 vbroadcastss %xmm2, %xmm7 vmovaps %xmm7, 0x1e0(%rsp) vbroadcastss %xmm3, %xmm7 vmovaps 0x110(%rsp), %xmm8 vxorps %xmm7, %xmm8, %xmm7 vmovaps %xmm7, 0x1d0(%rsp) vshufps $0x55, %xmm2, %xmm2, %xmm7 # xmm7 = xmm2[1,1,1,1] vmovaps %xmm7, 0x1c0(%rsp) vshufps $0x55, %xmm3, %xmm3, %xmm7 # xmm7 = xmm3[1,1,1,1] vxorps %xmm7, %xmm8, %xmm7 vmovaps %xmm7, 0x1b0(%rsp) vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vmovaps %xmm2, 0x1a0(%rsp) vshufps $0xaa, %xmm3, %xmm3, %xmm2 # xmm2 = xmm3[2,2,2,2] vxorps %xmm2, %xmm8, %xmm2 vmovaps %xmm2, 0x190(%rsp) vbroadcastss %xmm1, %xmm2 vmovaps %xmm2, 0x180(%rsp) vbroadcastss %xmm4, %xmm2 vxorps %xmm2, %xmm8, %xmm2 vmovaps %xmm2, 0x170(%rsp) vshufps $0x55, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,1,1,1] vmovaps %xmm2, 0x160(%rsp) vshufps $0x55, %xmm4, %xmm4, %xmm2 # xmm2 = xmm4[1,1,1,1] vxorps %xmm2, %xmm8, %xmm2 vmovaps %xmm2, 0x150(%rsp) vshufps $0xaa, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vmovaps %xmm1, 0x140(%rsp) vshufps $0xaa, %xmm4, %xmm4, %xmm1 # xmm1 = xmm4[2,2,2,2] vxorps %xmm1, %xmm8, %xmm1 vmovaps %xmm1, 0x130(%rsp) vbroadcastss %xmm6, %xmm1 vmovaps %xmm1, 0x120(%rsp) vmovaps 0x280(%rsp), %ymm1 vblendvps %ymm0, %ymm15, %ymm1, %ymm0 vmovaps %ymm0, 0xc0(%rsp) vmovaps 0x220(%rsp), %ymm15 leaq 0x3d0(%rsp), %rbx vbroadcastss %xmm5, %xmm0 vmovaps %xmm0, 0x1f0(%rsp) cmpq %r11, %rbx je 0x50e0d8 vbroadcastss -0x8(%rbx), %ymm0 addq $-0x10, %rbx vcmpltps 0xa0(%rsp), %ymm0, %ymm1 vtestps %ymm1, %ymm1 je 0x50dc0d movq (%rbx), %r15 testb $0x8, %r15b jne 0x50df28 movq 0x88(%rsp), %rax vmovaps 0x20(%r15,%rax), %xmm1 vmovaps 0x1d0(%rsp), %xmm2 vfmadd132ps 0x1e0(%rsp), %xmm2, %xmm1 # xmm1 = (xmm1 * mem) + xmm2 movq 0x98(%rsp), %rax vmovaps 0x20(%r15,%rax), %xmm2 vmovaps 0x1b0(%rsp), %xmm3 vfmadd132ps 0x1c0(%rsp), %xmm3, %xmm2 # xmm2 = (xmm2 * mem) + xmm3 vpmaxsd %xmm2, %xmm1, %xmm1 vmovaps 0x20(%r15,%r13), %xmm2 vmovaps 0x190(%rsp), %xmm3 vfmadd132ps 0x1a0(%rsp), %xmm3, %xmm2 # xmm2 = (xmm2 * mem) + xmm3 vpmaxsd 0x120(%rsp), %xmm2, %xmm2 vpmaxsd %xmm2, %xmm1, %xmm1 vmovaps 0x20(%r15,%r12), %xmm2 vmovaps 0x170(%rsp), %xmm3 vfmadd132ps 0x180(%rsp), %xmm3, %xmm2 # xmm2 = (xmm2 * mem) + xmm3 movq 0x90(%rsp), %rax vmovaps 0x20(%r15,%rax), %xmm3 vmovaps 0x150(%rsp), %xmm4 vfmadd132ps 0x160(%rsp), %xmm4, %xmm3 # xmm3 = (xmm3 * mem) + xmm4 vpminsd %xmm3, %xmm2, %xmm2 vmovaps 0x20(%r15,%r14), %xmm3 vmovaps 0x130(%rsp), %xmm4 vfmadd132ps 0x140(%rsp), %xmm4, %xmm3 # xmm3 = (xmm3 * mem) + xmm4 vpminsd 0x1f0(%rsp), %xmm3, %xmm3 vpminsd %xmm3, %xmm2, %xmm2 vmovdqa %xmm1, 0xe0(%rsp) vcmpleps %xmm2, %xmm1, %xmm1 vmovmskps %xmm1, %eax testl %eax, %eax je 0x50de65 movzbl %al, %edi vbroadcastss 0x19ddce0(%rip), %ymm0 # 0x1eeba20 movq %r8, %rax xorl %r9d, %r9d tzcntq %rdi, %rsi vbroadcastss 0x20(%r15,%rsi,4), %ymm1 vbroadcastss 0x40(%r15,%rsi,4), %ymm2 vfmadd213ps %ymm10, %ymm11, %ymm1 # ymm1 = (ymm11 * ymm1) + ymm10 vfmadd213ps %ymm15, %ymm12, %ymm2 # ymm2 = (ymm12 * ymm2) + ymm15 vbroadcastss 0x60(%r15,%rsi,4), %ymm3 vfmadd213ps %ymm9, %ymm13, %ymm3 # ymm3 = (ymm13 * ymm3) + ymm9 vbroadcastss 0x30(%r15,%rsi,4), %ymm4 vfmadd213ps %ymm10, %ymm11, %ymm4 # ymm4 = (ymm11 * ymm4) + ymm10 vbroadcastss 0x50(%r15,%rsi,4), %ymm5 vbroadcastss 0x70(%r15,%rsi,4), %ymm6 vfmadd213ps %ymm15, %ymm12, %ymm5 # ymm5 = (ymm12 * ymm5) + ymm15 vfmadd213ps %ymm9, %ymm13, %ymm6 # ymm6 = (ymm13 * ymm6) + ymm9 vpminsd %ymm4, %ymm1, %ymm7 vpminsd %ymm5, %ymm2, %ymm8 vpmaxsd %ymm8, %ymm7, %ymm7 vpminsd %ymm6, %ymm3, %ymm8 vpmaxsd 0xc0(%rsp), %ymm8, %ymm8 vpmaxsd %ymm8, %ymm7, %ymm7 vpmaxsd %ymm4, %ymm1, %ymm1 vpmaxsd %ymm5, %ymm2, %ymm2 vpminsd %ymm2, %ymm1, %ymm1 vpmaxsd %ymm6, %ymm3, %ymm2 vpminsd 0xa0(%rsp), %ymm2, %ymm2 vpminsd %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm7, %ymm1 vtestps %ymm1, %ymm1 je 0x50de2f vbroadcastss 0xe0(%rsp,%rsi,4), %ymm1 movq (%r15,%rsi,8), %rsi prefetcht0 (%rsi) prefetcht0 0x40(%rsi) vcmpltps %ymm0, %ymm1, %ymm2 vtestps %ymm2, %ymm2 jne 0x50de0f movq %rsi, (%rbx) vmovaps %xmm1, %xmm2 jmp 0x50de23 cmpq $0x8, %rax je 0x50de3c movq %rax, (%rbx) vmovaps %xmm0, %xmm2 vmovaps %ymm1, %ymm0 movq %rsi, %rax incq %r9 vmovss %xmm2, 0x8(%rbx) addq $0x10, %rbx blsrq %rdi, %rdi jne 0x50dd46 jmp 0x50de45 vmovaps %ymm1, %ymm0 movq %rsi, %rax jmp 0x50de2f cmpq $0x8, %rax je 0x50de6a movb $0x1, %r10b cmpq $0x2, %r9 jae 0x50de6f movq %rax, %r15 testb %r10b, %r10b jne 0x50dc34 jmp 0x50dc0d xorl %r10d, %r10d jmp 0x50de57 xorl %r10d, %r10d jmp 0x50de54 leaq -0x20(%rbx), %rsi movl -0x18(%rbx), %r8d leaq -0x10(%rbx), %rdi cmpl -0x8(%rbx), %r8d jae 0x50deaa vmovaps (%rsi), %xmm1 vmovaps %xmm1, 0x10(%rsp) movl 0x8(%rdi), %r8d movl %r8d, 0x8(%rsi) movq (%rdi), %r8 movq %r8, (%rsi) movq 0x10(%rsp), %r8 movq %r8, (%rdi) movl 0x18(%rsp), %r8d movl %r8d, 0x8(%rdi) cmpq $0x2, %r9 pushq $0x8 popq %r8 je 0x50de54 leaq -0x30(%rbx), %r9 movl -0x28(%rbx), %r8d cmpl -0x8(%rbx), %r8d jae 0x50deec vmovaps (%r9), %xmm1 vmovaps %xmm1, 0x10(%rsp) movl 0x8(%rdi), %r8d movl %r8d, 0x8(%r9) movq (%rdi), %r8 movq %r8, (%r9) movq 0x10(%rsp), %r8 movq %r8, (%rdi) movl 0x18(%rsp), %r8d movl %r8d, 0x8(%rdi) movl -0x28(%rbx), %edi cmpl -0x18(%rbx), %edi pushq $0x8 popq %r8 jae 0x50de54 vmovaps (%r9), %xmm1 vmovaps %xmm1, 0x10(%rsp) movl 0x8(%rsi), %edi movl %edi, 0x8(%r9) movq (%rsi), %rdi movq %rdi, (%r9) movq 0x10(%rsp), %rdi movq %rdi, (%rsi) movl 0x18(%rsp), %edi movl %edi, 0x8(%rsi) jmp 0x50de54 vmovaps 0xa0(%rsp), %ymm1 vcmpnleps %ymm0, %ymm1, %ymm2 vtestps %ymm2, %ymm2 je 0x50dc0d movl %r15d, %esi andl $0xf, %esi addq $-0x8, %rsi je 0x50e082 andq $-0x10, %r15 xorl %r9d, %r9d vmovaps %ymm2, 0x3a0(%rsp) movq %rsi, 0x78(%rsp) movq (%rcx), %rax movl (%r15,%r9,8), %edi movq 0x1e8(%rax), %rax movq (%rax,%rdi,8), %rax vpbroadcastd 0x34(%rax), %ymm0 vpand 0x120(%rdx), %ymm0, %ymm0 vpcmpeqd %ymm0, %ymm14, %ymm0 vtestps %ymm2, %ymm0 jb 0x50e076 vandnps %ymm2, %ymm0, %ymm0 movq %r9, 0x80(%rsp) movl 0x4(%r15,%r9,8), %r8d vmovaps %ymm0, 0xe0(%rsp) leaq 0xe0(%rsp), %rsi movq %rsi, 0x10(%rsp) movq 0x18(%rax), %rsi movq %rsi, 0x18(%rsp) movq 0x8(%rcx), %rsi movq %rsi, 0x28(%rsp) movq %rdx, 0x30(%rsp) movl $0x8, 0x38(%rsp) movl %edi, 0x3c(%rsp) movl %r8d, 0x20(%rsp) movq %rax, 0x40(%rsp) andq $0x0, 0x48(%rsp) movq 0x10(%rcx), %rcx movq %rcx, 0x50(%rsp) movq 0x18(%rcx), %rcx testq %rcx, %rcx jne 0x50e005 movq 0x60(%rax), %rcx leaq 0x10(%rsp), %rdi vzeroupper callq *%rcx movq 0x70(%rsp), %rcx movq 0x68(%rsp), %rdx vmovaps 0x380(%rsp), %ymm11 vmovaps 0x360(%rsp), %ymm12 vmovaps 0x340(%rsp), %ymm13 vxorps %xmm14, %xmm14, %xmm14 vmovaps 0x240(%rsp), %ymm10 vmovaps 0x220(%rsp), %ymm15 vmovaps 0x200(%rsp), %ymm9 leaq 0x3c0(%rsp), %r11 pushq $0x8 popq %r8 vmovaps 0x3a0(%rsp), %ymm2 movq 0x78(%rsp), %rsi movq 0x80(%rsp), %r9 incq %r9 cmpq %r9, %rsi jne 0x50df66 vmovaps 0x100(%rdx), %ymm0 vcmpltps 0xa0(%rsp), %ymm0, %ymm1 vtestps %ymm2, %ymm1 je 0x50dc0d vmovaps 0xa0(%rsp), %ymm1 vblendvps %ymm2, %ymm0, %ymm1, %ymm1 vshufps $0xb1, %ymm1, %ymm1, %ymm0 # ymm0 = ymm1[1,0,3,2,5,4,7,6] vmovaps %ymm1, 0xa0(%rsp) vmaxps %ymm0, %ymm1, %ymm0 vshufpd $0x5, %ymm0, %ymm0, %ymm1 # ymm1 = ymm0[1,0,3,2] vmaxps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vmaxss %xmm1, %xmm0, %xmm5 jmp 0x50dbff movq 0x60(%rsp), %rsi testq %rsi, %rsi jne 0x50d7b2 leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vint8_avx2.h
embree::avx2::BVHNIntersectorKHybrid<4, 8, 1, false, embree::avx2::ArrayIntersectorK_1<8, embree::avx2::InstanceIntersectorK<8>>, false>::intersectCoherent(embree::vint_impl<8>, embree::Accel::Intersectors, embree::RayHitK<8>&, embree::RayQueryContext*)	static __forceinline vboolf8 operator ==(const vint8& a, const vint8& b) { return _mm256_castsi256_ps(_mm256_cmpeq_epi32(a, b)); }	vpcmpeqd %ymm0, %ymm0, %ymm0 vpcmpeqd (%rdi), %ymm0, %ymm1 vmovmskps %ymm1, %eax testl %eax, %eax je 0x50f6b9 pushq %rbp movq %rsp, %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx andq $-0x20, %rsp subq $0x12e0, %rsp # imm = 0x12E0 movq (%rsi), %rsi movq %rsi, 0x58(%rsp) movzbl %al, %r9d vmovaps (%rdx), %ymm7 vmovaps 0x20(%rdx), %ymm8 vmovaps 0x40(%rdx), %ymm9 vmovaps 0x80(%rdx), %ymm2 vbroadcastss 0x1a12293(%rip), %ymm3 # 0x1f20ec4 vandps %ymm3, %ymm2, %ymm4 vbroadcastss 0x19e23aa(%rip), %ymm5 # 0x1ef0fe8 vcmpltps %ymm5, %ymm4, %ymm4 vblendvps %ymm4, %ymm5, %ymm2, %ymm11 vmovaps 0xa0(%rdx), %ymm4 vandps %ymm3, %ymm4, %ymm6 vcmpltps %ymm5, %ymm6, %ymm6 vblendvps %ymm6, %ymm5, %ymm4, %ymm12 vmovaps 0xc0(%rdx), %ymm4 vandps %ymm3, %ymm4, %ymm3 vcmpltps %ymm5, %ymm3, %ymm3 vblendvps %ymm3, %ymm5, %ymm4, %ymm13 vmovaps 0x60(%rdx), %ymm3 vrcpps %ymm11, %ymm4 vbroadcastss 0x19dda8a(%rip), %ymm5 # 0x1eec714 vfnmadd213ps %ymm5, %ymm4, %ymm11 # ymm11 = -(ymm4 * ymm11) + ymm5 vfmadd132ps %ymm4, %ymm4, %ymm11 # ymm11 = (ymm11 * ymm4) + ymm4 vrcpps %ymm12, %ymm4 vfnmadd213ps %ymm5, %ymm4, %ymm12 # ymm12 = -(ymm4 * ymm12) + ymm5 vfmadd132ps %ymm4, %ymm4, %ymm12 # ymm12 = (ymm12 * ymm4) + ymm4 vrcpps %ymm13, %ymm4 vfnmadd213ps %ymm5, %ymm4, %ymm13 # ymm13 = -(ymm4 * ymm13) + ymm5 vfmadd132ps %ymm4, %ymm4, %ymm13 # ymm13 = (ymm13 * ymm4) + ymm4 vxorps %xmm4, %xmm4, %xmm4 vmaxps %ymm4, %ymm3, %ymm3 vmovaps %ymm3, 0x200(%rsp) vmovaps 0x100(%rdx), %ymm3 vmaxps %ymm4, %ymm3, %ymm3 vmovaps %ymm3, 0x2e0(%rsp) vbroadcastss 0x1a03a23(%rip), %ymm3 # 0x1f12704 vcmpltps %ymm4, %ymm2, %ymm2 vandps %ymm3, %ymm2, %ymm2 vmovaps 0xa0(%rdx), %ymm3 vmovaps 0xc0(%rdx), %ymm5 vcmpltps %ymm4, %ymm3, %ymm3 vbroadcastss 0x1a121d4(%rip), %ymm6 # 0x1f20edc vandps %ymm6, %ymm3, %ymm3 vorps %ymm2, %ymm3, %ymm2 vcmpltps %ymm4, %ymm5, %ymm3 vbroadcastss 0x1a4bc56(%rip), %ymm4 # 0x1f5a974 vandps %ymm4, %ymm3, %ymm3 vpxor %ymm0, %ymm1, %ymm0 vpor %ymm3, %ymm0, %ymm0 vpor %ymm2, %ymm0, %ymm0 vmovdqa %ymm0, 0x2a0(%rsp) vbroadcastss 0x1a12180(%rip), %ymm0 # 0x1f20ec0 vmovaps %ymm7, 0x340(%rsp) vxorps %ymm0, %ymm7, %ymm1 vmulps %ymm1, %ymm11, %ymm15 vmovaps %ymm8, 0x320(%rsp) vxorps %ymm0, %ymm8, %ymm1 vmulps %ymm1, %ymm12, %ymm10 vmovaps %ymm9, 0x300(%rsp) vxorps %ymm0, %ymm9, %ymm0 vmulps %ymm0, %ymm13, %ymm14 vbroadcastss 0x1a12144(%rip), %xmm0 # 0x1f20ec0 vmovaps %xmm0, 0x110(%rsp) leaq 0x380(%rsp), %r11 pushq $0x8 popq %r8 leaq 0x20(%rsp), %rdi leaq 0x1f(%rsp), %rsi vbroadcastss 0x19dcc7c(%rip), %ymm0 # 0x1eeba20 vmovaps %ymm0, 0x2c0(%rsp) movq %rcx, 0x70(%rsp) movq %rdx, 0x68(%rsp) vmovaps %ymm11, 0x280(%rsp) vmovaps %ymm12, 0x260(%rsp) vmovaps %ymm13, 0x240(%rsp) vmovaps %ymm15, 0x220(%rsp) vmovaps %ymm10, 0xc0(%rsp) vmovaps %ymm14, 0xa0(%rsp) tzcntq %r9, %rax vpbroadcastd 0x2a0(%rsp,%rax,4), %ymm0 vpcmpeqd 0x2a0(%rsp), %ymm0, %ymm0 vmovmskps %ymm0, %eax andnq %r9, %rax, %r9 movq %r9, 0x60(%rsp) vbroadcastss 0x19dcc04(%rip), %ymm4 # 0x1eeba20 vblendvps %ymm0, %ymm11, %ymm4, %ymm1 vshufps $0xb1, %ymm1, %ymm1, %ymm2 # ymm2 = ymm1[1,0,3,2,5,4,7,6] vminps %ymm2, %ymm1, %ymm1 vshufpd $0x5, %ymm1, %ymm1, %ymm2 # ymm2 = ymm1[1,0,3,2] vminps %ymm2, %ymm1, %ymm1 vextractf128 $0x1, %ymm1, %xmm2 vminps %xmm2, %xmm1, %xmm1 vblendvps %ymm0, %ymm12, %ymm4, %ymm2 vshufps $0xb1, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2,5,4,7,6] vminps %ymm3, %ymm2, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vminps %ymm3, %ymm2, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vminps %xmm3, %xmm2, %xmm2 vunpcklps %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] vblendvps %ymm0, %ymm13, %ymm4, %ymm2 vmovaps %ymm4, %ymm7 vshufps $0xb1, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2,5,4,7,6] vminps %ymm3, %ymm2, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vminps %ymm3, %ymm2, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vminps %xmm3, %xmm2, %xmm2 vbroadcastss 0x19ddcf1(%rip), %ymm14 # 0x1eecb84 vblendvps %ymm0, %ymm11, %ymm14, %ymm3 vinsertps $0x28, %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0,1],xmm2[0],zero vshufps $0xb1, %ymm3, %ymm3, %ymm2 # ymm2 = ymm3[1,0,3,2,5,4,7,6] vmaxps %ymm2, %ymm3, %ymm2 vshufpd $0x5, %ymm2, %ymm2, %ymm3 # ymm3 = ymm2[1,0,3,2] vmaxps %ymm3, %ymm2, %ymm2 vextractf128 $0x1, %ymm2, %xmm3 vblendvps %ymm0, %ymm12, %ymm14, %ymm4 vmaxps %xmm3, %xmm2, %xmm2 vshufps $0xb1, %ymm4, %ymm4, %ymm3 # ymm3 = ymm4[1,0,3,2,5,4,7,6] vmaxps %ymm3, %ymm4, %ymm3 vshufpd $0x5, %ymm3, %ymm3, %ymm4 # ymm4 = ymm3[1,0,3,2] vmaxps %ymm4, %ymm3, %ymm3 vextractf128 $0x1, %ymm3, %xmm4 vmaxps %xmm4, %xmm3, %xmm3 vunpcklps %xmm3, %xmm2, %xmm2 # xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1] vblendvps %ymm0, %ymm13, %ymm14, %ymm3 vshufps $0xb1, %ymm3, %ymm3, %ymm4 # ymm4 = ymm3[1,0,3,2,5,4,7,6] vmaxps %ymm4, %ymm3, %ymm3 vshufpd $0x5, %ymm3, %ymm3, %ymm4 # ymm4 = ymm3[1,0,3,2] vmaxps %ymm4, %ymm3, %ymm3 vextractf128 $0x1, %ymm3, %xmm4 vmaxps %xmm4, %xmm3, %xmm3 vinsertps $0x28, %xmm3, %xmm2, %xmm3 # xmm3 = xmm2[0,1],xmm3[0],zero vmovaps %xmm1, 0xe0(%rsp) vcmpnltps 0x19dcaf5(%rip), %xmm1, %xmm4 # 0x1eeba10 vblendvps %xmm4, %xmm1, %xmm3, %xmm2 vmovshdup %xmm2, %xmm5 # xmm5 = xmm2[1,1,3,3] xorl %r9d, %r9d vxorps %xmm6, %xmm6, %xmm6 vucomiss %xmm5, %xmm6 seta %r9b shll $0x4, %r9d orq $0x20, %r9 vshufpd $0x1, %xmm2, %xmm2, %xmm5 # xmm5 = xmm2[1,0] xorl %r14d, %r14d vucomiss %xmm5, %xmm6 seta %r14b shll $0x4, %r14d orq $0x40, %r14 movq %r9, 0x88(%rsp) xorq $0x10, %r9 movq %r9, 0x80(%rsp) xorl %r12d, %r12d vucomiss %xmm2, %xmm6 seta %r12b vmovaps 0x340(%rsp), %ymm8 vmovaps %ymm7, %ymm1 vblendvps %ymm0, %ymm8, %ymm7, %ymm5 vshufps $0xb1, %ymm5, %ymm5, %ymm6 # ymm6 = ymm5[1,0,3,2,5,4,7,6] vminps %ymm6, %ymm5, %ymm5 vshufpd $0x5, %ymm5, %ymm5, %ymm6 # ymm6 = ymm5[1,0,3,2] vminps %ymm6, %ymm5, %ymm5 vextractf128 $0x1, %ymm5, %xmm6 vminps %xmm6, %xmm5, %xmm5 vmovaps 0x320(%rsp), %ymm9 vblendvps %ymm0, %ymm9, %ymm7, %ymm6 vshufps $0xb1, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2,5,4,7,6] vminps %ymm7, %ymm6, %ymm6 vshufpd $0x5, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2] vminps %ymm7, %ymm6, %ymm6 vextractf128 $0x1, %ymm6, %xmm7 vminps %xmm7, %xmm6, %xmm6 vunpcklps %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[0],xmm6[0],xmm5[1],xmm6[1] vmovaps 0x300(%rsp), %ymm10 vblendvps %ymm0, %ymm10, %ymm1, %ymm6 vshufps $0xb1, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2,5,4,7,6] vminps %ymm7, %ymm6, %ymm6 vshufpd $0x5, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2] vminps %ymm7, %ymm6, %ymm6 vextractf128 $0x1, %ymm6, %xmm7 vminps %xmm7, %xmm6, %xmm6 vinsertps $0x28, %xmm6, %xmm5, %xmm5 # xmm5 = xmm5[0,1],xmm6[0],zero vblendvps %ymm0, %ymm8, %ymm14, %ymm6 vshufps $0xb1, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2,5,4,7,6] vmaxps %ymm7, %ymm6, %ymm6 vshufpd $0x5, %ymm6, %ymm6, %ymm7 # ymm7 = ymm6[1,0,3,2] vmaxps %ymm7, %ymm6, %ymm6 vextractf128 $0x1, %ymm6, %xmm7 vmaxps %xmm7, %xmm6, %xmm6 vblendvps %ymm0, %ymm9, %ymm14, %ymm7 vshufps $0xb1, %ymm7, %ymm7, %ymm8 # ymm8 = ymm7[1,0,3,2,5,4,7,6] vmaxps %ymm8, %ymm7, %ymm7 vshufpd $0x5, %ymm7, %ymm7, %ymm8 # ymm8 = ymm7[1,0,3,2] vmaxps %ymm8, %ymm7, %ymm7 vextractf128 $0x1, %ymm7, %xmm8 vmaxps %xmm8, %xmm7, %xmm7 vblendvps %ymm0, %ymm10, %ymm14, %ymm8 vunpcklps %xmm7, %xmm6, %xmm6 # xmm6 = xmm6[0],xmm7[0],xmm6[1],xmm7[1] vshufps $0xb1, %ymm8, %ymm8, %ymm7 # ymm7 = ymm8[1,0,3,2,5,4,7,6] vmaxps %ymm7, %ymm8, %ymm7 vshufpd $0x5, %ymm7, %ymm7, %ymm8 # ymm8 = ymm7[1,0,3,2] vmaxps %ymm8, %ymm7, %ymm7 vextractf128 $0x1, %ymm7, %xmm8 vmaxps %xmm8, %xmm7, %xmm7 vinsertps $0x28, %xmm7, %xmm6, %xmm6 # xmm6 = xmm6[0,1],xmm7[0],zero vblendvps %xmm4, %xmm5, %xmm6, %xmm7 vblendvps %xmm4, %xmm6, %xmm5, %xmm6 vblendvps %ymm0, 0x2e0(%rsp), %ymm14, %ymm14 vshufps $0xb1, %ymm14, %ymm14, %ymm5 # ymm5 = ymm14[1,0,3,2,5,4,7,6] vmaxps %ymm5, %ymm14, %ymm5 vshufpd $0x5, %ymm5, %ymm5, %ymm8 # ymm8 = ymm5[1,0,3,2] vmaxps %ymm8, %ymm5, %ymm8 vblendvps %ymm0, 0x200(%rsp), %ymm1, %ymm5 vextractf128 $0x1, %ymm8, %xmm9 vshufps $0xb1, %ymm5, %ymm5, %ymm10 # ymm10 = ymm5[1,0,3,2,5,4,7,6] vminps %ymm10, %ymm5, %ymm5 vshufpd $0x5, %ymm5, %ymm5, %ymm10 # ymm10 = ymm5[1,0,3,2] vminps %ymm10, %ymm5, %ymm5 vextractf128 $0x1, %ymm5, %xmm10 vminps %xmm10, %xmm5, %xmm5 vmovaps 0xc0(%rsp), %ymm10 vmaxss %xmm9, %xmm8, %xmm8 vmovaps 0xe0(%rsp), %xmm1 vblendvps %xmm4, %xmm3, %xmm1, %xmm1 vmulps %xmm2, %xmm6, %xmm3 vmulps %xmm1, %xmm7, %xmm4 shll $0x4, %r12d movq %r12, 0x78(%rsp) xorq $0x10, %r12 movq 0x58(%rsp), %rax movq 0x70(%rax), %rax movq %rax, 0x380(%rsp) andl $0x0, 0x388(%rsp) movq %r14, %r15 xorq $0x10, %r14 vbroadcastss %xmm2, %xmm6 vmovaps %xmm6, 0x1e0(%rsp) vbroadcastss %xmm3, %xmm6 vmovaps 0x110(%rsp), %xmm7 vxorps %xmm7, %xmm6, %xmm6 vmovaps %xmm6, 0x1d0(%rsp) vshufps $0x55, %xmm2, %xmm2, %xmm6 # xmm6 = xmm2[1,1,1,1] vmovaps %xmm6, 0x1c0(%rsp) vshufps $0x55, %xmm3, %xmm3, %xmm6 # xmm6 = xmm3[1,1,1,1] vxorps %xmm7, %xmm6, %xmm6 vmovaps %xmm6, 0x1b0(%rsp) vshufps $0xaa, %xmm2, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] vmovaps %xmm2, 0x1a0(%rsp) vshufps $0xaa, %xmm3, %xmm3, %xmm2 # xmm2 = xmm3[2,2,2,2] vxorps %xmm7, %xmm2, %xmm2 vmovaps %xmm2, 0x190(%rsp) vbroadcastss %xmm1, %xmm2 vmovaps %xmm2, 0x180(%rsp) vbroadcastss %xmm4, %xmm2 vxorps %xmm7, %xmm2, %xmm2 vmovaps %xmm2, 0x170(%rsp) vshufps $0x55, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,1,1,1] vmovaps %xmm2, 0x160(%rsp) vshufps $0x55, %xmm4, %xmm4, %xmm2 # xmm2 = xmm4[1,1,1,1] vxorps %xmm7, %xmm2, %xmm2 vmovaps %xmm2, 0x150(%rsp) vshufps $0xaa, %xmm1, %xmm1, %xmm1 # xmm1 = xmm1[2,2,2,2] vmovaps %xmm1, 0x140(%rsp) vshufps $0xaa, %xmm4, %xmm4, %xmm1 # xmm1 = xmm4[2,2,2,2] vxorps %xmm7, %xmm1, %xmm1 vmovaps %xmm1, 0x130(%rsp) vbroadcastss %xmm5, %xmm1 vmovaps %xmm1, 0x120(%rsp) vmovaps %ymm14, %ymm5 vmovaps 0x2c0(%rsp), %ymm1 vblendvps %ymm0, 0x200(%rsp), %ymm1, %ymm0 vmovaps %ymm0, 0xe0(%rsp) vmovaps 0xa0(%rsp), %ymm14 leaq 0x390(%rsp), %rbx vmovaps %xmm8, 0x90(%rsp) vbroadcastss %xmm8, %xmm0 vmovaps %xmm0, 0x40(%rsp) cmpq %r11, %rbx je 0x50f69d vbroadcastss -0x8(%rbx), %ymm1 addq $-0x10, %rbx vcmpltps %ymm5, %ymm1, %ymm2 vtestps %ymm2, %ymm2 je 0x50f242 movq (%rbx), %r13 testb $0x8, %r13b jne 0x50f560 movq 0x78(%rsp), %rax vmovaps 0x20(%r13,%rax), %xmm2 vmovaps 0x1d0(%rsp), %xmm0 vfmadd132ps 0x1e0(%rsp), %xmm0, %xmm2 # xmm2 = (xmm2 * mem) + xmm0 movq 0x88(%rsp), %rax vmovaps 0x20(%r13,%rax), %xmm3 vmovaps 0x1b0(%rsp), %xmm0 vfmadd132ps 0x1c0(%rsp), %xmm0, %xmm3 # xmm3 = (xmm3 * mem) + xmm0 vpmaxsd %xmm3, %xmm2, %xmm2 vmovaps 0x20(%r13,%r15), %xmm3 vmovaps 0x190(%rsp), %xmm0 vfmadd132ps 0x1a0(%rsp), %xmm0, %xmm3 # xmm3 = (xmm3 * mem) + xmm0 vpmaxsd 0x120(%rsp), %xmm3, %xmm3 vpmaxsd %xmm3, %xmm2, %xmm2 vmovaps 0x20(%r13,%r12), %xmm3 vmovaps 0x170(%rsp), %xmm0 vfmadd132ps 0x180(%rsp), %xmm0, %xmm3 # xmm3 = (xmm3 * mem) + xmm0 movq 0x80(%rsp), %rax vmovaps 0x20(%r13,%rax), %xmm4 vmovaps 0x150(%rsp), %xmm0 vfmadd132ps 0x160(%rsp), %xmm0, %xmm4 # xmm4 = (xmm4 * mem) + xmm0 vpminsd %xmm4, %xmm3, %xmm3 vmovaps 0x20(%r13,%r14), %xmm4 vmovaps 0x130(%rsp), %xmm0 vfmadd132ps 0x140(%rsp), %xmm0, %xmm4 # xmm4 = (xmm4 * mem) + xmm0 vpminsd 0x40(%rsp), %xmm4, %xmm4 vpminsd %xmm4, %xmm3, %xmm3 vmovdqa %xmm2, 0x1f0(%rsp) vcmpleps %xmm3, %xmm2, %xmm2 vmovmskps %xmm2, %eax testl %eax, %eax je 0x50f49d movzbl %al, %edi vbroadcastss 0x19dc6b6(%rip), %ymm1 # 0x1eeba20 movq %r8, %rax xorl %r9d, %r9d tzcntq %rdi, %rsi vbroadcastss 0x20(%r13,%rsi,4), %ymm2 vbroadcastss 0x40(%r13,%rsi,4), %ymm3 vfmadd213ps %ymm15, %ymm11, %ymm2 # ymm2 = (ymm11 * ymm2) + ymm15 vfmadd213ps %ymm10, %ymm12, %ymm3 # ymm3 = (ymm12 * ymm3) + ymm10 vbroadcastss 0x60(%r13,%rsi,4), %ymm4 vfmadd213ps %ymm14, %ymm13, %ymm4 # ymm4 = (ymm13 * ymm4) + ymm14 vmovaps %ymm5, %ymm0 vbroadcastss 0x30(%r13,%rsi,4), %ymm5 vfmadd213ps %ymm15, %ymm11, %ymm5 # ymm5 = (ymm11 * ymm5) + ymm15 vbroadcastss 0x50(%r13,%rsi,4), %ymm6 vbroadcastss 0x70(%r13,%rsi,4), %ymm7 vfmadd213ps %ymm10, %ymm12, %ymm6 # ymm6 = (ymm12 * ymm6) + ymm10 vfmadd213ps %ymm14, %ymm13, %ymm7 # ymm7 = (ymm13 * ymm7) + ymm14 vpminsd %ymm5, %ymm2, %ymm8 vpminsd %ymm6, %ymm3, %ymm9 vpmaxsd %ymm9, %ymm8, %ymm8 vpminsd %ymm7, %ymm4, %ymm9 vpmaxsd 0xe0(%rsp), %ymm9, %ymm9 vpmaxsd %ymm9, %ymm8, %ymm8 vpmaxsd %ymm5, %ymm2, %ymm2 vmovaps %ymm0, %ymm5 vpmaxsd %ymm6, %ymm3, %ymm3 vpminsd %ymm3, %ymm2, %ymm2 vpmaxsd %ymm7, %ymm4, %ymm3 vpminsd %ymm0, %ymm3, %ymm3 vpminsd %ymm3, %ymm2, %ymm2 vcmpleps %ymm2, %ymm8, %ymm2 vtestps %ymm2, %ymm2 je 0x50f45d vbroadcastss 0x1f0(%rsp,%rsi,4), %ymm2 movq (%r13,%rsi,8), %rsi prefetcht0 (%rsi) prefetcht0 0x40(%rsi) vcmpltps %ymm1, %ymm2, %ymm3 vtestps %ymm3, %ymm3 jne 0x50f43d movq %rsi, (%rbx) vmovaps %xmm2, %xmm3 jmp 0x50f451 cmpq $0x8, %rax je 0x50f46a movq %rax, (%rbx) vmovaps %xmm1, %xmm3 vmovaps %ymm2, %ymm1 movq %rsi, %rax incq %r9 vmovss %xmm3, 0x8(%rbx) addq $0x10, %rbx blsrq %rdi, %rdi jne 0x50f370 jmp 0x50f473 vmovaps %ymm2, %ymm1 movq %rsi, %rax jmp 0x50f45d cmpq $0x8, %rax je 0x50f4a2 movb $0x1, %r10b cmpq $0x2, %r9 jae 0x50f4a7 movq %rax, %r13 leaq 0x20(%rsp), %rdi leaq 0x1f(%rsp), %rsi testb %r10b, %r10b jne 0x50f264 jmp 0x50f242 xorl %r10d, %r10d jmp 0x50f48f xorl %r10d, %r10d jmp 0x50f482 leaq -0x20(%rbx), %rsi movl -0x18(%rbx), %r8d leaq -0x10(%rbx), %rdi cmpl -0x8(%rbx), %r8d jae 0x50f4e2 vmovaps (%rsi), %xmm2 vmovaps %xmm2, 0x20(%rsp) movl 0x8(%rdi), %r8d movl %r8d, 0x8(%rsi) movq (%rdi), %r8 movq %r8, (%rsi) movq 0x20(%rsp), %r8 movq %r8, (%rdi) movl 0x28(%rsp), %r8d movl %r8d, 0x8(%rdi) cmpq $0x2, %r9 pushq $0x8 popq %r8 je 0x50f482 leaq -0x30(%rbx), %r9 movl -0x28(%rbx), %r8d cmpl -0x8(%rbx), %r8d jae 0x50f524 vmovaps (%r9), %xmm2 vmovaps %xmm2, 0x20(%rsp) movl 0x8(%rdi), %r8d movl %r8d, 0x8(%r9) movq (%rdi), %r8 movq %r8, (%r9) movq 0x20(%rsp), %r8 movq %r8, (%rdi) movl 0x28(%rsp), %r8d movl %r8d, 0x8(%rdi) movl -0x28(%rbx), %edi cmpl -0x18(%rbx), %edi pushq $0x8 popq %r8 jae 0x50f482 vmovaps (%r9), %xmm2 vmovaps %xmm2, 0x20(%rsp) movl 0x8(%rsi), %edi movl %edi, 0x8(%r9) movq (%rsi), %rdi movq %rdi, (%r9) movq 0x20(%rsp), %rdi movq %rdi, (%rsi) movl 0x28(%rsp), %edi movl %edi, 0x8(%rsi) jmp 0x50f482 vcmpnleps %ymm1, %ymm5, %ymm0 vmovaps %ymm0, 0x20(%rsp) vtestps %ymm0, %ymm0 je 0x50f68f vmovaps %ymm5, 0x360(%rsp) movl %r13d, %eax andl $0xf, %eax addq $-0x8, %rax je 0x50f5c1 andq $-0x10, %r13 movq %rax, 0x40(%rsp) movq %r13, %r8 vzeroupper callq 0x5f2ef4 movq 0x40(%rsp), %rax leaq 0x1f(%rsp), %rsi leaq 0x20(%rsp), %rdi movq 0x68(%rsp), %rdx movq 0x70(%rsp), %rcx addq $0x10, %r13 decq %rax jne 0x50f58f vmovaps 0x100(%rdx), %ymm0 vmovaps 0x360(%rsp), %ymm5 vcmpltps %ymm5, %ymm0, %ymm2 vmovaps 0x20(%rsp), %ymm1 vtestps %ymm1, %ymm2 je 0x50f64d vblendvps %ymm1, %ymm0, %ymm5, %ymm5 vshufps $0xb1, %ymm5, %ymm5, %ymm0 # ymm0 = ymm5[1,0,3,2,5,4,7,6] vmaxps %ymm0, %ymm5, %ymm0 vshufpd $0x5, %ymm0, %ymm0, %ymm1 # ymm1 = ymm0[1,0,3,2] vmaxps %ymm1, %ymm0, %ymm0 vextractf128 $0x1, %ymm0, %xmm1 vmaxss %xmm1, %xmm0, %xmm8 vmovaps 0x280(%rsp), %ymm11 vmovaps 0x260(%rsp), %ymm12 vmovaps 0x240(%rsp), %ymm13 vmovaps 0x220(%rsp), %ymm15 vmovaps 0xc0(%rsp), %ymm10 vmovaps 0xa0(%rsp), %ymm14 leaq 0x380(%rsp), %r11 pushq $0x8 popq %r8 jmp 0x50f22e vmovaps 0x280(%rsp), %ymm11 vmovaps 0x260(%rsp), %ymm12 vmovaps 0x240(%rsp), %ymm13 vmovaps 0x220(%rsp), %ymm15 vmovaps 0xc0(%rsp), %ymm10 vmovaps 0xa0(%rsp), %ymm14 leaq 0x380(%rsp), %r11 pushq $0x8 popq %r8 vmovaps 0x90(%rsp), %xmm8 jmp 0x50f22e movq 0x60(%rsp), %r9 testq %r9, %r9 jne 0x50eded leaq -0x28(%rbp), %rsp popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vint8_avx2.h
embree::avx512::BVHNIntersectorKHybrid<8, 4, 1, false, embree::avx512::ArrayIntersectorK_1<4, embree::avx512::QuadMvIntersectorKMoeller<4, 4, false>>, true>::occludedCoherent(embree::vint_impl<4>, embree::Accel::Intersectors, embree::RayK<4>&, embree::RayQueryContext*)	__forceinline vboolf4 operator !=(const vint4& a, const vint4& b) { return _mm_cmp_epi32_mask(a,b,_MM_CMPINT_NE); }	vpcmpeqd %xmm0, %xmm0, %xmm0 vpcmpeqd (%rdi), %xmm0, %k1 kortestb %k1, %k1 je 0x74b449 pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x25d8, %rsp # imm = 0x25D8 movq %rcx, -0x68(%rsp) kmovd %k1, %eax movq (%rsi), %r14 vmovaps (%rdx), %xmm8 vmovaps 0x10(%rdx), %xmm9 vmovaps 0x20(%rdx), %xmm10 vmovaps 0x30(%rdx), %xmm2 movl %eax, -0x6c(%rsp) movzbl %al, %ebp vmovaps 0x40(%rdx), %xmm1 vbroadcastss 0x17d68dd(%rip), %xmm11 # 0x1f20ec4 vandps %xmm1, %xmm11, %xmm3 vbroadcastss 0x17a69f4(%rip), %xmm7 # 0x1ef0fe8 vcmpltps %xmm7, %xmm3, %k2 vblendmps %xmm7, %xmm1, %xmm4 {%k2} vmovaps 0x50(%rdx), %xmm5 vandps %xmm5, %xmm11, %xmm3 vcmpltps %xmm7, %xmm3, %k2 vmovaps %xmm7, %xmm5 {%k2} vmovaps 0x60(%rdx), %xmm6 vandps %xmm6, %xmm11, %xmm3 vcmpltps %xmm7, %xmm3, %k2 vmovaps %xmm7, %xmm6 {%k2} vrcp14ps %xmm4, %xmm3 vxorps %xmm11, %xmm11, %xmm11 vbroadcastss 0x17a20d3(%rip), %xmm7 # 0x1eec714 vfnmadd213ps %xmm7, %xmm3, %xmm4 # xmm4 = -(xmm3 * xmm4) + xmm7 vfmadd132ps %xmm3, %xmm3, %xmm4 # xmm4 = (xmm4 * xmm3) + xmm3 vrcp14ps %xmm5, %xmm3 vfnmadd213ps %xmm7, %xmm3, %xmm5 # xmm5 = -(xmm3 * xmm5) + xmm7 vfmadd132ps %xmm3, %xmm3, %xmm5 # xmm5 = (xmm5 * xmm3) + xmm3 vrcp14ps %xmm6, %xmm3 vfnmadd213ps %xmm7, %xmm3, %xmm6 # xmm6 = -(xmm3 * xmm6) + xmm7 vfmadd132ps %xmm3, %xmm3, %xmm6 # xmm6 = (xmm6 * xmm3) + xmm3 vmaxps %xmm11, %xmm2, %xmm2 vmovaps %xmm2, 0x220(%rsp) leaq 0x80(%rdx), %rax movq %rax, 0xc8(%rsp) vmovaps 0x80(%rdx), %xmm2 vmaxps %xmm11, %xmm2, %xmm2 vmovaps %xmm2, 0x210(%rsp) kmovd %k1, %eax vcmpltps %xmm11, %xmm1, %xmm1 vandps 0x17c8052(%rip){1to4}, %xmm1, %xmm2 # 0x1f12704 vcmpgtps 0x50(%rdx), %xmm11, %k2 xorb $0xf, %al movl %eax, -0x74(%rsp) vpbroadcastd 0x17d6813(%rip), %xmm3 # 0x1f20edc vpsubd %xmm1, %xmm3, %xmm2 {%k2} movq %rdx, 0xd8(%rsp) vcmpgtps 0x60(%rdx), %xmm11, %k2 vpord 0x181028b(%rip){1to4}, %xmm2, %xmm2 {%k2} # 0x1f5a974 vmovdqa32 %xmm2, %xmm0 {%k1} vmovdqa %xmm0, 0x200(%rsp) vbroadcastss 0x17d67bf(%rip), %xmm1 # 0x1f20ec0 vmovaps %xmm8, 0x250(%rsp) vxorps %xmm1, %xmm8, %xmm0 vmulps %xmm0, %xmm4, %xmm12 vmovaps %xmm9, 0x240(%rsp) vxorps %xmm1, %xmm9, %xmm0 vmulps %xmm0, %xmm5, %xmm13 vmovaps %xmm10, 0x230(%rsp) vxorps %xmm1, %xmm10, %xmm0 vmulps %xmm0, %xmm6, %xmm14 movabsq $0x8, %rbx movq %r14, 0xd0(%rsp) tzcntq %rbp, %rax vpbroadcastd 0x200(%rsp,%rax,4), %xmm0 vpcmpeqd 0x200(%rsp), %xmm0, %k1 kmovb %k1, %edx vbroadcastss 0x17a12b5(%rip), %xmm20 # 0x1eeba20 vblendmps %xmm4, %xmm20, %xmm0 {%k1} vblendmps %xmm5, %xmm20, %xmm1 {%k1} vblendmps %xmm6, %xmm20, %xmm2 {%k1} vbroadcastss 0x17a23fd(%rip), %xmm18 # 0x1eecb84 vblendmps %xmm4, %xmm18, %xmm3 {%k1} vblendmps %xmm5, %xmm18, %xmm7 {%k1} vblendmps %xmm6, %xmm18, %xmm8 {%k1} vmovaps 0x250(%rsp), %xmm15 vblendmps %xmm15, %xmm20, %xmm9 {%k1} vmovaps 0x240(%rsp), %xmm16 vblendmps %xmm16, %xmm20, %xmm10 {%k1} vmovaps 0x230(%rsp), %xmm17 vblendmps %xmm17, %xmm20, %xmm11 {%k1} vblendmps %xmm15, %xmm18, %xmm15 {%k1} vblendmps %xmm16, %xmm18, %xmm16 {%k1} vblendmps %xmm17, %xmm18, %xmm17 {%k1} vblendmps 0x210(%rsp), %xmm18, %xmm21 {%k1} vmovaps 0x220(%rsp), %xmm19 vblendmps %xmm19, %xmm20, %xmm18 {%k1} vpbroadcastd 0x17a122a(%rip), %xmm20 # 0x1eeba20 vpblendmd %xmm19, %xmm20, %xmm22 {%k1} kmovd %k1, %eax movl %eax, -0x70(%rsp) andnq %rbp, %rdx, %rbp vshufps $0xb1, %xmm0, %xmm0, %xmm19 # xmm19 = xmm0[1,0,3,2] vminps %xmm0, %xmm19, %xmm0 vshufpd $0x1, %xmm0, %xmm0, %xmm19 # xmm19 = xmm0[1,0] vminps %xmm0, %xmm19, %xmm0 vshufps $0xb1, %xmm1, %xmm1, %xmm19 # xmm19 = xmm1[1,0,3,2] vminps %xmm1, %xmm19, %xmm1 vshufpd $0x1, %xmm1, %xmm1, %xmm19 # xmm19 = xmm1[1,0] vminps %xmm1, %xmm19, %xmm1 vinsertps $0x1c, %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[0],xmm1[0],zero,zero vshufps $0xb1, %xmm2, %xmm2, %xmm1 # xmm1 = xmm2[1,0,3,2] vminps %xmm2, %xmm1, %xmm1 vshufpd $0x1, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,0] vminps %xmm1, %xmm2, %xmm1 vinsertps $0x20, %xmm1, %xmm0, %xmm0 # xmm0 = xmm0[0,1],xmm1[0],xmm0[3] vshufps $0xb1, %xmm3, %xmm3, %xmm1 # xmm1 = xmm3[1,0,3,2] vmaxps %xmm3, %xmm1, %xmm1 vshufpd $0x1, %xmm1, %xmm1, %xmm2 # xmm2 = xmm1[1,0] vmaxps %xmm1, %xmm2, %xmm1 vshufps $0xb1, %xmm7, %xmm7, %xmm2 # xmm2 = xmm7[1,0,3,2] vmaxps %xmm7, %xmm2, %xmm2 vshufpd $0x1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0] vmaxps %xmm2, %xmm3, %xmm2 vinsertps $0x1c, %xmm2, %xmm1, %xmm1 # xmm1 = xmm1[0],xmm2[0],zero,zero vshufps $0xb1, %xmm8, %xmm8, %xmm2 # xmm2 = xmm8[1,0,3,2] vmaxps %xmm8, %xmm2, %xmm2 vshufpd $0x1, %xmm2, %xmm2, %xmm3 # xmm3 = xmm2[1,0] vmaxps %xmm2, %xmm3, %xmm2 vinsertps $0x20, %xmm2, %xmm1, %xmm3 # xmm3 = xmm1[0,1],xmm2[0],xmm1[3] vxorps %xmm1, %xmm1, %xmm1 vcmpnltps %xmm1, %xmm0, %k1 vblendmps %xmm0, %xmm3, %xmm1 {%k1} vmovshdup %xmm1, %xmm2 # xmm2 = xmm1[1,1,3,3] xorl %r15d, %r15d vxorps %xmm8, %xmm8, %xmm8 vucomiss %xmm2, %xmm8 seta %r15b shll $0x5, %r15d orq $0x40, %r15 vshufpd $0x1, %xmm1, %xmm1, %xmm7 # xmm7 = xmm1[1,0] xorl %eax, %eax vucomiss %xmm7, %xmm8 seta %al shll $0x5, %eax orq $0x80, %rax movq %r15, %r13 xorq $0x20, %r13 xorl %ecx, %ecx vucomiss %xmm1, %xmm8 seta %cl vshufps $0xb1, %xmm9, %xmm9, %xmm7 # xmm7 = xmm9[1,0,3,2] vminps %xmm9, %xmm7, %xmm7 vshufpd $0x1, %xmm7, %xmm7, %xmm8 # xmm8 = xmm7[1,0] vminps %xmm7, %xmm8, %xmm7 vshufps $0xb1, %xmm10, %xmm10, %xmm8 # xmm8 = xmm10[1,0,3,2] vminps %xmm10, %xmm8, %xmm8 vshufpd $0x1, %xmm8, %xmm8, %xmm9 # xmm9 = xmm8[1,0] vminps %xmm8, %xmm9, %xmm8 vinsertps $0x1c, %xmm8, %xmm7, %xmm7 # xmm7 = xmm7[0],xmm8[0],zero,zero vshufps $0xb1, %xmm11, %xmm11, %xmm8 # xmm8 = xmm11[1,0,3,2] vminps %xmm11, %xmm8, %xmm8 vshufpd $0x1, %xmm8, %xmm8, %xmm9 # xmm9 = xmm8[1,0] vminps %xmm8, %xmm9, %xmm8 vinsertps $0x20, %xmm8, %xmm7, %xmm7 # xmm7 = xmm7[0,1],xmm8[0],xmm7[3] vshufps $0xb1, %xmm15, %xmm15, %xmm8 # xmm8 = xmm15[1,0,3,2] vmaxps %xmm15, %xmm8, %xmm8 vshufpd $0x1, %xmm8, %xmm8, %xmm9 # xmm9 = xmm8[1,0] vmaxps %xmm8, %xmm9, %xmm8 vshufps $0xb1, %xmm16, %xmm16, %xmm9 # xmm9 = xmm16[1,0,3,2] vmaxps %xmm16, %xmm9, %xmm9 vshufpd $0x1, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,0] vmaxps %xmm9, %xmm10, %xmm9 vinsertps $0x1c, %xmm9, %xmm8, %xmm8 # xmm8 = xmm8[0],xmm9[0],zero,zero vshufps $0xb1, %xmm17, %xmm17, %xmm9 # xmm9 = xmm17[1,0,3,2] vmaxps %xmm17, %xmm9, %xmm9 vshufpd $0x1, %xmm9, %xmm9, %xmm10 # xmm10 = xmm9[1,0] vmaxps %xmm9, %xmm10, %xmm9 vinsertps $0x20, %xmm9, %xmm8, %xmm8 # xmm8 = xmm8[0,1],xmm9[0],xmm8[3] vblendmps %xmm7, %xmm8, %xmm9 {%k1} vmovaps %xmm8, %xmm7 {%k1} vshufps $0xb1, %xmm21, %xmm21, %xmm8 # xmm8 = xmm21[1,0,3,2] vmaxps %xmm21, %xmm8, %xmm8 vshufpd $0x1, %xmm8, %xmm8, %xmm10 # xmm10 = xmm8[1,0] vshufps $0xb1, %xmm18, %xmm18, %xmm11 # xmm11 = xmm18[1,0,3,2] vminps %xmm18, %xmm11, %xmm11 vshufpd $0x1, %xmm11, %xmm11, %xmm15 # xmm15 = xmm11[1,0] vminps %xmm11, %xmm15, %xmm11 vmaxps %xmm8, %xmm10, %xmm8 vmovaps %xmm3, %xmm0 {%k1} vmulps %xmm1, %xmm7, %xmm3 vmulps %xmm0, %xmm9, %xmm7 shll $0x5, %ecx movq %rcx, %r8 xorq $0x20, %r8 movq %rax, %r12 xorq $0x20, %r12 movq 0x70(%r14), %rsi movq %rsi, 0x290(%rsp) movq %rdx, 0x298(%rsp) vbroadcastss %xmm1, %ymm15 vbroadcastss %xmm3, %ymm9 vbroadcastss 0x17d64ad(%rip), %ymm10 # 0x1f20ec0 vxorps %ymm10, %ymm9, %ymm16 vbroadcastsd %xmm2, %ymm17 vbroadcastss 0x17c7cdc(%rip), %ymm9 # 0x1f12704 vpermps %ymm3, %ymm9, %ymm2 vxorps %ymm10, %ymm2, %ymm18 vbroadcastss 0x17d64a0(%rip), %ymm2 # 0x1f20edc vpermps %ymm1, %ymm2, %ymm19 vpermps %ymm3, %ymm2, %ymm1 vxorps %ymm10, %ymm1, %ymm20 vbroadcastss %xmm0, %ymm23 vbroadcastss %xmm7, %ymm1 vxorps %ymm10, %ymm1, %ymm24 vpermps %ymm0, %ymm9, %ymm25 vpermps %ymm7, %ymm9, %ymm1 vxorps %ymm10, %ymm1, %ymm26 vpermps %ymm0, %ymm2, %ymm27 vpermps %ymm7, %ymm2, %ymm0 vxorps %ymm10, %ymm0, %ymm28 vbroadcastss %xmm11, %ymm11 vbroadcastss %xmm8, %ymm29 leaq 0x2a0(%rsp), %rsi leaq 0x290(%rsp), %rdx cmpq %rdx, %rsi je 0x74b40c leaq -0x10(%rsi), %r10 movzbl -0x74(%rsp), %edx andnq -0x8(%rsi), %rdx, %rdx je 0x74b3f9 movq (%r10), %rsi testb $0x8, %sil jne 0x74ac52 vmovaps 0x40(%rsi,%rcx), %ymm0 vfmadd132ps %ymm15, %ymm16, %ymm0 # ymm0 = (ymm0 * ymm15) + ymm16 vmovaps 0x40(%rsi,%r15), %ymm1 vfmadd132ps %ymm17, %ymm18, %ymm1 # ymm1 = (ymm1 * ymm17) + ymm18 vmovaps 0x40(%rsi,%rax), %ymm2 vpmaxsd %ymm1, %ymm0, %ymm0 vfmadd132ps %ymm19, %ymm20, %ymm2 # ymm2 = (ymm2 * ymm19) + ymm20 vpmaxsd %ymm11, %ymm2, %ymm1 vpmaxsd %ymm1, %ymm0, %ymm0 vmovaps 0x40(%rsi,%r8), %ymm1 vfmadd132ps %ymm23, %ymm24, %ymm1 # ymm1 = (ymm1 * ymm23) + ymm24 vmovaps 0x40(%rsi,%r13), %ymm2 vfmadd132ps %ymm25, %ymm26, %ymm2 # ymm2 = (ymm2 * ymm25) + ymm26 vmovaps 0x40(%rsi,%r12), %ymm3 vpminsd %ymm2, %ymm1, %ymm1 vfmadd132ps %ymm27, %ymm28, %ymm3 # ymm3 = (ymm3 * ymm27) + ymm28 vpminsd %ymm29, %ymm3, %ymm2 vpminsd %ymm2, %ymm1, %ymm1 vcmpleps %ymm1, %ymm0, %k0 kortestb %k0, %k0 je 0x74ac4e kmovd %k0, %edx movzbl %dl, %edi movq %rbx, %r9 xorl %edx, %edx tzcntq %rdi, %r11 vbroadcastss 0x40(%rsi,%r11,4), %xmm0 vfmadd132ps %xmm4, %xmm12, %xmm0 # xmm0 = (xmm0 * xmm4) + xmm12 vbroadcastss 0x80(%rsi,%r11,4), %xmm1 vfmadd132ps %xmm5, %xmm13, %xmm1 # xmm1 = (xmm1 * xmm5) + xmm13 vbroadcastss 0xc0(%rsi,%r11,4), %xmm2 vfmadd132ps %xmm6, %xmm14, %xmm2 # xmm2 = (xmm2 * xmm6) + xmm14 vbroadcastss 0x60(%rsi,%r11,4), %xmm3 vfmadd132ps %xmm4, %xmm12, %xmm3 # xmm3 = (xmm3 * xmm4) + xmm12 vbroadcastss 0xa0(%rsi,%r11,4), %xmm7 vfmadd132ps %xmm5, %xmm13, %xmm7 # xmm7 = (xmm7 * xmm5) + xmm13 vbroadcastss 0xe0(%rsi,%r11,4), %xmm8 vfmadd132ps %xmm6, %xmm14, %xmm8 # xmm8 = (xmm8 * xmm6) + xmm14 vpminsd %xmm3, %xmm0, %xmm9 vpminsd %xmm7, %xmm1, %xmm10 vpmaxsd %xmm10, %xmm9, %xmm9 vpminsd %xmm8, %xmm2, %xmm10 vpmaxsd %xmm22, %xmm10, %xmm10 vpmaxsd %xmm10, %xmm9, %xmm9 vpmaxsd %xmm3, %xmm0, %xmm0 vpmaxsd %xmm7, %xmm1, %xmm1 vpminsd %xmm1, %xmm0, %xmm0 vpmaxsd %xmm8, %xmm2, %xmm1 vpminsd %xmm21, %xmm1, %xmm1 vpminsd %xmm1, %xmm0, %xmm0 vpcmpled %xmm0, %xmm9, %k0 kortestb %k0, %k0 je 0x74ac2a movq (%rsi,%r11,8), %r11 prefetcht0 (%r11) prefetcht0 0x40(%r11) prefetcht0 0x80(%r11) prefetcht0 0xc0(%r11) cmpq $0x8, %r9 je 0x74ac20 movq %r9, (%r10) movq %rdx, 0x8(%r10) addq $0x10, %r10 kmovd %k0, %edx movzbl %dl, %edx movq %r11, %r9 blsrq %rdi, %rdi jne 0x74ab4e cmpq $0x8, %r9 setne %dil movq %r9, %rsi testb %dil, %dil jne 0x74aabc jmp 0x74b3f9 xorl %edi, %edi jmp 0x74ac40 testq %rdx, %rdx je 0x74b3f9 movl %esi, %edi andl $0xf, %edi movl -0x74(%rsp), %edx movl %edx, %r9d andb $0xf, %r9b addq $-0x8, %rdi movq %rdi, 0xe8(%rsp) je 0x74b3cb movq %rbp, 0xe0(%rsp) andq $-0x10, %rsi xorb $0xf, %r9b movq 0xd8(%rsp), %rdx vmovaps (%rdx), %xmm0 vmovaps %xmm0, 0x1f0(%rsp) vmovaps 0x10(%rdx), %xmm0 vmovaps %xmm0, 0x1e0(%rsp) vmovaps 0x20(%rdx), %xmm0 vmovaps %xmm0, 0x1d0(%rsp) vmovaps 0x30(%rdx), %xmm0 vmovaps %xmm0, 0x110(%rsp) vmovaps 0x50(%rdx), %xmm0 vmovaps %xmm0, 0x1c0(%rsp) vmovaps 0x60(%rdx), %xmm0 vmovaps %xmm0, 0x1b0(%rsp) vmovaps 0x40(%rdx), %xmm0 vmovaps %xmm0, 0x1a0(%rsp) vmovaps 0x80(%rdx), %xmm0 vmovaps %xmm0, 0x100(%rsp) vmovaps 0x90(%rdx), %xmm0 vmovaps %xmm0, 0xf0(%rsp) addq $0xc0, %rsi xorl %edi, %edi vmovups %ymm15, 0xa0(%rsp) vmovups %ymm16, 0x80(%rsp) vmovups %ymm17, 0x60(%rsp) vmovups %ymm18, 0x40(%rsp) vmovups %ymm19, 0x20(%rsp) vmovups %ymm20, (%rsp) vmovups %ymm23, -0x20(%rsp) vmovups %ymm24, -0x40(%rsp) vmovups %ymm25, -0x60(%rsp) vmovups %ymm26, 0x180(%rsp) vmovups %ymm27, 0x160(%rsp) vmovups %ymm28, 0x140(%rsp) vmovups %ymm11, 0x270(%rsp) vmovups %ymm29, 0x120(%rsp) xorl %edx, %edx movb %r9b, -0x75(%rsp) movl %r9d, %ebp movl (%rsi,%rdx,4), %r9d movl $0xffffffff, %r11d # imm = 0xFFFFFFFF cmpq %r11, %r9 je 0x74b395 vbroadcastss -0xc0(%rsi,%rdx,4), %xmm3 vbroadcastss -0xb0(%rsi,%rdx,4), %xmm9 vbroadcastss -0xa0(%rsi,%rdx,4), %xmm17 vbroadcastss -0x90(%rsi,%rdx,4), %xmm11 vbroadcastss -0x80(%rsi,%rdx,4), %xmm15 vbroadcastss -0x70(%rsi,%rdx,4), %xmm7 vbroadcastss -0x30(%rsi,%rdx,4), %xmm10 vbroadcastss -0x20(%rsi,%rdx,4), %xmm0 vbroadcastss -0x10(%rsi,%rdx,4), %xmm1 vsubss %xmm11, %xmm3, %xmm19 vbroadcastss %xmm19, %xmm26 vsubss %xmm15, %xmm9, %xmm2 vbroadcastss %xmm2, %xmm29 vmovaps %xmm7, 0x260(%rsp) vsubss %xmm7, %xmm17, %xmm20 vbroadcastss %xmm20, %xmm31 vsubss %xmm3, %xmm10, %xmm23 vbroadcastss %xmm23, %xmm8 vsubss %xmm9, %xmm0, %xmm24 vbroadcastss %xmm24, %xmm18 vsubss %xmm17, %xmm1, %xmm25 vbroadcastss %xmm25, %xmm16 vmulss %xmm25, %xmm2, %xmm2 vbroadcastss %xmm2, %xmm2 vfmsub231ps %xmm31, %xmm18, %xmm2 # xmm2 = (xmm18 * xmm31) - xmm2 vmulss %xmm23, %xmm20, %xmm20 vbroadcastss %xmm20, %xmm23 vfmsub231ps %xmm26, %xmm16, %xmm23 # xmm23 = (xmm16 * xmm26) - xmm23 vmulss %xmm24, %xmm19, %xmm19 vbroadcastss %xmm19, %xmm27 vfmsub231ps %xmm29, %xmm8, %xmm27 # xmm27 = (xmm8 * xmm29) - xmm27 vsubps 0x1f0(%rsp), %xmm3, %xmm24 vsubps 0x1e0(%rsp), %xmm9, %xmm28 vsubps 0x1d0(%rsp), %xmm17, %xmm30 vmovaps 0x1c0(%rsp), %xmm3 vmulps %xmm30, %xmm3, %xmm20 vmovaps 0x1b0(%rsp), %xmm17 vfmsub231ps %xmm17, %xmm28, %xmm20 # xmm20 = (xmm28 * xmm17) - xmm20 vmulps %xmm24, %xmm17, %xmm9 vmovaps 0x1a0(%rsp), %xmm7 vfmsub231ps %xmm7, %xmm30, %xmm9 # xmm9 = (xmm30 * xmm7) - xmm9 vmulps %xmm28, %xmm7, %xmm19 vfmsub231ps %xmm3, %xmm24, %xmm19 # xmm19 = (xmm24 * xmm3) - xmm19 vmulps %xmm17, %xmm27, %xmm25 vfmadd231ps %xmm3, %xmm23, %xmm25 # xmm25 = (xmm23 * xmm3) + xmm25 vfmadd231ps %xmm7, %xmm2, %xmm25 # xmm25 = (xmm2 * xmm7) + xmm25 vandps 0x17d5fd1(%rip){1to4}, %xmm25, %xmm3 # 0x1f20ec0 vmulps %xmm19, %xmm16, %xmm16 vfmadd231ps %xmm18, %xmm9, %xmm16 # xmm16 = (xmm9 * xmm18) + xmm16 vfmadd231ps %xmm8, %xmm20, %xmm16 # xmm16 = (xmm20 * xmm8) + xmm16 vxorps %xmm16, %xmm3, %xmm17 vxorps %xmm8, %xmm8, %xmm8 vcmpnltps %xmm8, %xmm17, %k0 kmovd %k0, %r14d andb %bpl, %r14b jne 0x74b113 testb %bpl, %bpl vmovups 0x180(%rsp), %ymm26 vmovups 0x160(%rsp), %ymm27 vmovups 0x140(%rsp), %ymm28 vmovups 0x120(%rsp), %ymm29 je 0x74b0a0 vbroadcastss -0x60(%rsi,%rdx,4), %xmm2 vbroadcastss -0x50(%rsi,%rdx,4), %xmm3 vbroadcastss -0x40(%rsi,%rdx,4), %xmm8 vsubps %xmm10, %xmm2, %xmm23 vsubps %xmm0, %xmm3, %xmm24 vsubps %xmm1, %xmm8, %xmm25 vsubps %xmm2, %xmm11, %xmm16 vsubps %xmm3, %xmm15, %xmm17 vmovaps 0x260(%rsp), %xmm0 vsubps %xmm8, %xmm0, %xmm18 vmulps %xmm18, %xmm24, %xmm0 vfmsub231ps %xmm25, %xmm17, %xmm0 # xmm0 = (xmm17 * xmm25) - xmm0 vmulps %xmm16, %xmm25, %xmm1 vfmsub231ps %xmm23, %xmm18, %xmm1 # xmm1 = (xmm18 * xmm23) - xmm1 vmulps %xmm17, %xmm23, %xmm10 vfmsub231ps %xmm24, %xmm16, %xmm10 # xmm10 = (xmm16 * xmm24) - xmm10 vsubps 0x1f0(%rsp), %xmm2, %xmm2 vsubps 0x1e0(%rsp), %xmm3, %xmm11 vsubps 0x1d0(%rsp), %xmm8, %xmm15 vmovaps 0x1c0(%rsp), %xmm3 vmulps %xmm15, %xmm3, %xmm20 vmovaps 0x1b0(%rsp), %xmm7 vfmsub231ps %xmm7, %xmm11, %xmm20 # xmm20 = (xmm11 * xmm7) - xmm20 vmulps %xmm2, %xmm7, %xmm9 vmovaps 0x1a0(%rsp), %xmm8 vfmsub231ps %xmm8, %xmm15, %xmm9 # xmm9 = (xmm15 * xmm8) - xmm9 vmulps %xmm11, %xmm8, %xmm19 vfmsub231ps %xmm3, %xmm2, %xmm19 # xmm19 = (xmm2 * xmm3) - xmm19 vmulps %xmm7, %xmm10, %xmm7 vfmadd231ps %xmm3, %xmm1, %xmm7 # xmm7 = (xmm1 * xmm3) + xmm7 vfmadd231ps %xmm8, %xmm0, %xmm7 # xmm7 = (xmm0 * xmm8) + xmm7 vandps 0x17d5ea2(%rip){1to4}, %xmm7, %xmm3 # 0x1f20ec0 vmulps %xmm19, %xmm18, %xmm8 vfmadd231ps %xmm17, %xmm9, %xmm8 # xmm8 = (xmm9 * xmm17) + xmm8 vfmadd231ps %xmm16, %xmm20, %xmm8 # xmm8 = (xmm20 * xmm16) + xmm8 vxorps %xmm8, %xmm3, %xmm17 vxorps %xmm8, %xmm8, %xmm8 vcmpnltps %xmm8, %xmm17, %k0 kmovd %k0, %r14d andb %bpl, %r14b jne 0x74b204 vmovups 0xa0(%rsp), %ymm15 vmovups 0x80(%rsp), %ymm16 vmovups 0x60(%rsp), %ymm17 vmovups 0x40(%rsp), %ymm18 vmovups 0x20(%rsp), %ymm19 vmovups (%rsp), %ymm20 vmovups -0x20(%rsp), %ymm23 vmovups -0x40(%rsp), %ymm24 vmovups -0x60(%rsp), %ymm25 testb %bpl, %bpl setne %r9b jmp 0x74b0eb xorl %r9d, %r9d vmovups 0xa0(%rsp), %ymm15 vmovups 0x80(%rsp), %ymm16 vmovups 0x60(%rsp), %ymm17 vmovups 0x40(%rsp), %ymm18 vmovups 0x20(%rsp), %ymm19 vmovups (%rsp), %ymm20 vmovups -0x20(%rsp), %ymm23 vmovups -0x40(%rsp), %ymm24 vmovups -0x60(%rsp), %ymm25 vmovups 0x270(%rsp), %ymm11 testb %r9b, %r9b je 0x74b395 leaq 0x1(%rdx), %r9 cmpq $0x3, %rdx movq %r9, %rdx jb 0x74ad9c jmp 0x74b395 vmulps %xmm19, %xmm31, %xmm8 vfmadd213ps %xmm8, %xmm9, %xmm29 # xmm29 = (xmm9 * xmm29) + xmm8 vfmadd213ps %xmm29, %xmm20, %xmm26 # xmm26 = (xmm20 * xmm26) + xmm29 vxorps %xmm26, %xmm3, %xmm19 vxorps %xmm8, %xmm8, %xmm8 vcmpnltps %xmm8, %xmm19, %k0 kmovd %k0, %r11d andb %r11b, %r14b je 0x74af20 vandps 0x17d5d76(%rip){1to4}, %xmm25, %xmm9 # 0x1f20ec4 vsubps %xmm17, %xmm9, %xmm8 vcmpnltps %xmm19, %xmm8, %k0 kmovd %k0, %r11d andb %r11b, %r14b je 0x74af20 vmulps %xmm30, %xmm27, %xmm8 vfmadd213ps %xmm8, %xmm23, %xmm28 # xmm28 = (xmm23 * xmm28) + xmm8 vfmadd213ps %xmm28, %xmm2, %xmm24 # xmm24 = (xmm2 * xmm24) + xmm28 vxorps %xmm24, %xmm3, %xmm2 vmulps 0x110(%rsp), %xmm9, %xmm3 vmulps 0x100(%rsp), %xmm9, %xmm8 vcmpleps %xmm8, %xmm2, %k1 vcmpltps %xmm2, %xmm3, %k0 {%k1} kmovd %k0, %r11d andb %r11b, %r14b je 0x74af20 vxorps %xmm2, %xmm2, %xmm2 vcmpneqps %xmm2, %xmm25, %k0 kmovd %k0, %r11d andb %r11b, %r14b je 0x74af20 movq -0x68(%rsp), %r11 movq (%r11), %r11 movq 0x1e8(%r11), %r11 movq (%r11,%r9,8), %r11 vmovdqa 0xf0(%rsp), %xmm2 vptestmd 0x34(%r11){1to4}, %xmm2, %k0 kmovd %k0, %r11d andb %r11b, %r14b je 0x74af20 xorb $0xf, %r14b andb %bpl, %r14b movl %r14d, %ebp jmp 0x74af20 vmulps %xmm19, %xmm25, %xmm8 vfmadd213ps %xmm8, %xmm9, %xmm24 # xmm24 = (xmm9 * xmm24) + xmm8 vfmadd213ps %xmm24, %xmm20, %xmm23 # xmm23 = (xmm20 * xmm23) + xmm24 vxorps %xmm23, %xmm3, %xmm19 vxorps %xmm8, %xmm8, %xmm8 vcmpnltps %xmm8, %xmm19, %k0 kmovd %k0, %r11d andb %r11b, %r14b je 0x74b04f vandps 0x17d5c85(%rip){1to4}, %xmm7, %xmm9 # 0x1f20ec4 vsubps %xmm17, %xmm9, %xmm8 vcmpnltps %xmm19, %xmm8, %k0 kmovd %k0, %r11d andb %r11b, %r14b vmovups (%rsp), %ymm20 vmovups -0x20(%rsp), %ymm23 vmovups -0x40(%rsp), %ymm24 vmovups -0x60(%rsp), %ymm25 jne 0x74b2a2 vmovups 0xa0(%rsp), %ymm15 vmovups 0x80(%rsp), %ymm16 vmovups 0x60(%rsp), %ymm17 vmovups 0x40(%rsp), %ymm18 vmovups 0x20(%rsp), %ymm19 jmp 0x74b097 vmulps %xmm15, %xmm10, %xmm8 vfmadd213ps %xmm8, %xmm1, %xmm11 # xmm11 = (xmm1 * xmm11) + xmm8 vfmadd213ps %xmm11, %xmm0, %xmm2 # xmm2 = (xmm0 * xmm2) + xmm11 vxorps %xmm2, %xmm3, %xmm0 vmulps 0x110(%rsp), %xmm9, %xmm1 vmulps 0x100(%rsp), %xmm9, %xmm2 vcmpleps %xmm2, %xmm0, %k1 vcmpltps %xmm0, %xmm1, %k0 {%k1} kmovd %k0, %r11d andb %r11b, %r14b je 0x74b328 vxorps %xmm0, %xmm0, %xmm0 vcmpneqps %xmm0, %xmm7, %k0 kmovd %k0, %r11d andb %r11b, %r14b je 0x74b328 movq -0x68(%rsp), %r11 movq (%r11), %r11 movq 0x1e8(%r11), %r11 movq (%r11,%r9,8), %r9 vmovdqa 0xf0(%rsp), %xmm0 vptestmd 0x34(%r9){1to4}, %xmm0, %k0 kmovd %k0, %r9d andb %r9b, %r14b je 0x74b328 xorb $0xf, %r14b andb %bpl, %r14b movl %r14d, %ebp vmovups 0xa0(%rsp), %ymm15 vmovups 0x80(%rsp), %ymm16 vmovups 0x60(%rsp), %ymm17 vmovups 0x40(%rsp), %ymm18 vmovups 0x20(%rsp), %ymm19 vmovups (%rsp), %ymm20 vmovups -0x20(%rsp), %ymm23 vmovups -0x40(%rsp), %ymm24 vmovups -0x60(%rsp), %ymm25 vmovups 0x180(%rsp), %ymm26 vmovups 0x160(%rsp), %ymm27 vmovups 0x140(%rsp), %ymm28 vmovups 0x120(%rsp), %ymm29 jmp 0x74b097 movb -0x75(%rsp), %r9b andb %bpl, %r9b je 0x74b3b7 incq %rdi addq $0xe0, %rsi cmpq 0xe8(%rsp), %rdi jb 0x74ad92 xorb $0xf, %r9b movq 0xd0(%rsp), %r14 movq 0xe0(%rsp), %rbp movl -0x74(%rsp), %esi orb %r9b, %sil movl %esi, %edx notb %dl movl -0x70(%rsp), %edi andb %dl, %dil andb $0xf, %dil movl %edi, -0x70(%rsp) movl %esi, -0x74(%rsp) je 0x74b408 kmovd %esi, %k1 vbroadcastss 0x17a178b(%rip), %xmm21 {%k1} # 0x1eecb84 xorl %edx, %edx movq %r10, %rsi testb %dl, %dl je 0x74aa93 jmp 0x74b40c movb $0x1, %dl jmp 0x74b3fb testq %rbp, %rbp jne 0x74a746 movl -0x74(%rsp), %eax andb -0x6c(%rsp), %al kmovd %eax, %k1 movq 0xc8(%rsp), %rax vbroadcastss 0x17a1752(%rip), %xmm0 # 0x1eecb84 vmovaps %xmm0, (%rax) {%k1} addq $0x25d8, %rsp # imm = 0x25D8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp vzeroupper retq nop	/embree[P]embree/kernels/bvh/../common/../../common/sys/../math/../simd/vint4_sse2.h
void std::__move_median_to_first<embree::sse2::BVHBuilderMorton::BuildPrim, __gnu_cxx::__ops::_Iter_comp_iter<std::less<embree::sse2::BVHBuilderMorton::BuildPrim>>>(embree::sse2::BVHBuilderMorton::BuildPrim, embree::sse2::BVHBuilderMorton::BuildPrim, embree::sse2::BVHBuilderMorton::BuildPrim, embree::sse2::BVHBuilderMorton::BuildPrim*, __gnu_cxx::__ops::_Iter_comp_iter<std::less<embree::sse2::BVHBuilderMorton::BuildPrim>>)	__forceinline bool operator<(const BuildPrim &m) const { return code < m.code; }	movl (%rsi), %r8d movl (%rdx), %r9d movl (%rcx), %eax cmpl %r9d, %r8d jae 0x90e361 cmpl %eax, %r9d jae 0x90e373 movq (%rdi), %rax movq (%rdx), %rcx movq %rcx, (%rdi) movq %rax, (%rdx) retq cmpl %eax, %r8d jae 0x90e385 movq (%rdi), %rax movq (%rsi), %rcx movq %rcx, (%rdi) movq %rax, (%rsi) retq movq (%rdi), %rdx cmpl %eax, %r8d jae 0x90e397 movq (%rcx), %rax movq %rax, (%rdi) movq %rdx, (%rcx) retq movq (%rdi), %rsi cmpl %eax, %r9d jae 0x90e3a1 movq (%rcx), %rax movq %rax, (%rdi) movq %rsi, (%rcx) retq movq (%rsi), %rax movq %rax, (%rdi) movq %rdx, (%rsi) retq movq (%rdx), %rax movq %rax, (%rdi) movq %rsi, (%rdx) retq	/embree[P]embree/kernels/common/../builders/bvh_builder_morton.h
void embree::sse2::CurveNiMBIntersector1<4>::intersect_n<embree::sse2::OrientedCurve1Intersector1<embree::CatmullRomCurveT, 3, 4>, embree::sse2::Intersect1Epilog1<true>>(embree::sse2::CurvePrecalculations1 const&, embree::RayHitK<1>&, embree::RayQueryContext*, embree::CurveNiMB<4> const&)	static __forceinline void intersect_n(const Precalculations& pre, RayHit& ray, RayQueryContext* context, const Primitive& prim) { vfloat<M> tNear; vbool<M> valid = intersect(ray,prim,tNear); const size_t N = prim.N; size_t mask = movemask(valid); while (mask) { const size_t i = bscf(mask); STAT3(normal.trav_prims,1,1,1); const unsigned int geomID = prim.geomID(N); const unsigned int primID = prim.primID(N)[i]; const CurveGeometry* geom = context->scene->get<CurveGeometry>(geomID); const TensorLinearCubicBezierSurface3fa curve = geom->getNormalOrientedCurve<typename Intersector::SourceCurve3ff, typename Intersector::SourceCurve3fa, TensorLinearCubicBezierSurface3fa>(context, ray.org, primID,ray.time()); Intersector().intersect(pre,ray,context,geom,primID,curve,Epilog(ray,context,geomID,primID)); mask &= movemask(tNear <= vfloat<M>(ray.tfar)); } }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x3f8, %rsp # imm = 0x3F8 movq %rdx, %rbp movzbl 0x1(%rcx), %eax leaq (%rax,%rax,8), %r11 leaq (%rax,%r11,4), %r9 movups 0x6(%rcx,%r9), %xmm2 movaps (%rsi), %xmm3 subps %xmm2, %xmm3 shufps $0xff, %xmm2, %xmm2 # xmm2 = xmm2[3,3,3,3] mulps %xmm2, %xmm3 mulps 0x10(%rsi), %xmm2 movd 0x6(%rcx,%rax,4), %xmm0 punpcklbw %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3] psrad $0x18, %xmm0 cvtdq2ps %xmm0, %xmm13 leaq (%rax,%rax,4), %r10 movd 0x6(%rcx,%r10), %xmm1 punpcklbw %xmm1, %xmm1 # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm1, %xmm1 # xmm1 = xmm1[0,0,1,1,2,2,3,3] psrad $0x18, %xmm1 cvtdq2ps %xmm1, %xmm4 leaq (%rax,%rax,2), %rdx movd 0x6(%rcx,%rdx,2), %xmm1 punpcklbw %xmm1, %xmm1 # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm1, %xmm1 # xmm1 = xmm1[0,0,1,1,2,2,3,3] psrad $0x18, %xmm1 cvtdq2ps %xmm1, %xmm6 leaq (%r10,%r10,2), %r8 movd 0x6(%rcx,%r8), %xmm1 punpcklbw %xmm1, %xmm1 # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm1, %xmm1 # xmm1 = xmm1[0,0,1,1,2,2,3,3] psrad $0x18, %xmm1 cvtdq2ps %xmm1, %xmm0 movaps %xmm0, 0x30(%rsp) movl %eax, %r8d shll $0x4, %r8d movd 0x6(%rcx,%r8), %xmm5 punpcklbw %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3] psrad $0x18, %xmm5 cvtdq2ps %xmm5, %xmm7 addq %rax, %r8 movd 0x6(%rcx,%r8), %xmm5 punpcklbw %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3] psrad $0x18, %xmm5 cvtdq2ps %xmm5, %xmm9 leaq (%r10,%r10,4), %r8 addq %rax, %r8 movd 0x6(%rcx,%r8), %xmm5 punpcklbw %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3] psrad $0x18, %xmm5 cvtdq2ps %xmm5, %xmm8 leaq (%r11,%r11,2), %r8 movd 0x6(%rcx,%r8), %xmm5 punpcklbw %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3] psrad $0x18, %xmm5 cvtdq2ps %xmm5, %xmm10 addq %rax, %r8 movd 0x6(%rcx,%r8), %xmm5 punpcklbw %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] punpcklwd %xmm5, %xmm5 # xmm5 = xmm5[0,0,1,1,2,2,3,3] psrad $0x18, %xmm5 cvtdq2ps %xmm5, %xmm1 movaps %xmm2, %xmm5 shufps $0x0, %xmm2, %xmm5 # xmm5 = xmm5[0,0],xmm2[0,0] movaps %xmm2, %xmm14 shufps $0x55, %xmm2, %xmm14 # xmm14 = xmm14[1,1],xmm2[1,1] shufps $0xaa, %xmm2, %xmm2 # xmm2 = xmm2[2,2,2,2] movaps %xmm2, %xmm11 mulps %xmm6, %xmm11 movaps %xmm2, %xmm12 mulps %xmm9, %xmm12 mulps %xmm1, %xmm2 movaps %xmm14, %xmm15 mulps %xmm4, %xmm15 addps %xmm11, %xmm15 movaps %xmm14, %xmm0 mulps %xmm7, %xmm0 addps %xmm12, %xmm0 mulps %xmm10, %xmm14 addps %xmm2, %xmm14 movaps %xmm5, %xmm12 mulps %xmm13, %xmm12 addps %xmm15, %xmm12 movaps %xmm5, %xmm11 movaps 0x30(%rsp), %xmm15 mulps %xmm15, %xmm11 addps %xmm0, %xmm11 mulps %xmm8, %xmm5 addps %xmm14, %xmm5 movaps %xmm3, %xmm2 shufps $0x0, %xmm3, %xmm2 # xmm2 = xmm2[0,0],xmm3[0,0] movaps %xmm3, %xmm14 shufps $0x55, %xmm3, %xmm14 # xmm14 = xmm14[1,1],xmm3[1,1] shufps $0xaa, %xmm3, %xmm3 # xmm3 = xmm3[2,2,2,2] mulps %xmm3, %xmm6 mulps %xmm3, %xmm9 mulps %xmm1, %xmm3 mulps %xmm14, %xmm4 addps %xmm6, %xmm4 mulps %xmm14, %xmm7 addps %xmm9, %xmm7 mulps %xmm10, %xmm14 addps %xmm3, %xmm14 mulps %xmm2, %xmm13 addps %xmm4, %xmm13 movaps %xmm15, %xmm1 mulps %xmm2, %xmm1 addps %xmm7, %xmm1 mulps %xmm8, %xmm2 addps %xmm14, %xmm2 movaps 0x13bac5f(%rip), %xmm4 # 0x1eec6c0 movaps %xmm12, %xmm3 andps %xmm4, %xmm3 movaps 0x13c02d1(%rip), %xmm0 # 0x1ef1d40 cmpnltps %xmm0, %xmm3 andps %xmm3, %xmm12 andnps %xmm0, %xmm3 orps %xmm12, %xmm3 movaps %xmm11, %xmm6 andps %xmm4, %xmm6 cmpnltps %xmm0, %xmm6 andps %xmm6, %xmm11 andnps %xmm0, %xmm6 orps %xmm11, %xmm6 movaps %xmm5, %xmm7 andps %xmm4, %xmm7 cmpnltps %xmm0, %xmm7 andps %xmm7, %xmm5 andnps %xmm0, %xmm7 orps %xmm5, %xmm7 rcpps %xmm3, %xmm0 mulps %xmm0, %xmm3 movaps 0x13baf5b(%rip), %xmm8 # 0x1eeca10 movaps %xmm8, %xmm4 subps %xmm3, %xmm4 mulps %xmm0, %xmm4 addps %xmm0, %xmm4 rcpps %xmm6, %xmm0 mulps %xmm0, %xmm6 movaps %xmm8, %xmm5 subps %xmm6, %xmm5 mulps %xmm0, %xmm5 addps %xmm0, %xmm5 rcpps %xmm7, %xmm0 mulps %xmm0, %xmm7 movaps %xmm8, %xmm6 subps %xmm7, %xmm6 mulps %xmm0, %xmm6 movss 0x1c(%rsi), %xmm8 subss 0x16(%rcx,%r9), %xmm8 addps %xmm0, %xmm6 mulss 0x1a(%rcx,%r9), %xmm8 shufps $0x0, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] leaq (,%rax,8), %r9 subq %rax, %r9 movq 0x6(%rcx,%r9), %xmm0 punpcklwd %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3] psrad $0x10, %xmm0 cvtdq2ps %xmm0, %xmm0 leaq (%rax,%r10,2), %r9 movq 0x6(%rcx,%r9), %xmm3 punpcklwd %xmm3, %xmm3 # xmm3 = xmm3[0,0,1,1,2,2,3,3] psrad $0x10, %xmm3 cvtdq2ps %xmm3, %xmm3 subps %xmm0, %xmm3 mulps %xmm8, %xmm3 addps %xmm0, %xmm3 movq 0x6(%rcx,%r11), %xmm0 punpcklwd %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3] psrad $0x10, %xmm0 cvtdq2ps %xmm0, %xmm0 leaq (%rax,%rdx,4), %r9 movq 0x6(%rcx,%r9), %xmm7 punpcklwd %xmm7, %xmm7 # xmm7 = xmm7[0,0,1,1,2,2,3,3] psrad $0x10, %xmm7 cvtdq2ps %xmm7, %xmm7 subps %xmm0, %xmm7 mulps %xmm8, %xmm7 addps %xmm0, %xmm7 movq 0x6(%rcx,%r11,2), %xmm0 punpcklwd %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3] psrad $0x10, %xmm0 cvtdq2ps %xmm0, %xmm0 shll $0x2, %r10d leaq (%rax,%rax), %r9 addq %r10, %r9 movq 0x6(%rcx,%r9), %xmm9 punpcklwd %xmm9, %xmm9 # xmm9 = xmm9[0,0,1,1,2,2,3,3] psrad $0x10, %xmm9 cvtdq2ps %xmm9, %xmm9 subps %xmm0, %xmm9 mulps %xmm8, %xmm9 addps %xmm0, %xmm9 movq 0x6(%rcx,%r10), %xmm0 punpcklwd %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3] psrad $0x10, %xmm0 cvtdq2ps %xmm0, %xmm0 movq 0x6(%rcx,%rdx,8), %xmm10 punpcklwd %xmm10, %xmm10 # xmm10 = xmm10[0,0,1,1,2,2,3,3] psrad $0x10, %xmm10 cvtdq2ps %xmm10, %xmm10 subps %xmm0, %xmm10 mulps %xmm8, %xmm10 addps %xmm0, %xmm10 addq %rax, %r8 movq 0x6(%rcx,%r8), %xmm0 punpcklwd %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3] psrad $0x10, %xmm0 cvtdq2ps %xmm0, %xmm0 movl %eax, %r8d shll $0x5, %r8d leaq (%rax,%r8), %rdx movq 0x6(%rcx,%rdx), %xmm11 punpcklwd %xmm11, %xmm11 # xmm11 = xmm11[0,0,1,1,2,2,3,3] psrad $0x10, %xmm11 cvtdq2ps %xmm11, %xmm11 subps %xmm0, %xmm11 mulps %xmm8, %xmm11 addps %xmm0, %xmm11 subq %rax, %r8 movq 0x6(%rcx,%r8), %xmm0 punpcklwd %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1,2,2,3,3] psrad $0x10, %xmm0 cvtdq2ps %xmm0, %xmm0 imulq $0x23, %rax, %rdx movq %rcx, 0x138(%rsp) movq 0x6(%rcx,%rdx), %xmm12 punpcklwd %xmm12, %xmm12 # xmm12 = xmm12[0,0,1,1,2,2,3,3] psrad $0x10, %xmm12 cvtdq2ps %xmm12, %xmm12 subps %xmm0, %xmm12 mulps %xmm8, %xmm12 addps %xmm0, %xmm12 subps %xmm13, %xmm3 mulps %xmm4, %xmm3 subps %xmm13, %xmm7 mulps %xmm4, %xmm7 subps %xmm1, %xmm9 mulps %xmm5, %xmm9 subps %xmm1, %xmm10 mulps %xmm5, %xmm10 subps %xmm2, %xmm11 mulps %xmm6, %xmm11 subps %xmm2, %xmm12 mulps %xmm6, %xmm12 movaps %xmm3, %xmm2 minps %xmm7, %xmm2 movaps %xmm9, %xmm0 minps %xmm10, %xmm0 maxps %xmm0, %xmm2 movaps %xmm11, %xmm0 minps %xmm12, %xmm0 movups 0xc(%rsi), %xmm1 shufps $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] maxps %xmm1, %xmm0 maxps %xmm0, %xmm2 mulps 0x13c00b4(%rip), %xmm2 # 0x1ef1d80 maxps %xmm7, %xmm3 maxps %xmm10, %xmm9 minps %xmm9, %xmm3 maxps %xmm12, %xmm11 movaps 0x20(%rsi), %xmm0 shufps $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] minps %xmm0, %xmm11 minps %xmm11, %xmm3 mulps 0x13c009e(%rip), %xmm3 # 0x1ef1d90 movd %eax, %xmm0 pshufd $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] pcmpgtd 0x13befed(%rip), %xmm0 # 0x1ef0cf0 movaps %xmm2, 0x270(%rsp) movaps %xmm2, %xmm1 cmpleps %xmm3, %xmm1 andps %xmm0, %xmm1 movmskps %xmm1, %eax testl %eax, %eax je 0xb342b7 movq %rdi, %r12 movzbl %al, %ebx leaq 0x161e253(%rip), %rax # 0x214ff80 movaps 0xf0(%rax), %xmm1 xorps %xmm0, %xmm0 shufps $0x20, %xmm1, %xmm0 # xmm0 = xmm0[0,0],xmm1[2,0] shufps $0x24, %xmm0, %xmm1 # xmm1 = xmm1[0,1],xmm0[2,0] movaps %xmm1, 0x2c0(%rsp) movq %rsi, 0x48(%rsp) bsfq %rbx, %rax movq 0x138(%rsp), %rcx movl 0x2(%rcx), %edx movl 0x6(%rcx,%rax,4), %r15d movq (%rbp), %rax movq 0x1e8(%rax), %rax movq %rdx, 0x50(%rsp) movq (%rax,%rdx,8), %r14 movss 0x1c(%rsi), %xmm0 movss 0x28(%r14), %xmm3 movss %xmm3, 0x10(%rsp) movss 0x2c(%r14), %xmm2 movss 0x30(%r14), %xmm1 subss %xmm2, %xmm0 subss %xmm2, %xmm1 divss %xmm1, %xmm0 mulss %xmm3, %xmm0 movaps %xmm0, 0x30(%rsp) callq 0x6a790 movss 0x10(%rsp), %xmm1 addss 0x13bec13(%rip), %xmm1 # 0x1ef09cc minss %xmm1, %xmm0 xorps %xmm1, %xmm1 maxss %xmm0, %xmm1 movaps 0x30(%rsp), %xmm0 subss %xmm1, %xmm0 movaps %xmm0, 0x30(%rsp) cvttss2si %xmm1, %eax movslq %eax, %rcx movq 0x58(%r14), %rax movq %r15, 0xe8(%rsp) movq %r15, %rdx imulq 0x68(%r14), %rdx movl (%rax,%rdx), %eax movq 0x188(%r14), %r9 imulq $0x38, %rcx, %rdx movq (%r9,%rdx), %r8 movq 0x10(%r9,%rdx), %r10 movq %r10, %rcx imulq %rax, %rcx movaps (%r8,%rcx), %xmm1 leaq 0x1(%rax), %rcx movq %r10, %rsi imulq %rcx, %rsi movaps (%r8,%rsi), %xmm5 leaq 0x2(%rax), %r15 movq %r10, %rdi imulq %r15, %rdi movaps (%r8,%rdi), %xmm12 leaq 0x3(%rax), %rdi imulq %rdi, %r10 movaps (%r8,%r10), %xmm6 movq 0x1a8(%r14), %r8 movq (%r8,%rdx), %r10 movq 0x10(%r8,%rdx), %r11 movq %r11, %r14 imulq %rax, %r14 movups (%r10,%r14), %xmm4 movq %r11, %r14 imulq %rcx, %r14 movups (%r10,%r14), %xmm7 movq %r11, %r14 imulq %r15, %r14 movups (%r10,%r14), %xmm9 imulq %rdi, %r11 movups (%r10,%r11), %xmm8 movaps %xmm6, %xmm2 movaps 0x13ba84c(%rip), %xmm0 # 0x1eec6d0 mulps %xmm0, %xmm2 movaps %xmm12, %xmm3 xorps %xmm10, %xmm10 mulps %xmm10, %xmm3 movaps %xmm6, %xmm10 movaps 0x13ba841(%rip), %xmm13 # 0x1eec6e0 mulps %xmm13, %xmm6 addps %xmm3, %xmm6 addps %xmm2, %xmm3 movaps %xmm5, 0x20(%rsp) addps %xmm5, %xmm3 xorps %xmm11, %xmm11 mulps %xmm11, %xmm10 addps %xmm12, %xmm2 movaps %xmm2, (%rsp) movaps %xmm12, %xmm2 mulps %xmm13, %xmm2 addps %xmm10, %xmm2 movaps %xmm5, %xmm12 mulps %xmm11, %xmm12 addps %xmm12, %xmm2 movaps %xmm1, %xmm10 mulps %xmm13, %xmm10 movaps %xmm13, %xmm5 subps %xmm10, %xmm2 movaps %xmm8, %xmm10 mulps %xmm0, %xmm10 movaps %xmm9, %xmm13 mulps %xmm11, %xmm13 xorps %xmm15, %xmm15 movaps %xmm8, %xmm14 mulps %xmm5, %xmm8 addps %xmm13, %xmm8 addps %xmm10, %xmm13 mulps %xmm15, %xmm14 addps %xmm9, %xmm10 movaps %xmm9, %xmm11 mulps %xmm5, %xmm11 addps %xmm14, %xmm11 movaps %xmm7, %xmm9 mulps %xmm15, %xmm9 addps %xmm9, %xmm11 movaps %xmm4, %xmm14 mulps %xmm5, %xmm14 subps %xmm14, %xmm11 movaps %xmm1, %xmm14 movaps 0x13ba78b(%rip), %xmm15 # 0x1eec6d0 mulps %xmm15, %xmm14 addps %xmm14, %xmm3 movaps %xmm3, 0x10(%rsp) addps %xmm7, %xmm13 movaps (%rsp), %xmm0 addps %xmm12, %xmm0 movaps %xmm4, %xmm12 mulps %xmm15, %xmm12 addps %xmm12, %xmm13 addps %xmm14, %xmm0 movaps %xmm0, %xmm3 movaps %xmm5, %xmm0 movaps 0x20(%rsp), %xmm5 mulps %xmm0, %xmm5 subps %xmm5, %xmm6 xorps %xmm5, %xmm5 mulps %xmm5, %xmm1 addps %xmm6, %xmm1 movaps %xmm1, %xmm15 addps %xmm9, %xmm10 addps %xmm12, %xmm10 mulps %xmm0, %xmm7 subps %xmm7, %xmm8 mulps %xmm5, %xmm4 addps %xmm8, %xmm4 movaps %xmm2, %xmm8 shufps $0xc9, %xmm2, %xmm8 # xmm8 = xmm8[1,2],xmm2[0,3] movaps %xmm8, %xmm12 mulps %xmm13, %xmm12 shufps $0xc9, %xmm13, %xmm13 # xmm13 = xmm13[1,2,0,3] mulps %xmm2, %xmm13 subps %xmm13, %xmm12 movaps %xmm12, %xmm9 shufps $0xc9, %xmm12, %xmm9 # xmm9 = xmm9[1,2],xmm12[0,3] mulps %xmm11, %xmm8 shufps $0xc9, %xmm11, %xmm11 # xmm11 = xmm11[1,2,0,3] mulps %xmm2, %xmm11 subps %xmm11, %xmm8 shufps $0xc9, %xmm8, %xmm8 # xmm8 = xmm8[1,2,0,3] movaps %xmm1, %xmm5 shufps $0xc9, %xmm1, %xmm5 # xmm5 = xmm5[1,2],xmm1[0,3] movaps %xmm5, %xmm7 mulps %xmm10, %xmm7 shufps $0xc9, %xmm10, %xmm10 # xmm10 = xmm10[1,2,0,3] mulps %xmm1, %xmm10 subps %xmm10, %xmm7 movaps %xmm7, %xmm6 shufps $0xc9, %xmm7, %xmm6 # xmm6 = xmm6[1,2],xmm7[0,3] mulps %xmm4, %xmm5 shufps $0xc9, %xmm4, %xmm4 # xmm4 = xmm4[1,2,0,3] mulps %xmm1, %xmm4 subps %xmm4, %xmm5 shufps $0xc9, %xmm5, %xmm5 # xmm5 = xmm5[1,2,0,3] mulps %xmm12, %xmm12 movaps %xmm12, %xmm4 shufps $0x55, %xmm12, %xmm4 # xmm4 = xmm4[1,1],xmm12[1,1] movaps %xmm12, %xmm10 unpckhpd %xmm12, %xmm10 # xmm10 = xmm10[1],xmm12[1] addps %xmm4, %xmm10 addps %xmm12, %xmm10 xorps %xmm12, %xmm12 movss %xmm10, %xmm12 # xmm12 = xmm10[0],xmm12[1,2,3] movaps %xmm12, %xmm4 rsqrtss %xmm12, %xmm4 movaps %xmm4, %xmm11 movss 0x13ba6c9(%rip), %xmm13 # 0x1eec718 mulss %xmm13, %xmm11 movaps %xmm13, %xmm0 movaps %xmm10, %xmm13 movss 0x13bab1b(%rip), %xmm14 # 0x1eecb80 mulss %xmm14, %xmm13 mulss %xmm4, %xmm13 mulss %xmm4, %xmm4 mulss %xmm13, %xmm4 subss %xmm4, %xmm11 shufps $0x0, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] movaps %xmm9, %xmm13 mulps %xmm8, %xmm13 movaps %xmm13, %xmm4 shufps $0x55, %xmm13, %xmm4 # xmm4 = xmm4[1,1],xmm13[1,1] addss %xmm13, %xmm4 movhlps %xmm13, %xmm13 # xmm13 = xmm13[1,1] addss %xmm4, %xmm13 shufps $0x0, %xmm13, %xmm13 # xmm13 = xmm13[0,0,0,0] mulps %xmm9, %xmm13 movaps %xmm9, %xmm4 mulps %xmm11, %xmm4 movaps %xmm10, %xmm9 shufps $0x0, %xmm10, %xmm9 # xmm9 = xmm9[0,0],xmm10[0,0] mulps %xmm8, %xmm9 subps %xmm13, %xmm9 rcpss %xmm12, %xmm12 mulss %xmm12, %xmm10 movss 0x13bef22(%rip), %xmm13 # 0x1ef0ff8 movaps %xmm13, %xmm8 subss %xmm10, %xmm8 mulss %xmm12, %xmm8 shufps $0x0, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] mulps %xmm9, %xmm8 mulps %xmm11, %xmm8 mulps %xmm7, %xmm7 movaps %xmm7, %xmm10 shufps $0x55, %xmm7, %xmm10 # xmm10 = xmm10[1,1],xmm7[1,1] movaps %xmm7, %xmm9 unpckhpd %xmm7, %xmm9 # xmm9 = xmm9[1],xmm7[1] addps %xmm10, %xmm9 addps %xmm7, %xmm9 xorps %xmm10, %xmm10 movss %xmm9, %xmm10 # xmm10 = xmm9[0],xmm10[1,2,3] movaps %xmm10, %xmm11 rsqrtss %xmm10, %xmm11 movaps %xmm11, %xmm7 mulss %xmm0, %xmm7 movaps %xmm9, %xmm12 mulss %xmm14, %xmm12 mulss %xmm11, %xmm12 mulss %xmm11, %xmm11 mulss %xmm12, %xmm11 subss %xmm11, %xmm7 shufps $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] movaps %xmm6, %xmm11 mulps %xmm5, %xmm11 movaps %xmm11, %xmm12 shufps $0x55, %xmm11, %xmm12 # xmm12 = xmm12[1,1],xmm11[1,1] addss %xmm11, %xmm12 movhlps %xmm11, %xmm11 # xmm11 = xmm11[1,1] addss %xmm12, %xmm11 shufps $0x0, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] mulps %xmm6, %xmm11 mulps %xmm7, %xmm6 movaps %xmm9, %xmm12 shufps $0x0, %xmm9, %xmm12 # xmm12 = xmm12[0,0],xmm9[0,0] mulps %xmm5, %xmm12 subps %xmm11, %xmm12 rcpss %xmm10, %xmm10 mulss %xmm10, %xmm9 movaps %xmm13, %xmm5 subss %xmm9, %xmm5 mulss %xmm10, %xmm5 shufps $0x0, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] mulps %xmm12, %xmm5 mulps %xmm7, %xmm5 movaps 0x10(%rsp), %xmm0 movaps %xmm0, %xmm9 shufps $0xff, %xmm0, %xmm9 # xmm9 = xmm9[3,3],xmm0[3,3] movaps %xmm9, %xmm7 mulps %xmm4, %xmm7 movaps %xmm0, %xmm10 subps %xmm7, %xmm10 movaps %xmm10, 0xb0(%rsp) movaps %xmm7, %xmm10 movaps %xmm2, %xmm7 shufps $0xff, %xmm2, %xmm7 # xmm7 = xmm7[3,3],xmm2[3,3] mulps %xmm4, %xmm7 mulps %xmm8, %xmm9 addps %xmm7, %xmm9 movaps %xmm2, %xmm4 subps %xmm9, %xmm4 movaps %xmm4, 0xa0(%rsp) addps %xmm0, %xmm10 movaps %xmm10, 0xc0(%rsp) addps %xmm2, %xmm9 movaps %xmm9, 0xd0(%rsp) movaps %xmm3, %xmm1 shufps $0xff, %xmm3, %xmm3 # xmm3 = xmm3[3,3,3,3] movaps %xmm3, %xmm0 mulps %xmm6, %xmm0 movaps %xmm1, %xmm4 subps %xmm0, %xmm4 movaps %xmm4, %xmm7 movaps %xmm4, 0x70(%rsp) movaps %xmm15, %xmm4 shufps $0xff, %xmm15, %xmm4 # xmm4 = xmm4[3,3],xmm15[3,3] mulps %xmm6, %xmm4 mulps %xmm5, %xmm3 addps %xmm4, %xmm3 movaps %xmm15, %xmm4 subps %xmm3, %xmm4 addps %xmm1, %xmm0 movaps %xmm0, 0x10(%rsp) addps %xmm15, %xmm3 movaps 0x13bed2c(%rip), %xmm1 # 0x1ef0f80 mulps %xmm1, %xmm4 subps %xmm4, %xmm7 movaps %xmm7, 0x90(%rsp) mulps %xmm1, %xmm3 movaps %xmm0, %xmm1 subps %xmm3, %xmm1 movaps %xmm1, 0x60(%rsp) movq 0x38(%r9,%rdx), %r10 movq 0x48(%r9,%rdx), %r9 movq %r9, %r11 imulq %rax, %r11 movaps (%r10,%r11), %xmm8 movaps %xmm8, (%rsp) movq %r9, %r11 imulq %rcx, %r11 movaps (%r10,%r11), %xmm2 movq %r9, %r11 imulq %r15, %r11 movaps (%r10,%r11), %xmm4 movq 0x48(%rsp), %rsi imulq %rdi, %r9 movaps (%r10,%r9), %xmm6 movaps %xmm6, %xmm5 movaps 0x13ba414(%rip), %xmm9 # 0x1eec6d0 mulps %xmm9, %xmm5 movaps %xmm4, %xmm3 mulps 0x13b9746(%rip), %xmm3 # 0x1eeba10 movaps %xmm6, %xmm7 movaps 0x13ba40c(%rip), %xmm0 # 0x1eec6e0 mulps %xmm0, %xmm6 addps %xmm3, %xmm6 addps %xmm5, %xmm3 xorps %xmm1, %xmm1 mulps %xmm1, %xmm7 addps %xmm4, %xmm5 mulps %xmm0, %xmm4 addps %xmm7, %xmm4 movaps %xmm2, %xmm11 mulps %xmm1, %xmm11 addps %xmm11, %xmm4 movaps %xmm8, %xmm7 mulps %xmm0, %xmm7 movaps %xmm0, %xmm1 subps %xmm7, %xmm4 movq 0x38(%r8,%rdx), %r9 movq 0x48(%r8,%rdx), %rdx imulq %rdx, %rax imulq %rdx, %rcx imulq %rdx, %r15 imulq %rdi, %rdx movups (%r9,%rdx), %xmm8 movaps %xmm8, %xmm7 mulps %xmm9, %xmm7 movups (%r9,%r15), %xmm10 movaps %xmm10, %xmm12 xorps %xmm0, %xmm0 mulps %xmm0, %xmm12 movaps %xmm8, %xmm9 mulps %xmm1, %xmm8 addps %xmm12, %xmm8 addps %xmm7, %xmm12 mulps %xmm0, %xmm9 addps %xmm10, %xmm7 mulps %xmm1, %xmm10 addps %xmm9, %xmm10 movups (%r9,%rcx), %xmm13 movaps %xmm13, %xmm14 mulps %xmm0, %xmm14 addps %xmm14, %xmm10 movups (%r9,%rax), %xmm9 movaps %xmm9, %xmm15 mulps %xmm1, %xmm15 subps %xmm15, %xmm10 addps %xmm11, %xmm5 addps %xmm2, %xmm3 movaps (%rsp), %xmm0 movaps %xmm0, %xmm11 movaps 0x13ba33b(%rip), %xmm15 # 0x1eec6d0 mulps %xmm15, %xmm11 addps %xmm11, %xmm3 addps %xmm11, %xmm5 movaps %xmm5, 0x20(%rsp) mulps %xmm1, %xmm2 subps %xmm2, %xmm6 movaps %xmm0, %xmm5 xorps %xmm0, %xmm0 mulps %xmm0, %xmm5 addps %xmm6, %xmm5 movaps %xmm5, %xmm6 addps %xmm14, %xmm7 addps %xmm13, %xmm12 movaps %xmm9, %xmm5 mulps %xmm15, %xmm5 addps %xmm5, %xmm12 addps %xmm5, %xmm7 mulps %xmm1, %xmm13 subps %xmm13, %xmm8 mulps %xmm0, %xmm9 addps %xmm8, %xmm9 movaps %xmm4, %xmm8 shufps $0xc9, %xmm4, %xmm8 # xmm8 = xmm8[1,2],xmm4[0,3] movaps %xmm8, %xmm11 mulps %xmm12, %xmm11 shufps $0xc9, %xmm12, %xmm12 # xmm12 = xmm12[1,2,0,3] mulps %xmm4, %xmm12 subps %xmm12, %xmm11 mulps %xmm10, %xmm8 shufps $0xc9, %xmm10, %xmm10 # xmm10 = xmm10[1,2,0,3] mulps %xmm4, %xmm10 subps %xmm10, %xmm8 movaps %xmm6, %xmm1 movaps %xmm6, %xmm5 shufps $0xc9, %xmm6, %xmm5 # xmm5 = xmm5[1,2],xmm6[0,3] movaps %xmm5, %xmm6 mulps %xmm7, %xmm6 shufps $0xc9, %xmm7, %xmm7 # xmm7 = xmm7[1,2,0,3] mulps %xmm1, %xmm7 subps %xmm7, %xmm6 mulps %xmm9, %xmm5 shufps $0xc9, %xmm9, %xmm9 # xmm9 = xmm9[1,2,0,3] mulps %xmm1, %xmm9 subps %xmm9, %xmm5 movaps %xmm11, %xmm9 mulps %xmm11, %xmm11 movaps %xmm11, %xmm7 shufps $0x55, %xmm11, %xmm7 # xmm7 = xmm7[1,1],xmm11[1,1] movaps %xmm11, %xmm10 unpckhpd %xmm11, %xmm10 # xmm10 = xmm10[1],xmm11[1] addps %xmm7, %xmm10 addps %xmm11, %xmm10 xorps %xmm11, %xmm11 movss %xmm10, %xmm11 # xmm11 = xmm10[0],xmm11[1,2,3] movaps %xmm11, %xmm7 rsqrtss %xmm11, %xmm7 movaps %xmm10, %xmm13 movss 0x13ba703(%rip), %xmm14 # 0x1eecb80 mulss %xmm14, %xmm13 mulss %xmm7, %xmm13 movaps %xmm7, %xmm12 mulss %xmm7, %xmm7 mulss %xmm13, %xmm7 movss 0x13ba27c(%rip), %xmm0 # 0x1eec718 mulss %xmm0, %xmm12 subss %xmm7, %xmm12 shufps $0xc9, %xmm9, %xmm9 # xmm9 = xmm9[1,2,0,3] shufps $0xc9, %xmm8, %xmm8 # xmm8 = xmm8[1,2,0,3] movaps %xmm9, %xmm7 mulps %xmm8, %xmm7 movaps %xmm10, %xmm13 shufps $0x0, %xmm10, %xmm13 # xmm13 = xmm13[0,0],xmm10[0,0] mulps %xmm8, %xmm13 movaps %xmm7, %xmm8 shufps $0x55, %xmm7, %xmm8 # xmm8 = xmm8[1,1],xmm7[1,1] addss %xmm7, %xmm8 movhlps %xmm7, %xmm7 # xmm7 = xmm7[1,1] addss %xmm8, %xmm7 shufps $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] mulps %xmm9, %xmm7 subps %xmm7, %xmm13 rcpss %xmm11, %xmm11 mulss %xmm11, %xmm10 movss 0x13beafe(%rip), %xmm15 # 0x1ef0ff8 movaps %xmm15, %xmm7 subss %xmm10, %xmm7 mulss %xmm11, %xmm7 shufps $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] mulps %xmm13, %xmm7 shufps $0x0, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] movaps %xmm9, %xmm8 mulps %xmm12, %xmm8 mulps %xmm12, %xmm7 movaps %xmm6, %xmm9 mulps %xmm6, %xmm6 movaps %xmm6, %xmm11 shufps $0x55, %xmm6, %xmm11 # xmm11 = xmm11[1,1],xmm6[1,1] movaps %xmm6, %xmm10 unpckhpd %xmm6, %xmm10 # xmm10 = xmm10[1],xmm6[1] addps %xmm11, %xmm10 addps %xmm6, %xmm10 xorps %xmm6, %xmm6 movss %xmm10, %xmm6 # xmm6 = xmm10[0],xmm6[1,2,3] movaps %xmm6, %xmm12 rsqrtss %xmm6, %xmm12 movaps %xmm10, %xmm13 mulss %xmm14, %xmm13 mulss %xmm12, %xmm13 movaps %xmm12, %xmm11 mulss %xmm12, %xmm12 mulss %xmm13, %xmm12 mulss %xmm0, %xmm11 subss %xmm12, %xmm11 shufps $0xc9, %xmm9, %xmm9 # xmm9 = xmm9[1,2,0,3] shufps $0xc9, %xmm5, %xmm5 # xmm5 = xmm5[1,2,0,3] movaps %xmm9, %xmm12 mulps %xmm5, %xmm12 movaps %xmm10, %xmm13 shufps $0x0, %xmm10, %xmm13 # xmm13 = xmm13[0,0],xmm10[0,0] mulps %xmm5, %xmm13 movaps %xmm12, %xmm5 shufps $0x55, %xmm12, %xmm5 # xmm5 = xmm5[1,1],xmm12[1,1] addss %xmm12, %xmm5 movhlps %xmm12, %xmm12 # xmm12 = xmm12[1,1] addss %xmm5, %xmm12 shufps $0x0, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] mulps %xmm9, %xmm12 subps %xmm12, %xmm13 rcpss %xmm6, %xmm6 mulss %xmm6, %xmm10 movaps %xmm15, %xmm12 subss %xmm10, %xmm12 mulss %xmm6, %xmm12 shufps $0x0, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] mulps %xmm13, %xmm12 shufps $0x0, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] movaps %xmm9, %xmm10 mulps %xmm11, %xmm10 mulps %xmm11, %xmm12 movaps %xmm3, %xmm5 shufps $0xff, %xmm3, %xmm5 # xmm5 = xmm5[3,3],xmm3[3,3] movaps %xmm5, %xmm6 mulps %xmm8, %xmm6 movaps %xmm4, %xmm9 shufps $0xff, %xmm4, %xmm9 # xmm9 = xmm9[3,3],xmm4[3,3] mulps %xmm8, %xmm9 mulps %xmm7, %xmm5 addps %xmm9, %xmm5 movaps %xmm3, %xmm7 subps %xmm6, %xmm7 addps %xmm3, %xmm6 movaps %xmm4, %xmm8 subps %xmm5, %xmm8 addps %xmm4, %xmm5 movaps 0x20(%rsp), %xmm0 movaps %xmm0, %xmm9 shufps $0xff, %xmm0, %xmm9 # xmm9 = xmm9[3,3],xmm0[3,3] movaps %xmm9, %xmm3 mulps %xmm10, %xmm3 movaps %xmm1, %xmm4 shufps $0xff, %xmm1, %xmm4 # xmm4 = xmm4[3,3],xmm1[3,3] mulps %xmm10, %xmm4 mulps %xmm12, %xmm9 addps %xmm4, %xmm9 movaps %xmm0, %xmm4 subps %xmm3, %xmm4 addps %xmm0, %xmm3 movaps %xmm1, %xmm10 subps %xmm9, %xmm10 addps %xmm1, %xmm9 movaps 0x13be919(%rip), %xmm0 # 0x1ef0f80 mulps %xmm0, %xmm10 movaps %xmm4, %xmm2 subps %xmm10, %xmm2 mulps %xmm0, %xmm9 movaps %xmm3, %xmm1 subps %xmm9, %xmm1 movaps 0xa0(%rsp), %xmm9 mulps %xmm0, %xmm9 movaps 0xb0(%rsp), %xmm11 addps %xmm11, %xmm9 movaps %xmm9, %xmm12 mulps %xmm0, %xmm8 addps %xmm7, %xmm8 movss 0x13ba068(%rip), %xmm9 # 0x1eec714 movaps 0x30(%rsp), %xmm10 subss %xmm10, %xmm9 shufps $0x0, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] shufps $0x0, %xmm9, %xmm9 # xmm9 = xmm9[0,0,0,0] mulps %xmm10, %xmm7 mulps %xmm9, %xmm11 addps %xmm7, %xmm11 movaps %xmm11, 0xb0(%rsp) mulps %xmm10, %xmm8 mulps %xmm9, %xmm12 addps %xmm8, %xmm12 movaps %xmm12, 0xa0(%rsp) movaps 0xd0(%rsp), %xmm7 mulps %xmm0, %xmm7 movaps 0xc0(%rsp), %xmm8 addps %xmm8, %xmm7 movaps %xmm7, %xmm13 mulps %xmm0, %xmm5 addps %xmm6, %xmm5 mulps %xmm10, %xmm2 mulps %xmm10, %xmm4 movaps 0x90(%rsp), %xmm7 mulps %xmm9, %xmm7 addps %xmm2, %xmm7 movaps %xmm7, %xmm15 movaps %xmm7, 0x90(%rsp) movaps 0x70(%rsp), %xmm0 mulps %xmm9, %xmm0 addps %xmm4, %xmm0 movaps %xmm0, 0x70(%rsp) mulps %xmm10, %xmm6 mulps %xmm10, %xmm5 mulps %xmm10, %xmm1 mulps %xmm3, %xmm10 movaps %xmm8, %xmm2 mulps %xmm9, %xmm2 addps %xmm6, %xmm2 movaps %xmm2, %xmm6 movaps %xmm2, 0xc0(%rsp) mulps %xmm9, %xmm13 addps %xmm5, %xmm13 movaps %xmm13, 0xd0(%rsp) movaps 0x60(%rsp), %xmm0 mulps %xmm9, %xmm0 addps %xmm1, %xmm0 movaps %xmm0, 0x60(%rsp) mulps 0x10(%rsp), %xmm9 addps %xmm10, %xmm9 movaps (%rsi), %xmm2 movaps %xmm11, %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm14 shufps $0x0, %xmm4, %xmm14 # xmm14 = xmm14[0,0],xmm4[0,0] movaps %xmm4, %xmm5 shufps $0x55, %xmm4, %xmm5 # xmm5 = xmm5[1,1],xmm4[1,1] shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] movaps 0x10(%r12), %xmm0 movaps 0x20(%r12), %xmm1 movaps 0x30(%r12), %xmm3 mulps %xmm3, %xmm4 mulps %xmm1, %xmm5 addps %xmm4, %xmm5 mulps %xmm0, %xmm14 addps %xmm5, %xmm14 movaps %xmm12, %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm7 shufps $0x0, %xmm4, %xmm7 # xmm7 = xmm7[0,0],xmm4[0,0] movaps %xmm4, %xmm5 shufps $0x55, %xmm4, %xmm5 # xmm5 = xmm5[1,1],xmm4[1,1] shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] mulps %xmm3, %xmm4 mulps %xmm1, %xmm5 addps %xmm4, %xmm5 mulps %xmm0, %xmm7 addps %xmm5, %xmm7 movaps %xmm15, %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm5 shufps $0x55, %xmm4, %xmm5 # xmm5 = xmm5[1,1],xmm4[1,1] movaps %xmm4, %xmm8 shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] mulps %xmm3, %xmm4 mulps %xmm1, %xmm5 addps %xmm4, %xmm5 shufps $0x0, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] mulps %xmm0, %xmm8 addps %xmm5, %xmm8 movaps 0x70(%rsp), %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm5 shufps $0x55, %xmm4, %xmm5 # xmm5 = xmm5[1,1],xmm4[1,1] movaps %xmm4, %xmm10 shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] mulps %xmm3, %xmm4 mulps %xmm1, %xmm5 addps %xmm4, %xmm5 shufps $0x0, %xmm10, %xmm10 # xmm10 = xmm10[0,0,0,0] mulps %xmm0, %xmm10 addps %xmm5, %xmm10 movaps %xmm6, %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm5 shufps $0x55, %xmm4, %xmm5 # xmm5 = xmm5[1,1],xmm4[1,1] movaps %xmm4, %xmm11 shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] mulps %xmm3, %xmm4 mulps %xmm1, %xmm5 addps %xmm4, %xmm5 shufps $0x0, %xmm11, %xmm11 # xmm11 = xmm11[0,0,0,0] mulps %xmm0, %xmm11 addps %xmm5, %xmm11 movaps %xmm13, %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm5 shufps $0x55, %xmm4, %xmm5 # xmm5 = xmm5[1,1],xmm4[1,1] movaps %xmm4, %xmm12 shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] mulps %xmm3, %xmm4 mulps %xmm1, %xmm5 addps %xmm4, %xmm5 shufps $0x0, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] mulps %xmm0, %xmm12 addps %xmm5, %xmm12 movaps 0x60(%rsp), %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm5 shufps $0x55, %xmm4, %xmm5 # xmm5 = xmm5[1,1],xmm4[1,1] movaps %xmm4, %xmm13 shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] mulps %xmm3, %xmm4 mulps %xmm1, %xmm5 addps %xmm4, %xmm5 shufps $0x0, %xmm13, %xmm13 # xmm13 = xmm13[0,0,0,0] mulps %xmm0, %xmm13 addps %xmm5, %xmm13 movaps %xmm9, %xmm4 subps %xmm2, %xmm4 movaps %xmm4, %xmm5 movaps %xmm4, %xmm2 shufps $0xaa, %xmm4, %xmm4 # xmm4 = xmm4[2,2,2,2] mulps %xmm3, %xmm4 shufps $0x55, %xmm2, %xmm2 # xmm2 = xmm2[1,1,1,1] mulps %xmm1, %xmm2 addps %xmm4, %xmm2 shufps $0x0, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] mulps %xmm0, %xmm5 addps %xmm2, %xmm5 movaps %xmm14, %xmm15 movlhps %xmm11, %xmm15 # xmm15 = xmm15[0],xmm11[0] movaps %xmm7, %xmm3 movlhps %xmm12, %xmm3 # xmm3 = xmm3[0],xmm12[0] movaps %xmm8, %xmm4 movlhps %xmm13, %xmm4 # xmm4 = xmm4[0],xmm13[0] movaps %xmm10, %xmm6 movlhps %xmm5, %xmm6 # xmm6 = xmm6[0],xmm5[0] movaps %xmm15, %xmm0 minps %xmm3, %xmm0 movaps %xmm4, %xmm1 minps %xmm6, %xmm1 minps %xmm1, %xmm0 movaps %xmm15, %xmm1 maxps %xmm3, %xmm1 movaps %xmm4, %xmm2 maxps %xmm6, %xmm2 maxps %xmm2, %xmm1 movaps %xmm0, %xmm2 unpckhpd %xmm0, %xmm2 # xmm2 = xmm2[1],xmm0[1] minps %xmm2, %xmm0 movaps %xmm1, %xmm2 unpckhpd %xmm1, %xmm2 # xmm2 = xmm2[1],xmm1[1] maxps %xmm2, %xmm1 movaps 0x13b9d66(%rip), %xmm2 # 0x1eec6c0 andps %xmm2, %xmm0 andps %xmm2, %xmm1 movsd 0x13b9d88(%rip), %xmm2 # 0x1eec6f0 maxps %xmm1, %xmm0 movaps %xmm0, %xmm1 shufps $0x55, %xmm0, %xmm1 # xmm1 = xmm1[1,1],xmm0[1,1] maxss %xmm0, %xmm1 leaq 0xf(%rbx), %r11 mulss 0x13bf536(%rip), %xmm1 # 0x1ef1eb8 movlhps %xmm14, %xmm14 # xmm14 = xmm14[0,0] movaps %xmm14, 0x220(%rsp) movlhps %xmm7, %xmm7 # xmm7 = xmm7[0,0] movaps %xmm7, 0x210(%rsp) movlhps %xmm8, %xmm8 # xmm8 = xmm8[0,0] movaps %xmm8, 0x200(%rsp) movlhps %xmm10, %xmm10 # xmm10 = xmm10[0,0] movaps %xmm10, 0x1f0(%rsp) movlhps %xmm11, %xmm11 # xmm11 = xmm11[0,0] movaps %xmm11, 0x1e0(%rsp) movlhps %xmm12, %xmm12 # xmm12 = xmm12[0,0] movaps %xmm12, 0x1d0(%rsp) movlhps %xmm13, %xmm13 # xmm13 = xmm13[0,0] movaps %xmm13, 0x1c0(%rsp) movlhps %xmm5, %xmm5 # xmm5 = xmm5[0,0] movaps %xmm5, 0x1b0(%rsp) movaps %xmm1, 0x30(%rsp) shufps $0x0, %xmm1, %xmm1 # xmm1 = xmm1[0,0,0,0] movaps %xmm1, 0x2e0(%rsp) xorps 0x13b9cd2(%rip), %xmm1 # 0x1eec6d0 movaps %xmm1, 0x2d0(%rsp) movaps %xmm3, %xmm0 subps %xmm15, %xmm0 movaps %xmm0, 0x3a0(%rsp) movaps %xmm4, %xmm0 subps %xmm3, %xmm0 movaps %xmm0, 0x390(%rsp) movaps %xmm6, %xmm0 subps %xmm4, %xmm0 movaps %xmm0, 0x380(%rsp) movaps 0xc0(%rsp), %xmm0 subps 0xb0(%rsp), %xmm0 movaps %xmm0, 0x2b0(%rsp) movaps 0xd0(%rsp), %xmm0 subps 0xa0(%rsp), %xmm0 movaps %xmm0, 0x2a0(%rsp) movaps 0x60(%rsp), %xmm0 subps 0x90(%rsp), %xmm0 movaps %xmm0, 0x290(%rsp) movaps %xmm9, 0x170(%rsp) subps 0x70(%rsp), %xmm9 movaps %xmm9, 0x280(%rsp) xorl %r14d, %r14d movaps %xmm2, 0x10(%rsp) movaps %xmm2, 0x20(%rsp) movaps %xmm15, 0x120(%rsp) movaps %xmm3, 0x110(%rsp) movaps %xmm4, 0x100(%rsp) movaps %xmm6, 0x370(%rsp) movaps 0x10(%rsp), %xmm0 unpcklps %xmm0, %xmm0 # xmm0 = xmm0[0,0,1,1] movaps 0x13b9f44(%rip), %xmm15 # 0x1eeca10 movaps %xmm15, %xmm14 subps %xmm0, %xmm14 movaps 0x1e0(%rsp), %xmm1 mulps %xmm0, %xmm1 movaps 0x1d0(%rsp), %xmm2 mulps %xmm0, %xmm2 movaps 0x1c0(%rsp), %xmm3 mulps %xmm0, %xmm3 mulps 0x1b0(%rsp), %xmm0 movaps 0x220(%rsp), %xmm8 mulps %xmm14, %xmm8 addps %xmm1, %xmm8 movaps 0x210(%rsp), %xmm10 mulps %xmm14, %xmm10 addps %xmm2, %xmm10 movaps 0x200(%rsp), %xmm11 mulps %xmm14, %xmm11 addps %xmm3, %xmm11 mulps 0x1f0(%rsp), %xmm14 addps %xmm0, %xmm14 movaps 0x20(%rsp), %xmm6 movaps %xmm6, %xmm12 shufps $0x55, %xmm6, %xmm12 # xmm12 = xmm12[1,1],xmm6[1,1] movaps %xmm12, %xmm0 subss %xmm6, %xmm0 mulss 0x13bf355(%rip), %xmm0 # 0x1ef1eb0 movaps %xmm0, %xmm13 shufps $0x0, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] subps %xmm6, %xmm12 movaps %xmm8, %xmm2 shufps $0x0, %xmm8, %xmm2 # xmm2 = xmm2[0,0],xmm8[0,0] movaps %xmm8, %xmm4 shufps $0x55, %xmm8, %xmm4 # xmm4 = xmm4[1,1],xmm8[1,1] movaps %xmm10, %xmm3 shufps $0x0, %xmm10, %xmm3 # xmm3 = xmm3[0,0],xmm10[0,0] movaps %xmm10, %xmm9 shufps $0x55, %xmm10, %xmm9 # xmm9 = xmm9[1,1],xmm10[1,1] movaps %xmm11, %xmm5 shufps $0x0, %xmm11, %xmm5 # xmm5 = xmm5[0,0],xmm11[0,0] movaps %xmm11, %xmm1 shufps $0x55, %xmm11, %xmm1 # xmm1 = xmm1[1,1],xmm11[1,1] movaps %xmm14, %xmm7 shufps $0x0, %xmm14, %xmm7 # xmm7 = xmm7[0,0],xmm14[0,0] movaps %xmm14, %xmm0 shufps $0x55, %xmm14, %xmm0 # xmm0 = xmm0[1,1],xmm14[1,1] shufps $0x0, %xmm13, %xmm13 # xmm13 = xmm13[0,0,0,0] movaps %xmm13, 0xf0(%rsp) mulps 0x13bf1eb(%rip), %xmm12 # 0x1ef1db0 addps %xmm6, %xmm12 movaps %xmm15, %xmm13 subps %xmm12, %xmm13 movaps %xmm3, %xmm15 mulps %xmm12, %xmm15 mulps %xmm13, %xmm2 addps %xmm15, %xmm2 movaps %xmm9, %xmm15 mulps %xmm12, %xmm15 mulps %xmm13, %xmm4 addps %xmm15, %xmm4 movaps %xmm5, %xmm15 mulps %xmm12, %xmm15 mulps %xmm13, %xmm3 addps %xmm15, %xmm3 movaps %xmm1, %xmm15 mulps %xmm12, %xmm15 mulps %xmm13, %xmm9 addps %xmm15, %xmm9 mulps %xmm12, %xmm7 mulps %xmm12, %xmm0 mulps %xmm13, %xmm5 addps %xmm7, %xmm5 mulps %xmm13, %xmm1 addps %xmm0, %xmm1 movaps %xmm12, %xmm0 mulps %xmm3, %xmm0 movaps %xmm12, %xmm7 mulps %xmm9, %xmm7 mulps %xmm13, %xmm2 addps %xmm0, %xmm2 mulps %xmm13, %xmm4 addps %xmm7, %xmm4 mulps %xmm12, %xmm5 mulps %xmm12, %xmm1 mulps %xmm13, %xmm3 addps %xmm5, %xmm3 mulps %xmm13, %xmm9 addps %xmm1, %xmm9 movaps %xmm12, %xmm0 mulps %xmm3, %xmm0 movaps %xmm12, %xmm5 mulps %xmm9, %xmm5 movaps %xmm13, %xmm1 mulps %xmm2, %xmm1 addps %xmm0, %xmm1 movaps %xmm13, %xmm6 mulps %xmm4, %xmm6 addps %xmm5, %xmm6 subps %xmm2, %xmm3 subps %xmm4, %xmm9 movaps 0x13be2d4(%rip), %xmm4 # 0x1ef0f60 mulps %xmm4, %xmm3 mulps %xmm4, %xmm9 movaps 0xf0(%rsp), %xmm7 mulps %xmm7, %xmm3 mulps %xmm7, %xmm9 xorps %xmm0, %xmm0 shufps $0x30, %xmm1, %xmm0 # xmm0 = xmm0[0,0],xmm1[3,0] movaps %xmm1, %xmm2 movaps %xmm1, 0x180(%rsp) shufps $0x29, %xmm0, %xmm2 # xmm2 = xmm2[1,2],xmm0[2,0] movaps %xmm2, %xmm4 xorps %xmm0, %xmm0 shufps $0x30, %xmm6, %xmm0 # xmm0 = xmm0[0,0],xmm6[3,0] movaps %xmm6, %xmm2 movaps %xmm6, %xmm15 movaps %xmm6, 0x190(%rsp) shufps $0x29, %xmm0, %xmm2 # xmm2 = xmm2[1,2],xmm0[2,0] movaps %xmm2, %xmm5 xorps %xmm0, %xmm0 shufps $0x30, %xmm3, %xmm0 # xmm0 = xmm0[0,0],xmm3[3,0] xorps %xmm2, %xmm2 shufps $0x30, %xmm9, %xmm2 # xmm2 = xmm2[0,0],xmm9[3,0] movaps %xmm1, %xmm6 addps %xmm3, %xmm6 movaps %xmm6, 0x80(%rsp) shufps $0x29, %xmm0, %xmm3 # xmm3 = xmm3[1,2],xmm0[2,0] movaps %xmm15, %xmm0 addps %xmm9, %xmm0 movaps %xmm0, (%rsp) shufps $0x29, %xmm2, %xmm9 # xmm9 = xmm9[1,2],xmm2[2,0] movaps %xmm4, %xmm0 movaps %xmm4, %xmm2 subps %xmm3, %xmm0 movaps %xmm0, 0x1a0(%rsp) movaps %xmm5, %xmm3 movaps %xmm5, %xmm4 subps %xmm9, %xmm3 movaps %xmm8, %xmm9 shufps $0xaa, %xmm8, %xmm9 # xmm9 = xmm9[2,2],xmm8[2,2] movaps %xmm10, %xmm5 shufps $0xaa, %xmm10, %xmm5 # xmm5 = xmm5[2,2],xmm10[2,2] movaps %xmm5, %xmm0 mulps %xmm12, %xmm0 mulps %xmm13, %xmm9 addps %xmm0, %xmm9 shufps $0xff, %xmm8, %xmm8 # xmm8 = xmm8[3,3,3,3] shufps $0xff, %xmm10, %xmm10 # xmm10 = xmm10[3,3,3,3] movaps %xmm10, %xmm0 mulps %xmm12, %xmm0 mulps %xmm13, %xmm8 addps %xmm0, %xmm8 movaps %xmm11, %xmm15 shufps $0xaa, %xmm11, %xmm15 # xmm15 = xmm15[2,2],xmm11[2,2] movaps %xmm15, %xmm0 mulps %xmm12, %xmm0 mulps %xmm13, %xmm5 addps %xmm0, %xmm5 shufps $0xff, %xmm11, %xmm11 # xmm11 = xmm11[3,3,3,3] movaps %xmm11, %xmm0 mulps %xmm12, %xmm0 mulps %xmm13, %xmm10 addps %xmm0, %xmm10 movaps %xmm14, %xmm0 shufps $0xaa, %xmm14, %xmm0 # xmm0 = xmm0[2,2],xmm14[2,2] mulps %xmm12, %xmm0 mulps %xmm13, %xmm15 addps %xmm0, %xmm15 shufps $0xff, %xmm14, %xmm14 # xmm14 = xmm14[3,3,3,3] mulps %xmm12, %xmm14 mulps %xmm13, %xmm11 addps %xmm14, %xmm11 movaps %xmm12, %xmm0 mulps %xmm5, %xmm0 mulps %xmm13, %xmm9 addps %xmm0, %xmm9 movaps %xmm12, %xmm0 mulps %xmm10, %xmm0 mulps %xmm13, %xmm8 addps %xmm0, %xmm8 mulps %xmm12, %xmm15 mulps %xmm13, %xmm5 addps %xmm15, %xmm5 mulps %xmm12, %xmm11 mulps %xmm13, %xmm10 addps %xmm11, %xmm10 movaps %xmm12, %xmm0 mulps %xmm5, %xmm0 movaps %xmm13, %xmm6 mulps %xmm9, %xmm6 addps %xmm0, %xmm6 mulps %xmm10, %xmm12 mulps %xmm8, %xmm13 addps %xmm12, %xmm13 subps %xmm9, %xmm5 subps %xmm8, %xmm10 movaps 0x13be149(%rip), %xmm0 # 0x1ef0f60 mulps %xmm0, %xmm5 mulps %xmm0, %xmm10 mulps %xmm7, %xmm5 mulps %xmm7, %xmm10 xorps %xmm0, %xmm0 shufps $0x30, %xmm6, %xmm0 # xmm0 = xmm0[0,0],xmm6[3,0] movaps %xmm6, %xmm1 shufps $0x29, %xmm0, %xmm1 # xmm1 = xmm1[1,2],xmm0[2,0] xorps %xmm0, %xmm0 shufps $0x30, %xmm13, %xmm0 # xmm0 = xmm0[0,0],xmm13[3,0] movaps %xmm13, %xmm12 shufps $0x29, %xmm0, %xmm12 # xmm12 = xmm12[1,2],xmm0[2,0] xorps %xmm0, %xmm0 shufps $0x30, %xmm5, %xmm0 # xmm0 = xmm0[0,0],xmm5[3,0] movaps %xmm6, %xmm7 addps %xmm5, %xmm7 movaps %xmm7, 0xf0(%rsp) shufps $0x29, %xmm0, %xmm5 # xmm5 = xmm5[1,2],xmm0[2,0] xorps %xmm0, %xmm0 shufps $0x30, %xmm10, %xmm0 # xmm0 = xmm0[0,0],xmm10[3,0] movaps %xmm13, %xmm11 addps %xmm10, %xmm11 shufps $0x29, %xmm0, %xmm10 # xmm10 = xmm10[1,2],xmm0[2,0] movaps %xmm1, %xmm7 subps %xmm5, %xmm7 movaps %xmm12, %xmm14 subps %xmm10, %xmm14 movaps %xmm6, %xmm0 movaps 0x180(%rsp), %xmm15 subps %xmm15, %xmm0 movaps %xmm1, %xmm10 movaps %xmm2, %xmm8 subps %xmm2, %xmm10 addps %xmm0, %xmm10 movaps %xmm13, %xmm0 movaps 0x190(%rsp), %xmm5 subps %xmm5, %xmm0 movaps %xmm12, %xmm9 subps %xmm4, %xmm9 addps %xmm0, %xmm9 movaps %xmm15, %xmm0 mulps %xmm10, %xmm5 mulps %xmm9, %xmm0 subps %xmm0, %xmm5 movaps (%rsp), %xmm2 mulps %xmm10, %xmm2 movaps 0x80(%rsp), %xmm0 mulps %xmm9, %xmm0 subps %xmm0, %xmm2 movaps %xmm3, 0x350(%rsp) mulps %xmm10, %xmm3 movaps %xmm9, %xmm0 mulps 0x1a0(%rsp), %xmm0 subps %xmm0, %xmm3 movaps %xmm10, %xmm15 movaps %xmm4, 0x360(%rsp) mulps %xmm4, %xmm15 movaps %xmm9, %xmm0 movaps %xmm8, 0x340(%rsp) mulps %xmm8, %xmm0 subps %xmm0, %xmm15 movaps %xmm13, %xmm0 mulps %xmm10, %xmm0 movaps %xmm6, %xmm4 mulps %xmm9, %xmm6 subps %xmm6, %xmm0 movaps %xmm11, 0x320(%rsp) mulps %xmm10, %xmm11 movaps 0xf0(%rsp), %xmm8 mulps %xmm9, %xmm8 subps %xmm8, %xmm11 movaps %xmm14, 0x300(%rsp) movaps %xmm14, %xmm8 mulps %xmm10, %xmm8 movaps %xmm9, %xmm14 movaps %xmm7, 0x2f0(%rsp) mulps %xmm7, %xmm14 subps %xmm14, %xmm8 movaps %xmm12, 0x310(%rsp) mulps %xmm12, %xmm10 movaps %xmm1, 0x330(%rsp) mulps %xmm1, %xmm9 subps %xmm9, %xmm10 movaps %xmm5, %xmm14 minps %xmm2, %xmm14 maxps %xmm2, %xmm5 movaps %xmm3, %xmm2 minps %xmm15, %xmm2 minps %xmm2, %xmm14 maxps %xmm15, %xmm3 maxps %xmm3, %xmm5 movaps %xmm0, %xmm2 minps %xmm11, %xmm2 maxps %xmm11, %xmm0 movaps %xmm8, %xmm3 minps %xmm10, %xmm3 minps %xmm3, %xmm2 minps %xmm2, %xmm14 maxps %xmm10, %xmm8 maxps %xmm8, %xmm0 maxps %xmm0, %xmm5 movaps 0x2e0(%rsp), %xmm8 cmpleps %xmm8, %xmm14 movaps 0x2d0(%rsp), %xmm9 cmpnltps %xmm9, %xmm5 andps %xmm14, %xmm5 andps 0x2c0(%rsp), %xmm5 movmskps %xmm5, %ecx movl $0x0, %eax testl %ecx, %ecx je 0xb33151 movaps 0x340(%rsp), %xmm11 movaps %xmm11, %xmm0 movaps 0x180(%rsp), %xmm1 subps %xmm1, %xmm0 movaps 0x330(%rsp), %xmm6 movaps %xmm6, %xmm10 subps %xmm4, %xmm10 addps %xmm0, %xmm10 movaps 0x360(%rsp), %xmm3 movaps %xmm3, %xmm2 movaps 0x190(%rsp), %xmm15 subps %xmm15, %xmm2 movaps 0x310(%rsp), %xmm12 movaps %xmm12, %xmm0 subps %xmm13, %xmm0 addps %xmm2, %xmm0 mulps %xmm10, %xmm15 mulps %xmm0, %xmm1 subps %xmm1, %xmm15 movaps (%rsp), %xmm2 mulps %xmm10, %xmm2 movaps 0x80(%rsp), %xmm7 mulps %xmm0, %xmm7 subps %xmm7, %xmm2 movaps 0x350(%rsp), %xmm7 mulps %xmm10, %xmm7 movaps 0x1a0(%rsp), %xmm14 mulps %xmm0, %xmm14 subps %xmm14, %xmm7 mulps %xmm10, %xmm3 mulps %xmm0, %xmm11 subps %xmm11, %xmm3 mulps %xmm10, %xmm13 mulps %xmm0, %xmm4 subps %xmm4, %xmm13 movaps 0x320(%rsp), %xmm4 mulps %xmm10, %xmm4 movaps 0xf0(%rsp), %xmm1 mulps %xmm0, %xmm1 subps %xmm1, %xmm4 movaps 0x300(%rsp), %xmm11 mulps %xmm10, %xmm11 movaps 0x2f0(%rsp), %xmm1 mulps %xmm0, %xmm1 subps %xmm1, %xmm11 mulps %xmm12, %xmm10 mulps %xmm6, %xmm0 subps %xmm0, %xmm10 movaps %xmm15, %xmm0 minps %xmm2, %xmm0 maxps %xmm2, %xmm15 movaps %xmm7, %xmm2 minps %xmm3, %xmm2 minps %xmm2, %xmm0 maxps %xmm3, %xmm7 maxps %xmm7, %xmm15 movaps %xmm13, %xmm2 minps %xmm4, %xmm2 maxps %xmm4, %xmm13 movaps %xmm11, %xmm3 minps %xmm10, %xmm3 minps %xmm3, %xmm2 minps %xmm2, %xmm0 maxps %xmm10, %xmm11 maxps %xmm11, %xmm13 maxps %xmm13, %xmm15 cmpleps %xmm8, %xmm0 cmpnltps %xmm9, %xmm15 andps %xmm0, %xmm15 andps %xmm5, %xmm15 movmskps %xmm15, %eax testl %eax, %eax je 0xb3317c movl %r14d, %ecx movl %eax, 0x230(%rsp,%rcx,4) movaps 0x20(%rsp), %xmm0 movlps %xmm0, 0x3b0(%rsp,%rcx,8) movaps 0x10(%rsp), %xmm0 movlps %xmm0, 0x3d0(%rsp,%rcx,8) incl %r14d testl %r14d, %r14d je 0xb34295 leal -0x1(%r14), %eax movl 0x230(%rsp,%rax,4), %edx movsd 0x3d0(%rsp,%rax,8), %xmm3 bsfq %rdx, %rcx leal -0x1(%rdx), %edi andl %edx, %edi movl %edi, 0x230(%rsp,%rax,4) cmovel %eax, %r14d testq %rcx, %rcx js 0xb331bc xorps %xmm0, %xmm0 cvtsi2ss %rcx, %xmm0 jmp 0xb331d6 movq %rcx, %rdx shrq %rdx movl %ecx, %edi andl $0x1, %edi orq %rdx, %rdi xorps %xmm0, %xmm0 cvtsi2ss %rdi, %xmm0 addss %xmm0, %xmm0 incq %rcx js 0xb331e5 xorps %xmm1, %xmm1 cvtsi2ss %rcx, %xmm1 jmp 0xb331fd movq %rcx, %rdx shrq %rdx andl $0x1, %ecx orq %rdx, %rcx xorps %xmm1, %xmm1 cvtsi2ss %rcx, %xmm1 addss %xmm1, %xmm1 unpcklps %xmm1, %xmm0 # xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] mulps 0x13bec39(%rip), %xmm0 # 0x1ef1e40 movaps 0x13b9822(%rip), %xmm1 # 0x1eeca30 subps %xmm0, %xmm1 movss 0x3b0(%rsp,%rax,8), %xmm9 movss 0x3b4(%rsp,%rax,8), %xmm2 shufps $0x0, %xmm2, %xmm2 # xmm2 = xmm2[0,0,0,0] mulps %xmm0, %xmm2 shufps $0x0, %xmm9, %xmm9 # xmm9 = xmm9[0,0,0,0] mulps %xmm1, %xmm9 addps %xmm2, %xmm9 movaps %xmm9, %xmm2 shufps $0x55, %xmm9, %xmm2 # xmm2 = xmm2[1,1],xmm9[1,1] movaps %xmm2, %xmm1 subss %xmm9, %xmm1 movss 0x13bddaf(%rip), %xmm0 # 0x1ef1000 ucomiss %xmm1, %xmm0 movaps %xmm3, 0x10(%rsp) movaps %xmm9, 0x20(%rsp) jbe 0xb32abc movaps %xmm3, %xmm0 unpcklps %xmm3, %xmm0 # xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1] movaps 0x13b979e(%rip), %xmm4 # 0x1eeca10 subps %xmm0, %xmm4 movaps 0x1e0(%rsp), %xmm3 mulps %xmm0, %xmm3 movaps 0x1d0(%rsp), %xmm7 mulps %xmm0, %xmm7 movaps 0x1c0(%rsp), %xmm8 mulps %xmm0, %xmm8 mulps 0x1b0(%rsp), %xmm0 movaps 0x220(%rsp), %xmm5 mulps %xmm4, %xmm5 addps %xmm3, %xmm5 movaps 0x210(%rsp), %xmm6 mulps %xmm4, %xmm6 addps %xmm7, %xmm6 movaps 0x200(%rsp), %xmm7 mulps %xmm4, %xmm7 addps %xmm8, %xmm7 mulps 0x1f0(%rsp), %xmm4 addps %xmm0, %xmm4 movaps %xmm9, %xmm0 shufps $0x0, %xmm9, %xmm0 # xmm0 = xmm0[0,0],xmm9[0,0] subps %xmm7, %xmm4 movaps %xmm0, %xmm3 mulps %xmm4, %xmm3 addps %xmm7, %xmm3 mulps %xmm2, %xmm4 addps %xmm7, %xmm4 subps %xmm6, %xmm7 movaps %xmm0, %xmm8 mulps %xmm7, %xmm8 addps %xmm6, %xmm8 mulps %xmm2, %xmm7 addps %xmm6, %xmm7 subps %xmm5, %xmm6 movaps %xmm0, %xmm9 mulps %xmm6, %xmm9 addps %xmm5, %xmm9 subps %xmm8, %xmm3 mulps %xmm0, %xmm3 addps %xmm8, %xmm3 subps %xmm9, %xmm8 mulps %xmm0, %xmm8 addps %xmm9, %xmm8 subps %xmm8, %xmm3 mulps %xmm3, %xmm0 addps %xmm8, %xmm0 movaps 0x13bdc21(%rip), %xmm8 # 0x1ef0f60 mulps %xmm8, %xmm3 mulps %xmm2, %xmm6 addps %xmm5, %xmm6 subps %xmm7, %xmm4 mulps %xmm2, %xmm4 addps %xmm7, %xmm4 subps %xmm6, %xmm7 mulps %xmm2, %xmm7 addps %xmm6, %xmm7 subps %xmm7, %xmm4 mulps %xmm4, %xmm2 addps %xmm7, %xmm2 mulps %xmm8, %xmm4 movaps %xmm1, %xmm5 mulss 0x13beb49(%rip), %xmm5 # 0x1ef1ebc shufps $0x0, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] mulps %xmm5, %xmm3 addps %xmm0, %xmm3 mulps %xmm4, %xmm5 movaps %xmm2, %xmm4 subps %xmm5, %xmm4 movaps %xmm0, %xmm14 unpckhpd %xmm0, %xmm14 # xmm14 = xmm14[1],xmm0[1] movaps %xmm2, %xmm6 unpckhpd %xmm2, %xmm6 # xmm6 = xmm6[1],xmm2[1] movaps %xmm14, %xmm5 subps %xmm0, %xmm5 movaps %xmm6, %xmm8 subps %xmm2, %xmm8 addps %xmm5, %xmm8 movaps %xmm0, %xmm5 shufps $0xf5, %xmm0, %xmm5 # xmm5 = xmm5[1,1],xmm0[3,3] movaps %xmm3, %xmm9 shufps $0xf5, %xmm3, %xmm9 # xmm9 = xmm9[1,1],xmm3[3,3] movaps %xmm4, %xmm10 shufps $0xf5, %xmm4, %xmm10 # xmm10 = xmm10[1,1],xmm4[3,3] movaps %xmm2, %xmm11 shufps $0xf5, %xmm2, %xmm11 # xmm11 = xmm11[1,1],xmm2[3,3] movaps %xmm8, %xmm7 shufps $0x0, %xmm8, %xmm7 # xmm7 = xmm7[0,0],xmm8[0,0] shufps $0x55, %xmm8, %xmm8 # xmm8 = xmm8[1,1,1,1] mulps %xmm8, %xmm5 mulps %xmm8, %xmm9 mulps %xmm8, %xmm10 mulps %xmm11, %xmm8 movaps %xmm0, %xmm12 mulps %xmm7, %xmm12 addps %xmm5, %xmm12 movaps %xmm3, %xmm11 mulps %xmm7, %xmm11 addps %xmm9, %xmm11 movaps %xmm4, %xmm5 mulps %xmm7, %xmm5 addps %xmm10, %xmm5 mulps %xmm2, %xmm7 addps %xmm8, %xmm7 movaps %xmm11, %xmm8 movaps %xmm12, %xmm10 minss %xmm11, %xmm10 maxss %xmm12, %xmm11 movhlps %xmm12, %xmm12 # xmm12 = xmm12[1,1] movhlps %xmm8, %xmm8 # xmm8 = xmm8[1,1] movaps %xmm5, %xmm9 minss %xmm7, %xmm9 minss %xmm9, %xmm10 movaps %xmm7, %xmm9 maxss %xmm5, %xmm7 movhlps %xmm5, %xmm5 # xmm5 = xmm5[1,1] movhlps %xmm9, %xmm9 # xmm9 = xmm9[1,1] maxss %xmm11, %xmm7 movaps %xmm12, %xmm13 minss %xmm8, %xmm13 maxss %xmm12, %xmm8 movaps %xmm5, %xmm11 minss %xmm9, %xmm11 minss %xmm11, %xmm13 maxss %xmm5, %xmm9 maxss %xmm8, %xmm9 movss 0x13be5d0(%rip), %xmm5 # 0x1ef1a4c ucomiss %xmm1, %xmm5 seta %cl cmpl $0x4, %r14d setae %al movss 0x13bd547(%rip), %xmm5 # 0x1ef09d8 ucomiss %xmm10, %xmm5 jbe 0xb334a1 ucomiss 0x13bea21(%rip), %xmm9 # 0x1ef1ec0 ja 0xb334eb movss 0x13bea16(%rip), %xmm8 # 0x1ef1ec0 ucomiss %xmm8, %xmm9 setbe %dl movss 0x13bd51f(%rip), %xmm5 # 0x1ef09d8 ucomiss %xmm10, %xmm5 setbe %dil ucomiss %xmm13, %xmm5 setbe %r10b ucomiss %xmm8, %xmm7 setbe %r8b movl %r8d, %r9d orb %r10b, %r9b cmpb $0x1, %r9b jne 0xb334eb orb %r8b, %dil je 0xb334eb orb %dl, %r10b jne 0xb340f8 movaps %xmm14, 0x80(%rsp) movaps %xmm6, (%rsp) orb %cl, %al xorps %xmm5, %xmm5 movaps %xmm10, %xmm15 cmpltss %xmm5, %xmm15 movaps %xmm15, %xmm8 movss 0x13bd4b9(%rip), %xmm6 # 0x1ef09cc andps %xmm6, %xmm8 movss 0x13b91f4(%rip), %xmm11 # 0x1eec714 andnps %xmm11, %xmm15 orps %xmm8, %xmm15 movaps %xmm7, %xmm14 cmpltss %xmm5, %xmm14 movaps %xmm14, %xmm8 andps %xmm6, %xmm8 andnps %xmm11, %xmm14 orps %xmm8, %xmm14 ucomiss %xmm14, %xmm15 xorps %xmm8, %xmm8 jne 0xb33557 jp 0xb33557 movss 0x13b84c9(%rip), %xmm8 # 0x1eeba20 xorps %xmm12, %xmm12 jne 0xb33568 jp 0xb33568 movss 0x13b961c(%rip), %xmm12 # 0x1eecb84 movaps %xmm13, %xmm11 cmpltss %xmm5, %xmm11 movaps %xmm11, %xmm5 movss 0x13bd44e(%rip), %xmm6 # 0x1ef09cc andps %xmm6, %xmm5 movss 0x13b918b(%rip), %xmm6 # 0x1eec714 andnps %xmm6, %xmm11 orps %xmm5, %xmm11 ucomiss %xmm11, %xmm15 jne 0xb33599 jnp 0xb335e6 ucomiss %xmm10, %xmm13 movss 0x13b916e(%rip), %xmm15 # 0x1eec714 jne 0xb335f5 jp 0xb335f5 xorps %xmm13, %xmm13 ucomiss %xmm13, %xmm10 movss 0x13b8466(%rip), %xmm5 # 0x1eeba20 jne 0xb335c9 movss 0x13b845c(%rip), %xmm5 # 0x1eeba20 jp 0xb335c9 xorps %xmm5, %xmm5 cmpeqss %xmm13, %xmm10 movd %xmm10, %ecx andl $0x1, %ecx leaq 0x13be8e6(%rip), %rdx # 0x1ef1ec4 movss (%rdx,%rcx,4), %xmm10 jmp 0xb33622 xorps %xmm13, %xmm13 movss 0x13b9121(%rip), %xmm15 # 0x1eec714 jmp 0xb33630 subss %xmm10, %xmm13 xorps 0x13b90ce(%rip), %xmm10 # 0x1eec6d0 divss %xmm13, %xmm10 movaps %xmm15, %xmm5 subss %xmm10, %xmm5 xorps %xmm13, %xmm13 mulss %xmm13, %xmm5 addss %xmm10, %xmm5 movaps %xmm5, %xmm10 minss %xmm5, %xmm8 maxss %xmm12, %xmm10 movaps %xmm10, %xmm12 xorl %ecx, %ecx ucomiss %xmm9, %xmm13 seta %cl leaq 0x13be88c(%rip), %rdx # 0x1ef1ecc movss (%rdx,%rcx,4), %xmm10 ucomiss %xmm10, %xmm14 jne 0xb3364e jnp 0xb33690 ucomiss %xmm7, %xmm9 jne 0xb3369f jp 0xb3369f xorps %xmm9, %xmm9 ucomiss %xmm9, %xmm7 movss 0x13b83ba(%rip), %xmm5 # 0x1eeba20 jne 0xb33675 movss 0x13b83b0(%rip), %xmm5 # 0x1eeba20 jp 0xb33675 xorps %xmm5, %xmm5 cmpeqss %xmm9, %xmm7 movd %xmm7, %ecx andl $0x1, %ecx leaq 0x13be83b(%rip), %rdx # 0x1ef1ec4 movss (%rdx,%rcx,4), %xmm7 jmp 0xb336c4 unpcklps %xmm12, %xmm8 # xmm8 = xmm8[0],xmm12[0],xmm8[1],xmm12[1] movsd 0x13b9053(%rip), %xmm14 # 0x1eec6f0 jmp 0xb336db subss %xmm7, %xmm9 xorps 0x13b9025(%rip), %xmm7 # 0x1eec6d0 divss %xmm9, %xmm7 movaps %xmm15, %xmm5 subss %xmm7, %xmm5 mulss %xmm13, %xmm5 addss %xmm7, %xmm5 movaps %xmm5, %xmm7 movsd 0x13b9023(%rip), %xmm14 # 0x1eec6f0 minss %xmm5, %xmm8 maxss %xmm12, %xmm7 unpcklps %xmm7, %xmm8 # xmm8 = xmm8[0],xmm7[0],xmm8[1],xmm7[1] ucomiss %xmm10, %xmm11 jne 0xb336e3 jnp 0xb3370e movaps %xmm8, %xmm5 cmpltps 0x13b9341(%rip), %xmm5 # 0x1eeca30 movaps %xmm6, %xmm7 movlhps %xmm8, %xmm7 # xmm7 = xmm7[0],xmm8[0] shufps $0xe2, %xmm8, %xmm7 # xmm7 = xmm7[2,0],xmm8[2,3] movss %xmm6, %xmm8 # xmm8 = xmm6[0],xmm8[1,2,3] andps %xmm5, %xmm7 andnps %xmm8, %xmm5 orps %xmm7, %xmm5 movaps %xmm5, %xmm8 movaps %xmm8, %xmm9 cmpltps %xmm14, %xmm9 movaps %xmm8, %xmm5 movss %xmm13, %xmm5 # xmm5 = xmm13[0],xmm5[1,2,3] movaps %xmm6, %xmm7 movlhps %xmm8, %xmm7 # xmm7 = xmm7[0],xmm8[0] shufps $0xe2, %xmm8, %xmm7 # xmm7 = xmm7[2,0],xmm8[2,3] andps %xmm9, %xmm5 andnps %xmm7, %xmm9 orps %xmm5, %xmm9 pshufd $0x55, %xmm9, %xmm5 # xmm5 = xmm9[1,1,1,1] movb $0x1, %r13b ucomiss %xmm5, %xmm9 ja 0xb340e5 addps 0x13be6fd(%rip), %xmm9 # 0x1ef1e50 movlhps %xmm0, %xmm0 # xmm0 = xmm0[0,0] movaps %xmm3, %xmm7 movlhps %xmm3, %xmm7 # xmm7 = xmm7[0],xmm3[0] movaps %xmm4, %xmm8 movlhps %xmm4, %xmm8 # xmm8 = xmm8[0],xmm4[0] movlhps %xmm2, %xmm2 # xmm2 = xmm2[0,0] movhlps %xmm3, %xmm3 # xmm3 = xmm3[1,1] movhlps %xmm4, %xmm4 # xmm4 = xmm4[1,1] movaps %xmm9, %xmm10 cmpltps %xmm14, %xmm10 movaps %xmm9, %xmm5 movss %xmm13, %xmm5 # xmm5 = xmm13[0],xmm5[1,2,3] movaps %xmm15, %xmm11 movlhps %xmm9, %xmm11 # xmm11 = xmm11[0],xmm9[0] shufps $0xe2, %xmm9, %xmm11 # xmm11 = xmm11[2,0],xmm9[2,3] andps %xmm10, %xmm5 andnps %xmm11, %xmm10 orps %xmm5, %xmm10 pshufd $0x50, %xmm10, %xmm5 # xmm5 = xmm10[0,0,1,1] movaps 0x13b926a(%rip), %xmm9 # 0x1eeca10 subps %xmm5, %xmm9 movaps 0x80(%rsp), %xmm12 mulps %xmm5, %xmm12 mulps %xmm5, %xmm3 mulps %xmm5, %xmm4 movaps (%rsp), %xmm11 mulps %xmm5, %xmm11 mulps %xmm9, %xmm0 addps %xmm12, %xmm0 mulps %xmm9, %xmm7 addps %xmm3, %xmm7 mulps %xmm9, %xmm8 addps %xmm4, %xmm8 mulps %xmm2, %xmm9 addps %xmm11, %xmm9 movaps 0x13b9244(%rip), %xmm2 # 0x1eeca30 subps %xmm10, %xmm2 movaps 0x10(%rsp), %xmm12 movaps %xmm12, %xmm3 shufps $0x55, %xmm12, %xmm3 # xmm3 = xmm3[1,1],xmm12[1,1] mulps %xmm10, %xmm3 shufps $0x0, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] mulps %xmm2, %xmm12 addps %xmm3, %xmm12 movaps %xmm12, %xmm2 shufps $0x55, %xmm12, %xmm2 # xmm2 = xmm2[1,1],xmm12[1,1] movaps %xmm15, %xmm4 divss %xmm1, %xmm4 movaps %xmm9, %xmm1 unpckhpd %xmm9, %xmm1 # xmm1 = xmm1[1],xmm9[1] subps %xmm9, %xmm1 subps %xmm8, %xmm9 movaps %xmm8, %xmm3 unpckhpd %xmm8, %xmm3 # xmm3 = xmm3[1],xmm8[1] subps %xmm8, %xmm3 movaps %xmm8, %xmm5 subps %xmm7, %xmm5 movaps %xmm7, %xmm10 unpckhpd %xmm7, %xmm10 # xmm10 = xmm10[1],xmm7[1] subps %xmm7, %xmm10 subps %xmm0, %xmm7 movaps 0x13bd702(%rip), %xmm11 # 0x1ef0f60 mulps %xmm11, %xmm7 mulps %xmm11, %xmm5 mulps %xmm11, %xmm9 movaps %xmm5, %xmm6 minps %xmm9, %xmm6 maxps %xmm9, %xmm5 movaps %xmm7, %xmm8 minps %xmm6, %xmm8 maxps %xmm5, %xmm7 movaps %xmm8, %xmm5 unpckhpd %xmm8, %xmm5 # xmm5 = xmm5[1],xmm8[1] movaps %xmm7, %xmm6 unpckhpd %xmm7, %xmm6 # xmm6 = xmm6[1],xmm7[1] minps %xmm5, %xmm8 maxps %xmm6, %xmm7 shufps $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] mulps %xmm4, %xmm8 mulps %xmm7, %xmm4 subss %xmm12, %xmm2 movaps %xmm15, %xmm7 divss %xmm2, %xmm7 movaps %xmm0, %xmm2 unpckhpd %xmm0, %xmm2 # xmm2 = xmm2[1],xmm0[1] subps %xmm0, %xmm2 movaps %xmm2, %xmm6 minps %xmm10, %xmm6 maxps %xmm10, %xmm2 movaps %xmm3, %xmm0 minps %xmm1, %xmm0 minps %xmm0, %xmm6 maxps %xmm1, %xmm3 maxps %xmm3, %xmm2 shufps $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] mulps %xmm7, %xmm6 mulps %xmm2, %xmm7 movaps 0x20(%rsp), %xmm0 movaps %xmm0, %xmm1 unpcklps %xmm12, %xmm1 # xmm1 = xmm1[0],xmm12[0],xmm1[1],xmm12[1] movaps %xmm1, (%rsp) movaps %xmm12, 0x10(%rsp) movaps %xmm12, %xmm5 shufps $0x11, %xmm0, %xmm5 # xmm5 = xmm5[1,0],xmm0[1,0] shufps $0xe2, %xmm0, %xmm5 # xmm5 = xmm5[2,0],xmm0[2,3] movaps %xmm1, %xmm0 addps %xmm5, %xmm0 mulps 0x13be554(%rip), %xmm0 # 0x1ef1e60 movaps %xmm0, %xmm10 shufps $0x0, %xmm0, %xmm10 # xmm10 = xmm10[0,0],xmm0[0,0] movaps 0x3a0(%rsp), %xmm1 mulps %xmm10, %xmm1 addps 0x120(%rsp), %xmm1 movaps 0x390(%rsp), %xmm2 mulps %xmm10, %xmm2 addps 0x110(%rsp), %xmm2 movaps 0x380(%rsp), %xmm14 mulps %xmm10, %xmm14 addps 0x100(%rsp), %xmm14 subps %xmm2, %xmm14 mulps %xmm10, %xmm14 addps %xmm2, %xmm14 subps %xmm1, %xmm2 mulps %xmm10, %xmm2 addps %xmm1, %xmm2 subps %xmm2, %xmm14 mulps %xmm14, %xmm10 addps %xmm2, %xmm10 mulps %xmm11, %xmm14 movaps %xmm10, %xmm9 movlhps %xmm10, %xmm9 # xmm9 = xmm9[0],xmm10[0] movhlps %xmm10, %xmm10 # xmm10 = xmm10[1,1] movaps %xmm0, %xmm11 shufps $0x55, %xmm0, %xmm11 # xmm11 = xmm11[1,1],xmm0[1,1] subps %xmm9, %xmm10 movaps %xmm14, %xmm1 movlhps %xmm14, %xmm1 # xmm1 = xmm1[0],xmm14[0] movhlps %xmm14, %xmm14 # xmm14 = xmm14[1,1] subps %xmm1, %xmm14 mulps %xmm11, %xmm14 addps %xmm1, %xmm14 movaps %xmm10, %xmm1 movaps 0x13b8d1b(%rip), %xmm3 # 0x1eec6d0 xorps %xmm3, %xmm1 movaps %xmm10, %xmm12 shufps $0x55, %xmm10, %xmm12 # xmm12 = xmm12[1,1],xmm10[1,1] movaps %xmm12, %xmm2 movlhps %xmm14, %xmm1 # xmm1 = xmm1[0],xmm14[0] shufps $0x8, %xmm14, %xmm1 # xmm1 = xmm1[0,2],xmm14[0,0] mulss %xmm14, %xmm12 shufps $0x55, %xmm14, %xmm14 # xmm14 = xmm14[1,1,1,1] movaps %xmm14, %xmm15 mulss %xmm10, %xmm14 movaps %xmm10, %xmm13 mulps %xmm11, %xmm13 movaps (%rsp), %xmm10 subps %xmm0, %xmm10 movaps %xmm5, %xmm11 subps %xmm0, %xmm11 shufps $0x54, %xmm0, %xmm0 # xmm0 = xmm0[0,1,1,1] addps %xmm9, %xmm13 xorps %xmm3, %xmm15 unpcklps %xmm15, %xmm2 # xmm2 = xmm2[0],xmm15[0],xmm2[1],xmm15[1] shufps $0x4, %xmm15, %xmm2 # xmm2 = xmm2[0,1],xmm15[0,0] subss %xmm14, %xmm12 shufps $0x0, %xmm12, %xmm12 # xmm12 = xmm12[0,0,0,0] divps %xmm12, %xmm2 divps %xmm12, %xmm1 movaps %xmm13, %xmm9 shufps $0x0, %xmm13, %xmm9 # xmm9 = xmm9[0,0],xmm13[0,0] mulps %xmm2, %xmm9 shufps $0x55, %xmm13, %xmm13 # xmm13 = xmm13[1,1,1,1] mulps %xmm1, %xmm13 addps %xmm9, %xmm13 subps %xmm13, %xmm0 movaps %xmm2, %xmm9 shufps $0x55, %xmm2, %xmm9 # xmm9 = xmm9[1,1],xmm2[1,1] movaps %xmm8, %xmm12 unpcklps %xmm6, %xmm12 # xmm12 = xmm12[0],xmm6[0],xmm12[1],xmm6[1] movaps %xmm9, %xmm14 mulps %xmm12, %xmm14 movaps %xmm4, %xmm13 unpcklps %xmm7, %xmm13 # xmm13 = xmm13[0],xmm7[0],xmm13[1],xmm7[1] mulps %xmm13, %xmm9 movaps %xmm14, %xmm15 minps %xmm9, %xmm15 maxps %xmm14, %xmm9 movaps %xmm1, %xmm14 shufps $0x55, %xmm1, %xmm14 # xmm14 = xmm14[1,1],xmm1[1,1] shufps $0x11, %xmm8, %xmm6 # xmm6 = xmm6[1,0],xmm8[1,0] shufps $0xe2, %xmm8, %xmm6 # xmm6 = xmm6[2,0],xmm8[2,3] movaps %xmm14, %xmm3 mulps %xmm6, %xmm3 shufps $0x11, %xmm4, %xmm7 # xmm7 = xmm7[1,0],xmm4[1,0] shufps $0xe2, %xmm4, %xmm7 # xmm7 = xmm7[2,0],xmm4[2,3] mulps %xmm7, %xmm14 movaps %xmm3, %xmm8 minps %xmm14, %xmm8 addps %xmm15, %xmm8 maxps %xmm3, %xmm14 addps %xmm9, %xmm14 movsd 0x13b8c3e(%rip), %xmm4 # 0x1eec6f0 movaps %xmm4, %xmm3 subps %xmm14, %xmm3 subps %xmm8, %xmm4 movaps %xmm10, %xmm9 mulps %xmm3, %xmm9 movaps %xmm2, %xmm8 shufps $0x0, %xmm2, %xmm8 # xmm8 = xmm8[0,0],xmm2[0,0] mulps %xmm8, %xmm12 mulps %xmm13, %xmm8 movaps %xmm12, %xmm13 minps %xmm8, %xmm13 maxps %xmm12, %xmm8 movaps %xmm1, %xmm14 shufps $0x0, %xmm1, %xmm14 # xmm14 = xmm14[0,0],xmm1[0,0] mulps %xmm14, %xmm6 mulps %xmm7, %xmm14 movaps %xmm6, %xmm15 minps %xmm14, %xmm15 addps %xmm13, %xmm15 movaps %xmm10, %xmm12 mulps %xmm4, %xmm12 mulps %xmm11, %xmm3 mulps %xmm11, %xmm4 maxps %xmm6, %xmm14 addps %xmm8, %xmm14 movss 0x13b8bf5(%rip), %xmm7 # 0x1eec714 movaps %xmm7, %xmm6 subps %xmm14, %xmm6 subps %xmm15, %xmm7 movaps %xmm10, %xmm8 mulps %xmm6, %xmm8 mulps %xmm7, %xmm10 mulps %xmm11, %xmm6 mulps %xmm11, %xmm7 movaps %xmm8, %xmm11 minps %xmm10, %xmm11 movaps %xmm6, %xmm13 minps %xmm7, %xmm13 minps %xmm13, %xmm11 maxps %xmm8, %xmm10 maxps %xmm6, %xmm7 maxps %xmm10, %xmm7 movaps %xmm9, %xmm6 minps %xmm12, %xmm6 movaps %xmm3, %xmm8 minps %xmm4, %xmm8 minps %xmm8, %xmm6 maxps %xmm9, %xmm12 maxps %xmm3, %xmm4 maxps %xmm12, %xmm4 movaps %xmm11, %xmm3 unpcklps %xmm6, %xmm3 # xmm3 = xmm3[0],xmm6[0],xmm3[1],xmm6[1] shufps $0x11, %xmm11, %xmm6 # xmm6 = xmm6[1,0],xmm11[1,0] shufps $0xe2, %xmm11, %xmm6 # xmm6 = xmm6[2,0],xmm11[2,3] addps %xmm3, %xmm6 movaps %xmm7, %xmm3 unpcklps %xmm4, %xmm3 # xmm3 = xmm3[0],xmm4[0],xmm3[1],xmm4[1] shufps $0x11, %xmm7, %xmm4 # xmm4 = xmm4[1,0],xmm7[1,0] shufps $0xe2, %xmm7, %xmm4 # xmm4 = xmm4[2,0],xmm7[2,3] addps %xmm3, %xmm4 addps %xmm0, %xmm6 addps %xmm0, %xmm4 movaps (%rsp), %xmm7 maxps %xmm6, %xmm7 movaps %xmm4, %xmm3 minps %xmm5, %xmm3 cmpltps %xmm7, %xmm3 unpcklps %xmm3, %xmm3 # xmm3 = xmm3[0,0,1,1] movmskpd %xmm3, %ecx testl %ecx, %ecx jne 0xb340e5 ucomiss %xmm4, %xmm5 seta %dl xorl %ecx, %ecx ucomiss 0x20(%rsp), %xmm6 movaps 0x13b8ae4(%rip), %xmm8 # 0x1eec6c0 movaps 0x370(%rsp), %xmm12 jbe 0xb33c35 testb %dl, %dl xorps %xmm9, %xmm9 movss 0x13b8b1e(%rip), %xmm13 # 0x1eec714 movss 0x13bd3ed(%rip), %xmm14 # 0x1ef0fec movaps 0x120(%rsp), %xmm15 movaps 0x110(%rsp), %xmm10 movaps 0x100(%rsp), %xmm11 je 0xb33c66 shufps $0x55, %xmm6, %xmm6 # xmm6 = xmm6[1,1,1,1] ucomiss 0x10(%rsp), %xmm6 seta %dl cmpltps %xmm5, %xmm4 pextrw $0x2, %xmm4, %ecx andb %dl, %cl jmp 0xb33c66 xorps %xmm9, %xmm9 movss 0x13b8ad2(%rip), %xmm13 # 0x1eec714 movss 0x13bd3a1(%rip), %xmm14 # 0x1ef0fec movaps 0x120(%rsp), %xmm15 movaps 0x110(%rsp), %xmm10 movaps 0x100(%rsp), %xmm11 orb %cl, %al cmpb $0x1, %al jne 0xb340f3 movl $0xc8, %eax movaps %xmm13, %xmm4 subss %xmm0, %xmm4 movaps %xmm4, %xmm3 movaps %xmm0, %xmm5 mulss %xmm0, %xmm5 movaps %xmm0, %xmm6 mulss %xmm5, %xmm6 mulss %xmm14, %xmm5 mulss %xmm4, %xmm5 mulss %xmm4, %xmm4 mulss %xmm4, %xmm3 movaps %xmm0, %xmm7 mulss %xmm14, %xmm7 mulss %xmm4, %xmm7 shufps $0x0, %xmm3, %xmm3 # xmm3 = xmm3[0,0,0,0] shufps $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] shufps $0x0, %xmm5, %xmm5 # xmm5 = xmm5[0,0,0,0] shufps $0x0, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] mulps %xmm12, %xmm6 mulps %xmm11, %xmm5 addps %xmm6, %xmm5 mulps %xmm10, %xmm7 addps %xmm5, %xmm7 mulps %xmm15, %xmm3 addps %xmm7, %xmm3 movaps %xmm3, %xmm4 movlhps %xmm3, %xmm4 # xmm4 = xmm4[0],xmm3[0] movhlps %xmm3, %xmm3 # xmm3 = xmm3[1,1] movaps %xmm0, %xmm5 shufps $0x55, %xmm0, %xmm5 # xmm5 = xmm5[1,1],xmm0[1,1] subps %xmm4, %xmm3 mulps %xmm5, %xmm3 addps %xmm4, %xmm3 movaps %xmm3, %xmm4 shufps $0x0, %xmm3, %xmm4 # xmm4 = xmm4[0,0],xmm3[0,0] mulps %xmm2, %xmm4 movaps %xmm3, %xmm5 shufps $0x55, %xmm3, %xmm5 # xmm5 = xmm5[1,1],xmm3[1,1] mulps %xmm1, %xmm5 addps %xmm4, %xmm5 subps %xmm5, %xmm0 andps %xmm8, %xmm3 pshufd $0x55, %xmm3, %xmm4 # xmm4 = xmm3[1,1,1,1] maxss %xmm3, %xmm4 movaps 0x30(%rsp), %xmm3 ucomiss %xmm4, %xmm3 ja 0xb33d2c decq %rax jne 0xb33c75 jmp 0xb340e5 ucomiss %xmm9, %xmm0 jb 0xb340e5 ucomiss %xmm0, %xmm13 jb 0xb340e5 movaps %xmm0, %xmm2 shufps $0x55, %xmm0, %xmm2 # xmm2 = xmm2[1,1],xmm0[1,1] ucomiss %xmm9, %xmm2 jb 0xb340e5 ucomiss %xmm2, %xmm13 jb 0xb340e5 movss 0x18(%r12), %xmm4 movss 0x28(%r12), %xmm1 unpcklps %xmm1, %xmm4 # xmm4 = xmm4[0],xmm1[0],xmm4[1],xmm1[1] movss 0x38(%r12), %xmm1 movlhps %xmm1, %xmm4 # xmm4 = xmm4[0],xmm1[0] movaps (%rsi), %xmm6 movaps 0xb0(%rsp), %xmm1 subps %xmm6, %xmm1 mulps %xmm4, %xmm1 movaps %xmm1, %xmm3 shufps $0x55, %xmm1, %xmm3 # xmm3 = xmm3[1,1],xmm1[1,1] addss %xmm1, %xmm3 movhlps %xmm1, %xmm1 # xmm1 = xmm1[1,1] addss %xmm3, %xmm1 movaps 0xa0(%rsp), %xmm3 subps %xmm6, %xmm3 mulps %xmm4, %xmm3 movaps %xmm3, %xmm5 shufps $0x55, %xmm3, %xmm5 # xmm5 = xmm5[1,1],xmm3[1,1] addss %xmm3, %xmm5 movhlps %xmm3, %xmm3 # xmm3 = xmm3[1,1] addss %xmm5, %xmm3 movaps 0x90(%rsp), %xmm5 subps %xmm6, %xmm5 mulps %xmm4, %xmm5 movaps %xmm5, %xmm7 shufps $0x55, %xmm5, %xmm7 # xmm7 = xmm7[1,1],xmm5[1,1] addss %xmm5, %xmm7 movhlps %xmm5, %xmm5 # xmm5 = xmm5[1,1] addss %xmm7, %xmm5 movaps 0x70(%rsp), %xmm7 subps %xmm6, %xmm7 mulps %xmm4, %xmm7 movaps %xmm7, %xmm8 shufps $0x55, %xmm7, %xmm8 # xmm8 = xmm8[1,1],xmm7[1,1] addss %xmm7, %xmm8 movhlps %xmm7, %xmm7 # xmm7 = xmm7[1,1] addss %xmm8, %xmm7 movaps 0xc0(%rsp), %xmm8 subps %xmm6, %xmm8 mulps %xmm4, %xmm8 movaps %xmm8, %xmm9 shufps $0x55, %xmm8, %xmm9 # xmm9 = xmm9[1,1],xmm8[1,1] addss %xmm8, %xmm9 movhlps %xmm8, %xmm8 # xmm8 = xmm8[1,1] addss %xmm9, %xmm8 movaps 0xd0(%rsp), %xmm9 subps %xmm6, %xmm9 mulps %xmm4, %xmm9 movaps %xmm9, %xmm10 shufps $0x55, %xmm9, %xmm10 # xmm10 = xmm10[1,1],xmm9[1,1] addss %xmm9, %xmm10 movhlps %xmm9, %xmm9 # xmm9 = xmm9[1,1] addss %xmm10, %xmm9 movaps 0x60(%rsp), %xmm11 subps %xmm6, %xmm11 mulps %xmm4, %xmm11 movaps %xmm11, %xmm10 shufps $0x55, %xmm11, %xmm10 # xmm10 = xmm10[1,1],xmm11[1,1] addss %xmm11, %xmm10 movhlps %xmm11, %xmm11 # xmm11 = xmm11[1,1] addss %xmm10, %xmm11 movaps 0x170(%rsp), %xmm12 subps %xmm6, %xmm12 mulps %xmm4, %xmm12 movaps %xmm12, %xmm4 shufps $0x55, %xmm12, %xmm4 # xmm4 = xmm4[1,1],xmm12[1,1] addss %xmm12, %xmm4 movhlps %xmm12, %xmm12 # xmm12 = xmm12[1,1] addss %xmm4, %xmm12 movaps %xmm13, %xmm10 subss %xmm2, %xmm10 mulss %xmm2, %xmm8 mulss %xmm2, %xmm9 mulss %xmm2, %xmm11 mulss %xmm2, %xmm12 mulss %xmm10, %xmm1 addss %xmm8, %xmm1 mulss %xmm10, %xmm3 addss %xmm9, %xmm3 mulss %xmm10, %xmm5 addss %xmm11, %xmm5 mulss %xmm7, %xmm10 addss %xmm12, %xmm10 movaps %xmm13, %xmm9 subss %xmm0, %xmm9 movaps %xmm9, %xmm7 mulss %xmm9, %xmm7 movaps %xmm9, %xmm4 mulss %xmm7, %xmm4 movaps %xmm0, %xmm6 mulss %xmm14, %xmm6 mulss %xmm7, %xmm6 movaps %xmm0, %xmm7 mulps %xmm0, %xmm7 movaps %xmm0, %xmm8 mulps %xmm7, %xmm8 mulss %xmm14, %xmm7 mulss %xmm9, %xmm7 mulss %xmm8, %xmm10 mulss %xmm7, %xmm5 addss %xmm10, %xmm5 mulss %xmm6, %xmm3 addss %xmm5, %xmm3 mulss %xmm4, %xmm1 addss %xmm3, %xmm1 ucomiss 0xc(%rsi), %xmm1 jb 0xb340e5 movss 0x20(%rsi), %xmm13 ucomiss %xmm1, %xmm13 jb 0xb340e5 movaps 0x13b8ab7(%rip), %xmm5 # 0x1eeca10 subps %xmm2, %xmm5 movaps 0xc0(%rsp), %xmm3 mulps %xmm2, %xmm3 movaps 0xd0(%rsp), %xmm10 mulps %xmm2, %xmm10 movaps 0x60(%rsp), %xmm11 mulps %xmm2, %xmm11 movaps 0xb0(%rsp), %xmm12 mulps %xmm5, %xmm12 addps %xmm3, %xmm12 movaps 0xa0(%rsp), %xmm3 mulps %xmm5, %xmm3 addps %xmm10, %xmm3 movaps 0x90(%rsp), %xmm10 mulps %xmm5, %xmm10 addps %xmm11, %xmm10 mulps 0x170(%rsp), %xmm2 mulps 0x70(%rsp), %xmm5 addps %xmm2, %xmm5 subps %xmm10, %xmm5 subps %xmm3, %xmm10 subps %xmm12, %xmm3 movaps %xmm0, %xmm2 shufps $0x0, %xmm0, %xmm2 # xmm2 = xmm2[0,0],xmm0[0,0] movaps %xmm2, %xmm11 mulps %xmm10, %xmm11 shufps $0x0, %xmm9, %xmm9 # xmm9 = xmm9[0,0,0,0] mulps %xmm9, %xmm3 addps %xmm11, %xmm3 mulps %xmm2, %xmm5 mulps %xmm9, %xmm10 addps %xmm5, %xmm10 mulps %xmm2, %xmm10 mulps %xmm9, %xmm3 addps %xmm10, %xmm3 mulps 0x13bcf5b(%rip), %xmm3 # 0x1ef0f60 movq (%rbp), %rax movq 0x1e8(%rax), %rax movq %rbp, %rcx movq 0x50(%rsp), %rdx movq (%rax,%rdx,8), %rbp movl 0x24(%rsi), %eax testl %eax, 0x34(%rbp) je 0xb340e2 movq %rbx, %r15 movq %r12, %rbx shufps $0x0, %xmm8, %xmm8 # xmm8 = xmm8[0,0,0,0] mulps 0x280(%rsp), %xmm8 shufps $0x0, %xmm7, %xmm7 # xmm7 = xmm7[0,0,0,0] mulps 0x290(%rsp), %xmm7 addps %xmm8, %xmm7 shufps $0x0, %xmm6, %xmm6 # xmm6 = xmm6[0,0,0,0] mulps 0x2a0(%rsp), %xmm6 addps %xmm7, %xmm6 shufps $0x0, %xmm4, %xmm4 # xmm4 = xmm4[0,0,0,0] mulps 0x2b0(%rsp), %xmm4 addps %xmm6, %xmm4 movaps %xmm4, %xmm2 shufps $0xc9, %xmm4, %xmm2 # xmm2 = xmm2[1,2],xmm4[0,3] movaps %xmm3, %xmm5 shufps $0xc9, %xmm3, %xmm5 # xmm5 = xmm5[1,2],xmm3[0,3] mulps %xmm4, %xmm5 mulps %xmm3, %xmm2 subps %xmm5, %xmm2 movq %rcx, %r12 movq 0x10(%rcx), %rax cmpq $0x0, 0x10(%rax) jne 0xb340fd cmpq $0x0, 0x40(%rbp) jne 0xb340fd movss %xmm1, 0x20(%rsi) movaps %xmm2, %xmm1 shufps $0xe9, %xmm2, %xmm1 # xmm1 = xmm1[1,2],xmm2[2,3] movlps %xmm1, 0x30(%rsi) movss %xmm2, 0x38(%rsi) movlps %xmm0, 0x3c(%rsi) movq 0xe8(%rsp), %rax movl %eax, 0x44(%rsi) movq 0x50(%rsp), %rax movl %eax, 0x48(%rsi) movq %r12, %rbp movq 0x8(%r12), %rax movl (%rax), %eax movl %eax, 0x4c(%rsi) movq 0x8(%r12), %rax movl 0x4(%rax), %eax movl %eax, 0x50(%rsi) movq %rbx, %r12 movq %r15, %rbx jmp 0xb340e5 movq %rcx, %rbp testb %r13b, %r13b jne 0xb3317c jmp 0xb32abc xorl %r13d, %r13d jmp 0xb340e5 movb $0x1, %r13b jmp 0xb340e5 movq 0x8(%r12), %rax movaps %xmm2, %xmm3 shufps $0xe9, %xmm2, %xmm3 # xmm3 = xmm3[1,2],xmm2[2,3] movlps %xmm3, 0x240(%rsp) movss %xmm2, 0x248(%rsp) movlps %xmm0, 0x24c(%rsp) movq 0xe8(%rsp), %rcx movl %ecx, 0x254(%rsp) movq 0x50(%rsp), %rcx movl %ecx, 0x258(%rsp) movl (%rax), %ecx movl %ecx, 0x25c(%rsp) movl 0x4(%rax), %eax movl %eax, 0x260(%rsp) movss %xmm1, 0x20(%rsi) movl $0xffffffff, 0x5c(%rsp) # imm = 0xFFFFFFFF leaq 0x5c(%rsp), %rax movq %rax, 0x140(%rsp) movq 0x18(%rbp), %rax movq %rax, 0x148(%rsp) movq 0x8(%r12), %rax movq %rax, 0x150(%rsp) movq %rsi, 0x158(%rsp) leaq 0x240(%rsp), %rax movq %rax, 0x160(%rsp) movl $0x1, 0x168(%rsp) movq 0x40(%rbp), %rax testq %rax, %rax movss %xmm13, (%rsp) je 0xb341eb leaq 0x140(%rsp), %rdi movq %r11, 0x80(%rsp) callq %rax movss (%rsp), %xmm13 movq 0x80(%rsp), %r11 movq 0x48(%rsp), %rsi movq 0x140(%rsp), %rax cmpl $0x0, (%rax) je 0xb34287 movq 0x10(%r12), %rcx movq 0x10(%rcx), %rax testq %rax, %rax je 0xb3422c testb $0x2, (%rcx) jne 0xb34204 testb $0x40, 0x3e(%rbp) je 0xb3421f leaq 0x140(%rsp), %rdi movq %r11, %rbp callq %rax movss (%rsp), %xmm13 movq %rbp, %r11 movq 0x48(%rsp), %rsi movq 0x140(%rsp), %rax cmpl $0x0, (%rax) je 0xb34287 movq 0x158(%rsp), %rax movq 0x160(%rsp), %rcx movss (%rcx), %xmm0 movss %xmm0, 0x30(%rax) movss 0x4(%rcx), %xmm0 movss %xmm0, 0x34(%rax) movss 0x8(%rcx), %xmm0 movss %xmm0, 0x38(%rax) movss 0xc(%rcx), %xmm0 movss %xmm0, 0x3c(%rax) movss 0x10(%rcx), %xmm0 movss %xmm0, 0x40(%rax) movl 0x14(%rcx), %edx movl %edx, 0x44(%rax) movl 0x18(%rcx), %edx movl %edx, 0x48(%rax) movl 0x1c(%rcx), %edx movl %edx, 0x4c(%rax) movl 0x20(%rcx), %ecx movl %ecx, 0x50(%rax) jmp 0xb3428d movss %xmm13, 0x20(%rsi) movq %r12, %rbp jmp 0xb340da movaps 0x20(%rsi), %xmm0 shufps $0x0, %xmm0, %xmm0 # xmm0 = xmm0[0,0,0,0] movaps 0x270(%rsp), %xmm1 cmpleps %xmm0, %xmm1 movmskps %xmm1, %eax andl %r11d, %ebx andl %eax, %ebx jne 0xb31d4c addq $0x3f8, %rsp # imm = 0x3F8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq nop	/embree[P]embree/kernels/geometry/curveNi_mb_intersector.h
embree::sse2::createPrimRefArrayMBlur(embree::Scene*, embree::Geometry::GTypeMask, unsigned long, embree::vector_t<embree::PrimRef, embree::aligned_monitored_allocator<embree::PrimRef, 32ul>>&, embree::BuildProgressMonitor&, unsigned long)	PrimInfo createPrimRefArrayMBlur(Scene* scene, Geometry::GTypeMask types, const size_t numPrimRefs, mvector<PrimRef>& prims, BuildProgressMonitor& progressMonitor, size_t itime) { ParallelForForPrefixSumState<PrimInfo> pstate; Scene::Iterator2 iter(scene,types,true); /* first try / progressMonitor(0); pstate.init(iter,size_t(1024)); PrimInfo pinfo = parallel_for_for_prefix_sum0( pstate, iter, PrimInfo(empty), [&](Geometry mesh, const range<size_t>& r, size_t k, size_t geomID) -> PrimInfo { return mesh->createPrimRefArrayMB(prims,itime,r,k,(unsigned)geomID); }, [](const PrimInfo& a, const PrimInfo& b) -> PrimInfo { return PrimInfo::merge(a,b); }); /* if we need to filter out geometry, run again / if (pinfo.size() != numPrimRefs) { progressMonitor(0); pinfo = parallel_for_for_prefix_sum1( pstate, iter, PrimInfo(empty), [&](Geometry mesh, const range<size_t>& r, size_t k, size_t geomID, const PrimInfo& base) -> PrimInfo { return mesh->createPrimRefArrayMB(prims,itime,r,base.size(),(unsigned)geomID); }, [](const PrimInfo& a, const PrimInfo& b) -> PrimInfo { return PrimInfo::merge(a,b); }); } return pinfo; }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x2db8, %rsp # imm = 0x2DB8 movq %r8, 0x40(%rsp) movq %rcx, 0x58(%rsp) movq %rdi, %rbx movq $0x0, 0x5a0(%rsp) movq %rsi, 0x28(%rsp) movq %rdx, 0x30(%rsp) movb $0x1, 0x38(%rsp) movq (%r9), %rax xorl %r13d, %r13d movq %r9, 0x50(%rsp) movq %r9, %rdi xorl %esi, %esi callq (%rax) movq 0x28(%rsp), %rcx movq 0x1d8(%rcx), %rax testq %rax, %rax je 0xb3db68 movq 0x1e8(%rcx), %rdx movq 0x30(%rsp), %rsi movb 0x38(%rsp), %dil xorl %r13d, %r13d xorl %r8d, %r8d movq (%rdx,%r8,8), %r9 testq %r9, %r9 je 0xb3db5b movl 0x3c(%r9), %ecx btl $0x15, %ecx jae 0xb3db5b movl $0x1, %r10d shll %cl, %r10d movslq %r10d, %rcx testq %rcx, %rsi je 0xb3db5b cmpl $0x1, 0x24(%r9) setne %cl cmpb %cl, %dil jne 0xb3db5b movl 0x20(%r9), %ecx jmp 0xb3db5d xorl %ecx, %ecx addq %rcx, %r13 incq %r8 cmpq %r8, %rax jne 0xb3db24 movq %r13, 0x5a8(%rsp) xorl %r15d, %r15d xorl %edi, %edi callq 0x6ab80 cltq leaq 0x3ff(%r13), %rcx shrq $0xa, %rcx cmpq %rcx, %rax cmovbq %rax, %rcx cmpq $0x40, %rcx movl $0x40, %r14d cmovbq %rcx, %r14 cmpq $0x1, %r14 adcq %r15, %r14 movq %r14, 0x5a0(%rsp) movq %r15, 0x1a0(%rsp) movq %r15, 0x3a0(%rsp) cmpq $0x2, %rcx jb 0xb3dcb2 xorl %esi, %esi movq %r13, %rax xorl %edx, %edx divq %r14 movq 0x28(%rsp), %rdi movq 0x30(%rsp), %r8 movl $0x1, %r9d movb 0x38(%rsp), %r10b xorl %r11d, %r11d movq 0x1e8(%rdi), %rcx movq (%rcx,%r11,8), %rdx testq %rdx, %rdx je 0xb3dc21 movl 0x3c(%rdx), %ecx btl $0x15, %ecx jae 0xb3dc21 movl $0x1, %ebp shll %cl, %ebp movslq %ebp, %rcx testq %rcx, %r8 je 0xb3dc21 cmpl $0x1, 0x24(%rdx) setne %cl cmpb %cl, %r10b jne 0xb3dc21 movl 0x20(%rdx), %r14d jmp 0xb3dc24 xorl %r14d, %r14d testq %r14, %r14 je 0xb3dc8e leaq 0x1(%r9), %rcx imulq %r13, %rcx xorl %edx, %edx xorl %r15d, %r15d movq %rsi, %r12 movq %rax, %rsi leaq (%r12,%r14), %rax addq %rdx, %rax cmpq %rsi, %rax jb 0xb3dc92 cmpq 0x5a0(%rsp), %r9 jae 0xb3dc92 movq %r11, 0x1a0(%rsp,%r9,8) movq %rsi, %rax subq %r12, %rax addq %rax, %r15 movq %r15, 0x3a0(%rsp,%r9,8) incq %r9 movq %rcx, %rax xorl %edx, %edx divq 0x5a0(%rsp) movq %r15, %rdx negq %rdx addq %r13, %rcx movq %rsi, %r12 cmpq %r14, %r15 jb 0xb3dc39 jmp 0xb3dc98 xorl %edx, %edx jmp 0xb3dc98 movq %rsi, %rax movq %r12, %rsi addq %r14, %rsi addq %rdx, %rsi incq %r11 movq 0x5a0(%rsp), %r14 cmpq %r14, %r9 jb 0xb3dbe7 movaps 0x13add37(%rip), %xmm0 # 0x1eeb9f0 leaq 0x80(%rsp), %rbp movaps %xmm0, (%rbp) movaps 0x13add34(%rip), %xmm1 # 0x1eeba00 movaps %xmm1, 0x10(%rbp) movaps %xmm0, 0x20(%rbp) movaps %xmm1, 0x30(%rbp) xorps %xmm0, %xmm0 movaps %xmm0, 0x40(%rbp) leaq 0x70(%rsp), %r15 movq 0x40(%rsp), %rax movq %rax, (%r15) leaq 0x2df0(%rsp), %rax movq %rax, 0x8(%r15) leaq 0x28(%rsp), %rcx leaq 0x48(%rsp), %r13 movq %rcx, (%r13) leaq 0x60(%rsp), %rax movq %r15, (%rax) movq %rcx, 0x8(%rax) leaq 0x10(%rsp), %rcx movq %r14, (%rcx) leaq 0x1a0(%rsp), %rdx leaq 0xd0(%rsp), %r12 movq %rdx, (%r12) movq %rcx, 0x8(%r12) movq %rbp, 0x10(%r12) movq %r13, 0x18(%r12) leaq 0xf(%rsp), %rcx movq %rcx, 0x20(%r12) movq %rax, 0x28(%r12) leaq 0x120(%rsp), %rdi movw $0x401, 0xc(%rdi) # imm = 0x401 movups %xmm0, 0x20(%rdi) movq $0x8, 0x40(%rdi) callq 0x6a660 movq %r12, 0x18(%rsp) testq %r14, %r14 je 0xb3de53 leaq 0x20(%rsp), %rdi movq $0x0, (%rdi) movl $0xc0, %esi callq 0x6ac50 movq %rax, %r13 xorps %xmm0, %xmm0 movups %xmm0, 0x28(%rax) movups %xmm0, 0x18(%rax) movups %xmm0, 0x8(%rax) xorl %eax, %eax movq %rax, 0x38(%r13) leaq 0x15c91ca(%rip), %rcx # 0x2106f78 movq %rcx, (%r13) movq %r14, 0x40(%r13) movq %rax, 0x48(%r13) movl $0x1, %r14d movq %r14, 0x50(%r13) leaq 0x18(%rsp), %rcx movq %rcx, 0x58(%r13) movq %rax, 0x60(%r13) movq %r14, 0x68(%r13) xorl %edi, %edi callq 0x6ab80 cltq movabsq $0x3fffffffffffffff, %rcx # imm = 0x3FFFFFFFFFFFFFFF andq %rax, %rcx movl $0x0, 0x80(%r13) movb $0x5, 0x84(%r13) addq %rcx, %rcx movq %rcx, 0x78(%r13) movq 0x20(%rsp), %rax movq %rax, 0x88(%r13) leaq 0x110(%rsp), %rdx movq $0x0, -0x10(%rdx) movl $0x1, -0x8(%rdx) movq %r14, (%rdx) movq %r14, 0x8(%rdx) leaq 0x100(%rsp), %rax movq %rax, 0x70(%r13) leaq 0x120(%rsp), %rcx movq %r13, %rdi movq %rcx, %rsi callq 0x6a550 leaq 0x48(%rsp), %r13 leaq 0x120(%rsp), %rdi callq 0x6a770 testb %al, %al jne 0xb3e1b4 leaq 0x120(%rsp), %rdi callq 0x6aab0 leaq 0xc0(%rsp), %rax movaps -0x40(%rax), %xmm0 movaps %xmm0, (%rbx) movaps -0x30(%rax), %xmm1 movaps %xmm1, 0x10(%rbx) movaps -0x20(%rax), %xmm2 movaps %xmm2, 0x20(%rbx) movaps -0x10(%rax), %xmm3 movaps %xmm3, 0x30(%rbx) movdqa (%rax), %xmm4 movdqa %xmm4, 0x40(%rbx) movq 0x10(%rsp), %rax testq %rax, %rax je 0xb3df06 leaq 0x40(%rbx), %rcx movdqa (%rcx), %xmm4 leaq 0x19f0(%rsp), %rdx movaps %xmm0, -0x40(%rdx) movaps %xmm1, -0x30(%rdx) movaps %xmm2, -0x20(%rdx) movaps %xmm3, -0x10(%rdx) movaps (%rcx), %xmm5 movaps %xmm5, (%rdx) minps -0x1440(%rdx), %xmm0 maxps -0x1430(%rdx), %xmm1 minps -0x1420(%rdx), %xmm2 maxps -0x1410(%rdx), %xmm3 paddq -0x1400(%rdx), %xmm4 movdqa %xmm4, (%rcx) addq $0x50, %rdx decq %rax jne 0xb3debf movaps %xmm0, (%rbx) movaps %xmm1, 0x10(%rbx) movaps %xmm2, 0x20(%rbx) movaps %xmm3, 0x30(%rbx) movq 0x48(%rbx), %rax subq 0x40(%rbx), %rax cmpq 0x58(%rsp), %rax je 0xb3e19f movq 0x50(%rsp), %rdi movq (%rdi), %rax xorl %esi, %esi callq (%rax) movaps 0x13adab5(%rip), %xmm0 # 0x1eeb9f0 movaps %xmm0, 0x80(%rsp) movaps 0x13adab6(%rip), %xmm1 # 0x1eeba00 movaps %xmm1, 0x90(%rsp) movaps %xmm0, 0xa0(%rsp) movaps %xmm1, 0xb0(%rsp) xorps %xmm0, %xmm0 movaps %xmm0, 0xc0(%rsp) movq 0x40(%rsp), %rax movq %rax, 0x70(%rsp) leaq 0x2df0(%rsp), %rax movq %rax, 0x78(%rsp) leaq 0x28(%rsp), %rax movq %rax, 0x48(%rsp) movq %r15, 0x60(%rsp) movq %rax, 0x68(%rsp) movq 0x5a0(%rsp), %r15 movq %r15, 0x10(%rsp) leaq 0x1a0(%rsp), %rax movq %rax, 0xd0(%rsp) leaq 0x10(%rsp), %rax movq %rax, 0xd8(%rsp) movq %rbp, 0xe0(%rsp) movq %r13, 0xe8(%rsp) leaq 0xf(%rsp), %rax movq %rax, 0xf0(%rsp) leaq 0x60(%rsp), %rax movq %rax, 0xf8(%rsp) leaq 0x120(%rsp), %rdi movw $0x401, 0xc(%rdi) # imm = 0x401 movups %xmm0, 0x20(%rdi) movq $0x8, 0x40(%rdi) callq 0x6a660 movq %r12, 0x18(%rsp) testq %r15, %r15 je 0xb3e0ed leaq 0x20(%rsp), %rdi movq $0x0, (%rdi) movl $0xc0, %esi callq 0x6ac50 movq %rax, %r14 xorps %xmm0, %xmm0 movups %xmm0, 0x28(%rax) movups %xmm0, 0x18(%rax) movups %xmm0, 0x8(%rax) xorl %eax, %eax movq %rax, 0x38(%r14) leaq 0x15c8f72(%rip), %rcx # 0x2106fc0 movq %rcx, (%r14) movq %r15, 0x40(%r14) movq %rax, 0x48(%r14) movl $0x1, %r15d movq %r15, 0x50(%r14) leaq 0x18(%rsp), %rcx movq %rcx, 0x58(%r14) movq %rax, 0x60(%r14) movq %r15, 0x68(%r14) xorl %edi, %edi callq 0x6ab80 cltq movabsq $0x3fffffffffffffff, %rcx # imm = 0x3FFFFFFFFFFFFFFF andq %rax, %rcx movl $0x0, 0x80(%r14) movb $0x5, 0x84(%r14) addq %rcx, %rcx movq %rcx, 0x78(%r14) movq 0x20(%rsp), %rax movq %rax, 0x88(%r14) leaq 0x110(%rsp), %rdx movq $0x0, -0x10(%rdx) movl $0x1, -0x8(%rdx) movq %r15, (%rdx) movq %r15, 0x8(%rdx) leaq 0x100(%rsp), %rax movq %rax, 0x70(%r14) leaq 0x120(%rsp), %rcx movq %r14, %rdi movq %rcx, %rsi callq 0x6a550 leaq 0x120(%rsp), %rdi callq 0x6a770 testb %al, %al jne 0xb3e1e6 leaq 0x120(%rsp), %rdi callq 0x6aab0 movaps 0x80(%rsp), %xmm3 movaps 0x90(%rsp), %xmm2 movaps 0xa0(%rsp), %xmm1 movaps 0xb0(%rsp), %xmm0 movdqa 0xc0(%rsp), %xmm4 movq 0x10(%rsp), %rax testq %rax, %rax je 0xb3e18b leaq 0x19f0(%rsp), %rcx movaps %xmm3, -0x40(%rcx) movaps %xmm2, -0x30(%rcx) movaps %xmm1, -0x20(%rcx) movaps %xmm0, -0x10(%rcx) movdqa %xmm4, (%rcx) minps -0x1440(%rcx), %xmm3 maxps -0x1430(%rcx), %xmm2 minps -0x1420(%rcx), %xmm1 maxps -0x1410(%rcx), %xmm0 paddq -0x1400(%rcx), %xmm4 addq $0x50, %rcx decq %rax jne 0xb3e14a movaps %xmm3, (%rbx) movaps %xmm2, 0x10(%rbx) movaps %xmm1, 0x20(%rbx) movaps %xmm0, 0x30(%rbx) movdqa %xmm4, 0x40(%rbx) movq %rbx, %rax addq $0x2db8, %rsp # imm = 0x2DB8 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movl $0x10, %edi callq 0x6a3b0 movq %rax, %r14 leaq 0x13ad928(%rip), %rsi # 0x1eebaf0 movq %rax, %rdi callq 0x6a230 movq 0x15e6a31(%rip), %rsi # 0x2124c08 movq 0x15e67ea(%rip), %rdx # 0x21249c8 movq %r14, %rdi callq 0x6a5d0 movl $0x10, %edi callq 0x6a3b0 movq %rax, %r14 leaq 0x13ad8f6(%rip), %rsi # 0x1eebaf0 movq %rax, %rdi callq 0x6a230 movq 0x15e69ff(%rip), %rsi # 0x2124c08 movq 0x15e67b8(%rip), %rdx # 0x21249c8 movq %r14, %rdi callq 0x6a5d0 movq %rax, %rbx movq %r14, %rdi callq 0x6a8a0 jmp 0xb3e239 jmp 0xb3e266 movq %rax, %rbx movq %r14, %rdi callq 0x6a8a0 jmp 0xb3e251 jmp 0xb3e236 movq %rax, %rbx leaq 0x120(%rsp), %rdi callq 0x6aab0 jmp 0xb3e25e jmp 0xb3e266 jmp 0xb3e266 jmp 0xb3e24e movq %rax, %rbx leaq 0x120(%rsp), %rdi callq 0x6aab0 movq %rbx, %rdi callq 0x6a600 movq %rax, %rdi callq 0x8d6de8	/embree[P]embree/kernels/builders/primrefgen.cpp
unsigned long embree::sse2::createMortonCodeArray<embree::Instance>(embree::Instance*, embree::vector_t<embree::sse2::BVHBuilderMorton::BuildPrim, embree::aligned_monitored_allocator<embree::sse2::BVHBuilderMorton::BuildPrim, 8ul>>&, embree::BuildProgressMonitor&)	size_t createMortonCodeArray(Mesh* mesh, mvector<BVHBuilderMorton::BuildPrim>& morton, BuildProgressMonitor& progressMonitor) { size_t numPrimitives = morton.size(); /* compute scene bounds / std::pair<size_t,BBox3fa> cb_empty(0,empty); auto cb = parallel_reduce ( size_t(0), numPrimitives, size_t(1024), cb_empty, [&](const range<size_t>& r) -> std::pair<size_t,BBox3fa> { size_t num = 0; BBox3fa bounds = empty; for (size_t j=r.begin(); j<r.end(); j++) { BBox3fa prim_bounds = empty; if (unlikely(!mesh->buildBounds(j,&prim_bounds))) continue; bounds.extend(center2(prim_bounds)); num++; } return std::make_pair(num,bounds); }, [] (const std::pair<size_t,BBox3fa>& a, const std::pair<size_t,BBox3fa>& b) { return std::make_pair(a.first + b.first,merge(a.second,b.second)); }); size_t numPrimitivesGen = cb.first; const BBox3fa centBounds = cb.second; / compute morton codes / if (likely(numPrimitivesGen == numPrimitives)) { / fast path if all primitives were valid / BVHBuilderMorton::MortonCodeMapping mapping(centBounds); parallel_for( size_t(0), numPrimitives, size_t(1024), [&](const range<size_t>& r) -> void { BVHBuilderMorton::MortonCodeGenerator generator(mapping,&morton.data()[r.begin()]); for (size_t j=r.begin(); j<r.end(); j++) generator(mesh->bounds(j),unsigned(j)); }); } else { / slow path, fallback in case some primitives were invalid */ ParallelPrefixSumState<size_t> pstate; BVHBuilderMorton::MortonCodeMapping mapping(centBounds); parallel_prefix_sum( pstate, size_t(0), numPrimitives, size_t(1024), size_t(0), [&](const range<size_t>& r, const size_t base) -> size_t { size_t num = 0; BVHBuilderMorton::MortonCodeGenerator generator(mapping,&morton.data()[r.begin()]); for (size_t j=r.begin(); j<r.end(); j++) { BBox3fa bounds = empty; if (unlikely(!mesh->buildBounds(j,&bounds))) continue; generator(bounds,unsigned(j)); num++; } return num; }, std::plus<size_t>()); parallel_prefix_sum( pstate, size_t(0), numPrimitives, size_t(1024), size_t(0), [&](const range<size_t>& r, const size_t base) -> size_t { size_t num = 0; BVHBuilderMorton::MortonCodeGenerator generator(mapping,&morton.data()[base]); for (size_t j=r.begin(); j<r.end(); j++) { BBox3fa bounds = empty; if (!mesh->buildBounds(j,&bounds)) continue; generator(bounds,unsigned(j)); num++; } return num; }, std::plus<size_t>()); } return numPrimitivesGen; }	pushq %rbp pushq %r15 pushq %r14 pushq %r13 pushq %r12 pushq %rbx subq $0x568, %rsp # imm = 0x568 leaq 0x68(%rsp), %rax movq %rdi, (%rax) movq %rsi, %r14 movq 0x10(%rsi), %rbp xorl %r12d, %r12d leaq 0x130(%rsp), %r15 movq %r12, (%r15) movaps 0x13a09cd(%rip), %xmm0 # 0x1eeb9f0 movaps %xmm0, 0x10(%r15) movaps 0x13a09d1(%rip), %xmm0 # 0x1eeba00 movaps %xmm0, 0x20(%r15) leaq 0x28(%rsp), %r13 movq %rax, (%r13) leaq 0x168(%rsp), %rbx movw $0x401, 0xc(%rbx) # imm = 0x401 xorps %xmm0, %xmm0 movups %xmm0, 0x20(%rbx) movq $0x8, 0x40(%rbx) movq %rbx, %rdi callq 0x6a660 leaq 0x40(%rsp), %rdi movq %rbp, (%rdi) movq %r12, 0x8(%rdi) movq $0x400, 0x10(%rdi) # imm = 0x400 leaq 0x8(%rsp), %rax leaq 0x70(%rsp), %r12 movq %rax, (%r12) movq %r13, 0x8(%r12) leaq 0xb0(%rsp), %r13 movq %r15, (%r13) movq %r12, 0x8(%r13) movq %rax, 0x10(%r13) movq (%r15), %rax movq %rax, 0x20(%r13) movaps 0x10(%r15), %xmm0 movaps %xmm0, 0x30(%r13) movaps 0x20(%r15), %xmm0 movaps %xmm0, 0x40(%r13) leaq 0x20(%rsp), %rdx movq %r13, %rsi movq %rbx, %rcx callq 0xb54b10 movq 0xd0(%rsp), %rbx movaps 0xe0(%rsp), %xmm0 movaps %xmm0, 0x90(%rsp) movaps 0xf0(%rsp), %xmm0 movaps %xmm0, 0xa0(%rsp) leaq 0x168(%rsp), %rdi callq 0x6a770 testb %al, %al leaq 0x68(%rsp), %r15 jne 0xb4b488 leaq 0x168(%rsp), %rdi callq 0x6aab0 movaps 0xa0(%rsp), %xmm3 movaps 0x90(%rsp), %xmm0 subps %xmm0, %xmm3 cmpq %rbp, %rbx jne 0xb4b20b movaps %xmm0, 0xb0(%rsp) movaps %xmm3, %xmm0 cmpnleps 0x13a18ba(%rip), %xmm0 # 0x1eeca00 rcpps %xmm3, %xmm1 mulps %xmm1, %xmm3 movaps 0x13a18bd(%rip), %xmm2 # 0x1eeca10 subps %xmm3, %xmm2 mulps %xmm1, %xmm2 addps %xmm1, %xmm2 mulps 0x13a18bd(%rip), %xmm2 # 0x1eeca20 andps %xmm0, %xmm2 movaps %xmm2, 0xc0(%rsp) movq %r13, 0x70(%rsp) movq %r14, 0x78(%rsp) movq %r15, 0x80(%rsp) leaq 0x168(%rsp), %r14 movw $0x401, 0xc(%r14) # imm = 0x401 xorps %xmm0, %xmm0 movups %xmm0, 0x20(%r14) movq $0x8, 0x40(%r14) movq %r14, %rdi callq 0x6a660 leaq 0x40(%rsp), %rdi movq %rbp, (%rdi) movq $0x0, 0x8(%rdi) movq $0x400, 0x10(%rdi) # imm = 0x400 leaq 0x28(%rsp), %rsi movq %r12, (%rsi) leaq 0x20(%rsp), %rdx movq %r14, %rcx callq 0xb55564 leaq 0x168(%rsp), %rdi callq 0x6a770 testb %al, %al jne 0xb4b4ba leaq 0x168(%rsp), %rdi callq 0x6aab0 movq %rbx, %rax addq $0x568, %rsp # imm = 0x568 popq %rbx popq %r12 popq %r13 popq %r14 popq %r15 popq %rbp retq movaps %xmm0, 0x70(%rsp) movaps %xmm3, %xmm0 cmpnleps 0x13a17e5(%rip), %xmm0 # 0x1eeca00 rcpps %xmm3, %xmm1 mulps %xmm1, %xmm3 movaps 0x13a17e8(%rip), %xmm2 # 0x1eeca10 subps %xmm3, %xmm2 mulps %xmm1, %xmm2 addps %xmm1, %xmm2 mulps 0x13a17e8(%rip), %xmm2 # 0x1eeca20 andps %xmm0, %xmm2 movaps %xmm2, 0x80(%rsp) movq %r12, 0x28(%rsp) movq %r14, 0x30(%rsp) movq %r15, 0x38(%rsp) movq $0x0, 0x8(%rsp) leaq 0x18(%rsp), %r13 movq %rbp, (%r13) xorl %edi, %edi callq 0x6ab80 cltq movq (%r13), %rcx movq %r13, %rdx subq 0x8(%rsp), %rcx addq $0x3ff, %rcx # imm = 0x3FF shrq $0xa, %rcx cmpq %rcx, %rax cmovbq %rax, %rcx cmpq $0x40, %rcx movl $0x40, %r13d cmovbq %rcx, %r13 leaq 0x10(%rsp), %rcx movq %r13, (%rcx) leaq 0x8(%rsp), %rax movq %rax, 0x40(%rsp) movq %rdx, 0x48(%rsp) movq %rcx, 0x50(%rsp) leaq 0x168(%rsp), %rax movq %rax, 0x58(%rsp) leaq 0x28(%rsp), %rax movq %rax, 0x60(%rsp) leaq 0xb0(%rsp), %rdi movw $0x401, 0xc(%rdi) # imm = 0x401 xorps %xmm0, %xmm0 movups %xmm0, 0x20(%rdi) movq $0x8, 0x40(%rdi) callq 0x6a660 leaq 0x20(%rsp), %rcx leaq 0x40(%rsp), %rax movq %rax, (%rcx) leaq 0x7(%rsp), %r8 movl $0x1, %edx xorl %edi, %edi movq %r13, %rsi leaq 0xb0(%rsp), %r9 callq 0xb56103 leaq 0xb0(%rsp), %rdi callq 0x6a770 testb %al, %al jne 0xb4b4ec leaq 0xb0(%rsp), %rdi callq 0x6aab0 movq 0x10(%rsp), %rax testq %rax, %rax je 0xb4b35f xorl %ecx, %ecx xorl %edx, %edx movq %rdx, 0x368(%rsp,%rcx,8) addq 0x168(%rsp,%rcx,8), %rdx incq %rcx cmpq %rcx, %rax jne 0xb4b347 movq %r12, 0x28(%rsp) movq %r14, 0x30(%rsp) movq %r15, 0x38(%rsp) movq $0x0, 0x8(%rsp) movq %rbp, 0x18(%rsp) xorl %edi, %edi callq 0x6ab80 cltq movq 0x18(%rsp), %rcx subq 0x8(%rsp), %rcx addq $0x3ff, %rcx # imm = 0x3FF shrq $0xa, %rcx cmpq %rcx, %rax cmovbq %rax, %rcx cmpq $0x40, %rcx movl $0x40, %r14d cmovbq %rcx, %r14 movq %r14, 0x10(%rsp) leaq 0x8(%rsp), %rax movq %rax, 0x40(%rsp) leaq 0x18(%rsp), %rax movq %rax, 0x48(%rsp) leaq 0x10(%rsp), %rax movq %rax, 0x50(%rsp) leaq 0x168(%rsp), %rax movq %rax, 0x58(%rsp) leaq 0x28(%rsp), %rax movq %rax, 0x60(%rsp) leaq 0xb0(%rsp), %r12 movw $0x401, 0xc(%r12) # imm = 0x401 xorps %xmm0, %xmm0 movups %xmm0, 0x20(%r12) movq $0x8, 0x40(%r12) movq %r12, %rdi callq 0x6a660 leaq 0x20(%rsp), %rcx leaq 0x40(%rsp), %rax movq %rax, (%rcx) leaq 0x7(%rsp), %r8 movl $0x1, %edx xorl %edi, %edi movq %r14, %rsi movq %r12, %r9 callq 0xb56e35 leaq 0xb0(%rsp), %rdi callq 0x6a770 testb %al, %al jne 0xb4b51e leaq 0xb0(%rsp), %rdi callq 0x6aab0 movq 0x10(%rsp), %rax testq %rax, %rax je 0xb4b1f6 xorl %ecx, %ecx xorl %edx, %edx movq %rdx, 0x368(%rsp,%rcx,8) addq 0x168(%rsp,%rcx,8), %rdx incq %rcx cmpq %rcx, %rax jne 0xb4b46b jmp 0xb4b1f6 movl $0x10, %edi callq 0x6a3b0 movq %rax, %r14 leaq 0x13a0654(%rip), %rsi # 0x1eebaf0 movq %rax, %rdi callq 0x6a230 movq 0x15d975d(%rip), %rsi # 0x2124c08 movq 0x15d9516(%rip), %rdx # 0x21249c8 movq %r14, %rdi callq 0x6a5d0 movl $0x10, %edi callq 0x6a3b0 movq %rax, %r14 leaq 0x13a0622(%rip), %rsi # 0x1eebaf0 movq %rax, %rdi callq 0x6a230 movq 0x15d972b(%rip), %rsi # 0x2124c08 movq 0x15d94e4(%rip), %rdx # 0x21249c8 movq %r14, %rdi callq 0x6a5d0 movl $0x10, %edi callq 0x6a3b0 movq %rax, %r14 leaq 0x13a05f0(%rip), %rsi # 0x1eebaf0 movq %rax, %rdi callq 0x6a230 movq 0x15d96f9(%rip), %rsi # 0x2124c08 movq 0x15d94b2(%rip), %rdx # 0x21249c8 movq %r14, %rdi callq 0x6a5d0 movl $0x10, %edi callq 0x6a3b0 movq %rax, %r14 leaq 0x13a05be(%rip), %rsi # 0x1eebaf0 movq %rax, %rdi callq 0x6a230 movq 0x15d96c7(%rip), %rsi # 0x2124c08 movq 0x15d9480(%rip), %rdx # 0x21249c8 movq %r14, %rdi callq 0x6a5d0 movq %rax, %rbx movq %r14, %rdi callq 0x6a8a0 jmp 0xb4b575 movq %rax, %rbx movq %r14, %rdi callq 0x6a8a0 jmp 0xb4b589 jmp 0xb4b5e8 jmp 0xb4b572 jmp 0xb4b5e8 jmp 0xb4b586 movq %rax, %rbx leaq 0xb0(%rsp), %rdi callq 0x6aab0 jmp 0xb4b5e0 jmp 0xb4b5e8 movq %rax, %rbx leaq 0xb0(%rsp), %rdi callq 0x6aab0 jmp 0xb4b5e0 jmp 0xb4b5e8 movq %rax, %rbx movq %r14, %rdi callq 0x6a8a0 jmp 0xb4b5bf movq %rax, %rbx movq %r14, %rdi callq 0x6a8a0 jmp 0xb4b5d3 jmp 0xb4b5e8 jmp 0xb4b5bc jmp 0xb4b5e8 jmp 0xb4b5d0 movq %rax, %rbx leaq 0x168(%rsp), %rdi callq 0x6aab0 jmp 0xb4b5e0 jmp 0xb4b5e8 movq %rax, %rbx leaq 0x168(%rsp), %rdi callq 0x6aab0 movq %rbx, %rdi callq 0x6a600 movq %rax, %rdi callq 0x8d6de8	/embree[P]embree/kernels/builders/primrefgen.cpp
embree::FastAllocator::Block::getUsedBytes(embree::FastAllocator::AllocationType, bool) const	size_t getUsedBytes(AllocationType atype, bool huge_pages = false) const { size_t bytes = 0; for (const Block* block = this; block; block = block->next) { if (!block->hasType(atype,huge_pages)) continue; bytes += block->getBlockUsedBytes(); } return bytes; }	xorl %eax, %eax cmpl $0x3, %esi je 0xb5eee6 movl 0x28(%rdi), %ecx cmpl $0x1, %ecx jne 0xb5eedd cmpl $0x1, %esi sete %r8b cmpb %dl, 0x2c(%rdi) sete %cl andb %r8b, %cl jmp 0xb5eee2 cmpl %esi, %ecx sete %cl testb %cl, %cl je 0xb5eef7 movq (%rdi), %rcx movq 0x10(%rdi), %r8 cmpq %r8, %rcx cmovbq %rcx, %r8 addq %r8, %rax movq 0x18(%rdi), %rdi testq %rdi, %rdi jne 0xb5eebe retq nop	/embree[P]embree/kernels/bvh/../common/alloc.h

Subsets and Splits

SQL Console for LLM4Binary/decompile-bench

Filters out entries with file names ending in .cpp, providing a basic subset of the dataset that excludes C++ files.