org.bytedeco.bullet.Bullet3OpenCL

Class b3GpuParallelLinearBvh

java.lang.Object

org.bytedeco.javacpp.Pointer

org.bytedeco.bullet.Bullet3OpenCL.b3GpuParallelLinearBvh

All Implemented Interfaces:

AutoCloseable

@NoOffset
 @Properties(inherit=Bullet3OpenCL.class)
public class b3GpuParallelLinearBvh
extends Pointer

\brief GPU Parallel Linearized Bounding Volume Heirarchy(LBVH) that is reconstructed every frame \remarks See presentation in docs/b3GpuParallelLinearBvh.pdf for algorithm details. \par Related papers: \n "Fast BVH Construction on GPUs" [Lauterbach et al. 2009] \n "Maximizing Parallelism in the Construction of BVHs, Octrees, and k-d trees" [Karras 2012] \n \par The basic algorithm for building the BVH as presented in [Lauterbach et al. 2009] consists of 4 stages: - [fully parallel] Assign morton codes for each AABB using its center (after quantizing the AABB centers into a virtual grid) - [fully parallel] Sort morton codes - [somewhat parallel] Build binary radix tree (assign parent/child pointers for internal nodes of the BVH) - [somewhat parallel] Set internal node AABBs \par [Karras 2012] improves on the algorithm by introducing fully parallel methods for the last 2 stages. The BVH implementation here shares many concepts with [Karras 2012], but a different method is used for constructing the tree. Instead of searching for the child nodes of each internal node, we search for the parent node of each node. Additionally, a non-atomic traversal that starts from the leaf nodes and moves towards the root node is used to set the AABBs.

Nested Class Summary

Nested classes/interfaces inherited from class org.bytedeco.javacpp.Pointer

Pointer.CustomDeallocator, Pointer.Deallocator, Pointer.NativeDeallocator, Pointer.ReferenceCounter

Field Summary

Fields inherited from class org.bytedeco.javacpp.Pointer

address, capacity, limit, position

Constructor Summary

Constructors

Constructor and Description
b3GpuParallelLinearBvh(Pointer p) Pointer cast constructor.
b3GpuParallelLinearBvh(Pointer context, Pointer device, Pointer queue)

Method Summary

Modifier and Type	Method and Description
void	build(b3SapAabbOCLArray worldSpaceAabbs, b3IntOCLArray smallAabbIndices, b3IntOCLArray largeAabbIndices) Must be called before any other function
void	calculateOverlappingPairs(b3Int4OCLArray out_overlappingPairs) calculateOverlappingPairs() uses the worldSpaceAabbs parameter of b3GpuParallelLinearBvh::build() as the query AABBs.
void	testRaysAgainstBvhAabbs(b3RayInfoOCLArray rays, b3IntOCLArray out_numRayRigidPairs, b3Int2OCLArray out_rayRigidPairs)

Methods inherited from class org.bytedeco.javacpp.Pointer

address, asBuffer, asByteBuffer, availablePhysicalBytes, calloc, capacity, capacity, close, deallocate, deallocate, deallocateReferences, deallocator, deallocator, equals, fill, formatBytes, free, getDirectBufferAddress, getPointer, getPointer, getPointer, getPointer, hashCode, interruptDeallocatorThread, isNull, isNull, limit, limit, malloc, maxBytes, maxPhysicalBytes, memchr, memcmp, memcpy, memmove, memset, offsetAddress, offsetof, offsetof, parseBytes, physicalBytes, physicalBytesInaccurate, position, position, put, realloc, referenceCount, releaseReference, retainReference, setNull, sizeof, sizeof, toString, totalBytes, totalCount, totalPhysicalBytes, withDeallocator, zero

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Constructor Detail

b3GpuParallelLinearBvh

public b3GpuParallelLinearBvh(Pointer p)

Pointer cast constructor. Invokes Pointer(Pointer).

b3GpuParallelLinearBvh

public b3GpuParallelLinearBvh(@Cast(value="cl_context")
                              Pointer context,
                              @Cast(value="cl_device_id")
                              Pointer device,
                              @Cast(value="cl_command_queue")
                              Pointer queue)

Method Detail

build

public void build(@Const @ByRef
                  b3SapAabbOCLArray worldSpaceAabbs,
                  @Const @ByRef
                  b3IntOCLArray smallAabbIndices,
                  @Const @ByRef
                  b3IntOCLArray largeAabbIndices)

Must be called before any other function

calculateOverlappingPairs

public void calculateOverlappingPairs(@ByRef
                                      b3Int4OCLArray out_overlappingPairs)

calculateOverlappingPairs() uses the worldSpaceAabbs parameter of b3GpuParallelLinearBvh::build() as the query AABBs.

Parameters:

out_overlappingPairs - The size() of this array is used to determine the max number of pairs. If the number of overlapping pairs is < out_overlappingPairs.size(), out_overlappingPairs is resized.

testRaysAgainstBvhAabbs

public void testRaysAgainstBvhAabbs(@Const @ByRef
                                    b3RayInfoOCLArray rays,
                                    @ByRef
                                    b3IntOCLArray out_numRayRigidPairs,
                                    @ByRef
                                    b3Int2OCLArray out_rayRigidPairs)

Parameters:

out_numRigidRayPairs - Array of length 1; contains the number of detected ray-rigid AABB intersections; this value may be greater than out_rayRigidPairs.size() if out_rayRigidPairs is not large enough.

out_rayRigidPairs - Contains an array of rays intersecting rigid AABBs; x == ray index, y == rigid body index. If the size of this array is insufficient to hold all ray-rigid AABB intersections, additional intersections are discarded.