2017-08-01 20:30:58 +08:00
|
|
|
#ifndef B3_GPU_SAP_BROADPHASE_H
|
|
|
|
#define B3_GPU_SAP_BROADPHASE_H
|
|
|
|
|
|
|
|
#include "Bullet3OpenCL/ParallelPrimitives/b3OpenCLArray.h"
|
2019-01-03 21:26:51 +08:00
|
|
|
#include "Bullet3OpenCL/ParallelPrimitives/b3FillCL.h" //b3Int2
|
2017-08-01 20:30:58 +08:00
|
|
|
class b3Vector3;
|
|
|
|
#include "Bullet3OpenCL/ParallelPrimitives/b3RadixSort32CL.h"
|
|
|
|
|
|
|
|
#include "b3SapAabb.h"
|
|
|
|
#include "Bullet3Common/shared/b3Int2.h"
|
|
|
|
|
|
|
|
#include "b3GpuBroadphaseInterface.h"
|
|
|
|
|
|
|
|
class b3GpuSapBroadphase : public b3GpuBroadphaseInterface
|
|
|
|
{
|
2019-01-03 21:26:51 +08:00
|
|
|
cl_context m_context;
|
|
|
|
cl_device_id m_device;
|
|
|
|
cl_command_queue m_queue;
|
|
|
|
cl_kernel m_flipFloatKernel;
|
|
|
|
cl_kernel m_scatterKernel;
|
|
|
|
cl_kernel m_copyAabbsKernel;
|
|
|
|
cl_kernel m_sapKernel;
|
|
|
|
cl_kernel m_sap2Kernel;
|
|
|
|
cl_kernel m_prepareSumVarianceKernel;
|
2017-08-01 20:30:58 +08:00
|
|
|
|
|
|
|
class b3RadixSort32CL* m_sorter;
|
|
|
|
|
|
|
|
///test for 3d SAP
|
2019-01-03 21:26:51 +08:00
|
|
|
b3AlignedObjectArray<b3SortData> m_sortedAxisCPU[3][2];
|
|
|
|
b3AlignedObjectArray<b3UnsignedInt2> m_objectMinMaxIndexCPU[3][2];
|
|
|
|
b3OpenCLArray<b3UnsignedInt2> m_objectMinMaxIndexGPUaxis0;
|
|
|
|
b3OpenCLArray<b3UnsignedInt2> m_objectMinMaxIndexGPUaxis1;
|
|
|
|
b3OpenCLArray<b3UnsignedInt2> m_objectMinMaxIndexGPUaxis2;
|
|
|
|
b3OpenCLArray<b3UnsignedInt2> m_objectMinMaxIndexGPUaxis0prev;
|
|
|
|
b3OpenCLArray<b3UnsignedInt2> m_objectMinMaxIndexGPUaxis1prev;
|
|
|
|
b3OpenCLArray<b3UnsignedInt2> m_objectMinMaxIndexGPUaxis2prev;
|
|
|
|
|
|
|
|
b3OpenCLArray<b3SortData> m_sortedAxisGPU0;
|
|
|
|
b3OpenCLArray<b3SortData> m_sortedAxisGPU1;
|
|
|
|
b3OpenCLArray<b3SortData> m_sortedAxisGPU2;
|
|
|
|
b3OpenCLArray<b3SortData> m_sortedAxisGPU0prev;
|
|
|
|
b3OpenCLArray<b3SortData> m_sortedAxisGPU1prev;
|
|
|
|
b3OpenCLArray<b3SortData> m_sortedAxisGPU2prev;
|
|
|
|
|
|
|
|
b3OpenCLArray<b3Int4> m_addedHostPairsGPU;
|
|
|
|
b3OpenCLArray<b3Int4> m_removedHostPairsGPU;
|
|
|
|
b3OpenCLArray<int> m_addedCountGPU;
|
|
|
|
b3OpenCLArray<int> m_removedCountGPU;
|
|
|
|
|
|
|
|
int m_currentBuffer;
|
2017-08-01 20:30:58 +08:00
|
|
|
|
|
|
|
public:
|
|
|
|
b3OpenCLArray<int> m_pairCount;
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
b3OpenCLArray<b3SapAabb> m_allAabbsGPU;
|
|
|
|
b3AlignedObjectArray<b3SapAabb> m_allAabbsCPU;
|
2017-08-01 20:30:58 +08:00
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
virtual b3OpenCLArray<b3SapAabb>& getAllAabbsGPU()
|
2017-08-01 20:30:58 +08:00
|
|
|
{
|
|
|
|
return m_allAabbsGPU;
|
|
|
|
}
|
2019-01-03 21:26:51 +08:00
|
|
|
virtual b3AlignedObjectArray<b3SapAabb>& getAllAabbsCPU()
|
2017-08-01 20:30:58 +08:00
|
|
|
{
|
|
|
|
return m_allAabbsCPU;
|
|
|
|
}
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
b3OpenCLArray<b3Vector3> m_sum;
|
|
|
|
b3OpenCLArray<b3Vector3> m_sum2;
|
|
|
|
b3OpenCLArray<b3Vector3> m_dst;
|
2017-08-01 20:30:58 +08:00
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
b3OpenCLArray<int> m_smallAabbsMappingGPU;
|
2017-08-01 20:30:58 +08:00
|
|
|
b3AlignedObjectArray<int> m_smallAabbsMappingCPU;
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
b3OpenCLArray<int> m_largeAabbsMappingGPU;
|
2017-08-01 20:30:58 +08:00
|
|
|
b3AlignedObjectArray<int> m_largeAabbsMappingCPU;
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
b3OpenCLArray<b3Int4> m_overlappingPairs;
|
2017-08-01 20:30:58 +08:00
|
|
|
|
|
|
|
//temporary gpu work memory
|
2019-01-03 21:26:51 +08:00
|
|
|
b3OpenCLArray<b3SortData> m_gpuSmallSortData;
|
|
|
|
b3OpenCLArray<b3SapAabb> m_gpuSmallSortedAabbs;
|
2017-08-01 20:30:58 +08:00
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
class b3PrefixScanFloat4CL* m_prefixScanFloat4;
|
2017-08-01 20:30:58 +08:00
|
|
|
|
|
|
|
enum b3GpuSapKernelType
|
|
|
|
{
|
2019-01-03 21:26:51 +08:00
|
|
|
B3_GPU_SAP_KERNEL_BRUTE_FORCE_CPU = 1,
|
2017-08-01 20:30:58 +08:00
|
|
|
B3_GPU_SAP_KERNEL_BRUTE_FORCE_GPU,
|
|
|
|
B3_GPU_SAP_KERNEL_ORIGINAL,
|
|
|
|
B3_GPU_SAP_KERNEL_BARRIER,
|
|
|
|
B3_GPU_SAP_KERNEL_LOCAL_SHARED_MEMORY
|
|
|
|
};
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
b3GpuSapBroadphase(cl_context ctx, cl_device_id device, cl_command_queue q, b3GpuSapKernelType kernelType = B3_GPU_SAP_KERNEL_LOCAL_SHARED_MEMORY);
|
2017-08-01 20:30:58 +08:00
|
|
|
virtual ~b3GpuSapBroadphase();
|
2019-01-03 21:26:51 +08:00
|
|
|
|
|
|
|
static b3GpuBroadphaseInterface* CreateFuncBruteForceCpu(cl_context ctx, cl_device_id device, cl_command_queue q)
|
2017-08-01 20:30:58 +08:00
|
|
|
{
|
2019-01-03 21:26:51 +08:00
|
|
|
return new b3GpuSapBroadphase(ctx, device, q, B3_GPU_SAP_KERNEL_BRUTE_FORCE_CPU);
|
2017-08-01 20:30:58 +08:00
|
|
|
}
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
static b3GpuBroadphaseInterface* CreateFuncBruteForceGpu(cl_context ctx, cl_device_id device, cl_command_queue q)
|
2017-08-01 20:30:58 +08:00
|
|
|
{
|
2019-01-03 21:26:51 +08:00
|
|
|
return new b3GpuSapBroadphase(ctx, device, q, B3_GPU_SAP_KERNEL_BRUTE_FORCE_GPU);
|
2017-08-01 20:30:58 +08:00
|
|
|
}
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
static b3GpuBroadphaseInterface* CreateFuncOriginal(cl_context ctx, cl_device_id device, cl_command_queue q)
|
2017-08-01 20:30:58 +08:00
|
|
|
{
|
2019-01-03 21:26:51 +08:00
|
|
|
return new b3GpuSapBroadphase(ctx, device, q, B3_GPU_SAP_KERNEL_ORIGINAL);
|
2017-08-01 20:30:58 +08:00
|
|
|
}
|
2019-01-03 21:26:51 +08:00
|
|
|
static b3GpuBroadphaseInterface* CreateFuncBarrier(cl_context ctx, cl_device_id device, cl_command_queue q)
|
2017-08-01 20:30:58 +08:00
|
|
|
{
|
2019-01-03 21:26:51 +08:00
|
|
|
return new b3GpuSapBroadphase(ctx, device, q, B3_GPU_SAP_KERNEL_BARRIER);
|
2017-08-01 20:30:58 +08:00
|
|
|
}
|
2019-01-03 21:26:51 +08:00
|
|
|
static b3GpuBroadphaseInterface* CreateFuncLocalMemory(cl_context ctx, cl_device_id device, cl_command_queue q)
|
2017-08-01 20:30:58 +08:00
|
|
|
{
|
2019-01-03 21:26:51 +08:00
|
|
|
return new b3GpuSapBroadphase(ctx, device, q, B3_GPU_SAP_KERNEL_LOCAL_SHARED_MEMORY);
|
2017-08-01 20:30:58 +08:00
|
|
|
}
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
virtual void calculateOverlappingPairs(int maxPairs);
|
|
|
|
virtual void calculateOverlappingPairsHost(int maxPairs);
|
|
|
|
|
|
|
|
void reset();
|
2017-08-01 20:30:58 +08:00
|
|
|
|
|
|
|
void init3dSap();
|
|
|
|
virtual void calculateOverlappingPairsHostIncremental3Sap();
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
virtual void createProxy(const b3Vector3& aabbMin, const b3Vector3& aabbMax, int userPtr, int collisionFilterGroup, int collisionFilterMask);
|
|
|
|
virtual void createLargeProxy(const b3Vector3& aabbMin, const b3Vector3& aabbMax, int userPtr, int collisionFilterGroup, int collisionFilterMask);
|
2017-08-01 20:30:58 +08:00
|
|
|
|
|
|
|
//call writeAabbsToGpu after done making all changes (createProxy etc)
|
|
|
|
virtual void writeAabbsToGpu();
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
virtual cl_mem getAabbBufferWS();
|
|
|
|
virtual int getNumOverlap();
|
|
|
|
virtual cl_mem getOverlappingPairBuffer();
|
|
|
|
|
2017-08-01 20:30:58 +08:00
|
|
|
virtual b3OpenCLArray<b3Int4>& getOverlappingPairsGPU();
|
|
|
|
virtual b3OpenCLArray<int>& getSmallAabbIndicesGPU();
|
|
|
|
virtual b3OpenCLArray<int>& getLargeAabbIndicesGPU();
|
|
|
|
};
|
|
|
|
|
2019-01-03 21:26:51 +08:00
|
|
|
#endif //B3_GPU_SAP_BROADPHASE_H
|