parallel_kernels.h File Reference

Go to the source code of this file.

Defines
#define	B_ID(index, i)   (index+(bStride*(i)))
#define	C_ID(index, i)   (index+(cStride*(i)))
Functions
template<typename T >
dxGlobal void	cudaComputeAdcfmBT (int4 *bodyIDs, typename vec4< T >::Type *j0, typename vec4< T >::Type *j1, typename vec4< T >::Type *j2, typename vec4< T >::Type *j3, typename vec4< T >::Type *ij0, typename vec4< T >::Type *ij1, typename vec4< T >::Type *ij2, typename vec4< T >::Type *ij3, T *adcfm, T *rhs, T sorParam, int numConstraints)
template<typename T >
dxGlobal void	cudaComputeInvMJTT (int4 *bodyIDs, typename vec4< T >::Type *j0, typename vec4< T >::Type *j1, typename vec4< T >::Type *j2, typename vec4< T >::Type *j3, typename vec4< T >::Type *ij0, typename vec4< T >::Type *ij1, typename vec4< T >::Type *ij2, typename vec4< T >::Type *ij3, T *iMass, int numConstraints, typename vec4< T >::Type *ii0, typename vec4< T >::Type *ii1, typename vec4< T >::Type *ii2)
template<typename T >
dxGlobal void	cudaIntegrateT (typename vec4< T >::Type *pos, typename vec4< T >::Type *lVel, typename vec4< T >::Type *aVel, float deltaTime, int numConstraints)
template<typename T >
dxGlobal void	cudaReduceIterativeCompactT (typename vec4< T >::Type *fc0_reduction, typename vec4< T >::Type *fc1_reduction, const int treePower)
template<typename T >
dxGlobal void	cudaReduceLoopedCompactT (typename vec4< T >::Type *fc0_reduction, typename vec4< T >::Type *fc1_reduction, const int size, const int step)
template<typename T , unsigned int blockSize>
dxGlobal void	cudaReduceSequentialT (typename vec4< T >::Type *fc0, typename vec4< T >::Type *fc1, typename vec4< T >::Type *fc0_reduction, typename vec4< T >::Type *fc1_reduction, int n)
template<typename T >
dxGlobal void	cudaReduceStridedT (typename vec4< T >::Type *fc0, typename vec4< T >::Type *fc1, const typename vec4< T >::Type *fc0_reduction, const typename vec4< T >::Type *fc1_reduction, const int reductionStride, const int bodySize, const int reductionSize)
template<typename T >
dxGlobal void	cudaSORLCPT (typename vec4< T >::Type *fc0_reduction, typename vec4< T >::Type *fc1_reduction, T *lambda, const int4 *bodyIDs, const int *fIDs, const typename vec4< T >::Type *j, const typename vec4< T >::Type *ij, const typename vec4< T >::Type *fc0, const typename vec4< T >::Type *fc1, const T *adcfm, const T *rhs, const T *hilo, const int offset, const int numConstraints, const int bStride, const int cStride)
template<typename T >
dxGlobal void	cudaZeroT (T *buffer, const int bufferSize)
template<typename T >
static dxDevice T	parallel_inf ()
template<>
dxDevice double	parallel_inf< double > ()
template<>
dxDevice float	parallel_inf< float > ()
template<typename T >
static dxDevice T	parallel_zero ()
template<>
dxDevice double	parallel_zero< double > ()
template<>
dxDevice double4	parallel_zero< double4 > ()
template<>
dxDevice float	parallel_zero< float > ()
template<>
dxDevice float4	parallel_zero< float4 > ()
Variables
static uint2	s_blockDim
static uint2	s_blockIdx
static uint2	s_threadIdx

---

Define Documentation

#define B_ID	(	index,
		i		)	(index+(bStride*(i)))

Definition at line 5 of file parallel_kernels.h.

#define C_ID	(	index,
		i		)	(index+(cStride*(i)))

Definition at line 4 of file parallel_kernels.h.

---

Function Documentation

template<typename T >

dxGlobal void cudaComputeAdcfmBT	(	int4 *	bodyIDs,
		typename vec4< T >::Type *	j0,
		typename vec4< T >::Type *	j1,
		typename vec4< T >::Type *	j2,
		typename vec4< T >::Type *	j3,
		typename vec4< T >::Type *	ij0,
		typename vec4< T >::Type *	ij1,
		typename vec4< T >::Type *	ij2,
		typename vec4< T >::Type *	ij3,
		T *	adcfm,
		T *	rhs,
		T	sorParam,
		int	numConstraints
	)			[inline]

Definition at line 399 of file parallel_kernels.h.

template<typename T >

dxGlobal void cudaComputeInvMJTT	(	int4 *	bodyIDs,
		typename vec4< T >::Type *	j0,
		typename vec4< T >::Type *	j1,
		typename vec4< T >::Type *	j2,
		typename vec4< T >::Type *	j3,
		typename vec4< T >::Type *	ij0,
		typename vec4< T >::Type *	ij1,
		typename vec4< T >::Type *	ij2,
		typename vec4< T >::Type *	ij3,
		T *	iMass,
		int	numConstraints,
		typename vec4< T >::Type *	ii0,
		typename vec4< T >::Type *	ii1,
		typename vec4< T >::Type *	ii2
	)			[inline]

Definition at line 321 of file parallel_kernels.h.

template<typename T >

dxGlobal void cudaIntegrateT	(	typename vec4< T >::Type *	pos,
		typename vec4< T >::Type *	lVel,
		typename vec4< T >::Type *	aVel,
		float	deltaTime,
		int	numConstraints
	)			[inline]

Definition at line 457 of file parallel_kernels.h.

template<typename T >

dxGlobal void cudaReduceIterativeCompactT	(	typename vec4< T >::Type *	fc0_reduction,
		typename vec4< T >::Type *	fc1_reduction,
		const int	treePower
	)			[inline]

Definition at line 157 of file parallel_kernels.h.

template<typename T >

dxGlobal void cudaReduceLoopedCompactT	(	typename vec4< T >::Type *	fc0_reduction,
		typename vec4< T >::Type *	fc1_reduction,
		const int	size,
		const int	step
	)			[inline]

Definition at line 180 of file parallel_kernels.h.

template<typename T , unsigned int blockSize>

dxGlobal void cudaReduceSequentialT	(	typename vec4< T >::Type *	fc0,
		typename vec4< T >::Type *	fc1,
		typename vec4< T >::Type *	fc0_reduction,
		typename vec4< T >::Type *	fc1_reduction,
		int	n
	)			[inline]

Definition at line 240 of file parallel_kernels.h.

template<typename T >

dxGlobal void cudaReduceStridedT	(	typename vec4< T >::Type *	fc0,
		typename vec4< T >::Type *	fc1,
		const typename vec4< T >::Type *	fc0_reduction,
		const typename vec4< T >::Type *	fc1_reduction,
		const int	reductionStride,
		const int	bodySize,
		const int	reductionSize
	)			[inline]

Definition at line 209 of file parallel_kernels.h.

template<typename T >

dxGlobal void cudaSORLCPT	(	typename vec4< T >::Type *	fc0_reduction,
		typename vec4< T >::Type *	fc1_reduction,
		T *	lambda,
		const int4 *	bodyIDs,
		const int *	fIDs,
		const typename vec4< T >::Type *	j,
		const typename vec4< T >::Type *	ij,
		const typename vec4< T >::Type *	fc0,
		const typename vec4< T >::Type *	fc1,
		const T *	adcfm,
		const T *	rhs,
		const T *	hilo,
		const int	offset,
		const int	numConstraints,
		const int	bStride,
		const int	cStride
	)			[inline]

Definition at line 38 of file parallel_kernels.h.

template<typename T >

dxGlobal void cudaZeroT	(	T *	buffer,
		const int	bufferSize
	)			[inline]

Definition at line 27 of file parallel_kernels.h.

template<typename T >

static dxDevice T parallel_inf

(

)

[inline, static]

Definition at line 9 of file parallel_kernels.h.

template<>

dxDevice double parallel_inf< double >

(

)

[inline]

template<>

dxDevice float parallel_inf< float >

(

)

[inline]

template<typename T >

static dxDevice T parallel_zero

(

)

[inline, static]

Definition at line 10 of file parallel_kernels.h.

template<>

dxDevice double parallel_zero< double >

(

)

[inline]

template<>

dxDevice double4 parallel_zero< double4 >

(

)

[inline]

template<>

dxDevice float parallel_zero< float >

(

)

[inline]

template<>

dxDevice float4 parallel_zero< float4 >

(

)

[inline]

---

Variable Documentation

uint2 s_blockDim [static]

Definition at line 7 of file parallel_kernels.h.

uint2 s_blockIdx [static]

Definition at line 7 of file parallel_kernels.h.

uint2 s_threadIdx [static]

Definition at line 7 of file parallel_kernels.h.