Macros
#define	LUD_HIGH_PRECISION

Functions
template<typename dataType >
DSPLIB_STATUS	DSPLIB_lud_init_ci (DSPLIB_kernelHandle handle, const DSPLIB_bufParams2D_t bufParamsA, const DSPLIB_bufParams2D_t bufParamsL, const DSPLIB_bufParams2D_t bufParamsU, const DSPLIB_bufParams2D_t bufParamsP, const DSPLIB_ludInitArgs *pKerInitArgs)
	This function is the initialization function for the C7x implementation of the kernel. The function declaration conforms to the declaration of DSPLIB_lud_init. More...

template DSPLIB_STATUS	DSPLIB_lud_init_ci< float > (DSPLIB_kernelHandle handle, const DSPLIB_bufParams2D_t bufParamsA, const DSPLIB_bufParams2D_t bufParamsL, const DSPLIB_bufParams2D_t bufParamsU, const DSPLIB_bufParams2D_t bufParamsP, const DSPLIB_ludInitArgs *pKerInitArgs)

template DSPLIB_STATUS	DSPLIB_lud_init_ci< double > (DSPLIB_kernelHandle handle, const DSPLIB_bufParams2D_t bufParamsA, const DSPLIB_bufParams2D_t bufParamsL, const DSPLIB_bufParams2D_t bufParamsU, const DSPLIB_bufParams2D_t bufParamsP, const DSPLIB_ludInitArgs *pKerInitArgs)

template<typename dataType >
__vpred	getPMask (uint32_t idx)

template<>
__vpred	getPMask< float > (uint32_t idx)

template<>
__vpred	getPMask< double > (uint32_t idx)

template<typename V , typename dataType >
V	getIdxVec ()

template<>
c7x::float_vec	getIdxVec< c7x::float_vec, float > ()

template<>
c7x::double_vec	getIdxVec< c7x::double_vec, double > ()

template<typename dataType , typename vec = typename c7x::make_full_vector<dataType>::type>
static void	DSPLIB_lud_maxMinIndex_exec_ci (dataType pCol, int32_t nRows, vec idx_0_to_eleCount, int32_t k, int32_t colStride, int32_t max, int32_t *min, __SE_TEMPLATE_v1 se0Params)

template void	DSPLIB_lud_maxMinIndex_exec_ci< float > (float pCol, int32_t nRows, typename c7x::make_full_vector< float >::type idx_0_to_eleCount, int32_t k, int32_t colStride, int32_t max, int32_t *min, __SE_TEMPLATE_v1 se0Params)

template void	DSPLIB_lud_maxMinIndex_exec_ci< double > (double pCol, int32_t nRows, typename c7x::make_full_vector< double >::type idx_0_to_eleCount, int32_t k, int32_t colStride, int32_t max, int32_t *min, __SE_TEMPLATE_v1 se0Params)

template<typename dataType >
static void	DSPLIB_lud_array_swap_exec_ci (dataType pArray1, dataType pArray2, int32_t nCols, __SA_TEMPLATE_v1 sa1Params, __SE_TEMPLATE_v1 se1Params)

template void	DSPLIB_lud_array_swap_exec_ci< float > (float pArray1, float pArray2, int32_t nCols, __SA_TEMPLATE_v1 sa1Params, __SE_TEMPLATE_v1 se1Params)

template void	DSPLIB_lud_array_swap_exec_ci< double > (double pArray1, double pArray2, int32_t nCols, __SA_TEMPLATE_v1 sa1Params, __SE_TEMPLATE_v1 se1Params)

template void	DSPLIB_lud_array_swap_exec_ci< unsigned short > (unsigned short pArray1, unsigned short pArray2, int32_t nCols, __SA_TEMPLATE_v1 sa1Params, __SE_TEMPLATE_v1 se1Params)

template<typename dataType >
static void	DSPLIB_lud_U_colDiv_exec_ci (dataType *pCol, int32_t nRows, int32_t colStride, __SA_TEMPLATE_v1 saColParams, __SE_TEMPLATE_v1 se0ColParams, __SE_TEMPLATE_v1 se1ColParams)

template void	DSPLIB_lud_U_colDiv_exec_ci< float > (float *pCol, int32_t nRows, int32_t colStride, __SA_TEMPLATE_v1 saColParams, __SE_TEMPLATE_v1 se0ColParams, __SE_TEMPLATE_v1 se1ColParams)

template void	DSPLIB_lud_U_colDiv_exec_ci< double > (double *pCol, int32_t nRows, int32_t colStride, __SA_TEMPLATE_v1 saColParams, __SE_TEMPLATE_v1 se0ColParams, __SE_TEMPLATE_v1 se1ColParams)

template<typename dataType >
static void	DSPLIB_lud_U_generate_exec_ci (dataType pLocalU, int32_t order, int32_t colStrideU, uint8_t pBlock)

template void	DSPLIB_lud_U_generate_exec_ci< float > (float pLocalU, int32_t order, int32_t colStrideU, uint8_t pBlock)

template void	DSPLIB_lud_U_generate_exec_ci< double > (double pLocalU, int32_t order, int32_t colStrideU, uint8_t pBlock)

template<typename dataType >
void	DSPLIB_lud_extract_exec_ci (dataType pU, dataType pL, int32_t order, int32_t colStride, uint8_t *pBlock)

template void	DSPLIB_lud_extract_exec_ci< float > (float pU, float pL, int32_t order, int32_t colStride, uint8_t *pBlock)

template void	DSPLIB_lud_extract_exec_ci< double > (double pU, double pL, int32_t order, int32_t colStride, uint8_t *pBlock)

template<typename dataType >
DSPLIB_STATUS	DSPLIB_lud_exec_ci (DSPLIB_kernelHandle handle, void restrict pA, void restrict pL, void restrict pU, void restrict pP)
	This function is the main execution function for the C7x implementation of the kernel. The function declaration conforms to the declaration of DSPLIB_lud_exec. More...

template DSPLIB_STATUS	DSPLIB_lud_exec_ci< float > (DSPLIB_kernelHandle handle, void restrict pA, void restrict pL, void restrict pU, void restrict pP)

template DSPLIB_STATUS	DSPLIB_lud_exec_ci< double > (DSPLIB_kernelHandle handle, void restrict pA, void restrict pL, void restrict pU, void restrict pP)

Variables
float16	idx_float = float16(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)

double8	idx_double = double8(0, 1, 2, 3, 4, 5, 6, 7)

Macro Definition Documentation

◆ LUD_HIGH_PRECISION

#define LUD_HIGH_PRECISION

Definition at line 51 of file DSPLIB_lud_ci.cpp.

Function Documentation

◆ DSPLIB_lud_init_ci()

template<typename dataType >

DSPLIB_STATUS DSPLIB_lud_init_ci	(	DSPLIB_kernelHandle	handle,
		const DSPLIB_bufParams2D_t *	bufParamsA,
		const DSPLIB_bufParams2D_t *	bufParamsL,
		const DSPLIB_bufParams2D_t *	bufParamsU,
		const DSPLIB_bufParams2D_t *	bufParamsP,
		const DSPLIB_ludInitArgs *	pKerInitArgs
	)

This function is the initialization function for the C7x implementation of the kernel. The function declaration conforms to the declaration of DSPLIB_lud_init.

This function determines the configuration for the streaming engine resources based on the function call parameters, and the configuration is saved in bufPBlock array. In the kernel call sequence, DSPLIB_lud_exec_ci would be called later independently by the application. When DSPLIB_lud_exec_ci runs, it merely retrieves the configuration from the bufPBlock and uses it to set up the hardware resources. This arrangement is so that DSPLIB_lud_exec_ci does not lose cycles to determine the hardware configuration.

Parameters

[in]	handle	: Active handle to the kernel
[in]	bufParamsA	: Pointer to the structure containing dimensional information of input buffer A
[out]	bufParamsL	: Pointer to the structure containing dimensional information of output buffer L
[out]	bufParamsU	: Pointer to the structure containing dimensional information of output buffer U
[out]	bufParamsP	: Pointer to the structure containing dimensional information of output buffer P
[in]	pKerInitArgs	: Pointer to the structure holding init parameters

Returns: Status value indicating success or failure. Refer to DSPLIB_STATUS.

Definition at line 59 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_init_ci< float >()

template DSPLIB_STATUS DSPLIB_lud_init_ci< float >	(	DSPLIB_kernelHandle	handle,
		const DSPLIB_bufParams2D_t *	bufParamsA,
		const DSPLIB_bufParams2D_t *	bufParamsL,
		const DSPLIB_bufParams2D_t *	bufParamsU,
		const DSPLIB_bufParams2D_t *	bufParamsP,
		const DSPLIB_ludInitArgs *	pKerInitArgs
	)

◆ DSPLIB_lud_init_ci< double >()

template DSPLIB_STATUS DSPLIB_lud_init_ci< double >	(	DSPLIB_kernelHandle	handle,
		const DSPLIB_bufParams2D_t *	bufParamsA,
		const DSPLIB_bufParams2D_t *	bufParamsL,
		const DSPLIB_bufParams2D_t *	bufParamsU,
		const DSPLIB_bufParams2D_t *	bufParamsP,
		const DSPLIB_ludInitArgs *	pKerInitArgs
	)

◆ getPMask()

template<typename dataType >

__vpred getPMask ( uint32_t idx )

inline

◆ getPMask< float >()

template<>

__vpred getPMask< float > ( uint32_t idx )

inline

Definition at line 324 of file DSPLIB_lud_ci.cpp.

◆ getPMask< double >()

template<>

__vpred getPMask< double > ( uint32_t idx )

inline

Definition at line 326 of file DSPLIB_lud_ci.cpp.

◆ getIdxVec()

template<typename V , typename dataType >

V getIdxVec ( )

inline

◆ getIdxVec< c7x::float_vec, float >()

template<>

c7x::float_vec getIdxVec< c7x::float_vec, float > ( )

inline

Definition at line 333 of file DSPLIB_lud_ci.cpp.

◆ getIdxVec< c7x::double_vec, double >()

template<>

c7x::double_vec getIdxVec< c7x::double_vec, double > ( )

inline

Definition at line 335 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_maxMinIndex_exec_ci()

template<typename dataType , typename vec = typename c7x::make_full_vector<dataType>::type>

static void DSPLIB_lud_maxMinIndex_exec_ci	(	dataType *	pCol,
		int32_t	nRows,
		vec	idx_0_to_eleCount,
		int32_t	k,
		int32_t	colStride,
		int32_t *	max,
		int32_t *	min,
		__SE_TEMPLATE_v1	se0Params
	)

inlinestatic

Definition at line 338 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_maxMinIndex_exec_ci< float >()

template void DSPLIB_lud_maxMinIndex_exec_ci< float >	(	float *	pCol,
		int32_t	nRows,
		typename c7x::make_full_vector< float >::type	idx_0_to_eleCount,
		int32_t	k,
		int32_t	colStride,
		int32_t *	max,
		int32_t *	min,
		__SE_TEMPLATE_v1	se0Params
	)

◆ DSPLIB_lud_maxMinIndex_exec_ci< double >()

template void DSPLIB_lud_maxMinIndex_exec_ci< double >	(	double *	pCol,
		int32_t	nRows,
		typename c7x::make_full_vector< double >::type	idx_0_to_eleCount,
		int32_t	k,
		int32_t	colStride,
		int32_t *	max,
		int32_t *	min,
		__SE_TEMPLATE_v1	se0Params
	)

◆ DSPLIB_lud_array_swap_exec_ci()

template<typename dataType >

static void DSPLIB_lud_array_swap_exec_ci	(	dataType *	pArray1,
		dataType *	pArray2,
		int32_t	nCols,
		__SA_TEMPLATE_v1	sa1Params,
		__SE_TEMPLATE_v1	se1Params
	)

inlinestatic

Definition at line 436 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_array_swap_exec_ci< float >()

template void DSPLIB_lud_array_swap_exec_ci< float >	(	float *	pArray1,
		float *	pArray2,
		int32_t	nCols,
		__SA_TEMPLATE_v1	sa1Params,
		__SE_TEMPLATE_v1	se1Params
	)

◆ DSPLIB_lud_array_swap_exec_ci< double >()

template void DSPLIB_lud_array_swap_exec_ci< double >	(	double *	pArray1,
		double *	pArray2,
		int32_t	nCols,
		__SA_TEMPLATE_v1	sa1Params,
		__SE_TEMPLATE_v1	se1Params
	)

◆ DSPLIB_lud_array_swap_exec_ci< unsigned short >()

template void DSPLIB_lud_array_swap_exec_ci< unsigned short >	(	unsigned short *	pArray1,
		unsigned short *	pArray2,
		int32_t	nCols,
		__SA_TEMPLATE_v1	sa1Params,
		__SE_TEMPLATE_v1	se1Params
	)

◆ DSPLIB_lud_U_colDiv_exec_ci()

template<typename dataType >

static void DSPLIB_lud_U_colDiv_exec_ci	(	dataType *	pCol,
		int32_t	nRows,
		int32_t	colStride,
		__SA_TEMPLATE_v1	saColParams,
		__SE_TEMPLATE_v1	se0ColParams,
		__SE_TEMPLATE_v1	se1ColParams
	)

inlinestatic

Definition at line 489 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_U_colDiv_exec_ci< float >()

template void DSPLIB_lud_U_colDiv_exec_ci< float >	(	float *	pCol,
		int32_t	nRows,
		int32_t	colStride,
		__SA_TEMPLATE_v1	saColParams,
		__SE_TEMPLATE_v1	se0ColParams,
		__SE_TEMPLATE_v1	se1ColParams
	)

◆ DSPLIB_lud_U_colDiv_exec_ci< double >()

template void DSPLIB_lud_U_colDiv_exec_ci< double >	(	double *	pCol,
		int32_t	nRows,
		int32_t	colStride,
		__SA_TEMPLATE_v1	saColParams,
		__SE_TEMPLATE_v1	se0ColParams,
		__SE_TEMPLATE_v1	se1ColParams
	)

◆ DSPLIB_lud_U_generate_exec_ci()

template<typename dataType >

static void DSPLIB_lud_U_generate_exec_ci	(	dataType *	pLocalU,
		int32_t	order,
		int32_t	colStrideU,
		uint8_t *	pBlock
	)

inlinestatic

Definition at line 573 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_U_generate_exec_ci< float >()

template void DSPLIB_lud_U_generate_exec_ci< float >	(	float *	pLocalU,
		int32_t	order,
		int32_t	colStrideU,
		uint8_t *	pBlock
	)

◆ DSPLIB_lud_U_generate_exec_ci< double >()

template void DSPLIB_lud_U_generate_exec_ci< double >	(	double *	pLocalU,
		int32_t	order,
		int32_t	colStrideU,
		uint8_t *	pBlock
	)

◆ DSPLIB_lud_extract_exec_ci()

template<typename dataType >

void DSPLIB_lud_extract_exec_ci	(	dataType *	pU,
		dataType *	pL,
		int32_t	order,
		int32_t	colStride,
		uint8_t *	pBlock
	)

Definition at line 892 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_extract_exec_ci< float >()

template void DSPLIB_lud_extract_exec_ci< float >	(	float *	pU,
		float *	pL,
		int32_t	order,
		int32_t	colStride,
		uint8_t *	pBlock
	)

◆ DSPLIB_lud_extract_exec_ci< double >()

template void DSPLIB_lud_extract_exec_ci< double >	(	double *	pU,
		double *	pL,
		int32_t	order,
		int32_t	colStride,
		uint8_t *	pBlock
	)

◆ DSPLIB_lud_exec_ci()

template<typename dataType >

DSPLIB_STATUS DSPLIB_lud_exec_ci	(	DSPLIB_kernelHandle	handle,
		void *restrict	pA,
		void *restrict	pL,
		void *restrict	pU,
		void *restrict	pP
	)

This function is the main execution function for the C7x implementation of the kernel. The function declaration conforms to the declaration of DSPLIB_lud_exec.

The function uses streaming engine to load data. The result is then stored at separate location using a streaming address generator.

Parameters

[in]	handle	: Active handle to the kernel
[in]	pA	: Pointer to matrix A
[out]	pL	: Pointer to lower triangular matrix L
[out]	pU	: Pointer to upper triangular matrix U
[out]	pP	: Pointer to permutation matrix P

Returns: Status value indicating success or failure. Refer to DSPLIB_STATUS.

Performance Considerations:

For best performance,

the input and output data buffers are expected to be in L2 memory
the buffer pointers are assumed to be 64-byte aligned

Definition at line 976 of file DSPLIB_lud_ci.cpp.

◆ DSPLIB_lud_exec_ci< float >()

template DSPLIB_STATUS DSPLIB_lud_exec_ci< float >	(	DSPLIB_kernelHandle	handle,
		void *restrict	pA,
		void *restrict	pL,
		void *restrict	pU,
		void *restrict	pP
	)

◆ DSPLIB_lud_exec_ci< double >()

template DSPLIB_STATUS DSPLIB_lud_exec_ci< double >	(	DSPLIB_kernelHandle	handle,
		void *restrict	pA,
		void *restrict	pL,
		void *restrict	pU,
		void *restrict	pP
	)

Variable Documentation

◆ idx_float

float16 idx_float = float16(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)

Definition at line 328 of file DSPLIB_lud_ci.cpp.

◆ idx_double

double8 idx_double = double8(0, 1, 2, 3, 4, 5, 6, 7)

Definition at line 329 of file DSPLIB_lud_ci.cpp.

Macros

Functions

Variables

Macro Definition Documentation

◆ LUD_HIGH_PRECISION

Function Documentation

◆ DSPLIB_lud_init_ci()

◆ DSPLIB_lud_init_ci< float >()

◆ DSPLIB_lud_init_ci< double >()

◆ getPMask()

◆ getPMask< float >()

◆ getPMask< double >()

◆ getIdxVec()

◆ getIdxVec< c7x::float_vec, float >()

◆ getIdxVec< c7x::double_vec, double >()

◆ DSPLIB_lud_maxMinIndex_exec_ci()

◆ DSPLIB_lud_maxMinIndex_exec_ci< float >()

◆ DSPLIB_lud_maxMinIndex_exec_ci< double >()

◆ DSPLIB_lud_array_swap_exec_ci()

◆ DSPLIB_lud_array_swap_exec_ci< float >()

◆ DSPLIB_lud_array_swap_exec_ci< double >()

◆ DSPLIB_lud_array_swap_exec_ci< unsigned short >()

◆ DSPLIB_lud_U_colDiv_exec_ci()

◆ DSPLIB_lud_U_colDiv_exec_ci< float >()

◆ DSPLIB_lud_U_colDiv_exec_ci< double >()

◆ DSPLIB_lud_U_generate_exec_ci()

◆ DSPLIB_lud_U_generate_exec_ci< float >()

◆ DSPLIB_lud_U_generate_exec_ci< double >()

◆ DSPLIB_lud_extract_exec_ci()

◆ DSPLIB_lud_extract_exec_ci< float >()

◆ DSPLIB_lud_extract_exec_ci< double >()

◆ DSPLIB_lud_exec_ci()

◆ DSPLIB_lud_exec_ci< float >()

◆ DSPLIB_lud_exec_ci< double >()

Variable Documentation

◆ idx_float

◆ idx_double