Functions
void	init ()
	Create the BLAS context. More...

void	destroy ()
	Destroy the BLAS context. More...

long long	BatchInvertMatrix (void Ainv, void A, const int n, const uint64_t batch, QudaPrecision precision, QudaFieldLocation location)
	Batch inversion the matrix field using an LU decomposition method. More...

long long	stridedBatchGEMM (void A, void B, void *C, QudaBLASParam blas_param, QudaFieldLocation location)
	Strided Batch GEMM. This function performs N GEMM type operations in a strided batched fashion. If the user passes. More...

template<typename EigenMatrix , typename Float >
void	invertEigen (std::complex< Float > A_eig, std::complex< Float > Ainv_eig, int n, uint64_t batch)

template<typename EigenMat , typename T >
void	fillArray (EigenMat &EigenArr, T *arr, int rows, int cols, int ld, int offset, bool fill_eigen)

template<typename EigenMat , typename T >
void	GEMM (void A_h, void B_h, void *C_h, T alpha, T beta, int max_stride, QudaBLASParam &blas_param)

Detailed Description

The generic namespace is where we can deploy any target-independent blas/lapack operations that are not supported on the native target. To this end, we use Eigen on the host.

Function Documentation

◆ BatchInvertMatrix()

long long quda::blas_lapack::generic::BatchInvertMatrix	(	void *	Ainv,
		void *	A,
		const int	n,
		const uint64_t	batch,
		QudaPrecision	precision,
		QudaFieldLocation	location
	)

Batch inversion the matrix field using an LU decomposition method.

Parameters

[out]	Ainv	Matrix field containing the inverse matrices
[in]	A	Matrix field containing the input matrices
[in]	n	Dimension each matrix
[in]	batch	Problem batch size
[in]	precision	Precision of the input/output data
[in]	Location	of the input/output data

Returns: Number of flops done in this computation

Definition at line 52 of file blas_lapack_eigen.cpp.

◆ destroy()

void quda::blas_lapack::generic::destroy ( )

Destroy the BLAS context.

Definition at line 21 of file blas_lapack_eigen.cpp.

◆ fillArray()

template<typename EigenMat , typename T >

void quda::blas_lapack::generic::fillArray	(	EigenMat &	EigenArr,
		T *	arr,
		int	rows,
		int	cols,
		int	ld,
		int	offset,
		bool	fill_eigen
	)

Definition at line 115 of file blas_lapack_eigen.cpp.

◆ GEMM()

template<typename EigenMat , typename T >

void quda::blas_lapack::generic::GEMM	(	void *	A_h,
		void *	B_h,
		void *	C_h,
		T	alpha,
		T	beta,
		int	max_stride,
		QudaBLASParam &	blas_param
	)

Definition at line 131 of file blas_lapack_eigen.cpp.

◆ init()

void quda::blas_lapack::generic::init ( )

Create the BLAS context.

Definition at line 19 of file blas_lapack_eigen.cpp.

◆ invertEigen()

template<typename EigenMatrix , typename Float >

void quda::blas_lapack::generic::invertEigen	(	std::complex< Float > *	A_eig,
		std::complex< Float > *	Ainv_eig,
		int	n,
		uint64_t	batch
	)

Definition at line 26 of file blas_lapack_eigen.cpp.

◆ stridedBatchGEMM()

long long quda::blas_lapack::generic::stridedBatchGEMM	(	void *	A,
		void *	B,
		void *	C,
		QudaBLASParam	blas_param,
		QudaFieldLocation	location
	)

Strided Batch GEMM. This function performs N GEMM type operations in a strided batched fashion. If the user passes.

stride<A,B,C> = -1

it deduces the strides for the A, B, and C arrays from the matrix dimensions, leading dims, etc, and will behave identically to the batched GEMM. If any of the stride<A,B,C> values passed in the parameter structure are greater than or equal to 0, the routine accepts the user's values instead.

Example: If the user passes

a_stride = 0

the routine will use only the first matrix in the A array and compute

C_{n} <- a * A_{0} * B_{n} + b * C_{n}

where n is the batch index.