lama::CUDABLAS2 Class Reference

#include <CUDABLAS2.hpp>

Public Member Functions
template<>
void	gemv (const enum CBLAS_ORDER order, const enum CBLAS_TRANSPOSE trans, const IndexType m, const IndexType n, const float alpha, const float *const A, const IndexType lda, const float *const x, const IndexType incx, const float beta, float *const y, const IndexType incy, SyncToken *syncToken)
	gemv
template<>
void	gemv (const enum CBLAS_ORDER order, const enum CBLAS_TRANSPOSE trans, const IndexType m, const IndexType n, const double alpha, const double *const A, const IndexType lda, const double *const x, const IndexType incx, const double beta, double *const y, const IndexType incy, SyncToken *syncToken)
template<>
void	symv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const float alpha, const float *A, const IndexType lda, const float *x, const IndexType incX, const float beta, float *y, const IndexType incY, SyncToken *syncToken)
	symv
template<>
void	symv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const double alpha, const double *A, const IndexType lda, const double *x, const IndexType incX, const double beta, double *y, const IndexType incY, SyncToken *syncToken)
template<>
void	trmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const float *A, const IndexType lda, float *x, const IndexType incX, SyncToken *syncToken)
	trmv
template<>
void	trmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const double *A, const IndexType lda, double *x, const IndexType incX, SyncToken *syncToken)
template<>
void	trsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const float *A, const IndexType lda, float *x, const IndexType incX, SyncToken *syncToken)
	trsv
template<>
void	trsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const double *A, const IndexType lda, double *x, const IndexType incX, SyncToken *syncToken)
template<>
void	gbmv (const enum CBLAS_ORDER order, const enum CBLAS_TRANSPOSE trans, const IndexType m, const IndexType n, IndexType kl, IndexType ku, const float alpha, const float *A, const IndexType lda, const float *x, const IndexType incX, const float beta, float *y, const IndexType incY, SyncToken *syncToken)
	gbmv
template<>
void	gbmv (const enum CBLAS_ORDER order, const enum CBLAS_TRANSPOSE trans, const IndexType m, const IndexType n, IndexType kl, IndexType ku, const double alpha, const double *A, const IndexType lda, const double *x, const IndexType incX, const double beta, double *y, const IndexType incY, SyncToken *syncToken)
template<>
void	sbmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const IndexType k, const float alpha, const float *A, const IndexType lda, const float *x, const IndexType incX, const float beta, float *y, const IndexType incY, SyncToken *syncToken)
	sbmv
template<>
void	sbmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const IndexType k, const double alpha, const double *A, const IndexType lda, const double *x, const IndexType incX, const double beta, double *y, const IndexType incY, SyncToken *syncToken)
template<>
void	tbmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const IndexType k, const float *A, const IndexType lda, float *x, const IndexType incX, SyncToken *syncToken)
	tbmv
template<>
void	tbmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const IndexType k, const double *A, const IndexType lda, double *x, const IndexType incX, SyncToken *syncToken)
template<>
void	tbsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const IndexType k, const float *A, const IndexType lda, float *x, const IndexType incX, SyncToken *syncToken)
	tbsv
template<>
void	tbsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const IndexType k, const double *A, const IndexType lda, double *x, const IndexType incX, SyncToken *syncToken)
template<>
void	ger (const enum CBLAS_ORDER order, const IndexType m, const IndexType n, const float alpha, const float *x, const IndexType incX, const float *y, const IndexType incY, float *A, const IndexType lda, SyncToken *syncToken)
	ger
template<>
void	ger (const enum CBLAS_ORDER order, const IndexType m, const IndexType n, const double alpha, const double *x, const IndexType incX, const double *y, const IndexType incY, double *A, const IndexType lda, SyncToken *syncToken)
template<>
void	syr (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const float alpha, const float *x, const IndexType incX, float *A, const IndexType lda, SyncToken *syncToken)
	syr
template<>
void	syr (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const double alpha, const double *x, const IndexType incX, double *A, const IndexType lda, SyncToken *syncToken)
template<>
void	syr2 (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const float alpha, const float *x, const IndexType incX, const float *y, const IndexType incY, float *A, const IndexType lda, SyncToken *syncToken)
	syr2
template<>
void	syr2 (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const double alpha, const double *x, const IndexType incX, const double *y, const IndexType incY, double *A, const IndexType lda, SyncToken *syncToken)
template<>
void	spmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const float alpha, const float *AP, const float *x, const IndexType incX, const float beta, float *y, const IndexType incY, SyncToken *syncToken)
	spmv
template<>
void	spmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const double alpha, const double *AP, const double *x, const IndexType incX, const double beta, double *y, const IndexType incY, SyncToken *syncToken)
template<>
void	spr (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const float alpha, const float *x, const IndexType incX, float *AP, SyncToken *syncToken)
	spr
template<>
void	spr (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const double alpha, const double *x, const IndexType incX, double *AP, SyncToken *syncToken)
template<>
void	spr2 (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const float alpha, const float *x, const IndexType incX, const float *y, const IndexType incY, float *AP, SyncToken *syncToken)
	spr2
template<>
void	spr2 (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const double alpha, const double *x, const IndexType incX, const double *y, const IndexType incY, double *AP, SyncToken *syncToken)
template<>
void	tpmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const float *AP, float *x, const IndexType incX, SyncToken *syncToken)
	tpmv
template<>
void	tpmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const double *AP, double *x, const IndexType incX, SyncToken *syncToken)
template<>
void	tpsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const float *AP, float *x, const IndexType incX, SyncToken *syncToken)
	tpsv
template<>
void	tpsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const double *AP, double *x, const IndexType incX, SyncToken *syncToken)
Static Public Member Functions
template<typename T >
static void	gemv (const enum CBLAS_ORDER order, const enum CBLAS_TRANSPOSE trans, const IndexType m, const IndexType n, const T alpha, const T *A, const IndexType lda, const T *x, const IndexType incX, const T beta, T *y, const IndexType incY, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::gemv.
template<typename T >
static void	symv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const T alpha, const T *A, const IndexType lda, const T *x, const IndexType incX, const T beta, T *y, const IndexType incY, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::symv.
template<typename T >
static void	trmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const T *A, const IndexType lda, T *x, const IndexType incX, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::trmv.
template<typename T >
static void	trsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const T *A, const IndexType lda, T *x, const IndexType incX, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::trsv.
template<typename T >
static void	gbmv (const enum CBLAS_ORDER order, const enum CBLAS_TRANSPOSE trans, const IndexType m, const IndexType n, const IndexType kl, const IndexType ku, const T alpha, const T *A, const IndexType lda, const T *x, const IndexType incX, const T beta, T *y, const IndexType incY, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::gbmv.
template<typename T >
static void	sbmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const IndexType k, const T alpha, const T *A, const IndexType lda, const T *x, const IndexType incX, const T beta, T *y, const IndexType incY, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::sbmv.
template<typename T >
static void	tbmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const IndexType k, const T *A, const IndexType lda, T *x, const IndexType incX, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::tbmv.
template<typename T >
static void	tbsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const IndexType k, const T *A, const IndexType lda, T *x, const IndexType incX, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::tbsv.
template<typename T >
static void	ger (const enum CBLAS_ORDER order, const IndexType m, const IndexType n, const T alpha, const T *x, const IndexType incX, const T *y, const IndexType incY, T *A, const IndexType lda, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::ger.
template<typename T >
static void	syr (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const T alpha, const T *x, const IndexType incX, T *A, const IndexType lda, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::syr.
template<typename T >
static void	syr2 (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const T alpha, const T *x, const IndexType incX, const T *y, const IndexType incY, T *A, const IndexType lda, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::syr2.
template<typename T >
static void	spmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const T alpha, const T *AP, const T *x, const IndexType incX, const T beta, T *y, const IndexType incY, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::spmv.
template<typename T >
static void	spr (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const T alpha, const T *x, const IndexType incX, T *AP, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::spr.
template<typename T >
static void	spr2 (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const IndexType n, const T alpha, const T *x, const IndexType incX, const T *y, const IndexType incY, T *AP, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::spr2.
template<typename T >
static void	tpmv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const T *AP, T *x, const IndexType incX, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::tpmv.
template<typename T >
static void	tpsv (const enum CBLAS_ORDER order, const enum CBLAS_UPLO uplo, const enum CBLAS_TRANSPOSE trans, const enum CBLAS_DIAG diag, const IndexType n, const T *Ap, T *x, const IndexType incX, SyncToken *syncToken)
	This function is the CUDA implementation of lama::BLAS2Interface::tpsv.
Private Member Functions
	LAMA_LOG_DECL_STATIC_LOGGER (logger)

---

Member Function Documentation

template<typename T >

static void lama::CUDABLAS2::gbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_TRANSPOSE	trans,
		const IndexType	m,
		const IndexType	n,
		const IndexType	kl,
		const IndexType	ku,
		const T	alpha,
		const T *	A,
		const IndexType	lda,
		const T *	x,
		const IndexType	incX,
		const T	beta,
		T *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::gbmv.

template<>

void lama::CUDABLAS2::gbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_TRANSPOSE	trans,
		const IndexType	m,
		const IndexType	n,
		IndexType	kl,
		IndexType	ku,
		const float	alpha,
		const float *	A,
		const IndexType	lda,
		const float *	x,
		const IndexType	incX,
		const float	beta,
		float *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

gbmv

References lama::CblasNoTrans, lama::CblasRowMajor, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::gbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_TRANSPOSE	trans,
		const IndexType	m,
		const IndexType	n,
		IndexType	kl,
		IndexType	ku,
		const double	alpha,
		const double *	A,
		const IndexType	lda,
		const double *	x,
		const IndexType	incX,
		const double	beta,
		double *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

References lama::CblasNoTrans, lama::CblasRowMajor, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::gemv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_TRANSPOSE	trans,
		const IndexType	m,
		const IndexType	n,
		const float	alpha,
		const float *const	A,
		const IndexType	lda,
		const float *const	x,
		const IndexType	incx,
		const float	beta,
		float *const	y,
		const IndexType	incy,
		SyncToken *	syncToken
	)

gemv

References lama::CblasNoTrans, lama::CblasRowMajor, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::gemv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_TRANSPOSE	trans,
		const IndexType	m,
		const IndexType	n,
		const T	alpha,
		const T *	A,
		const IndexType	lda,
		const T *	x,
		const IndexType	incX,
		const T	beta,
		T *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::gemv.

template<>

void lama::CUDABLAS2::gemv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_TRANSPOSE	trans,
		const IndexType	m,
		const IndexType	n,
		const double	alpha,
		const double *const	A,
		const IndexType	lda,
		const double *const	x,
		const IndexType	incx,
		const double	beta,
		double *const	y,
		const IndexType	incy,
		SyncToken *	syncToken
	)

References lama::CblasNoTrans, lama::CblasRowMajor, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::ger	(	const enum CBLAS_ORDER	order,
		const IndexType	m,
		const IndexType	n,
		const T	alpha,
		const T *	x,
		const IndexType	incX,
		const T *	y,
		const IndexType	incY,
		T *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::ger.

template<>

void lama::CUDABLAS2::ger	(	const enum CBLAS_ORDER	order,
		const IndexType	m,
		const IndexType	n,
		const float	alpha,
		const float *	x,
		const IndexType	incX,
		const float *	y,
		const IndexType	incY,
		float *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)

ger

References lama::CblasRowMajor, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUBLAS_ERROR, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::ger	(	const enum CBLAS_ORDER	order,
		const IndexType	m,
		const IndexType	n,
		const double	alpha,
		const double *	x,
		const IndexType	incX,
		const double *	y,
		const IndexType	incY,
		double *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)

References lama::CblasRowMajor, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUBLAS_ERROR, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

lama::CUDABLAS2::LAMA_LOG_DECL_STATIC_LOGGER

(

logger

)

[private]

template<typename T >

static void lama::CUDABLAS2::sbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const IndexType	k,
		const T	alpha,
		const T *	A,
		const IndexType	lda,
		const T *	x,
		const IndexType	incX,
		const T	beta,
		T *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::sbmv.

template<>

void lama::CUDABLAS2::sbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const IndexType	k,
		const float	alpha,
		const float *	A,
		const IndexType	lda,
		const float *	x,
		const IndexType	incX,
		const float	beta,
		float *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

sbmv

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::sbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const IndexType	k,
		const double	alpha,
		const double *	A,
		const IndexType	lda,
		const double *	x,
		const IndexType	incX,
		const double	beta,
		double *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::spmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const T	alpha,
		const T *	AP,
		const T *	x,
		const IndexType	incX,
		const T	beta,
		T *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::spmv.

template<>

void lama::CUDABLAS2::spmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const float	alpha,
		const float *	AP,
		const float *	x,
		const IndexType	incX,
		const float	beta,
		float *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

spmv

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::spmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const double	alpha,
		const double *	AP,
		const double *	x,
		const IndexType	incX,
		const double	beta,
		double *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::spr	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const T	alpha,
		const T *	x,
		const IndexType	incX,
		T *	AP,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::spr.

template<>

void lama::CUDABLAS2::spr	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const float	alpha,
		const float *	x,
		const IndexType	incX,
		float *	AP,
		SyncToken *	syncToken
	)

spr

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::spr	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const double	alpha,
		const double *	x,
		const IndexType	incX,
		double *	AP,
		SyncToken *	syncToken
	)

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::spr2	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const T	alpha,
		const T *	x,
		const IndexType	incX,
		const T *	y,
		const IndexType	incY,
		T *	AP,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::spr2.

template<>

void lama::CUDABLAS2::spr2	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const float	alpha,
		const float *	x,
		const IndexType	incX,
		const float *	y,
		const IndexType	incY,
		float *	AP,
		SyncToken *	syncToken
	)

spr2

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::spr2	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const double	alpha,
		const double *	x,
		const IndexType	incX,
		const double *	y,
		const IndexType	incY,
		double *	AP,
		SyncToken *	syncToken
	)

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::symv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const T	alpha,
		const T *	A,
		const IndexType	lda,
		const T *	x,
		const IndexType	incX,
		const T	beta,
		T *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::symv.

template<>

void lama::CUDABLAS2::symv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const float	alpha,
		const float *	A,
		const IndexType	lda,
		const float *	x,
		const IndexType	incX,
		const float	beta,
		float *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

symv

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::symv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const double	alpha,
		const double *	A,
		const IndexType	lda,
		const double *	x,
		const IndexType	incX,
		const double	beta,
		double *	y,
		const IndexType	incY,
		SyncToken *	syncToken
	)

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::syr	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const T	alpha,
		const T *	x,
		const IndexType	incX,
		T *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::syr.

template<>

void lama::CUDABLAS2::syr	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const float	alpha,
		const float *	x,
		const IndexType	incX,
		float *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)

syr

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::syr	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const double	alpha,
		const double *	x,
		const IndexType	incX,
		double *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::syr2	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const T	alpha,
		const T *	x,
		const IndexType	incX,
		const T *	y,
		const IndexType	incY,
		T *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::syr2.

template<>

void lama::CUDABLAS2::syr2	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const float	alpha,
		const float *	x,
		const IndexType	incX,
		const float *	y,
		const IndexType	incY,
		float *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)

syr2

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::syr2	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const IndexType	n,
		const double	alpha,
		const double *	x,
		const IndexType	incX,
		const double *	y,
		const IndexType	incY,
		double *	A,
		const IndexType	lda,
		SyncToken *	syncToken
	)

References lama::CblasLower, lama::CblasRowMajor, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::tbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const IndexType	k,
		const T *	A,
		const IndexType	lda,
		T *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::tbmv.

template<>

void lama::CUDABLAS2::tbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const IndexType	k,
		const float *	A,
		const IndexType	lda,
		float *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

tbmv

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::tbmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const IndexType	k,
		const double *	A,
		const IndexType	lda,
		double *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::tbsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const IndexType	k,
		const T *	A,
		const IndexType	lda,
		T *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::tbsv.

template<>

void lama::CUDABLAS2::tbsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const IndexType	k,
		const float *	A,
		const IndexType	lda,
		float *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

tbsv

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::tbsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const IndexType	k,
		const double *	A,
		const IndexType	lda,
		double *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::tpmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const T *	AP,
		T *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::tpmv.

template<>

void lama::CUDABLAS2::tpmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const float *	AP,
		float *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

tpmv

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::tpmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const double *	AP,
		double *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::tpsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const T *	Ap,
		T *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::tpsv.

template<>

void lama::CUDABLAS2::tpsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const float *	AP,
		float *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

tpsv

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::tpsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const double *	AP,
		double *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::trmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const T *	A,
		const IndexType	lda,
		T *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::trmv.

template<>

void lama::CUDABLAS2::trmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const float *	A,
		const IndexType	lda,
		float *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

trmv

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::trmv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const double *	A,
		const IndexType	lda,
		double *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<typename T >

static void lama::CUDABLAS2::trsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const T *	A,
		const IndexType	lda,
		T *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)		[static]

This function is the CUDA implementation of lama::BLAS2Interface::trsv.

template<>

void lama::CUDABLAS2::trsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const float *	A,
		const IndexType	lda,
		float *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

trsv

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

template<>

void lama::CUDABLAS2::trsv	(	const enum CBLAS_ORDER	order,
		const enum CBLAS_UPLO	uplo,
		const enum CBLAS_TRANSPOSE	trans,
		const enum CBLAS_DIAG	diag,
		const IndexType	n,
		const double *	A,
		const IndexType	lda,
		double *	x,
		const IndexType	incX,
		SyncToken *	syncToken
	)

References lama::CblasNonUnit, lama::CblasRowMajor, lama::CblasTrans, lama::CblasUnit, lama::CblasUpper, lama::CUDAStreamSyncToken::getCUDAStream(), LAMA_ASSERT_DEBUG, LAMA_CHECK_CUDA_ACCESS, LAMA_CUBLAS_CALL, and LAMA_CUDA_RT_CALL.

---

The documentation for this class was generated from the following files:

• /home/brandes/workspace/LAMA/src/lama/cuda/CUDABLAS2.hpp
• /home/brandes/workspace/LAMA/src/lama/cuda/CUDABLAS2.cpp