Inheritance diagram for quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >:

Collaboration diagram for quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >:

[legend]

Public Member Functions
	MultiBlas (SpinorX X[], SpinorY Y[], SpinorZ Z[], SpinorW W[], Functor &f, const coeff_array< T > &a, const coeff_array< T > &b, const coeff_array< T > &c, std::vector< ColorSpinorField > &x, std::vector< ColorSpinorField > &y, std::vector< ColorSpinorField > &z, std::vector< ColorSpinorField > &w, int NYW, int length)

virtual	~MultiBlas ()

TuneKey	tuneKey () const

void	apply (const cudaStream_t &stream)

void	preTune ()

void	postTune ()

void	initTuneParam (TuneParam &param) const

void	defaultTuneParam (TuneParam &param) const

long long	flops () const

long long	bytes () const

int	tuningIter () const

Public Member Functions inherited from quda::TunableVectorY
	TunableVectorY (unsigned int vector_length_y)

bool	advanceBlockDim (TuneParam &param) const

void	resizeVector (int y) const

void	resizeStep (int y) const

Public Member Functions inherited from quda::Tunable
	Tunable ()

virtual	~Tunable ()

virtual std::string	paramString (const TuneParam &param) const

virtual std::string	perfString (float time) const

virtual bool	advanceTuneParam (TuneParam &param) const

void	checkLaunchParam (TuneParam &param)

CUresult	jitifyError () const

CUresult &	jitifyError ()

Private Member Functions
bool	tuneSharedBytes () const

Private Attributes
const int	NYW

const int	nParity

MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >	arg

const coeff_array< T > &	a

const coeff_array< T > &	b

const coeff_array< T > &	c

std::vector< ColorSpinorField * > &	x

std::vector< ColorSpinorField * > &	y

std::vector< ColorSpinorField * > &	z

std::vector< ColorSpinorField * > &	w

char *	Y_h [MAX_MULTI_BLAS_N]

char *	W_h [MAX_MULTI_BLAS_N]

char *	Ynorm_h [MAX_MULTI_BLAS_N]

char *	Wnorm_h [MAX_MULTI_BLAS_N]

Additional Inherited Members
Protected Member Functions inherited from quda::TunableVectorY
virtual unsigned int	sharedBytesPerThread () const

virtual unsigned int	sharedBytesPerBlock (const TuneParam &param) const

Protected Member Functions inherited from quda::Tunable
virtual unsigned int	minThreads () const

virtual bool	tuneGridDim () const

virtual bool	tuneAuxDim () const

virtual bool	advanceGridDim (TuneParam &param) const

virtual unsigned int	maxBlockSize (const TuneParam &param) const

virtual unsigned int	maxGridSize () const

virtual unsigned int	minGridSize () const

virtual int	gridStep () const
	gridStep sets the step size when iterating the grid size in advanceGridDim. More...

virtual int	blockStep () const

virtual int	blockMin () const

virtual void	resetBlockDim (TuneParam &param) const

unsigned int	maxBlocksPerSM () const
	For some reason this can't be queried from the device properties, so here we set set this. Based on Table 14 of the CUDA Programming Guide 10.0 (Technical Specifications per Compute Capability) More...

template<typename F >
void	setMaxDynamicSharedBytesPerBlock (F *func) const
	Enable the maximum dynamic shared bytes for the kernel "func" (values given by maxDynamicSharedBytesPerBlock()). More...

unsigned int	maxDynamicSharedBytesPerBlock () const
	This can't be correctly queried in CUDA for all architectures so here we set set this. Based on Table 14 of the CUDA Programming Guide 10.0 (Technical Specifications per Compute Capability). More...

virtual unsigned int	maxSharedBytesPerBlock () const
	The maximum shared memory that a CUDA thread block can use in the autotuner. This isn't necessarily the same as maxDynamicSharedMemoryPerBlock since that may need explicit opt in to enable (by calling setMaxDynamicSharedBytes for the kernel in question). If the CUDA kernel in question does this opt in then this function can be overloaded to return maxDynamicSharedBytesPerBlock. More...

virtual bool	advanceSharedBytes (TuneParam &param) const

virtual bool	advanceAux (TuneParam &param) const

int	writeAuxString (const char *format,...)

Protected Attributes inherited from quda::TunableVectorY
unsigned int	vector_length_y

unsigned int	step_y

bool	tune_block_x

Protected Attributes inherited from quda::Tunable
char	aux [TuneKey::aux_n]

CUresult	jitify_error

Detailed Description

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>
class quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >

Definition at line 47 of file multi_blas_quda.cu.

Constructor & Destructor Documentation

◆ MultiBlas()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::MultiBlas	(	SpinorX	X[],
		SpinorY	Y[],
		SpinorZ	Z[],
		SpinorW	W[],
		Functor &	f,
		const coeff_array< T > &	a,
		const coeff_array< T > &	b,
		const coeff_array< T > &	c,
		std::vector< ColorSpinorField *> &	x,
		std::vector< ColorSpinorField *> &	y,
		std::vector< ColorSpinorField *> &	z,
		std::vector< ColorSpinorField *> &	w,
		int	NYW,
		int	length
	)

inline

Definition at line 65 of file multi_blas_quda.cu.

References quda::blas::Amatrix_h, quda::blas::Bmatrix_h, quda::blas::Cmatrix_h, and quda::blas::coeff_array< T >::data.

◆ ~MultiBlas()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

virtual quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::~MultiBlas ( )

inlinevirtual

Definition at line 100 of file multi_blas_quda.cu.

Member Function Documentation

◆ apply()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

void quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::apply ( const cudaStream_t & stream )

inlinevirtual

Implements quda::Tunable.

Definition at line 111 of file multi_blas_quda.cu.

References quda::blas::Amatrix_d, quda::arg(), quda::blas::arg_buffer, quda::TuneParam::block, quda::blas::Bmatrix_d, quda::blas::Cmatrix_d, quda::blas::coeff_array< T >::data, quda::blas::getStream(), getTuning(), getVerbosity(), quda::TuneParam::grid, MAX_MATRIX_SIZE, MAX_MULTI_BLAS_N, quda::TuneParam::shared_bytes, quda::tuneLaunch(), and quda::blas::coeff_array< T >::use_const.

Referenced by quda::blas::multiBlas().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ bytes()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

long long quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::bytes ( ) const

inlinevirtual

Reimplemented from quda::Tunable.

Definition at line 234 of file multi_blas_quda.cu.

References quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::f.

Referenced by quda::blas::multiBlas().

Here is the caller graph for this function:

◆ defaultTuneParam()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

void quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::defaultTuneParam ( TuneParam & param ) const

inlinevirtual

sets default values for when tuning is disabled

Reimplemented from quda::TunableVectorY.

Definition at line 226 of file multi_blas_quda.cu.

References quda::TunableVectorY::defaultTuneParam(), and quda::TuneParam::grid.

Here is the call graph for this function:

◆ flops()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

long long quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::flops ( ) const

inlinevirtual

Implements quda::Tunable.

Definition at line 232 of file multi_blas_quda.cu.

References quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::f, and quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::length.

Referenced by quda::blas::multiBlas().

Here is the caller graph for this function:

◆ initTuneParam()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

void quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::initTuneParam ( TuneParam & param ) const

inlinevirtual

Reimplemented from quda::TunableVectorY.

Definition at line 220 of file multi_blas_quda.cu.

References quda::TuneParam::grid, and quda::TunableVectorY::initTuneParam().

Here is the call graph for this function:

◆ postTune()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

void quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::postTune ( )

inlinevirtual

Reimplemented from quda::Tunable.

Definition at line 212 of file multi_blas_quda.cu.

References quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::W, and quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::Y.

◆ preTune()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

void quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::preTune ( )

inlinevirtual

Reimplemented from quda::Tunable.

Definition at line 204 of file multi_blas_quda.cu.

References quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::W, and quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::Y.

◆ tuneKey()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

TuneKey quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::tuneKey ( ) const

inlinevirtual

Implements quda::Tunable.

Definition at line 102 of file multi_blas_quda.cu.

References quda::blas::MultiBlasArg< NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor >::f, and quda::TuneKey::name_n.

◆ tuneSharedBytes()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

bool quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::tuneSharedBytes ( ) const

inlineprivatevirtual

Reimplemented from quda::Tunable.

Definition at line 62 of file multi_blas_quda.cu.

◆ tuningIter()

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

int quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::tuningIter ( ) const

inlinevirtual

Reimplemented from quda::Tunable.

Definition at line 240 of file multi_blas_quda.cu.

Member Data Documentation

◆ a

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

const coeff_array<T>& quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::a

private

Definition at line 54 of file multi_blas_quda.cu.

◆ arg

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

MultiBlasArg<NXZ, SpinorX, SpinorY, SpinorZ, SpinorW, Functor> quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::arg

mutableprivate

Definition at line 53 of file multi_blas_quda.cu.

◆ b

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

const coeff_array<T> & quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::b

private

Definition at line 54 of file multi_blas_quda.cu.

◆ c

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

const coeff_array<T> & quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::c

private

Definition at line 54 of file multi_blas_quda.cu.

◆ nParity

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

const int quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::nParity

private

Definition at line 52 of file multi_blas_quda.cu.

◆ NYW

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

const int quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::NYW

private

Definition at line 51 of file multi_blas_quda.cu.

◆ w

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

std::vector<ColorSpinorField *> & quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::w

private

Definition at line 56 of file multi_blas_quda.cu.

◆ W_h

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

char * quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::W_h[MAX_MULTI_BLAS_N]

private

Definition at line 60 of file multi_blas_quda.cu.

◆ Wnorm_h

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

char * quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::Wnorm_h[MAX_MULTI_BLAS_N]

private

Definition at line 60 of file multi_blas_quda.cu.

◆ x

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

std::vector<ColorSpinorField *>& quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::x

private

Definition at line 56 of file multi_blas_quda.cu.

◆ y

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

std::vector<ColorSpinorField *> & quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::y

private

Definition at line 56 of file multi_blas_quda.cu.

◆ Y_h

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

char* quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::Y_h[MAX_MULTI_BLAS_N]

private

Definition at line 60 of file multi_blas_quda.cu.

◆ Ynorm_h

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

char * quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::Ynorm_h[MAX_MULTI_BLAS_N]

private

Definition at line 60 of file multi_blas_quda.cu.

◆ z

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>

std::vector<ColorSpinorField *> & quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >::z

private

Definition at line 56 of file multi_blas_quda.cu.

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

lib/multi_blas_quda.cu

Public Member Functions

Private Member Functions

Private Attributes

Additional Inherited Members

Detailed Description

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T> class quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >

Constructor & Destructor Documentation

◆ MultiBlas()

◆ ~MultiBlas()

Member Function Documentation

◆ apply()

◆ bytes()

◆ defaultTuneParam()

◆ flops()

◆ initTuneParam()

◆ postTune()

◆ preTune()

◆ tuneKey()

◆ tuneSharedBytes()

◆ tuningIter()

Member Data Documentation

◆ a

◆ arg

◆ b

◆ c

◆ nParity

◆ NYW

◆ w

◆ W_h

◆ Wnorm_h

◆ x

◆ y

◆ Y_h

◆ Ynorm_h

◆ z

template<int NXZ, typename FloatN, int M, typename SpinorX, typename SpinorY, typename SpinorZ, typename SpinorW, typename Functor, typename T>
class quda::blas::MultiBlas< NXZ, FloatN, M, SpinorX, SpinorY, SpinorZ, SpinorW, Functor, T >