RDBHilbert< T, MatrixType > Class Template Reference

The Beta Hilbert triplet scheme. More...

#include <RDBHilbert.hpp>

Inheritance diagram for RDBHilbert< T, MatrixType >:

Collaboration diagram for RDBHilbert< T, MatrixType >:

Public Member Functions
	RDBHilbert (const std::string file, T zero=0, std::vector< size_t > *_p_translate=NULL)
	Default file-based constructor. More...
	RDBHilbert (std::vector< Triplet< T > > &input, ULI m, ULI n, T zero=0, std::vector< size_t > *_p_translate=NULL)
	Default triplet-based constructor. More...
virtual	~RDBHilbert ()
	Base deconstructor. More...
void	wait ()
virtual void	load (std::vector< Triplet< T > > &input, const ULI m, const ULI n, const T zero)
virtual T *	mv (const T *x)
	Overloaded mv call; allocates output vector using numa_interleaved. More...
virtual void	zxa (const T *x, T *z)
virtual void	zxa (const T *x, T *z, const size_t repeat)
void	reset ()
virtual void	zax (const T *x, T *z)
virtual void	zax (const T *x, T *z, const size_t repeat, const clockid_t clock_id, double *elapsed_time)
virtual void	getFirstIndexPair (ULI &i, ULI &j)
virtual size_t	bytesUsed ()
Public Member Functions inherited from SparseMatrix< T, ULI >
	SparseMatrix ()
	Base constructor. More...
	SparseMatrix (const ULInzs, const ULInr, const ULInc, const T zero)
	Base constructor. More...
virtual	~SparseMatrix ()
	Base deconstructor. More...
void	loadFromFile (const std::string file, const T zero=0)
	Function which loads a matrix from a matrix market file. More...
virtual unsigned long int	m ()
	Queries the number of rows this matrix contains. More...
virtual unsigned long int	n ()
	Queries the number of columns this matrix contains. More...
virtual unsigned long int	nzs ()
	Queries the number of nonzeroes stored in this matrix. More...
virtual void	zax (const T *__restrict__ x, T *__restrict__ z)=0
	In-place z=Ax function. More...
virtual void	zxa (const T *__restrict__ x, T *__restrict__ z)=0
	In-place z=xA function. More...
Public Member Functions inherited from Matrix< T >
	Matrix ()
	Base constructor. More...
virtual	~Matrix ()
	Base deconstructor. More...
virtual void	zax (const T *__restrict__ x, T *__restrict__ z, const size_t k, const clockid_t clock_id=0, double *elapsed_time=NULL)
	Wrapper function to call the zax kernel multiple times successively, while timing the duration of the operation. More...
template<size_t k>
void	ZaX (const T *__restrict__ const *__restrict__ const X, T *__restrict__ const *__restrict__ const Z)
	In-place Z=AX function, where A is m x n, Z = m x k, and X is n x k. More...
template<size_t k>
void	ZXa (const T *__restrict__ const *__restrict__ const X, T *__restrict__ const *__restrict__ const Z)
	In-place Z=XA function, where A is m x n, Z = k x n, and X is k x m. More...
virtual void	zxa (const T *__restrict__ x, T *__restrict__ z, const unsigned long int repeat, const clockid_t clock_id=0, double *elapsed_time=NULL)
	Wrapper function to call the zxa kernel multiple times successively, while timing the operation duration. More...

Static Public Member Functions
static void	end (pthread_mutex_t *mutex, pthread_cond_t *cond, size_t *sync, const size_t P)
static void	synchronise (pthread_mutex_t *mutex, pthread_cond_t *cond, size_t *sync, const size_t P)
static void *	thread (void *data)
static void	collectY (RDB_shared_data< T > *shared)

Protected Member Functions
void	set_p_translate (std::vector< size_t > *_p_translate)
	Sets p_translate to 0..P-1 by default, or equal to the optionally supplied vector. More...

Protected Attributes
std::vector< size_t >	p_translate
	Which processors to pin threads to.
const T *	input
	Input vector.
T *	output
	Output vector.
pthread_t *	threads
	Input vectors for multiple-RHS operations. More...
RDB_shared_data< T > *	thread_data
	array of initial thread data
pthread_mutex_t	mutex
	Stop/continue mechanism: mutex.
pthread_cond_t	cond
	Stop/continue mechanism: condition.
pthread_mutex_t	end_mutex
	Wait for end mechanism: mutex.
pthread_cond_t	end_cond
	Wait for end mechanism: condition.
size_t	sync
	Used for synchronising threads.
size_t	end_sync
	Used for construction end signal.
Protected Attributes inherited from SparseMatrix< T, ULI >
ULI	nor
	Number of rows. More...
ULI	noc
	Number of columns.
ULI	nnz
	Number of non-zeros. More...

Static Protected Attributes
static size_t	P = 0
	Number of threads to fire up.
static clockid_t	global_clock_id = 0
	Clock type used for thread-local timing.
static const ULI	max_n = FBICRS< T >::beta_n
	Given FBICRS, the maximum value for columnwise matrix size.
static const ULI	max_m = FBICRS< T >::beta_m
	Given FBICRS, the maximum value for the rowwise matrix size, assuming short ints on ICRS at the lower level.

Additional Inherited Members
Public Attributes inherited from SparseMatrix< T, ULI >
T	zero_element
	The element considered to be zero. More...

Detailed Description

template<typename T, class MatrixType = FBICRS< T >>
class RDBHilbert< T, MatrixType >

The Beta Hilbert triplet scheme.

Full parallel SpMV, based on Blocked Hilbert and PThreads. Inspired by Aydin & Gilbert's CSB and comments by Patrick Amestoy on the BICRS Hilbert scheme (both pointed towards sparse blocking, for different reasons that this scheme combines).

Constructor & Destructor Documentation

template<typename T, class MatrixType = FBICRS< T >>

RDBHilbert< T, MatrixType >::RDBHilbert ( const std::string  file, T  zero = 0, std::vector< size_t > *  _p_translate = NULL  )

inline

Default file-based constructor.

References SparseMatrix< T, ULI >::loadFromFile(), and RDBHilbert< T, MatrixType >::set_p_translate().

template<typename T, class MatrixType = FBICRS< T >>

RDBHilbert< T, MatrixType >::RDBHilbert ( std::vector< Triplet< T > > &  input, ULI  m, ULI  n, T  zero = 0, std::vector< size_t > *  _p_translate = NULL  )

inline

Default triplet-based constructor.

References RDBHilbert< T, MatrixType >::input, RDBHilbert< T, MatrixType >::load(), and RDBHilbert< T, MatrixType >::set_p_translate().

template<typename T, class MatrixType = FBICRS< T >>

virtual RDBHilbert< T, MatrixType >::~RDBHilbert ( )

inlinevirtual

Base deconstructor.

References RDBHilbert< T, MatrixType >::cond, RDBHilbert< T, MatrixType >::mutex, RDBHilbert< T, MatrixType >::P, RDBHilbert< T, MatrixType >::thread_data, and RDBHilbert< T, MatrixType >::threads.

Member Function Documentation

template<typename T, class MatrixType = FBICRS< T >>

virtual size_t RDBHilbert< T, MatrixType >::bytesUsed ( )

inlinevirtual

See also

Matrix::bytesUsed

Implements Matrix< T >.

References RDBHilbert< T, MatrixType >::P, and RDBHilbert< T, MatrixType >::thread_data.

template<typename T, class MatrixType = FBICRS< T >>

static void RDBHilbert< T, MatrixType >::collectY ( RDB_shared_data< T > *  shared )

inlinestatic

See also

RDScheme::collectY

References RDB_shared_data< T >::id, RDB_shared_data< T >::local_y, RDB_shared_data< T >::output, RDB_shared_data< T >::output_vector_offset, and RDB_shared_data< T >::output_vector_size.

Referenced by RDBHilbert< T, MatrixType >::thread().

template<typename T, class MatrixType = FBICRS< T >>

static void RDBHilbert< T, MatrixType >::end ( pthread_mutex_t *  mutex, pthread_cond_t *  cond, size_t *  sync, const size_t  P  )

inlinestatic

See also

RDScheme::end

Referenced by RDBHilbert< T, MatrixType >::thread().

template<typename T, class MatrixType = FBICRS< T >>

virtual void RDBHilbert< T, MatrixType >::getFirstIndexPair ( ULI &  i, ULI &  j  )

inlinevirtual

See also

SparseMatrix::getFirstIndexPair

Implements SparseMatrix< T, ULI >.

template<typename T, class MatrixType = FBICRS< T >>

virtual void RDBHilbert< T, MatrixType >::load ( std::vector< Triplet< T > > &  input, const ULI  m, const ULI  n, const T  zero  )

inlinevirtual

See also

SparseMatrix::load

Implements SparseMatrix< T, ULI >.

References RDBHilbert< T, MatrixType >::cond, RDBHilbert< T, MatrixType >::end_cond, RDBHilbert< T, MatrixType >::end_mutex, RDBHilbert< T, MatrixType >::end_sync, RDBHilbert< T, MatrixType >::input, SparseMatrix< T, ULI >::m(), RDBHilbert< T, MatrixType >::mutex, SparseMatrix< T, ULI >::n(), SparseMatrix< T, ULI >::nnz, SparseMatrix< T, ULI >::noc, SparseMatrix< T, ULI >::nor, RDBHilbert< T, MatrixType >::output, RDBHilbert< T, MatrixType >::P, RDBHilbert< T, MatrixType >::p_translate, RDBHilbert< T, MatrixType >::sync, RDBHilbert< T, MatrixType >::thread(), RDBHilbert< T, MatrixType >::thread_data, RDBHilbert< T, MatrixType >::threads, RDBHilbert< T, MatrixType >::wait(), and SparseMatrix< T, ULI >::zero_element.

Referenced by RDBHilbert< T, MatrixType >::RDBHilbert().

template<typename T, class MatrixType = FBICRS< T >>

virtual T* RDBHilbert< T, MatrixType >::mv ( const T *  x )

inlinevirtual

Overloaded mv call; allocates output vector using numa_interleaved.

Reimplemented from SparseMatrix< T, ULI >.

References SparseMatrix< T, ULI >::nor, RDBHilbert< T, MatrixType >::reset(), RDBHilbert< T, MatrixType >::zax(), and SparseMatrix< T, ULI >::zero_element.

template<typename T, class MatrixType = FBICRS< T >>

void RDBHilbert< T, MatrixType >::reset ( )

inline

See also

RDScheme::reset

References RDBHilbert< T, MatrixType >::cond, RDBHilbert< T, MatrixType >::end_mutex, RDBHilbert< T, MatrixType >::mutex, RDBHilbert< T, MatrixType >::P, RDBHilbert< T, MatrixType >::thread_data, and RDBHilbert< T, MatrixType >::wait().

Referenced by RDBHilbert< T, MatrixType >::mv().

template<typename T, class MatrixType = FBICRS< T >>

void RDBHilbert< T, MatrixType >::set_p_translate ( std::vector< size_t > *  _p_translate )

inlineprotected

Sets p_translate to 0..P-1 by default, or equal to the optionally supplied vector.

References MachineInfo::cores(), MachineInfo::getInstance(), RDBHilbert< T, MatrixType >::P, and RDBHilbert< T, MatrixType >::p_translate.

Referenced by RDBHilbert< T, MatrixType >::RDBHilbert().

template<typename T, class MatrixType = FBICRS< T >>

static void RDBHilbert< T, MatrixType >::synchronise ( pthread_mutex_t *  mutex, pthread_cond_t *  cond, size_t *  sync, const size_t  P  )

inlinestatic

See also

RDScheme::synchronise

Referenced by RDBHilbert< T, MatrixType >::thread().

template<typename T, class MatrixType = FBICRS< T >>

static void* RDBHilbert< T, MatrixType >::thread ( void *  data )

inlinestatic

See also

RDScheme::thread

References RDB_shared_data< T >::bytes, RDBHilbert< T, MatrixType >::collectY(), RDB_shared_data< T >::cond, RDBHilbert< T, MatrixType >::cond, RDBHilbert< T, MatrixType >::end(), RDB_shared_data< T >::end_cond, RDB_shared_data< T >::end_mutex, RDB_shared_data< T >::end_sync, RDB_shared_data< T >::fillIn, RDBHilbert< T, MatrixType >::global_clock_id, RDB_shared_data< T >::id, RDB_shared_data< T >::input, Matrix2HilbertCoordinates::IntegerToHilbert(), RDB_shared_data< T >::local_y, SparseMatrix< T, ULI >::m(), RDBHilbert< T, MatrixType >::max_m, RDBHilbert< T, MatrixType >::max_n, RDB_shared_data< T >::mode, RDB_shared_data< T >::mutex, RDBHilbert< T, MatrixType >::mutex, SparseMatrix< T, ULI >::n(), SparseMatrix< T, ULI >::nnz, RDB_shared_data< T >::nzb, RDB_shared_data< T >::original, RDB_shared_data< T >::output, RDB_shared_data< T >::output_vector_offset, RDB_shared_data< T >::output_vector_size, RDB_shared_data< T >::P, RDBHilbert< T, MatrixType >::P, RDB_shared_data< T >::repeat, RDB_shared_data< T >::sync, RDBHilbert< T, MatrixType >::synchronise(), and RDB_shared_data< T >::time.

Referenced by RDBHilbert< T, MatrixType >::load().

template<typename T, class MatrixType = FBICRS< T >>

void RDBHilbert< T, MatrixType >::wait ( )

inline

See also

RDScheme::wait

References RDBHilbert< T, MatrixType >::end_cond, and RDBHilbert< T, MatrixType >::end_mutex.

Referenced by RDBHilbert< T, MatrixType >::load(), RDBHilbert< T, MatrixType >::reset(), RDBHilbert< T, MatrixType >::zax(), and RDBHilbert< T, MatrixType >::zxa().

template<typename T, class MatrixType = FBICRS< T >>

virtual void RDBHilbert< T, MatrixType >::zax ( const T *  x, T *  z  )

inlinevirtual

See also

Matrix::zax

Referenced by RDBHilbert< T, MatrixType >::mv().

template<typename T, class MatrixType = FBICRS< T >>

virtual void RDBHilbert< T, MatrixType >::zax ( const T *  x, T *  z, const size_t  repeat, const clockid_t  clock_id, double *  elapsed_time  )

inlinevirtual

See also

RDScheme::zax

References RDBHilbert< T, MatrixType >::cond, RDBHilbert< T, MatrixType >::end_mutex, RDBHilbert< T, MatrixType >::global_clock_id, RDBHilbert< T, MatrixType >::input, RDBHilbert< T, MatrixType >::mutex, RDBHilbert< T, MatrixType >::output, RDBHilbert< T, MatrixType >::P, RDBHilbert< T, MatrixType >::thread_data, and RDBHilbert< T, MatrixType >::wait().

template<typename T, class MatrixType = FBICRS< T >>

virtual void RDBHilbert< T, MatrixType >::zxa ( const T *  x, T *  z  )

inlinevirtual

See also

Matrix::zxa

template<typename T, class MatrixType = FBICRS< T >>

virtual void RDBHilbert< T, MatrixType >::zxa ( const T *  x, T *  z, const size_t  repeat  )

inlinevirtual

See also

RDScheme::zxa

References RDBHilbert< T, MatrixType >::cond, RDBHilbert< T, MatrixType >::end_mutex, RDBHilbert< T, MatrixType >::input, RDBHilbert< T, MatrixType >::mutex, RDBHilbert< T, MatrixType >::output, RDBHilbert< T, MatrixType >::P, RDBHilbert< T, MatrixType >::thread_data, and RDBHilbert< T, MatrixType >::wait().

Member Data Documentation

template<typename T, class MatrixType = FBICRS< T >>

pthread_t* RDBHilbert< T, MatrixType >::threads

protected

Input vectors for multiple-RHS operations.

Output vectors for multiple-RHS operations. p_threads associated to this data strcuture

Referenced by RDBHilbert< T, MatrixType >::load(), and RDBHilbert< T, MatrixType >::~RDBHilbert().

---

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

• RDBHilbert.hpp