OPM - Reference Documentation for opm-autodiff

CUDA compatiable variant of Dune::OwnerOverlapCopyCommunication. More...

#include <GpuOwnerOverlapCopy.hpp>

Public Types
using	X = GpuVector< field_type >

Public Member Functions
	GpuOwnerOverlapCopy (std::shared_ptr< GPUSender< field_type, OwnerOverlapCopyCommunicationType > > sender)

void	copyOwnerToAll (const X &source, X &dest) const

void	dot (const X &x, const X &y, field_type &output) const

field_type	norm (const X &x) const

void	project (X &x) const

const ::Dune::Communication< MPI_Comm > &	communicator () const
	communicator returns the MPI communicator used by this GpuOwnerOverlapCopy More...

Detailed Description

template<class field_type, class OwnerOverlapCopyCommunicationType>
class Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >

CUDA compatiable variant of Dune::OwnerOverlapCopyCommunication.

This class can essentially be seen as an adapter around Dune::OwnerOverlapCopyCommunication, and should work as a Dune::OwnerOverlapCopyCommunication on GpuVectors

Note: This currently only has the functionality to parallelize the linear solve.

Template Parameters

field_type	should be a field_type supported by GpuVector (double, float)
OwnerOverlapCopyCommunicationType	should mimic Dune::OwnerOverlapCopyCommunication.

Member Typedef Documentation

◆ X

template<class field_type , class OwnerOverlapCopyCommunicationType >

using Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >::X = GpuVector<field_type>

Constructor & Destructor Documentation

◆ GpuOwnerOverlapCopy()

template<class field_type , class OwnerOverlapCopyCommunicationType >

Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >::GpuOwnerOverlapCopy ( std::shared_ptr< GPUSender< field_type, OwnerOverlapCopyCommunicationType > > sender )

inlineexplicit

Member Function Documentation

◆ communicator()

template<class field_type , class OwnerOverlapCopyCommunicationType >

const ::Dune::Communication< MPI_Comm > & Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >::communicator ( ) const

inline

communicator returns the MPI communicator used by this GpuOwnerOverlapCopy

Returns: the MPI communicator

◆ copyOwnerToAll()

template<class field_type , class OwnerOverlapCopyCommunicationType >

void Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >::copyOwnerToAll	(	const X &	source,
		X &	dest
	)		const

inline

◆ dot()

template<class field_type , class OwnerOverlapCopyCommunicationType >

void Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >::dot	(	const X &	x,
		const X &	y,
		field_type &	output
	)		const

inline

◆ norm()

template<class field_type , class OwnerOverlapCopyCommunicationType >

field_type Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >::norm ( const X & x ) const

inline

◆ project()

template<class field_type , class OwnerOverlapCopyCommunicationType >

void Opm::gpuistl::GpuOwnerOverlapCopy< field_type, OwnerOverlapCopyCommunicationType >::project ( X & x ) const

inline

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

GpuOwnerOverlapCopy.hpp