#include <deal.II/matrix_free/cuda_fe_evaluation.h>

Public Types
using	value_type = Number

using	gradient_type = Tensor< 1, dim, Number >

using	data_type = typename MatrixFree< dim, Number >::Data

Public Member Functions
	FEEvaluation (const unsigned int cell_id, const data_type data, SharedData< dim, Number > shdata)

void	read_dof_values (const Number *src)

void	distribute_local_to_global (Number *dst) const

void	evaluate (const bool evaluate_val, const bool evaluate_grad)

void	integrate (const bool integrate_val, const bool integrate_grad)

value_type	get_value (const unsigned int q_point) const

value_type	get_value () const

value_type	get_dof_value (const unsigned int dof) const

value_type	get_dof_value () const

void	submit_value (const value_type &val_in, const unsigned int q_point)

void	submit_value (const value_type &val_in)

void	submit_dof_value (const value_type &val_in, const unsigned int dof)

void	submit_dof_value (const value_type &val_in)

gradient_type	get_gradient (const unsigned int q_point) const

gradient_type	get_gradient () const

void	submit_gradient (const gradient_type &grad_in, const unsigned int q_point)

void	submit_gradient (const gradient_type &grad_in)

template<typename Functor >
void	apply_quad_point_operations (const Functor &func)

template<typename Functor >
void	apply_for_each_quad_point (const Functor &func)

Static Public Attributes
static constexpr unsigned int	dimension = dim

static constexpr unsigned int	n_components = n_components_

static constexpr unsigned int	n_q_points

static constexpr unsigned int	tensor_dofs_per_cell

Private Attributes
types::global_dof_index *	local_to_global

unsigned int	n_cells

unsigned int	padding_length

const unsigned int	mf_object_id

const unsigned int	constraint_mask

const bool	use_coloring

Number *	inv_jac

Number *	JxW

Number *	values

Number *	gradients [dim]

Detailed Description

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>
class CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >

This class provides all the functions necessary to evaluate functions at quadrature points and cell integrations. In functionality, this class is similar to FEValues<dim>.

This class has five template arguments:

Template Parameters

dim	Dimension in which this class is to be used
fe_degree	Degree of the tensor prodict finite element with fe_degree+1 degrees of freedom per coordinate direction
n_q_points_1d	Number of points in the quadrature formular in 1D, defaults to fe_degree+1
n_components	Number of vector components when solving a system of PDEs. If the same operation is applied to several components of a PDE (e.g. a vector Laplace equation), they can be applied simultaneously with one call (and often more efficiently). Defaults to 1
Number	Number format, `double` or `float`. Defaults to `double`.

Definition at line 90 of file cuda_fe_evaluation.h.

Member Typedef Documentation

◆ value_type

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

using CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::value_type = Number

An alias for scalar quantities.

Definition at line 96 of file cuda_fe_evaluation.h.

◆ gradient_type

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

using CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::gradient_type = Tensor<1, dim, Number>

An alias for vectorial quantities.

Definition at line 101 of file cuda_fe_evaluation.h.

◆ data_type

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

using CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::data_type = typename MatrixFree<dim, Number>::Data

An alias to kernel specific information.

Definition at line 106 of file cuda_fe_evaluation.h.

Constructor & Destructor Documentation

◆ FEEvaluation()

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::FEEvaluation	(	const unsigned int	cell_id,
		const data_type *	data,
		SharedData< dim, Number > *	shdata
	)

Constructor.

Definition at line 339 of file cuda_fe_evaluation.h.

Member Function Documentation

◆ read_dof_values()

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::read_dof_values ( const Number * src )

For the vector src, read out the values on the degrees of freedom of the current cell, and store them internally. Similar functionality as the function DoFAccessor::get_interpolated_dof_values when no constraints are present, but it also includes constraints from hanging nodes, so once can see it as a similar function to AffineConstraints::read_dof_valuess as well.

Definition at line 366 of file cuda_fe_evaluation.h.

◆ distribute_local_to_global()

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::distribute_local_to_global ( Number * dst ) const

Take the value stored internally on dof values of the current cell and sum them into the vector dst. The function also applies constraints during the write operation. The functionality is hence similar to the function AffineConstraints::distribute_local_to_global.

Definition at line 391 of file cuda_fe_evaluation.h.

◆ evaluate()

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::evaluate	(	const bool	evaluate_val,
		const bool	evaluate_grad
	)

Evaluate the function values and the gradients of the FE function given at the DoF values in the input vector at the quadrature points on the unit cell. The function arguments specify which parts shall actually be computed. This function needs to be called before the functions get_value() or get_gradient() give useful information.

Definition at line 417 of file cuda_fe_evaluation.h.

◆ integrate()

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::integrate	(	const bool	integrate_val,
		const bool	integrate_grad
	)

This function takes the values and/or gradients that are stored on quadrature points, tests them by all the basis functions/gradients on the cell and performs the cell integration. The two function arguments integrate_val and integrate_grad are used to enable/disable some of the values or the gradients.

Definition at line 456 of file cuda_fe_evaluation.h.

◆ get_value() [1/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::value_type FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::get_value ( const unsigned int q_point ) const

Return the value of a finite element function at quadrature point number q_point after a call to evaluate(true,...).

Deprecated:: Use the version without parameters instead.

Definition at line 496 of file cuda_fe_evaluation.h.

◆ get_value() [2/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::value_type FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::get_value

Same as above, except that the quadrature point is computed from thread id.

Definition at line 514 of file cuda_fe_evaluation.h.

◆ get_dof_value() [1/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::value_type FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::get_dof_value ( const unsigned int dof ) const

Return the value of a finite element function at degree of freedom dof after a call to integrate() or before a call to evaluate().

Deprecated:: Use the version without parameters instead.

Definition at line 533 of file cuda_fe_evaluation.h.

◆ get_dof_value() [2/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::value_type FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::get_dof_value

Same as above, except that the local dof index is computed from the thread id.

Definition at line 551 of file cuda_fe_evaluation.h.

◆ submit_value() [1/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::submit_value	(	const value_type &	val_in,
		const unsigned int	q_point
	)

Write a value to the field containing the values on quadrature points with component q_point. Access to the same fields as through get_value(). This specifies the value which is tested by all basis function on the current cell and integrated over.

Deprecated:: Use the version without parameters instead.

Definition at line 566 of file cuda_fe_evaluation.h.

◆ submit_value() [2/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::submit_value ( const value_type & val_in )

Same as above, except that the quadrature point is computed from the thread id.

Definition at line 580 of file cuda_fe_evaluation.h.

◆ submit_dof_value() [1/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::submit_dof_value	(	const value_type &	val_in,
		const unsigned int	dof
	)

Write a value to the field containing the values for the degree of freedom with index dof after a call to integrate() or before calling evaluate(). Access through the same fields as through get_dof_value().

Deprecated:: Use the version without parameters instead.

Definition at line 595 of file cuda_fe_evaluation.h.

◆ submit_dof_value() [2/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::submit_dof_value ( const value_type & val_in )

Same as above, except that the local dof index is computed from the thread id.

Definition at line 609 of file cuda_fe_evaluation.h.

◆ get_gradient() [1/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::gradient_type FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::get_gradient ( const unsigned int q_point ) const

Return the gradient of a finite element function at quadrature point number q_point after a call to evaluate(...,true).

Deprecated:: Use the version without parameters instead.

Definition at line 628 of file cuda_fe_evaluation.h.

◆ get_gradient() [2/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::gradient_type FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::get_gradient

Same as above, except that the quadrature point is computed from the thread id.

Definition at line 660 of file cuda_fe_evaluation.h.

◆ submit_gradient() [1/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::submit_gradient	(	const gradient_type &	grad_in,
		const unsigned int	q_point
	)

Write a contribution that is tested by the gradient to the field containing the values on quadrature points with component q_point.

Deprecated:: Use the version without parameters instead.

Definition at line 690 of file cuda_fe_evaluation.h.

◆ submit_gradient() [2/2]

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::submit_gradient ( const gradient_type & grad_in )

Same as above, except that the quadrature point is computed from the thread id.

Definition at line 713 of file cuda_fe_evaluation.h.

◆ apply_quad_point_operations()

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

template<typename Functor >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::apply_quad_point_operations ( const Functor & func )

Apply the functor func on every quadrature point.

func needs to define

__device__ void operator()(
  CUDAWrappers::FEEvaluation<dim, fe_degree, n_q_points_1d, n_components, Number> *fe_eval,
  const unsigned int                                                               q_point) const;

Deprecated:: Use apply_for_each_quad_point() instead.

Definition at line 738 of file cuda_fe_evaluation.h.

◆ apply_for_each_quad_point()

template<int dim, int fe_degree, int n_q_points_1d, int n_components_, typename Number >

template<typename Functor >

void FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::apply_for_each_quad_point ( const Functor & func )

Same as above, except that the functor func only takes a single input argument (fe_eval) and computes the quadrature point from the thread id.

func needs to define

__device__ void operator()(

CUDAWrappers::FEEvaluation<dim, fe_degree, n_q_points_1d, n_components, Number> *fe_eval) const;

Definition at line 755 of file cuda_fe_evaluation.h.

Member Data Documentation

◆ dimension

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

constexpr unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::dimension = dim

staticconstexpr

Dimension.

Definition at line 111 of file cuda_fe_evaluation.h.

◆ n_components

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

constexpr unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::n_components = n_components_

staticconstexpr

Number of components.

Definition at line 116 of file cuda_fe_evaluation.h.

◆ n_q_points

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

constexpr unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::n_q_points

staticconstexpr

Initial value:

=

Utilities::pow(n_q_points_1d, dim)

Utilities::pow

constexpr T pow(const T base, const int iexp)

Definition: utilities.h:461

Number of quadrature points per cell.

Definition at line 121 of file cuda_fe_evaluation.h.

◆ tensor_dofs_per_cell

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

constexpr unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::tensor_dofs_per_cell

staticconstexpr

Initial value:

=

Utilities::pow(fe_degree + 1, dim)

Number of tensor degrees of freedoms per cell.

Definition at line 127 of file cuda_fe_evaluation.h.

◆ local_to_global

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

types::global_dof_index* CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::local_to_global

private

Definition at line 314 of file cuda_fe_evaluation.h.

◆ n_cells

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::n_cells

private

Definition at line 315 of file cuda_fe_evaluation.h.

◆ padding_length

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::padding_length

private

Definition at line 316 of file cuda_fe_evaluation.h.

◆ mf_object_id

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

const unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::mf_object_id

private

Definition at line 317 of file cuda_fe_evaluation.h.

◆ constraint_mask

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

const unsigned int CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::constraint_mask

private

Definition at line 319 of file cuda_fe_evaluation.h.

◆ use_coloring

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

const bool CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::use_coloring

private

Definition at line 321 of file cuda_fe_evaluation.h.

◆ inv_jac

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

Number* CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::inv_jac

private

Definition at line 323 of file cuda_fe_evaluation.h.

◆ JxW

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

Number* CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::JxW

private

Definition at line 324 of file cuda_fe_evaluation.h.

◆ values

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

Number* CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::values

private

Definition at line 327 of file cuda_fe_evaluation.h.

◆ gradients

template<int dim, int fe_degree, int n_q_points_1d = fe_degree + 1, int n_components_ = 1, typename Number = double>

Number* CUDAWrappers::FEEvaluation< dim, fe_degree, n_q_points_1d, n_components_, Number >::gradients[dim]

private

Definition at line 328 of file cuda_fe_evaluation.h.

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

include/deal.II/matrix_free/cuda_fe_evaluation.h

Public Types

Public Member Functions

Static Public Attributes

Private Attributes

Detailed Description

Member Typedef Documentation

◆ value_type

◆ gradient_type

◆ data_type

Constructor & Destructor Documentation

◆ FEEvaluation()

Member Function Documentation

◆ read_dof_values()

◆ distribute_local_to_global()

◆ evaluate()

◆ integrate()

◆ get_value() [1/2]

◆ get_value() [2/2]

◆ get_dof_value() [1/2]

◆ get_dof_value() [2/2]

◆ submit_value() [1/2]

◆ submit_value() [2/2]

◆ submit_dof_value() [1/2]

◆ submit_dof_value() [2/2]

◆ get_gradient() [1/2]

◆ get_gradient() [2/2]

◆ submit_gradient() [1/2]

◆ submit_gradient() [2/2]

◆ apply_quad_point_operations()

◆ apply_for_each_quad_point()

Member Data Documentation

◆ dimension

◆ n_components

◆ n_q_points

◆ tensor_dofs_per_cell

◆ local_to_global

◆ n_cells

◆ padding_length

◆ mf_object_id

◆ constraint_mask

◆ use_coloring

◆ inv_jac

◆ JxW

◆ values

◆ gradients