#include <transfer.h>

Collaboration diagram for quda::Transfer:

[legend]

Public Member Functions
	Transfer (const std::vector< ColorSpinorField > &B, int Nvec, int NblockOrtho, int geo_bs, int spin_bs, QudaPrecision null_precision, TimeProfile &profile)

virtual	~Transfer ()

void	reset ()
	for resetting the Transfer when the null vectors have changed More...

void	P (ColorSpinorField &out, const ColorSpinorField &in) const

void	R (ColorSpinorField &out, const ColorSpinorField &in) const

QudaPrecision	NullPrecision (QudaFieldLocation location) const
	The precision of the packed null-space vectors. More...

const ColorSpinorField &	Vectors (QudaFieldLocation location=QUDA_INVALID_FIELD_LOCATION) const

int	nvec () const

int	Spin_bs () const

const int *	Geo_bs () const

const int *	fineToCoarse (QudaFieldLocation location=QUDA_CPU_FIELD_LOCATION) const

const int *	coarseToFine (QudaFieldLocation location=QUDA_CPU_FIELD_LOCATION) const

void	setTransferGPU (bool use_gpu) const

void	setSiteSubset (QudaSiteSubset site_subset, QudaParity parity)
	Sets whether the transfer operator is to act on full fields or single parity fields, and if single-parity which parity. More...

double	flops () const

Private Member Functions
void	createV (QudaFieldLocation location) const
	Allocate V field. More...

void	createTmp (QudaFieldLocation location) const
	Allocate temporaries used when applying transfer operators. More...

void	createGeoMap (int *geo_bs)
	Creates the map between fine and coarse grids. More...

void	createSpinMap (int spin_bs)
	Creates the map between fine spin and parity to coarse spin dimensions. More...

void	initializeLazy (QudaFieldLocation location) const
	Lazy allocation of the transfer operator in a given location. More...

Private Attributes
const std::vector< ColorSpinorField * > &	B

const int	Nvec

const int	NblockOrtho

const QudaPrecision	null_precision

ColorSpinorField *	V_h

ColorSpinorField *	V_d

ColorSpinorField *	fine_tmp_h

ColorSpinorField *	fine_tmp_d

ColorSpinorField *	coarse_tmp_h

ColorSpinorField *	coarse_tmp_d

int *	geo_bs

int *	fine_to_coarse_h

int *	coarse_to_fine_h

int *	fine_to_coarse_d

int *	coarse_to_fine_d

int	spin_bs

int **	spin_map

const int	nspin_fine

QudaSiteSubset	site_subset

QudaParity	parity

bool	enable_gpu

bool	enable_cpu

bool	use_gpu

double	flops_

TimeProfile &	profile

Detailed Description

The transfer class defines the inter-grid operators that connect fine and coarse grids. This implements both restriction and prologation methods. The transfer operator is fully defined at object creation time, and defined by the null-space vectors and the coarsening pattern that are passed to it.

At present only the restriction (R) and prolongation (P) methods have been offloaded to run on the GPU, with the block orthogonlization yet to be offloaded.

Definition at line 29 of file transfer.h.

Constructor & Destructor Documentation

◆ Transfer()

quda::Transfer::Transfer	(	const std::vector< ColorSpinorField *> &	B,
		int	Nvec,
		int	NblockOrtho,
		int *	geo_bs,
		int	spin_bs,
		QudaPrecision	null_precision,
		TimeProfile &	profile
	)

The constructor for Transfer

Parameters

B	Array of null-space vectors
Nvec	Number of null-space vectors
NblockOrtho	Number of times to Gram-Schmidt within block ortho
d	The Dirac operator to which these null-space vectors correspond
geo_bs	The geometric block sizes to use
spin_bs	The spin block sizes to use
parity	For single-parity fields are these QUDA_EVEN_PARITY or QUDA_ODD_PARITY
null_precision	The precision to store the null-space basis vectors in
enable_gpu	Whether to enable this to run on GPU (as well as CPU)

Definition at line 19 of file transfer.cpp.

References coarse_to_fine_d, coarse_to_fine_h, createGeoMap(), createSpinMap(), createTmp(), createV(), enable_gpu, errorQuda, fine_to_coarse_d, fine_to_coarse_h, getVerbosity(), ndim, nspin_fine, pool_device_malloc, pool_pinned_malloc, postTrace, printfQuda, QUDA_CPU_FIELD_LOCATION, QUDA_VERBOSE, reset(), quda::s, safe_malloc, spin_map, warningQuda, and X.

Here is the call graph for this function:

◆ ~Transfer()

quda::Transfer::~Transfer ( )

virtual

The destructor for Transfer

Definition at line 205 of file transfer.cpp.

References coarse_tmp_d, coarse_tmp_h, coarse_to_fine_d, coarse_to_fine_h, fine_tmp_d, fine_tmp_h, fine_to_coarse_d, fine_to_coarse_h, geo_bs, host_free, nspin_fine, pool_device_free, pool_pinned_free, quda::s, spin_map, V_d, and V_h.

Member Function Documentation

◆ coarseToFine()

const int* quda::Transfer::coarseToFine ( QudaFieldLocation location = QUDA_CPU_FIELD_LOCATION ) const

inline

Returns: Pointer to the lookup table to the coarse-to-fine map

Definition at line 241 of file transfer.h.

References coarse_to_fine_d, coarse_to_fine_h, and QUDA_CPU_FIELD_LOCATION.

◆ createGeoMap()

void quda::Transfer::createGeoMap ( int * geo_bs )

private

Creates the map between fine and coarse grids.

Parameters

geo_bs An array storing the block size in each geometric dimension

Definition at line 248 of file transfer.cpp.

References B, checkCudaError, coarse_tmp_h, coarse_to_fine_d, coarse_to_fine_h, enable_gpu, fine_tmp_h, fine_to_coarse_d, fine_to_coarse_h, quda::ColorSpinorField::LatticeIndex(), quda::ColorSpinorField::Ndim(), quda::ColorSpinorField::OffsetIndex(), QUDA_MAX_DIM, qudaMemcpy, and quda::ColorSpinorField::Volume().

Referenced by Transfer().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ createSpinMap()

void quda::Transfer::createSpinMap ( int spin_bs )

private

Creates the map between fine spin and parity to coarse spin dimensions.

Parameters

spin_bs The spin block size

Definition at line 289 of file transfer.cpp.

References B, quda::s, spin_bs, and spin_map.

Referenced by Transfer().

Here is the caller graph for this function:

◆ createTmp()

void quda::Transfer::createTmp ( QudaFieldLocation location ) const

private

Allocate temporaries used when applying transfer operators.

Parameters

[in] location Where to allocate the temporaries

Definition at line 134 of file transfer.cpp.

References B, coarse_tmp_d, coarse_tmp_h, quda::ColorSpinorParam::create, quda::ColorSpinorField::Create(), quda::ColorSpinorField::CreateCoarse(), quda::ColorSpinorParam::fieldOrder, fine_tmp_d, fine_tmp_h, geo_bs, quda::ColorSpinorParam::location, Nvec, param, postTrace, quda::LatticeFieldParam::Precision(), QUDA_CUDA_FIELD_LOCATION, QUDA_FLOAT2_FIELD_ORDER, QUDA_NULL_FIELD_CREATE, QUDA_SINGLE_PRECISION, QUDA_SPACE_SPIN_COLOR_FIELD_ORDER, quda::ColorSpinorParam::setPrecision(), and spin_bs.

Referenced by initializeLazy(), and Transfer().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ createV()

void quda::Transfer::createV ( QudaFieldLocation location ) const

private

Allocate V field.

Parameters

[in] location Where to allocate the V field

Definition at line 101 of file transfer.cpp.

References B, quda::ColorSpinorParam::create, quda::ColorSpinorField::Create(), enable_cpu, enable_gpu, quda::ColorSpinorParam::fieldOrder, quda::ColorSpinorParam::location, quda::ColorSpinorParam::nColor, quda::ColorSpinorParam::nSpin, null_precision, Nvec, quda::ColorSpinorParam::nVec, param, postTrace, QUDA_CUDA_FIELD_LOCATION, QUDA_FLOAT2_FIELD_ORDER, QUDA_FULL_SITE_SUBSET, QUDA_NULL_FIELD_CREATE, QUDA_PARITY_SITE_SUBSET, QUDA_SPACE_SPIN_COLOR_FIELD_ORDER, quda::ColorSpinorParam::setPrecision(), quda::LatticeFieldParam::siteSubset, V_d, V_h, and quda::LatticeFieldParam::x.

Referenced by initializeLazy(), and Transfer().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ fineToCoarse()

const int* quda::Transfer::fineToCoarse ( QudaFieldLocation location = QUDA_CPU_FIELD_LOCATION ) const

inline

Returns: Pointer to the lookup table to the fine-to-coarse map

Definition at line 235 of file transfer.h.

References fine_to_coarse_d, fine_to_coarse_h, and QUDA_CPU_FIELD_LOCATION.

◆ flops()

double quda::Transfer::flops ( ) const

Return flops

Returns: flops expended by this operator

Definition at line 387 of file transfer.cpp.

References flops_.

Referenced by quda::MG::flops(), and setTransferGPU().

Here is the caller graph for this function:

◆ Geo_bs()

const int* quda::Transfer::Geo_bs ( ) const

inline

Returns the geometrical coarse grid blocking

Returns: geo_bs

Definition at line 230 of file transfer.h.

References geo_bs.

Referenced by quda::DiracCoarse::createY(), and quda::DiracCoarse::createYhat().

Here is the caller graph for this function:

◆ initializeLazy()

void quda::Transfer::initializeLazy ( QudaFieldLocation location ) const

private

Lazy allocation of the transfer operator in a given location.

Parameters

[in] location Where to allocate the temporaries

Definition at line 154 of file transfer.cpp.

References B, coarse_to_fine_d, coarse_to_fine_h, createTmp(), createV(), enable_cpu, enable_gpu, errorQuda, fine_to_coarse_d, fine_to_coarse_h, pool_device_malloc, QUDA_CPU_FIELD_LOCATION, QUDA_CUDA_FIELD_LOCATION, qudaMemcpy, V_d, and V_h.

Referenced by P(), and R().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ NullPrecision()

QudaPrecision quda::Transfer::NullPrecision ( QudaFieldLocation location ) const

inline

The precision of the packed null-space vectors.

Definition at line 195 of file transfer.h.

References null_precision, QUDA_CUDA_FIELD_LOCATION, and QUDA_SINGLE_PRECISION.

Referenced by quda::DiracCoarse::createY(), and quda::DiracCoarse::createYhat().

Here is the caller graph for this function:

◆ nvec()

int quda::Transfer::nvec ( ) const

inline

Returns the number of near nullvectors

Returns: Nvec

Definition at line 218 of file transfer.h.

References Nvec.

Referenced by quda::DiracCoarse::createY(), and quda::DiracCoarse::createYhat().

Here is the caller graph for this function:

◆ P()

void quda::Transfer::P	(	ColorSpinorField &	out,
		const ColorSpinorField &	in
	)		const

Apply the prolongator

Parameters

out	The resulting field on the fine lattice
in	The input field on the coarse lattice

Definition at line 305 of file transfer.cpp.

References coarse_tmp_d, enable_gpu, errorQuda, quda::ColorSpinorField::Even(), fine_tmp_d, fine_tmp_h, fine_to_coarse_d, fine_to_coarse_h, flops_, quda::ColorSpinorField::GammaBasis(), in, initializeLazy(), quda::LatticeField::Location(), quda::ColorSpinorField::Ncolor(), quda::ColorSpinorField::Nspin(), Nvec, out, parity, profile, quda::Prolongate(), QUDA_CPU_FIELD_LOCATION, QUDA_CUDA_FIELD_LOCATION, QUDA_FULL_SITE_SUBSET, QUDA_PARITY_SITE_SUBSET, quda::QUDA_PROFILE_COMPUTE, quda::ColorSpinorField::SiteSubset(), spin_map, use_gpu, V, V_d, V_h, and quda::ColorSpinorField::VolumeCB().

Referenced by quda::MG::operator()(), and quda::MG::verify().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ R()

void quda::Transfer::R	(	ColorSpinorField &	out,
		const ColorSpinorField &	in
	)		const

Apply the restrictor

Parameters

out	The resulting field on the coarse lattice
in	The input field on the fine lattice

Definition at line 344 of file transfer.cpp.

References coarse_tmp_d, coarse_to_fine_d, coarse_to_fine_h, enable_gpu, errorQuda, quda::ColorSpinorField::Even(), fine_tmp_d, fine_tmp_h, fine_to_coarse_d, fine_to_coarse_h, flops_, quda::ColorSpinorField::GammaBasis(), in, initializeLazy(), quda::LatticeField::Location(), quda::ColorSpinorField::Ncolor(), quda::ColorSpinorField::Nspin(), Nvec, out, parity, profile, QUDA_CPU_FIELD_LOCATION, QUDA_CUDA_FIELD_LOCATION, QUDA_FULL_SITE_SUBSET, QUDA_PARITY_SITE_SUBSET, quda::QUDA_PROFILE_COMPUTE, qudaDeviceSynchronize, quda::Restrict(), quda::ColorSpinorField::SiteSubset(), spin_map, use_gpu, V, V_d, V_h, and quda::ColorSpinorField::VolumeCB().

Referenced by quda::MG::generateNullVectors(), quda::MG::operator()(), quda::MG::reset(), and quda::MG::verify().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ reset()

void quda::Transfer::reset ( )

for resetting the Transfer when the null vectors have changed

Definition at line 182 of file transfer.cpp.

References B, quda::BlockOrthogonalize(), coarse_to_fine_d, coarse_to_fine_h, enable_cpu, enable_gpu, errorQuda, fine_to_coarse_d, fine_to_coarse_h, geo_bs, getVerbosity(), NblockOrtho, postTrace, printfQuda, QUDA_CUDA_FIELD_LOCATION, QUDA_VERBOSE, spin_bs, V_d, and V_h.

Referenced by quda::MG::reset(), and Transfer().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ setSiteSubset()

void quda::Transfer::setSiteSubset	(	QudaSiteSubset	site_subset,
		QudaParity	parity
	)

Sets whether the transfer operator is to act on full fields or single parity fields, and if single-parity which parity.

Parameters

[in]	site_subset	The site_subset of the fine-grid fields
[in]	parity	The parity of the single-parity fields (if applicable)

Definition at line 227 of file transfer.cpp.

References errorQuda, parity, QUDA_EVEN_PARITY, QUDA_ODD_PARITY, QUDA_PARITY_SITE_SUBSET, and site_subset.

Referenced by quda::MG::operator()(), quda::MG::reset(), and setTransferGPU().

Here is the caller graph for this function:

◆ setTransferGPU()

void quda::Transfer::setTransferGPU ( bool use_gpu ) const

inline

Sets where the prolongator / restrictor should take place

Parameters

location Location where the transfer operator should be computed

Definition at line 248 of file transfer.h.

References B, quda::BlockOrthogonalize(), flops(), geo_bs, in, n_block_ortho, Nvec, out, parity, quda::Prolongate(), QUDA_INVALID_PARITY, quda::Restrict(), setSiteSubset(), spin_bs, spin_map, use_gpu, and V.

Here is the call graph for this function:

◆ Spin_bs()

int quda::Transfer::Spin_bs ( ) const

inline

Returns the amount of spin blocking. Defined as zero when coarsening staggered.

Returns: spin_bs

Definition at line 224 of file transfer.h.

References spin_bs.

Referenced by quda::DiracCoarse::createY(), and quda::DiracCoarse::createYhat().

Here is the caller graph for this function:

◆ Vectors()

const ColorSpinorField& quda::Transfer::Vectors ( QudaFieldLocation location = QUDA_INVALID_FIELD_LOCATION ) const

inline

Returns a const reference to the V field

Parameters

location Which memory space are we requesting

Returns: The V field const reference

Definition at line 205 of file transfer.h.

References QUDA_CUDA_FIELD_LOCATION, QUDA_INVALID_FIELD_LOCATION, V_d, and V_h.

Referenced by quda::CoarseCoarseOp(), quda::CoarseOp(), quda::DiracWilson::createCoarseOp(), quda::DiracClover::createCoarseOp(), quda::DiracCloverPC::createCoarseOp(), quda::DiracTwistedClover::createCoarseOp(), quda::DiracTwistedCloverPC::createCoarseOp(), quda::DiracCoarse::createCoarseOp(), quda::DiracCoarsePC::createCoarseOp(), quda::DiracCoarse::createY(), quda::DiracCoarse::createYhat(), and quda::MG::verify().

Here is the caller graph for this function:

Member Data Documentation

◆ B

const std::vector<ColorSpinorField*>& quda::Transfer::B

private

The raw null space components

Definition at line 34 of file transfer.h.

Referenced by createGeoMap(), createSpinMap(), createTmp(), createV(), initializeLazy(), reset(), and setTransferGPU().

◆ coarse_tmp_d

ColorSpinorField* quda::Transfer::coarse_tmp_d

mutableprivate

A GPU temporary field with coarse geometry and coarse color we use for CPU input / output

Definition at line 61 of file transfer.h.

Referenced by createTmp(), P(), R(), and ~Transfer().

◆ coarse_tmp_h

ColorSpinorField* quda::Transfer::coarse_tmp_h

mutableprivate

A CPU temporary field with coarse geometry and coarse color

Definition at line 58 of file transfer.h.

Referenced by createGeoMap(), createTmp(), and ~Transfer().

◆ coarse_to_fine_d

int* quda::Transfer::coarse_to_fine_d

mutableprivate

The mapping onto fine sites from coarse sites. This has length equal to the fine-grid volume, and is sorted into lexicographical block order, with each value corresponding to a fine-grid offset. (GPU)

Definition at line 86 of file transfer.h.

Referenced by coarseToFine(), createGeoMap(), initializeLazy(), R(), reset(), Transfer(), and ~Transfer().

◆ coarse_to_fine_h

int* quda::Transfer::coarse_to_fine_h

mutableprivate

The mapping onto fine sites from coarse sites. This has length equal to the fine-grid volume, and is sorted into lexicographical block order, with each value corresponding to a fine-grid offset. (CPU)

Definition at line 75 of file transfer.h.

Referenced by coarseToFine(), createGeoMap(), initializeLazy(), R(), reset(), Transfer(), and ~Transfer().

◆ enable_cpu

bool quda::Transfer::enable_cpu

mutableprivate

Whether the CPU transfer operator has been constructed

Definition at line 107 of file transfer.h.

Referenced by createV(), initializeLazy(), and reset().

◆ enable_gpu

bool quda::Transfer::enable_gpu

mutableprivate

Whether the GPU transfer operator has been constructed

Definition at line 104 of file transfer.h.

Referenced by createGeoMap(), createV(), initializeLazy(), P(), R(), reset(), and Transfer().

◆ fine_tmp_d

ColorSpinorField* quda::Transfer::fine_tmp_d

mutableprivate

A GPU temporary field with fine geometry and fine color we use for changing gamma basis

Definition at line 55 of file transfer.h.

Referenced by createTmp(), P(), R(), and ~Transfer().

◆ fine_tmp_h

ColorSpinorField* quda::Transfer::fine_tmp_h

mutableprivate

A CPU temporary field with fine geometry and fine color we use for changing gamma basis

Definition at line 52 of file transfer.h.

Referenced by createGeoMap(), createTmp(), P(), R(), and ~Transfer().

◆ fine_to_coarse_d

int* quda::Transfer::fine_to_coarse_d

mutableprivate

The mapping onto coarse sites from fine sites. This has length equal to the fine-grid volume, and is sorted into lexicographical fine-grid order, with each value corresponding to a coarse-grid offset. (GPU)

Definition at line 81 of file transfer.h.

Referenced by createGeoMap(), fineToCoarse(), initializeLazy(), P(), R(), reset(), Transfer(), and ~Transfer().

◆ fine_to_coarse_h

int* quda::Transfer::fine_to_coarse_h

mutableprivate

The mapping onto coarse sites from fine sites. This has length equal to the fine-grid volume, and is sorted into lexicographical fine-grid order, with each value corresponding to a coarse-grid offset. (CPU)

Definition at line 70 of file transfer.h.

Referenced by createGeoMap(), fineToCoarse(), initializeLazy(), P(), R(), reset(), Transfer(), and ~Transfer().

◆ flops_

double quda::Transfer::flops_

mutableprivate

Internal flops accumulator

Definition at line 146 of file transfer.h.

Referenced by flops(), P(), and R().

◆ geo_bs

int* quda::Transfer::geo_bs

private

The geometrical coase grid blocking

Definition at line 64 of file transfer.h.

Referenced by createTmp(), Geo_bs(), reset(), setTransferGPU(), and ~Transfer().

◆ NblockOrtho

const int quda::Transfer::NblockOrtho

private

The number of times to Gram-Schmidt within block ortho

Definition at line 40 of file transfer.h.

Referenced by reset().

◆ nspin_fine

const int quda::Transfer::nspin_fine

private

Nspin for the fine level. Required for deallocating spin_map.

Definition at line 95 of file transfer.h.

Referenced by Transfer(), and ~Transfer().

◆ null_precision

const QudaPrecision quda::Transfer::null_precision

private

Precision to use for the GPU null-space components

Definition at line 43 of file transfer.h.

Referenced by createV(), and NullPrecision().

◆ Nvec

const int quda::Transfer::Nvec

private

The number of null space components

Definition at line 37 of file transfer.h.

Referenced by createTmp(), createV(), nvec(), P(), R(), and setTransferGPU().

◆ parity

QudaParity quda::Transfer::parity

private

The parity of any single-parity fine-grid fields that are passed into the transfer operator

Definition at line 101 of file transfer.h.

Referenced by P(), R(), setSiteSubset(), and setTransferGPU().

◆ profile

TimeProfile& quda::Transfer::profile

private

Reference to profile kept in the corresponding MG instance. Use this to record restriction and prolongation overhead.

Definition at line 152 of file transfer.h.

Referenced by P(), and R().

◆ site_subset

QudaSiteSubset quda::Transfer::site_subset

private

Whether the transfer operator is to be applied to full fields or single parity fields

Definition at line 98 of file transfer.h.

Referenced by setSiteSubset().

◆ spin_bs

int quda::Transfer::spin_bs

private

The spin blocking. Defined as zero when the fine operator is staggered.

Definition at line 89 of file transfer.h.

Referenced by createSpinMap(), createTmp(), reset(), setTransferGPU(), and Spin_bs().

◆ spin_map

int** quda::Transfer::spin_map

private

The mapping onto coarse spin from fine spin (inner) and fine parity (outer), for staggered

Definition at line 92 of file transfer.h.

Referenced by createSpinMap(), P(), R(), setTransferGPU(), Transfer(), and ~Transfer().

◆ use_gpu

bool quda::Transfer::use_gpu

mutableprivate

Whether to apply the transfer operaton the GPU (requires enable_gpu=true in the constructor)

Definition at line 111 of file transfer.h.

Referenced by P(), R(), and setTransferGPU().

◆ V_d

ColorSpinorField* quda::Transfer::V_d

mutableprivate

GPU copy of the block-normalized null-space components that define the prolongator

Definition at line 49 of file transfer.h.

Referenced by createV(), initializeLazy(), P(), R(), reset(), Vectors(), and ~Transfer().

◆ V_h

ColorSpinorField* quda::Transfer::V_h

mutableprivate

CPU copy of the block-normalized null-space components that define the prolongator

Definition at line 46 of file transfer.h.

Referenced by createV(), initializeLazy(), P(), R(), reset(), Vectors(), and ~Transfer().

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

include/transfer.h
lib/transfer.cpp

Public Member Functions

Private Member Functions

Private Attributes

Detailed Description

Constructor & Destructor Documentation

◆ Transfer()

◆ ~Transfer()

Member Function Documentation

◆ coarseToFine()

◆ createGeoMap()

◆ createSpinMap()

◆ createTmp()

◆ createV()

◆ fineToCoarse()

◆ flops()

◆ Geo_bs()

◆ initializeLazy()

◆ NullPrecision()

◆ nvec()

◆ P()

◆ R()

◆ reset()

◆ setSiteSubset()

◆ setTransferGPU()

◆ Spin_bs()

◆ Vectors()

Member Data Documentation

◆ B

◆ coarse_tmp_d

◆ coarse_tmp_h

◆ coarse_to_fine_d

◆ coarse_to_fine_h

◆ enable_cpu

◆ enable_gpu

◆ fine_tmp_d

◆ fine_tmp_h

◆ fine_to_coarse_d

◆ fine_to_coarse_h

◆ flops_

◆ geo_bs

◆ NblockOrtho

◆ nspin_fine

◆ null_precision

◆ Nvec

◆ parity

◆ profile

◆ site_subset

◆ spin_bs

◆ spin_map

◆ use_gpu

◆ V_d

◆ V_h