Plasma Class Reference

#include <plasma.hpp>

Collaboration diagram for Plasma:

Classes
struct	DecomposedRecalculableF0Arrays

Public Types
enum	DevicePtlOpt { UseDevicePtl =0, NoDevicePtl }
enum	ParticleType { Electrons =0, Ions }

Public Member Functions
	Plasma ()
	Plasma (NLReader::NamelistReader &nlr, bool use_f0_grid, const Grid< DeviceType > &grid, const MagneticField< DeviceType > &magnetic_field, const DomainDecomposition< DeviceType > &pol_decomp, const VelocityGrid &vgrid)
template<typename F >
void	for_all_species (F func, DevicePtlOpt device_ptl_opt=UseDevicePtl)
template<typename F >
void	for_all_nonadiabatic_species (F func, DevicePtlOpt device_ptl_opt=UseDevicePtl)
template<typename F >
void	for_electrons (F func, DevicePtlOpt device_ptl_opt=UseDevicePtl)
template<typename F >
void	for_all_ions (F func, DevicePtlOpt device_ptl_opt=UseDevicePtl)
template<typename F >
void	for_all (ParticleType particle_type, F func, DevicePtlOpt device_ptl_opt=UseDevicePtl)
template<typename F >
void	for_one_species (int isp, F func, DevicePtlOpt device_ptl_opt=UseDevicePtl)
int	largest_n_ptl (bool check_backup)
void	deallocate_device_ptl ()
void	update_decomposed_f0_calculations (const DomainDecomposition< DeviceType > &pol_decomp, const Grid< DeviceType > &grid, const MagneticField< DeviceType > &magnetic_field, const VelocityGrid &vgrid)
void	resize_f0_f0g (const DomainDecomposition< DeviceType > &pol_decomp, const VelocityGrid &vgrid)

Static Public Member Functions
static std::vector < MemoryPrediction >	estimate_memory_usage (NLReader::NamelistReader &nlr, const Grid< DeviceType > &grid, const DomainDecomposition< DeviceType > &pol_decomp)

Public Attributes
std::vector< Species < DeviceType > >	all_species
	Every particle species in the simulation. More...
VGridDistribution< HostType >	f0_f0g
DecomposedRecalculableF0Arrays	decomposed_recalculable_f0_arrays
	Contains f0 values that are poloidally decomposed but don't need to be transferred between ranks during load rebalancing. More...
Kokkos::View< double **, Kokkos::LayoutRight, HostType >	f0_delta_n
	Flux-surface averaged change of density. More...
Kokkos::View< double **, Kokkos::LayoutRight, HostType >	f0_delta_u
	Flux-surface averaged change of parallel flow. More...
Kokkos::View< double **, Kokkos::LayoutRight, HostType >	f0_delta_T
	Flux-surface averaged change of temperature. More...
View< double **, CLayout, HostType >	f0_den_global
	Equilibrium density at vertices. More...
View< double **, CLayout, HostType >	f0_temp_global
	Equilibrium temperature at vertices. More...
int	nspecies
	Number of species including electrons. More...
int	n_nonadiabatic_species
	Number of nonadiabatic species. More...

Private Member Functions
void	init_global_f0_arrays (int grid_nnode)
template<typename T_in , typename T_out >
void	set_unmanaged_f0_species_view (const T_in &view_in, int isp, T_out &view_out)
void	transfer_particles_d_ownership (int isp)
void	manage_particle_ownership (int isp, DevicePtlOpt device_ptl_opt)

Private Attributes
bool	species_share_particles_d_ownership
	Whether to use the device particles sharing scheme. More...
bool	particles_d_has_owner
	Whether a species owns the device particles allocation. More...
int	particles_d_owner
	Which species, if any, owns the device particles allocation. More...

Member Enumeration Documentation

enum Plasma::DevicePtlOpt

Enumerator
UseDevicePtl
NoDevicePtl

enum Plasma::ParticleType

Enumerator
Electrons
Ions

Constructor & Destructor Documentation

Plasma::Plasma ( )

inline

Plasma::Plasma	(	NLReader::NamelistReader &	nlr,
		bool	use_f0_grid,
		const Grid< DeviceType > &	grid,
		const MagneticField< DeviceType > &	magnetic_field,
		const DomainDecomposition< DeviceType > &	pol_decomp,
		const VelocityGrid &	vgrid
	)

Here is the call graph for this function:

Member Function Documentation

void Plasma::deallocate_device_ptl ( )

inline

Here is the caller graph for this function:

std::vector< MemoryPrediction > Plasma::estimate_memory_usage ( NLReader::NamelistReader &  nlr, const Grid< DeviceType > &  grid, const DomainDecomposition< DeviceType > &  pol_decomp  )

static

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename F >

void Plasma::for_all ( ParticleType  particle_type, F  func, DevicePtlOpt  device_ptl_opt = UseDevicePtl  )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename F >

void Plasma::for_all_ions ( F  func, DevicePtlOpt  device_ptl_opt = UseDevicePtl  )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename F >

void Plasma::for_all_nonadiabatic_species ( F  func, DevicePtlOpt  device_ptl_opt = UseDevicePtl  )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename F >

void Plasma::for_all_species ( F  func, DevicePtlOpt  device_ptl_opt = UseDevicePtl  )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename F >

void Plasma::for_electrons ( F  func, DevicePtlOpt  device_ptl_opt = UseDevicePtl  )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename F >

void Plasma::for_one_species ( int  isp, F  func, DevicePtlOpt  device_ptl_opt = UseDevicePtl  )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

void Plasma::init_global_f0_arrays ( int  grid_nnode )

inlineprivate

Here is the call graph for this function:

Here is the caller graph for this function:

int Plasma::largest_n_ptl ( bool  check_backup )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

void Plasma::manage_particle_ownership ( int  isp, DevicePtlOpt  device_ptl_opt  )

inlineprivate

Here is the caller graph for this function:

void Plasma::resize_f0_f0g ( const DomainDecomposition< DeviceType > &  pol_decomp, const VelocityGrid &  vgrid  )

inline

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename T_in , typename T_out >

void Plasma::set_unmanaged_f0_species_view ( const T_in &  view_in, int  isp, T_out &  view_out  )

inlineprivate

Here is the call graph for this function:

Here is the caller graph for this function:

void Plasma::transfer_particles_d_ownership ( int  isp )

inlineprivate

void Plasma::update_decomposed_f0_calculations	(	const DomainDecomposition< DeviceType > &	pol_decomp,
		const Grid< DeviceType > &	grid,
		const MagneticField< DeviceType > &	magnetic_field,
		const VelocityGrid &	vgrid
	)

Here is the call graph for this function:

Here is the caller graph for this function:

Member Data Documentation

std::vector<Species<DeviceType> > Plasma::all_species

Every particle species in the simulation.

DecomposedRecalculableF0Arrays Plasma::decomposed_recalculable_f0_arrays

Contains f0 values that are poloidally decomposed but don't need to be transferred between ranks during load rebalancing.

Kokkos::View<double**,Kokkos::LayoutRight, HostType> Plasma::f0_delta_n

Flux-surface averaged change of density.

Kokkos::View<double**,Kokkos::LayoutRight, HostType> Plasma::f0_delta_T

Flux-surface averaged change of temperature.

Kokkos::View<double**,Kokkos::LayoutRight, HostType> Plasma::f0_delta_u

Flux-surface averaged change of parallel flow.

View<double**,CLayout, HostType> Plasma::f0_den_global

Equilibrium density at vertices.

VGridDistribution<HostType> Plasma::f0_f0g

View<double**,CLayout, HostType> Plasma::f0_temp_global

Equilibrium temperature at vertices.

int Plasma::n_nonadiabatic_species

Number of nonadiabatic species.

int Plasma::nspecies

Number of species including electrons.

bool Plasma::particles_d_has_owner

private

Whether a species owns the device particles allocation.

int Plasma::particles_d_owner

private

Which species, if any, owns the device particles allocation.

bool Plasma::species_share_particles_d_ownership

private

Whether to use the device particles sharing scheme.

---

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

• /u/gitlab-xgc/builds/YGMz2TJ8/0/xgc/XGC-Devel/XGC_core/cpp/plasma.hpp
• /u/gitlab-xgc/builds/YGMz2TJ8/0/xgc/XGC-Devel/XGC_core/cpp/plasma.cpp