inciter::LohCG class
#include <src/Inciter/LohCG.hpp>

LohCG Charm++ chare array used to advance PDEs in time with LohCG.

Public static functions

static void registerReducers(): Configure Charm++ custom reduction types initiated from this chare array.

Constructors, destructors, conversion operators

LohCG(CkMigrateMessage* m) explicit: Migrate constructor.

Public functions

auto LohCG(const CProxy_Discretization& disc, const tk::CProxy_ConjugateGradients& cgpre, const std::map<int, std::vector<std::size_t>>& bface, const std::map<int, std::vector<std::size_t>>& bnode, const std::vector<std::size_t>& triinpoel) -> LohCG_SDAG_CODE explicit: Constructor.
void ResumeFromSync() override: Return from migration.
void setup(tk::real v): Start setup for solution.
void pinit(): Initialize Poisson solve.
void psolve(): Solve Poisson equation.
void psolved(): Continue after Poisson solve.
void advance(tk::real newdt): Advance equations to next time step.
void diag(): Evaluate diagnostics.
void feop(): Start (re-)computing domain and boundary integrals.
void comnorm(const std::unordered_map<int, std::unordered_map<std::size_t, std::array<tk::real, 4>>>& inbnd): Receive contributions to boundary point normals on chare-boundaries.
void comvgrad(const std::unordered_map<std::size_t, std::vector<tk::real>>& ingrad): Receive contributions to velocity gradients.
void comflux(const std::unordered_map<std::size_t, std::vector<tk::real>>& influx): Receive contributions to momentum flux on chare-boundaries.
void comsgrad(const std::unordered_map<std::size_t, std::vector<tk::real>>& ingrad): Receive contributions to conjugate gradients solution gradient.
void comgrad(const std::unordered_map<std::size_t, std::vector<tk::real>>& ingrad): Receive contributions to gradient.
void comrhs(const std::unordered_map<std::size_t, std::vector<tk::real>>& inrhs): Receive contributions to right-hand side vector on chare-boundaries.
void comdiv(const std::unordered_map<std::size_t, tk::real>& indiv): Receive contributions to velocity divergence on chare-boundaries.
void evalres(const std::vector<tk::real>& l2res): Evaluate residuals.
void resizePostAMR(const std::vector<std::size_t>& ginpoel, const tk::UnsMesh::Chunk& chunk, const tk::UnsMesh::Coords& coord, const std::unordered_map<std::size_t, tk::UnsMesh::Edge>& addedNodes, const std::unordered_map<std::size_t, std::size_t>& addedTets, const std::set<std::size_t>& removedNodes, const std::unordered_map<int, std::unordered_set<std::size_t>>& nodeCommMap, const std::map<int, std::vector<std::size_t>>& bface, const std::map<int, std::vector<std::size_t>>& bnode, const std::vector<std::size_t>& triinpoel): Receive new mesh from Refiner.
auto solution() const -> const tk::Fields&
void integrals(): Compute integral quantities for output.
void sgrad(): Compute recent conjugate gradients solution gradient.
void step(): Evaluate whether to continue with next time step.
void evalLB(int nrestart)

Charm++ pack/unpack serializer member functions

CProxy_Discretization m_disc: Discretization proxy.
tk::CProxy_ConjugateGradients m_cgpre: Conjugate Gradients Charm++ proxy for pressure solve.
std::size_t m_nrhs: Counter for right-hand side vector nodes updated.
std::size_t m_nnorm: Counter for receiving boundary point normals.
std::size_t m_ngrad: Counter for receiving gradient.
std::size_t m_nsgrad: Counter for receiving conjugrate gradient solution gradient.
std::size_t m_nvgrad: Counter for receiving velocity gradient.
std::size_t m_nflux: Counter for receiving momentum flux.
std::size_t m_ndiv: Counter for receiving boundary velocity divergences.
std::size_t m_nbpint: Counter for receiving boundary point integrals.
std::size_t m_np: Count number of Poisson solves during setup.
std::map<int, std::vector<std::size_t>> m_bnode: Boundary node lists mapped to side set ids used in the input file.
std::map<int, std::vector<std::size_t>> m_bface: Boundary face lists mapped to side set ids used in the input file.
std::vector<std::size_t> m_triinpoel: Boundary triangle face connecitivity where BCs are set by user.
tk::Fields m_u: Unknown/solution vector at mesh nodes.
tk::Fields m_un: Unknown/solution vector at mesh nodes at previous time step.
tk::Fields m_grad: Gradient in mesh nodes.
std::unordered_map<std::size_t, std::vector<tk::real>> m_gradc: Gradient receive buffer.
tk::Fields m_vgrad: Velocity gradient in mesh nodes.
std::unordered_map<std::size_t, std::vector<tk::real>> m_vgradc: Velocity gradient receive buffer.
tk::Fields m_flux: Momentum flux in mesh nodes.
std::unordered_map<std::size_t, std::vector<tk::real>> m_fluxc: Momentum flux receive buffer.
std::vector<tk::real> m_div: Velocity divergence.
std::unordered_map<std::size_t, tk::real> m_divc
tk::Fields m_sgrad: Conjugate gradient solution gradient in mesh nodes.
std::unordered_map<std::size_t, std::vector<tk::real>> m_sgradc: Conjugate gradient solution gradient receive buffer.
tk::Fields m_rhs: Right-hand side vector (for the high order system)
std::unordered_map<std::size_t, std::vector<tk::real>> m_rhsc
NodeDiagnostics m_diag: Diagnostics object.
std::unordered_map<int, std::unordered_map<std::size_t, std::array<tk::real, 4>>> m_bnorm
decltype(m_bnorm) m_bnormc
std::unordered_map<std::size_t, std::array<tk::real, 3>> m_bndpoinint
std::unordered_map<tk::UnsMesh::Edge, std::array<tk::real, 4>, tk::UnsMesh::Hash<2>, tk::UnsMesh::Eq<2>> m_domedgeint: Domain edge integrals.
std::vector<tk::real> m_bpint: Streamable boundary point integrals.
std::array<std::vector<std::size_t>, 3> m_dsupedge: Superedge (tet, face, edge) end points with local ids for domain edges.
std::array<std::vector<tk::real>, 3> m_dsupint: Superedge (tet, face, edge) domain edge integrals.
std::vector<std::size_t> m_dirbcmask: Nodes and their Dirichlet BC masks.
std::vector<double> m_dirbcval: Nodes and their Dirichlet BC values.
std::vector<std::size_t> m_dirbcmaskp: Nodes and their pressure Dirichlet BC masks.
std::vector<double> m_dirbcvalp: Nodes and their pressure Dirichlet BC values.
std::vector<std::size_t> m_symbcnodes: Streamable nodes at which symmetry BCs are set.
std::vector<tk::real> m_symbcnorms: Streamable normals at nodes at which symmetry BCs are set.
std::vector<std::size_t> m_noslipbcnodes: Streamable nodes at which noslip BCs are set.
std::map<int, std::pair<std::vector<std::size_t>, std::vector<tk::real>>> m_surfint: Streamable surface integral nodes and normals * dA on surfaces.
std::size_t m_stage: Runge-Kutta stage counter.
int m_finished: True in the last time step.
std::vector<tk::real> m_rkcoef: Runge-Kutta coefficients.
std::vector<tk::Timer> m_timer: Timer.
void pup(PUP::er& p) override: Pack/Unpack serialize member function.
void operator|(PUP::er& p, LohCG& i): Pack/Unpack serialize operator|.

Function documentation

static void inciter::LohCG::registerReducers()

Configure Charm++ custom reduction types initiated from this chare array.

Since this is a [initnode] routine, the runtime system executes the routine exactly once on every logical node early on in the Charm++ init sequence. Must be static as it is called without an object. See also: Section "Initializations at Program Startup" at in the Charm++ manual http://charm.cs.illinois.edu/manuals/html/charm++/manual.html.

LohCG_SDAG_CODE inciter::LohCG::LohCG(const CProxy_Discretization& disc, const tk::CProxy_ConjugateGradients& cgpre, const std::map<int, std::vector<std::size_t>>& bface, const std::map<int, std::vector<std::size_t>>& bnode, const std::vector<std::size_t>& triinpoel) explicit

Constructor.

Parameters
disc in	Discretization proxy
cgpre in	ConjugateGradients Charm++ proxy for initial pressure solve
bface in	Boundary-faces mapped to side sets used in the input file
bnode in	Boundary-node lists mapped to side sets used in input file
triinpoel in	Boundary-face connectivity where BCs set (global ids)

void inciter::LohCG::ResumeFromSync() override

Return from migration.

This is called when load balancing (LB) completes. The presence of this function does not affect whether or not we block on LB.

void inciter::LohCG::setup(tk::real v)

Start setup for solution.

Parameters
v in	Total volume within user-specified box

void inciter::LohCG::advance(tk::real newdt)

Advance equations to next time step.

Parameters
newdt in	The smallest dt across the whole problem

void inciter::LohCG::comnorm(const std::unordered_map<int, std::unordered_map<std::size_t, std::array<tk::real, 4>>>& inbnd)

Receive contributions to boundary point normals on chare-boundaries.

Parameters
inbnd in	Incoming partial sums of boundary point normals

void inciter::LohCG::comvgrad(const std::unordered_map<std::size_t, std::vector<tk::real>>& ingrad)

Receive contributions to velocity gradients.

Parameters
ingrad in	Partial contributions of momentum flux to chare-boundary nodes. Key: global mesh node IDs, values: contributions.

This function receives contributions to m_vgrad, which stores the velocity gradients at mesh nodes. While m_vgrad stores own contributions, m_vgradc collects the neighbor chare contributions during communication. This way work on m_vgrad and m_vgradc is overlapped.

void inciter::LohCG::comflux(const std::unordered_map<std::size_t, std::vector<tk::real>>& influx)

Receive contributions to momentum flux on chare-boundaries.

Parameters
influx in	Partial contributions of momentum flux to chare-boundary nodes. Key: global mesh node IDs, values: contributions.

This function receives contributions to m_flux, which stores the momentum flux at mesh nodes. While m_flux stores own contributions, m_fluxc collects the neighbor chare contributions during communication. This way work on m_flux and m_fluxc is overlapped.

void inciter::LohCG::comsgrad(const std::unordered_map<std::size_t, std::vector<tk::real>>& ingrad)

Receive contributions to conjugate gradients solution gradient.

Parameters
ingrad in	Partial contributions to chare-boundary nodes. Key: global mesh node IDs, value: contributions for all scalar components.

This function receives contributions to m_sgrad, which stores the gradients at mesh nodes. While m_sgrad stores own contributions, m_sgradc collects the neighbor chare contributions during communication. This way work on m_sgrad and m_sgradc is overlapped.

void inciter::LohCG::comgrad(const std::unordered_map<std::size_t, std::vector<tk::real>>& ingrad)

Receive contributions to gradient.

Parameters
ingrad in	Partial contributions to chare-boundary nodes. Key: global mesh node IDs, value: contributions for all scalar components.

This function receives contributions to m_grad, which stores the gradients at mesh nodes. While m_grad stores own contributions, m_gradc collects the neighbor chare contributions during communication. This way work on m_grad and m_gradc is overlapped.

void inciter::LohCG::comrhs(const std::unordered_map<std::size_t, std::vector<tk::real>>& inrhs)

Receive contributions to right-hand side vector on chare-boundaries.

Parameters
inrhs in	Partial contributions of RHS to chare-boundary nodes. Key: global mesh node IDs, value: contributions for all scalar components.

This function receives contributions to m_rhs, which stores the right hand side vector at mesh nodes. While m_rhs stores own contributions, m_rhsc collects the neighbor chare contributions during communication. This way work on m_rhs and m_rhsc is overlapped.

void inciter::LohCG::comdiv(const std::unordered_map<std::size_t, tk::real>& indiv)

Receive contributions to velocity divergence on chare-boundaries.

Parameters
indiv in	Partial contributions of velocity divergence to chare-boundary nodes. Key: global mesh node IDs, value: contribution.

This function receives contributions to m_div, which stores the velocity divergence at mesh nodes. While m_div stores own contributions, m_divc collects the neighbor chare contributions during communication. This way work on m_div and m_divc is overlapped.

void inciter::LohCG::resizePostAMR(const std::vector<std::size_t>& ginpoel, const tk::UnsMesh::Chunk& chunk, const tk::UnsMesh::Coords& coord, const std::unordered_map<std::size_t, tk::UnsMesh::Edge>& addedNodes, const std::unordered_map<std::size_t, std::size_t>& addedTets, const std::set<std::size_t>& removedNodes, const std::unordered_map<int, std::unordered_set<std::size_t>>& nodeCommMap, const std::map<int, std::vector<std::size_t>>& bface, const std::map<int, std::vector<std::size_t>>& bnode, const std::vector<std::size_t>& triinpoel)

Receive new mesh from Refiner.

Parameters
ginpoel in	Mesh connectivity with global node ids
chunk in	New mesh chunk (connectivity and global<->local id maps)
coord in	New mesh node coordinates
addedNodes in	Newly added mesh nodes and their parents (local ids)
addedTets in	Newly added mesh cells and their parents (local ids)
removedNodes in	Newly removed mesh node local ids
nodeCommMap in	New node communication map
bface in	Boundary-faces mapped to side set ids
bnode in	Boundary-node lists mapped to side set ids
triinpoel in	Boundary-face connectivity

const tk::Fields& inciter::LohCG::solution() const

Returns	Const-ref to current solution

Const-ref access to current solution

void inciter::LohCG::evalLB(int nrestart)

Parameters
nrestart in	Number of times restarted

void inciter::LohCG::pup(PUP::er& p) override

Pack/Unpack serialize member function.

Parameters
p in/out	Charm++'s PUP::er serializer object reference

void inciter::LohCG::operator|(PUP::er& p, LohCG& i)

Pack/Unpack serialize operator|.

Parameters
p in/out	Charm++'s PUP::er serializer object reference
i in/out	LohCG object reference

Variable documentation

std::unordered_map<std::size_t, tk::real> inciter::LohCG::m_divc

Receive buffer for communication of the velocity divergence

Key: global node id, value: velocity divergence

std::unordered_map<std::size_t, std::vector<tk::real>> inciter::LohCG::m_rhsc

Receive buffer for communication of the right hand side

Key: global node id, value: rhs for all scalar components per node.

std::unordered_map<int, std::unordered_map<std::size_t, std::array<tk::real, 4>>> inciter::LohCG::m_bnorm

Boundary point normals

Outer key: side set id. Inner key: global node id of boundary point, value: weighted normal vector, inverse distance square.

decltype(m_bnorm) inciter::LohCG::m_bnormc

Boundary point normals receive buffer

Outer key: side set id. Inner key: global node id of boundary point, value: weighted normals and inverse distance square.

std::unordered_map<std::size_t, std::array<tk::real, 3>> inciter::LohCG::m_bndpoinint

Boundary point integrals

Key: global node id of boundary point, value: boundary point integral contributions.

inciter::LohCG class #include <src/Inciter/LohCG.hpp>