diff options
Add hacky MPI support for grid refinement
Works but is nowhere near anything one could consider good.
Obvious issues:
* More than one cuboid per grid makes it harder to determine the next lattice cell to be coupled
* i.e. currently lattice positions are determined ad hoc by resolving their physical position
* Coupling is not actually parallelized
* All coupling lines are traversed by all processes, way to much communication
* Load balancing and cuboid decomposition doesn't care about refinement
* ideally refined cuboids should be computationally near their coarse _parent_ cuboids
The first two isses should be fixable with a reasonable amount of work.
This sadly doesn't apply in any form to the last issue.
Diffstat (limited to 'src/refinement')
-rw-r--r-- | src/refinement/coupler2D.hh | 77 | ||||
-rw-r--r-- | src/refinement/grid2D.hh | 7 |
2 files changed, 65 insertions, 19 deletions
diff --git a/src/refinement/coupler2D.hh b/src/refinement/coupler2D.hh index 8370707..ab082f4 100644 --- a/src/refinement/coupler2D.hh +++ b/src/refinement/coupler2D.hh @@ -35,10 +35,16 @@ template <typename T, template<typename> class DESCRIPTOR> Vector<int,3> Coupler2D<T,DESCRIPTOR>::getFineLatticeR(int y) const { if (_vertical) { - return _fineOrigin + Vector<int,3> {0, 0, y}; + const auto physR = _physOrigin + Vector<T,2>{0, y*_fine.getConverter().getPhysDeltaX()}; + Vector<int,3> latticeR{}; + _fine.getCuboidGeometry().getLatticeR(physR, latticeR); + return latticeR; } else { - return _fineOrigin + Vector<int,3> {0, y, 0}; + const auto physR = _physOrigin + Vector<T,2>{y*_fine.getConverter().getPhysDeltaX(), 0}; + Vector<int,3> latticeR{}; + _fine.getCuboidGeometry().getLatticeR(physR, latticeR); + return latticeR; } } @@ -46,10 +52,16 @@ template <typename T, template<typename> class DESCRIPTOR> Vector<int,3> Coupler2D<T,DESCRIPTOR>::getCoarseLatticeR(int y) const { if (_vertical) { - return _coarseOrigin + Vector<int,3> {0, 0, y}; + const auto physR = _physOrigin + Vector<T,2>{0, y*_coarse.getConverter().getPhysDeltaX()}; + Vector<int,3> latticeR{}; + _coarse.getCuboidGeometry().getLatticeR(physR, latticeR); + return latticeR; } else { - return _coarseOrigin + Vector<int,3> {0, y, 0}; + const auto physR = _physOrigin + Vector<T,2>{y*_coarse.getConverter().getPhysDeltaX(), 0}; + Vector<int,3> latticeR{}; + _coarse.getCuboidGeometry().getLatticeR(physR, latticeR); + return latticeR; } } @@ -97,12 +109,13 @@ void FineCoupler2D<T,DESCRIPTOR>::store() for (int y=0; y < this->_coarseSize; ++y) { const auto pos = this->getCoarseLatticeR(y); - T rho{}; T u[2] {}; T fNeq[DESCRIPTOR<T>::q] {}; - coarseLattice.get(pos).computeRhoU(rho, u); - coarseLattice.get(pos).computeFneq(fNeq); + Cell<T,DESCRIPTOR> coarseCell; + coarseLattice.get(pos, coarseCell); + lbHelpers<T,DESCRIPTOR>::computeRhoU(coarseCell, rho, u); + lbHelpers<T,DESCRIPTOR>::computeFneq(coarseCell, fNeq, rho, u); _c2f_rho[y] = rho; _c2f_u[y] = Vector<T,2>(u); @@ -134,17 +147,18 @@ void FineCoupler2D<T,DESCRIPTOR>::interpolate() auto& coarseLattice = this->_coarse.getSuperLattice(); for (int y=0; y < this->_coarseSize; ++y) { - const auto coarseCell = coarseLattice.get(this->getCoarseLatticeR(y)); + Cell<T,DESCRIPTOR> coarseCell; + coarseLattice.get(this->getCoarseLatticeR(y), coarseCell); T rho{}; T u[2] {}; - coarseCell.computeRhoU(rho, u); + lbHelpers<T,DESCRIPTOR>::computeRhoU(coarseCell, rho, u); _c2f_rho[y] = order2interpolation(rho, _c2f_rho[y]); _c2f_u[y][0] = order2interpolation(u[0], _c2f_u[y][0]); _c2f_u[y][1] = order2interpolation(u[1], _c2f_u[y][1]); T fNeq[DESCRIPTOR<T>::q] {}; - coarseCell.computeFneq(fNeq); + lbHelpers<T,DESCRIPTOR>::computeFneq(coarseCell, fNeq, rho, u); for (int iPop=0; iPop < DESCRIPTOR<T>::q; ++iPop) { _c2f_fneq[y][iPop] = order2interpolation(fNeq[iPop], _c2f_fneq[y][iPop]); } @@ -162,11 +176,16 @@ void FineCoupler2D<T,DESCRIPTOR>::couple() const auto finePos = this->getFineLatticeR(2*y); T fEq[DESCRIPTOR<T>::q] {}; - coarseLattice.get(coarsePos).computeFeq(fEq); + Cell<T,DESCRIPTOR> coarseCell; + coarseLattice.get(coarsePos, coarseCell); + lbHelpers<T,DESCRIPTOR>::computeFeq(coarseCell, fEq); + Cell<T,DESCRIPTOR> cell; + fineLattice.get(finePos, cell); for (int iPop=0; iPop < DESCRIPTOR<T>::q; ++iPop) { - fineLattice.get(finePos)[iPop] = fEq[iPop] + this->_coarse.getScalingFactor() * _c2f_fneq[y][iPop]; + cell[iPop] = fEq[iPop] + this->_coarse.getScalingFactor() * _c2f_fneq[y][iPop]; } + fineLattice.set(finePos, cell); } for (int y=1; y < this->_coarseSize-2; ++y) { @@ -192,6 +211,8 @@ void FineCoupler2D<T,DESCRIPTOR>::couple() const T uSqr = u[0]*u[0] + u[1]*u[1]; const auto finePos = this->getFineLatticeR(1+2*y); + Cell<T,DESCRIPTOR> fineCell; + fineLattice.get(finePos, fineCell); for (int iPop=0; iPop < DESCRIPTOR<T>::q; ++iPop) { const T fneq = order4interpolation( @@ -201,8 +222,10 @@ void FineCoupler2D<T,DESCRIPTOR>::couple() _c2f_fneq[y+2][iPop] ); - fineLattice.get(finePos)[iPop] = lbHelpers<T,DESCRIPTOR>::equilibrium(iPop, rho, u, uSqr) + this->_coarse.getScalingFactor() * fneq; + fineCell[iPop] = lbHelpers<T,DESCRIPTOR>::equilibrium(iPop, rho, u, uSqr) + this->_coarse.getScalingFactor() * fneq; } + + fineLattice.set(finePos, fineCell); } { @@ -225,6 +248,8 @@ void FineCoupler2D<T,DESCRIPTOR>::couple() const T uSqr = u[0]*u[0] + u[1]*u[1]; const auto finePos = this->getFineLatticeR(1); + Cell<T,DESCRIPTOR> fineCell; + fineLattice.get(finePos, fineCell); for (int iPop=0; iPop < DESCRIPTOR<T>::q; ++iPop) { const T fneq = order3interpolation( @@ -232,8 +257,10 @@ void FineCoupler2D<T,DESCRIPTOR>::couple() _c2f_fneq[1][iPop], _c2f_fneq[2][iPop] ); - fineLattice.get(finePos)[iPop] = lbHelpers<T,DESCRIPTOR>::equilibrium(iPop, rho, u, uSqr) + this->_coarse.getScalingFactor() * fneq; + fineCell[iPop] = lbHelpers<T,DESCRIPTOR>::equilibrium(iPop, rho, u, uSqr) + this->_coarse.getScalingFactor() * fneq; } + + fineLattice.set(finePos, fineCell); } { @@ -256,6 +283,8 @@ void FineCoupler2D<T,DESCRIPTOR>::couple() const T uSqr = u[0]*u[0] + u[1]*u[1]; const auto finePos = this->getFineLatticeR(this->_fineSize-2); + Cell<T,DESCRIPTOR> fineCell; + fineLattice.get(finePos, fineCell); for (int iPop=0; iPop < DESCRIPTOR<T>::q; ++iPop) { const T fneq = order3interpolation( @@ -263,8 +292,10 @@ void FineCoupler2D<T,DESCRIPTOR>::couple() _c2f_fneq[this->_coarseSize-2][iPop], _c2f_fneq[this->_coarseSize-3][iPop] ); - fineLattice.get(finePos)[iPop] = lbHelpers<T,DESCRIPTOR>::equilibrium(iPop, rho, u, uSqr) + this->_coarse.getScalingFactor() * fneq; + fineCell[iPop] = lbHelpers<T,DESCRIPTOR>::equilibrium(iPop, rho, u, uSqr) + this->_coarse.getScalingFactor() * fneq; } + + fineLattice.set(finePos, fineCell); } } @@ -276,7 +307,10 @@ void computeRestrictedFneq(const SuperLattice2D<T,DESCRIPTOR>& lattice, { for (int iPop=0; iPop < DESCRIPTOR<T>::q; ++iPop) { T fNeq[DESCRIPTOR<T>::q] {}; - lattice.get(pos[0], pos[1] + DESCRIPTOR<T>::c[iPop][0], pos[2] + DESCRIPTOR<T>::c[iPop][1]).computeFneq(fNeq); + Cell<T,DESCRIPTOR> cell; + const Vector<int,3> neighborPos {pos[0], pos[1] + DESCRIPTOR<T>::c[iPop][0], pos[2] + DESCRIPTOR<T>::c[iPop][1]}; + lattice.get(neighborPos, cell); + lbHelpers<T,DESCRIPTOR>::computeFneq(cell, fNeq); for (int jPop=0; jPop < DESCRIPTOR<T>::q; ++jPop) { restrictedFneq[jPop] += fNeq[jPop]; @@ -311,14 +345,21 @@ void CoarseCoupler2D<T,DESCRIPTOR>::couple() const auto coarsePos = this->getCoarseLatticeR(y); T fEq[DESCRIPTOR<T>::q] {}; - fineLattice.get(finePos).computeFeq(fEq); + Cell<T,DESCRIPTOR> fineCell; + fineLattice.get(finePos, fineCell); + lbHelpers<T,DESCRIPTOR>::computeFeq(fineCell, fEq); T fNeq[DESCRIPTOR<T>::q] {}; computeRestrictedFneq(fineLattice, finePos, fNeq); + Cell<T,DESCRIPTOR> coarseCell; + coarseLattice.get(coarsePos, coarseCell); + for (int iPop=0; iPop < DESCRIPTOR<T>::q; ++iPop) { - coarseLattice.get(coarsePos)[iPop] = fEq[iPop] + this->_coarse.getInvScalingFactor() * fNeq[iPop]; + coarseCell[iPop] = fEq[iPop] + this->_coarse.getInvScalingFactor() * fNeq[iPop]; } + + coarseLattice.set(coarsePos, coarseCell); } } diff --git a/src/refinement/grid2D.hh b/src/refinement/grid2D.hh index 71b943e..65dd84d 100644 --- a/src/refinement/grid2D.hh +++ b/src/refinement/grid2D.hh @@ -55,7 +55,12 @@ Grid2D<T,DESCRIPTOR>::Grid2D(IndicatorF2D<T>& domainF, int resolution, T tau, in _cuboids(new CuboidGeometry2D<T>( domainF, _converter->getConversionFactorLength(), - 1)), +#ifdef PARALLEL_MODE_MPI + singleton::mpi().getSize() +#else + 1 +#endif + )), _balancer(new HeuristicLoadBalancer<T>( *_cuboids)), _geometry(new SuperGeometry2D<T>( |