Skip to content

Replace DirectSolverGive macros with type-safe functions #168

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
julianlitz wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Replace DirectSolverGive macros with type-safe functions #168

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
edd5a5a
Remove macros
julianlitz Feb 3, 2026
dddd51f
Formatting
julianlitz Feb 3, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions include/DirectSolver/DirectSolver-COO-MUMPS-Give/directSolverGive.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,6 +86,10 @@ class DirectSolver_COO_MUMPS_Give : public DirectSolver

// Retrieves the stencil for the solver matrix at the given radial index.
const Stencil& getStencil(int i_r) const;

void nodeBuildSolverMatrixGive(int i_r, int i_theta, const PolarGrid& grid, bool DirBC_Interior,
SparseMatrixCOO<double>& solver_matrix, double arr, double att, double art,
double detDF, double coeff_beta);
};

#endif
4 changes: 4 additions & 0 deletions include/DirectSolver/DirectSolver-CSR-LU-Give/directSolverGiveCustomLU.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -58,4 +58,8 @@ class DirectSolver_CSR_LU_Give : public DirectSolver
const Stencil& getStencil(int i_r) const;

int getStencilSize(int global_index) const;

void nodeBuildSolverMatrixGive(int i_r, int i_theta, const PolarGrid& grid, const bool DirBC_Interior,
SparseMatrixCSR<double>& solver_matrix, double arr, double att, double art,
double detDF, double coeff_beta);
};
30 changes: 10 additions & 20 deletions src/DirectSolver/DirectSolver-COO-MUMPS-Give/applySymmetryShift.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -122,35 +122,25 @@ void DirectSolver_COO_MUMPS_Give::applySymmetryShiftOuterBoundary(Vector<double>
}
}

// clang-format off
void DirectSolver_COO_MUMPS_Give::applySymmetryShift(Vector<double> x) const
{
assert(std::ssize(x) == grid_.numberOfNodes());
assert(grid_.nr() >= 4);

if (num_omp_threads_ == 1) {
/* Single-threaded execution */
if (DirBC_Interior_) {
applySymmetryShiftInnerBoundary(x);
}
applySymmetryShiftOuterBoundary(x);
}
else {
#pragma omp parallel sections num_threads(num_omp_threads_)
#pragma omp parallel sections num_threads(num_omp_threads_)
{
#pragma omp section
{
#pragma omp section
{
if (DirBC_Interior_) {
applySymmetryShiftInnerBoundary(x);
}
if (DirBC_Interior_) {
applySymmetryShiftInnerBoundary(x);
Comment on lines +134 to +135
Copy link
Collaborator

@EmilyBourne EmilyBourne Feb 4, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I wonder if we should make DirBC_Interior_ constexpr in the future. Do you have an opinion?

Copy link
Collaborator Author

@julianlitz julianlitz Feb 4, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I am not that experienced with constexpr yet, but if you like to make this modification feel free to.

I would guess there is no performance gain due to the branch predictor disregarding the other path.

Copy link
Collaborator

@EmilyBourne EmilyBourne Feb 5, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I would guess there is no performance gain due to the branch predictor disregarding the other path.

I have never studied the branch predictor in detail but you may well be right, especially after the first iteration.

I am not that experienced with constexpr yet, but if you like to make this modification feel free to.

In any case this would be for a separate PR

}
}

#pragma omp section
{
applySymmetryShiftOuterBoundary(x);
}
#pragma omp section
{
applySymmetryShiftOuterBoundary(x);
}
}
}
// clang-format on

#endif
Loading