Skip to content

Bug: Loadbalancing doesn't compile under nvcc #223

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

Merged
LudwigBoess merged 4 commits into from
Aug 12, 2026
Merged

Bug: Loadbalancing doesn't compile under nvcc #223

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
ad2f3db
bugfix for loadbalancing when compiled with nvcc
LudwigBoess Aug 7, 2026
d31b1c4
Merge branch '1.5.0rc' into bug/loadbalancing_nvidia
LudwigBoess Aug 10, 2026
772b127
Merge branch '1.5.0rc' into bug/loadbalancing_nvidia
LudwigBoess Aug 10, 2026
765784e
replace the Rebalance particle-shift lambda with a ShiftPrtlIndices_k…
LudwigBoess Aug 11, 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
3 changes: 1 addition & 2 deletions src/framework/domain/metadomain.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -146,8 +146,7 @@ namespace ntt {
*/
void Rebalance(unsigned int dim_mask,
real_t tolerance,
ncells_t max_shift_cells)
requires(MetricClass<M>);
ncells_t max_shift_cells);

/* output-related ------------------------------------------------------- */
#if defined(OUTPUT_ENABLED)
Expand Down
135 changes: 91 additions & 44 deletions src/framework/domain/metadomain_loadbal.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,14 +105,87 @@ namespace ntt {
}
Kokkos::deep_copy(dst_dev, dst_h);
}

// Shift particle cell indices by dx_d after the local active-cell offset
// moved, and re-tag every particle whose new index falls outside the new
// active range [0, new_n_d) for the corresponding neighbor.
template <Dimension D>
class ShiftPrtlIndices_kernel {
array_t<int*> i1, i1_prev, i2, i2_prev, i3, i3_prev;
array_t<short*> tag;

const int dx1, dx2, dx3;
const int new_n1, new_n2, new_n3;

public:
ShiftPrtlIndices_kernel(array_t<int*>& i1,
array_t<int*>& i1_prev,
array_t<int*>& i2,
array_t<int*>& i2_prev,
array_t<int*>& i3,
array_t<int*>& i3_prev,
array_t<short*>& tag,
int dx1,
int dx2,
int dx3,
int new_n1,
int new_n2,
int new_n3)
: i1 { i1 }
, i1_prev { i1_prev }
, i2 { i2 }
, i2_prev { i2_prev }
, i3 { i3 }
, i3_prev { i3_prev }
, tag { tag }
, dx1 { dx1 }
, dx2 { dx2 }
, dx3 { dx3 }
, new_n1 { new_n1 }
, new_n2 { new_n2 }
, new_n3 { new_n3 } {}

Inline void operator()(prtlidx_t p) const {
if (tag(p) != ParticleTag::alive) {
return;
}
if constexpr (D == Dim::_1D or D == Dim::_2D or D == Dim::_3D) {
i1(p) += dx1;
i1_prev(p) += dx1;
}
if constexpr (D == Dim::_2D or D == Dim::_3D) {
i2(p) += dx2;
i2_prev(p) += dx2;
}
if constexpr (D == Dim::_3D) {
i3(p) += dx3;
i3_prev(p) += dx3;
}
if constexpr (D == Dim::_1D) {
tag(p) = mpi::SendTag(tag(p), i1(p) < 0, i1(p) >= new_n1);
} else if constexpr (D == Dim::_2D) {
tag(p) = mpi::SendTag(tag(p),
i1(p) < 0,
i1(p) >= new_n1,
i2(p) < 0,
i2(p) >= new_n2);
} else if constexpr (D == Dim::_3D) {
tag(p) = mpi::SendTag(tag(p),
i1(p) < 0,
i1(p) >= new_n1,
i2(p) < 0,
i2(p) >= new_n2,
i3(p) < 0,
i3(p) >= new_n3);
}
}
};
#endif // MPI_ENABLED

template <SimEngine::type S, MetricClass M>
void Metadomain<S, M>::Rebalance(unsigned int dim_mask,
real_t tolerance,
ncells_t max_shift_cells)
requires(MetricClass<M>)
{
ncells_t max_shift_cells) {
#if !defined(MPI_ENABLED)
(void)dim_mask;
(void)tolerance;
Expand Down Expand Up @@ -353,9 +426,6 @@ namespace ntt {
if (sp.npart() == 0) {
continue;
}
auto i1 = sp.i1, i2 = sp.i2, i3 = sp.i3;
auto i1p = sp.i1_prev, i2p = sp.i2_prev, i3p = sp.i3_prev;
auto tag = sp.tag;
const int dx1 = -delta[0];
int dx2 = 0;
int dx3 = 0;
Expand All @@ -370,44 +440,21 @@ namespace ntt {
dx3 = -delta[2];
new_n3 = static_cast<int>(new_local_ncells[2]);
}
Kokkos::parallel_for(
"RebalanceShiftPrtls",
sp.rangeActiveParticles(),
Lambda(prtlidx_t p) {
if (tag(p) != ParticleTag::alive) {
return;
}
if constexpr (M::Dim == Dim::_1D or M::Dim == Dim::_2D or
M::Dim == Dim::_3D) {
i1(p) += dx1;
i1p(p) += dx1;
}
if constexpr (M::Dim == Dim::_2D or M::Dim == Dim::_3D) {
i2(p) += dx2;
i2p(p) += dx2;
}
if constexpr (M::Dim == Dim::_3D) {
i3(p) += dx3;
i3p(p) += dx3;
}
if constexpr (M::Dim == Dim::_1D) {
tag(p) = mpi::SendTag(tag(p), i1(p) < 0, i1(p) >= new_n1);
} else if constexpr (M::Dim == Dim::_2D) {
tag(p) = mpi::SendTag(tag(p),
i1(p) < 0,
i1(p) >= new_n1,
i2(p) < 0,
i2(p) >= new_n2);
} else if constexpr (M::Dim == Dim::_3D) {
tag(p) = mpi::SendTag(tag(p),
i1(p) < 0,
i1(p) >= new_n1,
i2(p) < 0,
i2(p) >= new_n2,
i3(p) < 0,
i3(p) >= new_n3);
}
});
Kokkos::parallel_for("RebalanceShiftPrtls",
sp.rangeActiveParticles(),
ShiftPrtlIndices_kernel<M::Dim> { sp.i1,
sp.i1_prev,
sp.i2,
sp.i2_prev,
sp.i3,
sp.i3_prev,
sp.tag,
dx1,
dx2,
dx3,
new_n1,
new_n2,
new_n3 });
sp.set_unsorted();
}

Expand Down
Loading