Skip to content
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

ROCm workaround: Use ParallelFor instead of Reduce #2749

Merged
merged 5 commits into from
Feb 10, 2024
Merged

ROCm workaround: Use ParallelFor instead of Reduce #2749

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
d478f62
ROCm workaround: Use ParallelFor instead of Reduce
WeiqunZhang Feb 9, 2024
452ed6f
add Gpu::streamSynchronize
WeiqunZhang Feb 9, 2024
195d6af
fix typo
WeiqunZhang Feb 9, 2024
08a6670
fix cpu code
WeiqunZhang Feb 9, 2024
4b2f611
streamSync for HIP only
WeiqunZhang Feb 9, 2024
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
85 changes: 55 additions & 30 deletions Source/reactions/Castro_react.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -186,12 +186,14 @@ Castro::react_state(MultiFab& s, MultiFab& r, Real time, Real dt, const int stra
MultiFab tmp_mask_mf;
const MultiFab& mask_mf = mask_covered_zones ? getLevel(level+1).build_fine_mask() : tmp_mask_mf;

ReduceOps<ReduceOpSum> reduce_op;
ReduceData<Real> reduce_data(reduce_op);
using ReduceTuple = typename decltype(reduce_data)::Type;
#if defined(AMREX_USE_GPU)
Gpu::Buffer<int> d_num_failed({0});
auto* p_num_failed = d_num_failed.data();
#endif
int num_failed = 0;

#ifdef _OPENMP
#pragma omp parallel
#pragma omp parallel reduction(+:num_failed)
#endif
for (MFIter mfi(s, TilingIfNotGPU()); mfi.isValid(); ++mfi)
{
Expand All @@ -208,8 +210,11 @@ Castro::react_state(MultiFab& s, MultiFab& r, Real time, Real dt, const int stra
const auto problo = geom.ProbLoArray();
#endif

reduce_op.eval(bx, reduce_data,
[=] AMREX_GPU_HOST_DEVICE (int i, int j, int k) -> ReduceTuple
#if defined(AMREX_USE_GPU)
ParallelFor(bx, [=] AMREX_GPU_DEVICE (int i, int j, int k)
#else
LoopOnCpu(bx, [&] (int i, int j, int k) mutable
#endif
{

burn_t burn_state;
Expand All @@ -230,7 +235,7 @@ Castro::react_state(MultiFab& s, MultiFab& r, Real time, Real dt, const int stra

bool do_burn = true;
burn_state.success = true;
Real burn_failed = 0.0_rt;
int burn_failed = 0;

// Don't burn on zones inside shock regions, if the relevant option is set.

Expand Down Expand Up @@ -329,7 +334,7 @@ Castro::react_state(MultiFab& s, MultiFab& r, Real time, Real dt, const int stra
// If we were unsuccessful, update the failure count.

if (!burn_state.success) {
burn_failed = 1.0_rt;
burn_failed = 1;
}

// Add burning rates to reactions MultiFab, but be
Expand Down Expand Up @@ -399,19 +404,25 @@ Castro::react_state(MultiFab& s, MultiFab& r, Real time, Real dt, const int stra

}


return {burn_failed};

#if defined(AMREX_USE_GPU)
if (burn_failed) {
Gpu::Atomic::Add(p_num_failed, burn_failed);
}
#else
num_failed += burn_failed;
#endif
});

#if defined(AMREX_USE_HIP)
Gpu::streamSynchronize(); // otherwise HIP may faile to allocate the necessary resources.
#endif
}

ReduceTuple hv = reduce_data.value();
Real burn_failed = amrex::get<0>(hv);
#if defined(AMREX_USE_GPU)
num_failed = *(d_num_failed.copyToHost());
#endif

if (burn_failed != 0.0) {
burn_success = 0;
}
burn_success = !num_failed;

ParallelDescriptor::ReduceIntMin(burn_success);

Expand Down Expand Up @@ -516,11 +527,13 @@ Castro::react_state(Real time, Real dt)

int burn_success = 1;

ReduceOps<ReduceOpSum> reduce_op;
ReduceData<Real> reduce_data(reduce_op);

using ReduceTuple = typename decltype(reduce_data)::Type;
#if defined(AMREX_USE_GPU)
Gpu::Buffer<int> d_num_failed({0});
auto* p_num_failed = d_num_failed.data();
#endif
int num_failed = 0;

// why no omp here?
for (MFIter mfi(S_new, TilingIfNotGPU()); mfi.isValid(); ++mfi)
{
const Box& bx = mfi.growntilebox(ng);
Expand All @@ -542,8 +555,11 @@ Castro::react_state(Real time, Real dt)
const auto dx = geom.CellSizeArray();
const auto problo = geom.ProbLoArray();

reduce_op.eval(bx, reduce_data,
[=] AMREX_GPU_HOST_DEVICE (int i, int j, int k) -> ReduceTuple
#if defined(AMREX_USE_GPU)
ParallelFor(bx, [=] AMREX_GPU_DEVICE (int i, int j, int k)
#else
LoopOnCpu(bx, [&] (int i, int j, int k) mutable
#endif
{
burn_t burn_state;

Expand All @@ -563,7 +579,7 @@ Castro::react_state(Real time, Real dt)

bool do_burn = true;
burn_state.success = true;
Real burn_failed = 0.0_rt;
int burn_failed = 0;

// Don't burn on zones inside shock regions, if the
// relevant option is set.
Expand Down Expand Up @@ -687,7 +703,7 @@ Castro::react_state(Real time, Real dt)
// If we were unsuccessful, update the failure count.

if (!burn_state.success) {
burn_failed = 1.0_rt;
burn_failed = 1;
}

// update the state data.
Expand Down Expand Up @@ -780,16 +796,25 @@ Castro::react_state(Real time, Real dt)
}
}

return {burn_failed};
#if defined(AMREX_USE_GPU)
if (burn_failed) {
Gpu::Atomic::Add(p_num_failed, burn_failed);
}
#else
num_failed += burn_failed;
#endif
});

#if defined(AMREX_USE_HIP)
Gpu::streamSynchronize(); // otherwise HIP may faile to allocate the necessary resources.
#endif
}

ReduceTuple hv = reduce_data.value();
Real burn_failed = amrex::get<0>(hv);
#if defined(AMREX_USE_GPU)
num_failed = *(d_num_failed.copyToHost());
#endif

if (burn_failed != 0.0) {
burn_success = 0;
}
burn_success = !num_failed;

ParallelDescriptor::ReduceIntMin(burn_success);

Expand Down
Loading