Splines improve benchmark

benchmarks/splines.cpp

@@ -14,27 +14,37 @@
 
 #include <benchmark/benchmark.h>
 
+static const ddc::SplineSolver Backend = ddc::SplineSolver::GINKGO;
+
 namespace DDC_HIP_5_7_ANONYMOUS_NAMESPACE_WORKAROUND(SPLINES_CPP)
 {
-    static constexpr std::size_t s_degree_x = 3;
-
     struct X
     {
         static constexpr bool PERIODIC = true;
     };
 
-    struct BSplinesX : ddc::UniformBSplines<X, s_degree_x>
+    template <typename NonUniform, std::size_t s_degree_x>
+    struct BSplinesX
+        : std::conditional_t<
+                  NonUniform::value,
+                  ddc::NonUniformBSplines<X, s_degree_x>,
+                  ddc::UniformBSplines<X, s_degree_x>>
     {
     };
+
+    template <typename NonUniform, std::size_t s_degree_x>
     using GrevillePoints = ddc::GrevilleInterpolationPoints<
-            BSplinesX,
+            BSplinesX<NonUniform, s_degree_x>,
             ddc::BoundCond::PERIODIC,
             ddc::BoundCond::PERIODIC>;
-    struct DDimX : GrevillePoints::interpolation_mesh_type
+
+    template <typename NonUniform, std::size_t s_degree_x>
+    struct DDimX : GrevillePoints<NonUniform, s_degree_x>::interpolation_mesh_type
     {
     };
 
     struct Y;
+
     struct DDimY : ddc::UniformPointSampling<Y>
     {
     };
@@ -47,7 +57,8 @@ void monitorMemoryAsync(std::mutex& mutex, bool& monitorFlag, size_t& maxUsedMem
     size_t freeMem = 0;
     size_t totalMem = 0;
     while (monitorFlag) {
-        std::this_thread::sleep_for(std::chrono::milliseconds(1)); // Adjust the interval as needed
+        std::this_thread::sleep_for(
+                std::chrono::microseconds(100)); // Adjust the interval as needed
 
         // Acquire a lock to ensure thread safety when accessing CUDA functions
         std::lock_guard<std::mutex> lock(mutex);
@@ -59,8 +70,14 @@ void monitorMemoryAsync(std::mutex& mutex, bool& monitorFlag, size_t& maxUsedMem
     }
 }
 
-static void characteristics_advection(benchmark::State& state)
+template <typename ExecSpace, typename NonUniform, std::size_t s_degree_x>
+static void characteristics_advection_unitary(benchmark::State& state)
 {
+    std::size_t nx = state.range(3);
+    std::size_t ny = state.range(4);
+    int cols_per_chunk = state.range(5);
+    int preconditionner_max_block_size = state.range(6);
+
     size_t freeMem = 0;
     size_t totalMem = 0;
 #if defined(__CUDACC__)
@@ -80,76 +97,87 @@ static void characteristics_advection(benchmark::State& state)
             std::ref(monitorFlag),
             std::ref(maxUsedMem));
 
-    ddc::init_discrete_space<
-            BSplinesX>(ddc::Coordinate<X>(-1.), ddc::Coordinate<X>(1.), state.range(0));
-    ddc::init_discrete_space<DDimX>(ddc::GrevilleInterpolationPoints<
-                                    BSplinesX,
-                                    ddc::BoundCond::PERIODIC,
-                                    ddc::BoundCond::PERIODIC>::get_sampling<DDimX>());
+    if constexpr (!NonUniform::value) {
+        ddc::init_discrete_space<BSplinesX<
+                NonUniform,
+                s_degree_x>>(ddc::Coordinate<X>(0.), ddc::Coordinate<X>(1.), nx);
+    } else {
+        std::vector<ddc::Coordinate<X>> breaks(nx + 1);
+        for (std::size_t i(0); i < nx + 1; ++i) {
+            breaks[i] = ddc::Coordinate<X>(static_cast<double>(i) / nx);
+        }
+        ddc::init_discrete_space<BSplinesX<NonUniform, s_degree_x>>(breaks);
+    }
+    ddc::init_discrete_space<DDimX<NonUniform, s_degree_x>>(
+            ddc::GrevilleInterpolationPoints<
+                    BSplinesX<NonUniform, s_degree_x>,
+                    ddc::BoundCond::PERIODIC,
+                    ddc::BoundCond::PERIODIC>::
+                    template get_sampling<DDimX<NonUniform, s_degree_x>>());
     ddc::DiscreteDomain<DDimY> y_domain = ddc::init_discrete_space<DDimY>(DDimY::init<DDimY>(
             ddc::Coordinate<Y>(-1.),
             ddc::Coordinate<Y>(1.),
-            ddc::DiscreteVector<DDimY>(state.range(1))));
+            ddc::DiscreteVector<DDimY>(ny)));
 
     auto const x_domain = ddc::GrevilleInterpolationPoints<
-            BSplinesX,
+            BSplinesX<NonUniform, s_degree_x>,
             ddc::BoundCond::PERIODIC,
-            ddc::BoundCond::PERIODIC>::get_domain<DDimX>();
+            ddc::BoundCond::PERIODIC>::template get_domain<DDimX<NonUniform, s_degree_x>>();
     ddc::Chunk density_alloc(
-            ddc::DiscreteDomain<DDimX, DDimY>(x_domain, y_domain),
-            ddc::DeviceAllocator<double>());
+            ddc::DiscreteDomain<DDimX<NonUniform, s_degree_x>, DDimY>(x_domain, y_domain),
+            ddc::KokkosAllocator<double, typename ExecSpace::memory_space>());
     ddc::ChunkSpan const density = density_alloc.span_view();
     // Initialize the density on the main domain
-    ddc::DiscreteDomain<DDimX, DDimY> x_mesh
-            = ddc::DiscreteDomain<DDimX, DDimY>(x_domain, y_domain);
+    ddc::DiscreteDomain<DDimX<NonUniform, s_degree_x>, DDimY> x_mesh
+            = ddc::DiscreteDomain<DDimX<NonUniform, s_degree_x>, DDimY>(x_domain, y_domain);
     ddc::parallel_for_each(
+            ExecSpace(),
             x_mesh,
-            KOKKOS_LAMBDA(ddc::DiscreteElement<DDimX, DDimY> const ixy) {
-                double const x = ddc::coordinate(ddc::select<DDimX>(ixy));
+            KOKKOS_LAMBDA(ddc::DiscreteElement<DDimX<NonUniform, s_degree_x>, DDimY> const ixy) {
+                double const x = ddc::coordinate(ddc::select<DDimX<NonUniform, s_degree_x>>(ixy));
                 double const y = ddc::coordinate(ddc::select<DDimY>(ixy));
                 density(ixy) = 9.999 * Kokkos::exp(-(x * x + y * y) / 0.1 / 2);
                 // initial_density(ixy) = 9.999 * ((x * x + y * y) < 0.25);
             });
     ddc::SplineBuilder<
-            Kokkos::DefaultExecutionSpace,
-            Kokkos::DefaultExecutionSpace::memory_space,
-            BSplinesX,
-            DDimX,
+            ExecSpace,
+            typename ExecSpace::memory_space,
+            BSplinesX<NonUniform, s_degree_x>,
+            DDimX<NonUniform, s_degree_x>,
             ddc::BoundCond::PERIODIC,
             ddc::BoundCond::PERIODIC,
-            ddc::SplineSolver::GINKGO,
-            DDimX,
+            Backend,
+            DDimX<NonUniform, s_degree_x>,
             DDimY>
-            spline_builder(x_mesh, state.range(2), state.range(3));
+            spline_builder(x_mesh, cols_per_chunk, preconditionner_max_block_size);
     ddc::PeriodicExtrapolationRule<X> periodic_extrapolation;
     ddc::SplineEvaluator<
-            Kokkos::DefaultExecutionSpace,
-            Kokkos::DefaultExecutionSpace::memory_space,
-            BSplinesX,
-            DDimX,
+            ExecSpace,
+            typename ExecSpace::memory_space,
+            BSplinesX<NonUniform, s_degree_x>,
+            DDimX<NonUniform, s_degree_x>,
             ddc::PeriodicExtrapolationRule<X>,
             ddc::PeriodicExtrapolationRule<X>,
-            DDimX,
+            DDimX<NonUniform, s_degree_x>,
             DDimY>
             spline_evaluator(periodic_extrapolation, periodic_extrapolation);
     ddc::Chunk coef_alloc(
             spline_builder.batched_spline_domain(),
-            ddc::KokkosAllocator<double, Kokkos::DefaultExecutionSpace::memory_space>());
+            ddc::KokkosAllocator<double, typename ExecSpace::memory_space>());
     ddc::ChunkSpan coef = coef_alloc.span_view();
     ddc::Chunk feet_coords_alloc(
             spline_builder.batched_interpolation_domain(),
-            ddc::KokkosAllocator<
-                    ddc::Coordinate<X, Y>,
-                    Kokkos::DefaultExecutionSpace::memory_space>());
+            ddc::KokkosAllocator<ddc::Coordinate<X, Y>, typename ExecSpace::memory_space>());
     ddc::ChunkSpan feet_coords = feet_coords_alloc.span_view();
 
     for (auto _ : state) {
         Kokkos::Profiling::pushRegion("FeetCharacteristics");
         ddc::parallel_for_each(
+                ExecSpace(),
                 feet_coords.domain(),
-                KOKKOS_LAMBDA(ddc::DiscreteElement<DDimX, DDimY> const e) {
+                KOKKOS_LAMBDA(ddc::DiscreteElement<DDimX<NonUniform, s_degree_x>, DDimY> const e) {
                     feet_coords(e) = ddc::Coordinate<X, Y>(
-                            ddc::coordinate(ddc::select<DDimX>(e))
+                            ddc::coordinate(ddc::select<DDimX<NonUniform, s_degree_x>>(e))
                                     - ddc::Coordinate<X>(0.0176429863),
                             ddc::coordinate(ddc::select<DDimY>(e)));
                 });
@@ -163,9 +191,7 @@ static void characteristics_advection(benchmark::State& state)
     }
     monitorFlag = false;
     monitorThread.join();
-    state.SetBytesProcessed(
-            int64_t(state.iterations())
-            * int64_t(state.range(0) * state.range(1) * sizeof(double)));
+    state.SetBytesProcessed(int64_t(state.iterations()) * int64_t(nx * ny * sizeof(double)));
     state.counters["gpu_mem_occupancy"] = maxUsedMem - initUsedMem;
     ////////////////////////////////////////////////////
     /// --------------- HUGE WARNING --------------- ///
@@ -175,55 +201,147 @@ static void characteristics_advection(benchmark::State& state)
     /// The reason is it acts on underlying global   ///
     /// variables, which is always a bad idea.       ///
     ////////////////////////////////////////////////////
-    ddc::detail::g_discrete_space_dual<BSplinesX>.reset();
-    ddc::detail::g_discrete_space_dual<ddc::UniformBsplinesKnots<BSplinesX>>.reset();
-    ddc::detail::g_discrete_space_dual<DDimX>.reset();
+    ddc::detail::g_discrete_space_dual<BSplinesX<NonUniform, s_degree_x>>.reset();
+    if constexpr (!NonUniform::value) {
+        ddc::detail::g_discrete_space_dual<ddc::UniformBsplinesKnots<BSplinesX<NonUniform, s_degree_x>>>.reset();
+    } else {
+        ddc::detail::g_discrete_space_dual<ddc::NonUniformBsplinesKnots<BSplinesX<NonUniform, s_degree_x>>>.reset();
+    }
+    ddc::detail::g_discrete_space_dual<DDimX<NonUniform, s_degree_x>>.reset();
     ddc::detail::g_discrete_space_dual<DDimY>.reset();
     ////////////////////////////////////////////////////
 }
 
-// Tuning : 512 cols and 8 precond on CPU, 16384 cols and 1 precond on GPU
+static void characteristics_advection(benchmark::State& state)
+{
+    // Preallocate 12 unitary benchs for each combination of cpu/gpu execution space, uniform/non-uniform and spline degree we may want to benchmark (those are determined at compile-time, that's why we need to build explicitely 12 variants of the bench even if we call only one of them)
+    std::array<std::function<void(benchmark::State&)>, 12> benchs;
+    benchs[0] = characteristics_advection_unitary<
+            Kokkos::DefaultHostExecutionSpace,
+            std::false_type,
+            3>;
+    benchs[1] = characteristics_advection_unitary<
+            Kokkos::DefaultHostExecutionSpace,
+            std::false_type,
+            4>;
+    benchs[2] = characteristics_advection_unitary<
+            Kokkos::DefaultHostExecutionSpace,
+            std::false_type,
+            5>;
+    benchs[3] = characteristics_advection_unitary<
+            Kokkos::DefaultHostExecutionSpace,
+            std::true_type,
+            3>;
+    benchs[4] = characteristics_advection_unitary<
+            Kokkos::DefaultHostExecutionSpace,
+            std::true_type,
+            4>;
+    benchs[5] = characteristics_advection_unitary<
+            Kokkos::DefaultHostExecutionSpace,
+            std::true_type,
+            5>;
+    benchs[6]
+            = characteristics_advection_unitary<Kokkos::DefaultExecutionSpace, std::false_type, 3>;
+    benchs[7]
+            = characteristics_advection_unitary<Kokkos::DefaultExecutionSpace, std::false_type, 4>;
+    benchs[8]
+            = characteristics_advection_unitary<Kokkos::DefaultExecutionSpace, std::false_type, 5>;
+    benchs[9] = characteristics_advection_unitary<Kokkos::DefaultExecutionSpace, std::true_type, 3>;
+    benchs[10]
+            = characteristics_advection_unitary<Kokkos::DefaultExecutionSpace, std::true_type, 4>;
+    benchs[11]
+            = characteristics_advection_unitary<Kokkos::DefaultExecutionSpace, std::true_type, 5>;
+
+    // Run the desired bench
+    benchs[state.range(0) * 6 + state.range(1) * 3 + state.range(2) - 3](state);
+}
 
-#ifdef KOKKOS_ENABLE_CUDA
-std::string chip = "gpu";
+// Reference parameters: the benchmarks sweep on two parameters and fix all the others according to those reference parameters.
+bool on_gpu_ref = true;
+bool non_uniform_ref = false;
+std::size_t degree_x_ref = 3;
+#if (defined(KOKKOS_ENABLE_CUDA) or defined(KOKKOS_ENABLE_HIP))
 std::size_t cols_per_chunk_ref = 65535;
 unsigned int preconditionner_max_block_size_ref = 1u;
 #elif defined(KOKKOS_ENABLE_OPENMP)
-std::string chip = "cpu";
 std::size_t cols_per_chunk_ref = 8192;
-unsigned int preconditionner_max_block_size_ref = 32u;
+unsigned int preconditionner_max_block_size_ref = 1u;
 #elif defined(KOKKOS_ENABLE_SERIAL)
-std::string chip = "cpu";
 std::size_t cols_per_chunk_ref = 8192;
 unsigned int preconditionner_max_block_size_ref = 32u;
 #endif
+// std::size_t ny_ref = 100000;
+std::size_t ny_ref = 1000;
 
+// Sweep on spline order
+std::string name = "degree_x";
+BENCHMARK(characteristics_advection)
+        ->RangeMultiplier(2)
+        ->Ranges(
+                {{false, true},
+                 {false, true},
+                 {3, 5},
+                 {64, 1024},
+                 {ny_ref, ny_ref},
+                 {cols_per_chunk_ref, cols_per_chunk_ref},
+                 {preconditionner_max_block_size_ref, preconditionner_max_block_size_ref}})
+        ->MinTime(3)
+        ->UseRealTime();
+/*
+// Sweep on ny
+std::string name = "ny";
 BENCHMARK(characteristics_advection)
         ->RangeMultiplier(2)
         ->Ranges(
-                {{64, 1024},
+                {{false, true},
+                 {false, true},
+                 {degree_x_ref, degree_x_ref},
+                 {64, 1024},
                  {100, 200000},
                  {cols_per_chunk_ref, cols_per_chunk_ref},
                  {preconditionner_max_block_size_ref, preconditionner_max_block_size_ref}})
         ->MinTime(3)
         ->UseRealTime();
+*/
 /*
+// Sweep on cols_per_chunk
+std::string name = "cols_per_chunk";
 BENCHMARK(characteristics_advection)
         ->RangeMultiplier(2)
-        ->Ranges({{64, 1024}, {100000, 100000}, {64,65535}, {preconditionner_max_block_size_ref, preconditionner_max_block_size_ref}})
-        ->MinTime(3)->UseRealTime();
+        ->Ranges(
+                {{false, true},
+                 {false, true},
+                 {degree_x_ref, degree_x_ref},
+                 {64, 1024},
+                 {ny_ref, ny_ref},
+                 {64, 65535},
+                 {preconditionner_max_block_size_ref, preconditionner_max_block_size_ref}})
+        ->MinTime(3)
+        ->UseRealTime();
 */
 /*
+// Sweep on preconditionner_max_block_size
+std::string name = "preconditionner_max_block_size";
 BENCHMARK(characteristics_advection)
         ->RangeMultiplier(2)
-        ->Ranges({{64, 1024}, {100000, 100000}, {cols_per_chunk_ref, cols_per_chunk_ref}, {1, 32}})
-        ->MinTime(3)->UseRealTime();
+        ->Ranges(
+                {{on_gpu_ref, on_gpu_ref},
+                 {false, true},
+                 {degree_x_ref, degree_x_ref},
+                 {64, 1024},
+                 {ny_ref, ny_ref},
+                 {cols_per_chunk_ref, cols_per_chunk_ref},
+                 {1, 32}})
+        ->MinTime(3)
+        ->UseRealTime();
 */
 
 int main(int argc, char** argv)
 {
     ::benchmark::Initialize(&argc, argv);
-    ::benchmark::AddCustomContext("chip", chip);
+    ::benchmark::AddCustomContext("name", name);
+    ::benchmark::
+            AddCustomContext("backend", Backend == ddc::SplineSolver::GINKGO ? "GINKGO" : "LAPACK");
     ::benchmark::AddCustomContext("cols_per_chunk_ref", std::to_string(cols_per_chunk_ref));
     ::benchmark::AddCustomContext(
             "preconditionner_max_block_size_ref",