Dirichlet_double_precision_CUDA.cu

Go to the documentation of this file.

//======================================================================================================================
//
//  This file is part of waLBerla. waLBerla is free software: you can
//  redistribute it and/or modify it under the terms of the GNU General Public
//  License as published by the Free Software Foundation, either version 3 of
//  the License, or (at your option) any later version.
//
//  waLBerla is distributed in the hope that it will be useful, but WITHOUT
//  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
//  FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
//  for more details.
//
//  You should have received a copy of the GNU General Public License along
//  with waLBerla (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
//
//! \\file Dirichlet_double_precision_CUDA.cpp
//! \\author pystencils
//======================================================================================================================
 
// kernel generated with pystencils v1.4+1.ge851f4e, lbmpy v1.4+1.ge9efe34, sympy v1.12.1, lbmpy_walberla/pystencils_walberla from waLBerla commit 007e77e077ad9d22b5eed6f3d3118240993e553c
 
#include "Dirichlet_double_precision_CUDA.h"
#include "core/DataTypes.h"
#include "core/Macros.h"
#include "gpu/ErrorChecking.h"
 
#define FUNC_PREFIX __global__
 
using namespace std;
 
namespace walberla {
namespace pystencils {
 
#if defined(__NVCC__)
#define RESTRICT __restrict__
#if defined(__NVCC_DIAG_PRAGMA_SUPPORT__)
#pragma nv_diagnostic push
#pragma nv_diag_suppress 177 // unused variable
#else
#pragma push
#pragma diag_suppress 177 // unused variable
#endif                    // defined(__NVCC_DIAG_PRAGMA_SUPPORT__)
#elif defined(__clang__)
#if defined(__CUDA__)
#if defined(__CUDA_ARCH__)
// clang compiling CUDA code in device mode
#define RESTRICT __restrict__
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wstrict-aliasing"
#pragma clang diagnostic ignored "-Wunused-variable"
#pragma clang diagnostic ignored "-Wconversion"
#pragma clang diagnostic ignored "-Wsign-compare"
#else
// clang compiling CUDA code in host mode
#define RESTRICT __restrict__
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wstrict-aliasing"
#pragma clang diagnostic ignored "-Wunused-variable"
#pragma clang diagnostic ignored "-Wconversion"
#pragma clang diagnostic ignored "-Wsign-compare"
#endif // defined(__CUDA_ARCH__)
#endif // defined(__CUDA__)
#elif defined(__GNUC__) or defined(__GNUG__)
#define RESTRICT __restrict__
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wstrict-aliasing"
#pragma GCC diagnostic ignored "-Wunused-variable"
#pragma GCC diagnostic ignored "-Wconversion"
#elif defined(_MSC_VER)
#define RESTRICT __restrict
#else
#define RESTRICT
#endif
 
// NOLINTBEGIN(readability-non-const-parameter*)
namespace internal_dirichlet_double_precision_cuda_boundary_Dirichlet_double_precision_CUDA {
static FUNC_PREFIX __launch_bounds__(256) void dirichlet_double_precision_cuda_boundary_Dirichlet_double_precision_CUDA(double *RESTRICT _data_field, uint8_t *RESTRICT const _data_indexVector, int64_t const _stride_field_0, int64_t const _stride_field_1, int64_t const _stride_field_2, int32_t indexVectorSize) {
  if (blockDim.x * blockIdx.x + threadIdx.x < indexVectorSize) {
    uint8_t *RESTRICT _data_indexVector_10 = _data_indexVector;
    const int32_t x = *((int32_t *)(&_data_indexVector_10[24 * blockDim.x * blockIdx.x + 24 * threadIdx.x]));
    uint8_t *RESTRICT _data_indexVector_14 = _data_indexVector + 4;
    const int32_t y = *((int32_t *)(&_data_indexVector_14[24 * blockDim.x * blockIdx.x + 24 * threadIdx.x]));
    uint8_t *RESTRICT _data_indexVector_18 = _data_indexVector + 8;
    const int32_t z = *((int32_t *)(&_data_indexVector_18[24 * blockDim.x * blockIdx.x + 24 * threadIdx.x]));
 
    double *RESTRICT _data_field_10_20 = _data_field + _stride_field_1 * y + _stride_field_2 * z;
    uint8_t *RESTRICT _data_indexVector_116 = _data_indexVector + 16;
    _data_field_10_20[_stride_field_0 * x] = *((double *)(&_data_indexVector_116[24 * blockDim.x * blockIdx.x + 24 * threadIdx.x]));
  }
}
} // namespace internal_dirichlet_double_precision_cuda_boundary_Dirichlet_double_precision_CUDA
 
// NOLINTEND(readability-non-const-parameter*)
 
#if defined(__NVCC__)
#if defined(__NVCC_DIAG_PRAGMA_SUPPORT__)
#pragma nv_diagnostic pop
#else
#pragma pop
#endif // defined(__NVCC_DIAG_PRAGMA_SUPPORT__)
#elif defined(__clang__)
#if defined(__CUDA__)
#if defined(__CUDA_ARCH__)
// clang compiling CUDA code in device mode
#pragma clang diagnostic pop
#else
// clang compiling CUDA code in host mode
#pragma clang diagnostic pop
#endif // defined(__CUDA_ARCH__)
#endif // defined(__CUDA__)
#elif defined(__GNUC__) or defined(__GNUG__)
#pragma GCC diagnostic pop
#endif
 
void Dirichlet_double_precision_CUDA::run_impl(IBlock *block, IndexVectors::Type type, gpuStream_t stream) {
  auto *indexVectors = block->getData<IndexVectors>(indexVectorID);
  int32_t indexVectorSize = int32_c(indexVectors->indexVector(type).size());
  if (indexVectorSize == 0)
    return;
 
  auto pointer = indexVectors->pointerGpu(type);
 
  uint8_t *_data_indexVector = reinterpret_cast<uint8_t *>(pointer);
 
  auto field = block->getData<gpu::GPUField<double>>(fieldID);
 
  WALBERLA_ASSERT_GREATER_EQUAL(0, -int_c(field->nrOfGhostLayers()))
  double *RESTRICT _data_field = field->dataAt(0, 0, 0, 0);
  const int64_t _stride_field_0 = int64_t(field->xStride());
  const int64_t _stride_field_1 = int64_t(field->yStride());
  const int64_t _stride_field_2 = int64_t(field->zStride());
  dim3 _block(uint32_c(((256 < indexVectorSize) ? 256 : indexVectorSize)), uint32_c(1), uint32_c(1));
  dim3 _grid(uint32_c(((indexVectorSize) % (((256 < indexVectorSize) ? 256 : indexVectorSize)) == 0 ? (int64_t)(indexVectorSize) / (int64_t)(((256 < indexVectorSize) ? 256 : indexVectorSize)) : ((int64_t)(indexVectorSize) / (int64_t)(((256 < indexVectorSize) ? 256 : indexVectorSize))) + 1)), uint32_c(1), uint32_c(1));
  internal_dirichlet_double_precision_cuda_boundary_Dirichlet_double_precision_CUDA::dirichlet_double_precision_cuda_boundary_Dirichlet_double_precision_CUDA<<<_grid, _block, 0, stream>>>(_data_field, _data_indexVector, _stride_field_0, _stride_field_1, _stride_field_2, indexVectorSize);
}
 
void Dirichlet_double_precision_CUDA::run(IBlock *block, gpuStream_t stream) {
  run_impl(block, IndexVectors::ALL, stream);
}
 
void Dirichlet_double_precision_CUDA::inner(IBlock *block, gpuStream_t stream) {
  run_impl(block, IndexVectors::INNER, stream);
}
 
void Dirichlet_double_precision_CUDA::outer(IBlock *block, gpuStream_t stream) {
  run_impl(block, IndexVectors::OUTER, stream);
}
 
} // namespace pystencils
} // namespace walberla