matrix_field.h

Go to the documentation of this file.

#pragma once
 
// trove requires the warp shuffle instructions introduced with Kepler
#if __COMPUTE_CAPABILITY__ >= 300
#include <trove/ptr.h>
#else
#define DISABLE_TROVE
#endif
 
#include <quda_matrix.h>
 
namespace quda
{
 
  template <typename T, int n> struct matrix_field {
    T *field;
    int volume_cb;
 
    matrix_field(T *field, int volume_cb) : field(field), volume_cb(volume_cb) {}
 
    __device__ __host__ inline void load(Matrix<T, n> &A, int x_cb, int parity) const
    {
      int idx = parity * volume_cb + x_cb;
#ifdef __CUDA_ARCH__
      const trove::coalesced_ptr<Matrix<T, n>> field_((Matrix<T, n> *)field);
      A = field_[idx];
#else
#pragma unroll
      for (int i = 0; i < n; i++)
#pragma unroll
        for (int j = 0; j < n; j++) A(i, j) = field[(n * idx + i) * n + j] = A(i, j);
#endif
    }
 
    __device__ __host__ inline void save(const Matrix<T, n> &A, int x_cb, int parity)
    {
      int idx = parity * volume_cb + x_cb;
#ifdef __CUDA_ARCH__
      trove::coalesced_ptr<Matrix<T, n>> field_((Matrix<T, n> *)field);
      field_[idx] = A;
#else
#pragma unroll
      for (int i = 0; i < n; i++)
#pragma unroll
        for (int j = 0; j < n; j++) field[(n * idx + i) * n + j] = A(i, j);
#endif
    }
  };
 
} // namespace quda