axpyz.hpp

Go to the documentation of this file.

#pragma once
 
namespace monolish {
 
namespace {
template <typename F1, typename F2, typename F3, typename F4>
void Daxpyz_core(const F1 alpha, const F2 &x, const F3 &y, F4 &z) {
  Logger &logger = Logger::get_instance();
  logger.func_in(monolish_func);
 
  // err
  assert(util::is_same_size(x, y, z));
  assert(util::is_same_device_mem_stat(x, y, z));
 
  const double *xd = x.data();
  const double *yd = y.data();
  double *zd = z.data();
  size_t size = x.size();
  const size_t xoffset = x.get_offset();
  const size_t yoffset = y.get_offset();
  const size_t zoffset = z.get_offset();
 
  if (x.get_device_mem_stat() == true) {
#if MONOLISH_USE_GPU
#pragma omp target teams distribute parallel for
    for (size_t i = 0; i < size; i++) {
      zd[i + zoffset] = alpha * xd[i + xoffset] + yd[i + yoffset];
    }
#else
    throw std::runtime_error(
        "error USE_GPU is false, but get_device_mem_stat() == true");
#endif
  } else {
#pragma omp parallel for
    for (size_t i = 0; i < size; i++) {
      zd[i + zoffset] = alpha * xd[i + xoffset] + yd[i + yoffset];
    }
  }
  logger.func_out();
}
 
template <typename F1, typename F2, typename F3, typename F4>
void Saxpyz_core(const F1 alpha, const F2 &x, const F3 &y, F4 &z) {
  Logger &logger = Logger::get_instance();
  logger.func_in(monolish_func);
 
  // err
  assert(util::is_same_size(x, y, z));
  assert(util::is_same_device_mem_stat(x, y, z));
 
  const float *xd = x.data();
  const float *yd = y.data();
  float *zd = z.data();
  size_t size = x.size();
  const size_t xoffset = x.get_offset();
  const size_t yoffset = y.get_offset();
  const size_t zoffset = z.get_offset();
 
  if (x.get_device_mem_stat() == true) {
#if MONOLISH_USE_GPU
#pragma omp target teams distribute parallel for
    for (size_t i = 0; i < size; i++) {
      zd[i + zoffset] = alpha * xd[i + xoffset] + yd[i + yoffset];
    }
#else
    throw std::runtime_error(
        "error USE_GPU is false, but get_device_mem_stat() == true");
#endif
  } else {
#pragma omp parallel for
    for (size_t i = 0; i < size; i++) {
      zd[i + zoffset] = alpha * xd[i + xoffset] + yd[i + yoffset];
    }
  }
  logger.func_out();
}
 
} // namespace
 
} // namespace monolish