Question

我使用openmp将我的c ++代码并行化，如下所示：

....//some code
   double S_theta = 0.0, S_x = 0.0, S_y = 0.0;
#pragma omp parallel for private(dx,dy,theta_new) reduction(+ : S_x,S_y,S_theta)
    for(int j = 0; j < N; j++)
      {
        if (j==i) continue;
           double delta_x = x[i]-x[j],  

        for(int ky = -1; ky<= 1; ky++)
            {
            for(int kx = -1; kx<= 1; kx++)
               {

               if (r_ij_square > l0_two)
                  { 
                  double r_ij = sqrt(r_ij_square);
                  r_x_ij/= r_ij; 
                  r_y_ij/= r_ij;
                  double rdote = r_x_ij * e_x[j] + r_y_ij * e_y[j];
                  S_theta += e_dot_e * ( e_cross_e -  rdote * (e_x[i] * r_y_ij - e_y[i] * r_x_ij) ) / (r_ij_square * r_ij);
                  double S = rdote  /r_ij_square;
                  S_x += r_x_ij * S;
                  S_y += r_y_ij * S;
                  }
               }
        }

    double zeta = -eta / 2.0 ; 
    theta_new[i] += zeta + dt * dlthet;
    }
  ....//some code

但每次运行时结果都会改变！当我删除这一行

 #pragma omp parallel for reduction(+ : S_x,S_y,S_theta)

，结果在不同的运行中是相同的！我的代码出了什么问题？

Answer 1

您的变量S在循环外声明，并将在运行循环的所有线程之间共享。鉴于它是如何使用的，它可以在使用它的循环中声明为double S = rdote / r_ij_square;。

使用OPENMP并行化c ++代码

1 个答案: