How to tune numerical D? (matrix multiplication is faster in g++ vs gdc)

[Walter Bright]

On 3/3/2013 7:48 PM, J wrote:
> void mmult(int[][] m1, int[][] m2, int[][] m3)
> {
>      foreach(int i, int[] m1i; m1)
>      {
>          foreach(int j, ref int m3ij; m3[i])
>          {
>              int val;
>              foreach(int k, int[] m2k; m2)
>              {
>                  val += m1i[k] * m2k[j];
>              }
>              m3ij = val;
>          }
>      }
> }
[...]
> ////// C++ version
> int **mmult(int rows, int cols, int **m1, int **m2, int **m3) {
>      int i, j, k, val;
>      for (i=0; i<rows; i++) {
>      for (j=0; j<cols; j++) {
>          val = 0;
>          for (k=0; k<cols; k++) {
>          val += m1[i][k] * m2[k][j];
>          }
>          m3[i][j] = val;
>      }
>      }
>      return(m3);
> }

One difference that jumps out at me is you have extra variables and ref types in 
the D version, and in the C++ version you have "cached" the row & column loop 
limits. (I.e. the C++ version assumes a rectangular matrix, while the D one has 
a (presumably) different length for each column.)