matrix transpose tiling

1
2  const int N_r = 56;
3  const int N_c = 75;
4
5  const int TILE_DIM = 16;
6
7  const int outer_Dimc = (N_c - 1) / TILE_DIM + 1;
8  const int outer_Dimr = (N_r - 1) / TILE_DIM + 1;
9
10  int** dest; 
11  int** src;
12
13//
14  // (0) Outer loops to iterate over tiles
15  //
16  for (int by = 0; by < outer_Dimr; ++by) {
17    for (int bx = 0; bx < outer_Dimc; ++bx) {
18      //
19      // (1) Loops to iterate over tile entries
20      //
21      for (int ty = 0; ty < TILE_DIM; ++ty) {
22        for (int tx = 0; tx < TILE_DIM; ++tx) {
23
24          int col = bx * TILE_DIM + tx;  // Matrix column index
25          int row = by * TILE_DIM + ty;  // Matrix row index
26
27          // Bounds check
28          if (row < N_r && col < N_c) {
29            dest[col][row] = src[row][col];
30          }
31          
32        }
33      }
34      
35    }
36  }