void rotate4(int dim, pixel *src, pixel *dst)
 {
     int i, j, inR;
     int dimN = (dim * dim) - dim;
     pixel *buf=malloc(4);
     // [i, j] = i * n + j
     // #define RIDX(i,j,n) ( (i)*(n)+(j) )
 
     for (i = 0; i < dim; i++){
 
     	inR = dimN - (dim * i);
 
     	for (j = 0; j < dim; j+=4){
     		buf[0] = src[ (j * dim) + i ];
     		buf[1] = src[ ( (j + 1) * dim) + i ];
     		buf[2] = src[ ( (j + 2) * dim) + i ];
     		buf[3] = src[ ( (j + 3) * dim) + i ];
     		
     		memcpy((dst+inR), buf, 4*sizeof(buf)); 	
 		}
	
 		
 
 	}
 
 }