(d) A C function for the matrix-vector multiply using Compressed Row Storage is shown below: 1...

  

Transcribed Image Text:

(d) A C function for the matrix-vector multiply using Compressed Row Storage is shown below: 1 int mat_vec(float *v, struct CRS_mat M, float *u){ 2 int i,j; 3 4 5 6 for(i=0;i<M.nrow;i++){ for(j=M.row_start[i]; j<M.row_start[i+1]; j++){ v[i]+= M.val[j] * u[ M.col[j] ]; } 7 } 8 return 0; 9} How would you parallelise this function using OpenMP? Annotate the changes needed to the code (providing line numbers). (6 marks) (d) A C function for the matrix-vector multiply using Compressed Row Storage is shown below: 1 int mat_vec(float *v, struct CRS_mat M, float *u){ 2 int i,j; 3 4 5 6 for(i=0;i<M.nrow;i++){ for(j=M.row_start[i]; j<M.row_start[i+1]; j++){ v[i]+= M.val[j] * u[ M.col[j] ]; } 7 } 8 return 0; 9} How would you parallelise this function using OpenMP? Annotate the changes needed to the code (providing line numbers). (6 marks)