Inconsistent behavior for tensors in RHS

[AjayBrahmakshatriya]

For the expression - y(i) = A(i,j) * x(j) + z(i), the documentation suggests z(i) is added exactly one for each i outside the j loop. This is consistent with the result from taco "y(i) = A(i,j) * x(j) + z(i)" -f=y:d -f=A:dd -f=x:d -f=z:d -

int compute(taco_tensor_t *y, taco_tensor_t *A, taco_tensor_t *x, taco_tensor_t *z) {
  int y1_dimension = (int)(y->dimensions[0]);
  double* restrict y_vals = (double*)(y->vals);
  int A1_dimension = (int)(A->dimensions[0]);
  int A2_dimension = (int)(A->dimensions[1]);
  double* restrict A_vals = (double*)(A->vals);
  int x1_dimension = (int)(x->dimensions[0]);
  double* restrict x_vals = (double*)(x->vals);
  int z1_dimension = (int)(z->dimensions[0]);
  double* restrict z_vals = (double*)(z->vals);

  #pragma omp parallel for schedule(runtime)
  for (int32_t i = 0; i < z1_dimension; i++) {
    double tj_val = 0.0;
    for (int32_t j = 0; j < x1_dimension; j++) {
      int32_t jA = i * A2_dimension + j;
      tj_val += A_vals[jA] * x_vals[j];
    }
    y_vals[i] = tj_val + z_vals[i];
  }
  return 0;
}

However adding any more terms pushes z(i) inside the loop. For example - taco "y(i) = A(i,j) * x(j) + z(i) + 0 " -f=y:d -f=A:dd -f=x:d -f=z:d generates -

int compute(taco_tensor_t *y, taco_tensor_t *A, taco_tensor_t *x, taco_tensor_t *z) {
  int y1_dimension = (int)(y->dimensions[0]);
  double* restrict y_vals = (double*)(y->vals);
  int A1_dimension = (int)(A->dimensions[0]);
  int A2_dimension = (int)(A->dimensions[1]);
  double* restrict A_vals = (double*)(A->vals);
  int x1_dimension = (int)(x->dimensions[0]);
  double* restrict x_vals = (double*)(x->vals);
  int z1_dimension = (int)(z->dimensions[0]);
  double* restrict z_vals = (double*)(z->vals);

  #pragma omp parallel for schedule(runtime)
  for (int32_t i = 0; i < z1_dimension; i++) {
    double tj_val = 0.0;
    for (int32_t j = 0; j < x1_dimension; j++) {
      int32_t jA = i * A2_dimension + j;
      tj_val += A_vals[jA] * x_vals[j] + z_vals[i];
    }
    y_vals[i] = tj_val;
  }
  return 0;
}

Similar behavior is seen when z(i) is simply added twice.