84     int i,j,k,n,cols;  in arm_mat_solve_lower_triangular_f32()  local
104       for(j=0; j+3 < cols; j += 4)  in arm_mat_solve_lower_triangular_f32()
106             vecA = vld1q_f32(&pA[i * cols + j]);  in arm_mat_solve_lower_triangular_f32()
110                 vecX = vld1q_f32(&pX[cols*k+j]);  in arm_mat_solve_lower_triangular_f32()
121             vst1q(&pX[i*cols+j],vecA);  in arm_mat_solve_lower_triangular_f32()
125        for(; j < cols; j ++)  in arm_mat_solve_lower_triangular_f32()
127             a_col = &pA[j];  in arm_mat_solve_lower_triangular_f32()
134                 tmp -= lt_row[k] * pX[cols*k+j];  in arm_mat_solve_lower_triangular_f32()
142             pX[i*cols+j] = tmp;  in arm_mat_solve_lower_triangular_f32()
185     int i,j,k,n,cols;  in arm_mat_solve_lower_triangular_f32()  local
205       for(j=0; j+3 < cols; j += 4)  in arm_mat_solve_lower_triangular_f32()
207             vecA = vld1q_f32(&pA[i * cols + j]);  in arm_mat_solve_lower_triangular_f32()
211                 vecX = vld1q_f32(&pX[cols*k+j]);  in arm_mat_solve_lower_triangular_f32()
222             vst1q_f32(&pX[i*cols+j],vecA);  in arm_mat_solve_lower_triangular_f32()
226        for(; j < cols; j ++)  in arm_mat_solve_lower_triangular_f32()
228             a_col = &pA[j];  in arm_mat_solve_lower_triangular_f32()
235                 tmp -= lt_row[k] * pX[cols*k+j];  in arm_mat_solve_lower_triangular_f32()
243             pX[i*cols+j] = tmp;  in arm_mat_solve_lower_triangular_f32()
284     int i,j,k,n,cols;  in arm_mat_solve_lower_triangular_f32()  local
297     for(j=0; j < cols; j ++)  in arm_mat_solve_lower_triangular_f32()
299        a_col = &pA[j];  in arm_mat_solve_lower_triangular_f32()
309                 tmp -= lt_row[k] * pX[cols*k+j];  in arm_mat_solve_lower_triangular_f32()
317             pX[i*cols+j] = tmp;  in arm_mat_solve_lower_triangular_f32()