Question:

In P30 of F28035 CLA document spruge6b.pdf, there is a CLA register/bit MSTF.RND32.

My customer and I are confused with what this rounding is because this is for floating point multiplication, addition and subtraction.

In our understanding, rounding is only used in computation in fixed-point. Would you please make it clearer or show us an example how this rounding affects?

 

Answer:

This is from www.ti.com/lit/spraan9 which is the C28x FPU Primer - but it is the same thing for CLA.