Floating point arithmetic error propagation

Question

I have a function $$y=ln(\frac{x_1}{x_2})$$, $$x_1, x_2 > 1$$ and two ways to calculate it:

1. $$v = \frac{x_1}{x_2} => y_1=ln(v)$$
2. $$v_1 = ln(x_1)$$ $$v_2 = ln(x_2)$$ $$y_2 = v_1 - v_2$$

How error is propagated and how it affects end result? Which way is better?