Distribution of the supremum $M=\sup\limits_{t\leqslant\tau}W_t$ where $\tau=\inf\{t>0\mid W_t=-1\}$.
I've tried to use theorem which states $\sup\limits_{t\leq s}W_t \sim |W_s|$ and compute distribution of $\tau$ in order to rewrite task as $$P(M \leq y) = \int_{-\infty}^{\infty} P(\sup\limits_{t\leq s}W_t|\tau=s)p_\tau(s) \, ds $$ but it is useless because $\tau$ and $M$ are not independent.