I do not know how to convince B::Deparse (or B::Xref for that matter) to give you the starting line number. What do you plan to do with that information once you have it? Maybe the ideas in Re: 'extract' subs from source code can help. Perhaps if you poke around PPI you can figure out how to get the line numbers.
The goal is to use B::Deparse to check some properties of the code (kind of lint). The line numbers are needed to report precise information to the developer.
FWIW, yes, I'm aware that B::Lint and Perl::Critic exist.
I will look at PPI but this seems to be quite large...