[RFA 3/3] Move logic out of symbol_find_demangled_name

Tom Tromey tom@tromey.com
Wed Jun 8 02:14:00 GMT 2016


>>>>> "Pedro" == Pedro Alves <palves@redhat.com> writes:

Pedro> So IMHO, something like "la_sniff_from_mangled_name" would be
Pedro> even clearer.

I made this change.

Pedro> BTW, what do we do with the resulting demangled name?  Is
Pedro> there some sort of guideline for options to pass to the demangler?
Pedro> Some sort of output that we're aiming for?

The resulting name, if any, is put into a symbol by symbol_set_names.
(I didn't change this area, just made it so the existing code could be
factored into per-language methods.)

So, it should be whatever is appropriate for a symbol.

I think this area is still a bit messy.  IIRC there are some weird
demangling differences between psymbols and full symbols, and there is
still a hack for Java in symbol_set_names.

Tom



More information about the Gdb-patches mailing list