Re: [RFC] While processing a struct die, store the method's address in its fn_field

[Doug Evans <dje@google.com>]

On Tue, Nov 25, 2014 at 11:31 PM, Siva Chandra <sivachandra@google.com> wrote:
> Phew...
>
> On Tue, Nov 25, 2014 at 7:22 PM, Doug Evans <dje@google.com> wrote:
>> fwiw, GDB is bad enough that I often can't trust just reading the code.
>> I often punt on that and actually single step through
>> the relevant code to REALLY see what is happening.
>
> I single stepped through the code and now see everything that you said here!
>
> I believed this in gdbtypes.h:
>
>  889             /* * If is_stub is clear, this is the mangled name which
>  890                we can look up to find the address of the method
>  891                (FIXME: it would be cleaner to have a pointer to the
>  892                struct symbol here instead).
>  893
>  894                If is_stub is set, this is the portion of the mangled
>  895                name which specifies the arguments.  For example, "ii",
>  896                if there are two int arguments, or "" if there are no
>  897                arguments.  See gdb_mangle_name for the conversion from
>  898                this format to the one used if is_stub is clear.  */
>  899
>  900             const char *physname;
>
> Rest of what I thought I understood fell out from this belief! And the
> names look so enticing.
>
> Resetting my mind ...

Zoinks!  Yeah, that comment needs some TLC.

[btw, the term "physname" is in a transition mode.
In some places it means "mangled name".
In other places it means "demangled name".
It's not clear to me what we want to transition to though.]