This is the mail archive of the
mailing list for the GDB project.
Re: [Ada/dwarf] New DW_AT_GNAT_descriptive_type DWARF attribute...
- From: Tom Tromey <tromey at redhat dot com>
- To: Joel Brobecker <brobecker at adacore dot com>
- Cc: gdb-patches at sourceware dot org
- Date: Mon, 27 Apr 2009 18:14:17 -0600
- Subject: Re: [Ada/dwarf] New DW_AT_GNAT_descriptive_type DWARF attribute...
- References: <20090417161934.GW7585@adacore.com> <20090423234844.GE7570@adacore.com>
- Reply-to: tromey at redhat dot com
>>>>> "Joel" == Joel Brobecker <firstname.lastname@example.org> writes:
Joel> 1. Detection of the use of this attribute: Right now, the FSF GCC
Joel> does not produce this attribute, so we need to be able to
Joel> detect when the attribute is in use and when it's not.
Joel> The way AdaCore did it was through a special keyword in
Joel> the DW_AT_producer CU attribute.
We talked about this a bit on irc. IIUC the reason that some flag is
needed is that a given type may or may not have a parallel type, and
the searches are expensive.
Joel> 2. The dreaded gnat_stuff addition to the type_specific area
Joel> in struct main_type. Pretty kludgy, borderlining on gross :-).
We talked about this too. main_type has a bunch of extra bits; a
couple can be taken for the discriminant. This should clear up the
gross parts here.
Joel> What I'd like to do, however, is get a consensus on whether we're OK
Joel> with supporting this attribute, and how GCC is supposed to tell GDB
Joel> that the attribute is being used. That way, we can start on the GCC
Joel> end as well.
I think it is important to note that you already have the GCC work
done, just not in the FSF GCC. If you were starting from scratch, I
would much prefer a DWARF-based solution.
Given this, I think that if GCC generates it, then it makes sense for
GDB to read it.