This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Debug information for types questions
- To: Neil Booth <neil at daikokuya dot demon dot co dot uk>
- Subject: Re: Debug information for types questions
- From: Daniel Berlin <dan at cgsoftware dot com>
- Date: Sun, 22 Jul 2001 14:29:51 -0400
- Cc: gcc at gcc dot gnu dot org
- References: <20010722182418.A3747@daikokuya.demon.co.uk>
Neil Booth <neil@daikokuya.demon.co.uk> writes:
> In toplev.c, we have rest_of_type_compilation output type info for
> dbx, xcoff, sdb and dwarf2. This appears to be restricted to enum,
> struct and union types.
Yup.
>
> Where is dwarf?
No clue.
>
> Just above, in rest_of_decl_compilation, we have debug output for
> TYPE_DECLs for dbx, xcoff and sdb.
Yup.
>
> Why is there no overlap with rest_of_type_compilation?
Because it was badly designed?
No idea.
> Where are
> dwarf / dwarf2 handled?
Well, for the most part, gen_decl_die in dwarf2out.c for DWARF2.
In reality, it'll output the types as needed (since most things go
through gen_decl_die, it's the main place the type outputting happens)
Global typedefs and whatnot happen in the global decl loop.
> In the global_decl loop?
>
> Because of the above, I'm not clear on exactly how to rationalise this
> into using a type_decl hook. It would be nice to have all the debug
> info treated in the same way at the same time. Is there a fundamental
> reason that is not possible here?
>
> Neil.
--
"Why is the alphabet in that order? Is it because of that song?
The guy who wrote that song wrote everything.
"-Steven Wright