This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: New problem compiling g-exctra.adb
- From: Diego Novillo <dnovillo at redhat dot com>
- To: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Cc: "gcc at gcc dot gnu dot org" <gcc at gcc dot gnu dot org>
- Date: Tue, 07 Sep 2004 20:58:27 -0400
- Subject: Re: New problem compiling g-exctra.adb
- Organization: Red Hat Canada
- References: <10409072308.AA29094@vlsi1.ultra.nyu.edu>
On Tue, 2004-09-07 at 19:08, Richard Kenner wrote:
> Yes. And in the case of a pointer initialized to 0, we should
> probably set it PT_ANYTHING, unless we can equate 0 with "can't point
> to a user variable".
>
> I'd say that zero means it can't point to a user variable.
>
I'm not well versed in standards and such, but if that's the case, you
want to create a name tag for such a pointer.
To solicit name tag creation, you have to tell create_name_tags() about
the pointer. Currently it creates name tags for pointers with PT_VARS
and pointers with PT_MALLOC.
One approach is to set the PT_MALLOC attribute on these pointers with
value 0. If you do that, things should just work. The name would
probably need to be changed, but PT_MALLOC has the semantics you are
looking for. It means 'points to a chunk of memory that cannot be used
by a regular user variable'.
Diego.