This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: [cxx-abi-dev] Re: C++ ABI: Reuse of tail padding
- From: Jim Dehnert <dehnert at transmeta dot com>
- To: Nathan Sidwell <nathan at codesourcery dot com>
- Cc: Mark Mitchell <mark at codesourcery dot com>, cxx-abi-dev at codesourcery dot com, gcc at gcc dot gnu dot org
- Date: Thu, 15 Aug 2002 21:06:43 -0700
- Subject: Re: [cxx-abi-dev] Re: C++ ABI: Reuse of tail padding
- Organization: Transmeta Corp.
- References: <151750000.1029358418@warlock.codesourcery.com> <3D5AC85B.8E510A5A@codesourcery.com>
Nathan Sidwell wrote:
>
> Mark Mitchell wrote:
>
> > We have a problem here.
>
> > Intel's compiler does not pack things into the tail padding.
> > Neither, I expect, does HP's.
> I would like to get hard data from more than the two compilers
> you've verified.
>
> > Technically, the G++ version is superior. (It reduces space used by
> > objects.) On the other hand, it was basically impossible for
> > implementors to figure out what was intended from the spec.
> >
> > I think I would prefer to change G++, and drop this idea from the spec,
> > even though it is an optimization.
>
> Although using tail padding is neat, how many bytes would it
> save in real programs? You only get it with structs of the form
> int i;
> char c;
> or (on some non-i86 machines)
> double d;
> int i;
That's true. We did, however, have people during the meetings (I
don't recall who) who felt strongly about reusing the tail padding as
suggested. It was intended to be done that way -- I apologize for not
making it clearer in the writeup.
Jim
--
- Jim Dehnert dehnert@transmeta.com
(408)919-6984 dehnertj@acm.org