Meeting agenda

Jim Dehnert dehnert at baalbek.engr.sgi.com
Thu Feb 24 07:09:54 UTC 2000


Time to clarify...

It's amazing how much goes by when one's email is out of commission for
a couple of days.  Obviously, a couple of issues (at least) have become
a significant concern -- vcall offsets and virtual function calls, and
pointers to members.  At best, there's significant confusion.  I'd like
to use this meeting to slog through these as first priority.  Since the
vcall discussions are best done with pictures, it would be helpful if
Christophe and Jason were physically present (though if you can't, you
can't...).  Mark, your participation would probably be helpful too,
even just on the phone.

Following is an updated agenda.  The status pages are updated (as of
last Friday) and on the web at:

	http://reality.sgi.com/dehnert_engr/cxx/cxx-summary.html

The contact info page is now also there, in password-protected PDF only
for privacy.  The password is "not an orc".

The agenda is based on these priorities:

  A) Data layout and virtual function calls.

  B) Mangling.

  C) Exception handling.

So, take a look at the material associated with these issues:

  1) Quickie -- are the changes to the Data Layout document in red (e.g.
     ABI namespace writeup) OK?  I will make further clarifications in
     the excess-size bitfield writeup for Mark, and the ctor vtable and
     mangling discussions are still open.

  2) B-1:  I would like to work through the virtual function call
     protocol very carefully, so I understand it well enough to write
     it up (which will probably mean everyone else understands it well,
     too :-).  I see Mark found the description unclear, and Christophe
     doesn't see what he thought was there, so let's work through the
     options once more.  Let's try to understand both the mechanisms
     and the associated costs first, then deal with decisions.

  3) A-14:  Pointers-to-members -- clarify intent, and make necessary
     choices.

  4) C-4:  Construction vtables -- discuss Coleen's proposal.

  5) F-*:  Mangling -- How do we proceed with issues?

Issues that are active and awaiting proposals:

  a)  A-24:  Incomplete type RTTI (alternate proposal from Christophe).

  b) D-12:  Unwind table location.

Jim

-		Jim Dehnert  x3-4272




More information about the cxx-abi-dev mailing list