This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: C++ ABI Issues
- From: Per Bothner <per at bothner dot com>
- To: David Edelsohn <dje at watson dot ibm dot com>
- Cc: mark at codesourcery dot com, gcc at gcc dot gnu dot org
- Date: Mon, 26 Aug 2002 20:40:31 -0700
- Subject: Re: C++ ABI Issues
- References: <200208270224.WAA31450@makai.watson.ibm.com>
David Edelsohn wrote:
As was discussed during the GCC 3.2 transition, my personal
opinion is that GCC needs a binary -fabi-correctly switch which defaults
to *OFF*. We maintain the GCC 3.2 ABI until a well-defined transition
point in the future when the switch is changed to default to *ON*. This
allows GCC stability for the time being and interoperability with
proprietary compilers outside the GNU/Linux and *BSD environments.
Propritary compilers which run on GNU/Linux will probably have a GCC
compatibility switch.
I agree. The default should be to keep the current ABI. The default
should be no warnings. What does it gain us or our users to change the
ABI again at this point? What does it gain us or our users to warn
about ABI incompatibility (unless explicitly requested)? Nothing.
Our policy at this position should be that the ABI is done, for now.
At some point we may feel the need to make binary incompatible
changes. My guess is those may be library-related, or there may
be some feature that we want that can't be implemented without
breaking compatibility. However, compliance with some external
specifcation is not a sufficient reason for changing, unless it is
really important to be compatible. And it isn't.
Of course an *option* -fabi-correctly is good, but should not be the
default, at this point, or anytime soon.
--
--Per Bothner
per@bothner.com http://www.bothner.com/per/