This is the mail archive of the
mailing list for the GCC project.
Re: GCC Cauldron: Notes from the C++ ABI BOF
- From: Jonathan Wakely <jwakely dot gcc at gmail dot com>
- To: Jakub Jelinek <jakub at redhat dot com>
- Cc: Michael Matz <matz at suse dot de>, Ian Lance Taylor <iant at google dot com>, Cary Coutant <ccoutant at google dot com>, gcc <gcc at gcc dot gnu dot org>
- Date: Thu, 16 Aug 2012 09:10:13 +0100
- Subject: Re: GCC Cauldron: Notes from the C++ ABI BOF
- References: <CAHACq4oSX8a730MtDgQwyZBw1Z7dqMD-pJc2PoWTBfz5k6HL_A@mail.gmail.com> <CAKOQZ8w53hkfLdfW5w=ag+7vUzGF_JR-DK4X0N=xixHok08eYg@mail.gmail.com> <alpine.LNX.firstname.lastname@example.org> <20120720114353.GA26330@sunsite.ms.mff.cuni.cz> <CAH6eHdR+rnR9eb+aj32C_jmEoHvTGRH56Ss88rMG0sUX=WVVrw@mail.gmail.com>
On 20 July 2012 14:04, Jonathan Wakely wrote:
> On 20 July 2012 12:43, Jakub Jelinek wrote:
>> using a new std::list and std::pair, or the old one (well, std::pair is a
>> bad example probably, std::string and std::list is good, what other types we
>> are going to change?).
> I need to add a new virtual function and rename an existing virtual
> function for a base class in <future>. That's for C++11 types only,
> but doing the change will create an incompatibility with previous
> releases. (I don't know whether it's better to just make that change
> while we still call C++11 support experimental or to wait and use the
> new type attribute to make it a different type with the "_cxx11" tag.)
> I think there are some changes needed to the hierarchy of exception
> classes, to add std::system_error as a base class of std::ios_failure.
> I don't know if adding move semantics to iostream classes can be done
> without ABI changes (I haven't looked into it.)
I've just remembered that std::reverse_iterator has been in need of an
ABI change for years now, see LWG DR 198 and PR 51823.