This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: _G_USING_THUNKS + Sparc
- From: Jason Merrill <jason at redhat dot com>
- To: Mark Mitchell <mark at codesourcery dot com>
- Cc: "David S. Miller" <davem at redhat dot com>,"pme at sources dot redhat dot com" <pme at sources dot redhat dot com>,"libstdc++ at gcc dot gnu dot org" <libstdc++ at gcc dot gnu dot org>
- Date: Tue, 23 Apr 2002 16:38:53 +0100
- Subject: Re: _G_USING_THUNKS + Sparc
- References: <20020422.235924.75075329.davem@redhat.com><12980000.1019550264@gandalf.codesourcery.com>
>>>>> "Mark" == Mark Mitchell <mark@codesourcery.com> writes:
> --On Monday, April 22, 2002 11:59:24 PM -0700 "David S. Miller"
> <davem@redhat.com> wrote:
>> Why do we set this to zero on Solaris systems?
> We don't do the bit where libio's idea of a stream and the V3 idea
> are the same object anymore -- that hack doesn't work with the new
> C++ ABI. (It is difficult, even in principle, to make it work with
> the new ABI, but it is theoretically possible. Nobody has implemented
> it, though...)
Would it be that difficult? The streambufs only use single inheritance, so
I wouldn't expect the vtables to be any more complicated than in v2.
AFAIK, the main reason for separating the implementation was to avoid the
glibc/libstdc++ version lock problems we had in v2.
None of which is relevant to Solaris, of course. Nowadays, _G_USING_THUNKS
is meaningless; there's no reason for any of the config headers to set it.
Jason