This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: libstc++-v3 vector not convertible to pointer
- To: gdr at codesourcery dot com (Gabriel Dos Reis)
- Subject: Re: libstc++-v3 vector not convertible to pointer
- From: Joe Buck <jbuck at racerx dot synopsys dot com>
- Date: Thu, 30 Nov 2000 13:26:47 -0800 (PST)
- Cc: aoliva at redhat dot com (Alexandre Oliva), nbecker at fred dot net, gcc at gcc dot gnu dot org, oon-list at oonumerics dot org
> Alexandre Oliva <aoliva@redhat.com> writes:
>
> [...]
>
> | ... IIRC, vectors have to be contiguous, even
> | though the Standard doesn't explicitly state that.
>
> Strictly speaking there was no such requirement in the Standard.
> However that is issue is addressed in the first TC
> (yet-to-b-=published).
TC == "technical correction", I believe.
Right, when it was pointed out all the standards committee people realized
it was an accident; it was intended all along to be a requirement that
the elements of vector are contiguous, which means that &v[0] or &*v.begin()
can be passed to a C routine that wants a T* that points to v.size()
consecutive elements.