This is the mail archive of the libstdc++@gcc.gnu.org mailing list for the libstdc++ project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch] : Improve std::search on v7

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Paolo Carlini wrote:
> 
> As I said, there is no point in working out those details, because we
> don't know how to deal with unions. 

(sorry about my last e-mail... not sure how it got half-eaten..)

I haven't thought about unions in a very long time. I'm going to have a
look now. I can see now how EBO can go wrong, and also how to fix it
(same as vector already does). Thank you!

I can tell you how I'd like to deal with unions, which is by dispatching
based on tr1::is_union<_Comp> ;) However, until that comes along, it
might be necessary to come up with a temporary solution.

Out of interest, why isn't this union business also a problem for
allocators on the standard containers? Are they just not allowed to be
unions?

Chris
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)

iD8DBQFCta3b3FpDzErifpIRAmgXAJ0aBYYzLv5KapzsKbGXFdrtjXo2egCfZH24
qQjSpTw9GIncFL1mUG0yOik=
=C1V3
-----END PGP SIGNATURE-----
Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]