This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Problem with "<<" operator
- To: Joe Buck <jbuck at racerx dot synopsys dot com>
- Subject: Re: Problem with "<<" operator
- From: Alexandre Oliva <aoliva at redhat dot com>
- Date: 25 Aug 2000 16:49:18 -0300
- Cc: pfrancq at ulb dot ac dot be (Pascal Francq), gcc at gcc dot gnu dot org
- Organization: GCC Team, Red Hat
- References: <200008251932.MAA00634@racerx.synopsys.com>
On Aug 25, 2000, Joe Buck <jbuck@racerx.synopsys.com> wrote:
>> On Aug 25, 2000, Joe Buck <jbuck@racerx.synopsys.com> wrote:
>>
>> > I believe that the C++ standard now requires an explicit & to indicate
>> > a pointer-to-function
>>
>> Nope, `&' is only required for pointer-to-members. A function& decays
>> to function* and, from that, I believe it decays to bool.
> A name of a function can be converted to a pointer to function. But can
> it be silently converted to a bool?
[conv.bool] allows pointers to be converted to bool. It doesn't
exclude pointers to functions. Since function-to-pointer is an lvalue
transformation, it can be followed by a boolean conversion. Other
implicit conversion sequences not involving a boolean conversion would
be preferred, but that doesn't make the sequence invalid for purposes
of overload resolution. I think it's a bug in aCC.
--
Alexandre Oliva Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer aoliva@{cygnus.com, redhat.com}
CS PhD student at IC-Unicamp oliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist *Please* write to mailing lists, not to me