Re: [boost] variant visitation assertion Failure

From: Eric Friedman (ebf_at_[hidden])
Date: 2006-01-11 20:02:58

Hi Felipe,

The reason the code should never reach the point you indicate is because 
the which_ index of the variant should always be valid, and so it should 
always be the case that the other specialization of visitation_impl 
finds a match. If this is not happening, which appears to be the case, 
it may be due to memory corruption that is overwriting the which_ field 
of the variant, or it may be due to a bug in the variant code. Without 
more detail from you about your problem though, I won't be able to help 
more.

Regards,

Eric

Felipe Magno de Almeida wrote:
> Hi!
> 
> I'm having an assertion failure at visition_impl.hpp:201
> But the assertion doesnt state what could be wrong, there's only a
> comment saying: // should never be here at runtime:
> 
> Here is the function that is being called, that shouldnt...
> 
> template <
>       typename W, typename S
>     , typename Visitor, typename VPCV
>     , typename NBF
>     >
> inline
>     BOOST_VARIANT_AUX_GENERIC_RESULT_TYPE(typename Visitor::result_type)
> visitation_impl(
>       int, int, Visitor&, VPCV
>     , mpl::true_ // is_apply_visitor_unrolled
>     , NBF, W* = 0, S* = 0
>     )
> {
>     // should never be here at runtime:
>     BOOST_ASSERT(false);
>     typedef typename Visitor::result_type result_type;
>     return ::boost::detail::variant::forced_return< result_type >();
> }
> 
> 
> Could anyone help me?
> Thanks in advance,
> 
> --
> Felipe Magno de Almeida
> 
> _______________________________________________
> Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
>