[PATCH 1/6] Remove OP_LABELED.

[Tom Tromey]

>>>>> "Pedro" == Pedro Alves <palves@redhat.com> writes:

Pedro> Yep, that was the plan.  I'll take that as confirmation that designated
Pedro> initializers are not known to be, or should be working through some
Pedro> other means.  :-)  Thanks.

Yeah, sorry about this -- I actually thought about the C designated
initializer issue when ok'ing the series, but I neglected to mention
this in my review.

It isn't totally clear that we want to support this feature.  In C it is
just for initializers, not in arbitrary expressions.

In gdb terms you could only use it to assign to an existing struct
object, because the evaluator would need an "expected type".

So, it seemed marginal to me; and on top of that it is unclear whether
the old code is at all useful to C anyway -- hence my ok.

I don't object to a PR of course, and if somebody implements it I won't
reject it -- I think it is just marginal, but not truly bad :)

Tom