[llvm-commits] Edge dominance in the presence of multiple edges between two BB

[Eli Friedman]

On Fri, Aug 10, 2012 at 6:53 AM, Rafael Espíndola
<rafael.espindola at gmail.com> wrote:
>>> My understanding is that it is invalid, as %y1 is used in the unwind
>>> edge of the invoke.
>>
>> I would just say it's invalid because the normal destination has a
>> landing pad, without trying to reason about dominance.  That way, you
>> don't have to complicate the representation of an edge.
>
> I could do that, but having dominance handle it means that the logic
> is shared with gvn. Having said that, I noticed that we reject

The point is that we don't need to share the logic with GVN; if we
reject this case in the verifier, the existing dominance code is
sufficient everywhere else.

-Eli