[vta, graphite?] propagate degenerate phi nodes into debug stmts

[Richard Guenther]

On Fri, Nov 20, 2009 at 10:25 AM, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Nov 19, 2009, Richard Guenther <richard.guenther@gmail.com> wrote:
>
>> Ah, hm.  The num_ssa_operands and delink_stmt_imm_use
>> changes were necessary because they triggered the assert?
>
> Yup.
>
>> I wonder if the callers were expecting them to be a no-op...
>
> That will be something to be looked into.
>
>> The tre-ssa-pre.c change is ok.  I think we should rather
>> let num_ssa_operands and delink_stmt_imm_use ICE on PHIs,
>> but I'd rather do this in stage1 - can you queue this patch until then?
>
> You mean queue it all up, including the tree-ssa-pre.c change?

No, the tree-ssa-pre.c change is fine anyways.

Richard.