[patch] fix pr22550

[Diego Novillo]

On Tue, Jul 26, 2005 at 11:33:20AM -0600, Jeffrey A Law wrote:

> I think the big question we need to answer is whether or not
> merging blocks with these kind of degenerate PHIs is really
> worth the effort.
> 
Yes.  My opinion is that it is not worth the effort.

In principle, the CFG manipulation functions should not be
dealing with data flow.  As this and another PR show, they are
not well equipped to handle all the consequences.

I will be testing this approach over the next couple of days.

> > I forgot about the third alternative, which is to just make
> > optimizers handle no-op conditionals (as in James' original
> > patch).
> Which is also a possibility.  The downside is whatever compensation
> code we need would probably be needed in a variety of optimizers.
> This may open us up to a long term maintenance headache.
> 
Exactly my thoughts.