paradoxical subreg problem

[law@redhat.com]

In message <20020128115045.A26034@redhat.com>, Richard Henderson writes:
 > On Mon, Jan 28, 2002 at 12:26:13PM -0700, law@redhat.com wrote:
 > >  > That would be a bug.  A failure to call simplify_subreg or
 > >  > simplify_gen_subreg.
 > > I disagree given the above docs.
 > 
 > Given the problem you're seeing, wouldn't it be better to 
 > disallow it anyway and fix the docs?  I can't imagine what
 > good you could do by allowing such a thing.
Possibly -- I don't even known why it's allowed in this case.  There
may be a good reason for it, then again, there may not.  My gut
instinct is to disallow it, given the semantic issues that arise.

However, even if we do that I think we need to clarify precisely what
the semantics of a paradoxical subreg really are.    Clearly there is
some confusion.

[ Actually my preference would be to zap paradoxical subregs, but I
  don't think we're in a position to do that right now. ]

jeff