[PATCH] Improve PR30911 and PR31023, Ada and VRP

[Richard Kenner]

> Now what I ask the Ada guys for:  please define the semantics
> of sub-types for the middle-end.  Stick to _one_ semantics, independent
> of whatever mode the Ada frontend is in, and adjust the Ada frontend
> to emit code that is safe under these semantics.

I've got to run to a plane, but I want to add one more thing.

Right now, we *do* have very close to the exactly-needed semantics. 
Most of the cases where it's not exact are very pathological corner cases
of the language, about semantics that even language designers aren't sure
of.  There are a few cases that occur in real programs (mostly the use
of 'Valid) and I'd much prefer to fix the 5% that doesn't work correct
than throw away a mechanism that works right 95% of the time.