[patch] for PRs 27639 and 26719

[Eric Botcazou]

> I do not see anything principially wrong with the code that ivopts
> produces; I think this is simply a VRP bug.

I disagree: IVOPTs should preserve the semantics of subtypes, that is to say 
the arithmetics operations are performed in the base type and the result is 
casted to the subtype.  IOW, it should not generate IVs in subtypes.  IIUC 
it's Jeff's position too: http://gcc.gnu.org/ml/gcc/2006-02/msg00590.html

> Both the invalid range <2,0x80000000> and the "integer" type of the upper
> bound of the range of ivtmp.86D.1158_14 seem suspicious to me, and probably
> closer to the real reason for the problem than the changes in scev
> analysis.

I think they are only the symptoms, not the cause of the problem.

-- 
Eric Botcazou