Per discussion with Dan G, inbounds geps *certainly* can have

Per discussion with Dan G, inbounds geps *certainly* can have
unsigned overflow (e.g. "gep P, -1"), and while they can have
signed wrap in theoretical situations, modelling an AddRec as
not having signed wrap is going enough for any case we can
think of today.  In the future if this isn't enough, we can
revisit this.  Modeling them as having NUW isn't causing any
known problems either FWIW.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125410 91177308-0d34-0410-b5e6-96231b3b80d8