Array Access Region Analysis in LLVM?

Hi,

I am new to LLVM and was wondering if there exists a compiler pass
that implements array access region analysis (single and/or
multithreaded analysis) in the released version.

I took a look at:

http://llvm.org/docs/Passes.html

But I did not see anything directly, although certain passes such as
induction variable passes (in which the pass determines that the
induction variable used corresponds to a thread/processor id would be
one step in the direction).

If someone could indicate if such an analysis exists, and to what
functional degree I would appreciate it.

Thanks.

Malek