With LLVM 3.1 it was the first release where Polly became an official LLVM project for being an experimental optimizer for data locality and parallelism. Polly can be loaded directly into Clang and there was an automatic scheduling optimizer as well for performing loop transformations. Since the LLVM 3.1 release, there's been more Polly improvements building up for LLVM 3.2 this December.
In July, the Polly Polyhderal optimizer was modified so that it could be directly linked to the PLUTO optimizer. The PLUTO Compiler project page describes itself as "an automatic parallelization tool based on the polyhedral model. The polyhedral model for compiler optimization provides an abstraction to perform high-level transformations such as loop-nest optimization and parallelization on affine loop nests. Pluto transforms C programs from source to source for coarse-grained parallelism and data locality simultaneously. The core transformation framework mainly works by finding affine transformations for efficient tiling and fusion, but not limited to those."
Polly could already handle PLUTO-like optimizations with a similar algorithm having existed within the project's code-base, but now PLUTO can be directly linked to from Polly. "Being able to directly compare with the original implementation will not only bring in competition in the optimizer field. It will also allow new experiments with a cutting edge research tool."
Another milestone for Polly happened last month when experimental support was committed for the new ISL code generator. The new ISL code generator gives direct control over items like loop unrolling, control overhead, and code size. ISL was also relicensed to MIT in the process, but does depend upon the GPL-licensed GMP library. However, developers do hope to replace the GPL GMP code with something more liberally licensed. "The use of GMP is limited to a well defined interface. Replacing it with a BSD licensed replacement is a tractable engineering project we would be very interested in."
The Integer Set Library (ISL) work is described in the manual (PDF). "isl is a thread-safe C library for manipulating sets and relations of integer points bounded by affine constraints. The descriptions of the sets and relations may involve both parameters and existentially quantified variables. All computations are performed in exact integer arithmetic using GMP. The isl library offers functionality that is similar to that offered by the Omega and Omega+ libraries, but the underlying algorithms are in most cases completely different."
More information on the LLVM Polly work can be found via its project page.