Improved Loop Distribution in LLVM Using Polyhedral
Dependences
Author/Presenters
Event Type
Workshop
Compiler Analysis and Optimization
Compilers
Debugging
Parallel Programming Languages, Libraries, Models
and Notations
Program Transformation
SIGHPC Workshop
TimeMonday, November 13th5:14pm -
5:21pm
Location710
DescriptionWe propose a framework that can be used for improving
loop-optimizations in LLVM using the Polyhedral
framework of Polly. In our framework, we use the precise
polyhedral dependences from Polly (provided by
PolyhedralInfo), to construct a dependence graph, and
perform loop transformations. As the first
transformation case study of such a framework, we
implemented loop distribution targeting improvement of
inner-loop vectorization. Our loop distribution pass
shows promising results on the TSVC benchmark; it is
able to distribute 11 loops, while the earlier
distribution pass is unable to distribute at all. We
also have preliminary performance numbers from SPEC
2006. We believe that our work is the first step towards
scalable and pre-defined loop-transformations in LLVM
using exact dependences from Polly.
Author/Presenters
