Monday, 15:15 - 15:40 h, Room: H 3003A


Roberto Castaneda Lozano
Robust code generation using constraint programming

Coauthors: Mats Carlsson, Frej Drejhammar, Christian Schulte


Code generation in a compiler transforms an intermediate program
representation into assembly code for a particular
architecture. It has tremendous impact on the resulting code:
optimal assembly code can be several times more efficient than
naive assembly code. However, optimization techniques are
typically not used for code generation as they are considered as
non-scalable. Instead, traditional optimizing compilers
compromise code quality by addressing code generation with
heuristic algorithms and phase decomposition.
Two central phases in code generation are instruction scheduling
and register allocation which are strongly interdependent. In
this presentation, we introduce combinatorial models that
naturally capture both phases. These models are easier to analyze
and reuse than traditional heuristic algorithms. Then, we present
an integrated model that captures the dependencies between phases
and hence enables the generation of possibly optimal
code. Finally, we illustrate why constraint programming with
features such as flexible search and global constraints is a good
candidate for robustly solving code generation problems.


Talk 1 of the invited session Mon.3.H 3003A
"Constraint programming standard and industrial applications" [...]
Cluster 5
"Constraint programming" [...]


  cash advance . But at the same time, it acts only with sexual arousal. Viagra has a number of advantages in comparison with injections in the sexual organ or other procedures aimed at treatment of impotency.