Homeostasis: Design and Implementation of a Self-Stabilizing Compiler

Mainstream compilers perform a multitude of analyses and optimizations on the given input program. Each analysis pass may generate a program-abstraction. Each optimization pass is typically composed of multiple alternating phases of inspection of program-abstractions and transformations of the program. Upon transformation of a program, the program-abstractions generated by various analysis passes may become inconsistent with the program's modified state. Consequently, the downstream transformations may be considered unsafe until the relevant program-abstractions are stabilized, i.e., the program-abstractions are made consistent with the modified program. In general, the existing compiler frameworks do not perform automated stabilization of the program-abstractions and instead leave it to the optimization writer to deal with the complex task of identifying the relevant program-abstractions to stabilize, the points where the stabilization is to be performed, and the exact procedure of stabilization. Similarly, adding new analyses becomes a challenge as one has to understand which all existing optimizations may impact the newly added program-abstractions. In this paper, we address these challenges by providing the design and implementation of a novel generalized compiler-design framework called Homeostasis. Homeostasis can be used to guarantee the trigger of automated stabilization of relevant program-abstractions under every possible transformation of the program. Interestingly, Homeostasis provides such guarantees not only for the existing optimization passes but also for any future optimizations that may be added to the framework. We have implemented our proposed ideas in the IMOP compiler framework, for OpenMP C programs. We present an evaluation which shows that Homeostasis is efficient and easy to use.