PROFILE-GUIDED RECEIVER CLASS PREDICTION

The use of dynamically-dispatched procedure calls is a key mechanism f or writing extensible and flexible code in object-oriented languages. Unfortunately, dynamic dispatching imposes a runtime performance penal ty. Some recent implementations of pure object-oriented languages have utilized profile-guided receiver class prediction to reduce this perf ormance penalty, and some researchers have argued for applying receive r class prediction in hybrid languages like C++. We performed a detail ed examination of the dynamic profiles of eight large object-oriented applications written in C++ and Cecil, determining that the receiver c lass distributions are strongly peaked and stable across both inputs a nd program versions through time. We describe techniques for gathering and manipulating profile information at varying degrees of precision, particularly in the presence of optimizations such as inlining. Our i mplementation of profile-guided receiver class prediction improves the performance of large Cecil applications by more than a factor of two over solely static optimizations.