Using Contaminated Garbage Collection and Reference Counting Garbage Collection to Provide Automatic Storage Reclamation for Real-Time Systems

USING CONTAMINATED GARBAGE COLLECTION AND REFERENCE COUNTING GARBAGE COLLECTION TO PROVIDE AUTOMATIC STORAGE RECLAMATION FOR REAL TIME SYSTEMS by Matthew P Hampton ADVISOR Dr Ron K Cytron May Saint Louis Missouri Language support for dynamic storage management simpli es the application programming task immensely As a result dynamic storage allocation and garbage collection have become common in general purpose computing Garbage collection research has led to the development of algorithms for locating program memory that is no longer in use and returning that unused memory to the run time system for later use by the program While many programming languages have adopted automatic memory recla mation features this has not been the trend in Real Time systems Many garbage collection methods involve some form of marking the objects in memory This mark ing requires time proportional to the size of the heap to complete As a result the predictability constraints of Real Time are often not satis ed by such approaches In this thesis we present an analysis of several approaches for program garbage collection We examine two approximate collection strategies Reference Counting and Contaminated Garbage Collection and one complete collection approach Mark and Sweep Garbage Collection Additionally we analyze the relative success of each approach for meeting the demands of Real Time computing In addition we present an algorithm that attempts to classify object types as good candidates for reference counting Our approach is conservative and uses static analysis of an application s type system Our analysis of these three collection strategies leads to the observation that there could be bene ts to using multiple garbage collectors in parallel Consequently we address challenges associated with using multiple garbage collectors in one appli cation