Close Encounters of Java Memory Model Kind

It's been two years since "Java Memory Model Pragmatics". Yet, even among those who studied that talk, many weird superstitions are left unresolved. In this talk, we will try to dissect and disband some of those superstitions: on the power of barriers, reorderings as mental model, semi-synchronization as performance optimization, and others.

This talk is based on already published article, and would include narrative and examples from there. We would not stop at the Java Memory Model basics, and therefore attendees are required to know the JMM at least at "JMM Pragmatics" level.