Flexible Hardware Acceleration for Instruction-Grain Lifeguards

Instruction-grain lifeguards monitor executing programs at the granularity of individual instructions to quickly detect bugs and security attacks, but their fine-grain nature incurs high monitoring overheads. This article identifies three common sources of these overheads and proposes three techniques that together constitute a general-purpose hardware acceleration framework for lifeguards.