Reactive Techniques for Controlling Software Speculation

Zilles, Craig; Neelakantam, Naveen

doi:10.1109/cgo.2005.30

Cited by 8 publications

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By not profiling the optimized code, JVMs miss opportunities to re-optimize their code as program behavior changes [29]. Furthermore, continuous profiling mitigates the downside of aggressive speculative optimizations; we can accurately track the ratio of correct to incorrect speculations and use that to guide recompilation [34].…”

Section: Introductionmentioning

confidence: 99%

Branch-on-random

Lee

Zilles

2008

Proceedings of the 6th Annual IEEE/ACM International Symposium on Code Generation and Optimization

View full text Add to dashboard Cite

We propose a new instruction, branch-on-random, that is like a standard conditional branch, except rather than specifying the condition on which the branch should be taken, it specifies a frequency at which the branch should be taken. We show that branchon-random is useful for reducing the overhead of program instrumentation, via sampling. Specifically, branch-on-random provides an order-of-magnitude reduction in execution time overhead compared to previously proposed software-only frameworks for instrumentation sampling.Furthermore, we demonstrate that branch-on-random can be cleanly architected and implemented simply and efficiently. For simple processors, we estimate that branch-on-random can be implemented with 20 bits of state and less than 100 gates; for aggressive superscalars, this grows to less than 100 bits of state and at most a few hundred gates.

show abstract

Section: Introductionmentioning

confidence: 99%