An Ideal Cache Model Implemented in C++

Authors:

• Wei Song (SKLOIS, Institute of Information Engineering, Chinese Academy of Sciences)

Cotributors:

• Zhenzhen Li (SKLOIS, Institute of Information Engineering, Chinese Academy of Sciences)
• Boya Li (SKLOIS, Institute of Information Engineering, Chinese Academy of Sciences)

Features

• A pure C++ (std-11) and modular design.
• A hierarchical cache supporting the MSI coherence protocol.
• An ideal cache model:
  • Universal access latency: Accessing a cache block has the same latency disregarding to its location (level, set, way or slice) and status (hit or miss).
  • Ideal hit/miss status: Search algorithms can directly and accurately inquire the status of a cache block (hit or miss) with no time penalty.
  • No TLB noise: Virtual addresses are translated into physical addresses before cache accesses.
• Supporting user defined cache set mapping, tagging, replacement functions.
• Supporting the Intel complex addressing scheme.
• Sophisticated tracing and reporting system.

References

• Wei Song. 支持一致性缓存的Spike仿真器. CRVA联盟技术研讨会, 中国北京, 2020年7月18日. [PDF, Video]
• Wei Song and Peng Liu. Dynamically finding minimal eviction sets can be quicker than you think for side-channel attacks against the LLC. In Proc. of the International Symposium on Research in Attacks, Intrusions and Defenses (RAID), Beijing, China, pp. 427–442, September 2019. [Web]
• Zhenzhen Li and Wei Song. 升级RISC-V的指令级仿真器Spike的缓存模型. In 中国RISC-V论坛, 中国深圳, 2019年11月13日. [PDF]

License

GPL.

Notice

For the model used in the RAID-2019 paper, refer to the raid-2019 branch.