Home

Speculator Documentation

Speculator is a tool for studying speculative execution through Performance Counters. It allows to easily create proof-of-concepts implementations of speculative execution attacks and tests to verify their mitigations. Furthermore, speculator allows very precise measures (u-ops level) of events which gives the ability to reverse engineer speculative execution behavior. Using performance counters makers, it is possible to deterministically observe speculative execution without rely on very noisy and convoluted side-channels.

In this wiki, you can find all the information related to speculator:

• Design: Basic info on how tests are designed and run within the tool.
• Project Structure: Describe the repository structure.
• Installation Requirements: Gives the info to the basic requirements.
• Build Information: Gives info on speculator cmake specific options and how to build the project.
• speculator_mon usage: Man page for speculator_mon for manually run tests.
• Add new test[s]: Describe how to add write/generate new tests.
• post_processing usage: Describe how to generate and interpret the final report
• Examples: Describe how to run the already added examples
• FAQ: Answers to common problems in getting speculator to work