Developer Documentation

This section provides technical documentation for developers contributing to Revizor.

Development Guidelines

General Guidelines: Development environment setup, testing procedures, contribution workflow
Code Style: Formatting conventions for Python and C code, naming conventions
Git Workflow: Branch management, commit message format, merge procedures

Overview: High-level system architecture and component interaction
Code Structure: Organization of the source code directory and key modules
Orchestration: Main fuzzing loop and coordination between components
ISA Specification: Instruction set architecture definitions and JSON-based specification format
Test Case Code Generation: Program generation algorithm and relevant classes
Test Case Data Generation: Data generation algorithm and relevant classes
Hardware Tracing: Execution of test cases on the target HW and hardware trace collection
Contract Tracing: Leakage modeling and contract trace generation (high-level overview; implementation details in backend-specific pages)
Trace Analysis: Comparison of contract and hardware traces to detect violations
Minimization: Post-detection reduction of test cases to minimal reproducing examples
Logging: Logging infrastructure and debugging facilities

Revizor supports two different backends for contract-based leakage modeling. They are documented in the following pages:

Unicorn Backend: Backend based on the Unicorn CPU emulator
DynamoRIO Backend: Backend based on the DynamoRIO dynamic binary instrumentation engine

Register Allocation: Executor reserves a subset of registers for its own use; this page documents their purpose
Test Case Sandbox: Memory layout of the sandbox environment in which test cases are executed
Macros: Implementation of macros in Executor and Models
Binary Formats: Serialized binary formats for test case programs and data