AUTOSAR E2E Protection Library

A high-performance, memory-safe Rust implementation of the AUTOSAR E2E (End-to-End) Protection Protocol for safety-critical automotive communication systems.

Overview

This library implements the AUTOSAR E2E protection mechanism which provides end-to-end data protection for safety-critical automotive communication. The E2E protection helps detect:

Features

Supported Profiles

Key Features

• Zero-copy operations - In-place data modification
• Thread-safe - All operations are safe for concurrent use
• High performance - Optimized common operations with shared helpers
• Memory safe - 100% safe Rust, no unsafe code
• Configurable - Extensive configuration options per AUTOSAR spec
• Well tested - Comprehensive test coverage including edge cases
• Well documented - Extensive API documentation with examples

Installation

Add this to your Cargo.toml:

[dependencies]

autosar-e2e = "0.6.4"

Quick Start

Basic Usage

use autosar_e2e::{E2EProfile, E2EResult, E2EStatus};
use autosar_e2e::profile4::{Profile4, Profile4Config};

fn main() -> E2EResult<()> {
    // Configure Profile 4 for large data packets
    let config = Profile4Config {
        data_id: 0x12345678,
        max_delta_counter: 1,
        min_data_length: 96,    // 12 bytes minimum
        max_data_length: 4096,  // 512 bytes maximum
        ..Default::default()
    };

    // Create sender and receiver instances
    let mut sender = Profile4::new(config.clone());
    let mut receiver = Profile4::new(config);

    // Prepare data buffer (12 bytes minimum for Profile 4)
    let mut data = vec![0u8; 16]; // [length, counter, data_id, crc, user_data...]

    // Add E2E protection
    sender.protect(&mut data)?;
    println!("Protected data: {:02X?}", data);

    // Simulate network transmission...

    // Verify E2E protection at receiver
    match receiver.check(&data)? {
        E2EStatus::Ok => println!("Data integrity verified!"),
        E2EStatus::CrcError => println!("CRC error detected!"),
        E2EStatus::DataIdError => println!("Data ID mismatch!"),
        E2EStatus::WrongSequence => println!("Counter sequence error!"),
        E2EStatus::OkSomeLost => println!("Some messages lost but within tolerance"),
        E2EStatus::Repeated => println!("Repeated message detected"),
        E2EStatus::DataLengthError => println!("Data length mismatch!"),
    }

    Ok(())
}

Advanced Configuration

use autosar_e2e::profile7::{Profile7, Profile7Config};

// High-integrity protection with 64-bit CRC
let config = Profile7Config {
    data_id: 0x0a0b0c0d,
    offset: 64,                    // Header at bit offset 64
    min_data_length: 20 * 8,       // 20 bytes minimum
    max_data_length: 4096 * 8,     // 4KB maximum
    max_delta_counter: 5,          // Allow up to 5 lost messages
};

let mut profile = Profile7::new(config);

Architecture

Clean Module Organization

src/
├── lib.rs              # Main library interface
├── profiles/           # All E2E profile implementations
│   ├── profile4.rs     # Large packets, 32-bit CRC
│   ├── profile4m.rs    # Profile 4 + message metadata
│   ├── profile5.rs     # Small packets, 16-bit CRC
│   ├── profile6.rs     # Dynamic size, 16-bit CRC
│   ├── profile7.rs     # High integrity, 64-bit CRC
│   ├── profile7m.rs    # Profile 7 + message metadata
│   ├── profile8.rs     # Flexible protection, 32-bit CRC
│   ├── profile11.rs    # Nibble/Both variants
│   └── profile22.rs    # Enhanced protection
└── profiles.rs         # Profiles module

Trait-Based Design

The library follows a clean trait-based design for extensibility:

pub trait E2EProfile {
    type Config;

    /// Create a new profile instance with configuration
    fn new(config: Self::Config) -> Self;

    /// Add E2E protection to data buffer (in-place)
    fn protect(&mut self, data: &mut [u8]) -> E2EResult<()>;

    /// Verify E2E protection on received data
    fn check(&mut self, data: &[u8]) -> E2EResult<E2EStatus>;
}

CI/CD Pipeline

Build & Test Status

Platform	Rust Version	Build	Test	Status
Linux	Stable	Pass	Pass
Windows	Stable	Pass	Pass
macOS	Stable	Pass	Pass

Automated Quality Checks

• Code Formatting: rustfmt ensures consistent code style
• Linting: clippy with zero warnings tolerance
• Security Audit: cargo-audit scans for known vulnerabilities
• Test Coverage: Comprehensive coverage reporting via tarpaulin
• Performance: Automated benchmark regression detection

Testing

Run the comprehensive test suite:

# Run all tests

cargo test


# Run with output

cargo test -- --nocapture


# Run specific profile tests

cargo test profile4


# Run tests in release mode for performance

cargo test --release

Test Coverage

# Generate coverage report

cargo tarpaulin --out Html


# View coverage

open tarpaulin-report.html

Current test coverage: 83% with 32 test cases covering:

• Basic protection/check cycles
• Counter wraparound scenarios
• Error detection (CRC, sequence, length)
• Edge cases and boundary conditions
• Configuration validation

Performance

The library is optimized for automotive real-time constraints with continuous benchmarking in CI:

Latest Benchmark Results

Benchmarks automatically run on every commit using GitHub Actions with Criterion.rs

Performance Scaling by Data Size

CI Performance Features

• Automated Benchmarking: Every commit triggers comprehensive performance tests
• Trend Analysis: Performance regression detection with 200% alert threshold
• Historical Data: Track performance changes over time via GitHub Actions
• Multiple Profiles: All E2E profiles benchmarked across different data sizes
• Cross-Platform: Benchmarks run on Ubuntu, Windows, and macOS
• Real-time Updates: Performance badges update automatically

The library demonstrates excellent scalability with sub-microsecond performance for typical automotive message sizes, making it suitable for real-time safety-critical systems. Performance regressions are automatically detected and reported in CI.

Configuration Examples

Profile Selection Guide

Sample Configurations

// Minimal overhead for CAN (Profile 5)
let can_config = Profile5Config {
    data_length: 8 * 8,    // 8 bytes total
    data_id: 0x123,
    max_delta_counter: 1,
    offset: 0,
};

// High-integrity Ethernet (Profile 7)
let eth_config = Profile7Config {
    min_data_length: 20 * 8,
    max_data_length: 1500 * 8,  // MTU size
    data_id: 0xdeadbeef,
    max_delta_counter: 10,      // Higher tolerance
};

Safety and Correctness

• Memory safety: 100% safe Rust, no unsafe code
• Static analysis: Passes clippy with zero warnings
• Fuzz tested: Robust against malformed inputs
• AUTOSAR compliant: Follows specification exactly
• Verified CRCs: Uses industry-standard polynomials
• Correct wraparound: Handles counter overflow properly

Security Considerations

// Data ID should be unique per message type
let config = ProfileConfig {
    data_id: 0x12345678,  // Good: unique per message
    // data_id: 0x00000000,  // Bad: not unique
};

// Configure appropriate counter tolerance
let config = ProfileConfig {
    max_delta_counter: 5,  // Good: allows some loss
    // max_delta_counter: 255, // Bad: too permissive
};

Roadmap

Completed

• Core E2E profiles (4, 5, 6, 7, 8, 11, 22)
• Comprehensive test coverage
• Code refactoring and optimization
• Documentation and examples
• Profile 7M implementation
• Profile 4M implementation
• Peformance benchmarks

Future

• code optimization

Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

Quick Start for Contributors

# Clone the repository

git clone https://github.com/handohun/autosar-e2e

cd autosar-e2e


# Run the complete CI pipeline locally

cargo bench                  # Run performance benchmarks

cargo test                   # Run all tests

cargo fmt --check           # Check code formatting

cargo clippy -- -D warnings # Run linter

cargo tarpaulin             # Generate coverage report

cargo audit                 # Security audit


# Generate documentation

cargo doc --open

Local Development Workflow

# Run benchmarks with HTML reports

cargo bench


# Run specific benchmark

cargo bench profile4


# Compare benchmark results

cargo bench -- --baseline main


# Watch for changes and run tests

cargo watch -x test


# Check all quality gates (like CI)

cargo fmt --check && cargo clippy -- -D warnings && cargo test && cargo audit

License

This project is dual-licensed under either:

• Apache License, Version 2.0 (LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0)
• MIT license (LICENSE-MIT or http://opensource.org/licenses/MIT)

at your option.

References and Resources

• AUTOSAR E2E Protocol Specification
• AUTOSAR CRC Routines Specification
• AUTOSAR Classic Platform
• Rust Embedded Working Group

Disclaimer

This is an independent implementation and is not officially affiliated with or endorsed by AUTOSAR GbR. The implementation follows the publicly available AUTOSAR specifications but has not undergone official AUTOSAR certification.

For production safety-critical systems, please ensure appropriate validation and testing according to your functional safety requirements (ISO 26262 or similar).

Made with care for the automotive industry

Report Bug • Request Feature • Discussions