Optimize CI/CD pipeline with improved caching strategies

[lnbc1QWFyb24]

Summary

This PR introduces improvements to the build and CI/CD pipeline by implementing better caching strategies and optimizing the Docker build process. These changes will reduce build times and improve resource utilization across both GitHub Actions workflows and Docker builds.

Changes

GitHub Actions Improvements

• Replaced manual cargo caching with Swatinem/rust-cache: The workflows now use a specialized Rust caching action that provides more intelligent and efficient caching compared to the previous manual approach
• Enabled failure-resilient caching: Set cache-on-failure: "true" to preserve cache even when builds fail, preventing unnecessary cache rebuilds
• Disabled target caching: Set cache-targets: "false" to focus on dependency caching rather than build artifacts

Docker Build Optimization

• Implemented multi-stage build with cargo-chef: The Dockerfile now uses cargo-chef to create cacheable layers for dependencies, significantly reducing rebuild times when only source code changes
• Optimized runtime image: Switched from debian:bookworm to debian:bookworm-slim for a smaller final image size
• Removed BuildKit cache mounts: Replaced with cargo-chef's layer-based caching approach for better Docker layer reusability

Code Formatting

• Minor formatting improvement in src/interrupts.rs for cargo fmt