ajitjadhav28/kernel-bench repository overview

⁠Linux Kernel Extraction & Build Benchmark Dockerfile

This repository provides a minimal Docker setup to benchmark Linux kernel source extraction and kernel compilation performance.

Only the compressed kernel source tarball (linux-6.12.10.tar.xz) is stored in the Docker image. When the container runs, both extraction and compilation are performed and timed at runtime.

⁠Docker Hub Image

The pre-built benchmark image is available on Docker Hub:

• ajitjadhav28/kernel-bench:6.12.10
• ajitjadhav28/kernel-bench:latest

⁠Features

• Platform Target: Configurable via --platform (defaults to linux/amd64).
• Standardized Build Target: Uses vmlinux as the default build target across all platforms for identical, fair benchmarking.
• Runtime Tarball Extraction: Measures the time required to un-tar the Linux kernel source archive (/usr/src/linux.tar.xz).
• Runtime Compilation: Measures kernel configuration & parallel compilation time.
• Unrestricted Memory by Default: Runs without memory restrictions by default (RAM limit can be set via MEMORY_LIMIT).
• Combined Benchmark Summary: Displays breakdown of:
  1. Extraction Time (seconds)
  2. Compilation Time (seconds)
  3. Total Runtime (Extraction + Compilation)
  4. CPU utilization & Max RAM consumption (RSS)

⁠Quick Start

⁠1. Run Benchmark directly from Docker Hub

docker run --rm --platform=linux/amd64 ajitjadhav28/kernel-bench:6.12.10