Linus Torvalds Uses AI in Linux Kernel Development
Linus Torvalds has used AI tools like Claude Code to assist in developing new Linux kernel features, including creating code, test programs, and draft cover letters under his direction.
Linus Torvalds' Use of AI in Kernel Development
Linus Torvalds, the creator of Linux, has been using AI tools such as Claude Code to assist in the development of new Linux kernel features. In a recent patch message, Torvalds confirmed that the code, test programs, and initial draft of the cover letter were created with the help of an LLM coding assistant under his direction. This marks a notable shift in how kernel development is being approached, integrating AI into the process for tasks like debugging and feature implementation. While Torvalds has previously used AI for kernel debugging and even experimented with Google's Antigravity outside the kernel tree, his current use of Claude Code highlights a growing trend of leveraging AI in complex software development. The use of AI in this context is attributed to Torvalds himself, with no universal guarantees or claims of testing beyond what is explicitly stated.
New Kernel Branch and AI Integration
Linus Torvalds is currently working on a new Linux kernel branch named noref-filemap-read-with-llm, which suggests the involvement of AI or large language models (LLMs) in the development process. The branch is focused on improving memory management by adapting the code to read from the page cache without relying on folio references. This change could potentially enhance performance and reduce overhead in memory-intensive operations. The patches associated with this effort are currently available on Torvalds' Git branch, as noted in the source. While the integration of AI in kernel development is a novel approach, the specifics of how LLMs are being utilized in this context remain unclear. The work is still in progress and has not yet been merged into the mainline kernel.
Performance Impact of the Patch Series
Linus Torvalds is working on a significant Linux kernel improvement aimed at addressing performance bottlenecks related to small reads of a single file from multiple threads. The patch series focuses on reducing contention in the folio reference counting mechanism, which currently causes inefficiencies when many threads access the same file. The goal is to enable reads from the page cache without requiring folio references, potentially improving concurrency and reducing overhead. According to measurements conducted on the developer's machine, the impact of this change is notable, with a marked difference in performance observed. While the exact magnitude of the improvement is not quantified, the results suggest that this patch series could have a meaningful effect on systems handling high-concurrency file access scenarios. The work is part of an ongoing effort to optimize the Linux kernel for modern workloads.
Sources and reporting notes
Based on published source material. No independent hands-on testing is claimed.
