kernel

A tweet shares a free book called 539kernel that teaches building an entire operating system from the x86 boot sector at address 0x7C00, progressing from the first instruction to a working kernel.

Linus Torvalds describes Linux Kernel 7.2 as "huge" because AI tools are now reviewing kernel code faster than humans, despite his earlier criticism of low-quality AI-generated patches and bug reports. He accepts this as the new normal, noting Linux is not an anti-AI project.

ATFlash introduces a per-RoPE-wavelength distance window that prunes query-key inner-product terms proportional to each frequency pair's wavelength, cutting 37-48% of attention compute with minimal quality loss and up to 1.31x speedups on long-context LLM inference.

Cursor is open-sourcing Mixture-of-Kittens (MoK), a production MoE training megakernel for NVL72s that fuses communication and computation, delivering a 1.41x end-to-end training throughput improvement for their Composer model.

A new AMD Linux kernel patch improves low-end gaming performance on Steam Deck by monitoring core workloads and temporarily setting busy cores to performance mode, resulting in up to 31.8% improvement in 1% low FPS in Civilization VI.

The article discusses the integration of AI tools like LLMs in Linux kernel development, highlighting Linus Torvalds' support for AI-assisted code and the controversy around the new Sashiko code review tool, while critiquing Torvalds' dismissal of ethical concerns.

This article demonstrates how to profile eBPF code performance by creating a simple C test harness that measures file open latency, enabling developers to compare overhead before and after attaching eBPF hooks.

Apple's MIE exploitation challenge details two vulnerabilities in WebDAV and SMBClient that allow kernel exploitation, with a call for community solutions before Black Hat USA presentation.

An analysis of Linus Torvalds' unique role as a leader who stopped writing code early to focus on coordination and direction of the Linux kernel project, contrasting with other maintainers.

An experimental Linux driver has been developed for the Fairphone 6's ultra-wide camera, enabling QR scanning capability on mainline Linux via the qcom-camss driver and libcamera.

432 Linux kernel CVEs were published in the last 24 hours, indicating a significant batch of security vulnerabilities.

Introduces pve-microvm, a Debian package that integrates QEMU's microvm machine type into Proxmox VE, enabling sub-300ms boot times and hardware isolation with minimal overhead, supporting various guest OS types.

A filtering engine and database that scans public kernel commits to identify security patches that have not yet been propagated to stable branches, helping developers track potential vulnerabilities.

LiteTopK is a fused indexer-top-k kernel for long-context sparse attention that exploits the curse of dimensionality to reduce memory traffic and improve efficiency, accelerating GLM 5.2 by 1.2x during prefill.

A modern Rust rewrite of the Linux 0.11 kernel that boots in QEMU on i386, featuring a full userland with shell and coreutils, with one-command image building and devcontainer support.

This article describes how researchers at Praetorian used Claude Opus (via Claude Code) to discover and exploit vulnerabilities in the FreeBSD kernel, including a stack overflow (CVE-2026-3038) that allows escape from FreeBSD jails. Part one focuses on methodology for finding bugs.

A developer shares their day 6 of inference engineering, writing a CUDA kernel for 1D convolution, explaining PagedAttention's memory efficiency, and providing an overview of GPU memory hierarchy (global, register, local, constant, shared).

This article provides a detailed walkthrough of what happens when a CUDA kernel is compiled and executed on an NVIDIA GPU, covering compilation to PTX and SASS, and the underlying hardware interaction.

AI tool VEGA from Nebula Security discovered a 15-year-old use-after-free bug in the Linux kernel (GhostLock) that allows any logged-in user to gain root access. The flaw, present since 2011, was patched in April but rollout is uneven.

Tweet recommending a blog post that explains the CPU-GPU communication details required when launching a CUDA kernel.