Imagine a world where the beating heart of your computer—the Linux kernel—gets a fresh upgrade, but not without a bit of drama from its creator himself. That's the story behind Linux 6.18, the newest long-term support kernel, which promises stability and innovation but also sparks debates on maintenance and innovation. If you're curious about how this affects your devices, from servers to smartphones, stick around—because the details might surprise you.
Linus Torvalds, the visionary behind the Linux kernel, expressed some frustration about the final stages of the 6.18 release. In a message to the Linux kernel mailing list on November 30, he mentioned wishing there had been fewer bug fixes in the last week before launch. Still, he concluded that the kernel was ready, declaring it 'tagged and pushed out.' Now that it's here, let's dive into what's inside this release and unpack what 'long-term support' really entails for the Linux ecosystem.
First off, for beginners, think of the Linux kernel as the core software that manages your computer's hardware and runs the operating system. Versions like 6.18 are snapshots of this kernel, and 'long-term support' (LTS) means it's a stable version that gets updates and fixes for a set period, unlike regular releases that might become obsolete faster. Here's a quick overview of the current kernel landscape as of December 2023:
- 6.18: Brand new LTS, supported until December 2027. It was officially designated as LTS in early December 2023, becoming the sixth active LTS branch.
- 6.12: LTS until December 2026, also chosen for the Civil Infrastructure Platform (CIP) as a Super-LTS with up to 10 years of support.
- 6.6: LTS until December 2026, selected in late 2023 and commonly used in enterprise and desktop setups.
- 6.1: LTS until December 2027, adopted by various distributions and embedded systems.
- 5.15: LTS until December 2026, popular in enterprise distros and for long-term hardware compatibility.
- 5.10: LTS until December 2026, one of the older ones still in use, like in Debian 11 and some embedded devices.
Back in the day, LTS kernels received six years of support, but that's changed. In 2023, kernel developers shortened it to just two years. Why the shift? It's all about sustainability. Maintainers, the unsung heroes who fix bugs and add features, have been experiencing burnout. As Josef Bacik, a key file system developer, pointed out in a 2022 summit talk, 'Maintainers are burning out because maintainers don’t scale.' On top of that, this demanding role is often unpaid, making it hard to sustain long-term.
And this is the part most people miss: The clock started ticking for 6.18 on December 3, 2023, when Greg Kroah-Hartman, the stable kernel maintainer, announced its LTS status. Interestingly, this meant that the older LTS 5.4 dropped out of support at the same time. So, while the official support window is now shorter, it reflects a real need to keep the community healthy.
But what if two years isn't enough for your business or project? Fear not—enterprise Linux distributions often extend that support dramatically for their customers. For instance, Red Hat backs its Red Hat Enterprise Linux (RHEL) kernels for over 10 years, by carefully applying security fixes and features to a fixed version, even after upstream support ends. They even offer Extended Life Cycle Support (ELS) for older releases, ensuring kernel updates continue.
Red Hat clones like AlmaLinux and Rocky Linux mirror this by rebuilding RHEL's LTS kernels, giving users the same extended timelines. OpenELA, supported by companies like Oracle, SUSE, and CIQ/Rocky Linux, also maintains older kernels from the RHEL tree. They even took over support for the former LTS 4.14 until December 2024, and releases were still happening in 2023.
Amazon Web Services (AWS) provided support for 4.14 on Amazon Linux 2.0 until October 2025, and for the outdated 5.10 until June 2026. SUSE does similar for SUSE Linux Enterprise Server (SLES), with long lifecycles and extended support, now offering 16 years for SLES 16 and its 6.12 kernel.
Canonical, behind Ubuntu, supports LTS kernels for up to 15 years with add-on packages, and even maintains old ones like 4.14. For example, they still update the 11-and-a-half-year-old Ubuntu 14.04.
Why bother with LTS or these commercial versions? As Greg Kroah-Hartman explained in a 2020 interview, they offer a stable application binary interface (ABI), which means software runs reliably without breaking, and a steady flow of security patches. To clarify for newcomers, ABI is like a promise that the kernel's interfaces won't change unexpectedly, keeping things compatible. This is much safer than sticking with an unsupported kernel or trying to patch it yourself, which could leave vulnerabilities open.
Now, onto the highlights of Linux 6.18. At its foundation is a major overhaul of the slab memory allocator with 'sheaves,' a per-CPU caching system that cuts down on competition for resources, speeding up memory tasks. For a simple analogy, think of it as organizing a busy kitchen where cooks (processes) grab ingredients (memory) more efficiently. Plus, there are tweaks to swapping and virtual machine performance, especially under high memory load on servers and desktops.
But here's where it gets controversial: One big change is the removal of the experimental Bcachefs file system from the mainline kernel. Bcachefs is a versatile filesystem aimed at high-integrity storage, similar to Btrfs or ZFS, ideal for large local drives, multi-disk setups, and hybrid SSD-HDD combos for speed and space. It was pulled due to conflicts between its maintainer, Kent Overstreet, and kernel leaders like Torvalds over submission timing and review processes. The core issue was that fixes often arrived late, disrupting the kernel's stabilization phase. Critics argue that late changes signal immaturity for something as crucial as a filesystem, so Bcachefs now lives as a DKMS module outside the kernel. Distributions wanting it must add it separately. Is this the right call, or does it stifle innovation? Opinions diverge—some see it as necessary for stability, while others worry it sidelines promising tech.
On the networking front, 6.18 introduces Accurate Explicit Congestion Notification (AccECN) for TCP, providing detailed feedback on network traffic to boost performance during congestion. There's also PSP-encrypted TCP, a hardware-efficient alternative to IPsec or TLS in certain setups.
Security gets a boost with cryptographically signed BPF programs, allowing verified eBPF payloads for safer runtime operations, along with other subsystem improvements. Infrastructure-wise, processes can now use file-handle-like objects for namespaces, making container environments and tools more secure against race conditions—think of it as adding locks to prevent accidental overwrites in a shared workspace.
The kernel's gradual Rust adoption continues, with a Rust-based Binder driver for Android's inter-process communication, rewritten by Google to enhance reliability on Android devices.
As usual, 6.18 includes a plethora of new drivers for architectures like x86_64, ARM, and RISC-V, plus better support for GPUs, SoCs, storage, handheld gaming PCs (like the Asus ROG Ally and Lenovo Legion Go 2), laptops, and embedded boards. This means smoother out-of-the-box experiences on new distributions adopting 6.18.
For everyday Linux users, the perks include quicker memory handling, better networking, and enhanced hardware compatibility, especially on modern GPUs and ARM systems. For servers and clouds, focus on features like dm-pcache for hybrid storage, new TCP options, and signed BPF. With support through 2027 and no flashy 'killer' feature, 6.18 is a solid, dependable base for distributions in the coming years.
What do you think? Should the LTS support period be extended again to ease maintainer burnout, or is two years a fair balance? And was booting Bcachefs out of the kernel a smart move for stability, or does it hinder filesystem innovation? Share your views in the comments—we'd love to hear your perspective!
