LINUX NEWS · October 19, 2026 · 6–8 minute read
Linux Kernel 7.3 Released: FailFS, Scheduler Changes, and More
Linux kernel 7.3 is now available, bringing a new FailFS pseudo-filesystem, scheduler improvements, expanded sched_ext functionality, better virtual-machine memory tracking, and updates across storage, networking, graphics, Rust, and hardware support.
fchroot(), expands sched_ext, changes CPU group scheduling, and adds VM working-set tracking alongside numerous filesystem and driver updates.Linux kernel 7.3 is now available
Linus Torvalds released Linux kernel 7.3 after a development cycle that set records for contribution activity, according to the release coverage from Linuxiac and LWN.net. The cycle included 18,765 non-merge commits from 3,021 developers, including 807 first-time contributors.
Kernel 7.3 is a mainline kernel release, not a guarantee that every Linux distribution will immediately ship it as its default kernel. Rolling-release distributions may package it sooner; enterprise and long-term-support distributions typically follow their own testing and maintenance schedules. Check your distribution’s supported kernel policy before upgrading a production host.
Key changes in Linux 7.3
FailFS and fchroot()
FailFS is a deliberately restrictive pseudo-filesystem where filesystem operations fail. Combined with the new file-descriptor-based fchroot() system call, it offers building blocks for sandboxing processes that should not perform ordinary pathname lookups.
CPU scheduling and sched_ext
Group-scheduling changes aim to improve CPU-time distribution and reduce overhead in relevant workloads. The sched_ext framework also completes more of its nested sub-scheduler functionality, allowing BPF-based scheduling policies to be delegated to cgroup subtrees.
Virtual-machine memory tracking
Working-set tracking for guest memory improves visibility into which pages virtual machines actively use. This can help inform memory-management and capacity decisions, but it does not by itself guarantee lower memory usage or better performance for every VM.
Storage and filesystem updates
Changes include OverlayFS support for id-mapped mounts, removal of Btrfs’s legacy free-space cache v1 implementation, initial NTFS3 support for Windows alternative data streams, and performance work involving ext4, FUSE, and NVMe paths.
SMB server enhancements
The in-kernel ksmbd server adds support for Apple Time Machine backup workflows and per-share SMB3 encryption, along with improvements for encrypted SMB Direct connections.
Graphics and hardware enablement
The release brings further AMD Zen 6 preparation, AMD and Intel graphics changes, including work related to future Nova Lake GPUs, and updates for selected wireless adapters, ARM and RISC-V platforms, and gaming handhelds.
What FailFS means for sandboxing
FailFS is unusual because its purpose is not to provide normal file storage. Its intentionally failing filesystem semantics can help construct a restricted environment in which ordinary absolute-path and current-working-directory-relative lookups do not work as they normally would.
The new fchroot() call changes a process root using an already-open file descriptor rather than resolving a target pathname at the time of the change. This is useful as a lower-level building block for carefully designed isolation mechanisms, but it is not a complete sandbox by itself. Applications still need a deliberate threat model and correct handling of file descriptors, capabilities, namespaces, and other security boundaries.
Scheduler changes and performance expectations
Linux 7.3 continues work on CPU scheduling, including changes to cgroup weight distribution and run-queue handling. The kernel’s extensible scheduling framework, sched_ext, also gains more complete support for nested sub-schedulers, making it possible to delegate scheduling decisions for parts of a cgroup hierarchy to BPF-based schedulers.
These changes are particularly relevant to systems running many competing workloads, containers, or custom scheduling policies. Actual results depend on the CPU topology, workload mix, configuration, and distribution kernel options. Do not interpret upstream scheduler changes as a universal benchmark result.
Other notable kernel updates
| Area | What to know | Who may care |
|---|---|---|
| Containers and overlays | OverlayFS adds id-mapped mount support. | Container platform engineers and Linux administrators. |
| Windows file compatibility | NTFS3 gains initial support for Windows alternative data streams. | Users accessing NTFS volumes from Linux; validate requirements before relying on it. |
| File sharing | ksmbd gains Time Machine and SMB encryption improvements. | Administrators using the kernel SMB server. |
| FUSE and I/O | More zero-copy I/O work through io_uring. | FUSE developers and storage application maintainers. |
| Memory observability | VM working-set tracking and DAMON enhancements. | Virtualization and memory-performance teams. |
| Security | Updates include AppArmor policy-loading improvements, additional Landlock capabilities, and stricter signed-BPF handling. | Linux security and platform teams. |
| Rust in the kernel | Kernel Rust abstractions and architecture support continue to expand. | Kernel developers and maintainers of Rust-based components. |
How to download Linux kernel 7.3
The official Linux kernel archives publish the source tarballs, signatures, patches, and related release files. Most desktop and server users should install kernels through their Linux distribution’s supported update mechanism instead of manually replacing a production kernel with a source-tree build.
Visit kernel.orgOpen Linux 7.x source archive
- The official archive lists
linux-7.3.tar.xzand its signature file. Verify downloaded source against the official signature and checksum guidance. - For distro-managed systems, consult the distribution’s package repository and release notes for availability and support status.
- For production servers, test boot, drivers, storage, networking, virtualization, and monitoring in a non-production environment first. Keep a known-good kernel available for rollback.
Should you upgrade immediately?
For developers and testers who need the newest upstream features, Linux 7.3 is worth evaluating in a lab. For virtualization hosts, enterprise servers, and systems that prioritize vendor support, the better choice may be the distribution-supported kernel until the new release is packaged and validated for that platform.
fchroot() if you work on Linux isolation, review scheduler changes for heavily loaded systems, and test VM working-set tracking where memory observability matters. Upgrade production systems according to your distribution’s lifecycle and change-control process, not solely because a new upstream version is available.Official sources and further reading
- The Linux Kernel Archives — latest releases and downloads
- Linux Kernel Newbies — Linux 7.3 changelog and feature notes
- LWN.net — The 7.3 kernel has been released
- Linux kernel scheduler documentation
- Linuxiac — original release coverage
VMoreCloud verdict
Linux kernel 7.3 is a broad upstream release with meaningful changes for sandboxing, CPU scheduling, virtualization observability, storage, and hardware enablement. FailFS and fchroot() stand out as new isolation building blocks, while scheduler and VM-memory updates will interest teams managing dense workloads. As always, the practical value depends on how your distribution packages the kernel and how your workload behaves in testing.
This article summarizes published release information and does not claim independent kernel compilation, performance benchmarking, or hands-on testing.
Editorial note: Release facts were cross-checked against the Linux kernel archives, Linux Kernel Newbies, LWN.net, and the supplied Linuxiac article on October 19, 2026. Distribution availability and support may differ.