Linux 3342 Published by

Linux Kernel 7.1 RC1 has arrived with a massive merge window where an AMD GPU register header sync artificially inflates the patch count. The update actively strips out legacy i486 configurations and obsolete SoC support to keep maintenance overhead manageable. Beyond the inflated stats, developers packed in meaningful improvements across VFS handling, NTFS and SMB compatibility, system tracing tools, and ongoing Rust integration. Testers should expect typical release candidate instability, run make oldconfig to handle removed options, and report bugs promptly to help stabilize the kernel before the final launch.



Linux Kernel 7.1 RC1 Released: AMD GPU Sync Dominates, Old Hardware Gets the Axe

The Linux Kernel 7.1 release candidate one is out for testing. Linus Torvalds closed the merge window after two weeks of patching. This build brings a massive influx of changes, though the stats look skewed by a single vendor update. Users looking to test new features or help stabilize the kernel should jump in now before the final release locks things down.

Kernel

AMD GPU Header Sync Skews the Numbers

The diffstat shows a lopsided patch distribution because of a bulk sync for AMD GPU register headers. This single change accounts for about twenty-five percent of all patches in this window. While that number looks alarming to anyone scanning the stats, it is mostly metadata and definitions rather than functional logic changes. Veterans of kernel development have likely seen merge windows where the diffstat looks scary because of one vendor syncing headers, only to find the actual functional changes are minimal. Once you ignore that bulk update, the kernel development pace returns to normal with roughly half the work going into drivers and the rest spread across architecture updates, filesystems, networking, and core subsystems. This kind of header sync happens when a driver maintainer needs fresh register definitions to support newer hardware or fix existing bugs. The underlying code remains stable, but testers should expect the usual rc turbulence as these new definitions get exercised.

Cleaning House for i486 and Dead SoCs

Linus flagged several projects aimed at removing old hardware support to reduce maintenance burden. The kernel is phasing out i486 configurations, with code deletions following the config removals. There is also independent work stripping out really old networking hardware support and removing System on Chip support that never gained traction. This cleanup is necessary. Keeping legacy code alive just adds complexity and potential security risks without providing value to modern systems. Anyone still trying to compile kernels for a 486 processor or an obscure embedded board from the early two thousands should expect these features to vanish in upcoming releases. The net result of all this removal is still positive because the new feature additions far outweigh the deletions, so the total line count continues to grow.

Notable Changes Beyond the Noise

Scanning through the contributor list reveals several areas worth watching beyond the AMD sync. Christian Brauner handled a significant chunk of VFS updates, including writeback improvements and directory handling changes that affect how files are managed across different filesystems. NTFS support saw attention with updates to NTFS3 and what looks like resurrection efforts for older NTFS variants, which helps Windows partition compatibility on Linux systems. Steve French pushed multiple SMB client and server updates, improving network file sharing reliability. The kernel also continues its gradual integration of Rust code, as noted by Miguel Ojeda's update, signaling that the language is becoming a permanent fixture in subsystem development. Steven Rostedt contributed heavily to tracing and ftrace tools, which helps developers debug performance issues more effectively.

Testing Strategy for RC 1

The merge window produced about thirteen thousand non-merge commits along with a thousand merge commits, making this one of the larger windows in recent memory. Users who want to test 7.1 should clone the repository and apply the standard kernel build process. Running make oldconfig is essential here because the removal of i486 configs and other legacy options will likely trigger prompts or defaults that need review. Power users might notice changes in scheduler behavior or filesystem performance due to the VFS and block layer updates, so benchmarking before and after the upgrade provides concrete data on stability gains. If a system fails to boot or exhibits random crashes during rc testing, reporting the issue with full dmesg logs helps Linus and the maintainers calm things down faster. The goal is to catch regressions now rather than waiting for the final release when users are less likely to report bugs.

Linux kernel 7.1-rc1 released

Linux kernel version 7.1-rc1 is now available:

Full source: https://git.kernel.org/torvalds/t/linux-7.1-rc1.tar.gz
Patch: https://git.kernel.org/torvalds/p/v7.1-rc1/v7.0

You can view the summary of the changes at the following URL:
https://git.kernel.org/torvalds/ds/v7.1-rc1/v7.0

Grab a coffee and start compiling if you enjoy living on the edge, or wait for your distro to pick up 7.1 once rc2 settles things down. The kernel community is moving fast, and this release keeps that momentum going while trimming the fat from ancient hardware support.