Bcachefs Faces Potential Removal from Linux Kernel

Figure 1: Visual representation of the BeyondTrust vulnerability chain

Tensions between bcachefs development and kernel maintainers have been building throughout 2025. According to LWN.net reporting on June 28, 2025, the situation reached a critical point when kernel maintainers raised formal concerns about the development process during the 6.12 merge window discussions.

Kent Overstreet, who has led bcachefs development since its inception, submitted a substantial number of patches for the upcoming kernel release. Kernel maintainers, including those responsible for the overall filesystem subsystem, questioned whether the patches met the standards expected for mainline kernel code.

The LKML discussions reveal several specific concerns. Maintainers cited instances where patches were submitted without adequate review periods. Questions arose about regression testing procedures for bcachefs changes. Some maintainers expressed frustration with communication patterns during the review process.

Overstreet responded to these concerns on the mailing list, arguing that bcachefs requires rapid iteration to address bugs and implement features necessary for production use. He contended that the filesystem's complexity and relative newness justify a different development cadence than mature kernel subsystems.

The discussion expanded to include broader questions about how experimental or rapidly-evolving code should be handled within the kernel. Some participants suggested that bcachefs might benefit from continued development outside the mainline tree until it reaches greater stability.

Kernel maintainers have documented specific instances that inform their concerns. According to LKML archives, patches were submitted that introduced regressions in existing functionality. Review comments were sometimes addressed with minimal discussion. The volume of changes between kernel releases exceeded what maintainers considered manageable for proper review.

Overstreet has countered that bcachefs users depend on timely bug fixes and that delaying patches causes real harm to those running the filesystem. He pointed to the filesystem's growing user base as evidence that the development approach produces working code. Performance benchmarks and feature comparisons with other filesystems were cited as indicators of bcachefs progress.

Technical analysis from independent observers suggests that bcachefs code quality has improved since its initial merge but remains more volatile than established kernel filesystems. The filesystem's test suite has expanded, though coverage gaps exist in some areas.

The kernel's stable release maintainers have noted challenges with bcachefs backports. Patches that work in the development kernel sometimes require significant modification for stable releases, indicating potential issues with the original patch structure.

Figure 2: How the authentication bypass vulnerability works

Keeping bcachefs in the kernel provides users with access to a modern filesystem through standard distribution channels. The filesystem offers features not available in ext4, the most common Linux filesystem, including native encryption, compression, and snapshot capabilities.

Mainline inclusion ensures that bcachefs receives testing across diverse hardware configurations through kernel continuous integration systems. Bug reports from a broader user base accelerate the identification of issues that might not appear in smaller testing environments.

The visibility of mainline inclusion attracts contributors who might not engage with out-of-tree projects. Documentation, tooling, and distribution support improve when code resides in the official kernel tree.

For users already running bcachefs, continued mainline presence provides confidence in long-term support and reduces the complexity of maintaining systems with out-of-tree filesystem modules.

The current development pace creates risks for kernel stability. Filesystems interact with many kernel subsystems, and bugs can cause data loss or system crashes. Rapid changes increase the probability of introducing such bugs.

Kernel maintainers bear responsibility for code they accept into the tree. Accepting patches that do not meet review standards creates precedents that could affect other subsystems. The time spent reviewing bcachefs patches reduces capacity for other kernel work.

Users who have deployed bcachefs face uncertainty about the filesystem's future. If removed from the kernel, they would need to maintain out-of-tree modules or migrate to other filesystems. Data migration from bcachefs to other filesystems requires careful planning and carries inherent risks.

The dispute has created friction within the kernel community. Public disagreements between maintainers can discourage participation and affect the collaborative culture that kernel development depends upon.

Figure 3: Privilege escalation from user to SYSTEM level

Bcachefs implements a copy-on-write (COW) architecture where modifications create new data blocks rather than overwriting existing ones. The filesystem maintains a B-tree structure for metadata, enabling efficient lookups and modifications. Multiple devices can be combined into a single filesystem with configurable redundancy and tiering policies.

The filesystem's checksumming feature calculates cryptographic hashes for data and metadata blocks. On read operations, bcachefs verifies these checksums and can automatically repair corrupted data when redundant copies exist. Compression operates transparently, with multiple algorithms available including lz4 and zstd.

Encryption in bcachefs uses authenticated encryption modes, protecting both data confidentiality and integrity. Keys can be managed through kernel keyring integration, supporting various authentication mechanisms.

Snapshots in bcachefs leverage the COW architecture to create point-in-time copies without duplicating data. Only blocks that change after snapshot creation require additional storage. The snapshot implementation supports hierarchical relationships between snapshots.

Technical context for expert readers: Bcachefs uses a novel approach to space allocation that differs from both btrfs and ZFS. The allocator maintains per-device freelists with bucket-based allocation, designed to reduce fragmentation and improve write performance on SSDs. The journal implementation uses a ring buffer with configurable size, balancing write amplification against crash recovery time.

The bcachefs situation reflects broader tensions in open source development between innovation speed and stability requirements. Large projects like the Linux kernel must balance the desire to incorporate new features against the need to maintain reliability for millions of users.

The outcome will influence how other experimental subsystems approach kernel inclusion. A removal could discourage developers from seeking mainline status for ambitious projects, potentially fragmenting the Linux ecosystem. Alternatively, it could reinforce the importance of meeting established development standards.

Linux distributions face practical decisions regardless of the outcome. Those offering bcachefs as an installation option must evaluate whether to continue support, recommend migration paths, or wait for resolution. Enterprise distributions with long support cycles have particular exposure to filesystem stability concerns.

The filesystem market on Linux remains competitive. Btrfs continues development with Red Hat and SUSE support. ZFS, while not in the mainline kernel due to licensing, maintains a substantial user base through out-of-tree modules. Ext4 remains the default for most distributions despite its age.

Confirmed:

• Discussions about bcachefs removal are occurring on the Linux Kernel Mailing List
• Kernel maintainers have raised concerns about development practices
• Kent Overstreet has defended the current development approach
• Bcachefs was merged into kernel 6.7 in October 2023
• No final decision has been announced as of June 28, 2025

Unresolved:

• Whether bcachefs will actually be removed from the kernel
• Timeline for any potential decision
• What conditions might allow bcachefs to remain in the kernel
• Impact on users currently running bcachefs in production
• Whether a compromise on development practices can be reached

The Linux Kernel Mailing List will be the primary venue for any announcements regarding bcachefs status. Statements from Linus Torvalds or the filesystem subsystem maintainers would carry particular weight in determining the outcome.

The next kernel merge window will reveal whether bcachefs patches continue to be accepted. A reduction in accepted patches or explicit holds on bcachefs changes would signal movement toward removal.

Distribution responses will indicate industry assessment of the situation. Announcements from major distributions about bcachefs support status would affect user planning.

Kent Overstreet's responses to maintainer concerns and any proposed changes to development practices will influence the resolution. Community discussion about governance and development standards may extend beyond bcachefs to affect kernel processes more broadly.

Independent technical analysis of bcachefs stability and performance will inform user decisions regardless of kernel status. Benchmark comparisons and reliability reports from production deployments provide data points for evaluating the filesystem's maturity.