CVE-2026-98975
UnknownSummary
In the Linux kernel, XFS exchange-range operations incorrectly moved the reflink flag between files even though a file can still own shared blocks beyond i_disk_size. This caused in-place writes instead of CoW, corrupting the other file sharing those blocks. The fix stops moving the flags and sets them on both inodes, including for XFS_IOC_SWAPEXT.
Risk Assessment
The vulnerability can lead to silent data corruption in files sharing blocks on XFS filesystems with reflink enabled, especially during exchange-range or SWAPEXT operations. It requires local file operations but compromises data integrity.
Recommendation
Update the Linux kernel to a version containing the XFS fix described in the CVE. Until patched, avoid exchange-range and XFS_IOC_SWAPEXT operations on reflink-enabled files.
Other vulnerabilities in Linux kernel XFS
See all- CVE-2026-98974Unknown
A bug was fixed in the Linux kernel's XFS involving the XFS_IOC_EXCHANGE_RANGE operation with the TO_EOF flag. Incorrect swapping of file sizes at nonzero offsets caused a file to have shared blocks beyond EOF, and subsequently the reflink flag was incorrectly cleared, allowing a shared block to be overwritten.
- CVE-2026-97549Low risk· EPSS 9%
A vulnerability in the Linux kernel's XFS filesystem causes under-reservation of blocks when repairing shortform directories. Incorrectly setting args->total to 1 instead of the actual directory block count (fsbcount) triggers an assertion failure and can lead to a filesystem shutdown.
- CVE-2026-43365High
A vulnerability in the Linux kernel's XFS filesystem causes incorrect l_iclog_roundoff values (512 instead of expected 4096) when the superblock lacks a log stripe unit, leading to corrupted logs and unmountable filesystems. This occurs on disks with 4k physical sectors.
Original NVD description (English source)
In the Linux kernel, the following vulnerability has been resolved: xfs: stop exchanging reflink flags when exchanging mappings When an exchange covers both files from offset 0 to i_disk_size, xmi_can_exchange_reflink_flags() assumes every mapping moved and hands the reflink flag over to the other file. That assumption is wrong. A file can still own shared blocks past i_disk_size, e.g. after EXCHANGE_RANGE_TO_EOF sets a smaller size without unmapping anything. The flag is cleared anyway, and the next write to such a block happens in place instead of CoW, corrupting the other file that shares it. Commit a23eca88448e ("xfs: fix exchange-range reflink flag clearing issue with INO1_WRITTEN") fixed one instance of this. Rather than add another special case, stop moving the flags. xfs_exchmaps_ensure_reflink() already sets the flag on both inodes before the exchange, so both just keep it. Nothing sets CLEAR_INO{1,2}_REFLINK anymore. Intents logged by older kernels may still carry them; recovery already masks them off, so they stay in the log format mask and are ignored. XFS_IOC_SWAPEXT has the same problem: it moves the reflink flag from one inode to the other along with the extents. Just removing that is not enough here: the exchange path flags both inodes up front in xfs_exchmaps_ensure_reflink(), but swapext has no such step, so the inode receiving the shared extents would be left without the flag. Set it on both inodes instead, and make sure both have a CoW fork, since an inode with the reflink flag is expected to have one. The flag now stays until FALLOC_FL_UNSHARE_RANGE, xfs_scrub, or a truncate to zero removes it.
Vulnerability data from NVD (NIST) · CISA KEV · EPSS