[xfs: resample the data fork mapping after cycling ILOCK]

A flaw was found in the XFS filesystem. A race condition in the copy-on-write mechanism for reflinked files can cause writes to bypass the copy-on-write process and modify shared data blocks directly. As a result, data intended for a private copy may be written to the original shared location, corrupting the contents of other files that reference those blocks. A local attacker with read access to a file on an XFS filesystem with reflink enabled could exploit this flaw to corrupt files they do not have write access to, allowing content to be added to the target file. If properly exploited this vulnerability may lead to privilege escalations or arbitrary code execution.

In the Linux kernel, the following vulnerability has been resolved: xfs: resample the data fork mapping after cycling ILOCK xfs_reflink_fill_{cow_hole,delalloc} are both presented with an inode, a data fork mapping, and a cow fork mapping. Unfortunately, these two helpers cycle the ILOCK to grab a transaction, which means that the mappings are stale as soon as we reacquire the ILOCK. Currently we refresh the cow fork mapping by re-calling xfs_find_trim_cow_extent, but we don't refresh the data fork mapping beforehand, which means that the xfs_bmap_trim_cow in that function queries the refcount btree about the wrong physical blocks and returns an inaccurate value in *shared. If *shared is now false, the directio write proceeds with a stale data fork mapping. Fix this by querying the data fork mapping if the sequence counter changes across the ILOCK cycle.

xfs: resample the data fork mapping after cycling ILOCK

In the Linux kernel, the following vulnerability has been resolved: x ...

In the Linux kernel, the following vulnerability has been resolved: xfs: resample the data fork mapping after cycling ILOCK xfs_reflink_fill_{cow_hole,delalloc} are both presented with an inode, a data fork mapping, and a cow fork mapping. Unfortunately, these two helpers cycle the ILOCK to grab a transaction, which means that the mappings are stale as soon as we reacquire the ILOCK. Currently we refresh the cow fork mapping by re-calling xfs_find_trim_cow_extent, but we don't refresh the data fork mapping beforehand, which means that the xfs_bmap_trim_cow in that function queries the refcount btree about the wrong physical blocks and returns an inaccurate value in *shared. If *shared is now false, the directio write proceeds with a stale data fork mapping. Fix this by querying the data fork mapping if the sequence counter changes across the ILOCK cycle.

Important: kernel security, bug fix, and enhancement update

ELSA-2026-39494: kernel security, bug fix, and enhancement update (IMPORTANT)

Important: kernel security update