CVE-2025-52565

runc 1.1.13 and earlier, as well as 1.2.0-rc2 and earlier, can be tricked into creating empty files or directories in arbitrary locations in the host filesystem by sharing a volume between two containers and exploiting certain operations.

A vulnerability in runc 1.1.11 and earlier is caused by an internal file descriptor leak, allowing an attacker to set the working directory of a container process to the host filesystem namespace. This enables container escape and access to the host filesystem via runc exec or runc run. Additional attack variants allow overwriting host binaries, leading to complete container escape.

In runc prior to versions 1.3.6, 1.4.3, and 1.5.0, setupPtmx and setupDevSymlinks use filepath.Join with os.Remove and os.Symlink, allowing an image with /dev as a symlink to delete the ptmx file on the host or create symlinks in an arbitrary host directory. This is not exploitable under Docker, but other container tooling built on runc remains exposed.

In runc versions 1.2.7, 1.3.2, and 1.4.0-rc.2, an attacker can trick runc into misdirecting writes to /proc to other procfs files through the use of a racing container with shared mounts. This redirect could be through symbolic links in a tmpfs or other methods such as regular bind-mounts. This issue is fixed in versions 1.2.8, 1.3.3, and 1.4.0-rc.3.

In runc versions 1.2.7 and below, 1.3.0-rc.1 through 1.3.1, 1.4.0-rc.1 and 1.4.0-rc.2, insufficient verification of the bind-mount source (e.g., container's /dev/null) allows attacks: arbitrary mount gadget leading to host information disclosure, host DoS, container escape, or bypassing maskedPaths.