CVE-2025-21864

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

tcp: drop secpath at the same time as we currently drop dst

Xiumei reported hitting the WARN in xfrm6_tunnel_net_exit while running tests that boil down to:

• create a pair of netns
• run a basic TCP test over ipcomp6
• delete the pair of netns

The xfrm_state found on spi_byaddr was not deleted at the time we delete the netns, because we still have a reference on it. This lingering reference comes from a secpath (which holds a ref on the xfrm_state), which is still attached to an skb. This skb is not leaked, it ends up on sk_receive_queue and then gets defer-freed by skb_attempt_defer_free.

The problem happens when we defer freeing an skb (push it on one CPUs defer_list), and dont flush that list before the netns is deleted. In that case, we still have a reference on the xfrm_state that we dont expect at this point.

We already drop the skbs dst in the TCP receive path when its no longer needed, so lets also drop the secpath. At this point, tcp_filter has already called into the LSM hooks that may require the secpath, so it should not be needed anymore. However, in some of those places, the MPTCP extension has just been attached to the skb, so we cannot simply drop all extensions.

Weakness

The product dereferences a pointer that it expects to be valid but is NULL.

Affected Software

Potential Mitigations

References

• https://git.kernel.org/stable/c/69cafd9413084cd5012cf5d7c7ec6f3d493726d9
• https://git.kernel.org/stable/c/87858bbf21da239ace300d61dd209907995c0491
• https://git.kernel.org/stable/c/9b6412e6979f6f9e0632075f8f008937b5cd4efd
• https://git.kernel.org/stable/c/cd34a07f744451e2ecf9005bb7d24d0b2fb83656
• https://git.kernel.org/stable/c/f1d5e6a5e468308af7759cf5276779d3155c5e98
• https://lists.debian.org/debian-lts-announce/2025/05/msg00045.html