CVE-2026-53259
HighCVSS 7.8Exploitation Probability (EPSS)
Low risk2th percentile — higher than 2% of all known CVEs
Summary
In the Linux kernel, a vulnerability in IPv6 anycast implementation allows a use-after-free when a struct ifacaddr6 (aca) is freed while still linked in the global hash table. The race condition occurs because the insertion into idev->ac_list and the global hash are not atomic with respect to interface teardown.
Risk Assessment
An attacker could exploit this vulnerability to achieve remote code execution in kernel context or cause a denial of service (DoS) by sending crafted IPv6 anycast packets, potentially leading to system compromise or crash.
Recommendation
Immediately update the Linux kernel to a version containing the fix that moves ipv6_add_acaddr_hash() inside the idev->lock section, ensuring atomicity of the operations.
Original NVD description (English source)
In the Linux kernel, the following vulnerability has been resolved: ipv6: anycast: insert aca into global hash under idev->lock syzbot reported a splat [1]: a slab-use-after-free in ipv6_chk_acast_addr(), which walks the global inet6_acaddr_lst[] hash under RCU and dereferences a struct ifacaddr6 that has already been freed while still linked in the hash, so a later reader walks into a dangling node. In __ipv6_dev_ac_inc() the aca is allocated with refcount 1, then aca_get() bumps it to 2 to keep it alive across the unlocked region. It is published to idev->ac_list under idev->lock, but ipv6_add_acaddr_hash() runs after write_unlock_bh(). A concurrent teardown (ipv6_ac_destroy_dev() from addrconf_ifdown(), under RTNL) can slip into that window: CPU0 __ipv6_dev_ac_inc CPU1 ipv6_ac_destroy_dev (RTNL) ------------------------------ ------------------------------------ aca_alloc() refcnt 1 aca_get() refcnt 2 write_lock_bh(idev->lock) add aca to ac_list write_unlock_bh(idev->lock) write_lock_bh(idev->lock) pull aca off ac_list write_unlock_bh(idev->lock) ipv6_del_acaddr_hash(aca) hlist_del_init_rcu() is a no-op, aca is not in the hash yet aca_put() refcnt 2->1 ipv6_add_acaddr_hash(aca) aca now inserted into the hash aca_put() refcnt 1->0 call_rcu(aca_free_rcu) -> kfree(aca) The hash removal becomes a no-op because the insertion has not happened yet, so once CPU0 inserts and drops the last reference, the aca is freed while still linked in inet6_acaddr_lst[], and readers dereference freed memory after the slab slot is reused. This window opened once RTNL stopped serializing the join path against device teardown. Move ipv6_add_acaddr_hash() inside the idev->lock section so the ac_list and hash insertions are atomic with respect to teardown: a racing remover now either misses the aca entirely or finds it in both lists. acaddr_hash_lock is now nested under idev->lock, which is acquired in softirq context, so switch all acaddr_hash_lock sites to spin_lock_bh() to avoid the irq lock inversion reported in [2]. [1] https://syzkaller.appspot.com/bug?extid=a01df04303c131efbf3a [2] https://lore.kernel.org/netdev/[email protected]/