PT-2026-27669 · Linux · Linux

Published 2026-03-25 · Updated 2026-03-25 · CVE-2026-23304

NVD

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

ipv6: fix NULL pointer deref in ip6 rt get dev rcu()

l3mdev master dev rcu() can return NULL when the slave device is being un-slaved from a VRF. All other callers deal with this, but we lost the fallback to loopback in ip6 rt pcpu alloc() -> ip6 rt get dev rcu() with commit 4832c30d5458 ("net: ipv6: put host and anycast routes on device with address").

KASAN: null-ptr-deref in range [0x0000000000000108-0x000000000000010f] RIP: 0010:ip6 rt pcpu alloc (net/ipv6/route.c:1418) Call Trace: ip6 pol route (net/ipv6/route.c:2318) fib6 rule lookup (net/ipv6/fib6 rules.c:115) ip6 route output flags (net/ipv6/route.c:2607) vrf process v6 outbound (drivers/net/vrf.c:437)

I was tempted to rework the un-slaving code to clear the flag first and insert synchronize rcu() before we remove the upper. But looks like the explicit fallback to loopback dev is an established pattern. And I guess avoiding the synchronize rcu() is nice, too.

Related Identifiers

CVE-2026-23304

Affected Products

Linux

References · 7

• https://nvd.nist.gov/vuln/detail/CVE-2026-23304 · Security Note
• https://git.kernel.org/stable/c/3310fc11fc47387d1dd4759b0bc961643ea11c7f · Note
• https://git.kernel.org/stable/c/581800298313c9fd75e94985e6d37d21b7e35d34 · Note
• https://git.kernel.org/stable/c/4a48fe59f29f673a3d042d679f26629a9c3e29d4 · Note
• https://git.kernel.org/stable/c/0b5a7826020706057cc5a9d9009e667027f221ee · Note
• https://git.kernel.org/stable/c/ae88c8256547b63980770a9ea7be73a15900d27e · Note
• https://git.kernel.org/stable/c/2ffb4f5c2ccb2fa1c049dd11899aee7967deef5a · Note