KernelScan.io

HIGH Public PoC

ppp Channel UAF

CVE-2026-68398

CVSS 7.8 / 10.0 NVD

CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

KernelScan AI6.9MEDIUM

01

In the Linux kernel, the following vulnerability has been resolved: ppp: defer channel free to an RCU grace period to fix pppol2tp RX UAF pppol2tp_recv() runs in the L2TP UDP-encap softirq RX path: l2tp_udp_encap_recv() -> l2tp_recv_common() -> pppol2tp_recv() -> ppp_input(&po->chan) It runs under rcu_read_lock() holding only an l2tp_session reference and takes NO reference on the internal PPP channel (struct channel, chan->ppp) that ppp_input() dereferences. The pppox socket is SOCK_RCU_FREE, so 'po' and the embedded ppp_channel are RCU-safe. But the internal struct channel is a separate allocation that ppp_release_channel() frees with a plain kfree(): close(data socket) -> pppol2tp_release() -> pppox_unbind_sock() -> ppp_unregister_channel() -> ppp_release_channel() -> kfree(pch) For a channel that is bound (PPPIOCGCHAN) but not attached to a ppp unit (no PPPIOCCONNECT, pch->ppp == NULL) and not bridged, teardown skips both ppp_disconnect_channel()'s synchronize_net() and ppp_unbridge_channels()'s synchronize_rcu(), so the kfree() has no grace period. rcu_read_lock() in pppol2tp_recv() does not protect against a plain kfree(), so an in-flight ppp_input() on one CPU can dereference the channel just freed by close() on another CPU. The bug is reachable by an unprivileged user. Defer the channel free to an RCU callback via call_rcu() so the grace period fences any in-flight ppp_input(). The disconnect and unbridge teardown paths already fence with synchronize_net()/synchronize_rcu(); call_rcu() does the same here without stalling the close() path.

02

Engine v0.6.0

Risk summary

An unprivileged local user can trigger a use-after-free in the PPP channel subsystem by racing socket close against in-flight L2TP packet reception. The freed kernel object can be dereferenced by the receive path on another CPU, leading to memory corruption that is potentially exploitable for code execution or a kernel panic. Any system allowing untrusted local users to create PPPoL2TP sockets is at risk.

Affecteddrivers/net/ppp/ppp_generic.c (PPP generic layer)

Vulnerability analysis

When a PPP channel is bound but not attached to a PPP unit and not bridged, closing the data socket frees the channel's internal structure immediately without waiting for any RCU grace period. At the same time, the L2TP UDP-encapsulated receive path runs in softirq context under an RCU read lock and dereferences that same channel structure to process incoming packets. Because the free is not RCU-deferred, a packet being processed on one CPU can access memory that close() on another CPU has already freed, producing a use-after-free. The fix defers the channel free to an RCU callback so that any in-flight receive operation drains before the memory is reclaimed, and it adds a barrier at module unload to wait for pending callbacks. The bug is reachable by an unprivileged local user who creates and closes PPPoL2TP sockets while packets are flowing through the L2TP receive path.

Exploit availability

KernelScan found public exploit code for this CVE. Open it in the CVE browser to see what we found, where, and how strong the evidence is — that needs a free account with a confirmed email address.

03

BranchIntroducedFixed inPatch commit
3.23.2.993.34e47f1ac188e
3.163.16.543.17110b765744b1
4.44.4.2254.54bb84e964ff0
5.15—5.15.21606213c85d8c0
6.6—6.6.148—
6.12—6.12.101—
6.18—6.18.42—
7.1—7.1.6—
mainline—7.2—
4.94.9.2254.103ab32218d718
4.144.14.1824.15c9574b8a8ede
6.1—6.1.183ec4215683e47