In the Linux kernel, the following vulnerability has been resolved:
rxrpc: serialize kernel accept preallocation with socket teardown
rxrpc_kernel_charge_accept() reads rx->backlog without any
socket/backlog synchronization and passes that raw pointer into
rxrpc_service_prealloc_one(). A concurrent rxrpc_discard_prealloc()
sets rx->backlog = NULL and frees the backlog rings, so a kernel
preallocation worker can keep using a freed struct rxrpc_backlog
while updating *_backlog_head/tail and array slots.
Serialize the state check and backlog lookup with the socket lock,
and reject kernel preallocation once teardown has disabled
listening or discarded the service backlog.
CVSS Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
CVSS Score: 9.8
AV:N - RxRPC is a UDP network protocol; remote peers can send packets to AFS callback-manager listeners (port 7001) that instantiate incoming calls and queue kernel preallocation work reaching rxrpc_kernel_charge_accept().
AC:L - The attacker controls the preallocation side by flooding concurrent rxrpc/AFS callback traffic while a peer thread tears down the same socket via listen(0), shutdown, or close during AFS netns exit, creating a winnable race.
PR:N - No local account, capability, or authentication is required to reach the vulnerable preallocation path; any remote host that can deliver rxrpc/UDP packets to a listening AFS callback manager can drive the charge_accept worker.
UI:N - Exploitation requires no victim interaction beyond normal background AFS service operation; the attacker triggers the bug purely through network traffic timed against automatic or administrative socket teardown.
S:U - The use-after-free corrupts kernel heap metadata and rxrpc backlog structures within kernel memory, enabling local privilege escalation but not crossing a VM, container, or IOMMU security boundary by itself.
C:H - Concurrent use of a freed struct rxrpc_backlog allows reads and corruption of recycled slab memory, providing a standard kmalloc UAF primitive that can leak kernel pointers and sensitive data.
I:H - The race writes peer, connection, and call pointers plus head/tail indices into freed backlog ring slots, enabling heap grooming and arbitrary kernel memory corruption exploitable for code execution.
A:H - Use-after-free on the backlog structure during concurrent head/tail and pointer-array updates can cause kernel oops, panic, or hang, and repeated triggering enables sustained denial of service.
| Attack Vector |
Network |
Scope |
Unchanged |
| Attack Complexity |
Low |
Confidentiality Impact |
High |
| Privileges Required |
None |
Integrity Impact |
High |
| User Interaction |
None |
Availability Impact |
High |
AV:N - RxRPC is a UDP network protocol; remote peers can send packets to AFS callback-manager listeners (port 7001) that instantiate incoming calls and queue kernel preallocation work reaching rxrpc_kernel_charge_accept().
AC:L - The attacker controls the preallocation side by flooding concurrent rxrpc/AFS callback traffic while a peer thread tears down the same socket via listen(0), shutdown, or close during AFS netns exit, creating a winnable race.
PR:N - No local account, capability, or authentication is required to reach the vulnerable preallocation path; any remote host that can deliver rxrpc/UDP packets to a listening AFS callback manager can drive the charge_accept worker.
UI:N - Exploitation requires no victim interaction beyond normal background AFS service operation; the attacker triggers the bug purely through network traffic timed against automatic or administrative socket teardown.
S:U - The use-after-free corrupts kernel heap metadata and rxrpc backlog structures within kernel memory, enabling local privilege escalation but not crossing a VM, container, or IOMMU security boundary by itself.
C:H - Concurrent use of a freed struct rxrpc_backlog allows reads and corruption of recycled slab memory, providing a standard kmalloc UAF primitive that can leak kernel pointers and sensitive data.
I:H - The race writes peer, connection, and call pointers plus head/tail indices into freed backlog ring slots, enabling heap grooming and arbitrary kernel memory corruption exploitable for code execution.
A:H - Use-after-free on the backlog structure during concurrent head/tail and pointer-array updates can cause kernel oops, panic, or hang, and repeated triggering enables sustained denial of service.
CVSS 3.1