Description

The kernel queue helper z_queue_node_peek() in kernel/queue.c dereferences a node taken from a queue's data_q list, reading the node's flag byte and, for items enqueued via k_queue_alloc_append/alloc_prepend, the data pointer of an internally allocated alloc_node struct. The implementations of z_impl_k_queue_peek_head() and z_impl_k_queue_peek_tail() performed this read-and-dereference without holding the queue's spinlock, while every other accessor of the same list — including k_queue_get(), which unlinks a node and k_free()s its backing alloc_node — operates under that lock. Because peek was unsynchronized, a concurrent k_queue_get() on the same queue (on an SMP build, or under preemption/ISR concurrency) can free the node between the moment peek obtains the node pointer and the moment it dereferences it. The peek then reads flag bits and a data pointer out of freed, potentially re-allocated heap memory and returns a stale or dangling pointer to its caller. k_fifo and k_lifo are thin wrappers over k_queue, so this affects buffer queues used throughout the net_buf, Bluetooth, USB, and networking subsystems; the peek operations are also system calls reachable from CONFIG_USERSPACE threads. The consequences are a use-after-free read that can leak stale heap contents (one pointer word) and, when the returned dangling pointer is subsequently consumed as a live buffer, a dereference that can crash the system or corrupt memory. Exploitation requires winning a small race window with local access (e.g. a userspace process racing k_queue_peek_* against k_queue_get on a shared queue, or two CPUs), so practical impact is bounded and of low severity. The fix wraps both peek implementations with k_spin_lock/k_spin_unlock on the queue lock, making the read-and-dereference atomic with respect to the concurrent unlink-and-free and bringing peek into line with the rest of the queue's locking discipline.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "3.6",
  "pubDate": "2026-08-07T22:16:57.180Z",
  "pubdate": "2026-08-07T22:16:57.180Z",
  "executiveSummary": "A use-after-free vulnerability exists in the kernel queue helper z_queue_node_peek() within Zephyr, specifically affecting the z_impl_k_queue_peek_head() and z_impl_k_queue_peek_tail() implementations. This flaw arises due to a lack of proper synchronization when accessing the queue's data_q list.\nThe vulnerability allows a local attacker with access to userspace threads or system calls to exploit a race condition between a peek operation and a concurrent k_queue_get() operation on the same queue. If successfully exploited, the vulnerability results in a use-after-free read condition, which can leak stale heap memory contents or lead to memory corruption and system crashes when the returned dangling pointer is subsequently dereferenced.\nBecause k_fifo and k_lifo act as wrappers over k_queue, the issue impacts buffer queues utilized across core networking, Bluetooth, USB, and net_buf subsystems. Exploitation requires winning a small race window under SMP builds or preemption/ISR concurrency, meaning practical impact is bounded and evaluated as low severity.",
  "technicalDetails": "The root cause of the vulnerability lies in unsynchronized memory access within the kernel queue helper z_queue_node_peek() located in kernel/queue.c. Specifically, z_impl_k_queue_peek_head() and z_impl_k_queue_peek_tail() read and dereference a node taken from a queue's data_q list without acquiring the queue's spinlock. Every other list accessor—including k_queue_get(), which unlinks a node and invokes k_free() on its backing alloc_node—correctly operates under this lock.\nThe attack flow proceeds as follows: First, a thread initiates a peek operation via z_impl_k_queue_peek_head() or z_impl_k_queue_peek_tail(), obtaining a pointer to a node within the queue's data_q list. Before the peek operation can read the node's flag byte and data pointer (for items enqueued via k_queue_alloc_append or k_queue_alloc_prepend), a concurrent k_queue_get() operation executes on the same queue under SMP, preemption, or ISR concurrency. This concurrent operation unlinks the node and frees its backing alloc_node via k_free().\nSubsequently, the unsynchronized peek operation reads flag bits and a data pointer from the freed and potentially re-allocated heap memory. It then returns a stale or dangling pointer to the caller. When the caller attempts to consume this returned dangling pointer as a live buffer, the resulting dereference can crash the system or corrupt memory. Additionally, the vulnerability can leak a single pointer word of stale heap contents.\nThe vulnerable components include the z_queue_node_peek() function and the unsynchronized z_impl_k_queue_peek_head() and z_impl_k_queue_peek_tail() implementations in kernel/queue.c. These operations are exposed as system calls reachable from CONFIG_USERSPACE threads. Exploitation requires local access and the ability to win a narrow race window against concurrent queue removal operations, such as k_queue_get."
}