AZL-112562: CVE-2026-99229 affecting package kernel 6.6.159.1-1
In the Linux kernel, the following vulnerability has been resolved:
iouring/wait: don't spin on a stuck overflow list
As long as IOCHECKCQOVERFLOWBIT is set, iocqringwaitschedule() returns right away, before it gets to look at signals or arms the timeout, for iocqringwait() to flush the overflow list. If that doesn't empty the list, and the CQ ring doesn't hold the number of events the caller wants, the loop starts over, with the same result. The task spins in the kernel, and can't be killed.
The assumption is that this can't happen: if CQEs can't be flushed, the ring is full, and minevents is capped to the size of the ring. But:
- With IORINGSETUPCQEMIXED a 32b CQE needs two contiguous entries. It is stuck on the list if a single one is free, and then the ring holds less than cqentries events. A single threaded application that asks for as many events as the ring has entries never gets out of iouringenter().
- Whether the ring is full is up to the CQ head, which the application can write to. While one thread waits for events on an empty ring, another one can move the head back so that the ring looks full and have a CQE overflow. The waiter can't be stopped any more, not even by SIGKILL.
Only the application can make room in the CQ ring. Return to it if the list could not be flushed.
Affected Software
Event History
Frequently Asked Questions
Which workloads are exposed to this issue?
Workloads that use io_uring completion queues can be affected when completion events remain on the overflow list and the requested minimum event count cannot be met. The described cases include rings created with IORING_SETUP_CQE_MIXED and applications where another thread can modify the CQ head.
What conditions can trigger the uninterruptible spin?
With IORING_SETUP_CQE_MIXED, a 32-bit CQE requires two contiguous CQ entries; it can remain overflowed when only one entry is free. A separate thread can also move the CQ head so an empty ring appears full, causing a CQE overflow while another thread waits for events.
How would an affected process behave?
A task can repeatedly return from the wait path without checking signals or arming its timeout. It spins in the kernel, remains stuck in io_uring_enter(), and cannot be killed.