On Fri, 25 Sept 2026 at 11:45, Mathieu Poirier <[email protected]> wrote:
> We could lose the CPU right here and meanwhile, __rproc_virtio_del_vqs() > deletes > @vq. I don't see how READ_ONCE/WRITE_ONCE would change that. You're right, the single load only closes the NULL window, not the use-after-free. I'll send a v2 that protects rvring->vq with SRCU, so that __rproc_virtio_del_vqs() waits for rproc_vq_interrupt() to finish before freeing the virtqueues.
