During a concurrent fork scenario on the Linux kernel, the scheduler registers a newly created task as an MMCID user before the task becomes visible in the thread and global task lists. Because the per‑CPU threshold handling runs before this visibility, the scheduler may fail to fix the task’s CID, causing a stall when it later attempts to acquire a CID. The flaw is a classic race condition (CWE‑821) that can freeze the system, representing a denial‑of‑service. This also makes it symmetrical vs. exit() where the task is removed as a CID user before the task is removed from the thread and task lists.

The flaw is present in the Linux kernel itself; any kernel version that contains the old MMCID handling logic is potentially vulnerable. No specific distribution or version numbers are provided, but the patch that corrects the issue is available in recent kernel releases that reorder sched_mm_cid_fork to execute after the new task is made visible.

Risk is significant because a failure to acquire a CID can halt the scheduler and freeze the machine, a CVSS score of 7.0 indicating high severity. While exploitation probability is unknown due to missing EPSS and no KEV listing, the denial‑of‑service nature of the flaw warrants priority mitigation.