CVE-2021-47382

What it actually does

Missed error path in qeth_do_reset() still acquires discipline_mutex, creating deadlock potential with ccwgroup_remove during offline channel path configuration and concurrent reset operations.

Why it matters to the business

Intermittent system hangs or unresponsiveness on s390 mainframes using qeth networking hardware can cause unplanned downtime and operational disruption, especially in environments with frequent path configuration changes.

How an attack unfolds

An administrator or automated process configures a qeth channel path offline, triggering a race between qeth_do_reset() and ccwgroup_remove that leads to deadlock when the missed mutex acquisition occurs in the error path.

How to fix it

Update the Linux kernel to a version containing the fix that calls qeth_set_offline() directly in the qeth_do_reset() error case and uses a mutex-free variant of ccwgroup_set_offline().

How to confirm the fix worked

Confirm kernel version includes the referenced commit; test qeth offline configuration and recovery failure scenarios to verify absence of deadlock.

Typical attack path

1. Trigger qeth channel path offline configuration
2. Initiate concurrent qeth_do_reset() during recovery failure
3. Hit missed error path that acquires discipline_mutex
4. Deadlock with ccwgroup_remove causes system hang

How to detect it

Review kernel logs for messages involving qeth_do_reset, discipline_mutex, or deadlocks during ccwgroup operations; correlate with channel path offline events.

What your security team should do

• Apply the kernel fix during the next maintenance window
• Monitor kernel logs for qeth recovery-related hangs
• Limit automated or frequent qeth path reconfiguration in production