CVE-2025-68177

What it actually does

longhaul_exit() invokes cpufreq_cpu_get(0) and immediately dereferences the returned policy pointer without a NULL check, resulting in a kernel NULL pointer dereference that produces a warning or panic.

Why it matters to the business

Affected Ubuntu systems may suffer unexpected kernel crashes or warnings during cpufreq longhaul driver unload or certain hardware initialization sequences, causing downtime and potential data loss on production workloads.

How an attack unfolds

An attacker with local access triggers longhaul_exit under conditions where cpufreq_cpu_get returns NULL (specific hardware or configuration states), causing an immediate kernel panic and denial of service; no remote or unauthenticated exploitation path exists.

How to fix it

Upgrade affected packages to the listed fixed versions: linux 5.15.0-23.23 (Ubuntu 22.04), linux-aws 5.15.0-1003.5, linux-azure 5.15.0-1002.3, linux-aws-5.15 5.15.0-1014.18~20.04.1 and equivalent updates for all other listed variants; reboot after installation.

How to confirm the fix worked

Verify installed kernel package versions match the fixed releases, confirm absence of the vulnerable longhaul_exit code path via source inspection or kernel symbol checks, and test module load/unload on representative hardware.

Typical attack path

1. Obtain local access on a system using the longhaul cpufreq driver
2. Force conditions that invoke longhaul_exit without a valid policy pointer
3. Trigger NULL dereference leading to kernel panic

How to detect it

Search kernel logs and dmesg for NULL dereference oops mentioning longhaul_exit or cpufreq_cpu_get; inspect crash dumps for policy pointer dereference in the longhaul exit path.

• longhaul_exit
• cpufreq_cpu_get
• NULL dereference in longhaul

What your security team should do

• Prioritize patching all listed Ubuntu 20.04 and 22.04 kernel packages within the next maintenance window
• Monitor kernel logs for longhaul-related oops or panic messages until patches are deployed
• Restrict local access on systems using cpufreq longhaul where possible until remediation completes