CVE-2021-47093

What it actually does

Failure to call platform_device_put() on intel_pmc_core device registration error during module init leaves the platform device and its name string allocated, causing a memory leak.

Why it matters to the business

Repeated module load failures on affected systems can gradually exhaust kernel memory, leading to degraded performance or instability that requires reboot to recover.

How an attack unfolds

An attacker with local access triggers repeated failures of the intel_pmc_core module initialization path, each time leaking a small amount of kernel memory until resource pressure builds.

How to fix it

Update to a Linux kernel containing commit that adds the missing platform_device_put() call on registration failure; rebuild and reboot with the fixed kernel.

How to confirm the fix worked

Confirm the presence of the platform_device_put() call in intel_pmc_core_init() or verify kernel version is newer than the vulnerable commit; inspect dmesg for registration failure paths without corresponding leaks.

Typical attack path

1. Obtain local execution on target system
2. Cause or force intel_pmc_core module load attempts that hit the registration failure path
3. Repeat to accumulate leaked platform device structures
4. Observe system-wide memory pressure or OOM conditions

How to detect it

Search kernel logs and module load traces for repeated intel_pmc_core registration failures; correlate with rising slab or platform_device allocations without corresponding frees.

What your security team should do

• Prioritize kernel updates on systems using intel_pmc_core
• Monitor for repeated module initialization failures in system logs
• Apply the fix during next maintenance window given low EPSS and no KEV listing