CVE-2021-47617

What it actually does

The flaw causes an infinite loop in the pciehp hardirq handler when a sticky Power Fault Detected bit is set on a PCIe hotplug slot, because the handler clears the bit without setting the power_fault_detected flag that the IRQ thread requires, preventing the thread from ever running and starving other processing.

Why it matters to the business

Systems may experience sudden unresponsiveness or require manual reboot when certain PCIe hotplug power events occur, leading to unplanned downtime on servers or devices using affected kernel versions.

How an attack unfolds

An attacker with physical or hardware-level access triggers a main power fault on a hot-plug PCIe slot; the sticky PFD bit repeatedly fires the ISR which loops indefinitely without ever waking the thread that would handle the event, resulting in denial of service.

How to fix it

Update to a Linux kernel containing the fix that sets the power_fault_detected flag inside the hardirq handler itself; apply the stable-tree backport or later mainline release that resolves the reintroduced loop.

How to confirm the fix worked

Confirm the running kernel contains commit 8edf5332c393 or its equivalent backport; verify by inspecting pciehp_isr() source to ensure the flag is set before returning from the hardirq context.

Typical attack path

1. Gain ability to induce a power fault on a PCIe hotplug-capable slot
2. Trigger the fault condition so the Slot Status register sets the sticky PFD bit
3. Observe the hardirq handler enter an infinite clear-and-reassert loop
4. System resources are consumed and normal interrupt processing halts

How to detect it

Search system logs and kernel traces for repeated pciehp ISR entries without corresponding thread activity or for messages referencing power fault handling; examine dmesg or serial console output around the time of any spontaneous reboot or hang for pciehp-related stack traces.

What your security team should do

• Prioritize kernel upgrades on all systems exposing PCIe hotplug controllers
• Review hardware configurations for unnecessary hotplug-capable slots
• Monitor vendor security advisories for distribution-specific backports