CVE-2024-50196

What it actually does

The pinctrl ocelot driver fails to invoke chained_irq_enter() and chained_irq_exit() when a level-triggered GPIO interrupt is cleared by hardware before the parent edge-triggered interrupt is serviced, leaving the parent interrupt pending and causing a system hang during interrupt handling.

Why it matters to the business

Systems using affected Linux kernel pinctrl drivers on certain hardware can experience unexpected unavailability and require manual intervention or reboot, disrupting operations on servers and embedded devices with GPIO-based interrupts.

How an attack unfolds

An attacker or malicious hardware configuration triggers a short-duration level-based interrupt on a GPIO pin while the parent interrupt controller is in edge mode; the interrupt clears before the handler loop executes, bypassing the chained enter/exit calls and hanging the system while servicing the parent interrupt. General Linux kernel exploitation campaigns by espionage and financially motivated actors could leverage similar interrupt-handling flaws for denial of service.

How to fix it

Update the Linux kernel to a version containing commit 655f5d4662b958122b260be05aa6dfdf8768efe6. For Ubuntu 14.04 LTS systems upgrade libgcrypt11 to 1.5.3-2ubuntu4.3; for Ubuntu 16.04 LTS upgrade libgcrypt20 to 1.6.5-2. Apply the same fix pattern (moving chained_irq_enter/exit outside the loop) to any other pinctrl drivers exhibiting the same pattern.

How to confirm the fix worked

Confirm the running kernel contains the referenced commit or package versions above; reproduce the level-vs-edge interrupt scenario in a test environment and verify the system no longer hangs; inspect dmesg and interrupt logs for successful chained handler invocation.

Typical attack path

1. Identify or configure a GPIO pin in level-triggered mode with parent controller in edge mode
2. Generate a brief interrupt signal that the hardware clears rapidly
3. Trigger the parent interrupt while the inner loop skips chained_irq_enter due to cleared register
4. Cause the kernel to hang attempting to service the unacknowledged parent interrupt

How to detect it

Review kernel logs and interrupt traces for repeated parent interrupt servicing without corresponding child handler entry; correlate system hangs with GPIO level-triggered devices; search for evidence of the unpatched loop pattern in loaded kernel modules.

What your security team should do

• Prioritize kernel and package updates on all systems exposing GPIO or pinctrl hardware within the next maintenance window
• Audit custom pinctrl drivers for similar chained_irq patterns and apply the outer-loop fix
• Monitor for system hangs correlated with interrupt activity and maintain offline recovery procedures