CVE-2024-46808

What it actually does

Missing NULL pointer check after kcalloc in dpcd_extend_address_range within AMD display driver can lead to assertion failure or kernel crash on allocation failure.

Why it matters to the business

Causes unexpected system instability and downtime on Linux servers or desktops with AMD GPUs, disrupting operations and requiring restarts in environments where graphics hardware is present.

How an attack unfolds

An attacker with local access induces memory pressure to force kcalloc failure in the AMD display DPCD code path, triggering the missing check and resulting in a kernel assertion or panic.

How to fix it

Upgrade the Linux kernel to a version containing commit ca0b0b0a22306f2e51105ac48f4a09c2fbbb504e. Apply the fix from stable kernel trees; unrelated packages such as runc 1.0.0-rc8.0.20190930145003-cad42f6e0932 and selinux 1.3.1-0.20190929122143-5215b1806f52 are listed but do not address this kernel issue.

How to confirm the fix worked

Confirm the running kernel includes the referenced commit; inspect drm/amd/display source for the added NULL check after kcalloc in dpcd_extend_address_range; monitor for absence of prior assertion behavior.

Typical attack path

1. Obtain local user access on target Linux system
2. Trigger code path calling dpcd_extend_address_range under memory exhaustion
3. Force kcalloc to return NULL
4. Trigger assertion failure leading to kernel crash

How to detect it

Examine kernel logs and dmesg for assertion messages or NULL dereference traces originating from amd/display/dpcd or related modules during crashes or display initialization.

What your security team should do

• Apply kernel updates containing the fix on all AMD GPU systems as highest priority
• Reboot affected hosts after patching and validate graphics functionality
• Review kernel crash logs for related display driver failures
• Test patches in non-production environments before broad rollout