Linux Kernel AMD GPU KFD Race Condition Vulnerability Leading to Kernel Panic

A race condition vulnerability has been identified in the Linux kernel's handling of AMD GPU KFD (Kernel Fusion Driver) operations. This vulnerability occurs in the 'amdgpu_amdkfd_device_fini_sw' function, where a KFD cleanup process can be interrupted by a KGD (Graphics Debugger) interrupt. If this happens, the cleanup may not complete properly, leading to a NULL pointer dereference and a subsequent kernel panic. The issue has been observed in the stable Linux kernel versions 5.10.134-010.a1i5000.a18.x86_64.

Impact

Exploitation of this vulnerability causes a kernel panic due to a NULL pointer dereference, disrupting system operations and potentially leading to a system crash.

Reproduction

The vulnerability can be reproduced by triggering the 'amdgpu_amdkfd_device_fini_sw' function while a KGD interrupt is generated. This creates a race condition where the KFD cleanup process is interrupted, causing the KFD nodes to be improperly managed and leading to a NULL pointer dereference.

Remediation

Users can upgrade to the latest stable version of the Linux kernel where this vulnerability has been addressed. Instructions for downloading the latest version can be found on the official Linux kernel website.