Arm’s Mali GPU Vulnerabilities: A Critical Security Alert

Listen to this Post

2025-02-03

In recent security reports, Arm has unveiled several serious vulnerabilities impacting its Mali GPU Kernel and Firmware drivers across various versions. These vulnerabilities, identified under multiple CVEs (Common Vulnerabilities and Exposures), pose significant risks to system integrity, potentially allowing attackers to perform unauthorized operations on GPU memory. The affected GPU architectures include Midgard, Bifrost, Valhall, and Arm’s 5th Gen GPU architecture, with improper memory handling and GPU processing being the primary security concerns. With updates already rolled out, it’s crucial for users to upgrade to the latest versions to safeguard against potential exploitation.

the Key Vulnerabilities

Arm has disclosed vulnerabilities within several Mali GPU Kernel and Firmware drivers, which could be exploited by local non-privileged users to access unauthorized memory or cause system crashes. Key vulnerabilities, such as CVE-2025-0015, CVE-2024-6790, and CVE-2024-3655, affect GPU architectures like Valhall, Bifrost, and Arm 5th Gen, and can result in memory corruption, crashes, and exposure of sensitive data.

The most critical issues include improper GPU processing operations and improper memory handling, with CVE-2025-0015 affecting Valhall and Arm’s 5th Gen drivers (r48p0–r52p0). CVE-2024-6790 impacts multiple architectures and could lead to system unresponsiveness due to memory flaws induced by WebGL or WebGPU. These vulnerabilities have been patched in new driver versions, namely r49p2, r53p0, and r49p1.

Additionally, CVE-2024-0153 affects the firmware of Valhall and 5th Gen GPUs, opening up a path for attackers to exploit system memory. Some of these vulnerabilities have already been exploited in active attacks, highlighting the urgency for users to update their drivers.

What Undercode Says:

Arm’s recent disclosure of security vulnerabilities within its Mali GPU drivers emphasizes the growing importance of GPU security in modern computing environments. While the vulnerabilities span several architectures and affect both Kernel and Firmware drivers, the most critical flaws are centered around improper memory handling and unsafe GPU processing. These weaknesses provide opportunities for attackers to compromise memory security, crash systems, or gain access to sensitive data.

One of the most concerning vulnerabilities is CVE-2025-0015, which affects the Valhall and Arm 5th Gen GPU Kernel drivers. Attackers exploiting this vulnerability can access freed memory, potentially exposing critical system information. Memory management in GPUs has always been a complex task due to the high-performance demands placed on these components, making such flaws particularly hard to identify and patch. While the impact of these vulnerabilities is substantial, Arm’s swift patching of the affected driver versions (r49p2 and r53p0) is a step in the right direction toward mitigating the risks.

The issue is compounded by CVE-2024-6790, which could make entire systems unresponsive. This vulnerability demonstrates the increasing interdependence between GPUs and other software frameworks like WebGL and WebGPU. As these technologies become more integrated into web-based applications, the potential for exploitation grows, posing significant challenges for both developers and security teams. Fortunately, the patches for CVE-2024-6790 (r49p1 and r52p0) provide much-needed protection against this threat.

Additionally, CVE-2024-3655 and CVE-2024-2937 highlight the dangers of improper memory operations, such as race conditions and out-of-bounds accesses, which are common in high-performance computing systems. These types of vulnerabilities are particularly difficult to address because they require precise timing or execution conditions to trigger. The existence of these flaws underscores the complexity of designing secure systems that involve both software and hardware components.

CVE-2024-0153’s impact on firmware is another notable concern. Firmware vulnerabilities are often overlooked, yet they can be just as damaging as those affecting the operating system or applications. The risk of attackers gaining access to system memory via firmware exploits makes these vulnerabilities an area of particular concern for system integrity. Firmware vulnerabilities, if left unpatched, can lead to full system compromise, as attackers can gain control of low-level hardware operations.

Arm has demonstrated its commitment to addressing these issues by releasing patches and continuing collaboration with security researchers. For users of older GPU architectures like Midgard, contacting Arm support for guidance is recommended. Given that some vulnerabilities are already being actively exploited, staying up to date with the latest security advisories from Arm is essential.

This situation highlights a broader trend in the tech industry: the increasing complexity of hardware security. As GPUs and other hardware components become more powerful and integrated with cloud services and web-based applications, the security risks multiply. Security professionals must adopt a proactive approach to patching, monitoring, and securing their systems, ensuring that both hardware and software are regularly updated to address the latest vulnerabilities.

For users, the critical takeaway is clear: regular updates are essential. As Arm continues to release patches for its drivers, users should ensure they’re running the latest versions to minimize the risk of exploitation. Whether you are a developer working with GPUs, a system administrator managing hardware resources, or a regular user of Arm-powered devices, taking the time to update your drivers can prevent serious security breaches down the line.

References:

Reported By: https://cyberpress.org/arm-mali-gpu-kernel-driver-0-day-vulnerability/
https://stackoverflow.com
Wikipedia: https://www.wikipedia.org
Undercode AI: https://ai.undercodetesting.com

Image Source:

OpenAI: https://craiyon.com
Undercode AI DI v2: https://ai.undercode.helpFeatured Image