Monthly Archives: February 2025

AI and ML, Potential Risk of CVE

CVE-2024-36353: CROSS-PROCESS GPU MEMORY DISCLOSURE (27-02-2025)

Leave a comment

Preface: Regarding its use in HPC clusters, the Radeon PRO V710 is indeed suitable. It is supported by AMD’s ROCm platform, which is optimized for HPC and AI workloads. Additionally, it is used in Azure’s NVads V710 v5-series virtual machines, which are designed for GPU-accelerated applications, including HPC.

Background: The global memory of the AMD Radeon™ PRO V710 is the 28 GB of GDDR6 memory. This memory is connected via a 224-bit memory interface and operates at an effective speed of 18 Gbps1. The memory is used for storing data that the GPU processes, such as textures, frame buffers, and other computational data.

The NVIDIA Container Toolkit is specifically designed to work with NVIDIA GPUs and their CUDA framework. It is not compatible with AMD GPUs. For AMD GPUs, you should use the ROCm (Radeon Open Compute) platform, which provides similar functionality for containerized environments.

OpenCL (Open Computing Language) in AMD ROCm (Radeon Open Compute) serves as a framework for writing programs that execute across heterogeneous platforms, including CPUs, GPUs, and other processors. Specifically, in the context of AMD ROCm, OpenCL allows developers to harness the computational power of AMD GPUs for high-performance, data-parallel computing tasks.

Vulnerability details: Insufficient clearing of GPU global memory could allow a malicious process running on the same GPU to read left over memory values potentially leading to loss of confidentiality.

Official announcement: Please refer to the link for details https://www.amd.com/en/resources/product-security/bulletin/amd-sb-6019.html

AI and ML, Potential Risk of CVE

CVE-2024-0148: NVIDIA Jetson Linux and IGX OS image contains a vulnerability in the UEFI firmware RCM boot mode (25-02-2025)

Leave a comment

Preface: NVIDIA IGX Orin software is used by a variety of organizations, particularly those in industrial and medical environments. This platform is designed to support AI applications at the edge, providing high performance, advanced functional safety, and security.

Some specific use cases include:

  • Industrial Automation: Companies use IGX Orin to enhance manufacturing processes with AI-driven automation and predictive maintenance.
  • Healthcare: Medical institutions leverage IGX Orin for AI-powered diagnostics, medical imaging, and patient monitoring.
  • Robotics: Robotics companies utilize IGX Orin for developing intelligent robots that can operate safely alongside humans.

The platform’s versatility and robust support make it suitable for any organization looking to deploy AI solutions in demanding environments.

Background: The NVIDIA IGX Orin Developer Kit runs the Holopack 2.0 Developer Preview software. Holopack is a comprehensive solution for end-to-end GPU accelerated AI application development and testing. Holopack supports two GPU modes:

iGPU – Holopack deploys drivers and libraries to support NVIDIA Ampere sets on NVIDIA IGX Orin modules into the GPU.

dGPU – Holopack deploys drivers and libraries to support optional NVIDIA RTX A6000 connected to PCIe slot Discrete GPU.

Its high-performance, low-power computing for deep learning, and computer vision makes Jetson the ideal platform for compute-intensive projects. The Jetson platform includes a variety of Jetson modules with NVIDIA JetPack™ SDK.

Vulnerability details: NVIDIA Jetson Linux and IGX OS image contains a vulnerability in the UEFI firmware RCM boot mode, where an unprivileged attacker with physical access to the device could load untrusted code. A successful exploit might lead to code execution, escalation of privileges, data tampering, denial of service, and information disclosure. The scope of the impacts can extend to other components.

Remark: UEFI supply-chain allows for many of these shared libraries to be integrated in various ways, including compiled from source, licensed for modification and reuse and finally as a dynamic or static linked executable.

Official announcement: Please refer to the vendor announcement for details – https://nvidia.custhelp.com/app/answers/detail/a_id/5617

Science, System

Python , have ever you though? (25-02-2025)

Leave a comment

Preface: Maintaining a satellite’s orbit involves a combination of precise calculations and regular adjustments. Here are the key factors:

  1. Velocity and Gravity: A satellite stays in orbit by balancing its velocity (speed in a straight line) with the gravitational pull of the Earth. The satellite must travel fast enough to counteract the pull of gravity, which keeps it in a stable orbit.
  2. Orbital Station-Keeping: This involves small adjustments using thrusters to correct any deviations in the satellite’s path. These maneuvers ensure the satellite remains in its designated orbit.
  3. Fuel Management: Satellites carry a limited amount of fuel for these adjustments. Efficient fuel management is crucial for prolonging the satellite’s operational life.
  4. Monitoring and Control: Ground stations continuously monitor satellites and send commands to perform necessary adjustments. This helps in maintaining the satellite’s orbit and addressing any potential issues.

Background: The PyEphem module provides highly precise data on the planets and our solar system. This module leverages an extremely robust C library that allows you to pinpoint planets, perform interplanetary calculations and discover more data than you’ll ever know what to do with.

Best practice: If you’re using PyEphem, it’s a good idea to keep your Python environment and libraries up to date and to check the module’s GitHub repository for any reported issues or updates.

AI and ML

When the Sumerian advanced civilization met Cray HPC

Leave a comment

(24-02-2025)

Preface: Is linear algebra used in real life? An example of where there is a lot of research on these things is in sparse matrix analysis, which comes up a lot in real world applications of linear algebra. For some buzzwords, popular topics like machine learning, neural networks, and computer graphics all use huge amounts of linear algebra.

Since a box’s length is independent of its width and breadth, space has three dimensions. Since any point in space may be described by a linear combination of three independent vectors, space is considered to be three-dimensional in the technical language of linear algebra.

In Einstein’s special relativity theory we live in 4 dimensional spacetime. Though the way we normally “imagine” the world, we tend to believe that we live in a 3 dimensional Newtonian space with a separate absolute time dimension.

Introduction: AI calculations often rely on various mathematical techniques, including linear algebra, Fourier transforms, and sparse matrix operations.

Some of the key math libraries in ROCm include:

  • rocBLAS: A library for basic linear algebra subprograms.
  • rocFFT: A library for fast Fourier transforms.
  • rocRAND: A library for random number generation.
  • rocSOLVER: A library for solving linear algebra problems.
  • rocSPARSE: A library for sparse matrix operations

These libraries are optimized for AMD hardware and provide similar functionality to NVIDIA’s cuBLAS, cuFFT, cuRAND, etc., making it easier for developers to port their applications between different hardware platforms.

What does ROCm stand for? ROCm initially stood for Radeon Open Compute platform; however, due to Open Compute being a registered trademark, ROCm is no longer an acronym — it is simply AMD’s open-source stack designed for GPU compute.

Official reference: If you are interested in ROCm, please refer to the following link – https://rocm.docs.amd.com/en/docs-5.7.1/reference/gpu_libraries/math.html

Potential Risk of CVE

CVE-2024-46975 : GPU DDK – rgxfw_write_robustness_buffer allows arbirtary catreg set mapping (23rd Feb 2025)

Leave a comment

Preface: A Memory Management Unit (MMU) is the hardware system which performs both virtual memory mapping and checks the current privilege to keep user processes separated from the operating system — and each other. In addition it helps to prevent caching of ‘volatile’ memory regions (such as areas containing I/O peripherals.

Background: Generally speaking, GPU firmware and driver functionality do utilize the L2 cache. The L2 cache in a GPU is a larger, shared cache that helps improve memory access speeds and reduce latency for various operations. It plays a crucial role in optimizing the performance of GPU-accelerated tasks by storing frequently accessed data closer to the GPU cores.

The L2 cache is particularly important for managing memory access across different Streaming Multiprocessors (SMs) within the GPU. By efficiently handling memory requests and reducing the need for crossbar communication, the L2 cache helps minimize latency and improve overall task performance.

Vulnerability details: Kernel software installed and running inside a Guest VM may exploit memory shared with the GPU Firmware to write data into another Guest’s virtualised GPU memory.

Official announcement: Please refer to the vendor announcement for details – https://nvd.nist.gov/vuln/detail/CVE-2024-46975

AI and ML, Potential Risk of CVE

CVE‑2024‑53870, CVE‑2024‑53871, CVE‑2024‑53872, CVE‑2024‑53873, CVE‑2024‑53874, CVE‑2024‑53875, CVE‑2024‑53876, CVE‑2024‑53877, CVE‑2024‑53878 and CVE‑2024‑53879 (21-02-2025)

Leave a comment

Released on February 18, 2025

Preface: In NVIDIA CUDA, cuobjdump and nvdisasm are two binary utilities used for examining and disassembling CUDA binaries (cubin files).

cuobjdump

  • Purpose: It can disassemble CUDA binaries and extract PTX (Parallel Thread Execution) code from host binaries, executables, object files, static libraries, and external fatbinary files.
  • Usage: cuobjdump is versatile as it accepts both cubin files and host binaries.
  • Features: It provides basic disassembly and extraction capabilities but lacks advanced display options and control flow analysis.

nvdisasm

  • Purpose: It is specifically designed to disassemble cubin files.
  • Usage: Unlike cuobjdump, nvdisasm only accepts cubin files.
  • Features: It offers richer output options, including advanced display options and control flow analysis.

These tools are essential for developers who need to inspect and debug the compiled CUDA code.

Background: Parallel processing is a method in computing of running two or more processors (CPUs) to handle separate parts of an overall task. Breaking up different parts of a task among multiple processors will help reduce the amount of time to run a program. GPUs render images more quickly than a CPU because of its parallel processing architecture, which allows it to perform multiple calculations across streams of data simultaneously. The CPU is the brain of the operation, responsible for giving instructions to the rest of the system, including the GPU(s).

NVIDIA CUDA provides a simple C/C++ based interface. The CUDA compiler leverages parallelism built into the CUDA programming model as it compiles your program into code.
CUDA is a parallel computing platform and programming interface model created by Nvidia for the development of software which is used by parallel processors. It serves as an alternative to running simulations on traditional CPUs.

Vulnerability details:

The following two design flaws are associated with these CVEs:

CVE‑2024‑53870, CVE‑2024‑53871, CVE‑2024‑53872, CVE‑2024‑53873, CVE‑2024‑53874, CVE‑2024‑53875, CVE‑2024‑53876, CVE‑2024‑53877, CVE‑2024‑53878 and CVE‑2024‑53879

NVIDIA CUDA toolkit for Linux and Windows contains a vulnerability in the cuobjdump binary, where a user could cause a crash by passing a malformed ELF file to cuobjdump. A successful exploit of this vulnerability might lead to a partial denial of service.

NVIDIA CUDA toolkit for all platforms contains a vulnerability in the nvdisasm binary, where a user could cause an out-of-bounds read by passing a malformed ELF file to nvdisasm. A successful exploit of this vulnerability might lead to a partial denial of service.

Official announcement: Please refer to the vendor announcement for details – https://nvidia.custhelp.com/app/answers/detail/a_id/5594

Potential Risk of CVE

CVE-2024-57258 – Integer overflows in memory allocation in Das U-Boot  (19-02-2025)

Leave a comment

Preface: U-Boot is both a first-stage and second-stage bootloader. It is loaded by the system’s ROM (e.g. on-chip ROM of an ARM CPU) from a supported boot device, such as an SD card, SATA drive, NOR flash (e.g. using SPI or I²C), or NAND flash.

Background: Das U-Boot is an open source, primary boot loader used in embedded devices to package the instructions to boot the device’s operating system kernel. U-Boot uses commands similar to the BASH shell to manipulate environment variables. U-Boot supports TFTP (Trivial FTP), a stripped down FTP. So that user authentication is not required for downloading images into the board’s RAM

LK is the abbreviation of Little Kernel. LK is commonly used as bootloader in the Android system of Qualcomm platform. It is an open source project. LK is the boot part of the whole system, so it is not independent. However, LK currently only supports arm and x86 architectures. The notable feature of LK is that it implements a simple thread mechanism. And deeply customized and used with Qualcomm’s processors.

Vulnerability details: Integer overflows in memory allocation in Das U-Boot before 2025.01-rc1 occur for a crafted squashfs filesystem via sbrk, via request2size, or because ptrdiff_t is mishandled on x86_64.

Remark: An integer overflow is a type of software vulnerability that occurs when a variable, such as an integer, exceeds its assigned memory space. This can result in unexpected behavior or security issues, such as allowing an attacker to execute arbitrary code.

Official announcement: Please refer to the link for details – https://nvd.nist.gov/vuln/detail/CVE-2024-57258

Potential Risk of CVE

nodejs: GOAWAY HTTP/2 frames cause memory leak outside heap (CVE-2025-23085) 17-02-2025

Leave a comment

Preface: If artificial intelligence could create the world. Do you know how his creation differs from Genesis? Artificial intelligence focuses on efficiency, and everything needs to be fast.

But God is concerned with the balance of nature. Therefore, the development of everything is not rapid.

Background: HTTP/2 enables full request and response multiplexing. In practice, this means a connection made to a web server from your browser can be used to send multiple requests and receive multiple responses. This eliminates some of the time it takes to establish a new connection for each request.

The GOAWAY frame in HTTP/2 (type=0x7) is used to initiate the shutdown of a connection or to signal serious error conditions. When a server sends a GOAWAY frame, it tells the client to stop creating new streams on the connection. However, it allows the server to finish processing any streams that were already in progress. This mechanism is useful for administrative actions, such as server maintenance, as it allows for a graceful shutdown without abruptly terminating ongoing request.

Vulnerability details: A memory leak could occur when a remote peer abruptly closes the socket without sending a GOAWAY notification. Additionally, if an invalid header was detected by nghttp2, causing the connection to be terminated by the peer, the same leak was triggered. This flaw could lead to increased memory consumption and potential denial of service under certain conditions. This vulnerability affects HTTP/2 Server users on Node.js v18.x, v20.x, v22.x and v23.x.

Official announcement: Please refer to the link for details – https://access.redhat.com/errata/RHSA-2025:1613

Potential Risk of CVE

Cache-based Side-Channel Attack Against SEV (18th Feb 2024)

Leave a comment

Originally posted by AMD 3rd Feb 2025

Updated Acknowledgement – 2025-02-17

Preface: FIPS 186-5 removes DSA as an approved digital signature algorithm “due to a lack of use by industry and based on academic analyses that observed that implementations of DSA may be vulnerable to attacks if domain parameters are not properly generated.

February 3, 2023 – NIST published Federal Information Processing Standard (FIPS) 186-5, Digital Signature Standard (DSS), along with NIST Special Publication (SP) 800-186, Recommendations for Discrete Logarithm-based Cryptography: Elliptic Curve Domain Parameters.  

Background: The SEV feature relies on elliptic-curve cryptography for its secure key generation, which runs when a VM is launched. The VM initiates the elliptic-curve algorithm by providing points along its NIST (National Institute of Standards and Technology) curve and relaying the data based on the private key of the machine.

Vulnerability details: AMD has received a report from researchers at National Taiwan University detailing cache-based side-channel attacks against Secure Encrypted Virtualization (SEV).

Remedy: AMD recommends software developers employ existing best practices for prime and probe attacks (including constant-time algorithms) and avoid secret-dependent data accesses where appropriate.  AMD also recommends following previously published guidance regarding Spectre type attacks (refer to the link in the reference section below), as it believes the previous guidance remains applicable to mitigate these vulnerabilities.

Supplement: The lack of authentication in the memory encryption is one major drawback of the Secure Memory Encryption (SME) design, which has been demonstrated in fault injection attacks. SEV inherits this security issue. Therefore, a malicious hypervisor may alter the ciphertext of the encrypted memory without triggering faults in the guest VM.

Office announcement: Please refer to the link for details – https://www.amd.com/en/resources/product-security/bulletin/amd-sb-3010.html

AI and ML, Potential Risk of CVE

CVE-2024-0112 – about NVIDIA Jetson AGX Orin™ and NVIDIA IGX Orin software  (17th Feb 2025)

Leave a comment

The original security advisory was issued on February 11, 2025.

Preface: NVIDIA IGX Orin software is used by a variety of organizations, particularly those in industrial and medical environments. This platform is designed to support AI applications at the edge, providing high performance, advanced functional safety, and security.

Some specific use cases include:

  • Industrial Automation: Companies use IGX Orin to enhance manufacturing processes with AI-driven automation and predictive maintenance.
  • Healthcare: Medical institutions leverage IGX Orin for AI-powered diagnostics, medical imaging, and patient monitoring.
  • Robotics: Robotics companies utilize IGX Orin for developing intelligent robots that can operate safely alongside humans.

The platform’s versatility and robust support make it suitable for any organization looking to deploy AI solutions in demanding environments.

Background: The NVIDIA IGX Orin Developer Kit runs the Holopack 2.0 Developer Preview software. Holopack is a comprehensive solution for end-to-end GPU accelerated AI application development and testing. Holopack supports two GPU modes:

iGPU – Holopack deploys drivers and libraries to support NVIDIA Ampere sets on NVIDIA IGX Orin modules into the GPU.

dGPU – Holopack deploys drivers and libraries to support optional NVIDIA RTX A6000 connected to PCIe slot Discrete GPU.

Its high-performance, low-power computing for deep learning, and computer vision makes Jetson the ideal platform for compute-intensive projects. The Jetson platform includes a variety of Jetson modules with NVIDIA JetPack™ SDK.

Vulnerability details: NVIDIA Jetson AGX Orin™ and NVIDIA IGX Orin software contain a vulnerability where an attacker can cause an improper input validation issue by escalating certain permissions to a limited degree. A successful exploit of this vulnerability might lead to code execution, denial of service, data corruption, information disclosure, or escalation of privilege.

Official announcement: Please refer to the vendor announcement for details – https://nvidia.custhelp.com/app/answers/detail/a_id/5611