CVE-2026-84286

Description

A flaw was found in ExLlamaV3. The EXL3 loader processes a model-supplied index to select a CUDA kernel without verifying that the index is within valid bounds. An attacker can exploit this vulnerability by enticing a user to load a maliciously crafted model file, causing the application to invoke an out-of-bounds function pointer. This issue may result in arbitrary code execution under specific memory conditions or cause an application crash, resulting in a Denial of Service (DoS).

Statement

Red Hat Product Security has rated this vulnerability as Moderate because exploitation requires convincing a user or process to ingest an untrusted model file, and code execution is contingent on specific host memory layouts. ExLlamaV3 is not packaged or supported in Red Hat Enterprise Linux. In environments where inference workloads are restricted to vetted models from trusted internal registries, practical exposure is minimized.

Understanding the Weakness (CWE)

Confidentiality

Technical Impact: Read Memory

An attacker could get secret values such as cryptographic keys, PII, memory addresses, or other information that could be used in additional attacks.

Confidentiality

Technical Impact: Bypass Protection Mechanism

Out-of-bounds memory could contain memory addresses or other information that can be used to bypass ASLR and other protection mechanisms in order to improve the reliability of exploiting a separate weakness for code execution.

Availability

Technical Impact: DoS: Crash, Exit, or Restart

An attacker could cause a segmentation fault or crash by causing memory to be read outside of the bounds of the buffer. This is especially likely when the code reads a variable amount of data and assumes that a sentinel exists to stop the read operation, such as a NUL in a string.

Other

Technical Impact: Varies by Context

The read operation could produce other undefined or unexpected results.

Frequently Asked Questions

Want to get errata notifications? Sign up here.