vLLM is an inference and serving engine for large language models. Prior to 0.27.0, an integer overflow in blockIdx.x * 2 * d in activation_kernels.cu can cause act_and_mul_kernel to consume another batched users input, allowing a request processed in the same inference batch to receive a partial or complete copy of another users inference result. This issue is fixed in version 0.27.0.
Weakness
The product performs a calculation that can produce an integer overflow or wraparound when the logic assumes that the resulting value will always be larger than the original value. This occurs when an integer value is incremented to a value that is too large to store in the associated representation. When this occurs, the value may become a very small or negative number.
Affected Software
| Name | Vendor | Start Version | End Version |
|---|---|---|---|
| Vllm | Vllm | * | 0.27.0 (excluding) |
| Red Hat AI Inference Server 3.5 | RedHat | rhaii/vllm-rocm-rhel9:1791425869 | * |
| Red Hat AI Inference Server 3.5 | RedHat | rhaii/vllm-gaudi-rhel9:1791427182 | * |
| Red Hat AI Inference Server 3.5 | RedHat | rhaii/vllm-cuda-rhel9:1791473631 | * |
| Red Hat AI Inference Server 3.5 | RedHat | rhaii/vllm-spyre-rhel9:1791818357 | * |
| Red Hat AI Inference Server 3.5 | RedHat | rhaii/vllm-cpu-rhel9:1791473614 | * |
| Red Hat Enterprise Linux AI 3.5 | RedHat | rhelai3/disk-image-cuda-rhel9:1793206758 | * |
| Red Hat Enterprise Linux AI 3.5 | RedHat | rhelai3/bootc-aws-cuda-rhel9:1793194817 | * |
| Red Hat Enterprise Linux AI 3.5 | RedHat | rhelai3/bootc-azure-cuda-rhel9:1793194815 | * |
| Red Hat Enterprise Linux AI 3.5 | RedHat | rhelai3/bootc-azure-rocm-rhel9:1793194834 | * |
| Red Hat Enterprise Linux AI 3.5 | RedHat | rhelai3/bootc-cuda-rhel9:1793135861 | * |
| Red Hat Enterprise Linux AI 3.5 | RedHat | rhelai3/bootc-gcp-cuda-rhel9:1793194815 | * |
| Red Hat Enterprise Linux AI 3.5 | RedHat | rhelai3/bootc-rocm-rhel9:1793135863 | * |
Potential Mitigations
- Use a language that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid.
- If possible, choose a language or compiler that performs automatic bounds checking.
- Use a vetted library or framework that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid [REF-1482].
- Use libraries or frameworks that make it easier to handle numbers without unexpected consequences.
- Examples include safe integer handling packages such as SafeInt (C++) or IntegerLib (C or C++). [REF-106]
- Perform input validation on any numeric input by ensuring that it is within the expected range. Enforce that the input meets both the minimum and maximum requirements for the expected range.
- Use unsigned integers where possible. This makes it easier to perform validation for integer overflows. When signed integers are required, ensure that the range check includes minimum values as well as maximum values.
- Understand the programming language’s underlying representation and how it interacts with numeric calculation (CWE-681). Pay close attention to byte size discrepancies, precision, signed/unsigned distinctions, truncation, conversion and casting between types, “not-a-number” calculations, and how the language handles numbers that are too large or too small for its underlying representation. [REF-7]
- Also be careful to account for 32-bit, 64-bit, and other potential differences that may affect the numeric representation.
Related Attack Patterns
References
- https://github.com/vllm-project/vllm/commit/451227cb3ff07989698fed982c2d3e4300257924
- https://github.com/vllm-project/vllm/issues/42860
- https://github.com/vllm-project/vllm/pull/49660
- https://github.com/vllm-project/vllm/releases/tag/v0.27.0
- https://github.com/vllm-project/vllm/security/advisories/GHSA-7m6h-x95x-82q5
- https://github.com/vllm-project/vllm/security/advisories/GHSA-7m6h-x95x-82q5