Skip to content
Fencer Intel
high 7.5

CVE-2026-53923 json

vLLM GGUF Kernels: int64_t to int truncation of tensor dimensions causes GPU buffer overflow

published 4mo ago modified 17d ago also GHSA-5jv2-g5wq-cmr4 , PYSEC-2026-3403 , EUVD-2026-38400

cvss
7.5
version 3.1, NVD
cisa kev
Not listed
not in CISA's catalog
epss
0.5%
EPSS 0.5% chance of exploitation in 30 days, percentile 42%
public exploits
0
none known

Summary

vLLM is an inference and serving engine for large language models (LLMs). From 0.5.5 until 0.23.1rc0, integer truncation of tensor dimensions in vLLM's GGUF dequantize kernels (csrc/quantization/gguf/gguf_kernel.cu) causes partial tensor processing. The output tensor is allocated at full size via torch::empty (uninitialized memory), but the dequantize CUDA kernel processes only a truncated number of elements. The unfilled portion of the output tensor retains whatever was previously in GPU memory. In multi-tenant inference deployments, this residual GPU memory may contain tensor data from other users' inference requests, constituting information disclosure. This vulnerability is fixed in 0.23.1rc0.

Patch / fix

  • pypi vllm 0.24.0
  • vllm vllm 0.23.1

Is CVE-2026-53923 in your dependencies? Fencer shows which of your repositories are affected as soon as an advisory lands, and opens the fix as a pull request.

Check with Fencer

Vendor advisories

What the vendors themselves say about their products: affected, fixed or not affected, and what to do.

red hat product security vex CVE-2026-53923 updated cvss 4.3

vllm: vLLM: Information disclosure via integer truncation

  • Red Hat AI Inference Server 3.3 rhaiis/vllm-cuda-rhel9 1787161382, rhaiis/vllm-rocm-rhel9 1787161803, rhaiis/vllm-spyre-rhel9 1787161776
    fixed: vendor fix “No mitigation is required for unaffected deployments. Restrict untrusted access to inference APIs as a general hardening measure.” RHSA-2026:59138
  • Red Hat Enterprise Linux AI 3.3 rhelai3/bootc-cuda-rhel9 1788260684, rhelai3/bootc-rocm-rhel9 1788260620, rhelai3/bootc-aws-cuda-rhel9 1788273908, rhelai3/bootc-gcp-cuda-rhel9 1788273977, rhelai3/disk-image-cuda-rhel9 1788290314, rhelai3/bootc-azure-cuda-rhel9 1788273909, +1 more
    fixed: vendor fix “No mitigation is required for unaffected deployments. Restrict untrusted access to inference APIs as a general hardening measure.” RHSA-2026:62336
  • Red Hat AI Inference Server rhaii/vllm-cpu-rhel9, rhaii/vllm-tpu-rhel9, rhaiis/vllm-cpu-rhel9, rhaiis/vllm-tpu-rhel9, rhaii/vllm-gaudi-rhel9, rhaii/vllm-neuron-rhel9, +1 more Red Hat Enterprise Linux AI (RHEL AI) 3 rhelai3/bootc-gaudi-rhel9 Red Hat OpenShift AI (RHOAI) rhoai/odh-vllm-gaudi-rhel9, rhoai/odh-kserve-agent-rhel9, rhoai/odh-kserve-router-rhel9, rhoai/odh-llm-d-kv-cache-rhel9, rhoai/odh-kserve-controller-rhel9, rhoai/odh-kserve-storage-initializer-rhel9
    not affected: vulnerable code not present

Red Hat CSAF VEX © Red Hat, Inc., from https://security.access.redhat.com/data/csaf/v2/vex/; changed: reduced to product statuses and remediations. CC BY 4.0 · csaf

CISA SSVC

Exploitation
none
Automatable
no
Technical impact
partial
Assessed
4mo ago

Exploit prediction

0.48% chance of exploitation in the next 30 days, higher than 42% of all scored CVEs (FIRST EPSS, 2026-11-11).

CVSS scores

assigned by version score vector
CNA 4.0 5.3 CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N
CNA 4.0 5.3 CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:L/VI:L/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X
NVD 3.1 7.5 CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N
GitHub 3.1 7.5 CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N

Affected packages

ecosystem package affected fixed in
pypi vllm
>=0.5.5, <0.24.0
>=0.5.5, <=0.23.0
0.24.0

Affected products

  • vllm-project vllm

Classification

Weakness
  • CWE-681 Incorrect Conversion between Numeric Types
  • CWE-200 Exposure of Sensitive Information to an Unauthorized Actor

References

Sources

Fencer Intel keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.