CVE-2026-116181
개요
A flaw was found in vllm. An authenticated remote attacker could cause a Denial of Service (DoS) or memory corruption by submitting requests containing invalid stop token identifiers. Because the application does not verify whether provided stop tokens fall within the model's vocabulary range, indexing these values can crash the inference engine or trigger an out-of-bounds write in GPU memory.
내용
This vulnerability is rated as Moderate severity because exploitation is confined to service availability disruption and requires direct network access with permission to submit inference requests. In Red Hat OpenShift AI and Red Hat Enterprise Linux AI environments, supplying out-of-bounds stop token values in standard chat or completion queries causes the underlying inference worker to terminate abnormally due to an unhandled index exception. Although this disrupts active model serving workloads, the issue does not permit code execution or data exfiltration across container boundaries, and standard container orchestration will automatically restart the failing pod.
CVSS (Common Vulnerability Scoring System) 점수 세부 사항
Info alert:Important note
CVSS scores for open source components depend on vendor-specific factors (e.g. version or build chain). Therefore, Red Hat's score and impact rating can be different from NVD and other vendors. Red Hat remains the authoritative CVE Naming Authority (CNA) source for its products and services (see Red Hat classifications).
다음의 CVSS 지표 및 점수는 예비 정보로 검토 대상입니다.
CVSS v3 점수 분석
| Red Hat | NVD | cve.org | |
|---|---|---|---|
| 기본 점수 | 6.5 | N/A | 6.5 |
| 공격 벡터 | Network | N/A | Network |
| 공격 복잡성 | Low | N/A | Low |
| 필요한 권한 | Low | N/A | Low |
| 사용자 상호 작용 | None | N/A | None |
| 범위 | Unchanged | N/A | Unchanged |
| 기밀성 | None | N/A | None |
| 무결성에 미치는 영향 | None | N/A | None |
| 가용성에 미치는 영향 | High | N/A | High |
벡터
Red Hat: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
cve.org: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
취약점 이해 (CWE)
Integrity
Technical Impact: Modify Memory; Execute Unauthorized Code or Commands
Write operations could cause memory corruption. In some cases, an adversary can modify control data such as return addresses in order to execute unexpected code.
Availability
Technical Impact: DoS: Crash, Exit, or Restart
Attempting to access out-of-range, invalid, or unauthorized memory could cause the product to crash.
Other
Technical Impact: Unexpected State
Subsequent write operations can produce undefined or unexpected results.
