| ID |
CVE-2026-54234
|
| Sažetak |
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Prior to 0.24.0, a frontend-legal multi-request speculative decoding workload can cause the rejection sampler to produce a recovered token equal to the model vocabulary size boundary value, which is then converted to negative one when the engine selects the next live token for a request and is written back into the drafter's input ids; that out-of-vocabulary value is later consumed by the model's embedding and attention path and crashes the engine worker with a GPU device-side assertion. The same triggering request sequence is reachable through the public gRPC Generate and Abort endpoints, so a remote client that can send generation requests can crash the shared engine worker, aborting concurrent requests and causing a service-wide denial of service for other clients of the deployment until the worker is restarted. This issue is fixed in version 0.24.0. |
| Reference |
|
| CVSS |
| Base: | 7.5 |
| Impact: | 3.6 |
| Exploitability: | 3.9 |
|
| Pristup |
| Vektor | Složenost | Autentikacija |
| NETWORK |
LOW |
NONE |
|
| Impact |
| Povjerljivost | Cjelovitost | Dostupnost |
| NONE |
NONE |
HIGH |
|
| CVSS vektor |
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H |
| Zadnje važnije ažuriranje |
06-07-2026 - 21:16 |
| Objavljeno |
06-07-2026 - 21:16 |
