https://cve.radiocsirt.org Contains only the most 10 recent entries. http://www.rssboard.org/rss-specification python-feedgen en Wed, 07 Oct 2026 18:28:35 +0000 https://cve.radiocsirt.org/vuln/euvd-2026-333532 EUVD-2026-333532 https://cve.radiocsirt.org/vuln/euvd-2026-333532 https://cve.radiocsirt.org/vuln/fkie_cve-2026-55514 <p>vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is authorized to make a /v1/completions request can make such a request and induce a crash. This issue is fixed in version 0.24.0.</p> <p>vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is authorized to make a /v1/completions request can make such a request and induce a crash. This issue is fixed in version 0.24.0.</p> https://cve.radiocsirt.org/vuln/fkie_cve-2026-55514 https://cve.radiocsirt.org/vuln/ghsa-33cg-gxv8-3p8g <p><strong>Affected:</strong> PyPI: vllm</p> <p>### Summary _Short summary of the problem. Make the impact and severity as clear as possible. For example: An unsafe deserialization vulnerability allows any unauthenticated user to execute arbitrary code on the server._</p> <p>Sending a pure prompt embeds payload in a `/v1/completions` request with a model using M-RoPE causes the EngineCore to fail an assertion and fatally crash, shutting down the entire server application.</p> <p>Any remote user who is authorized to make a `/v1/completions` endpoint can trivially make such a request and induce a crash.</p> <p>### Details _Give all details on the vulnerability. Pointing to the incriminated source code is very helpful for the maintainer._</p> <p>In commit [56669c1](https://github.com/vllm-project/vllm/commit/56669c1f293d5c53b6a19ddf2f78802fa9fff2c2), a simple assert intended to be a type-narrowing assert was added to the `_init_mrope_positions` method in `GPUModelRunner` (the offending line on main at the time of writing: https://github.com/vllm-project/vllm/blob/2d481f8a946ee0521872af0f098674a8ee01ce4a/vllm/v1/worker/gpu_model_runner.py#L1588-L1607).</p> <p>```python assert req_state.prompt_token_ids is not None, ( &#34;M-RoPE requires prompt_token_ids to be available.&#34; ) ```</p> <p>This type narrowing assert is to prevent mypy errors later in the function because `None` is not a valid type for `mrope_model.get_mrope_input_positions`. Unfortunately, this assertion is not always true. `/v1/completions` requests that specify `prompt=None` and `p…</p> <p><strong>Affected:</strong> PyPI: vllm</p> <p>### Summary _Short summary of the problem. Make the impact and severity as clear as possible. For example: An unsafe deserialization vulnerability allows any unauthenticated user to execute arbitrary code on the server._</p> <p>Sending a pure prompt embeds payload in a `/v1/completions` request with a model using M-RoPE causes the EngineCore to fail an assertion and fatally crash, shutting down the entire server application.</p> <p>Any remote user who is authorized to make a `/v1/completions` endpoint can trivially make such a request and induce a crash.</p> <p>### Details _Give all details on the vulnerability. Pointing to the incriminated source code is very helpful for the maintainer._</p> <p>In commit [56669c1](https://github.com/vllm-project/vllm/commit/56669c1f293d5c53b6a19ddf2f78802fa9fff2c2), a simple assert intended to be a type-narrowing assert was added to the `_init_mrope_positions` method in `GPUModelRunner` (the offending line on main at the time of writing: https://github.com/vllm-project/vllm/blob/2d481f8a946ee0521872af0f098674a8ee01ce4a/vllm/v1/worker/gpu_model_runner.py#L1588-L1607).</p> <p>```python assert req_state.prompt_token_ids is not None, ( &#34;M-RoPE requires prompt_token_ids to be available.&#34; ) ```</p> <p>This type narrowing assert is to prevent mypy errors later in the function because `None` is not a valid type for `mrope_model.get_mrope_input_positions`. Unfortunately, this assertion is not always true. `/v1/completions` requests that specify `prompt=None` and `p…</p> https://cve.radiocsirt.org/vuln/ghsa-33cg-gxv8-3p8g https://cve.radiocsirt.org/vuln/pysec-2026-2303 <p><strong>Affected:</strong> PyPI: vllm</p> <p>vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is authorized to make a /v1/completions request can make such a request and induce a crash. This issue is fixed in version 0.24.0.</p> <p><strong>Affected:</strong> PyPI: vllm</p> <p>vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is authorized to make a /v1/completions request can make such a request and induce a crash. This issue is fixed in version 0.24.0.</p> https://cve.radiocsirt.org/vuln/pysec-2026-2303 https://cve.radiocsirt.org/vuln/wid-sec-w-2026-2176 <p>Ein Angreifer kann mehrere Schwachstellen in vllm ausnutzen, um einen Denial of Service Angriff durchzuführen.</p> <p>Ein Angreifer kann mehrere Schwachstellen in vllm ausnutzen, um einen Denial of Service Angriff durchzuführen.</p> https://cve.radiocsirt.org/vuln/wid-sec-w-2026-2176