CVE-2026-90543
Description détaillée
WWBN AVideo at commit c3edcc274c389816d434acadac07ee78eaf330c1 and earlier, with the Live plugin enabled, contains a missing authentication vulnerability in plugin/Live/socketMessageLiveOwner.json.php. The script reads the `key` and `msg` parameters from $_REQUEST, resolves the stream owner via LiveTransmition::keyExists, and verifies that the stream owner (not the caller) is an administrator; it performs no User::isLogged() check and enforces no CSRF token. As a result, an unauthenticated remote attacker who knows an administrator's Live stream key can call sendSocketSuccessMessageToUsers_id and deliver arbitrary in-app/socket notification messages to that administrator. The issue was unpatched at the time of publication.
Vecteur d'attaque (CVSS)
Dernières Vulnérabilités
CVE-2026-90555
vLLM versions before 0.28.0 fail to validate audio sample rate headers in the transcription endpoint, allowing authenticated clients to bypass duration checks. Attackers can submit forged FLAC headers with inflated sample rates to trigger excessive memory allocation and crash the API server process affecting all tenants.
CVE-2026-90554
vLLM versions >=0.10.2 and <0.28.0 do not apply any audio decode-size or duration limit when extracting audio from video input for NanoNemotronVL models. In nano_nemotron_vl.py, _extract_audio_from_videos calls load_audio_pyav(BytesIO(video_bytes)) without the max_duration_s or max_decode_bytes parameters, so neither VLLM_MAX_AUDIO_DECODE_DURATION_S nor VLLM_MAX_AUDIO_DECODE_BYTES is enforced (unlike the direct audio upload path in AudioMediaIO). When a NanoNemotronVL model is served with use_audio_in_video=True, an attacker who supplies a small, highly compressed video as multimodal input can force the server to allocate gigabytes of memory during audio decoding, resulting in a denial of service. Fixed in vLLM 0.28.0.
CVE-2026-90553
vLLM before 0.28.0 contains a remote code execution vulnerability in the LlavaOnevision2 processor loader that ignores the trust_remote_code parameter when loading remote processor classes. Attackers can craft a malicious model with arbitrary code in processing_llava_onevision2.py that executes with vLLM process authority even when trust_remote_code is set to False.