vLLM versions >= 0.10.2 and < 0.13.0 are missing sparse tensor validation in multimodal embeddings processing. Because PyTorch disables sparse tensor invariant checks by default, an attacker can submit crafted embedding requests with malformed (negative or out-of-bounds) tensor indices, when the prompt-embeds feature is enabled, to trigger crashes or resource exhaustion (denial of service), with potential for out-of-bounds/write-what-where memory corruption. This continues CVE-2025-62164, whose prior fix only disabled the feature by default rather than addressing the root cause.
Review the upstream advisory, identify the affected product in your inventory, and apply the vendor update when one is available.
CVE-2026-56340 is listed in the HOL Guard supply-chain feed, but affected software is not mapped to a package. Review the upstream advisory for the affected product and vendor guidance. vLLM versions >= 0.10.2 and < 0.13.0 are missing sparse tensor validation in multimodal embeddings processing. Because PyTorch disables sparse tensor invariant checks by default, an attacker can submit crafted embedding requests with malformed (negative or out-of-bounds) tensor indices, when the prompt-embeds feature is enabled, to trigger crashes or resource exhaustion (denial of service), with potential for out-of-bounds/write-what-where memory corruption. This continues CVE-2025-62164, whose prior fix only disabled the feature by default rather than addressing the root cause.
Affected software not mapped. Review the upstream record for vendor-specific product and version guidance.
vLLM versions >= 0.10.2 and < 0.13.0 are missing sparse tensor validation in multimodal embeddings processing. Because PyTorch disables sparse tensor invariant checks by default, an attacker can submit crafted embedding requests with malformed (negative or out-of-bounds) tensor indices, when the prompt-embeds feature is enabled, to trigger crashes or resource exhaustion (denial of service), with potential for out-of-bounds/write-what-where memory corruption. This continues CVE-2025-62164, whose prior fix only disabled the feature by default rather than addressing the root cause.
Review the upstream advisory, identify the affected product in your inventory, and apply the vendor update when one is available.
CVE-2026-56340 is listed in the HOL Guard supply-chain feed, but affected software is not mapped to a package. Review the upstream advisory for the affected product and vendor guidance. vLLM versions >= 0.10.2 and < 0.13.0 are missing sparse tensor validation in multimodal embeddings processing. Because PyTorch disables sparse tensor invariant checks by default, an attacker can submit crafted embedding requests with malformed (negative or out-of-bounds) tensor indices, when the prompt-embeds feature is enabled, to trigger crashes or resource exhaustion (denial of service), with potential for out-of-bounds/write-what-where memory corruption. This continues CVE-2025-62164, whose prior fix only disabled the feature by default rather than addressing the root cause.
Affected software not mapped. Review the upstream record for vendor-specific product and version guidance.
Reported by NVD (nvd).
HOL Guard can help your team monitor supply-chain activity while the upstream record is clarified.
Explore HOL GuardReported by NVD (nvd).
HOL Guard can help your team monitor supply-chain activity while the upstream record is clarified.
Explore HOL Guard