vLLM: Incomplete CVE-2025-62164 remediation can be bypassed by concurrent prompt parts

Exploitation

No confirmed exploitation recorded. Patch Next marks exploitation only from CISA KEV.

CVSS 4.0

6.3 · CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N · score source

EPSS

• CVE-2026-73557: 0.25% estimated exploitation activity in the next 30 days; percentile 16.12 (FIRST EPSS, 2026-11-03). This is a model estimate, not proof of exploitation or customer exposure.

As published by the source; long descriptions are shortened.

Executive Summary The follow-up protection for CVE-2025-62164 is incomplete at vLLM revision 26587f9519e22a5c4549ead7595ad9ca3229c4fd. It wraps serialized prompt-embedding reconstruction and dense conversion in torch.sparse.checksparsetensorinvariants(), but PyTorch 2.11.0 implements that context with save/enable/restore operations over process-global state. Two prompt-embedding parts in one /v1/chat/completions request are gathered concurrently on the event loop's default executor. When one context exits before the other loads its tensor, it can restore the global flag to False while the second part remains inside its guard. In a deterministic run against hash-verified source from the affected revision, the actual target loader rejected an invalid sparse payload as a negative control. The frozen chat tracker then scheduled benign and malicious parts on distinct asyncio0 and asyncio1 threads. The benign context exited, the malicious loader observed the invariant flag disabled, and…

Published by the source

4 Sep 2026, 21:39 UTC

Last modified by the source

12 Oct 2026, 11:22 UTC

Source last checked

26 Oct 2026, 13:55 UTC

1. Revision 1 · created · 26 Oct 2026

   Published from ghsa record GHSA-pr7f-p5mw-fc87

Preview the exact text before you copy it. Commands are read-only suggestions; review them before running anything.

# Patch Next advisory brief: CVE-2026-73557

Advisory: vLLM: Incomplete CVE-2025-62164 remediation can be bypassed by concurrent prompt parts
IDs: CVE-2026-73557, GHSA-pr7f-p5mw-fc87
Record: https://patchnext.com/advisories/cve-2026-73557 (revision 1, source checked 26 Oct 2026, 13:55 UTC)
Inventory: none attached. This brief comes from a public page and does not say whether any deployment is affected.
Severity: Medium (CVSS 6.3)
Exploitation: no confirmed exploitation recorded. Patch Next confirms exploitation only from CISA KEV.
Vendor response: Vendor fix available; fixed versions by package: vllm 0.26.0. A vendor fix does not mean a deployment is fixed.

## Affected components
- PyPI vllm: affected >= 0.21.0, < 0.26.0; fixed in 0.26.0

## Suggested inspection (read-only)
- `python -m pip show vllm`: prints the installed version. Run with the environment that serves production active.

## Supported action
Upgrade each affected component to a fixed version on its release line. For a transitive dependency, upgrade the parent package or add an override or constraint, then regenerate the lockfile. Redeploy every environment that ran an affected version.

## Required regression checks
- Run the test suite and a production build.
- Exercise the features that depend on vllm.
- Repeat the inspection and confirm that only fixed versions remain installed.

## Rollback considerations
- Keep the previous lockfile and deployment artifact. Rolling back reinstates the vulnerable version, so apply any documented mitigation while you fix forward.

## Unresolved questions
- Which deployments run an affected version?
- Exploitation status can change. Check again before lowering the priority.
- Mark the work resolved only after the fixed version is verified in production.

## Sources
- Verified: https://github.com/advisories/GHSA-pr7f-p5mw-fc87 (GitHub-reviewed advisory GHSA-pr7f-p5mw-fc87 lists affected ranges: PyPI vllm >=0.21.0 <0.26.0)

Commands are suggestions: review each one before running it.