GHSA-RWXX-MRJM-WC2M

Vulnerability from github – Published: 2026-07-17 17:10 – Updated: 2026-07-17 17:10
VLAI
Summary
vLLM: ReDoS via structured_outputs.regex compiled without timeout in xgrammar and outlines backends
Details

Summary

The structured_outputs.regex API parameter passes a user-supplied regex string directly to grammar compiler backends with no compilation timeout. In the xgrammar backend, the string reaches compile_regex() with no guard. In the outlines backend, validate_regex_is_buildable() blocks structural issues (lookarounds, backreferences) but provides zero protection against exponential DFA state-space explosion. Patterns like (a+)+b pass all checks and hang the inference worker.

Root Cause

backend_xgrammar.py:91 — no timeout:

ctx = self.compiler.compile_regex(grammar_spec)

backend_outlines.py:299–330 — structural checks only, no complexity analysis:

def validate_regex_is_buildable(regex: str) -> None:
    sre_parse.parse(regex)   # AST parse only — does not detect exponential patterns
    _check_unsupported(...)  # blocks lookarounds/backrefs, not nested quantifiers

backend_outlines.py:64 — no timeout:

oc.Index(regex_string, vocabulary.inner)

Impact

Denial of service — one request with an adversarial regex pattern hangs an inference worker indefinitely.

Remediation

Wrap compile_regex() and oc.Index() calls in a thread with a deadline (e.g., 5 seconds). Add complexity analysis to validate_regex_is_buildable() to detect nested quantifier patterns before compilation.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
8.7 (High)
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "vllm"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "0.24.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-55574"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-07-17T17:10:37Z",
    "nvd_published_at": "2026-07-06T21:16:57Z",
    "severity": "HIGH"
  },
  "details": "## Summary\n\nThe `structured_outputs.regex` API parameter passes a user-supplied regex string directly to grammar compiler backends with no compilation timeout. In the xgrammar backend, the string reaches `compile_regex()` with no guard. In the outlines backend, `validate_regex_is_buildable()` blocks structural issues (lookarounds, backreferences) but provides zero protection against exponential DFA state-space explosion. Patterns like `(a+)+b` pass all checks and hang the inference worker.\n\n## Root Cause\n\n`backend_xgrammar.py:91` \u2014 no timeout:\n```python\nctx = self.compiler.compile_regex(grammar_spec)\n```\n\n`backend_outlines.py:299\u2013330` \u2014 structural checks only, no complexity analysis:\n```python\ndef validate_regex_is_buildable(regex: str) -\u003e None:\n    sre_parse.parse(regex)   # AST parse only \u2014 does not detect exponential patterns\n    _check_unsupported(...)  # blocks lookarounds/backrefs, not nested quantifiers\n```\n\n`backend_outlines.py:64` \u2014 no timeout:\n```python\noc.Index(regex_string, vocabulary.inner)\n```\n\n## Impact\n\nDenial of service \u2014 one request with an adversarial regex pattern hangs an inference worker indefinitely.\n\n## Remediation\n\nWrap `compile_regex()` and `oc.Index()` calls in a thread with a deadline (e.g., 5 seconds). Add complexity analysis to `validate_regex_is_buildable()` to detect nested quantifier patterns before compilation.",
  "id": "GHSA-rwxx-mrjm-wc2m",
  "modified": "2026-07-17T17:10:37Z",
  "published": "2026-07-17T17:10:37Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-rwxx-mrjm-wc2m"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-55574"
    },
    {
      "type": "WEB",
      "url": "https://github.com/vllm-project/vllm/pull/45118"
    },
    {
      "type": "WEB",
      "url": "https://github.com/vllm-project/vllm/commit/2b3006076c5e9bc4cda9e03e3641388de3c5c286"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2026-2304.yaml"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/vllm-project/vllm"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "vLLM: ReDoS via structured_outputs.regex compiled without timeout in xgrammar and outlines backends"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.





Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.



Related by attack behaviour

Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.