vllm
Affected
- introduced
0, fixed0.30.0(ecosystem)
Fixed versions
0.30.0
Fix
- Fixed version
0.30.0pip install --upgrade vllm==0.30.0
Advisory
Also known as: CVE-2026-105757, PYSEC-2026-4199
vLLM: Structured-output request errors escape the request boundary and terminate the shared EngineCore — engine-fatal denial of service (3 sites)
Affected
vllmSummary
Three structured-output request paths let an ordinary request reach a condition that raises an uncaught, engine-fatal exception instead of a per-request validation error. The failure is not confined to the request that caused it: it escapes into EngineCore's busy loop and triggers a fatal _send_engine_dead(), so one malformed structured-output request denies service to all concurrent and subsequent tenants of that engine. The requests are ordinary API calls; the client does not need any privileged configuration beyond the (often default) structured-output feature.
The shared root cause is the absence of a per-request exception boundary around structured-output grammar/token handling: a value that should fail as a request-scoped validation error instead propagates as an uncaught exception (or bypasses frontend validation entirely) and reaches the scheduler/engine loop, which treats the failure as fatal.
These sites are distinct from the published fixes for GHSA-6qc9-v4r8-22xg and GHSA-8wr5-jm2h-8r4f: GHSA-6qc9 (PR #17623) added frontend json-schema/regex/type validation but the completion check on v0.25.1 still catches only TimeoutError, so Site 1's valid duplicate-root EBNF via the latched xgrammar backend still re-raises and kills EngineCore; GHSA-8wr5 (PR #44744) fixes recovered-token state in the eagle spec-decode path and explicitly does not treat trailing -1 padding as the fault, so Site 2's -1 reaching backend_guidance.validate_tokens() via ngram_gpu survives it.
Affected code
Links pinned to the confirmed commit `752a3a504485` (v0.25.1). The three sites are:
Site 1 — latched-backend compile exception escapes. StructuredOutputManager permanently latches one backend on the first structured-output request and always compiles through it, ignoring the per-request auto backend selection. A grammar that auto accepts only via a fallback backend (e.g. a duplicate-root EBNF) then makes the latched xgrammar compiler raise; the exception is stored in the grammar Future, and the completion check catches only TimeoutError, so Future.result() re-raises during scheduler promotion and kills EngineCore.
"auto": `vllm/config/structured_outputs.py#L21`.auto validator tries xgrammar and silently falls back, recording the resolved backend on the request: `vllm/sampling_params.py#L1004-L1035` (fields at `#L85-L88`)._create_grammar() at `#L173-L184`.TimeoutError: `vllm/v1/structured_output/request.py#L48-L63` (result(timeout=0.0001) at `#L55`, except TimeoutError at `#L57`)._send_engine_dead() at `#L1470`).The escaping-exception sink: the completion check only handles TimeoutError, so any compile exception stored in the Future re-raises out of result() during scheduler promotion.
# vllm/v1/structured_output/request.py Lines 48-59
def _check_grammar_completion(self) -> bool:
# NOTE: We have to lazy import to gate circular imports
from vllm.v1.request import RequestStatus
if isinstance(self._grammar, Future):
try:
# We will check whether the future is ready within 100 us
self._grammar = self._grammar.result(timeout=0.0001)
self.status = RequestStatus.WAITING
except TimeoutError:
return False
return True
Site 2 — guidance treats `ngram_gpu` `-1` padding as a token id. The ngram_gpu proposer pads a fixed-width draft-token row with -1 and separately records num_valid_draft_tokens, but the scheduler forwards the untrimmed padded row into GuidanceGrammar.validate_tokens(), which passes -1 to the native llguidance matcher; the matcher cannot convert a negative value to its unsigned token type and raises an OverflowError, uncaught and fatal.
-1 and computes num_valid_draft_tokens: `vllm/v1/spec_decode/ngram_proposer_gpu.py#L189-L209`._get_draft_token_ids_cpu() materializes the full padded row, not trimmed to the valid count: `vllm/v1/worker/gpu_model_runner.py#L4835-L4849`.validate_tokens(...): `vllm/v1/core/sched/scheduler.py#L1967` and `#L1997`.validate_tokens(tokens) at `#L192`).OverflowError is treated as fatal: `vllm/v1/engine/core.py#L1229-L1234`.Medium Basis: The source's own band: MODERATE
0, fixed 0.30.0 (ecosystem)0.30.0
0.30.0pip install --upgrade vllm==0.30.0Fix commands are the export's own generic commands. Ojuju never runs them. Check them against your setup first.
Check your own computer: guard packages check
Source: GitHub Advisory Database (reviewed advisories) · CC BY 4.0