Advisory

GHSA-grg2-63fw-f2qr

Also known as: CVE-2026-22773, PYSEC-2026-143

Summary

vLLM is vulnerable to DoS in Idefics3 vision models via image payload with ambiguous dimensions

Details

Summary
Users can crash the vLLM engine serving multimodal models that use the _Idefics3_ vision model implementation by sending a specially crafted 1x1 pixel image. This causes a tensor dimension mismatch that results in an unhandled runtime error, leading to complete server termination.

Details
The vulnerability is triggered when the image processor encounters a 1x1 pixel image with shape (1, 1, 3) in HWC (Height, Width, Channel) format. Due to the ambiguous dimensions, the processor incorrectly assumes the image is in CHW (Channel, Height, Width) format with shape (3, H, W). This misinterpretation causes an incorrect calculation of the number of image patches, resulting in a fatal tensor split operation failure.

Crash location: vllm/model_executor/models/idefics3.py line 672:

def _process_image_input(self, image_input: ImageInputs) -> torch.Tensor | list[torch.Tensor]:
    # ...
    num_patches = image_input["num_patches"]
    return [e.flatten(0, 1) for e in image_features.split(num_patches.tolist())]

The split() call fails because the computed num_patches value (17) does not match the actual tensor dimension (9):

RuntimeError: split_with_sizes expects split_sizes to sum exactly to 9 
(input tensor's size at dimension 0), but got split_sizes=[17]

This unhandled exception terminates the EngineCore process, crashing the server.

Affected Models
Any model using the Idefics3 architecture. The vulnerability was tested with HuggingFaceTB/SmolVLM-Instruct.

Impact
Denial of service by crashing the engine

Mitigation
Validating the input:

def _validate_image_dimensions(self, image_shape):
    h, w = image_shape[:2] if len(image_shape) == 3 else image_shape
    if h < MIN_IMAGE_SIZE or w < MIN_IMAGE_SIZE:
        raise ValueError(f"Image dimensions too small: {h}x{w}")

Managing the exception:

try:
    return [e.flatten(0, 1) for e in image_features.split(num_patches.tolist())]
except RuntimeError as e:
    logger.error(f"Image processing failed: {e}")
    raise InvalidImageError("Failed to process image features") from e

Fixes

  • https://github.com/vllm-project/vllm/pull/29881

Severity

Medium Basis: The source's own band: MODERATE

Dates

Published
2026-01-13 18:44 UTC
Modified
2026-09-10 03:50 UTC

Affected packages and fixes

vllm

PyPI · All systems

Affected

  • introduced 0.6.4, fixed 0.12.0 (ecosystem)

Fixed versions

0.12.0

Fix

  • Fixed version 0.12.0
    pip install --upgrade vllm==0.12.0

Fix 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

View this record on OSV.dev

Source

Source: GitHub Advisory Database (reviewed advisories) · CC BY 4.0

Sources & licences