Summary
In the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently, it serializes PIL.Image.Image objects using only obj.tobytes(), which returns only the raw pixel data, without including metadata such as the image’s shape (width, height, mode). As a result, two images of different sizes (e.g., 30x100 and 100x30) with the same pixel byte sequence could generate the same hash value. This may lead to hash collisions, incorrect cache hits, and even data leakage or security risks.
Details
- Affected file:
vllm/multimodal/hasher.py
- Affected method:
MultiModalHasher.serialize_item
https://github.com/vllm-project/vllm/blob/9420a1fc30af1a632bbc2c66eb8668f3af41f026/vllm/multimodal/hasher.py#L34-L35
- Current behavior: For
Image.Image instances, only obj.tobytes() is used for hashing.
- Problem description:
obj.tobytes() does not include the image’s width, height, or mode metadata.
- Impact: Two images with the same pixel byte sequence but different sizes could be regarded as the same image by the cache and hashing system, which may result in:
- Incorrect cache hits, leading to abnormal responses
- Deliberate construction of images with different meanings but the same hash value
Recommendation
In the serialize_item method, serialization of Image.Image objects should include not only pixel data, but also all critical metadata—such as dimensions (size), color mode (mode), format, and especially the info dictionary. The info dictionary is particularly important in palette-based images (e.g., mode 'P'), where the palette itself is stored in info. Ignoring info can result in hash collisions between visually distinct images with the same pixel bytes but different palettes or metadata. This can lead to incorrect cache hits or even data leakage.
Summary:
Serializing only the raw pixel data is insecure. Always include all image metadata (size, mode, format, info) in the hash calculation to prevent collisions, especially in cases like palette-based images.
Impact for other modalities
For the influence of other modalities, since the video modality is transformed into a multi-dimensional array containing the length, width, time, etc. of the video, the same problem exists due to the incorrect sequence of numpy as well.
For audio, since the momo function is not enabled in librosa.load, the loaded audio is automatically encoded into single channels by librosa and returns a one-dimensional array of numpy, thus keeping the structure of numpy fixed and not affected by this issue.
Fixes
References
kexinoh Reporter
DarkLight1337 Remediation developer
russellb Coordinator
Summary
In the file
vllm/multimodal/hasher.py, theMultiModalHasherclass has a security and data integrity issue in its image hashing method. Currently, it serializesPIL.Image.Imageobjects using onlyobj.tobytes(), which returns only the raw pixel data, without including metadata such as the image’s shape (width, height, mode). As a result, two images of different sizes (e.g., 30x100 and 100x30) with the same pixel byte sequence could generate the same hash value. This may lead to hash collisions, incorrect cache hits, and even data leakage or security risks.Details
vllm/multimodal/hasher.pyMultiModalHasher.serialize_itemhttps://github.com/vllm-project/vllm/blob/9420a1fc30af1a632bbc2c66eb8668f3af41f026/vllm/multimodal/hasher.py#L34-L35
Image.Imageinstances, onlyobj.tobytes()is used for hashing.obj.tobytes()does not include the image’s width, height, or mode metadata.Recommendation
In the
serialize_itemmethod, serialization ofImage.Imageobjects should include not only pixel data, but also all critical metadata—such as dimensions (size), color mode (mode), format, and especially theinfodictionary. Theinfodictionary is particularly important in palette-based images (e.g., mode'P'), where the palette itself is stored ininfo. Ignoringinfocan result in hash collisions between visually distinct images with the same pixel bytes but different palettes or metadata. This can lead to incorrect cache hits or even data leakage.Summary:
Serializing only the raw pixel data is insecure. Always include all image metadata (
size,mode,format,info) in the hash calculation to prevent collisions, especially in cases like palette-based images.Impact for other modalities
For the influence of other modalities, since the video modality is transformed into a multi-dimensional array containing the length, width, time, etc. of the video, the same problem exists due to the incorrect sequence of numpy as well.
For audio, since the momo function is not enabled in librosa.load, the loaded audio is automatically encoded into single channels by librosa and returns a one-dimensional array of numpy, thus keeping the structure of numpy fixed and not affected by this issue.
Fixes
References