CWE-502
AllowedDeserialization of Untrusted Data
Abstraction: Base · Status: Draft
The product deserializes untrusted data without sufficiently ensuring that the resulting data will be valid.
6035 vulnerabilities reference this CWE, most recent first.
GHSA-668Q-QRV7-99FM
Vulnerability from github – Published: 2021-12-17 20:00 – Updated: 2021-12-17 19:25In logback version 1.2.7 and prior versions, an attacker with the required privileges to edit configurations files could craft a malicious configuration allowing to execute arbitrary code loaded from LDAP servers.
{
"affected": [
{
"package": {
"ecosystem": "Maven",
"name": "ch.qos.logback:logback-core"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "1.2.9"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2021-42550"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": true,
"github_reviewed_at": "2021-12-17T19:25:11Z",
"nvd_published_at": "2021-12-16T19:15:00Z",
"severity": "MODERATE"
},
"details": "In logback version 1.2.7 and prior versions, an attacker with the required privileges to edit configurations files could craft a malicious configuration allowing to execute arbitrary code loaded from LDAP servers.",
"id": "GHSA-668q-qrv7-99fm",
"modified": "2021-12-17T19:25:11Z",
"published": "2021-12-17T20:00:50Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2021-42550"
},
{
"type": "WEB",
"url": "https://github.com/qos-ch/logback/commit/87291079a1de9369ac67e20dc70a8fdc7cc4359c"
},
{
"type": "WEB",
"url": "https://github.com/qos-ch/logback/commit/ef4fc4186b74b45ce80d86833820106ff27edd42"
},
{
"type": "WEB",
"url": "https://cert-portal.siemens.com/productcert/pdf/ssa-371761.pdf"
},
{
"type": "WEB",
"url": "https://github.com/cn-panda/logbackRceDemo"
},
{
"type": "PACKAGE",
"url": "https://github.com/qos-ch/logback"
},
{
"type": "WEB",
"url": "https://github.com/qos-ch/logback/blob/1502cba4c1dfd135b2e715bc0cf80c0045d4d128/logback-site/src/site/pages/news.html"
},
{
"type": "WEB",
"url": "https://jira.qos.ch/browse/LOGBACK-1591"
},
{
"type": "WEB",
"url": "https://security.netapp.com/advisory/ntap-20211229-0001"
},
{
"type": "WEB",
"url": "http://logback.qos.ch/news.html"
},
{
"type": "WEB",
"url": "http://packetstormsecurity.com/files/167794/Open-Xchange-App-Suite-7.10.x-Cross-Site-Scripting-Command-Injection.html"
},
{
"type": "WEB",
"url": "http://seclists.org/fulldisclosure/2022/Jul/11"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:H/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
],
"summary": "Deserialization of Untrusted Data in logback"
}
GHSA-66GJ-6M2V-RP63
Vulnerability from github – Published: 2026-08-18 15:31 – Updated: 2026-08-18 15:31Unauthenticated PHP Object Injection in Popup by Supsystic <= 1.13.0 versions.
{
"affected": [],
"aliases": [
"CVE-2026-73380"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-08-18T15:17:06Z",
"severity": "CRITICAL"
},
"details": "Unauthenticated PHP Object Injection in Popup by Supsystic \u003c= 1.13.0 versions.",
"id": "GHSA-66gj-6m2v-rp63",
"modified": "2026-08-18T15:31:48Z",
"published": "2026-08-18T15:31:48Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-73380"
},
{
"type": "WEB",
"url": "https://patchstack.com/database/wordpress/plugin/popup-by-supsystic/vulnerability/wordpress-popup-by-supsystic-plugin-1-13-0-php-object-injection-vulnerability?_s_id=cve"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-66J7-CGJC-VFW6
Vulnerability from github – Published: 2026-10-10 09:30 – Updated: 2026-10-10 09:30Deserialization of Untrusted Data vulnerability in ThemeREX Group Edema edema allows Object Injection.This issue affects Edema: from n/a through 1.2.2.2.
{
"affected": [],
"aliases": [
"CVE-2026-93941"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-10-10T08:17:06Z",
"severity": "CRITICAL"
},
"details": "Deserialization of Untrusted Data vulnerability in ThemeREX Group Edema edema allows Object Injection.This issue affects Edema: from n/a through 1.2.2.2.",
"id": "GHSA-66j7-cgjc-vfw6",
"modified": "2026-10-10T09:30:29Z",
"published": "2026-10-10T09:30:29Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-93941"
},
{
"type": "WEB",
"url": "https://patchstack.com/database/wordpress/theme/edema/vulnerability/wordpress-edema-theme-1-2-2-2-php-object-injection-vulnerability?_s_id=cve"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-66Q7-4WCM-7R85
Vulnerability from github – Published: 2026-02-20 18:31 – Updated: 2026-02-24 21:31Deserialization of Untrusted Data vulnerability in Dotstore Woocommerce Category Banner Management banner-management-for-woocommerce allows Object Injection.This issue affects Woocommerce Category Banner Management: from n/a through <= 2.5.1.
{
"affected": [],
"aliases": [
"CVE-2026-22354"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-02-20T16:22:34Z",
"severity": "HIGH"
},
"details": "Deserialization of Untrusted Data vulnerability in Dotstore Woocommerce Category Banner Management banner-management-for-woocommerce allows Object Injection.This issue affects Woocommerce Category Banner Management: from n/a through \u003c= 2.5.1.",
"id": "GHSA-66q7-4wcm-7r85",
"modified": "2026-02-24T21:31:38Z",
"published": "2026-02-20T18:31:38Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-22354"
},
{
"type": "WEB",
"url": "https://patchstack.com/database/Wordpress/Plugin/banner-management-for-woocommerce/vulnerability/wordpress-woocommerce-category-banner-management-plugin-2-5-1-php-object-injection-vulnerability?_s_id=cve"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-66VM-GG8J-3M49
Vulnerability from github – Published: 2025-09-03 15:30 – Updated: 2026-04-01 18:36Deserialization of Untrusted Data vulnerability in enituretechnology LTL Freight Quotes – Daylight Edition allows Object Injection. This issue affects LTL Freight Quotes – Daylight Edition: from n/a through 2.2.7.
{
"affected": [],
"aliases": [
"CVE-2025-58643"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2025-09-03T15:15:48Z",
"severity": "HIGH"
},
"details": "Deserialization of Untrusted Data vulnerability in enituretechnology LTL Freight Quotes \u2013 Daylight Edition allows Object Injection. This issue affects LTL Freight Quotes \u2013 Daylight Edition: from n/a through 2.2.7.",
"id": "GHSA-66vm-gg8j-3m49",
"modified": "2026-04-01T18:36:03Z",
"published": "2025-09-03T15:30:35Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2025-58643"
},
{
"type": "WEB",
"url": "https://patchstack.com/database/wordpress/plugin/ltl-freight-quotes-daylight-edition/vulnerability/wordpress-ltl-freight-quotes-daylight-edition-plugin-2-2-7-php-object-injection-vulnerability?_s_id=cve"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-66VP-RVJR-P222
Vulnerability from github – Published: 2025-01-25 09:30 – Updated: 2025-01-25 09:30The Custom Product Tabs Lite for WooCommerce plugin for WordPress is vulnerable to PHP Object Injection in all versions up to, and including, 1.9.0 via deserialization of untrusted input from the 'frs_woo_product_tabs' parameter. This makes it possible for authenticated attackers, with Shop Manager-level access and above, to inject a PHP Object. No known POP chain is present in the vulnerable software. If a POP chain is present via an additional plugin or theme installed on the target system, it could allow the attacker to delete arbitrary files, retrieve sensitive data, or execute code.
{
"affected": [],
"aliases": [
"CVE-2024-12600"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2025-01-25T07:15:07Z",
"severity": "HIGH"
},
"details": "The Custom Product Tabs Lite for WooCommerce plugin for WordPress is vulnerable to PHP Object Injection in all versions up to, and including, 1.9.0 via deserialization of untrusted input from the \u0027frs_woo_product_tabs\u0027 parameter. This makes it possible for authenticated attackers, with Shop Manager-level access and above, to inject a PHP Object. No known POP chain is present in the vulnerable software. If a POP chain is present via an additional plugin or theme installed on the target system, it could allow the attacker to delete arbitrary files, retrieve sensitive data, or execute code.",
"id": "GHSA-66vp-rvjr-p222",
"modified": "2025-01-25T09:30:54Z",
"published": "2025-01-25T09:30:54Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-12600"
},
{
"type": "WEB",
"url": "https://plugins.trac.wordpress.org/browser/woocommerce-custom-product-tabs-lite/trunk/woocommerce-custom-product-tabs-lite.php#L372"
},
{
"type": "WEB",
"url": "https://plugins.trac.wordpress.org/changeset/3226839"
},
{
"type": "WEB",
"url": "https://www.wordfence.com/threat-intel/vulnerabilities/id/33c16b47-3202-4f26-bf45-98172b8cac45?source=cve"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-673Q-MJM6-R5XR
Vulnerability from github – Published: 2026-10-22 16:42 – Updated: 2026-10-22 16:42Impact
A vulnerability in applications using App router with Cache Component can lead to remote code execution. The application needs to have a use cache function that returns multiple typed arrays.
Workaround
There is no known workaround for affected applications with such use cache functions. You should upgrade immediately even if your application does not have similar functions.
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 16.4.0"
},
"package": {
"ecosystem": "npm",
"name": "next"
},
"ranges": [
{
"events": [
{
"introduced": "16.0.0"
},
{
"fixed": "16.4.1"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": true,
"github_reviewed_at": "2026-10-22T16:42:31Z",
"nvd_published_at": null,
"severity": "CRITICAL"
},
"details": "## Impact\n\nA vulnerability in applications using App router with Cache Component can lead to remote code execution. The application needs to have a `use cache` function that returns multiple typed arrays.\n\n## Workaround\n\nThere is no known workaround for affected applications with such `use cache` functions. You should upgrade immediately even if your application does not have similar functions.",
"id": "GHSA-673q-mjm6-r5xr",
"modified": "2026-10-22T16:42:32Z",
"published": "2026-10-22T16:42:31Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/vercel/next.js/security/advisories/GHSA-673q-mjm6-r5xr"
},
{
"type": "WEB",
"url": "https://github.com/vercel/next.js/commit/a4cea6aec5da89d116b8960e113189cf51745d2f"
},
{
"type": "PACKAGE",
"url": "https://github.com/vercel/next.js"
},
{
"type": "WEB",
"url": "https://github.com/vercel/next.js/releases/tag/v16.4.1"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:N",
"type": "CVSS_V4"
}
],
"summary": "Next.js: Unauthenticated Remote Code Execution in `use cache` functions returning typed arrays"
}
GHSA-674M-75XM-CGV7
Vulnerability from github – Published: 2024-03-13 18:31 – Updated: 2024-03-13 18:31The Product Carousel Slider & Grid Ultimate for WooCommerce plugin for WordPress is vulnerable to PHP Object Injection in all versions up to, and including, 1.9.7 via deserialization of untrusted input via shortcode. This makes it possible for authenticated attackers, with contributor access and above, to inject a PHP Object. No POP chain is present in the vulnerable plugin. If a POP chain is present via an additional plugin or theme installed on the target system, it could allow the attacker to delete arbitrary files, retrieve sensitive data, or execute code.
{
"affected": [],
"aliases": [
"CVE-2024-1950"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2024-03-13T16:15:28Z",
"severity": "HIGH"
},
"details": "The Product Carousel Slider \u0026 Grid Ultimate for WooCommerce plugin for WordPress is vulnerable to PHP Object Injection in all versions up to, and including, 1.9.7 via deserialization of untrusted input via shortcode. This makes it possible for authenticated attackers, with contributor access and above, to inject a PHP Object. No POP chain is present in the vulnerable plugin. If a POP chain is present via an additional plugin or theme installed on the target system, it could allow the attacker to delete arbitrary files, retrieve sensitive data, or execute code.",
"id": "GHSA-674m-75xm-cgv7",
"modified": "2024-03-13T18:31:34Z",
"published": "2024-03-13T18:31:34Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-1950"
},
{
"type": "WEB",
"url": "https://plugins.trac.wordpress.org/browser/woo-product-carousel-slider-and-grid-ultimate/tags/1.9.7/includes/classes/class-meta-box.php"
},
{
"type": "WEB",
"url": "https://plugins.trac.wordpress.org/browser/woo-product-carousel-slider-and-grid-ultimate/tags/1.9.7/includes/classes/class-shortcode.php"
},
{
"type": "WEB",
"url": "https://plugins.trac.wordpress.org/changeset?old_path=/woo-product-carousel-slider-and-grid-ultimate/tags/1.9.7\u0026old=3045923\u0026new_path=/woo-product-carousel-slider-and-grid-ultimate/tags/1.9.8\u0026new=3045923\u0026sfp_email=\u0026sfph_mail="
},
{
"type": "WEB",
"url": "https://www.wordfence.com/threat-intel/vulnerabilities/id/ed8636bf-229a-42a5-a19c-332679613dd2?source=cve"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-674W-PVV4-C79W
Vulnerability from github – Published: 2026-01-23 06:31 – Updated: 2026-01-23 06:31Anritsu VectorStar CHX File Parsing Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Anritsu VectorStar. User interaction is required to exploit this vulnerability in that the target must visit a malicious page or open a malicious file.
The specific flaw exists within the parsing of CHX files. The issue results from the lack of proper validation of user-supplied data, which can result in deserialization of untrusted data. An attacker can leverage this vulnerability to execute code in the context of the current process. Was ZDI-CAN-27039.
{
"affected": [],
"aliases": [
"CVE-2025-15350"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-01-23T04:16:01Z",
"severity": "HIGH"
},
"details": "Anritsu VectorStar CHX File Parsing Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Anritsu VectorStar. User interaction is required to exploit this vulnerability in that the target must visit a malicious page or open a malicious file.\n\nThe specific flaw exists within the parsing of CHX files. The issue results from the lack of proper validation of user-supplied data, which can result in deserialization of untrusted data. An attacker can leverage this vulnerability to execute code in the context of the current process. Was ZDI-CAN-27039.",
"id": "GHSA-674w-pvv4-c79w",
"modified": "2026-01-23T06:31:23Z",
"published": "2026-01-23T06:31:23Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2025-15350"
},
{
"type": "WEB",
"url": "https://www.zerodayinitiative.com/advisories/ZDI-25-1201"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.0/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-6757-HGGM-HP2W
Vulnerability from github – Published: 2026-10-28 21:15 – Updated: 2026-10-28 21:15Affected
- Ecosystem / package: pip /
vllm - Affected versions: vLLM < 0.32.0.
Summary
When --mm-processor-cache-type shm is enabled, vLLM's multimodal SHM cache represents a cached item with an address and monotonic_id pair and treats that pair as authority to dereference shared memory in the worker. A public scale-out request can supply its own multimodal features.kwargs_data, so an authenticated caller can forge an inner SHM carrier that points at attacker-controlled bytes already staged in the ring buffer. The worker then reaches pickle.loads() while deserializing the forged object.
This is reachable through ordinary public HTTP routes and does not require VLLM_ALLOW_INSECURE_SERIALIZATION=1. The live chain used /v1/chat/completions/render to obtain a valid feature envelope, /v1/chat/completions to stage attacker-chosen bytes into SHM, and /inference/v1/generate to submit the forged inner carrier.
Affected code
Links pinned to the confirmed commit 752a3a504485 (v0.25.1):
vllm/distributed/device_communicators/shm_object_storage.py#L375-L400— SHM object deserialization reachespickle.loads(serialized_data)for byte objects.vllm/multimodal/cache.py#L567-L581— sender emits an SHM carrier containing onlyaddressandmonotonic_id.vllm/multimodal/cache.py#L711-L737— worker dereferences the untrusted pair and reads the SHM object.vllm/entrypoints/scale_out/token_in_token_out/serving.py#L146-L173— publicfeatures.kwargs_datais decoded into engine input without authenticating SHM handles before handoff.
The public scale-out route decodes attacker-controlled features.kwargs_data into MultiModalKwargsItems and hands them straight to the engine — no authentication of the SHM handle:
# vllm/entrypoints/scale_out/token_in_token_out/serving.py Lines 155-165
# Deserialize tensor data when present; None → cache hit.
mm_kwargs: dict[str, list[MultiModalKwargsItem | None]] = {}
if features.kwargs_data is not None:
for modality, items in features.kwargs_data.items():
mm_kwargs[modality] = [
decode_mm_kwargs_item(item) if item is not None else None
for item in items
]
else:
for modality, hashes in features.mm_hashes.items():
mm_kwargs[modality] = [None] * len(hashes)
In the worker, a decoded item that carries an address is treated as authority to dereference shared memory: the (address, monotonic_id) pair alone drives the read, with no binding to the cache key or to who issued the handle:
# vllm/multimodal/cache.py Lines 711-723
@override
def get_and_update_item(
self,
mm_item: MultiModalKwargsItem | None,
mm_hash: str,
) -> MultiModalKwargsItem:
assert mm_item is not None, f"Expected an address item for {mm_hash=}"
if "address" in mm_item:
address = cast(int, mm_item["address"].data)
monotonic_id = cast(int, mm_item["monotonic_id"].data)
return self._shm_cache.get(address, monotonic_id)
return mm_item
The dereference lands in the SHM deserializer, which pickle.loads() any chunk tagged as bytes — so attacker bytes staged in the ring buffer are unpickled in the worker process:
# vllm/distributed/device_communicators/shm_object_storage.py Lines 375-400
def deserialize(self, data_view: memoryview) -> Any:
# pickle.loads do not read past the end of a pickled object
# within a large buffer, so we can skip storing the metadata size
type_name, nbytes, len_arr = pickle.loads(data_view)
serialized_data = data_view[-nbytes:]
if type_name == torch.Tensor.__name__:
obj = []
start_idx = 0
for length in len_arr:
item_bytes = serialized_data[start_idx : start_idx + length]
obj.append(item_bytes)
start_idx += length
obj = self.tensor_decoder.decode(obj)
elif type_name == self._mm_kwargs_item_cls.__name__:
obj = []
start_idx = 0
for length in len_arr:
item_bytes = serialized_data[start_idx : start_idx + length]
obj.append(item_bytes)
start_idx += length
obj = self.mm_decoder.decode(obj)
elif type_name == bytes.__name__:
obj = pickle.loads(serialized_data)
else:
raise ValueError(f"Unsupported object type '{type_name}' in metadata")
return obj
Impact
An ordinary authenticated inference caller can execute arbitrary Python in the worker process on a deployment that uses SHM multimodal caching. The live proof used a benign marker gadget and also caused the EngineCore health endpoint to fall from 200 to 503 after the forged request. The exploit is reachable with VLLM_ALLOW_INSECURE_SERIALIZATION=0; it abuses the built-in SHM object deserializer, not the separate optional msgpack ext-hook flag.
Suggested Fix
Authenticate every SHM carrier with a server-issued capability bound to the cache key, address, and monotonic id; rotate the capability secret when the cache is cleared; and reject malformed or forged SHM carriers at the public scale-out boundary before engine handoff — so a forged pair never reaches the worker deref / pickle.loads().
The sender emits an HMAC capability (keyed by a 32-byte secret stored in the ring buffer and rotated on clear()) over (cache_key, address, monotonic_id) alongside the pair. The public scale-out path validates it before engine handoff:
# vllm/entrypoints/scale_out/token_in_token_out/serving.py
cache = self.online_renderer.renderer.mm_processor_cache
for mm_hash, item in zip(hashes, items):
decoded_item = decode_mm_kwargs_item(item) if item is not None else None
if decoded_item is not None and cache is not None:
# Rejects a forged (address, monotonic_id) that lacks a valid
# server-issued capability for this cache key.
cache.validate_input_item(decoded_item, mm_hash)
decoded_items.append(decoded_item)
The worker get() re-checks the capability on the reader side before it ever dereferences the caller-supplied address:
# vllm/distributed/device_communicators/shm_object_storage.py
def get(self, address, monotonic_id, capability=None, key=None):
if not self.is_writer:
# Reject forged handles before dereferencing the supplied address.
self.validate_capability(key, address, monotonic_id, capability)
with self.ring_buffer.access_buf(address) as (data_view, buf_metadata):
self._validate_monotonic_id(address, monotonic_id, buf_metadata)
obj = self.ser_de.deserialize(data_view[self.flag_bytes :])
...
where validate_capability recomputes the HMAC with hmac.compare_digest and raises Missing server-issued SHM capability for cache key. / Invalid server-issued SHM capability for cache key. on any mismatch. The fix applies cleanly to v0.25.1 and ships regression tests asserting the generate path rejects an invalid handle and that the receiver requires a matching sender handle (and a capability issued after the current clear()) for the cache key.
Credit
Reported by: Patch the Planet (Trail of Bits + OpenAI collaboration)
This vulnerability was discovered using GPT-5.5-Cyber as part of the Patch the Planet security initiative.
Proposed fix: a proposed change for this issue is up for review in this advisory's temporary private fork: https://github.com/vllm-project/vllm-ghsa-6757-hggm-hp2w/pull/1
This advisory was fixed in MR59357.
{
"affected": [
{
"package": {
"ecosystem": "PyPI",
"name": "vllm"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.32.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-115912"
],
"database_specific": {
"cwe_ids": [
"CWE-502"
],
"github_reviewed": true,
"github_reviewed_at": "2026-10-28T21:15:15Z",
"nvd_published_at": "2026-10-26T17:17:18Z",
"severity": "HIGH"
},
"details": "## Affected\n\n- **Ecosystem / package:** pip / `vllm`\n- **Affected versions:** vLLM \u003c 0.32.0.\n\n## Summary\n\nWhen `--mm-processor-cache-type shm` is enabled, vLLM\u0027s multimodal SHM cache represents a cached item with an `address` and `monotonic_id` pair and treats that pair as authority to dereference shared memory in the worker. A public scale-out request can supply its own multimodal `features.kwargs_data`, so an authenticated caller can forge an inner SHM carrier that points at attacker-controlled bytes already staged in the ring buffer. The worker then reaches `pickle.loads()` while deserializing the forged object.\n\nThis is reachable through ordinary public HTTP routes and does not require `VLLM_ALLOW_INSECURE_SERIALIZATION=1`. The live chain used `/v1/chat/completions/render` to obtain a valid feature envelope, `/v1/chat/completions` to stage attacker-chosen bytes into SHM, and `/inference/v1/generate` to submit the forged inner carrier.\n\n## Affected code\n\nLinks pinned to the confirmed commit [`752a3a504485`](https://github.com/vllm-project/vllm/tree/752a3a504485790a2e8491cacbb35c137339ad34) (v0.25.1):\n\n- [`vllm/distributed/device_communicators/shm_object_storage.py#L375-L400`](https://github.com/vllm-project/vllm/blob/752a3a504485790a2e8491cacbb35c137339ad34/vllm/distributed/device_communicators/shm_object_storage.py#L375-L400) \u2014 SHM object deserialization reaches `pickle.loads(serialized_data)` for byte objects.\n- [`vllm/multimodal/cache.py#L567-L581`](https://github.com/vllm-project/vllm/blob/752a3a504485790a2e8491cacbb35c137339ad34/vllm/multimodal/cache.py#L567-L581) \u2014 sender emits an SHM carrier containing only `address` and `monotonic_id`.\n- [`vllm/multimodal/cache.py#L711-L737`](https://github.com/vllm-project/vllm/blob/752a3a504485790a2e8491cacbb35c137339ad34/vllm/multimodal/cache.py#L711-L737) \u2014 worker dereferences the untrusted pair and reads the SHM object.\n- [`vllm/entrypoints/scale_out/token_in_token_out/serving.py#L146-L173`](https://github.com/vllm-project/vllm/blob/752a3a504485790a2e8491cacbb35c137339ad34/vllm/entrypoints/scale_out/token_in_token_out/serving.py#L146-L173) \u2014 public `features.kwargs_data` is decoded into engine input without authenticating SHM handles before handoff.\n\nThe public scale-out route decodes attacker-controlled `features.kwargs_data` into `MultiModalKwargsItem`s and hands them straight to the engine \u2014 no authentication of the SHM handle:\n\n```python\n# vllm/entrypoints/scale_out/token_in_token_out/serving.py Lines 155-165\n # Deserialize tensor data when present; None \u2192 cache hit.\n mm_kwargs: dict[str, list[MultiModalKwargsItem | None]] = {}\n if features.kwargs_data is not None:\n for modality, items in features.kwargs_data.items():\n mm_kwargs[modality] = [\n decode_mm_kwargs_item(item) if item is not None else None\n for item in items\n ]\n else:\n for modality, hashes in features.mm_hashes.items():\n mm_kwargs[modality] = [None] * len(hashes)\n```\n\nIn the worker, a decoded item that carries an `address` is treated as authority to dereference shared memory: the `(address, monotonic_id)` pair alone drives the read, with no binding to the cache key or to who issued the handle:\n\n```python\n# vllm/multimodal/cache.py Lines 711-723\n @override\n def get_and_update_item(\n self,\n mm_item: MultiModalKwargsItem | None,\n mm_hash: str,\n ) -\u003e MultiModalKwargsItem:\n assert mm_item is not None, f\"Expected an address item for {mm_hash=}\"\n if \"address\" in mm_item:\n address = cast(int, mm_item[\"address\"].data)\n monotonic_id = cast(int, mm_item[\"monotonic_id\"].data)\n return self._shm_cache.get(address, monotonic_id)\n\n return mm_item\n```\n\nThe dereference lands in the SHM deserializer, which `pickle.loads()` any chunk tagged as `bytes` \u2014 so attacker bytes staged in the ring buffer are unpickled in the worker process:\n\n```python\n# vllm/distributed/device_communicators/shm_object_storage.py Lines 375-400\n def deserialize(self, data_view: memoryview) -\u003e Any:\n # pickle.loads do not read past the end of a pickled object\n # within a large buffer, so we can skip storing the metadata size\n type_name, nbytes, len_arr = pickle.loads(data_view)\n serialized_data = data_view[-nbytes:]\n\n if type_name == torch.Tensor.__name__:\n obj = []\n start_idx = 0\n for length in len_arr:\n item_bytes = serialized_data[start_idx : start_idx + length]\n obj.append(item_bytes)\n start_idx += length\n obj = self.tensor_decoder.decode(obj)\n elif type_name == self._mm_kwargs_item_cls.__name__:\n obj = []\n start_idx = 0\n for length in len_arr:\n item_bytes = serialized_data[start_idx : start_idx + length]\n obj.append(item_bytes)\n start_idx += length\n obj = self.mm_decoder.decode(obj)\n elif type_name == bytes.__name__:\n obj = pickle.loads(serialized_data)\n else:\n raise ValueError(f\"Unsupported object type \u0027{type_name}\u0027 in metadata\")\n\n return obj\n```\n\n## Impact\n\nAn ordinary authenticated inference caller can execute arbitrary Python in the worker process on a deployment that uses SHM multimodal caching. The live proof used a benign marker gadget and also caused the EngineCore health endpoint to fall from 200 to 503 after the forged request. The exploit is reachable with `VLLM_ALLOW_INSECURE_SERIALIZATION=0`; it abuses the built-in SHM object deserializer, not the separate optional msgpack ext-hook flag.\n\n\n## Suggested Fix\n\nAuthenticate every SHM carrier with a server-issued capability bound to the cache key, address, and monotonic id; rotate the capability secret when the cache is cleared; and reject malformed or forged SHM carriers at the public scale-out boundary before engine handoff \u2014 so a forged pair never reaches the worker deref / `pickle.loads()`.\n\nThe sender emits an HMAC capability (keyed by a 32-byte secret stored in the ring buffer and rotated on `clear()`) over `(cache_key, address, monotonic_id)` alongside the pair. The public scale-out path validates it before engine handoff:\n\n```python\n# vllm/entrypoints/scale_out/token_in_token_out/serving.py\n cache = self.online_renderer.renderer.mm_processor_cache\n for mm_hash, item in zip(hashes, items):\n decoded_item = decode_mm_kwargs_item(item) if item is not None else None\n if decoded_item is not None and cache is not None:\n # Rejects a forged (address, monotonic_id) that lacks a valid\n # server-issued capability for this cache key.\n cache.validate_input_item(decoded_item, mm_hash)\n decoded_items.append(decoded_item)\n```\n\nThe worker `get()` re-checks the capability on the reader side before it ever dereferences the caller-supplied address:\n\n```python\n# vllm/distributed/device_communicators/shm_object_storage.py\n def get(self, address, monotonic_id, capability=None, key=None):\n if not self.is_writer:\n # Reject forged handles before dereferencing the supplied address.\n self.validate_capability(key, address, monotonic_id, capability)\n with self.ring_buffer.access_buf(address) as (data_view, buf_metadata):\n self._validate_monotonic_id(address, monotonic_id, buf_metadata)\n obj = self.ser_de.deserialize(data_view[self.flag_bytes :])\n ...\n```\n\nwhere `validate_capability` recomputes the HMAC with `hmac.compare_digest` and raises `Missing server-issued SHM capability for cache key.` / `Invalid server-issued SHM capability for cache key.` on any mismatch. The fix applies cleanly to `v0.25.1` and ships regression tests asserting the generate path rejects an invalid handle and that the receiver requires a matching sender handle (and a capability issued after the current `clear()`) for the cache key.\n\n## Credit\n\n**Reported by:** Patch the Planet (Trail of Bits + OpenAI collaboration)\n\nThis vulnerability was discovered using GPT-5.5-Cyber as part of the Patch the Planet security initiative.\n\n---\n\n**Proposed fix:** a proposed change for this issue is up for review in this advisory\u0027s temporary private fork: https://github.com/vllm-project/vllm-ghsa-6757-hggm-hp2w/pull/1\n\nThis advisory was fixed in [MR59357](https://github.com/vllm-project/vllm/pull/59357).",
"id": "GHSA-6757-hggm-hp2w",
"modified": "2026-10-28T21:15:15Z",
"published": "2026-10-28T21:15:15Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-6757-hggm-hp2w"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-115912"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/pull/59357"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/commit/3d5625cc3c2474994107047d7a6ecb8e4008f9b0"
},
{
"type": "PACKAGE",
"url": "https://github.com/vllm-project/vllm"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/releases/tag/v0.32.0"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
],
"summary": "vLLM: Public multimodal SHM handle forgery reaches worker-side `pickle.loads()` with insecure serialization disabled"
}
Mitigation
If available, use the signing/sealing features of the programming language to assure that deserialized data has not been tainted. For example, a hash-based message authentication code (HMAC) could be used to ensure that data has not been modified.
Mitigation
When deserializing data, populate a new object rather than just deserializing. The result is that the data flows through safe input validation and that the functions are safe.
Mitigation
Explicitly define a final object() to prevent deserialization.
Mitigation
- Make fields transient to protect them from deserialization.
- An attempt to serialize and then deserialize a class containing transient fields will result in NULLs where the transient data should be. This is an excellent way to prevent time, environment-based, or sensitive variables from being carried over and used improperly.
Mitigation
Avoid having unnecessary types or gadgets (a sequence of instances and method invocations that can self-execute during the deserialization process, often found in libraries) available that can be leveraged for malicious ends. This limits the potential for unintended or unauthorized types and gadgets to be leveraged by the attacker. Add only acceptable classes to an allowlist. Note: new gadgets are constantly being discovered, so this alone is not a sufficient mitigation.
Mitigation
Employ cryptography of the data or code for protection. However, it's important to note that it would still be client-side security. This is risky because if the client is compromised then the security implemented on the client (the cryptography) can be bypassed.
Mitigation MIT-29
Strategy: Firewall
Use an application firewall that can detect attacks against this weakness. It can be beneficial in cases in which the code cannot be fixed (because it is controlled by a third party), as an emergency prevention measure while more comprehensive software assurance measures are applied, or to provide defense in depth [REF-1481].
CAPEC-586: Object Injection
An adversary attempts to exploit an application by injecting additional, malicious content during its processing of serialized objects. Developers leverage serialization in order to convert data or state into a static, binary format for saving to disk or transferring over a network. These objects are then deserialized when needed to recover the data/state. By injecting a malformed object into a vulnerable application, an adversary can potentially compromise the application by manipulating the deserialization process. This can result in a number of unwanted outcomes, including remote code execution.