GCVE Workshop - 22 September 2026 (14:00-18:00), Luxembourg Before The Vulnopticon Conference - Registration
Common Weakness Enumeration

CWE-674

Allowed-with-Review

Uncontrolled Recursion

Abstraction: Class · Status: Draft

API response cwe.mitre.org

The product does not properly control the amount of recursion that takes place, consuming excessive resources, such as allocated memory or the program stack.

751 vulnerabilities reference this CWE, most recent first.

GHSA-CWV3-863G-39VX

Vulnerability from github – Published: 2021-05-21 14:26 – Updated: 2024-10-31 21:22
VLAI
Summary
Stack overflow due to looping TFLite subgraph
Details

Impact

TFlite graphs must not have loops between nodes. However, this condition was not checked and an attacker could craft models that would result in infinite loop during evaluation. In certain cases, the infinite loop would be replaced by stack overflow due to too many recursive calls.

For example, the While implementation could be tricked into a scneario where both the body and the loop subgraphs are the same. Evaluating one of the subgraphs means calling the Eval function for the other and this quickly exhaust all stack space.

Patches

We have patched the issue in GitHub commit 9c1dc920d8ffb4893d6c9d27d1f039607b326743 (for the While operator) and in GitHub commit c6173f5fe66cdbab74f4f869311fe6aae2ba35f4 (in general).

The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

For more information

Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by members of the Aivul Team from Qihoo 360.

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

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.1.4"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.2.0"
            },
            {
              "fixed": "2.2.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.3.0"
            },
            {
              "fixed": "2.3.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.4.0"
            },
            {
              "fixed": "2.4.2"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-cpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.1.4"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-cpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.2.0"
            },
            {
              "fixed": "2.2.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-cpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.3.0"
            },
            {
              "fixed": "2.3.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-cpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.4.0"
            },
            {
              "fixed": "2.4.2"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-gpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.1.4"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-gpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.2.0"
            },
            {
              "fixed": "2.2.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-gpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.3.0"
            },
            {
              "fixed": "2.3.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "tensorflow-gpu"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.4.0"
            },
            {
              "fixed": "2.4.2"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2021-29591"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-674",
      "CWE-835"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2021-05-18T16:46:07Z",
    "nvd_published_at": "2021-05-14T20:15:00Z",
    "severity": "HIGH"
  },
  "details": "### Impact\nTFlite graphs must not have loops between nodes. However, this condition was not checked and an attacker could craft models that would result in infinite loop during evaluation. In certain cases, the infinite loop would be replaced by stack overflow due to too many recursive calls.\n\nFor example, the [`While` implementation](https://github.com/tensorflow/tensorflow/blob/106d8f4fb89335a2c52d7c895b7a7485465ca8d9/tensorflow/lite/kernels/while.cc) could be tricked into a scneario where both the body and the loop subgraphs are the same. Evaluating one of the subgraphs means calling the `Eval` function for the other and this quickly exhaust all stack space.\n    \n### Patches \nWe have patched the issue in GitHub commit [9c1dc920d8ffb4893d6c9d27d1f039607b326743](https://github.com/tensorflow/tensorflow/commit/9c1dc920d8ffb4893d6c9d27d1f039607b326743) (for the `While` operator) and in GitHub commit [c6173f5fe66cdbab74f4f869311fe6aae2ba35f4](https://github.com/tensorflow/tensorflow/commit/c6173f5fe66cdbab74f4f869311fe6aae2ba35f4) (in general).\n    \nThe fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.\n\n### For more information\nPlease consult [our security guide](https://github.com/tensorflow/tensorflow/blob/master/SECURITY.md) for more information regarding the security model and how to contact us with issues and questions.\n\n### Attribution \nThis vulnerability has been reported by members of the Aivul Team from Qihoo 360.",
  "id": "GHSA-cwv3-863g-39vx",
  "modified": "2024-10-31T21:22:52Z",
  "published": "2021-05-21T14:26:56Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/tensorflow/tensorflow/security/advisories/GHSA-cwv3-863g-39vx"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2021-29591"
    },
    {
      "type": "WEB",
      "url": "https://github.com/tensorflow/tensorflow/commit/9c1dc920d8ffb4893d6c9d27d1f039607b326743"
    },
    {
      "type": "WEB",
      "url": "https://github.com/tensorflow/tensorflow/commit/c6173f5fe66cdbab74f4f869311fe6aae2ba35f4"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2021-519.yaml"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2021-717.yaml"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow/PYSEC-2021-228.yaml"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/tensorflow/tensorflow"
    },
    {
      "type": "WEB",
      "url": "https://github.com/tensorflow/tensorflow/blob/106d8f4fb89335a2c52d7c895b7a7485465ca8d9/tensorflow/lite/kernels/while.cc"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:H",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:L/AC:L/AT:N/PR:L/UI:N/VC:H/VI:L/VA:H/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Stack overflow due to looping TFLite subgraph"
}

GHSA-F2FG-5M3G-HQWV

Vulnerability from github – Published: 2026-02-18 18:30 – Updated: 2026-02-18 18:30
VLAI
Details

mayswind ezbookkeeping versions 1.2.0 and earlier contain a critical vulnerability in JSON and XML file import processing. The application fails to validate nesting depth during parsing operations, allowing authenticated attackers to trigger denial of service conditions by uploading deeply nested malicious files. This results in CPU exhaustion, service degradation, or complete service unavailability.

Severity
6.5 (Medium)
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2025-65519"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-674"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-02-18T16:22:28Z",
    "severity": "MODERATE"
  },
  "details": "mayswind ezbookkeeping versions 1.2.0 and earlier contain a critical vulnerability in JSON and XML file import processing. The application fails to validate nesting depth during parsing operations, allowing authenticated attackers to trigger denial of service conditions by uploading deeply nested malicious files. This results in CPU exhaustion, service degradation, or complete service unavailability.",
  "id": "GHSA-f2fg-5m3g-hqwv",
  "modified": "2026-02-18T18:30:39Z",
  "published": "2026-02-18T18:30:39Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2025-65519"
    },
    {
      "type": "WEB",
      "url": "https://github.com/ictrun/EBK-SA-2025-001"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-F38Q-MGVJ-VPH7

Vulnerability from github – Published: 2026-06-15 17:27 – Updated: 2026-07-15 22:06
VLAI
Summary
protobufjs : Schema-derived names can shadow runtime-significant properties
Details

Summary

protobufjs accepted certain schema-derived names that could collide with properties used by protobufjs runtime helpers. The known affected names are fields named hasOwnProperty, field or oneof names such as $type when loaded through protobufjs JSON/reflection descriptors, and service methods whose generated helper name is rpcCall.

When affected message or service types were used, protobufjs could read schema-controlled data where it expected an own-property helper, reflected type metadata, or the base RPC helper. This could cause deterministic exceptions or recursive calls in affected decode post-checks, verification, object conversion, reflected JSON serialization, or protobufjs RPC helper invocation.

Impact

An attacker who can provide or influence protobuf schemas or protobufjs JSON descriptors may be able to make affected message or service types unusable, resulting in denial of service for the affected processing path.

Applications using only trusted schemas are affected only if those schemas contain one of the problematic names and the application reaches the affected API path.

The issue is not known to allow code execution by itself.

Preconditions

  • The application must use an affected protobufjs version.
  • The application must load or use a schema or protobufjs JSON descriptor containing one of the problematic names:
  • a field named hasOwnProperty,
  • a field or oneof named $type through protobufjs JSON/reflection descriptor input,
  • or a service method whose generated helper name is rpcCall.
  • The application must reach the affected API path for that name: required-field decode post-checks, verify, or toObject for hasOwnProperty; reflected message JSON serialization for $type; or protobufjs RPC service invocation for rpcCall.

Workarounds

Do not load protobuf schemas or protobufjs JSON descriptors from untrusted sources with affected versions. If untrusted schemas or descriptors must be accepted, validate schema-derived field, oneof, and service method names before loading and reject the problematic names described above.

Applications using trusted schemas can avoid the issue by renaming affected fields or service methods, or by avoiding the affected API path.

Severity
5.3 (Medium)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L
Show details on source website
To clipboard 

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 7.6.2"
      },
      "package": {
        "ecosystem": "npm",
        "name": "protobufjs"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "7.6.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 2.5.0"
      },
      "package": {
        "ecosystem": "npm",
        "name": "protobufjs-cli"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.0.0"
            },
            {
              "fixed": "2.5.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 1.3.2"
      },
      "package": {
        "ecosystem": "npm",
        "name": "protobufjs-cli"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.3.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 8.5.0"
      },
      "package": {
        "ecosystem": "npm",
        "name": "protobufjs"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "8.0.0"
            },
            {
              "fixed": "8.6.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-54269"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-674",
      "CWE-754"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-06-15T17:27:18Z",
    "nvd_published_at": "2026-06-22T18:16:45Z",
    "severity": "MODERATE"
  },
  "details": "## Summary\n\nprotobufjs accepted certain schema-derived names that could collide with properties used by protobufjs runtime helpers. The known affected names are fields named `hasOwnProperty`, field or oneof names such as `$type` when loaded through protobufjs JSON/reflection descriptors, and service methods whose generated helper name is `rpcCall`.\n\nWhen affected message or service types were used, protobufjs could read schema-controlled data where it expected an own-property helper, reflected type metadata, or the base RPC helper. This could cause deterministic exceptions or recursive calls in affected decode post-checks, verification, object conversion, reflected JSON serialization, or protobufjs RPC helper invocation.\n\n## Impact\n\nAn attacker who can provide or influence protobuf schemas or protobufjs JSON descriptors may be able to make affected message or service types unusable, resulting in denial of service for the affected processing path.\n\nApplications using only trusted schemas are affected only if those schemas contain one of the problematic names and the application reaches the affected API path.\n\nThe issue is not known to allow code execution by itself.\n\n## Preconditions\n\n* The application must use an affected protobufjs version.\n* The application must load or use a schema or protobufjs JSON descriptor containing one of the problematic names:\n  * a field named `hasOwnProperty`,\n  * a field or oneof named `$type` through protobufjs JSON/reflection descriptor input,\n  * or a service method whose generated helper name is `rpcCall`.\n* The application must reach the affected API path for that name: required-field decode post-checks, `verify`, or `toObject` for `hasOwnProperty`; reflected message JSON serialization for `$type`; or protobufjs RPC service invocation for `rpcCall`.\n\n## Workarounds\n\nDo not load protobuf schemas or protobufjs JSON descriptors from untrusted sources with affected versions. If untrusted schemas or descriptors must be accepted, validate schema-derived field, oneof, and service method names before loading and reject the problematic names described above.\n\nApplications using trusted schemas can avoid the issue by renaming affected fields or service methods, or by avoiding the affected API path.",
  "id": "GHSA-f38q-mgvj-vph7",
  "modified": "2026-07-15T22:06:05Z",
  "published": "2026-06-15T17:27:18Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/protobufjs/protobuf.js/security/advisories/GHSA-f38q-mgvj-vph7"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-54269"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/protobufjs/protobuf.js"
    }
  ],
  "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:L",
      "type": "CVSS_V3"
    }
  ],
  "summary": "protobufjs : Schema-derived names can shadow runtime-significant properties"
}

GHSA-F3PG-HPPR-JF6F

Vulnerability from github – Published: 2024-11-21 21:33 – Updated: 2024-12-24 15:30
VLAI
Details

In the Linux kernel, the following vulnerability has been resolved:

afs: Fix lock recursion

afs_wake_up_async_call() can incur lock recursion. The problem is that it is called from AF_RXRPC whilst holding the ->notify_lock, but it tries to take a ref on the afs_call struct in order to pass it to a work queue - but if the afs_call is already queued, we then have an extraneous ref that must be put... calling afs_put_call() may call back down into AF_RXRPC through rxrpc_kernel_shutdown_call(), however, which might try taking the ->notify_lock again.

This case isn't very common, however, so defer it to a workqueue. The oops looks something like:

BUG: spinlock recursion on CPU#0, krxrpcio/7001/1646 lock: 0xffff888141399b30, .magic: dead4ead, .owner: krxrpcio/7001/1646, .owner_cpu: 0 CPU: 0 UID: 0 PID: 1646 Comm: krxrpcio/7001 Not tainted 6.12.0-rc2-build3+ #4351 Hardware name: ASUS All Series/H97-PLUS, BIOS 2306 10/09/2014 Call Trace: dump_stack_lvl+0x47/0x70 do_raw_spin_lock+0x3c/0x90 rxrpc_kernel_shutdown_call+0x83/0xb0 afs_put_call+0xd7/0x180 rxrpc_notify_socket+0xa0/0x190 rxrpc_input_split_jumbo+0x198/0x1d0 rxrpc_input_data+0x14b/0x1e0 ? rxrpc_input_call_packet+0xc2/0x1f0 rxrpc_input_call_event+0xad/0x6b0 rxrpc_input_packet_on_conn+0x1e1/0x210 rxrpc_input_packet+0x3f2/0x4d0 rxrpc_io_thread+0x243/0x410 ? __pfx_rxrpc_io_thread+0x10/0x10 kthread+0xcf/0xe0 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x24/0x40 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30

Severity
5.5 (Medium)
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2024-53090"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-674"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2024-11-21T19:15:12Z",
    "severity": "MODERATE"
  },
  "details": "In the Linux kernel, the following vulnerability has been resolved:\n\nafs: Fix lock recursion\n\nafs_wake_up_async_call() can incur lock recursion.  The problem is that it\nis called from AF_RXRPC whilst holding the -\u003enotify_lock, but it tries to\ntake a ref on the afs_call struct in order to pass it to a work queue - but\nif the afs_call is already queued, we then have an extraneous ref that must\nbe put... calling afs_put_call() may call back down into AF_RXRPC through\nrxrpc_kernel_shutdown_call(), however, which might try taking the\n-\u003enotify_lock again.\n\nThis case isn\u0027t very common, however, so defer it to a workqueue.  The oops\nlooks something like:\n\n  BUG: spinlock recursion on CPU#0, krxrpcio/7001/1646\n   lock: 0xffff888141399b30, .magic: dead4ead, .owner: krxrpcio/7001/1646, .owner_cpu: 0\n  CPU: 0 UID: 0 PID: 1646 Comm: krxrpcio/7001 Not tainted 6.12.0-rc2-build3+ #4351\n  Hardware name: ASUS All Series/H97-PLUS, BIOS 2306 10/09/2014\n  Call Trace:\n   \u003cTASK\u003e\n   dump_stack_lvl+0x47/0x70\n   do_raw_spin_lock+0x3c/0x90\n   rxrpc_kernel_shutdown_call+0x83/0xb0\n   afs_put_call+0xd7/0x180\n   rxrpc_notify_socket+0xa0/0x190\n   rxrpc_input_split_jumbo+0x198/0x1d0\n   rxrpc_input_data+0x14b/0x1e0\n   ? rxrpc_input_call_packet+0xc2/0x1f0\n   rxrpc_input_call_event+0xad/0x6b0\n   rxrpc_input_packet_on_conn+0x1e1/0x210\n   rxrpc_input_packet+0x3f2/0x4d0\n   rxrpc_io_thread+0x243/0x410\n   ? __pfx_rxrpc_io_thread+0x10/0x10\n   kthread+0xcf/0xe0\n   ? __pfx_kthread+0x10/0x10\n   ret_from_fork+0x24/0x40\n   ? __pfx_kthread+0x10/0x10\n   ret_from_fork_asm+0x1a/0x30\n   \u003c/TASK\u003e",
  "id": "GHSA-f3pg-hppr-jf6f",
  "modified": "2024-12-24T15:30:31Z",
  "published": "2024-11-21T21:33:32Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2024-53090"
    },
    {
      "type": "WEB",
      "url": "https://git.kernel.org/stable/c/610a79ffea02102899a1373fe226d949944a7ed6"
    },
    {
      "type": "WEB",
      "url": "https://git.kernel.org/stable/c/d7cbf81df996b1eae2dee8deb6df08e2eba78661"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-F4QM-VJ5J-9XPW

Vulnerability from github – Published: 2026-04-14 18:48 – Updated: 2026-04-14 18:48
VLAI
Summary
ImageMagick has a Stack Overflow via Recursive FX Expression Parsing
Details

A stack overflow vulnerability in ImageMagick's FX expression parser allows an attacker to crash the process by providing a deeply nested expression.

Severity
5.5 (Medium)
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-AnyCPU"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-HDRI-AnyCPU"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-HDRI-OpenMP-arm64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-HDRI-arm64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-HDRI-x64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-HDRI-x86"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-OpenMP-arm64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-OpenMP-x64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-arm64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-x64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q16-x86"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q8-AnyCPU"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q8-OpenMP-arm64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q8-OpenMP-x64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q8-arm64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q8-x64"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Magick.NET-Q8-x86"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "14.12.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-33902"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-674"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-04-14T18:48:06Z",
    "nvd_published_at": "2026-04-13T22:16:28Z",
    "severity": "MODERATE"
  },
  "details": "A stack overflow vulnerability in ImageMagick\u0027s FX expression parser allows an attacker to crash the process by providing a deeply nested expression.",
  "id": "GHSA-f4qm-vj5j-9xpw",
  "modified": "2026-04-14T18:48:07Z",
  "published": "2026-04-14T18:48:06Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/ImageMagick/ImageMagick/security/advisories/GHSA-f4qm-vj5j-9xpw"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-33902"
    },
    {
      "type": "WEB",
      "url": "https://github.com/ImageMagick/ImageMagick/commit/d3c0a37485314c5ccef72efb18f3847cd53868ba"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/ImageMagick/ImageMagick"
    },
    {
      "type": "WEB",
      "url": "https://github.com/ImageMagick/ImageMagick/releases/tag/7.1.2-19"
    },
    {
      "type": "WEB",
      "url": "https://github.com/dlemstra/Magick.NET/releases/tag/14.12.0"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": " ImageMagick has a Stack Overflow via Recursive FX Expression Parsing"
}

GHSA-F5V8-V6Q3-Q4H6

Vulnerability from github – Published: 2026-04-16 22:50 – Updated: 2026-04-16 22:50
VLAI
Summary
Meridian: Multiple defense-in-depth gaps (collection/depth caps, telemetry, retry, fan-out)
Details

Summary

Meridian v2.1.0 (Meridian.Mapping and Meridian.Mediator) shipped with nine defense-in-depth gaps reachable through its public APIs. Two are HIGH severity — the advertised DefaultMaxCollectionItems and DefaultMaxDepth safety caps are silently bypassed on the IMapper.Map(source, destination) overload and anywhere .UseDestinationValue() is configured on a collection-typed property. Four are MEDIUM (constructor invariant bypass, OpenTelemetry stack-trace info disclosure, retry amplification, notification fan-out amplification). Three are LOW (exception message disclosure, dictionary duplicate-key echo, static mediator cache growth under closed-generic types).

All nine are patched in v2.1.1. Upgrade is a drop-in NuGet bump; see the v2.1.1 CHANGELOG for the four behavioural changes (constructor selection, OTel default, publisher fan-out cap, retry caps).

Severity Matrix

# Severity CWE Finding Fix
1 HIGH CWE-770 MappingEngine.TryMapCollectionOntoExisting enumerated the source without enforcing DefaultMaxCollectionItems. Reachable via Mapper.Map<TSrc,TDst>(src, dst) and any .ForMember(..., o => o.UseDestinationValue()) on a collection member through a plain Map(src) call. Shared cap enforcement helper between MapCollection and TryMapCollectionOntoExisting.
2 HIGH CWE-674 Collection-item recursion in the existing-destination path did not increment ResolutionContext.Depth, so self-referential collection graphs could reach stack overflow before DefaultMaxDepth fired. Depth increments at every collection-item boundary.
3 MEDIUM CWE-665 ObjectCreator.CreateWithConstructorMapping always invoked the widest public constructor, silently filling unresolved parameters with default(T) and bypassing narrower-ctor invariants. Widest-ctor selection now requires every parameter to be bound via explicit ctor mapping, source-name match, or a C# optional default.
4 MEDIUM CWE-532 Mediator.MarkActivityFailure emitted the full ex.ToString() (stack + inner chain) to the OpenTelemetry exception.stacktrace activity tag by default, leaking context to any shared trace sink. Gated on MediatorTelemetryOptions.RecordExceptionStackTrace — opt-in, default false.
5 MEDIUM CWE-400 RetryBehavior retried every exception type with unbounded MaxRetries; the exponential-backoff delay overflowed TimeSpan at ~30 attempts. No cancellation exclusion. Server-side MaxRetriesCap = 10, MaxBackoff = 5 min, OperationCanceledException short-circuit, recommended RetryPolicy.TransientOnly helper.
6 MEDIUM CWE-400 TaskWhenAllPublisher started every registered handler concurrently with no bound on fan-out. New constructor parameter maxDegreeOfParallelism (default 16; -1 restores legacy unbounded).
7 LOW CWE-209 Public mapping exceptions leaked FullName of source/destination types and concatenated inner exception messages into top-level property-mapping errors. Scrubbed to type Name; inner details only via InnerException chain.
8 LOW CWE-209 Dictionary materialization threw ArgumentException on duplicate keys, echoing the attacker-supplied key's .ToString(). Last-write-wins indexer semantics.
9 LOW CWE-1325 Static mediator handler caches grow monotonically under closed-generic request types. Doc-only mitigation; no code change — consumers must not allow attacker-controlled runtime type materialization to reach Send, Publish, or CreateStream. Documented in docs/security-model.md.

Exploitation

Finding 1 / 2 (headline): A consumer that maps user-supplied collection payloads onto an existing destination list via mapper.Map(userCollection, existingList) — a documented and commonly used AutoMapper-style idiom — processes the full attacker-supplied collection with no size cap and no depth cap. An attacker sending a single request with a large (or self-referential) collection payload can block the worker thread for seconds and exhaust the managed heap or the call stack. Equivalent exposure through .UseDestinationValue() on a collection-typed destination member, reachable via a plain Map(src) call whose destination type default-initializes that member.

Finding 3: A destination type with multiple public constructors that differ only in their parameter-binding invariants (e.g., new UserAccount(string name, Email email) enforcing a non-default Email) could be instantiated with the narrower ctor's invariants silently bypassed if any source field was absent — the widest ctor was always picked, with unbound parameters replaced by default(T).

Findings 4 / 5 / 6: Amplification / information-disclosure vectors described in the matrix above. Each requires moderate integration context (telemetry sink trust, handler count, retry policy) to weaponize, but each is reachable through public APIs without authentication.

Patches

  • Meridian.Mapping 2.1.1 (published 2026-04-16)
  • Meridian.Mediator 2.1.1 (published 2026-04-16)

Verified via: - GitHub Release assets at https://github.com/UmutKorkmaz/meridian/releases/tag/v2.1.1 - Sigstore attestation (actions/attest-build-provenance@v2 → gh attestation verify green on both .nupkg from the GitHub Release) - NuGet.org indexed both packages within the release workflow run

Workarounds

Users who cannot upgrade immediately may: 1. Avoid mapper.Map(src, dst) and .UseDestinationValue() on collection-typed destination members. 2. Wrap input collection deserialization with an explicit size limit before handing the payload to Meridian. 3. Register TaskWhenAllPublisher with maxDegreeOfParallelism ≤ 16 manually (v2.1.1+ only). 4. Disable OpenTelemetry exception.stacktrace tag emission at the trace exporter level if your trace sink is less trusted than your application.

These are defense-in-depth; the only complete mitigation is upgrading to 2.1.1.

Supported Versions

As of this advisory the supported security branch is 2.1.x. The 2.0.x line (published 2026-04-15) is not receiving the Phase 1 safety-defaults infrastructure needed to carry the HIGH-severity fixes, so 2.0.x is deprecated in favor of 2.1.x. See SECURITY.md for the updated supported-versions table.

Credits

  • UmutKorkmaz (reporter and maintainer)

References

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Meridian.Mapping"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.0.0"
            },
            {
              "fixed": "2.1.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "Meridian.Mediator"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.0.0"
            },
            {
              "fixed": "2.1.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [],
  "database_specific": {
    "cwe_ids": [
      "CWE-1325",
      "CWE-209",
      "CWE-400",
      "CWE-532",
      "CWE-665",
      "CWE-674",
      "CWE-770"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-04-16T22:50:37Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "## Summary\n\nMeridian v2.1.0 (`Meridian.Mapping` and `Meridian.Mediator`) shipped with nine defense-in-depth gaps reachable through its public APIs. Two are HIGH severity \u2014 the advertised `DefaultMaxCollectionItems` and `DefaultMaxDepth` safety caps are silently bypassed on the `IMapper.Map(source, destination)` overload and anywhere `.UseDestinationValue()` is configured on a collection-typed property. Four are MEDIUM (constructor invariant bypass, OpenTelemetry stack-trace info disclosure, retry amplification, notification fan-out amplification). Three are LOW (exception message disclosure, dictionary duplicate-key echo, static mediator cache growth under closed-generic types).\n\nAll nine are patched in **v2.1.1**. Upgrade is a drop-in NuGet bump; see the v2.1.1 CHANGELOG for the four behavioural changes (constructor selection, OTel default, publisher fan-out cap, retry caps).\n\n## Severity Matrix\n\n| # | Severity | CWE | Finding | Fix |\n|---|---|---|---|---|\n| 1 | **HIGH** | CWE-770 | `MappingEngine.TryMapCollectionOntoExisting` enumerated the source without enforcing `DefaultMaxCollectionItems`. Reachable via `Mapper.Map\u003cTSrc,TDst\u003e(src, dst)` and any `.ForMember(..., o =\u003e o.UseDestinationValue())` on a collection member through a plain `Map(src)` call. | Shared cap enforcement helper between `MapCollection` and `TryMapCollectionOntoExisting`. |\n| 2 | **HIGH** | CWE-674 | Collection-item recursion in the existing-destination path did not increment `ResolutionContext.Depth`, so self-referential collection graphs could reach stack overflow before `DefaultMaxDepth` fired. | Depth increments at every collection-item boundary. |\n| 3 | MEDIUM | CWE-665 | `ObjectCreator.CreateWithConstructorMapping` always invoked the widest public constructor, silently filling unresolved parameters with `default(T)` and bypassing narrower-ctor invariants. | Widest-ctor selection now requires every parameter to be bound via explicit ctor mapping, source-name match, or a C# optional default. |\n| 4 | MEDIUM | CWE-532 | `Mediator.MarkActivityFailure` emitted the full `ex.ToString()` (stack + inner chain) to the OpenTelemetry `exception.stacktrace` activity tag by default, leaking context to any shared trace sink. | Gated on `MediatorTelemetryOptions.RecordExceptionStackTrace` \u2014 opt-in, default `false`. |\n| 5 | MEDIUM | CWE-400 | `RetryBehavior` retried every exception type with unbounded `MaxRetries`; the exponential-backoff delay overflowed `TimeSpan` at ~30 attempts. No cancellation exclusion. | Server-side `MaxRetriesCap = 10`, `MaxBackoff = 5 min`, `OperationCanceledException` short-circuit, recommended `RetryPolicy.TransientOnly` helper. |\n| 6 | MEDIUM | CWE-400 | `TaskWhenAllPublisher` started every registered handler concurrently with no bound on fan-out. | New constructor parameter `maxDegreeOfParallelism` (default 16; `-1` restores legacy unbounded). |\n| 7 | LOW | CWE-209 | Public mapping exceptions leaked `FullName` of source/destination types and concatenated inner exception messages into top-level property-mapping errors. | Scrubbed to type `Name`; inner details only via `InnerException` chain. |\n| 8 | LOW | CWE-209 | Dictionary materialization threw `ArgumentException` on duplicate keys, echoing the attacker-supplied key\u0027s `.ToString()`. | Last-write-wins indexer semantics. |\n| 9 | LOW | CWE-1325 | Static mediator handler caches grow monotonically under closed-generic request types. **Doc-only mitigation**; no code change \u2014 consumers must not allow attacker-controlled runtime type materialization to reach `Send`, `Publish`, or `CreateStream`. | Documented in `docs/security-model.md`. |\n\n## Exploitation\n\n**Finding 1 / 2 (headline):** A consumer that maps user-supplied collection payloads onto an existing destination list via `mapper.Map(userCollection, existingList)` \u2014 a documented and commonly used AutoMapper-style idiom \u2014 processes the full attacker-supplied collection with no size cap and no depth cap. An attacker sending a single request with a large (or self-referential) collection payload can block the worker thread for seconds and exhaust the managed heap or the call stack. Equivalent exposure through `.UseDestinationValue()` on a collection-typed destination member, reachable via a plain `Map(src)` call whose destination type default-initializes that member.\n\n**Finding 3:** A destination type with multiple public constructors that differ only in their parameter-binding invariants (e.g., `new UserAccount(string name, Email email)` enforcing a non-default `Email`) could be instantiated with the narrower ctor\u0027s invariants silently bypassed if any source field was absent \u2014 the widest ctor was always picked, with unbound parameters replaced by `default(T)`.\n\n**Findings 4 / 5 / 6:** Amplification / information-disclosure vectors described in the matrix above. Each requires moderate integration context (telemetry sink trust, handler count, retry policy) to weaponize, but each is reachable through public APIs without authentication.\n\n## Patches\n\n- `Meridian.Mapping` **2.1.1** (published 2026-04-16)\n- `Meridian.Mediator` **2.1.1** (published 2026-04-16)\n\nVerified via:\n- GitHub Release assets at \u003chttps://github.com/UmutKorkmaz/meridian/releases/tag/v2.1.1\u003e\n- Sigstore attestation (`actions/attest-build-provenance@v2` \u2192 `gh attestation verify` green on both `.nupkg` from the GitHub Release)\n- NuGet.org indexed both packages within the release workflow run\n\n## Workarounds\n\nUsers who cannot upgrade immediately may:\n1. Avoid `mapper.Map(src, dst)` and `.UseDestinationValue()` on collection-typed destination members.\n2. Wrap input collection deserialization with an explicit size limit before handing the payload to Meridian.\n3. Register `TaskWhenAllPublisher` with `maxDegreeOfParallelism` \u2264 16 manually (v2.1.1+ only).\n4. Disable OpenTelemetry `exception.stacktrace` tag emission at the trace exporter level if your trace sink is less trusted than your application.\n\nThese are defense-in-depth; the only complete mitigation is upgrading to 2.1.1.\n\n## Supported Versions\n\nAs of this advisory the supported security branch is **2.1.x**. The 2.0.x line (published 2026-04-15) is not receiving the Phase 1 safety-defaults infrastructure needed to carry the HIGH-severity fixes, so 2.0.x is deprecated in favor of 2.1.x. See `SECURITY.md` for the updated supported-versions table.\n\n## Credits\n\n- UmutKorkmaz (reporter and maintainer)\n\n## References\n\n- v2.1.1 CHANGELOG section: \u003chttps://github.com/UmutKorkmaz/meridian/blob/main/CHANGELOG.md#211---2026-04-16\u003e\n- `docs/security-model.md` threat model: \u003chttps://github.com/UmutKorkmaz/meridian/blob/main/docs/security-model.md\u003e\n- `SECURITY.md` disclosure policy: \u003chttps://github.com/UmutKorkmaz/meridian/blob/main/SECURITY.md\u003e\n- AutoMapper CVE-2026-32933 (motivating precedent for Meridian\u0027s safety-defaults)",
  "id": "GHSA-f5v8-v6q3-q4h6",
  "modified": "2026-04-16T22:50:37Z",
  "published": "2026-04-16T22:50:37Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/UmutKorkmaz/meridian/security/advisories/GHSA-f5v8-v6q3-q4h6"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/UmutKorkmaz/meridian"
    },
    {
      "type": "WEB",
      "url": "https://github.com/UmutKorkmaz/meridian/blob/main/CHANGELOG.md#211---2026-04-16"
    },
    {
      "type": "WEB",
      "url": "https://github.com/UmutKorkmaz/meridian/releases/tag/v2.1.1"
    }
  ],
  "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"
    }
  ],
  "summary": "Meridian: Multiple defense-in-depth gaps (collection/depth caps, telemetry, retry, fan-out)"
}

GHSA-FCGC-C6HR-PQM2

Vulnerability from github – Published: 2023-01-05 18:30 – Updated: 2023-01-11 21:30
VLAI
Details

GPAC MP4Box 2.1-DEV-rev649-ga8f438d20 has a segment fault (/stack overflow) due to infinite recursion in Media_GetSample isomedia/media.c:662

Severity
5.5 (Medium)
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2022-47662"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-674"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2023-01-05T16:15:00Z",
    "severity": "MODERATE"
  },
  "details": "GPAC MP4Box 2.1-DEV-rev649-ga8f438d20 has a segment fault (/stack overflow) due to infinite recursion in Media_GetSample isomedia/media.c:662",
  "id": "GHSA-fcgc-c6hr-pqm2",
  "modified": "2023-01-11T21:30:40Z",
  "published": "2023-01-05T18:30:30Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2022-47662"
    },
    {
      "type": "WEB",
      "url": "https://github.com/gpac/gpac/issues/2359"
    },
    {
      "type": "WEB",
      "url": "https://www.debian.org/security/2023/dsa-5411"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-FF5C-CP5C-9WJF

Vulnerability from github – Published: 2026-09-02 14:33 – Updated: 2026-09-02 14:33
VLAI
Summary
NLTK: Uncontrolled resource consumption in RecursiveDescentParser via ambiguous or left-recursive grammars
Details

nltk.parse.RecursiveDescentParser (and SteppingRecursiveDescentParser) enumerate parses top-down with no bound on the number of recursive steps. A small, crafted context-free grammar makes a short input consume unbounded CPU (and/or exhaust the Python recursion stack), pinning a process indefinitely — a denial of service.

Proof of concept

Both of the following hang on a 24-token input (killed after 8s; growth is super-linear in input length), on NLTK develop:

from nltk import CFG
from nltk.parse import RecursiveDescentParser

# (a) left recursion -> unbounded recursion
g = CFG.fromstring("S -> S S | 'a'")
list(RecursiveDescentParser(g).parse(["a"] * 24))   # hangs

# (b) ambiguous grammar -> exponential number of parses
g = CFG.fromstring("S -> 'a' S | 'a' S S | 'a'")
list(RecursiveDescentParser(g).parse(["a"] * 24))   # hangs

Impact

An application that runs RecursiveDescentParser on a grammar (or an input) drawn from an untrusted source can be driven into an unbounded CPU / stack-exhaustion loop by a tiny payload. No confidentiality or integrity impact; single-process availability only.

Sibling

The RegexpTokenizer ReDoS reported alongside this (CVE-2026-12875) is a different class (caller-supplied regex) and is addressed under GHSA-w3v8-gmh9-3wv7.

Severity
6.9 (Medium)
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 3.10.2"
      },
      "package": {
        "ecosystem": "PyPI",
        "name": "nltk"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "3.10.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-12876"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-407",
      "CWE-674"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-09-02T14:33:38Z",
    "nvd_published_at": null,
    "severity": "MODERATE"
  },
  "details": "`nltk.parse.RecursiveDescentParser` (and `SteppingRecursiveDescentParser`) enumerate parses top-down with no bound on the number of recursive steps. A small, crafted context-free grammar makes a short input consume unbounded CPU (and/or exhaust the Python recursion stack), pinning a process indefinitely \u2014 a denial of service.\n\n## Proof of concept\n\nBoth of the following hang on a 24-token input (killed after 8s; growth is super-linear in input length), on NLTK develop:\n\n```python\nfrom nltk import CFG\nfrom nltk.parse import RecursiveDescentParser\n\n# (a) left recursion -\u003e unbounded recursion\ng = CFG.fromstring(\"S -\u003e S S | \u0027a\u0027\")\nlist(RecursiveDescentParser(g).parse([\"a\"] * 24))   # hangs\n\n# (b) ambiguous grammar -\u003e exponential number of parses\ng = CFG.fromstring(\"S -\u003e \u0027a\u0027 S | \u0027a\u0027 S S | \u0027a\u0027\")\nlist(RecursiveDescentParser(g).parse([\"a\"] * 24))   # hangs\n```\n\n## Impact\n\nAn application that runs `RecursiveDescentParser` on a grammar (or an input) drawn from an untrusted source can be driven into an unbounded CPU / stack-exhaustion loop by a tiny payload. No confidentiality or integrity impact; single-process availability only.\n\n## Sibling\n\nThe RegexpTokenizer ReDoS reported alongside this (CVE-2026-12875) is a different class (caller-supplied regex) and is addressed under GHSA-w3v8-gmh9-3wv7.",
  "id": "GHSA-ff5c-cp5c-9wjf",
  "modified": "2026-09-02T14:33:39Z",
  "published": "2026-09-02T14:33:38Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/nltk/nltk/security/advisories/GHSA-ff5c-cp5c-9wjf"
    },
    {
      "type": "WEB",
      "url": "https://github.com/nltk/nltk/pull/3649"
    },
    {
      "type": "WEB",
      "url": "https://github.com/nltk/nltk/commit/43aaca1b9024138421c97f970bf13ee19ac8129d"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/nltk/nltk"
    },
    {
      "type": "WEB",
      "url": "https://github.com/nltk/nltk/releases/tag/v3.10.3"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "NLTK: Uncontrolled resource consumption in RecursiveDescentParser via ambiguous or left-recursive grammars"
}

GHSA-FF5X-7QG5-VWF2

Vulnerability from github – Published: 2023-12-13 13:32 – Updated: 2023-12-13 13:32
VLAI
Summary
Denial of service caused by infinite recursion when parsing SVG document
Details

Summary

When parsing the attributes passed to a use tag inside an svg document, we can cause the system to go to an infinite recursion. Depending on the system configuration and attack pattern this could exhaust the memory available to the executing process and/or to the server itself.

Details

Inside Svg\Tag\UseTag::before, php-svg-lib parses the attributes passed to an use tag inside an svg document. When it finds a href or xlink:href, it will try to retrieve the object representing this tag:

$link = $attributes["href"] ?? $attributes["xlink:href"];
$this->reference = $document->getDef($link);

if ($this->reference) {
    $this->reference->before($attributes);
}

$document->getDef is implemented as follow:

public function getDef($id) {
    $id = ltrim($id, "#");

    return isset($this->defs[$id]) ? $this->defs[$id] : null;
}

Note: the $id in the above method is actually the link being used in use tag. This part is important, because this behaviour here actually leads to the vulnerability. It will be mentioned later on in this report.

If it finds the referenced object, it will try to call the before method on the referenced object (this is still inside Svg\Tag\UseTag::before) :

if ($this->reference) {
    $this->reference->before($attributes);
}

In order to cause an infinte loop, we need to be able to control the $id used in the $this->defs[$id] code above. This defs property (Svg\Document::defs) is being populated when Svg\Document::_tagStart is called. This is the handler being used when the php-svg-lib is parsing the svg structure:

// Svg\Document line 343
if ($tag) {
    if (isset($attributes["id"])) {
        $this->defs[$attributes["id"]] = $tag;
    }
    else {
        // ...
    }

    // ...
}

So if the use tag contains an id, then that use tag will be added to the $defs array with it's id as the key.

Now as noted before, when there is a link inside the use tag, the library uses that link as the id to actually find the object or tag that has been added to the Svg\Document::defs.

So if the id attribute is equal to the link attribute inside the use tag, then the referenced object (in this case it is the Use tag object) will be called recursively until the memory given to the script is exhausted.

PoC

This is an example svg file that can be used to demonstrate the vulnerability.

<svg width="200" height="200"
  xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
  <use id="selfref" xlink:href="#selfref" />
</svg>

Impact

When the lib parses the above payload, it will crash:

PHP Fatal error:  Allowed memory size of 536870912 bytes exhausted (tried to allocate 262144 bytes) in /xxx/dompdf/vendor/phenx/php-svg-lib/src/Svg/Tag/UseTag.php on line 37

An attacker sending multiple request to a system to render the above payload can potentially cause resource exhaustion to the point that the system is unable to handle incoming request.

Severity
5.3 (Medium)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "phenx/php-svg-lib"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "0.5.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2023-50251"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-674"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2023-12-13T13:32:21Z",
    "nvd_published_at": "2023-12-12T21:15:08Z",
    "severity": "MODERATE"
  },
  "details": "### Summary\nWhen parsing the attributes passed to a `use` tag inside an svg document, we can cause the system to go to an infinite recursion. Depending on the system configuration and attack pattern this could exhaust the memory available to the executing process and/or to the server itself.\n\n### Details\nInside `Svg\\Tag\\UseTag::before`, php-svg-lib parses the attributes passed to an `use` tag inside an svg document. When it finds a `href` or `xlink:href`, it will try to retrieve the object representing this tag:\n\n```\n$link = $attributes[\"href\"] ?? $attributes[\"xlink:href\"];\n$this-\u003ereference = $document-\u003egetDef($link);\n\nif ($this-\u003ereference) {\n    $this-\u003ereference-\u003ebefore($attributes);\n}\n```\n\n`$document-\u003egetDef` is implemented as follow:\n\n```\npublic function getDef($id) {\n    $id = ltrim($id, \"#\");\n\n    return isset($this-\u003edefs[$id]) ? $this-\u003edefs[$id] : null;\n}\n```\n\n_Note:_ the `$id` in the above method is actually the _link_ being used in `use` tag. This part is important, because this behaviour  here actually leads to the vulnerability. It will be mentioned later on in this report.\n\nIf it finds the referenced object, it will try to call the `before` method on the referenced object (this is still inside `Svg\\Tag\\UseTag::before`) :\n\n```\nif ($this-\u003ereference) {\n    $this-\u003ereference-\u003ebefore($attributes);\n}\n```\n\nIn order to cause an infinte loop, we need to be able to control the `$id` used in the `$this-\u003edefs[$id]` code above. This `defs` property (`Svg\\Document::defs`) is being populated when `Svg\\Document::_tagStart` is called. This is the handler being used when the php-svg-lib is parsing the svg structure:\n\n```\n// Svg\\Document line 343\nif ($tag) {\n    if (isset($attributes[\"id\"])) {\n        $this-\u003edefs[$attributes[\"id\"]] = $tag;\n    }\n    else {\n        // ...\n    }\n\n    // ...\n}\n```\n\nSo if the `use` tag contains an `id`, then that `use` tag will be added to the `$defs` array with it\u0027s `id` as the key.\n\nNow as noted before, when there is a link inside the `use` tag, the library uses that link as the `id` to actually find the object or `tag` that has been added to the `Svg\\Document::defs`.\n\nSo if the `id` attribute is equal to the link attribute inside the `use` tag, then the referenced object (in this case it is the `Use` tag object) will be called recursively until the memory given to the script is exhausted.\n\n### PoC\n\nThis is an example svg file that can be used to demonstrate the vulnerability.\n\n```\n\u003csvg width=\"200\" height=\"200\"\n  xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"\u003e\n  \u003cuse id=\"selfref\" xlink:href=\"#selfref\" /\u003e\n\u003c/svg\u003e\n```\n\n### Impact\n\nWhen the lib parses the above payload, it will crash:\n\n```\nPHP Fatal error:  Allowed memory size of 536870912 bytes exhausted (tried to allocate 262144 bytes) in /xxx/dompdf/vendor/phenx/php-svg-lib/src/Svg/Tag/UseTag.php on line 37\n```\n\nAn attacker sending multiple request to a system to render the above payload can potentially cause resource exhaustion to the point that the system is unable to handle incoming request.",
  "id": "GHSA-ff5x-7qg5-vwf2",
  "modified": "2023-12-13T13:32:21Z",
  "published": "2023-12-13T13:32:21Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/dompdf/php-svg-lib/security/advisories/GHSA-ff5x-7qg5-vwf2"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2023-50251"
    },
    {
      "type": "WEB",
      "url": "https://github.com/dompdf/php-svg-lib/commit/88163cbe562d9b391b3a352e54d9c89d02d77ee0"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/dompdf/php-svg-lib"
    }
  ],
  "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:L",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Denial of service caused by infinite recursion when parsing SVG document"
}

GHSA-FJ27-X67R-V4G5

Vulnerability from github – Published: 2026-07-16 00:31 – Updated: 2026-07-20 18:32
VLAI
Details

A stack overflow in the evaluate() function (editors/awk.c) of BusyBox commit 371fe9 allows attackers to cause a Denial of Service (DoS) via supplying a crafted AWK script.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2026-38752"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-121",
      "CWE-674"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-07-15T22:16:47Z",
    "severity": "HIGH"
  },
  "details": "A stack overflow in the evaluate() function (editors/awk.c) of BusyBox commit 371fe9 allows attackers to cause a Denial of Service (DoS) via supplying a crafted AWK script.",
  "id": "GHSA-fj27-x67r-v4g5",
  "modified": "2026-07-20T18:32:26Z",
  "published": "2026-07-16T00:31:33Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-38752"
    },
    {
      "type": "WEB",
      "url": "https://busybox.net"
    },
    {
      "type": "WEB",
      "url": "https://lists.busybox.net/pipermail/busybox/2026-June/092351.html"
    },
    {
      "type": "WEB",
      "url": "http://busybox.com"
    }
  ],
  "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"
    }
  ]
}
Mitigation
Implementation

Ensure that an end condition will be reached under all logic conditions. The end condition may include checking against the depth of recursion and exiting with an error if the recursion goes too deep. The complexity of the end condition contributes to the effectiveness of this action.

Mitigation
Implementation

Increase the stack size.

CAPEC-230: Serialized Data with Nested Payloads

Applications often need to transform data in and out of a data format (e.g., XML and YAML) by using a parser. It may be possible for an adversary to inject data that may have an adverse effect on the parser when it is being processed. Many data format languages allow the definition of macro-like structures that can be used to simplify the creation of complex structures. By nesting these structures, causing the data to be repeatedly substituted, an adversary can cause the parser to consume more resources while processing, causing excessive memory consumption and CPU utilization.

CAPEC-231: Oversized Serialized Data Payloads

An adversary injects oversized serialized data payloads into a parser during data processing to produce adverse effects upon the parser such as exhausting system resources and arbitrary code execution.

← Back to the CWE list