Skip to content
Heist/ feed
Back to feed
Medium

Security advisory

Type: advisorySources: OSVCVE: CVE-2021-37646

TensorFlow is an end-to-end open source platform for machine learning. In affected versions the implementation of `tf.raw_ops.StringNGrams` is vulnerable to an integer overflow issue caused by converting a signed integer value to an unsigned one and then allocating memory based on this value. The [implementation](https://github.com/tensorflow/tensorflow/blob/8d72537c6abf5a44103b57b9c2e22c14f5f49698/tensorflow/core/kernels/string_ngrams_op.cc#L184) calls `reserve` on a `tstring` with a value that sometimes can be negative if user supplies negative `ngram_widths`. The `reserve` method calls `TF_TString_Reserve` which has an `unsigned long` argument for the size of the buffer. Hence, the implicit conversion transforms the negative value to a large integer. We have patched the issue in GitHub commit c283e542a3f422420cfdb332414543b62fc4e4a5. The fix will be included in TensorFlow 2.6.0. We will also cherrypick this commit on TensorFlow 2.5.1, TensorFlow 2.4.3, and TensorFlow 2.3.4, as these are also affected and still in supported range.

Products
Not specified
Affected versions
Not specified
Fixed versions
Not specified
CVEs
CVE-2021-37646
Confirmed fixed CVEs
Not specified
Updated in this feed
31 Oct 2026, 12:39 UTC · Revision 1

Source evidence

CVE-2021-37646

core-osv · Original source

Published 12 Aug 2021, 21:15 UTC

TensorFlow is an end-to-end open source platform for machine learning. In affected versions the implementation of `tf.raw_ops.StringNGrams` is vulnerable to an integer overflow issue caused by converting a signed integer value to an unsigned one and then allocating memory based on this value. The [implementation](https://github.com/tensorflow/tensorflow/blob/8d72537c6abf5a44103b57b9c2e22c14f5f49698/tensorflow/core/kernels/string_ngrams_op.cc#L184) calls `reserve` on a `tstring` with a value that sometimes can be negative if user supplies negative `ngram_widths`. The `reserve` method calls `TF_TString_Reserve` which has an `unsigned long` argument for the size of the buffer. Hence, the implicit conversion transforms the negative value to a large integer. We have patched the issue in GitHub commit c283e542a3f422420cfdb332414543b62fc4e4a5. The fix will be included in TensorFlow 2.6.0. We will also cherrypick this commit on TensorFlow 2.5.1, TensorFlow 2.4.3, and TensorFlow 2.3.4, as these are also affected and still in supported range.

osv:CVE-2021-37646
Complete source record
{
  "id": "c44a553086b7f44585d6c285a6d30a1fa276e63ffcdb97bafc196151c9bb88d8",
  "url": "https://osv.dev/vulnerability/CVE-2021-37646",
  "method": "structured",
  "payload": {
    "kind": "advisory",
    "title": "CVE-2021-37646",
    "summary": "TensorFlow is an end-to-end open source platform for machine learning. In affected versions the implementation of `tf.raw_ops.StringNGrams` is vulnerable to an integer overflow issue caused by converting a signed integer value to an unsigned one and then allocating memory based on this value. The [implementation](https://github.com/tensorflow/tensorflow/blob/8d72537c6abf5a44103b57b9c2e22c14f5f49698/tensorflow/core/kernels/string_ngrams_op.cc#L184) calls `reserve` on a `tstring` with a value that sometimes can be negative if user supplies negative `ngram_widths`. The `reserve` method calls `TF_TString_Reserve` which has an `unsigned long` argument for the size of the buffer. Hence, the implicit conversion transforms the negative value to a large integer. We have patched the issue in GitHub commit c283e542a3f422420cfdb332414543b62fc4e4a5. The fix will be included in TensorFlow 2.6.0. We will also cherrypick this commit on TensorFlow 2.5.1, TensorFlow 2.4.3, and TensorFlow 2.3.4, as these are also affected and still in supported range.",
    "evidence": "osv:CVE-2021-37646",
    "products": [],
    "severity": "medium",
    "structured": {
      "id": "CVE-2021-37646",
      "aliases": [
        "BIT-tensorflow-2021-37646",
        "CVE-2021-37646",
        "GHSA-h6jh-7gv5-28vg",
        "PYSEC-2021-268",
        "PYSEC-2021-559",
        "PYSEC-2021-757"
      ],
      "metrics": [
        {
          "type": "CVSS_V3",
          "score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H"
        }
      ],
      "affected": [
        {
          "ranges": [
            {
              "repo": "https://github.com/tensorflow/tensorflow",
              "type": "GIT",
              "events": [
                {
                  "introduced": "b36436b087bd8e8701ef51718179037cccdfc26e"
                },
                {
                  "fixed": "7462dcaae1e8cfe1dfd0c62dd6083f9749a9d827"
                },
                {
                  "introduced": "582c8d236cb079023657287c318ff26adb239002"
                },
                {
                  "fixed": "4c0b84bf2a714bcdd18da1f1f94d533d72399d52"
                },
                {
                  "introduced": "a4dfb8d1a71385bd6d122e4f27f86dcebb96712d"
                },
                {
                  "last_affected": "5368d50428b30b7c9ccd038aec65d09252d16596"
                },
                {
                  "fixed": "c283e542a3f422420cfdb332414543b62fc4e4a5"
                }
              ],
              "database_specific": {
                "extracted_events": [
                  {
                    "cpe": "cpe:2.3:a:google:tensorflow:*:*:*:*:*:*:*:*",
                    "range": [
                      {
                        "introduced": "2.3.0"
                      },
                      {
                        "fixed": "2.3.4"
                      }
                    ],
                    "source": "CPE_RANGE"
                  },
                  {
                    "cpe": "cpe:2.3:a:google:tensorflow:*:*:*:*:*:*:*:*",
                    "range": [
                      {
                        "introduced": "2.4.0"
                      },
                      {
                        "fixed": "2.4.3"
                      }
                    ],
                    "source": "CPE_RANGE"
                  },
                  {
                    "cpe": "cpe:2.3:a:google:tensorflow:2.5.0:*:*:*:*:*:*:*",
                    "range": [
                      {
                        "introduced": "2.5.0"
                      },
                      {
                        "last_affected": "2.5.0"
                      }
                    ],
                    "source": "CPE_STRING"
                  },
                  {
                    "cpe": "cpe:2.3:a:google:tensorflow:2.6.0:rc0:*:*:*:*:*:*",
                    "range": [
                      {
                        "introduced": "2.6.0-rc0"
                      },
                      {
                        "last_affected": "2.6.0-rc0"
                      }
                    ],
                    "source": "CPE_STRING"
                  },
                  {
                    "cpe": "cpe:2.3:a:google:tensorflow:2.6.0:rc1:*:*:*:*:*:*",
                    "range": [
                      {
                        "introduced": "2.6.0-rc1"
                      },
                      {
                        "last_affected": "2.6.0-rc1"
                      }
                    ],
                    "source": "CPE_STRING"
                  },
                  {
                    "cpe": "cpe:2.3:a:google:tensorflow:2.6.0:rc2:*:*:*:*:*:*",
                    "range": [
                      {
                        "introduced": "2.6.0-rc2"
                      },
                      {
                        "last_affected": "2.6.0-rc2"
                      }
                    ],
                    "source": "CPE_STRING"
                  },
                  {
                    "range": [
                      {
                        "fixed": "c283e542a3f422420cfdb332414543b62fc4e4a5"
                      }
                    ],
                    "source": "REFERENCES_COMMIT"
                  }
                ]
              }
            }
          ],
          "versions": [
            "2.5.0",
            "2.6.0-rc0",
            "2.6.0-rc1",
            "2.6.0-rc2",
            "v2.3.3",
            "v2.4.2",
            "v2.3.2",
            "v2.4.1",
            "v2.4.0",
            "v2.3.1",
            "v2.3.0"
          ],
          "database_specific": {
            "source": "https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2021-37646.json",
            "vanir_signatures": [
              {
                "id": "CVE-2021-37646-f8967422",
                "digest": {
                  "threshold": 0.9,
                  "line_hashes": [
                    "285621891283101487319147839348099093086",
                    "298920372003464589597498879960503880894",
                    "276814498817984533453586335083049431762",
                    "182428655347205365162753629093306997039"
                  ]
                },
                "source": "https://github.com/tensorflow/tensorflow/commit/c283e542a3f422420cfdb332414543b62fc4e4a5",
                "target": {
                  "file": "tensorflow/core/kernels/string_ngrams_op.cc"
                },
                "deprecated": false,
                "signature_type": "Line",
                "signature_version": "v1"
              }
            ],
            "vanir_signatures_modified": "2026-10-20T14:10:54Z"
          }
        }
      ],
      "modified": "2026-10-20T14:10:54.681127539Z",
      "provider": "osv",
      "references": [
        "https://github.com/tensorflow/tensorflow/security/advisories/GHSA-h6jh-7gv5-28vg",
        "https://github.com/tensorflow/tensorflow/commit/c283e542a3f422420cfdb332414543b62fc4e4a5"
      ],
      "provider_metadata": {}
    },
    "related_cves": [
      {
        "value": "CVE-2021-37646",
        "evidence": "/id"
      }
    ],
    "security_relevant": true,
    "severity_evidence": "/severity/0/score"
  },
  "source_id": "core-osv",
  "document_id": 388674,
  "published_at": "2021-08-12T21:15:07.983+00:00"
}