{"dataType": "CVE_RECORD", "dataVersion": "5.0", "cveMetadata": {"state": "PUBLISHED", "cveId": "CVE-2022-41894", "assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa", "assignerShortName": "GitHub_M", "dateUpdated": "2022-11-18T00:00:00", "dateReserved": "2022-09-30T00:00:00", "datePublished": "2022-11-18T00:00:00"}, "containers": {"cna": {"title": "Buffer overflow in `CONV_3D_TRANSPOSE` on TFLite", "providerMetadata": {"orgId": "a0819718-46f1-4df5-94e2-005712e83aaa", "shortName": "GitHub_M", "dateUpdated": "2022-11-18T00:00:00"}, "descriptions": [{"lang": "en", "value": "TensorFlow is an open source platform for machine learning. The reference kernel of the `CONV_3D_TRANSPOSE` TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result. Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels > output_num_channels. An attacker can craft a model with a specific number of input channels. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter. We have patched the issue in GitHub commit 72c0bdcb25305b0b36842d746cc61d72658d2941. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range."}], "affected": [{"vendor": "tensorflow", "product": "tensorflow", "versions": [{"version": ">= 2.10.0, < 2.10.1", "status": "affected"}, {"version": ">= 2.9.0, < 2.9.3", "status": "affected"}, {"version": "< 2.8.4", "status": "affected"}]}], "references": [{"url": "https://github.com/tensorflow/tensorflow/security/advisories/GHSA-h6q3-vv32-2cq5"}, {"url": "https://github.com/tensorflow/tensorflow/commit/72c0bdcb25305b0b36842d746cc61d72658d2941"}, {"url": "https://github.com/tensorflow/tensorflow/blob/091e63f0ea33def7ecad661a5ac01dcafbafa90b/tensorflow/lite/kernels/internal/reference/conv3d_transpose.h#L121"}], "metrics": [{"cvssV3_1": {"version": "3.1", "vectorString": "CVSS:3.1/AV:N/AC:H/PR:L/UI:R/S:U/C:H/I:H/A:H", "attackVector": "NETWORK", "attackComplexity": "HIGH", "privilegesRequired": "LOW", "userInteraction": "REQUIRED", "scope": "UNCHANGED", "confidentialityImpact": "HIGH", "integrityImpact": "HIGH", "availabilityImpact": "HIGH", "baseScore": 7.1, "baseSeverity": "HIGH"}}], "problemTypes": [{"descriptions": [{"type": "CWE", "lang": "en", "description": "CWE-120: Buffer Copy without Checking Size of Input ('Classic Buffer Overflow')", "cweId": "CWE-120"}]}], "source": {"advisory": "GHSA-h6q3-vv32-2cq5", "discovery": "UNKNOWN"}}}}

{"configurations": [{"nodes": [{"cpeMatch": [{"criteria": "cpe:2.3:a:google:tensorflow:*:*:*:*:*:*:*:*", "matchCriteriaId": "A694EEE1-BFB9-4E6C-B275-02DC2731961C", "versionEndExcluding": "2.8.4", "vulnerable": true}, {"criteria": "cpe:2.3:a:google:tensorflow:*:*:*:*:*:*:*:*", "matchCriteriaId": "9057B403-719C-4F10-BAB6-67F84786A89E", "versionEndExcluding": "2.9.3", "versionStartIncluding": "2.9.0", "vulnerable": true}, {"criteria": "cpe:2.3:a:google:tensorflow:*:*:*:*:*:*:*:*", "matchCriteriaId": "793BC396-7686-47FA-A107-DA6FC90704A2", "versionEndExcluding": "2.10.1", "versionStartIncluding": "2.10.0", "vulnerable": true}], "negate": false, "operator": "OR"}]}], "descriptions": [{"lang": "en", "value": "TensorFlow is an open source platform for machine learning. The reference kernel of the `CONV_3D_TRANSPOSE` TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result. Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels > output_num_channels. An attacker can craft a model with a specific number of input channels. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter. We have patched the issue in GitHub commit 72c0bdcb25305b0b36842d746cc61d72658d2941. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range."}, {"lang": "es", "value": "TensorFlow es una plataforma de c\u00f3digo abierto para aprendizaje autom\u00e1tico. El n\u00facleo de referencia del operador `CONV_3D_TRANSPOSE` de TensorFlow Lite incrementa err\u00f3neamente data_ptr al agregar el sesgo al resultado. En lugar de `data_ptr += num_channels;` deber\u00eda ser `data_ptr += output_num_channels;` ya que si el n\u00famero de canales de entrada es diferente al n\u00famero de canales de salida, se devolver\u00e1 un resultado incorrecto y se producir\u00e1 un desbordamiento del b\u00fafer si num_channels &gt; output_num_channels. Un atacante puede crear un modelo con un n\u00famero espec\u00edfico de canales de entrada. Entonces es posible escribir valores espec\u00edficos a trav\u00e9s del sesgo de la capa fuera de los l\u00edmites del b\u00fafer. Este ataque solo funciona si se utiliza el solucionador del n\u00facleo de referencia en el int\u00e9rprete. Hemos solucionado el problema en el compromiso de GitHub 72c0bdcb25305b0b36842d746cc61d72658d2941. La soluci\u00f3n se incluir\u00e1 en TensorFlow 2.11. Tambi\u00e9n seleccionaremos este commit en TensorFlow 2.10.1, 2.9.3 y TensorFlow 2.8.4, ya que estos tambi\u00e9n se ven afectados y a\u00fan se encuentran en el rango admitido."}], "id": "CVE-2022-41894", "lastModified": "2022-11-22T21:02:17.037", "metrics": {"cvssMetricV31": [{"cvssData": {"attackComplexity": "HIGH", "attackVector": "NETWORK", "availabilityImpact": "HIGH", "baseScore": 8.1, "baseSeverity": "HIGH", "confidentialityImpact": "HIGH", "integrityImpact": "HIGH", "privilegesRequired": "NONE", "scope": "UNCHANGED", "userInteraction": "NONE", "vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H", "version": "3.1"}, "exploitabilityScore": 2.2, "impactScore": 5.9, "source": "nvd@nist.gov", "type": "Primary"}, {"cvssData": {"attackComplexity": "HIGH", "attackVector": "NETWORK", "availabilityImpact": "HIGH", "baseScore": 7.1, "baseSeverity": "HIGH", "confidentialityImpact": "HIGH", "integrityImpact": "HIGH", "privilegesRequired": "LOW", "scope": "UNCHANGED", "userInteraction": "REQUIRED", "vectorString": "CVSS:3.1/AV:N/AC:H/PR:L/UI:R/S:U/C:H/I:H/A:H", "version": "3.1"}, "exploitabilityScore": 1.2, "impactScore": 5.9, "source": "security-advisories@github.com", "type": "Secondary"}]}, "published": "2022-11-18T22:15:17.523", "references": [{"source": "security-advisories@github.com", "tags": ["Third Party Advisory"], "url": "https://github.com/tensorflow/tensorflow/blob/091e63f0ea33def7ecad661a5ac01dcafbafa90b/tensorflow/lite/kernels/internal/reference/conv3d_transpose.h#L121"}, {"source": "security-advisories@github.com", "tags": ["Patch", "Third Party Advisory"], "url": "https://github.com/tensorflow/tensorflow/commit/72c0bdcb25305b0b36842d746cc61d72658d2941"}, {"source": "security-advisories@github.com", "tags": ["Exploit", "Patch", "Third Party Advisory"], "url": "https://github.com/tensorflow/tensorflow/security/advisories/GHSA-h6q3-vv32-2cq5"}], "sourceIdentifier": "security-advisories@github.com", "vulnStatus": "Analyzed", "weaknesses": [{"description": [{"lang": "en", "value": "CWE-120"}], "source": "security-advisories@github.com", "type": "Primary"}]}

{}